2024 Clifford Symposium Digital Binder by DePaulLaw

THE 30TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

F O Y C A THE LEG WIDE Y R T S U T IND C U D N O C S I M Y DEADL June 6-7, 2024

THE 30TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

F O Y C A G THE LE E D I W Y R INDUST ISCONDUCT M Y L D A E D At least three times in the past hundred years, entire American industries have united in a coordinated and sustained effort to advance their financial interests by knowingly promoting products that have deadly consequences for enormous numbers of individuals exposed to them. The three industries involved in these activities have been the asbestos manufacturers, the tobacco sellers and the opioid purveyors. There have been striking parallels both in their conduct and the inadequate societal response to the harm done. Accordingly, this year’s Clifford Symposium will explore: • The historical legacy of these events • The ethical questions raised by the legal representation offered the industries • The behavior of those who enabled widespread harm • The American industries that appear to be following the same deadly roadmap • The procedures we might adopt to address these activities

THE 30 TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

SYMPOSIUM FACULTY

In 1994, Robert A. Clifford (’76) endowed a faculty chair in tort law and social policy. The chair gives meaningful expression to his belief that the civil justice system serves a number of vital interests in American society. The Clifford Chair at DePaul provides a vehicle for exploration of the civil justice system in an intellectually rigorous fashion. In addition to providing support for faculty research and teaching, the endowment makes possible an annual symposium addressing a timely issue in the civil justice area. The purpose of the symposium is to bring the latest scholarship and advances in legal practice to lawyers and scholars who specialize in tort law, civil justice and related fields. Professor Stephan Landsman is the current organizer and director of the symposium underwritten by the Clifford Chair. Details from George Grosz, The Pillars of Society, 1926. Staatliche Museen zu Berlin (Berlin, Germany)

Elizabeth Chamblee Burch University of Georgia

Elise Maizel Michigan State University

Seth Endo Seattle University

Gregory Mark DePaul University

David Freeman Engstrom Stanford University

Thomas McGarity University of Texas

Myriam Gilles Yeshiva University Steve Gold Rutgers University John Goldberg Harvard University Samuel Issacharoff New York University Douglas Kysar Yale University Alexandra Lahav Cornell University Stephan Landsman DePaul University Colleen LanierChristensen Harvard University Hillel Levin University of Georgia Adam Littlestone-Luria Member, New York Bar Timothy Lytton Georgia State University

Robert Rabin Stanford University David Rosner Columbia University Noah Rosenblum New York University Rachel Rothschild University of Michigan Michael Saks Arizona State University Catherine Sharkey New York University Sarah Swan Rutgers University Wendy Wagner University of Texas Bradley Wendel Cornell University Maggie Wittlin Fordham University Benjamin Zipursky Fordham University

THE 30 TH ANNUAL

CLIFFORD SYMPOSIUM

AGENDA

ON TORT LAW AND SOCIAL POLICY

THURSDAY, JUNE 6, 2024

FRIDAY, JUNE 7, 2024

9:00

REGISTRATION AND CONTINENTAL BREAKFAST

8:00

9:30

OPENING REMARKS

9:00 S ESSION IV: ETHICAL DILEMMAS REGARDING LEGAL REPRESENTATION

Stephan Landsman Emeritus Professor, DePaul College of Law; Director, Clifford Symposium

9:45

Deal making and the Ethically Slippery Slope in the Quest for Global Peace Elizabeth Chamblee Burch, University of Georgia

SESSION I: HISTORICAL PERSPECTIVES

Corporate Lawyers, Disloyalty and the Opioid Crisis Elise Maizel, Michigan State University

Remedy Becomes Regulation Samuel Issacharoff, New York University Adam Littlestone-Luria, Member, New York Bar

What About Purdue’s Lawyers? A Brief History of Lawyer Gatekeeping From Social Trustee Professionalism Through the 2023 ABA Model Rules Amendment Bradley Wendel, Cornell University

Toxic Standards: Regulatory Epistemology and the Legacy of Endocrine Disruption Chemicals Colleen Lanier-Christensen, Harvard University Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University Physicists as Environmental Experts: The Case of Richard Wilson Rachel Rothschild, University of Michigan

Discussant: Myriam Gilles, Yeshiva University

11:00 BREAK 11:15 SESSION V: RESPONSES On the Relation Between Causation and Misconduct Alexandra Lahav, Cornell University Using Public Nuisance Litigation to Address Industry Misconduct: Common Law Statutes, Delegation Doctrine and Systems Theory Timothy Lytton, Georgia State University Hillel Levin, University of Georgia

Discussant: Noah Rosenblum, New York University

12:00 LUNCH (provided) 1:00

SESSION II: OTHER INDUSTRIES?

Microlocal Litigation for a Mass Tort World Sarah Swan, Rutgers University

Farm Until It’s Gone: Industrial Animal Agriculture and the Limits of Law Douglas Kysar, Yale University Public Health Nuisances Catherine Sharkey, New York University Tragic Exposure, Mass Litigation and Regulatory Failure: Does It Have to Be This Way? Wendy Wagner, University of Texas Steve Gold, Rutgers University Thomas McGarity, University of Texas Evidence of Compliance Maggie Wittlin, Fordham University Discussant: Gregory Mark, DePaul University

3:10

BREAK

3:20

SESSION III: ENABLERS iscovery Gamesmanship in Mass Torts D Seth Endo, Seattle University New Empirics on Litigation Secrecy (Virtual Presentation) David Freeman Engstrom, Stanford University* Toxic Torts and the Shadowland of Responsibility John Goldberg, Harvard University Benjamin Zipursky, Fordham University Effects of Sponsorship on Research Results Michael Saks, Arizona State University Discussant: Stephan Landsman * C ollaborators concerning the reported upon research include Nora Freeman Engstrom, Stanford University; Jonah Gelbach, University of California, Berkeley; Austin Peters, Stanford University

5:30

END OF DAY ONE

REGISTRATION AND CONTINENTAL BREAKFAST

Discussant: Robert Rabin, Stanford University

1:15

CONCLUDING REMARKS

SESSION I: HISTORICAL PERSPECTIVES Remedy Becomes Regulation Samuel Issacharoff, New York University Adam Littlestone-Luria, Member, New York Bar Toxic Standards: Regulatory Epistemology and the Legacy of Endocrine Disruption Chemicals Colleen Lanier-Christensen, Harvard University Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University Physicists as Environmental Experts: The Case of Richard Wilson Rachel Rothschild, University of Michigan Discussant: Noah Rosenblum, New York University

(Back to main agenda)

THE 30 TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

REMEDY BECOMES REGULATION: STATE MAKING AFTER THE FACT

Samuel Issacharoff1 and Adam Littlestone-Luria2

I. INTRODUCTION

II. PUBLIC LAW IN PRIVATE HANDS

ASBESTOS: EXPERIMENTING TOWARDS A NEW PARADIGM MODERN CRISES, EX POST ANSWERS i. Vioxx ii. Deepwater Horizon iii. Takata C. OPIOIDS: EX POST INSTITUTIONS AS FEDERAL GOVERNANCE

A. B.

8 13 13 15 18 20

III. MINDING THE GAP

CONCLUSION

I. INTRODUCTION State capacity is under siege. Terrorism, pandemics, climate change, runaway technology, and the ever-shifting pressures of the modern mass market all stretch the orderly allocation of power. Twenty years ago, Bruce Ackerman advanced the thesis that these new challenges would so often demand unconventional answers that the American constitutional order needed to adopt formal emergency powers, mirroring the “state of exception” that has been incorporated into the fabric of modern constitutions across the globe3—a formal mechanism for recognizing the extraordinary in ordinary law. Under a conventional understanding of the state building process, we build the state before we need it. But Ackerman’s insight highlights the limits of this simplistic vision in the interconnected, dynamic world of our modern political economy. According to this orthodox conception of American governance, it is legislators who construct our institutions. The Constitution 1

Reiss Professor of Constitutional Law, New York University School of Law. University of California, Berkeley, PhD; New York University School of Law, J.D., Furman Academic Scholar; [FILL IN POSITION WHEN PUBLISHED – either Associate, Paul, Weiss, Rifkind, Wharton & Garrison OR Judicial Law Clerk, Southern District of New York]. 3 Bruce Ackerman, The Emergency Constitution, 113 YALE L.J. 1029 (2004). 2

Remedy Becomes Regulation

offers these elected lawmakers a template—a basic framework for how government is supposed to work. Beginning from these bare constitutional bones, they create the structures of administration and enforcement that make up our state. Once built, these institutions then address society’s problems as they arise. Institutional design is supposed to take place before the fact. In prior work, including a significant piece one of us published in a Clifford Symposium volume seventeen years ago,4 we both stressed that the need for institutional responses outside the patterns of ordinary law is far from occasional. These “states of exception” are no longer the exception. This pattern has become a central organizing principle of American governance. On this approach, after-the-fact American responses fit into a general theory of regulatory delegation—a “delegation state” that deputizes a range of actors—from agency rule makers, enforcers, and adjudicators to private attorneys general.5 Most theories of delegation begin with the rarely-spoken premise that the basic institutional structure is already in place; the act of delegation simply helps to determine who should wield the whip hand. When private lawyers sue to enforce antitrust or securities laws, for instance, they operate within a paradigm that would allow state regulators to take up matters directly, were the resources for public enforcement not lacking. But our present account advances the concept of the delegation state another critical step. Instead of focusing on private-public enforcement partnerships, we highlight a new kind of substitution for state capabilities—the creation and administration of institutions, built after the fact at the behest of private parties. A domain of litigation has emerged where litigants and courts assume the authority to create their own institutions by means of court-based remedies. Through a combination of private design and judicial power, new institutions take shape. These remedial structures take on functions that would normally be assigned to the state. Such ex post institutions reflect more than simply a deficit of enforcement resources, or some sense that private ordering would be more efficient. There are certain types of modern societal challenge that the state repeatedly proves incapable of addressing. After the fact—after catastrophe strikes—the legal process offers solutions. And these solutions are not simply enforcement actions, as would be the case in antitrust or securities fraud cases. They are something far more inventive. In this new form of “remedial regulation,” courts and lawyers may find themselves drawn into a creative act of institution-making. Aggregate litigation and settlement may be dragooned as tools of ex post state building. Together, litigants, lawyers, and judges take on a role that is usually viewed as the province of the legislature—sometimes moving in concert with executive- branch authorities, and sometimes acting alone. But private litigation drives 4 5

Samuel Issacharoff, Regulating After the Fact, CITE Adam Littlestone-Luria, Nondelegation’s Two Faces.

Remedy Becomes Regulation

the ship, leading to the formation of substantial, often highly formalized administrative programs that in many ways mirror the function of executive branch administration. Consider the allocation scheme for opioid remediation created in the Opioids MDL, which drew in local and state governments, and then created an oversight mechanism above them all—a species of administration that looks suspiciously like the oversight of states and localities by the federal government.6 Think about the bankruptcy structure created in Takata that sought to remedy a disastrous defect in automobile air bags by channeling compensation to victims, while keeping afloat the manufacture of properly functioning air bags and funding the compensation trusts.7 Or consider the massive recall administration created to process a half million returned Volkswagens after the emissions litigation.8 Examples abound. Asbestos litigation was one of the opening acts, beginning in the final decades of the twentieth century.9 Examples proliferate across the decades between asbestos and opioids, including drug and automotive cases, the Deepwater Horizon oil spill, and the NFL Concussion Settlement. Private litigants play a critical role in creating these ad hoc institutions. But there is a second step. These non- state actors are also critical to their execution and oversight. This private, ex post state making may not be a first best world. And it is not a panacea capable of solving all of the deficits of the modern mass economy. But, as we illustrate in a selection of examples from the past two decades, it is a critical element of how our system of governance now responds to crisis and unforeseen harm. If only by default,10 we have delegated the power to make the state to private litigants and their judicial partners. The observation that aggregate litigation can play some public or quasi- public role in American governance is not new. Scholars have long understood that class actions, multidistrict litigation, and bankruptcy have ramifications that ripple far beyond the named parties before the court. Complex cases do not just test the boundary between private and public law, they also crack open the divide between enforcement and administration. Writing just three years after Rule 23 was minted, Harry Kalven Jr. and Maurice Rosenfield captured this insight almost at once, noting the interactive relationship between class 6

CITE CITE 8 CITE 9 Judge Jack Weinstein of Eastern District of New York was an early pioneer of this phenomenon—what Martha Minow described, as early as 1997, as a “creator of temporary administrative agencies.” See Martha Minow, Judge for the Situation: Judge Jack Weinstein, Creator of Temporary Administrative Agencies, 97 COLUM. L. REV. 2010 (1997). Writing at the beginning of the events we describe, Minow could only begin to hint at their shape, scope, and critical place in our institutional culture. 10 Cf. Jennifer Nou, Subdelegating Powers at 501-02 (observing a related phenomenon of delegation by “default” in the agency context). 7

Remedy Becomes Regulation

actions and the administrative state—with the two serving both to check and complement each other.11 There is little doubt that aggregation helps transform otherwise private claims into a species of public law litigation. The corollary of this claim, as we have both explored in depth, is the insight that the “basic regulatory model” of the United States “is ex post rather than ex ante, a form of regulation that draws heavily on its common-law tradition.”12 Abstracting from these literatures, the relationship between private aggregation and state authority yields three basic phenotypes: cases that challenge the misdeeds of public governance; cases that complement public authority; and cases that supplement the limited reach of public authority, which may be viewed as a delegation of the Article II enforcement power to private agents. But these three models fail to capture something essential about the function of private power in today’s legal landscape. In our contemporary political and economic universe, when state power fails—due either to the vicissitudes of politics or to the inherent deficits of ex ante institutional design—there are many occasions where the legal process becomes a vehicle for institutional creation. The first category includes cases such as Brown v. Board of Education— alongside decades of litigation organized primarily under Rule 23(b)(2)—in which private collective power is harnessed to challenge perceived state misbehavior. As a matter of political economy, such cases set up the combined voice of litigants as a challenge to the claim of the state to speak as the sole representative of public, democratic authority. Second, there are circumstances in which private aggregation is used to complement public power. This model developed with the maturation of Rule 23(b)(3). Thus, while the SEC can force certain types of registration of corporate information, the private securities bar takes that information to challenge potential fraud in the securities market. Similarly, the FTC and DOJ have the power of enforcement under the antitrust laws, but they rarely have either the resources, or sometimes even the expertise, to secure actual damages—or, as is most relevant to the story we tell here, to distribute those funds to the affected population. Finally, there are cases where private authority is neither a challenge to public power (as in the civil rights injunction cases) nor a complement for the limitations of the administrative capacity of regulators (as with the enforcement side of economic harm in regulated markets); instead, in this third 11

Harry Kalven Jr. & Maurice Rosenfield, The Contemporary Function of the Class Suit, 8 U. Chi. L. Rev. 684 (1941); Richard A. Nagareda, Class Actions in the Administrative State: Kalven and Rosenfield Revisited, 75 U. Chi. L. Rev. 603, 604 (2008). In Vasquez v. Superior Court, 4 Cal. 3d 800, 812 (1971), Justice Mosk, writing for the California Supreme Court in recognizing the necessity of consumer fraud class actions, quoted extensively from Kaleb and Rosenfeld, extending their insights on the need for collective/ representative action by investors to unsophisticated consumers, who need such an infrastructure even more. 12 After the Fact, supra, at .

Remedy Becomes Regulation

category, private power acts as a substitute for the finite resources or scope of public authority. Much of litigated consumer law can be conceptualized in these terms, substituting for regulators in overcoming the small stakes of the disputes. As the Supreme Court put it in one of the icons of its late twentieth- century class action jurisprudence, private aggregate litigation can turn “relatively paltry potential recoveries into something worth someone's (usually an attorney's) labor.”13 In this article, our concern is not with a rivalrous challenge to the state, with the simple insufficiency of state resources, or with a domain too distant from the main activity of the state to merit the attention of public regulators. Instead, we highlight contexts in which litigation reveals the need for court- based remedies to become administration—to create remedial institutions with the institutionalized capacity to address harms that were beyond the boundaries of preexisting state authority. The mass harm legal system—quietly and in discrete and limited circumstances—assumes the presumptively legislative task of institution- building. It creates new structures of quasi-administration to address the harms caused by particular calamities. These institutions remain outside the normal boundaries of the administrative state and its complex formal rules of administrative procedure. Prior literature has focused primarily on the question of liability and the incentives to galvanize the power of private attorneys general to litigate claims as if they had been vested with public authority. As Stephen Burbank and Sean Farhang have detailed, beginning in the 1960s, Congress expanded the role of private enforcers in this “litigation state,” spurred by growing skepticism of the capacity of the President’s branch to regulate the nation and hold back the excesses of a modernizing economy.14 Complex cases do not just test the boundary between private and public law, they also crack open the divide between enforcement and administration.15 Thus, we do not focus on regulating after the fact through legal liability—the conventional focus of the literature on private enforcement. Instead, we highlight the creation of administrative entities after the fact through the process of litigation, settlement, and bankruptcy – “one-shot delegations” of a kind of governing power to aggregate entities16 These bureaucratic institutions look more to a bootstrapped form of public administrative law than to the doctrines of liability.

Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (quoting Mace v. Van Ru Credit Corp., 109 F. 3d 338, 344 (1997)). See Stephen B. Burbank & Sean Farhang, Rights and Retrenchment in the Trump Era, 87 Fordham L. Rev. 37, 38 (2018); Sean Farhang, The Litigation State: Public Regulation and Private Lawsuits in the United States 13, 24 (2010). 15 For this and what follows see NFT at 422 16 Nagareda, Class Actions in the Administrative State at 625. 14

Remedy Becomes Regulation

II. PUBLIC LAW IN PRIVATE HANDS Here, we arrive at this article’s critical contribution. Although it may be underreported in public discourse, complex litigation insiders are well aware that, at least in recent decades, the judicial process has often been used to generate persistent, quasi-administrative institutions. Courts are delegated powers that may more naturally be the province of another branch. Justice Kagan has derided this formalist fantasy of a strict separation of powers as the “Schoolhouse Rock definition”—a fictional government where the “atom of sovereignty” is neatly split among Congress, the President, and the courts.17 But what has not yet been fully captured is the fact that litigants are using the judicial process as a venue for institutional creation. Prompted either by legislative failure or by insuperable legislative incapacity, courts and litigants construct institutions after the fact that operate adjacent to the state and fill gaps in its scope and reach. To understand what this phenomenon is, it helps to emphasize what it is not. It is not just private enforcement—litigants bringing cases to implement the mandate of the underlying law. That is to say, we are not talking about a delegation of Article II enforcement power. Nor is it a delegation of the legislative capacity to make rules with the force of law—the complaint that the champions of a more muscular nondelegation doctrine lodge against administrative rulemaking.18 The institutions we describe are not, for the most part, lawmaking bodies, since their actions lack both the generalizability and the prospective focus of legislation. Finally, it is not a simple delegation of administrative power, although it contains elements of administration. Instead, we are seeing private litigants, usually represented by repeat- player plaintiffs’ counsel, using the courts to both design and implement bureaucratic institutions. Expressed in the conventional idiom of separated powers, this looks like a combination of legislative and executive power organized under the oversight of the judiciary. Certain key tendencies characterize this form of ex post state making. First, the process is often driven by the litigants and, especially, by repeat- player lawyers with broad experience in aggregate litigation. Judges and courts regularly participate in this creative process. But when they do, their role tend to be more responsive than proactive. Private actors design and build the programs—usually within the judicial process and but occasionally outside of it. Second, the court provides a venue that allows this privately driven institutional design to take shape and to gain the official stamp of government 17 18

Seila Law.

It is no surprise to anyone that courts make rules with the force of law. That is, after all, one of the fundamental premises of the common law. To the extent that American courts embody that tradition, that it a large part of their basic roInle.

Remedy Becomes Regulation

power. Private contracts can have some power, but the participation of the court has certain advantages— both pragmatic and sociological. Practically, courts can draw on certain powers unavailable in private contracts, such as the devices available under the Bankruptcy Code and the various tools contained in Rule 23 or the coordination procedures common within an MDL. These processes may offer defendants a greater degree of closure—approaching nearly to the “global peace” ideal—making potentially higher transaction costs worthwhile and perhaps even motivating them to offer greater rewards to injured parties.19 Third, although the judge does not conventionally serve as the primary driver, the court may act as a critical participant in the ex post institution, not merely as a venue for its creation. Judges and court personnel often remain involved in the administration of the institution, monitoring its progress and, at times, even playing a role in bureaucratic administration. For instance, the Northern District of California has created an add-on to Rule 23, instituting a detailed and demanding post-class settlement accounting and enforcement infrastructure to specify the ways in which the courts’ pre-existing continuing jurisdiction over class settlements is given teeth.20 And since the VW Clean Diesel MDL, the office and remit of the “Settlement Master” has become a recurring feature of MDLs, with these officials administering and regulating the private settlement on behalf of the court, through its appointed designee, who can regularly reports on progress to the judicial overseer. Fourth, other public actors may also play a complementary role. Lawyers from the federal executive branch may be pursuing parallel suits—whether civil or criminal. State attorneys general and litigants from state agencies and local governments may also join in the broader litigation landscape. But as much as government lawyers play a role, it is often the private actors who spur, design, and even administer the bureaucratic apparatus. Fifth, these institutions necessarily mix elements of legislation, execution, and administration. Typically, this is covered by an internal dispute resolution mechanism that oversees specific applications of remedial goals.21 In this way, they are a lot like administrative agencies, which are well known for simultaneously wielding the power to make rules, bring enforcement actions, and adjudicate disputes. Our ex post institutions bear many of the same features as executive branch agencies, but they add another of their own. By contractual agreement of the affected parties, they also feature something almost precisely analogous to the legislative capacity for institutional design.

See, e.g., Issacharoff & Rave, BP article. See Guidance on Class Action Settlements. 21 Such dispute resolution sometimes allows for judicial review, as in the Deepwater Horizon settlement (CITE to 5th Cir opinion on appealability), and sometimes provides for final resolution within the internal review structure. CITE. 20

Remedy Becomes Regulation

What are seeing then, in a series of high-profile cases across the past few decades, is a delegation to private litigants—and to some extent to courts—of the basic legislative function of institutional design and creation. Usually, we talk about delegation of power to make rules that are a lot like laws, or at times of enforcement power. The delegation at issue here is a form of legislative function we usually do not think of as delegable, or at least as conventionally delegated. And in contrast with the forms of delegation more common in conventional discourse, this is a delegation by default. It is not the outcome of some legislative choice, but the result of incapacity and failure on behalf of the actors formally tasked with this role within the American constitutional framework. But it is a delegation nonetheless. A look at some of the critical mass harm cases of recent years illustrates the characteristics of this new post-hoc approach to building state capacity. This story begins with the asbestos litigation in the latter decades of the twentieth century, which helped shape both the possibilities and the limits of ex post institution-making. The phenomenon proliferates across a series of examples from the first and second decades of the new millennium, and culminates in the opioids litigation—a grand, if halting, experiment in private governance.

A. Asbestos: Experimenting Towards a New Paradigm Asbestos seemed like a miracle. Once described as a “magic mineral”— tougher than steel, but so flexible that it could be woven into cloth—asbestos is waterproof, fireproof, and resistant to corrosion.22 The substance occurs naturally, found in abundance in both rock and soil, and its use goes back to the ancients, probably since the domestication of fire.23 Manufacturers embraced this marvelous resource, as a cheap and effective solution to thousands of production needs. First, asbestos was used as insulation. By the mid-twentieth century, it was everywhere: in hair dryers and children’s modeling clay; missile silos and ships.24 During the second World War, it became the lifeblood of the U.S. Navy, deployed in abundance across our wartime shipyards. At the height of industrial asbestos use in 1974, the nation consumed nearly 800,000 metric tons in a single year.25 But asbestos can kill. The substance lingers in the ambient air, and trace amounts lurk in much of the world's drinking water, in food, and in commercial

Barnes at 12. Robreno at 101. 24 Barnes at 12. 25 Robreno at 102. 23

Remedy Becomes Regulation

products. But above a certain threshold, asbestos is deadly.26 Asbestosis is a progressive pathology that has laced the lungs of many asbestos victims with potentially lethal scars, and various forms of cancer have attacked the lungs, chests, and abdomens.27 Pleural plaques can become lodged in the body—not always a source of visible symptoms, but quietly constricting the function of the lungs. Since asbestos played such a center role in the twentieth century economy, it touched millions, leaving a trail of destruction across America. As the disaster began to build in the 1960s, the thousands of workers exposed to asbestos as part wartime production and military service during World War II turned to workers’ compensation.28 These programs created absolute liability for workers’ medical costs and a percentage of wages. But they proved both stingy and uncertain, and they could not cover anyone who was exposed who had no employer to cover costs. Early efforts in Congress to create a legislative compensation program showed little sign that they could garner enough legislative buy-in. As the insufficiency of both the ex ante institutions and the legislative response became apparent, workers turned, instead, to tort. Litigation began slowly, but it grew into a flood. British studies first linked asbestos with asbestosis in 1924, and a few American lawyers then brought suits on behalf of U.S. victims.29 Refusing to give up such a valuable source of profit, however, manufacturers worked to suppress or minimize medical findings, settling the small number of suits under the condition that plaintiffs’ lawyers would hand over the evidence they had gathered and agree to not file further claims.30 A 1964 study broke through this wall of silence, definitively linking asbestos exposure to cancer.31 Then, a wave of lawsuits began. Filings started almost at once after the study was released, but the dam broke after the Fifth Circuit upheld a strict liability theory in Borel v. Fibreboard in 1973.32 By the 1980s, the number of cases had grown into an “avalanche.”33 The challenge was to find an institutional vehicle to address this “elephantine mass” of litigation.34 26

Robreno at 101-02 (detailing the wide presence of asbestos and describing its dangers when the fibers become “friable,” or damaged, and start to float through the air in certain dangerous quantities). 27 See Barnes at 12 (highlighting mesothelioma—a lethal cancer that attacks the chest or abdomen—and lung cancer). 28 See Barnes at 13 for this and what follows. 29 Macchiarola at 592. 30 Macchiarola at 592. 31 Macchiarola at 592. 32 See Robreno at 105 (citing Borel v. Fibreboard Paper Prods: Corp., 493 F.2d 1076, 1092 (5th Cir. 1973)). 33 Robreno at 106 (quoting Jenkins v. Raymark Indus., Inc., 782 F.2d 468,470 (5th Cir. 1986)). 34 Ortiz v. Fibreboard Corp., 527 U.S. 815, 821 (1999).

Remedy Becomes Regulation

Conventional aggregation mechanisms foundered. Standard pretrial and trial management processes at the state level enjoyed limited success.35 Experiments with consolidated trials became mired in due process challenges.36 Attempts to use collateral estoppel and alternative dispute resolution to resolve cases also prove largely unavailing.37 Class actions were some of the most prominent failures. After multiple refusals, the Judicial Panel on Multidistrict Litigation agreed to consolidate and transfer tens of thousands of pending cases for pretrial proceedings to the Eastern District of Pennsylvania.38 This move inspired hopes for the kind of comprehensive global settlement that would streamline and institutionalize the claims process for present cases, as well as for the deluge of claims that would inevitably arise in future.39 In Amchem v. Windsor in 1997, the Supreme Court affirmed the Third Circuit’s rejection of a prominent asbestos class action on Rule 23 grounds, and contemporary commentators highlighted the concerns with individual justice and the unwarranted advantages to defendants.40 Private plaintiffs had negotiated an agreement with a consortium of twenty companies, which would have created an institution known at the Center for Claims Resolution.41 But the Supreme Court rejected this attempt at ex post institution making, asserting that this kind of “legislative” action was not the province of the courts—at least not as a product of the Rule 23 process. Ortiz v. Fibreboard gave double confirmation to this restriction on the use of Rule 23 as a means of making bureaucracies to address the crises of the modern mass market. The justices held that the courts could not create a “limited fund” settlement class under Rule 23(b)(1) that would constrain recovery for present and future claimants.42 Although the Supreme Court exhorted the legislature to author a solution, Congress proved incapable of passing a global legislative fix. Legislators expressed ardent interest, introducing more than fifteen bills and prompting vigorous advocacy from individual senators and representatives.43 But these efforts proved unavailing. Consensus among stakeholders about the proper structure and financing of a solution remained out of reach. One of the features of the legislative process is the need for widespread consensus and buy in. 35

Robreno at 107-08. Robreno 108-09. 37 Robreno at 110. 38 Robreno at 112. 39 Robreno at 112. 40 See Robreno at 109-10 (citing Amchem Prods. Inc. v. Windsor, 521 U.S. 591, 597 (1997)). 41 See Robreno at 112-13 (citing Georgine v. Amchem Prod., Inc., 83 F.3d 610, 618 (3d Cir. 1996), aff'd sub nom. Amchem Prod., Inc. v. Windsor, 521 U.S. 591, 117 S. Ct. 2231, 138 L. Ed. 2d 689 (1997)). 42 See Ortiz v. Fibreboard Corp., 527 U.S. 815, 831-32, 863 (1999). 43 See Robreno at 114-17. 36

Remedy Becomes Regulation

Here, however, without the coercive powers and the negotiating venue of the courts, agreement was impossible to reach. Bankruptcy trusts proved far more effective both in terms of aggregating all claims and in creating an alternative administrative framework. Johns- Manville led the way—a Fortune 500 company that built its assets by manufacturing asbestos products.44 It filed for bankruptcy in 1982, and the bankruptcy court approved its plan in 1988. This plan created a trust that would allow the company to continue to function, premised on the understanding that the trust would be able to cover all claims. After Congress specifically authorized asbestos bankruptcy trusts under new Section 524(g) of the Bankruptcy Code,45 more the one hundred companies followed the Manville model and established their own bankruptcy trusts.46 Once these trusts are set up, tort litigation stops. Together, these institutions have aimed to get compensation into the hands of victims while doing as much as possible to keep the companies going. The trigger for the creation of a trust is a voting scheme that requires approval by seventy-five percent of the current claimants to support the bankruptcy workout. That means that lawyers for the claimants drive the process, since they are the gatekeeper for trust formation. In turn, these repeat players bring to the process substantial domain expertise.47 As the late Francis McGovern, a court appointed expert for more than ten of these trusts, put it: “These lawyers are the legislators, judges, and juries in deciding the appropriate mechanism for paying current and future asbestos personal injury plaintiffs.”48 Each trust becomes its own administrative institution, employing an expert staff and tasked with equitably distributing funds to the people that the manufacturer’s products has affected. After a claimant files a form detailing the exposure, medical, and personal history that are required to evaluate their claim, plan administrators attempt to process this claim by applying principles of horizontal equity.49 Most trusts offer claimants either an expedited review process or a longer, individual review that has the potential to render greater recovery.50 This results in an offer, which a claimant can either accept or further negotiate. These trusts have a substantial and ongoing administrative organization, employing between ten and one hundred employees and funded 44

Macchiarolo at 591-92. CITE 46 More the sixty remain active https://www.mesotheliomafund.com/asbestos-trusts/. 47 McGovern at 165-66. 48 McGovern at 166. 49 McGovern and 167-68. 45

2024.

See,

e.g.,

https://www.worthingtoncaron.com/Library/Legal/Asbestos-Bankruptcy-Trusts-The- Good-The-Bad-And-.aspx.

Remedy Becomes Regulation

by the trust assets with annual budgets that can total millions of dollars.51 Each trust follows complex internal guidelines and criteria—a set of internal regulations that systematizes the bureaucratic practice.52 Although many of the trusts retained access to jury trials if the internal negotiations mechanisms failed, they were designed to head off traditional tort litigation in favor of an internalized practice of administrative adjudication.53 Congress approved these new devices when it amended the Bankruptcy Code in 1994.54 The “channeling injunction” device was key to the success of the bankruptcy trusts, mandating that all claims against the corporation had to be settled through the trust and prohibiting suits against the corporation and its insurance carriers.55 Even nondebtor insurance companies were released as part of this process,56 setting up the bankruptcy trusts as the sole source of relief for those with claims against the debtor and its insurers. Although Section 524(g) only applies, by its terms, to asbestos cases, courts have come to rely on these same channeling injunctions and nondebtor releases in other mass tort contexts, relying on the equitable powers imbued in the bankruptcy court by the broad discretionary grant in the Code’s Section 105(a).57 These trusts are not without issues. Limited trust assets often compress recoveries over time, especially if the early projection of future claims understates the eventual tally. The institutional designers have struggled to figure out how best to protect the interests of future claimants whose diseases have not yet manifested without undercutting the interests of those already suffering. Critics highlight the due process concerns and the lack of public hearing for sufferers,58 and the questionable use of bankruptcy by solvent, nondebtor companies.59 But this bankruptcy option was the only solution at scale that made an appreciable difference in resolving the asbestos morass for a long time. Eventually, global aggregation in MDL-875 turned into its own claims processing institution— transforming the court into a mini-agency. Responding to the observation that aggregation and consolidation had proven ineffective in the asbestos context, the MDL judge and the parties realized that it was stopping progress on individual cases while the parties and the court 51

McGovern at 169-70. See McGovern at 170-75 (detailing exposure, medical, and valuation criteria). 53 McGovern at 175. Issacharoff serves as a Future Claims Representative in one of the asbestos trusts. 52

54 55 56 57

§ 524(g).

Simon at 1172. Simon at 1173.

Simon at 1174 (noting that § 105(a) was also the purported source of the power in the early asbestos bankruptcy cases). See, e.g., Foohey & Odinet. 59 See, e.g., Simon. 58

Remedy Becomes Regulation

labored to construct global solutions. The court “disaggregated” the cases into the lowest common denominator, embracing a “one plaintiff-one claim” model.60 Rather than holding to the single-judge paradigm, the MDL’s judge brought in a wider sweep of personnel, including four magistrate judges, additional district judges, the chambers staff of each, the Clerk’s Office, and assistance from the lawyers on the various cases.61 This MDL also employed Francis McGovern as a special master to facilitate a higher remand percentage than is conventional in the MDL context, where remand is usually discouraged.62 This MDL had to reject the “lone wol[f]” culture that is so common among district judges and to understand that “it takes a village” to cope with the administrative responsibilities of such a large MDL.63 Ultimately, this MDL managed to process more than 180,000 cases, and only a few thousand remain.

B. Modern Crises, Ex Post Answers Asbestos was an early augur of a new pattern: ex post institution making in courts—driven by private lawyers—in the face of the crises of the modern mass market. Three examples will help illustrate the evolution of this phenomenon: Vioxx, Deepwater Horizon, and Takata Airbags—a list that is by no means comprehensive but that can give a sense of the scope and variety of the institutions private litigants and courts have been crafting.

Vioxx

The Vioxx settlement strategy combined private settlement with judicial oversight, and it was a direct reaction to the successes and failures of the asbestos litigation that came before. Vioxx is a prescription painkiller that proved a blockbuster drug. In 2004 it became clear that prolonged exposure over an extended period causes some increase in the number of heart attacks and strokes. But heart attacks and strokes are not a signature harm unique to this drug. In each individual case, the only proof connecting Vioxx to the particular event was epidemiological.64 Litigation results were incredibly discordant, with some plaintiffs receiving nothing while others were awarded

Robreno at 127. Robreno at 127-30. 62 Robreno at 147. 63 Robreno at 188. 64 Issacharoff at 216. 61

Remedy Becomes Regulation

millions.65 Congress held hearings,66 but no legislative action was forthcoming. After Amchem and Ortiz, no class settlement seemed plausible, so litigants experimented with a novel settlement process that would allow the company to escape from under the shadow of litigation threat while offering the plaintiffs reasonable compensation. This settlement that the lawyers designed was not a contract between Merck and the plaintiffs in their own capacity. Instead, it was a contractual agreement between the manufacturers and the law firms, requiring any lawyers to recommend the settlement structure to their clients and, consistent with professional rules on representation, to withdraw from representing any nonsettling clients and release any claim for referral to new lawyers that might agree to represent their erstwhile clients.67 Any client who declined to settle would lose her lawyer and might then discover that every other lawyer with preexisting experience in the litigation was unavailable to represent her.68 In terms of practical compensation and closure, this settlement was highly satisfactory for nearly all of the parties.69 But its innovative structure triggered questions about the ethical constraints on the treatment of unwilling plaintiffs.70 To distribute the $4.85 billion settlement fund, the litigants erected a special-purpose claims facility. It was set up to process claims according to streamlined procedures negotiated by defendant’s lawyers and the plaintiffs’ leadership.71 The process was exceptionally public, with a defined eligibility process scored by the claims administrator and staff.72 One critical feature of the settlement was that any claimant had to pass through a series of “gates,” which determined proper length of exposure, past medical history, and actual compensable injury. Unsuccessful claimants would be have their claims disallowed within the settlement process. But they remained formally free to pursue their claims on an individual basis. The mass settlement was structured as a set of unilateral offers by Merck to qualifying claimants, and to them alone. Parties contracted with the settlement administration company BrownGreer to handle day-to-day administration.73 But the management was not all private. Vioxx may have been resolved by a formally private contract—a nonclass aggregate settlement—but courts and judges played critical roles throughout the process. Litigation was 65

Issacharoff at 216. See Erichson & Zipursky at 277. 67 Erichson & Zipursky at 266. 68 Erichson & Zipursky at 266. 69 Erichson & Zipursky at 267. 70 Bradt & Rave at 114. 71 Noll at 410, n.104 72 See Erichson & Zipursky at 279. 73 Erichson & Zipursky at 266. 66

Remedy Becomes Regulation

centered in an MDL in the Eastern District of Louisiana and the MDL court selected the leadership group and allowed for juries to test the likely value of the claims in a series of bellwether trials. There were nineteen trials that helped determine the settlement terms, with six in the MDL court.74 Experienced plaintiffs attorneys shepherded these bellwether cases through the process, conceived as a mechanism for determining the size, scope, and distribution of compensation. The Vioxx MDL furnished a venue for the coordination of the settlement, but Judge Fallon was also pressed into service as “chief administrator” of this claims facility, overseeing the settlement, granting it a stamp of judicial legitimacy, and even taking on a day-to-day management role.75 This settlement was extraordinarily successful in its stated aims, settling 99.79% of eligible claims.76 Moreover, it benefited from review by an experienced and respected judge, who remained involved even after the moment of institutional design—present to resolve disputes about interpretation and compliance.

ii.

Deepwater Horizon

On April 20, 2010, an explosion on the Deepwater Horizon oil platform sent more than 200,000 gallons of oil flooding into the Gulf of Mexico, in addition to deaths and injuries among crewmen. The federal government showed keen interest in this disaster, with personal involvement from President Obama, a variety of government investigations and commissions, and civil and criminal lawsuits from the Department of Justice. But nothing in this maelstrom of state action promised to make whole the communities of the Gulf that faced the economic dislocation resulting from the spill. The dispute was extraordinarily complex—precisely the form of dispute that Lon Fuller described as “polycentric”—not readily susceptible to traditional forms of adjudication.77 Although the government response addressed certain limited aspects of the problem, the main gap remained unfilled: the need for compensation of all those whose lives and livelihoods had been disrupted. Instead, it was the private institution making that worked to address the hole in our safety net. This ex-governmental, ex post response came in two chapters. First, BP set up the Gulf Coast Claims Facility—a fully private administrative mechanism designed to short circuit the litigation process, avoid its attendant transaction costs, and deflect the massive negative publicity

Baker & Bradt at n.96. See Baker & Silver at 70. 76 Erichson & Zipursky at 266. 77 Rave (2023) at 298. 75

Remedy Becomes Regulation

such a disaster would inevitably entail.78 Drawing on the example from Aqua Dots, where the Seventh Circuit had held that a private settlement program could obviate the need for litigation, BP designed an elaborate bureaucracy to channel funds to spill victims.79 The oil company paid all expenses and set aside $20 billion in a trust fund following the commands of the Oil Pollution Act, with committing to add more should that prove insufficient. In consultation with the Obama administration, they chose Ken Feinberg, who had already worked as a special master on the September 11th Victim Compensation Fund, to manage the massive undertaking. Using a compensation scheme designed around strict liability standard and a simplified claims procedure, GCCF paid out $6.2 billion to more than 220,000 claimant in a mere eighteen months.80 But the GCCF approach soon revealed significant limitations. While GCCF could process claims, each proposed resolution was in the form of a one-off contractual offer, with no ability to ensure any closure, or even that some claimants might not hold out for more than others. However effective the private offer and acceptance model may have been at distributing funds on a one-by-one basis, it lacked something critical. It did not have the ability of court oversight or of private engagements with representatives of the many affected communities by the Gulf oil spill to harness, channel, and coordinate the resolution. Litigation stepped in to fill this void, creating a second administrative mechanism that could offer both more closure and greater compensation. A slew of lawsuits filed on behalf of individuals and groups of clients were consolidated in an MDL in the Eastern District of Louisiana under Judge Carl Barbier.81 Transaction costs may have been higher, with complex proceedings involving a spiderweb of defendants and third parties, overlapping bodies of law, dozens of reports from experts, hundreds of depositions, and over ninety million pages of discovery documents.82 But finality had immense value for BP.83 This peace premium made it worth their while to offer higher payments to injured parties under the MDL class settlement than they had through the GCCF.84 To channel these payments to the thousands of injured parties, the litigants set up a massive, multistate agency. It was run by a court-appointed Claims Administrator, Patrick Juneau, along with his personal staff of twenty- five who, in turn, hired five claims vendors, employing more than 3,200 local 78

Dodge at 377 (noting that many high-profile uses of uses of private settlement funds have resulted from public relations disasters). 79 On the Aqua Dots connection, see Rave at 92. See also Issacharoff & Rave at 429 (noting that the 9/11 Compensation Fund serves as a model for private claims facilities) 80 Issacharoff & Rave at 399. 81 Issacharoff & Rave at 400. 82 Issacharoff & Rave at 401. 83

Note that BP ultimately challenged the settlement, although it originally supported both class certification and settlement approval. Issacharoff & Cabraser at n.83. 84 See Issacharoff & Rave at 424 (explaining why plaintiffs were able to credibly offer peace).

Remedy Becomes Regulation

personnel in offices throughout the country.85 Headquartered in Louisiana, the organization was spread across nineteen “Claims Assistance Centers” throughout the Gulf Coast region.86 In a trend that has become common in some of the more prominent recent MDL settlements, the process took on a highly participatory cast. Plaintiffs remained in close contact with class counsel, and they often maintained independent representation of their own.87 They were capable of making active demands, and maintained structures of intermediate organization through digital channels such as Facebook and Twitter that allowed them to function as a group and develop their own coherent voice.88 Nuances of litigants’ personal interests were integrated into this process in a kind of ad hoc minidemocracy—a notable contrast with the usual expectation that class actions must, by their nature, shut out litigant participation.89 Through their own organization and through monitoring by personal counsel, the class members were able to help the lawyers and the court to structure an administrative mechanism that was better designed to take their actual interests into account. As a consequence, perhaps with more specificity than would have been possible from any ex ante program, the private lawyers were able to build an institution fitted for the moment. This participatory model was not unique to BP. There are at least two other close parallels: the NFL Concussion settlement and the Volkswagen Clean Diesel trust.90 In all three of these examples, a class settlement emerged from an MDL. Private lawyers—often many of the same players—led the charge, with their efforts influenced by input from a highly engaged class, with many of the individual litigants still represented by their own private lawyers. Institutional design was informed by these miniature, timebound, single-issue democracies, and it resulted in expansive and lasting administrative institutions custom-built for the problem and its moment. An additional value from the claimant perspective that the formalized, court-approved, supervised private settlement offered was transparency. All claim forms and criteria were in public view on a website, and the public loss calculation formulas ensured consistency of awards among claimants who operated in the same business or employment category, ensuring consistent damages. Administration and distribution of funds was similarly transparent. The Settlement Administrator filed and posted monthly reports, often exceeding sixty pages, with charts and graphs describing every aspect of the 85

Noll at 420. Noll at 420. 87 Cabraser & Issacharoff at 865. 88 Cabraser & Issacharoff at 865. 89 See Cabraser & Issacharoff at 877. 86

See Issacharoff & Cabraser at 850-51 (connecting these cases). Issacharoff served as counsel to the claimants in all three of these cases.

Remedy Becomes Regulation

process and payments in exhaustive detail. The class settlement structure even had its own investigation into potentially fraudulent claims. By contrast, the prior Feinberg regime was a black box. It merely led to a series of private negotiations, without affording the claimants or the public any window into what their neighbors were getting. In the end, then, the private settlement offered more compensation, greater participation, and enhanced visibility.

iii.

Takata

Beginning in the early 2000s, a sequence of revelations about the safety of Takata’s airbags led car manufacturers like Honda to recall vehicles equipped with the these components.91 Concerns increased following a series of fatal accidents involving Takata airbags, such as the death of 18-year-old, Ashley Parham, whose carotid artery was severed by shrapnel from her exploding airbag after her car bumped into another vehicle in a parking lot.92 By 2015, Takata had issued a recall for 34 million vehicles. The economic and human costs of Takata’s defective airbags have been staggering. Sixty-seven million vehicles have been recalled between 2013 and 2023—the largest recall program in U.S. history.93 As of 2021, defective Takata airbags accounted for over 400 injuries and 18 deaths.94 By 2017, T.K. Holdings, Takata’s U.S. arm, had filed for Chapter 11 bankruptcy citing liabilities between $10 and $50 billion.95 The response to Takata’s exploding airbags shows how the asbestos legacy continues to inform the administrative response to mass harms. As with the asbestos bankruptcy trusts, in Takata, the manufacturer’s bankruptcy became a vehicle for remedial institutional design. Takata had been a leading player in the development and manufacture of car safety equipment for more than fifty years.96 The litigation demands of the injured quickly threatened to overwhelm Takata’s limited resources. At the same time, replacing the defective airbags was impossible without Takata itself taking a lead role in ongoing airbag production. Supply chains were limited, and without Takata’s

91 Events Lead to Air Bag Maker Takata’s Bankruptcy Filing, ASSOC. PRESS (June 26, 2017) https://apnews.com/bc1276ef8522483194e336d3f151c15e/Events-leading-to-air-bag-maker- Takata%27s-bankruptcy-filing 92 Id. 93 Erica Werner, BMW Issues Warning, Ford Recalls Thousands of Older Vehicles Over Air Bags, WASH. POST (May 5, 2023) https://www.washingtonpost.com/business/2023/05/05/bmw- ford-takata-air-bags-recall/ 94 Kurdi, supra note Error! Bookmark not defined.. 95 Events Lead to Air Bag Maker Takata’s Bankruptcy Filing, supra 91 96 Fouad Kurdi, Channeling Injunctions in Hindsight: Lessons Learned from the Takata Airbag Trust, in NORTON ANN. SURV. OF BANKR. L. 157 (2022).

Remedy Becomes Regulation

participation, the industry could not have produced enough of the components to fill the demand.97 Part of the initiative relied on formal government enforcement. Both civil and criminal laws were implicated by Takata’s conduct and the NHTSA was able to mandate recalls and other after-the-fact remedial orders.98 The Department of Justice and U.S. Attorney for the Eastern District of Michigan negotiated a plea deal with Takata that required the company to pay $850 million in restitution to car manufacturers, and $125 million to individuals injured or killed by exploding airbags. These funds were placed in a trust called the Takata Individual Restitution Fund (IRF).99 But private institutional design may have been even more important to the resolution of this crisis than these actions by the state. Takata’s bankruptcy culminated with the sale of the company’s assets to competitor Key Safety Systems for $1.6 billion. The new owner did not acquire the liabilities related Takata’s faulty airbags, however.100 To address these liabilities, Takata established the Takata Airbag Tort Compensation Trust Fund (TATCT), initially funded with $125 million.101 Through a channeling injunction that echoed the asbestos precedent, all existing and future personal injury and wrongful death claims would be channeled towards the trust fund.102 The TATCTF also increased compensation available for victims by establishing a mechanism for other automakers who faced joint and several liability to participate in the trust fund. Participating OEMs were required to make an initial payment to the trust and contribute to the compensation of claimants who had been injured while driving a vehicle they had manufactured.103 In exchange they were able to benefit from the channeling injunction. Thus, the TATCTF created space for institutional flexibility—a mechanism that could coordinate contributions among multiple responsible parties who might be subject to suit. By securing compensation for harmed individuals, these trusts have satisfied a need that social and health insurance schemes address in societies that opt for a different institutional arrangement. To provide financial compensation to the victims, the courts overseeing the Takata cases had to work with the litigants—both government enforcers and private attorneys—to 97 In fact, even with Takata involved, the industry has still struggled to manufacture enough of the airbags to fill the new requirements. 98 For a discussion of the transformation of the NHTSA’s approach to regulating the auto industry, see generally Jerry L. Mashaw, David L. Harfst, From Command and Control to Collaboration and Deference: The Transformation of Auto Safety Regulation, 34 Yale J. on Reg. 167 (2017). 99 Kurdi. 100 Lindsey D. Simon, Bankruptcy Grifters, 131 YALE L.J. 1154, 1177 (2022) 101 Kurdi. 102 Simon at 1177. 103 Simon at 1178.

Remedy Becomes Regulation

create new structures and processes that replicated bureaucratic and administrative functions often viewed as the province of the state. Both Takata trusts adopted an administrative approach which standardized and streamlined the factfinding process, working to avoid the substantial resources litigants usually have to expend proving causation and damages.104 Instead, claimants could submit a standardized form to both trusts, which would be processed using a highly developed set of regularized procedures and grids for evaluating compensation.105 Guided by principles of equity and efficiency, the trustees are tasked with administering these procedures, and they are vested with substantial discretion and responsibility. The trusts even built in a substantial administrative appeals process106—an echo of the institutional design common across many agencies.

C. Opioids: Ex Post Institutions as Federal Governance In 2017, the Department of Health and Human Services labeled the opioid crisis a “public health emergency.”107 During the decades prior, the calamity had grown, swelling from a minor nuisance into a nationwide scourge. Between 1999 and 2020, opioids stole 450,000 lives.108 Over the course of 2019 alone, 50,000 victims succumbed, dwarfing the death toll from car crashes or gun violence.109 By 2020, the federal government estimates 68,000 people died from opioid overdoses110—a menace far more deadly than terrorism or war. Fatalities only hint at the cost. Just on the economic front, a 2017 report from the Council of Economic Advisers, the crisis costs the nation $500 billion each year—roughly 3% of U.S. gross domestic product.111 The public 104

Kurdi. See, e.g., PSAN/WD Trust Distribution Procedures at 20-22, In re TK Holdings Inc., No. 17-11375 (BLS), 2018 WL 1306271 (Bankr. D. Del. Mar. 13, 2018) 106 Kurdi (noting that the only issues that can be contested through the claimants lawsuit are injury causation and damages and that punitive damages are also not available to opt-out claimants). 107 Catherine M. Sharkey, The Opioid Litigation: The FDA is MIA, 124 DICK. L. REV. 669, 670 (2020). 108 Nora Freeman Engstrom & Robert L. Rabin, Pursuing Public Health Through Litigation: Lessons from Tobacco and Opioids, 73 STAN. L. REV. 285, 287 (2021). 109 Nora Freeman Engstrom & Michelle M. Mello, Litigation is Critical to Opioid Crisis Response, DAILY JOURNAL, https://www.dailyjournal.com/articles/351533-litigation-is-critical- to-opioid-crisis-response (Mar. 13, 2019) 110 Overdose Death Rates, NAT’L INST. ON DRUG ABUSE figs. 1 & 4 (Jan. 20, 2022), https://nida.nih.gov/drug-topics/trends-statistics/overdose-death-rates (16,000 of these were from prescription drug overdoses). 111 See COUNCIL OF ECON. ADVISERS, THE UNDERESTIMATED COST OF THE OPIOID CRISIS 1 (2017). 105

Remedy Becomes Regulation

response was haphazard at best and uniformly ineffectual. As Catherine Sharkey argued in 2020, federal agencies such as the FDA have been conspicuous largely in their absence, implicated in licensing the drugs, but “MIA” in the efforts to restrict them.112 Congress did nothing until 2016. Then, members started workshopping proposals.113 But the bills it passed did nothing to address the core issues. The executive branch has proved just as feeble. Many individual victims fell into a “coverage gap.” State Medicaid programs typically cover medication-assisted treatment for opioid use disorder.114 But Medicaid only covers households with severely constrained resources. According to the Census Bureau, twenty five million Americans lacked health insurance for an entire calendar year in 2022.115 Fiscal pressure from the COVID-19 pandemic led many states to slash budgets for opioid programs.116 As the National Association of State Alcohol and Drug Abuse Directors put it, these budget cuts “tak[e] an already at-risk population and really cut[] the legs off from under them.”117 The breakthrough came with the 2017 centralization of litigation in an MDL in front of Judge Dan Polster of the Northern District of Ohio. State and local governments were key movers. These entities had filed suits that included claims sounding in public nuisance that sought funds to abate opioid addiction and provide increased resources for the strain on public services such as law enforcement and emergency response.118 For example, Ohio’s Summit County claimed it had spent $66 million responding to the catastrophe.119 New York City filed a similar suit, alleging “hundreds of millions of dollars” in costs associated with responding to opioid scourge.120 Although the states as such were not formally in the MDL, that was the undisputed locus of activity. Under Judge Polster’s supervision, forty six states agreed to a $26 billion settlement with Janssen—the parent company of opioids manufacturer 112

Sharkey, supra note . Ryan Mutter & Noelia Duchovny, Recent Federal Responses to the Opioid Crisis, NEJM CATALYST (Dec. 13, 2022), https://catalyst.nejm.org/doi/full/10.1056/CAT.22.0345 (detailing three recently passed bills). 114 All state Medicaid programs cover at least one medication used as part of MAT and most cover all three of these medications. Kendal Orgera & Jennifer Tolbert, The Opioid Epidemic and Medicaid’s Role in Facilitating Access to Treatment, KFF (May 24, 2019). 115 Katherine Keisler-Starkey et al., U.S. CENSUS BUREAU, HEALTH INSURANCE COVERAGE IN THE UNITED STATES: 2022 (2023) 116 Amy Sokolow, Opioid Overdoses Have Skyrocketed Amid the Coronavirus, But States Are Nevertheless Slashing Addiction Treatment Program Budgets, STAT (July 16, 2020) https://www.statnews.com/2020/07/16/opioid-overdoses-have-skyrocketed-amid-the- coronavirus-but-states-are-neverthelessslashing-addiction-treatment-program-budgets/. 117 Id. 118 See Ausness, supra note Error! Bookmark not defined., at 203. 119 Gluck et al., supra note Error! Bookmark not defined., at 355. 120 Id. 113

Remedy Becomes Regulation

Johnson & Johnson—and three Opioid distributors: AmerisourceBergen, Cardinal Health, and McKesson.121 These defendants agreed to a separate $590 million settlement with native tribes and tribal health organizations.122 In late 2022, CVS, Walgreens, and Walmart—three of the largest pharmacy chains—agreed to a $13.8 billion settlement to resolve lawsuits with state, local, and tribal governments.123 The parties announced in early 2023 that enough states had agreed to the settlement to move forward, even though those states were not formally part of the MDL.124 In total, states and local governments are slated to receive over $50 billion from pharmaceutical opioid manufacturers, distributors, and retailers.125 Most striking from our perspective is not simply the scale of this settlement, but its form. The parties—led by lawyers on both sides who had participated in many of the earlier mass administrative settlements—quickly recreated an elaborate administrative process to guarantee that the settlement’s resources were directed to remediation of the opioid crisis, and not repurposed for other uses, and to ensure that the funds were utilized efficiently. Learning from past experience, these lawyers design an administrative system with ongoing authority to monitor compliance with the aims of opioid abatement. These moves were prompted, in part, by perceived failures with the Tobacco Master Settlement, which was widely seen as ill-targeted and unconstrained in the use of funds.126 As of November 2023, at least eighteen states had approved spending plans, ten had awarded funds to specific abatement programs, twelve had published annual reports on settlement-related activities, and nine were reviewing applications from local government and treatment programs.127 121

Sam Mermin et al., NAT’L ACAD. FOR STATE HEALTH POL’Y, Understanding Opioid Settlement Spending Plans Across States: Key Components and Approaches (Dec. 8, 2022) https://nashp.org/understanding-opioid-settlement-spending-plans-acrossstates-key- components-and- approaches/#:~:text=Allowable%20uses%20of%20funding%3A%20At,expenditures%20exce pt%20where%20this%20Agreement 122 Id. 123 Id. 124 Executive Summary of National Opioid Settlements, NAT’L OPIOIDS SETTLEMENT (Sept. 8, 2023) https://nationalopioidsettlement.com/executivesummary/#:~:text=In%20all%2C%20the%20Distributors%20will,additional%20manufacturer s—Allergan%20and%20Teva. 125 Christine Minhee, OPIOIDSETTLEMENTTRACKER.COM, Global Settlement Tracker, (last accessed Dec. 1, 2023) https://www.opioidsettlementtracker.com/globalsettlementtracker. Although Issacharoff helped represented the city and county plaintiffs in the MDL and was involved in the opioid settlement efforts, the discussion that follows is based exclusively on publicly available information. 126 See Gluck at 358. 127 Sam Mermin & Katie Greene, An Early Look at State Opioid Settlement Spending Decisions, NAT’L OPIOIDS SETTLEMENT (JUNE 12, 2023) https://nashp.org/an-early-look-at- state-opioid-settlement-spending-decisions/

Remedy Becomes Regulation

To prevent misdirection of the resources, the Global Settlement contains provisions designed to focus their use. Exhibit E of the 2021 Settlement stipulates that states should spend settlement funds on nine core strategies, such as reversing opioid overdoses with Naloxone and broadening the scope of recovery services to include co-occurring substance use disorders or other mental health conditions.128 The core strategies provide a comprehensive approach that addresses the opioid epidemic from many angles. These range from addressing opioid use disorder among incarcerated individuals, to supporting pregnant and postpartum women and targeting neonatal abstinence syndrome. In addition to their breadth, the core strategies provide actionable policy recommendations. For example, under Core Strategy C—“Pregnant and Postpartum Women”—the settlement prescribes highly specific policy interventions for states, detailing screening methods, particular recovery services, and priority populations.129 Mirroring the distribution of funds by the federal government to states, the Global Settlement pairs defined conditions with flexibility. Its drafters took care to provide jurisdictions with “significant discretion in how they spend the funds.”130 The settlement stipulates that “priority shall be given to the . . . core abatement strategies.”131 But it also allows states to spend settlement funds on any policies that support the treatment of opioid use disorder or co-occurring substance use or mental health disorders, as long as these policies are either evidence-based or informed by reputable evidence.132 These programs “may include, but are not limited to”133 twelve broad categories of approved uses.134 Approved uses are divided into three thematic areas: treatment—of opioid use disorder, for instance, or support people in treatment and recovery; prevention—of overprescription, for example; and other strategies such as supporting first responders and improving training and research.135 Like the core strategies, each approved use is accompanied by concrete steps states can take to ameliorate the opioid epidemic. With the first approved use, “Treat Opioid use Disorder,” states are offered fourteen potential strategies. States may “[e]xpand availability of treatment for OUD and any co-occurring [substance use disorder/mental health] conditions, including all forms of Medication-Assisted Treatment (“MAT”) approved by the U.S. Food and 128

DISTRIBUTOR SETTLEMENT AGREEMENT E-4 to -3 (2022). Id. at E-2. 130 Johns Hopkins Bloomberg School of Public Health,Primer on Spending Funds from the Opioid Litigation: A Guide for State and Local Decision Makers 3 (2022). 131 DISTRIBUTOR SETTLEMENT AGREEMENT, supra note Error! Bookmark not defined., at E-1. 132 Id. at E-4. 133 Id. 134 Id. 135 Id. at E-4 to -15. 129

Remedy Becomes Regulation

Drug Administration,” for instance, and “[o]ffer fellowships for addiction medicine specialists for direct patient care, instructors, and clinical research for treatments.”136 In another echo of federal government-style oversight, the settlements aim to put in place a systematic approach to the distribution of funds among government subunits. Private settlements have long utilized compensation grids to address patterns of harm in which fine-grained individual determinations are either impossible or too costly within the limited resources of a particular case resolution. Necessarily, the grid approach is less attuned to each harmed individual’s unique circumstances. But it is more efficient. The opioids settlements allocated among states and local governments using a grid, following the path created in asbestos, Deepwater Horizon, VW emissions, and the other exemplars of this private administration trend. Rather than requiring each government to document the fiscal burden of the opioids crisis, the settlements divided available funds using pre-determined criteria. These criteria aim to capture each states’ proportionate share of the impact of the epidemic.137 Under the 2021 Settlement, for instance, the maximum potential allocation for each state is determined by the state’s population, adjusted for “(1) the amount of opioids shipped to the state; (2) the number of opioid-related deaths that occurred in the state; and (3) the number of people who suffer opioid use disorder in the state.”138 The settlement also created a default allocation of funds between states and subdivisions, although states were able to change the agreement through a qualifying agreement, statute, or statutory trust.139 The opioid settlements foster efficiency by replicating the decentralized approach of a federal agency to oversight and monitoring. To minimize transaction costs, the settlement does not require that states report all spending of settlement funds to the Settlement Administrator. States only need to report spending that is not used for opioid remediation.140 To provide supervision, the settlement requires that states establish advisory committees, or other similar entities, to “provide input on spending decisions from statewide Abatement Accounts.”141 States must adopt formal guidelines on advisory group functioning— detailing their composition and process, equalizing state and local representation, creating a process for receiving input from 136 137

Id. at E-4. Frequently

Asked

Questions about the National Opioid Settlement, https://nationalopioidsettlement.com/faq-explanatorycharts/faq/#:~:text=Similarly%2C%20the%20sub%2Dpercentages%20for,%3B%20and%20(3 )%20the%20n0 umber (last accessed Dec. 1, 2023). 138 Id. 139 Mermin et al. 140 Mermin et al., supra note 121. 141 Id.

NAT’L

OPIOIDS

SETTLEMENT

Remedy Becomes Regulation

subdivisions and from the public, and putting into place a mechanism through which appropriate state agencies will consider their recommendations.142 To date, nineteen states have set up advisory committees that provide non-binding recommendations, while fourteen have advisory committees that directly disburse or award grants from their settlement funding.143 This decentralized approach to managing settlement spending minimizes administrative costs by shifting the burden of monitoring and guiding state spending to state- sponsored entities. Again, as with every one of the prior examples, it has been the repeat player lawyers and experts driving the ship, designing creative devices to move the litigation through the courts and creating the parameters of the settlement that will guide the nation’s response. Here, the most radical procedural innovation was the “negotiation class”—a procedural device created specifically for the opioids litigation to resolve the intractable challenge of complete closure. Because of the massive number of localities with claims in the opioid litigation, defendants worried that even a settlement that addressed the claims of the thousands of parties before the court would not create global peace, since it could not preclude the other 30,000 localities from suing.144 Instead, it would spur more litigation by proving that recovery might be attainable.145 To solve this problem, two professors, Francis McGovern, who was serving as a special master, and William Rubenstein, a court-appointed expert, championed a novel device to facilitate binding negotiations, taking advantage of the Rule 23 class-action framework.146 Following the contractual aggregation model championed by Section 3.17 of the ALI’s Principles of Aggregate Litigation,147 McGovern and Rubenstein used a voting mechanism mandating that seventy-five percent of each voting pool must approve of any deal class counsel would negotiate for the class to create “a voting arrangement” that would allow all of the implicated localities “to participate collectively, through their representatives, in any settlement discussions.”148 By invoking some of Rule 23’s structural protections, such as notice to the

142

Id. Id. 144 See Report on Negotiation Class at 2, In re Nat’l Prescription Opiate Litig., No. 17- md- 02804 (N.D. Ohio Dec. 2, 2019), ECF No. 2959 (stating that there are 34,458 class members). 145 Gluck & Burch, supra note , at 29. 146 Francis E. McGovern & William B. Rubenstein, The Negotiation Class: A Cooperative Approach to Class Actions Involving Large Stakeholders, 99 TEX. L. REV. 73, 74 (2020). 147 CITE 148 Plaintiffs’ Notice of Motion and Motion for Certification of Rule 23(b)(3) Cities/ Counties Negotiation Class at 2, In re Nat’l Prescription Opiate Litig., No. 17-md-02804 (N.D. Ohio June 14, 2019), ECF No. 1683. 143

Remedy Becomes Regulation

class, opt-out opportunities, and objections, the class arrangement might create the legitimacy to bind localities that had not yet sued.149 Although Judge Polster embraced the device,150 a divided panel of the Sixth Circuit reversed, relying on a formal reading of Rule 23.151 Despite the formal reversal, the negotiation class framework served to structure the subsequent allocation of settlement benefits and the aims of the private ex post institutions that emerged from the massive settlements.152 With buy-in from parties, even the unforgiving gaze of the ordinary legal process was not been able to halt the quiet quest for creative solutions. The negotiation class might have lost two to one in the Sixth Circuit, but it created a template for private institutional design. Even in the face of appellate courts’ due process-induced hesitancy, private design emerged triumphant—if quiet.153 In spite of due process formalism, ex post state building won out, making the power of the courts into a vehicle for private statecraft.

III. MINDING THE GAP The burden of this article is to flesh out how seemingly ad hoc institutional arrangements fit into the overall regulatory picture in the United States. It has long been understood that, “[c]ompared to other economically advanced democracies, the United States more often relies on courts, lawyers, legal threats, and legal contestation in making and implementing public policies, in compensating accident victims, in striving to hold governmental officials accountable, and in regulating economic behavior.”154 Accountability under tort law and a predictable private pattern of enforcement are both central within this universe of ex post regulation. In mature areas of tort liability, from auto accidents to asbestos exposure, this reliance on private enforcement will “inevitably” yield a private compensation system that substitutes for the insufficiency of public regulation and public compensation.155 Regulating After the Fact detailed the benefits of this liberalized economic order. It suggested that “[n]o country seems to realize the benefits of wide open markets, of relaxed 149

All localities were furnished with an allocation map that showed each entity’s settlement portion. Allocation Map, OPIOIDS NEGOTIATION CLASS, https://allocationmap.iclaimsonline.com (last updated June 18, 2019, 12:53 PM). 150 Gluck & Burch, supra note , at 31. 151 In re Nat’l Prescription Opiate Litig., 976 F.3d at 676 (quoting Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 620 (1997)). The Sixth Circuit denied the petition for en banc rehearing. Order, In re Nat’l Prescription Opiate Litig., No. 19-4097/19-4099 (6th Cir. Dec. 29, 2020). 152 These terms can be found informing the settlements negotiated through the National Opioids Settlement website. https://nationalopioidsettlement.com/ 153 Of course, this triumph has limits. The opioids crisis is ongoing, and the remedial settlement has only begun to scratch the surface. 154 ROBERT KAGAN, ADVERSARIAL LEGALISM 3 (2d ed. 2019). 155 Issacharoff & Witt, Inevitability.

Remedy Becomes Regulation

entry into the world of commerce, and of economic dynamism as fully as the United States.”156 Since Regulating After the Fact was published in 2007, U.S. economic growth has dramatically outpaced its peers in other advanced economies. Between 2007 and 2022, U.S. GDP per capita increased by fifty-nine percent, a substantially more impressive gain than the ten percent growth experienced by European Union countries in the same period (see

Figure 1).157 In 2007, the gap between U.S. and EU countries GDP per capita was $14,113. By 2022 the gap grew to $38,992, an increase of 176 percent (see Figure 2).158 The causal chain been loose barriers of entry and economic growth is extraordinarily complex, of course. But Europe’s disappointing economic results have led to renewed calls for Western European democracies to reduce regulation and cut bureaucracy, even as some areas of American economic life are becoming more regulated.159

Figure 1: U.S. and Eurozone GDP Per Capita: 2007 and 2022

Gap Between U.S. and Eurozone GDP Per Cap. (PPP) $45,000.00 $40,000.00 $35,000.00 $30,000.00 $25,000.00 $20,000.00 $15,000.00 $10,000.00 $5,000.00 $0.00 2007

156 157

2022

Issacharoff, supra note Error! Bookmark not defined., at 377. Calculations are based on the IMF’s dataset contained in GDP Per Capita, Current Prices,

IMF https://www.imf.org/external/datamapper/PPPPC@WEO/OEMDC/ADVEC/WEOWORLD (last accessed Oct. 4, 2023). 158 Id. 159 See, e.g., Prince Michael of Liechtenstein, GIS, Europe’s Regulatory-Technocratic Suicide (Feb. 8, 2023) https://www.gisreportsonline.com/r/free-europes-economy/ (arguing that technocracy and onerous regulation stifles economic growth in Europe). But see Rana Foroohar, America is toppling the EU from its regulatory throne, FINANCIAL TIMES (Mar. 6, 2023) https://www.ft.com/content/ed66a361-58a6-46e9-96c5-43a38f10ac15 (arguing that the U.S. has recently surpassed Europe in regulation and capital controls).

Remedy Becomes Regulation

Figure 2: U.S. and Eurozone GDP Per Capita (2007 – 2022)

GDP per capita, current prices (U.S. dollars per capita) 90000 80000 70000 60000 50000 40000 30000 20000 10000

United States

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

European Union

Our prior work focused on the role of ex post, and largely private, enforcement in the American model—regulation after the fact as deterrence— and as an effort to make potential defendants internalize some of the regulatory burden. What defines the inquiry in this article is not so much the role of after- thefact litigation in providing deterrence and compensation, but in overcoming the lack of an administrative apparatus that was in place before disaster strikes—capable of processing competing injuries and claims. Even if the problems that we highlight might, in theory, have been capable of anticipation, within the American culture of governance, they were not susceptible to ex ante institutional design. No legislators had either the foresight or the political will to address the looming need for urgent policy solutions. Even once the problem becomes evident, political will may still be too weak to address it. Sometimes, a crisis may be acute, but legislators cannot muster the consensus necessary to address it. Consider the BP oil spill, which happened in an instant but did not motivate Congress to craft its own institutional solution. On other occasions, a problem may be more chronic, but either too slow moving to prompt the kind of urgency that is born of fear—or limited to a subset of people too small or too poorly connected to rouse the political will to address it. Here, think of the opioids crisis, which is spread throughout the nation but has failed to generate the political impulse that can spur meaningful congressional solutions. The absence of a preexisting institutional mechanism triggers a search for a remedial forum; and given the institutional priors in the United State, it is no

Remedy Becomes Regulation

surprise the judiciary is the point of recourse. Tocqueville would not have been surprised. Consider his well-worn observation that “[s]carcely any political question arises in the United States that is not resolved, sooner or later, into a judicial question.”160 Famously, Judge Dan Polster of the Northern District of Ohio wrote to the Sixth Circuit in the spring of 2022: “[t]he responsibility to address the long-standing opioid epidemic should rest upon the executive and legislative branches, but they have failed to do their job. The judicial branch is not equipped to do so, but the nation’s States, cities and counties have nevertheless turned to the courts.” But, as this article has shown, courts do not act alone—in fact, much of the time, they are not even really in the driver’s seat. It is often private litigants and, especially, their lawyers who do the creative work to design the institutional solutions. Judges and courts often participate in this process of invention. When they do, however, their role tends to be more responsive than proactive. Private actors design and build the programs—usually inside the judicial process and but occasionally outside of it. For many of these modern mass harms, it is a tightly connected network of repeat player lawyers—on both plaintiff and defense sides, working with experts and academics—who design and then administer these institutions. True, judicial power is central, helping private contract into public law through the tools of the Bankruptcy Code, the MDL Statute, and Rule 23. And private litigation and private design often go hand in hand with the action of public enforcers, as lawyers from the Department of Justice or other federal agencies bring their own civil or criminal suits and state attorneys general and litigants from state agencies and local governments also join this litigation landscape. But in practice, it is the private, repeat-player lawyers driving the process—inventing the creative institutions—and figuring out how to build a form of governance after the fact that we did not have the will or the capacity to put into place before disaster came.

CONCLUSION This article adds another element to our understanding of our delegation state as it expresses itself in practice. Institutional design does not take place merely in the dialog between the legislature and the executive branch. Instead, courts may be the venue for institutional creation. By intention, or more often by default, judges and litigants become state makers. Here, we add a new layer to a perspective on governance that we have articulated in past work—a “delegation state” premised on the junction of ex post and ex ante regulation, public and private enforcement.161 This article observers something further. 160 161

CITE Littlestone-Luria, Nondelegation’s Two Faces.

Remedy Becomes Regulation

The American system is not just engaged in ex post regulation—and using private agents as delegates in this regulatory process—it is crafting its institutions after the fact. And private actors are joining, and often even leading, this ex post institution-making. As one of us explained in the pages of this law review in 2007, the American model is characterized by a penchant for “regulating after the fact.”162 By some measures, we have a big state. But the United States has often been lauded—and has sometimes been criticized—as an exemplar of the gifts of liberalized deregulation.163 That article noted, however, that it is not so much the lack of regulation that distinguishes the United States from other advanced economies. Instead, it is the choice to undertake a large proportion of the necessary regulation after the fact. Drawing on a common law tradition inherited from England, the United States privileges punishment and corrective justice over the strictures of ex ante rules. As any observer of American institutions well knows, this choice to privilege ex post governance is far from consistent.164 Nonetheless, it is a powerful trend, distinguishing the U.S. governance paradigm from comparators across the developed world. This ex post model has benefits. Market entry is generally simpler when potential entrants face fewer hurdles. Ex post regulation allows new businesses to make early investments in their own growth, rather than spending precious seed capital on regulatory compliance. An absence of barriers does not imply an absence of constraint. Enforcement actions punish deviance from the society’s standards, and the threat of retribution can force sophisticated market actors to internalize the risk of liability and save transaction costs as they regulate themselves. By keeping regulatory authority in reserve, this ex post enforcement model creates flexibility, allowing regulators to permit more, while giving them greater capacity to rein in unforeseen expressions of deviance. At a basic level, this is the beauty of regulation by adjudication— whether in courts or in agencies. In our prior work, both of us emphasized the enforcement side of this ex post governance. No doubt, enforcement is critical—and it has long been the most prominent feature of this after-the-fact state. Regulators are empowered to bring actions to punish, and this turns our systems of administrative law and tort law into a framework of de facto rules, even if they are only triggered in the breach. As detailed above, however, in recent decades, this ex post governance has often developed into more than mere enforcement. The maturation of mass harm litigation into a species of private administration furthers an insight that has roots in one of the fields classics— Abram Chayes’s The Role of the Judge in Public Law Litigation—a pathbreaking 1976 article that tried to make sense of the evolving role of the 162

Samuel Issacharoff, Regulating After the Fact, 56 DePaul L. Rev. 375, 377 (2007) RATF at 377; NTF at 408. 164 SEC and FDA examples from RATF and NTF. See NTF at 408-10. 163

Remedy Becomes Regulation

judiciary in “public law litigation.” Chayes observed that, in the “received tradition” of civil litigation, a lawsuit was a vehicle for private parties to settle disputes over private rights. But this model failed to capture critical characteristics of the emerging field of public law litigation that had become increasingly prominent during Chayes’s career. It could not explain cases such as constitutional civil rights suits, structural injunctions, and challenges to statutory policies. In public law litigation, the structure of the parties was “sprawling and amorphous”165 and the case’s outcome could reach “well beyond the particular litigants who had to appear in court.”166 Chayes’s observation was only the beginning. It could not stop merely at the role of the judge. The next step was to identify the social role played by other forms of mass litigation, even if not formally in the domain of public law. 167 Such complex litigation of private law disputes has a “manifestly public dimension.”168 Initially, the public law analogy spoke mainly to the role of “private attorneys general,” private litigants who sued alongside employees of the state as part of America’s team of enforcers. This article identifies a significant step beyond mere enforcement. Now, the judicial process play host to institutional creation. It creates administrative institutions that take charge of remediation programs for problems that would otherwise have gone unaddressed. This is the litigation state at a new level of sophistication, if you will. But it is also a revolution in institutional pragmatism. Ex post remedial administration throws our conventional vision of separation of powers into confusion, as private litigants and judges adopt a form of action that is usually the exclusive province of the legislature.169 They make the state.

165

Chayes, supra note Error! Bookmark not defined., at 1284. Issacharoff, supra note Error! Bookmark not defined. at 204. 167 See Samuel Issacharoff, Private Claims, Aggregate Rights, 2008 SUP. CT. REV. 183, 204 (2008) (“As mass harm cases have come to assume a more central role in the modern litigation landscape, some of the bright-line divisions between the role of courts in private and public law have been eroded.”). 168 Id. at . 169 Consider, here, that the administrative state engages in a form of “state making” itself, creating internal delegations, see Jennifer Nou, Subdelegating Powers, or developing relationships with external contractors who become part of their institution. See Dooling & Potter, Regulatory Body Shops. 166

Colleen Lanier-Christensen

Toxic Standards: Regulatory Epistemology and the Legacy of Endocrine Disrupting Chemicals Endocrine-disrupting chemicals (EDCs) are everywhere: in the food we eat, the air we breathe, and the water we drink, as well as in numerous products of modern life, from children’s toys and cosmetics to building materials and electronics. Nearly every person has detectable levels of EDCs in their bodies, where they can interfere with the hormonal processes that regulate essential bodily functions, including growth, metabolism, reproduction, and lactation. EDCs—including glyphosate, DDT, bisphenol A, phthalates, arsenic, triclosan, and per- and polyfluoroalkyl substances (PFAS, also known as “forever chemicals”)—have been implicated in a dizzying array of health effects, including neurological disorders, infertility, type 2 diabetes, and cardiovascular disease. Given their likely significant impact on disease and dysfunction across the life course, EDC exposures have been estimated to cost $340 billion in the U.S. and €163 in the EU annually.1 Yet, regulatory action has generally been weak and slow, despite robust evidence demonstrating the environmental and health consequences of these chemicals. Without government action, responsibility is effectively devolved from the state to individuals to navigate a sea of confusing labels and dueling toxicity claims while chemicals continue to pollute water and airways, contaminate food systems, and contribute to biodiversity loss. Why have governments failed to regulate chemicals that are widely known to be toxic? One reason is that regulatory agencies including the U.S. Food and Drug Administration (FDA) and Environmental Protection Agency (EPA) have relied on partial knowledge. Instead of

1 Christopher D Kassotis et al., “Endocrine-Disrupting Chemicals: Economic, Regulatory, and Policy Implications,” The Lancet. Diabetes & Endocrinology 8, no. 8 (2020): 719–30; Teresa M. Attina et al., “Exposure to EndocrineDisrupting Chemicals in the USA: A Population-Based Disease Burden and Cost Analysis,” The Lancet. Diabetes & Endocrinology 4, no. 12 (2016): 996–1003; L. Trasande et al., “Burden of Disease and Costs of Exposure to Endocrine Disrupting Chemicals in the European Union: An Updated Analysis,” Andrology 4, no. 4 (2016): 565–72.

Lanier-Christensen

considering the full array of knowledge about toxicity, regulators have prioritized knowledge produced in standardized studies performed according to toxicity testing standards known as Good Laboratory Practices (GLPs) and Test Guidelines. These standards were originally introduced the late 1970s to early 1980s to prevent fraud, increase bureaucratic efficiency, and facilitate trade; in the decades since regulators have used non-compliance with the standards as a reason to exclude knowledge produced by other means. The rationales of procedural fairness and rational management have produced what I call regulatory epistemology—the practices of managing knowledge, expertise, and evidence in regulatory decision-making—based on standard protocols and credentialed experts. Standardized approaches make a certain sense for business and government. For business, it can aid efficiency, which supports profitability. For government, justice requires fairness, which requires consistency. Standardized protocols are also beneficial because they provide a reason to trust the data produced by inherently biased actors: when companies are responsible for producing safety data about their products, as they generally have been in the U.S., they have a built-in interest in producing data that demonstrate safety.2 The idea here is that if regulators tell industry how to do the testing, they can be done without bias—or, at the very least, the standards make it harder for investigators to deliberately (or even subconsciously) design biased studies.3 But there is a problem: standardized tests can promote ignorance. By encouraging scientists and regulators to

See generally Samuel S. Epstein, “Corporate Crime: Why We Cannot Trust Industry-Derived Safety Studies,” International Journal of Health Services 20, no. 3 (July 1990): 443–58; Colleen Lanier-Christensen, “Toxic Rules: Chemical Regulation, International Trade, and the Epistemic Consequences of Standardized Practices” (PhD Thesis, Harvard University, 2023), Chapter 1. 3 Of course, that is far from a given. Standards prescribe action, but they also leave things out. Nor do they eliminate leeway. Critics of standard contend that industry can make judgment calls that effectively game the standards.

Lanier-Christensen

focus their attention on some forms of risk information over others, standardized tests selectively direct attention, ignoring other potentially valuable information. But do standardized approaches make scientific sense, and do they protect people or the environment—which is supposed to be the point of health and environmental regulation? I argue the answers are “no.” In the case of EDCs, reliance on standardized ways of knowing has inappropriately constrained the evidence and expertise used to govern by impeding the uptake of knowledge. Most regulatory safety standards in the twentieth were derived from testing premised on the fundamental toxicological assumption “the dose makes the poison”: every chemical would be toxic at some level and the relationship between dose and response was linear. In the early 1990s, scientists began to challenge this assumption. They found that exogenous chemicals could affect endocrine processes at very low doses, with non-linear dose-response relationships, and that response was sensitive to the timing of the exposure, not just the dose. Since the late 1990s, regulators have worked to incorporate this new type of risk into testing standards. But the results produced by those tests systemically differ from the results produced by the non-standard studies conducted in academic labs from which the science of endocrine disruption emerged. In this paper, I examine the history of GLPs and Test Guidelines, as developed and implemented by the U.S. FDA, the U.S. EPA, and the Organisation for Economic Co-operation and Development (OECD). These standards were adopted across the industrialized world beginning in the late 1970s to make chemicals governable and facilitate international trade. Over time, government reliance on them has restricted the evidence for regulatory action to a subset of studies: standardized studies, generally performed and funded by industry. Most academic

Lanier-Christensen

studies have been excluded. In many cases, this has meant ignoring evidence from precisely those studies that suggest a need for stronger regulation.

A Breach of Faith: U.S. Chemicals Testing in the 1970s Since the advent of U.S. testing laws in the early 20th century, decisions about the safety of chemical products have been based on industry-generated data, with little oversight. As science journalist William Broad asked in the New York Times in 1983, “Who tests the producttesting labs?”4 Often the answer was no one. Regulators, including the U.S. FDA and EPA, gave manufacturers the benefit of the doubt, despite their obvious profit motive and sometimes in the face of overwhelming evidence to the contrary. The FDA assumed reports were “accurate accounts of well-controlled scientific studies” and conducted few audits or inspections.5 Given the long history of unethical practices in the pharmaceutical industry—after all, testing fraud is as old as the laws mandating testing—that assumption seems naïve at best.6 By the mid 1970s, it, had become untenable. During routine reviews, FDA inspectors found both deliberate and inadvertent data “irregularities,” study designs that would obscure toxic effects, and “creative penmanship” where animals would appear and disappear during the study. Cases of the latter made one FDA official “wonder who [was] running the show, a toxicologist or a magician”?7 For example, the

William J. Broad, “Who Tests the Product-Testing Labs?,” New York Times, May 22, 1983, https://www.nytimes.com/1983/05/22/weekinreview/who-tests-the-product-testing-labs.html. 5 R Jeffrey Smith, “Creative Penmanship in Animal Testing Prompts FDA Controls,” Science 198, no. 4323 (1977): 1227. 6 Bennett Harvey Holman, “The Fundamental Antagonism: Science and Commerce in Medical Epistemology” (Ph.D. dissertation, UC Irvine, 2015); Joseph M. Gabriel and Bennett Holman, “Clinical Trials and the Origins of Pharmaceutical Fraud: Parke, Davis & Company, Virtue Epistemology, and the History of the Fundamental Antagonism,” History of Science 58, no. 4 (December 1, 2020): 533–58. 7 R Jeffrey Smith, “Creative Penmanship in Animal Testing Prompts FDA Controls,” Science 198, no. 4323 (1977): 1227.

Lanier-Christensen

companies told the FDA that “samples had been examined when … [they] had never been collected,” “Animals were recorded as normal … when in fact [they] were dead,” and “examinations of… slides were conducted by more than one pathologist… yet only the conclusions favorable to the drug were submitted.”8 In one case, involving an Illinois lab whose practices proved particularly egregious, Industrial Bio-Test (IBT, owned by Nalco Chemical), an FDA official became suspicious after seeing data that were “unbelievably clean.” The lab had made a rookie mistake of scientific fraud: their studies were too convincing.9 Upon examining the raw data, the official was perplexed by an unfamiliar term: “TBD, TBD, I kept seeing it and I wondered, what the hell is this?” He soon learned it stood for “too badly decomposed”—meaning that before IBT could derive data from animals, they had died and rotted.10 The investigations were an enormous undertaking. In addition to IBT, another major focus was the drug company G.D. Searle, best known for the first commercial oral contraceptive and the artificial sweetener aspartame, which they were trying to get through FDA approval at the time. The FDA did a two-month on-site investigation, involving more than twenty FDA staff.11 FDA salary and travel costs were estimated at $750,000 (over $4 million in 2023 dollars)

“Nonclinical Laboratory Studies: Proposed Regulations for Good Laboratory Practice Regulations (Proposed Rules),” Federal Register 41, no. 225 (November 19, 1976): 15207–8. 9 Eliot Marshall, “The Murky World of Toxicity Testing,” Science 220, no. 4602 (1983): 1131. As Jerome Halperin, then the Deputy Director, Bureau of Drugs, explained in a 2013 oral history, “People who aren’t thinking very well about how they want to fudge data, if they intend to fudge data, usually make it look too good.” But anyone with training in biology “knows that biological variation in human and any kind of animal system suggests that the world is more random than orderly … there’s always some scatter, some points that wind up in left field, that are just totally unexplainable, but they’re real.” Jerome Halperin, Oral History Interview, interview by Suzanne Junod, January 22, 2013, History of the U.S. Food and Drug Administration, https://www.fda.gov/media/129173/download, 43, 44. 10 Eliot Marshall, “The Murky World of Toxicity Testing,” Science 220, no. 4602 (1983): 1131. 11 For an account of the investigation from the company’s perspective, see William M Merino, “Good Laboratory Practices (GLPs) and the Bioresearch Monitoring Program,” in Good Laboratory Practice, ed. G. E. Paget (Baltimore: University Park Press, 1979), 67–79.

Lanier-Christensen and fifteen person-years.12 For Searle, it reportedly required 300 staff to gather records and cost several million dollars.13 It may have been the most thorough investigation of drug company testing ever conducted, but the FDA still had trouble piecing together what actually happened because the records were such a mess. The EPA found similar issues in studies supporting the safety of products it regulated— which is unsurprising because the many of the tests were done by the same labs. Industrial BioTest was estimated to have conducted 35-40% of all U.S. toxicity testing in the late 1970s, and 15% of all pesticides in the U.S. had been approved based on their data.14 Its main facility in Northbrook, IL performed over 1,500 studies annually. Before its collapse, the firm had conducted over 22,000 studies for U.S. chemical and drug companies, federal agencies, and foreign firms; more than half of IBT studies were reported to have been used in product registrations for drugs, food additives, pesticides, cosmetics, and more.15 Industry and regulators alike had seen them as not just the biggest but the best independent testing lab.16 In some cases, EPA officials were inclined to overlook scientific deficiencies in their testing simply based on the lab’s reputation.17 But IBT’s practices proved to be some of the worst. Employees even

Anthony D Cortese to Alvin Alm, “Retrospective Study of Pesticide Safety Data,” February 1976, in U.S. Senate, Appendix: The Environmental Protection Agency and the Regulation of Pesticides: Appendix to Staff Report to the Subcommittee on Administrative Practice and Procedure of the Committee on the Judiciary, December 1976 [hereafter, Appendix: The EPA and the Regulation of Pesticides], 172. 13 Exact figure illegible. Anthony D Cortese to Alvin Alm, “Retrospective Study of Pesticide Safety Data,” February 1976 (exact date illegible), in Appendix: The EPA and the Regulation of Pesticides, 173. 14 Keith Schneider, “Faking It: The Case against Industrial Bio-Test Laboratories,” The Amicus Journal 4 (1983): 14–26; “Currents,” Environmental Science & Technology 17, no. 7 (July 1, 1983): 278–80. 15 Eliot Marshall, “The Murky World of Toxicity Testing,” Science 220, no. 4602 (1983): 1130–32; Keith Schneider, “Lab Fraud -- or Government Harassment?: Researchers Accused of Faking Data at One of Nation’s Largest Chemical and Drug Testing Facilities,” Boston Globe, April 3, 1983. 16 The basis for this belief is unclear, as it is not apparent that regulators had ever bothered to check their work. 17 EPA Memo. Dr Ronald L Baron to Dr Donna Kuroda, “Phosvel (Leptophos) Delayed Neurotoxicity” July 31, 1974, in Appendix: The EPA and the Regulation of Pesticides, 547-550.

Lanier-Christensen

referred to one lab as a “swamp” because of plumbing problems and standing water; several executives were later convicted for conducting and concealing fake research.18 From 1975 to 1977, Massachusetts Senator Ted Kennedy held a series of hearings on these issues, ensuring that the scandal made headlines.19 Kennedy made the import of these findings clear: “if the … data is questioned, then the whole regulatory process is questioned. If data are ... false and misleading, ... decisions may be tragically wrong.”20 There was a two-fold threat. First, there were direct implications to health and safety. Second, the findings undermined the whole regulatory process—whether the FDA and EPA were doing their jobs, or if they even had the appropriate tools to do so. The scandal was not just an indictment of testing labs, but also of regulators. An obvious solution would have been to establish a system of third-party testing to eliminate industry’s built-in bias. Wisconsin Senator Gaylord Nelson—founder of Earth Day and a critic of the drug industry—advocated this.21 But it is not what happened. In fact, it was not even seriously considered.

Keith Schneider, “Lab Fraud -- or Government Harassment?: Researchers Accused of Faking Data at One of Nation’s Largest Chemical and Drug Testing Facilities,” Boston Globe, April 3, 1983. 19 For example: David Burnham, “Drug Maker Is Accused Of Falsifying Test Reports,” The New York Times, July 11, 1975, https://www.nytimes.com/1975/07/11/archives/drug-maker-is-accused-of-falsifying-test-reports.html; “FDA Calls for Grand Jury Investigation Of G.D. Searle’s Drug-Testing Practices,” Wall Street Journal, April 9, 1976, sec. 1; William C. Rempel and Ronald B. Taylor, “Public Being Imperiled by Fraudulent Lab Tests: Drugs and Dangerous Substances Certified for Use on Basis of Falsified or Slovenly Research,” Los Angeles Times, June 6, 1983, sec. Part I. 20 Senator Edward Kennedy in U.S. Senate, Preclinical and Clinical Testing by the Pharmaceutical Industry, 1976, Part II: Joint Hearings before the Subcommittee on Health of the Committee on Labor and Public Welfare and the Subcommittee on Administrative Practice and Procedure of the Committee on the Judiciary, 94th Cong, 2nd Sess, 1976 [hereinafter Preclinical and Clinical Testing Hearings, Part II], January 20, 1976, 1. 21 See Gaylord Nelson in Preclinical and Clinical Testing Hearings, Part II, 159. See also generally: “Comments: The Food and Drug Administration: Law, Science and Politics in the Evaluation and Control of New Drug Technology,” Northwestern University Law Review 67, no. 6 (1973): 886n110; Sidney A. Shapiro, “Divorcing Profit Motivation from New Drug Research: A Consideration of Proposals to Provide the FDA With Reliable Test Data,” Duke Law Journal 27, no. 1 (1978): 179.

Lanier-Christensen

The FDA had already settled on a solution before the scandal came to light, and their testimony to Congress was in large part about building the case for why their solution was appropriate.22 The FDA Commissioner argued that these problems “graphically” illustrated “the lack of and the need for detailed, explicit, written practices for laboratories, … which … if followed, would go a long way toward guaranteeing the integrity and value of this research. We just have to put these in place and be sure that they are followed.”23 As they saw it, it was less a scientific problem and more an administrative one, so they argued for a “process-oriented” approach. That is, if the process of data collection was adequate, then the data that accrued from that process would also be adequate.24 The FDA had an exemplar: Good Manufacturing Practice (GMP) regulations. GMPs originated in the 1930s as part of FDA’s effort to institute safety checks on drug production. They were seen by both regulators and industry as successful and quality assurance had become an established need in manufacturing.25 So the FDA sought to apply “the same concept in the ‘production’ of safety studies,” yielding Good Laboratory Practices (GLPs). In manufacturing, it

FDA officials had been planning since at least 1974 to adopt a suite of regulations for preclinical and clinical laboratories, institutional review committees, and patient follow-up, and officials began thinking of modeling laboratory practices on manufacturing regulations at that time. “FDA ‘Bio-Research Monitoring Program’ Plans Nearing Completion; Studies Backing Pending New Drug Applications Get Top Priority”; T.E. Byers, “Memorandum of Conference,” August 29, 1974, 1, in U.S. Senate, Preclinical and Clinical Testing by the Pharmaceutical Industry, 1975: Joint Hearings before the Subcommittee on Health of the Committee on Labor and Public Welfare and the Subcommittee on Administrative Practice and Procedure of the Committee on the Judiciary, 94th Cong, 1st Sess, 1975, 969. The Kennedy hearings provided an opportunity for the agency to broaden and expedite regulations they had already planned. See Daniel Carpenter, Reputation and Power: Organizational Image and Pharmaceutical Regulation at the FDA, (Princeton: Princeton University Press, 2010), 552-54. One official went so far as to say that the GLP rules would simply codify existing FDA policy, arguing that the current program would be “put into regulatory form so it’s not as loose as it is now.” “FDA Sets Nov. 1 for Start of Big New Scientific Monitoring Program; 42 Animal Labs Are First Targets; Data Sought for Good Lab Practices,” The Blue Sheet: Drug Research Reports 19, no. 41 (October 13, 1976): 9. 23 FDA Commissioner Alexander Schmidt in Preclinical and Clinical Testing Hearings, Part II, January 20, 1976, 49. 24 C. D. Van Houweling, M. A. Norcross, and P. D. Lepore, “An Overview of Good Laboratory Practices,” Clinical Toxicology 15, no. 5 (1979): 518. 25 Dale E. Cooper, “Adequate Controls for New Drugs: Good Manufacturing Practice and the 1938 Federal Food, Drug, and Cosmetic Act,” Pharmacy in History 44, no. 1 (2002): 13; John P. Swann, “The 1941 Sulfathiazole Disaster and the Birth of Good Manufacturing Practices,” Pharmacy in History 41, no. 1 (1999): 16–25.

Lanier-Christensen

was assumed that if you guaranteed the integrity of the manufacturing process, it would yield good products. The idea here was the same: guarantee the integrity of the testing process and you will get good data, which would reliably tell you if the chemical was safe. GLPs required that labs follow pre-determined, written procedures, that absolutely everything be written down—and often, signed and dated. It also required that original specimens and records be kept for a long time. If this seems reasonable, it arguably was. As a director from Oak Ridge National Labs noted, the rules seemed to state what was “obvious” to any well trained scientist or administrator.26 Still, there was a lot of pushback, both in FDA hearings with industry, and in comments they received on the rules.27 The biggest critique pertained to a general tension between bureaucracy and science: bringing rational administrative operations to labs would not produce high quality knowledge, merely high quality paperwork. Paperwork enabled labs’ work to be witnessed by regulators. Everyone, including the Oak Ridge director, commented on that. But this was not merely a complaint about red tape. Industry and academic scientists alike thought that the increased bureaucratic burden would divert scientists from the actual work of research. As the director of a UK contract lab stressed, “GLPs represent administrative procedural checks and do not go to the substantive scientific merit of the data.”28 Another quipped, “It is very easy to construct a beautiful study, of which every datum can be justified right back to the birthdate of the rat’s grandmother, and for the

LJ Serrano (Director of Laboratory Animals Resources at the Oak Ridge National Laboratory) to FDA, February 11, 1977, FDA Docket 1976-N-0476-0084. 27 I obtained these records via a Freedom of Information Act request that yielded over 600 documents, which are now publicly available. See Docket ID FDA-1976-N-0476, available at https://www.regulations.gov/docket?D=FDA-1976-N-0476. In this paper, sources are referenced with their document ID, which can be used to locate the files on regulations.gov. 28 Ralph Engel, Executive Director of the Chemical Specialties Manufacturers Association, to FDA, March 21, 1977, 2, FDA Docket 1976-N-0476-0171.

Lanier-Christensen study to mean absolutely nothing because nobody has thought about it as a totality.”29 In the hundreds of comments submitted and two days of hearings, this was reiterated time and again: that the procedural checks would not ensure scientific merit. The routinization involved was seen as posing a particular threat to toxicology, the study of the adverse effects of chemical substances on living organisms. The head of pathology at the Wellcome Research Laboratories warned: toxicology “must not degenerate into a checklist approach” or single-minded observance of routine, because “it is detection of the [non-routine] that is the very essence of good toxicology.”30 In short, Good Laboratory Practices would not ensure good laboratory practice. Perhaps surprisingly, neither the FDA nor the EPA disagreed, even as the agencies issued final GLPs between 1978 and 1983.31 In the FDA’s proposed regulations, they conceded that “compliance … does not guarantee the quality and integrity of a … study.” The stated aim of the regulations was to guarantee the quality and integrity of studies, but the FDA admitted the rules would not actually do that. So what would the rules do? The FDA offered this: that “failure to comply increases substantially the probability that the results will not be of high quality and integrity.”32 How are we to make sense of the agencies offering rules that would not do what they were supposedly intended to do. I argue that guaranteeing the quality and integrity of studies was not, actually, the point. The occasional bad test was not a terrible thing because it did not

G.E. Paget in “Transcript of Proceedings, Good Laboratory Practices For Nonclinical Laboratory Studies,” February 15, 1977, 133, FDA Docket 1976-N-0476-0008. 30 A.D. Dayan, “The Special GLP Problems of a Pathology Laboratory,” in Good Laboratory Practice, ed. G. E. Paget (Baltimore: University Park Press, 1979), 132. 31 The FDA published draft GLPs in 1976, and final rules in 1978. The EPA published draft GLPs under the Toxic Substances Control Act (TSCA) in 1979, and under the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) in 1980. The EPA published final regulations for both in 1983. The EPA rules were largely modeled on the FDA’s. 32 “Nonclinical Laboratory Studies: Proposed Regulations,” 51219.

Lanier-Christensen

threaten the regulatory system. In fact, the occasional bad test, assuming it was identified, could reinforce the norms and strengthen the regulatory system. What threatened the system was regulators’ inability to know if a test was poorly done. The efficacy of GLPs was premised on documentation and a particular form of auditable verification, as characterized in Michael Power’s concept of an “audit society” and Marilyn Strathern’s “audit culture.”33 The FDA reasoned that documentation and verification were critical, because the problems they found called into question the reproducibility of data and accountability of laboratories. Specific GLP elements were intended to facilitate inspections, such as retention of laboratory data, which would establish a data trail and offer “assurance that the data are what they say they are.”34 As expressed by the National Toxicology Program’s James Huff, the data on which conclusions were based must be “reconstructible” to make the regulators confident that the results and conclusions reflected the data.35 Ultimately this was about making elements of the work visible in place of trusting toxicologists to appropriately apply their expertise. The logic of GLPs was that non-compliance would be an indicator of possible problems, and systematic, witnessed documentation would enable officials to reconstruct studies, and if warranted hold people accountable. So, it was not that the standardization, documentation, and

Power illustrated how instituting audits did not necessarily precipitate greater effectiveness or accountability. Michael Power, The Audit Society: Rituals of Verification (Oxford University Press, 1999); Marilyn Strathern, “Introduction: New Accountabilities,” in Audit Cultures. Anthropological Studies in Accountability, Ethics and the Academy, ed. Marilyn Strathern, EASA Series in Social Anthropology (London: Routledge, 2000), 1–18. 34 Smith, “Creative Penmanship in Animal Testing,” 1228; Marshall Steinberg, “The Role of the Study Director as Seen by the Sponsor,” in Managing Conduct and Data Quality of Toxicology Studies: Sharing Perspectives, Expanding Horizons : Conference Proceedings, Raleigh, North Carolina, November 18-20, 1985, ed. B. Kristine Hoover et al. (Princeton, N.J.: Princeton Scientific Pub. Co., 1986), 209. 35 James E. Huff, “The Value of In-Life and Retrospective Data Audits,” in Managing Conduct and Data Quality of Toxicology Studies: Sharing Perspectives, Expanding Horizons : Conference Proceedings, Raleigh, North Carolina, November 18-20, 1985, ed. B. Kristine Hoover et al. (Princeton, N.J.: Princeton Scientific Pub. Co., 1986), 99.

Lanier-Christensen

audibility entailed by GLPs promised to make practice good. Rather, the standards would make lab practices governable.

Harmonizing Chemical Testing at the OECD The standardized approach to toxicity testing gained full force in 1981 when the U.S. came together with other industrialized nations to agree on common testing rules, a process referred to as “harmonization.” This happened not at an environmental or health organization, but an economic one: the Organisation for Economic Co-operation and Development (OECD). Established in 1961, the OECD grew out of the Organisation for European Economic Cooperation, which facilitated Marshall Plan implementation in post-WWII Western Europe. Its mission was premised on the idea that knitting together national economies would prevent another world war. With the addition of Canada and the U.S., the European organization became a North Atlantic one. Between 1962 and 1973, Italy, Japan, Finland, Australia, and New Zealand joined the group, creating a stable membership that persisted through the Cold War.36 The OECD quickly became a pre-eminent site of international governance, where high-level officials from member countries deliberated over wide-ranging issues. Operating under an “economic growth paradigm,” the OECD has produced regulatory tools in a variety of arenas including education, agriculture, finance, and environment.37 It has become a “restricted forum on virtually

The full membership in 1973 (in order of ratification): Canada, the U.S., the U.K., Denmark, Iceland, Norway, Turkey, Spain, Portugal, France, Ireland, Belgium, Greece, the Federal Republic of Germany, Switzerland, Sweden, Austria, the Netherlands, Luxembourg, Italy, Japan, Finland, Australia, New Zealand. In the 1990s, Mexico, Eastern Europe countries, and South Korea joined. 37 Matthias Schmelzer, The Hegemony of Growth: The OECD and the Making of the Economic Growth Paradigm (Cambridge University Press, 2016).

Lanier-Christensen unrestricted topics.”38 Tens of thousands of government officials and nongovernmental experts participate in OECD activities annually. In 1970 the OECD created an Environment Committee. It was the first official international environmental initiative, coming as it did 2 years before the UN Conference on the Human Environment in Stockholm. But the OECD’s interest in the environment was still framed within the concern with economic development. As the first Environment Committee chairman, explained in 1971: “Our responsibility is to see that stable economic growth continues to occur within the market economy and that institutional arrangements are made so that pollution control works towards a stronger and more effective market economy rather than towards a reduction of market forces.”39 Those who participated were deeply committed to environmental protection. But to be successful, the work had to be couched in economic terms. As Peter Crawford, an Australian official who led their Chemicals program from 1979 to 1982 told me: “You never would get things through the OECD unless you had a recognition of the economic implications.”40 He argued this did not undermine the environmental goals. But at the end of the day, this meant that their foremost concern was not the cost of pollution, but rather the cost of pollution control. When the OECD began its environmental work in the early 1970s, chemical manufacturers were facing heightened government and public scrutiny over their products’ health and environmental consequences. Before the 1970s, chemical regulations were fragmentary, with pre-market scrutiny largely limited to substances designed to directly affect 38

James Salzman, “Decentralized Administrative Law in the Organization for Economic Cooperation and Development,” Law and Contemporary Problems 68, no. 3–4 (2005): 192. 39 Draft Answers for the Secretary-General on Questions on Pollution (Draft by Roderick), February 2, 1971, OECD Archives, Box 36482, as quoted in Matthias Schmelzer, “The Crisis before the Crisis: The ‘Problems of Modern Society’ and the OECD, 1968–74,” European Review of History: Revue Européenne d’histoire 19, no. 6 (2012): 1009. 40 Peter Crawford, “Transcript of an Interview Conducted by Colleen T. Lanier-Christensen” (March 19, 2019), 8.

Lanier-Christensen

biological systems—namely, pesticides and pharmaceuticals. Through the 1960s, the unanticipated and global impacts of chemicals became increasingly apparent, particularly after episodes such as mercury poisoning in Minimata, Japan, and the realization of the ubiquity of compounds including polychlorinated biphenyls (PCBs).41 In response, policymakers increasingly recognized a need for pre-market screening. Among OECD nations, eight member countries adopted legislation by 1978: Switzerland (1969), Japan (1973), Sweden (1973), the U.K. (1974), Canada (1975), the U.S. (1976), France (1977), Norway (1977).42 One estimate indicated that as of 1980, nearly 65% of chemical production worldwide was or would shortly be under legislative control.43 The laws were “consistent in intent,” but they differed in their requirements, threatening “serious economic distortions” and non-tariff trade barriers.44 For example, laws varied in covering both new and existing substances; the minimum production volume requiring government notification; who was responsible for testing; and whether data submission was required before or after manufacturing. In terms of testing specifically, they varied as to which tests were required, as well as minute things about how these tests were conducted, like the number of days of a particular animal test, or the species of animal to use.

Mercury and PCBs (especially the latter) proved particularly significant for the OECD. On mercury, see B.L. Walker, Toxic Archipelago: A History of Industrial Disease in Japan (University of Washington Press, 2010). On PCBs, see Ellen Griffith Spears, Baptized in PCBs: Race, Pollution, and Justice in an All-American Town (Univ Of North Carolina Pr, 2014); Gerald Markowitz and David Rosner, “Monsanto, PCBs, and the Creation of a ‘WorldWide Ecological Problem,’” Journal of Public Health Policy 39, no. 4 (November 2018): 463–540. 42 “Environment Committee: Preparation of the Meeting of the Environment Committee at Ministerial Level: Report on Measures Taken Related to OECD Recommendations Concerning the Environment Adopted by the Council as a Result of the 1st Meeting of the Environment Committee at Ministerial Level in November, 1974,” ENV(79)7, February 19, 1979, 4, OECD Archives. The Council of the European Economic Community adopted the Sixth Amendment (so called because it was the sixth amendment to the Directive on the Classification, Packaging and Labeling of Dangerous Substances) in 1979. 43 A. W. Klein et al., “OECD Chemicals Testing Programme Physico-Chemical Test,” Chemosphere 10, no. 2 (January 1, 1981): 155. 44 “Special Programme on the Control of Chemicals: Chemicals Group: The International Control of Chemicals within the OECD Context,” ENV/CHEM/79.22, November 7, 1979, 7, OECD Archives.

Lanier-Christensen

The question was, as one outlet succinctly stated, “[Would] toxic substances laws be trade barriers?”45 The OECD’s fear was that a patchwork of national standards would require companies to test products to each country’s standards, raising trade barriers and inhibiting economic growth. The numerous new laws proposed in a short timeframe made the trade implications particularly salient and urgent.46 The OECD and similar institutions viewed the prospect of divergence among countries’ environmental and protection regulations as a threat to already-fragile economic development. In the organization’s view, since “the chemical industry forms an important part of the economies of Member countries, this is… not the time to impose undue restrictions.” They framed international harmonization as “a rational means of ensuring the development of practicable assessment schemes.”47 At its core, these OECD efforts should be seen as a means of rationalizing chemical controls in a way consonant with market economies. This concern reflects a view of the OECD economic space similar to that of the European Community and later European Union, as sociologist Andrew Barry has argued: fractured by linguistic, legal, and technical barriers inhibiting the movement of goods and capital across borders.48 Such differences fragmented markets and drew OECD scrutiny. They sought to harmonize regulatory research—an important thread in the late 20th century globalization of

“Will Toxic Substances Laws Be Trade Barriers?,” Chemical Week, March 7, 1979, 39–40. In contrast, for example, no one argued that the new drug testing requirements introduced by the 1962 KefauverHarris Amendments to the U.S. Federal Food, Drug, and Cosmetic Act created non-tariff barriers to trade. David Vogel, “The Globalization of Pharmaceutical Regulation,” Governance 11, no. 1 (January 1, 1998): 17. 47 Addendum to “1973 Annual Report and Work Proposed for 1974: Screening of Chemicals for Potential Effects to the Environment.” Environment Committee, Sector Group on Unintended Consequences of Chemicals in the Environment,” Addendum to ENV(74)27, February 4, 1975, 4, OECD Archives. 48 Andrew Barry, “The European Community and European Government: Harmonization, Mobility and Space,” Economy and Society 22, no. 3 (1993): 314–26; Andrew Barry, Political Machines: Governing a Technological Society (London ; New York: Athlone Press, 2001). 46

Lanier-Christensen

science—so that goods produced for and regulated in one market could be efficiently bought, sold, and regulated in another. They did so by creating a bureaucratic process for attesting to the acceptability of chemical data: what they called the Mutual Acceptance of Data. This was a legal agreement akin to a treaty, where all OECD regulators agreed to accept data from studies that were conducted according to GLPs and Test Guideline protocols.49 A company could be confident that if they tested according to a common template, their data would be accepted by everyone. GLPs were modeled on the regulations from the U.S.; the Test Guidelines were part of a “cookbook” of test methods the OECD began working on in 1977. Descriptions in the “how to” manual would “enable any properly equipped and staffed laboratory to carry out tests required by governments.”50 They included methods for evaluating physical properties of chemicals, environmental effects, how chemicals accumulated and degraded, and short and long term toxicity. For example, to evaluate effects on aquatic organisms, the pertinent Test Guideline would state the type of organism to use, how to administer doses, and how to measure toxicity. None of the guidelines constituted new techniques per se, but rather established methods now documented in standardized protocols, the details of which were worked out by consensus. In

The OECD’s work has primarily been executed via non-binding “Recommendations,” which ask members to consider and make good faith efforts at given objectives. But it has also served as a site to negotiate binding “Decisions” akin to treaties, which commit members to taking given actions (they are not legally the same as treaties, but entail the same types of obligations). Bill L. Long, International Environmental Issues and the OECD 1950-2000: An Historical Perspective (Paris: Organisation for Economic Co-operation and Development, 2000). A disproportionate number of OECD Decisions and Recommendations relate to the environment. Richard Woodward, The Organisation for Economic Co-Operation and Development (New York: Routledge, 2009). Across eighteen issue areas (e.g., Economics, Education, Finance and Investment, Nuclear Energy, Trade, and Taxation), the OECD has 251 legal instruments in force in 2023; 69 of them—nearly a third—pertain to the Environment. See OECD Legal Instruments, https://legalinstruments.oecd.org/en/, accessed November 4, 2023. 50 “Towards Effective International Control of Toxic Chemicals,” The OECD Observer, no. 105 (July 1980): 30.

Lanier-Christensen

some cases they had to take several protocols used in different countries and collectively decide which should be the base of the OECD Guideline.51 Participants saw a lot of value in the Test Guidelines in particular. Hundreds of government, academic, and industry scientists—and a few from environmental groups— participated. They participated because they genuinely thought it would contribute to better regulation: pooling their expertise would yield the best tools for evaluating chemicals. This was valued not just for its trade effects, but also its knowledge effects. In addition to the “practical political purpose,” a former official from the Netherlands recalled, it also had a learning purpose.52 Regulators knew they did not know everything and wanted to benefit from international expertise—including from industry—as they implemented their laws. Implementing the Decision did not mean that regulators had to make the same decisions about how to act on that data. That was left to states, according to their political preferences. Rather, if the data were compliant, regulators would be required to accept them as the basis for decision making. This is part of how they tried to thread the needle of coordinating laws while respecting national sovereignty. But it was not uncontested. A prominent U.S. environmental health scientist warned that “the new science of risk evaluation [was] incomplete and fragile.” He was afraid that the “standardization of tests … might lead to the ‘freezing’ of the science at its present level.”53 His concerns are echoed in recent critiques of this system, and central to debates about EDCs. But most others did not share this concern, instead prioritizing “the administrative need to adopt

As discussed below, at times they did additional work to “validate” methods—which ideally meant testing them in multiple labs to ensure they generated sufficiently similar results. 52 Rob Visser, “Transcript of an Interview Conducted by Colleen T. Lanier-Christensen” (June 28, 2018), 2. 53 Norton Nelson in Thomas Palme, “Laying the Groundwork for International Control of Toxic Substances,” Ambio 7, no. 3 (1978): 126.

Lanier-Christensen standardized tests.”54 As much as it was about facilitating trade, harmonizing standards was also about making testing amenable to administrative bureaucracy. OECD participants saw this as a major breakthrough. The Swedish delegate, who was a key leader in the program, argued that the Decision would provide “the tools to make [national] legislation really effective.” Countries considering new laws or updating old ones would “wrap them around” the OECD framework in order to comply.55 In practice, the agreement has made chemical regulation efficient, but not necessarily effective. The OECD envisioned its chemicals program as taking a “preventive” approach, by which they meant intervening before chemical in the market, the environment, and bodies.56 But the key rationale was that preventing environmental damage before it was done would be least costly—it was primarily desirable because of its economic rationale, not because it was desirable in itself. The work was limited to shaping what kind of data would be produced, which would increase the efficiency and reduce cost. While in principle efficiency could help to better protect public health, there is no evidence to support the idea that the OECD rules did that. The OECD has argued that by making chemical safety assessments more efficient, it has enabled governments to “better, and more rapidly, address and minimise impacts to health and the environment”—presumably with the idea that with greater efficiency, governments can evaluate more chemicals and if warranted restrict their use. But it does not follow that governments necessarily make more protective decisions. They conflated efficient regulation with effective regulation. To my knowledge the OECD has made no claims about how their work

Palme, “Laying the Groundwork,” 126. “Towards Effective International Control of Toxic Chemicals,” OECD Observer, July 1980. Any country that wishes to join the OECD needs to abide by all OECD legal instruments, about third of which are environmental, before becoming official members. 56 See “Environment Committee: Sector Group on Unintended Consequences of Chemicals in the Environment: 1973 Annual Report and Work Proposed for 1974,” ENV(74)5, February 6, 1974, OECD Archives. 55

Lanier-Christensen

has led to government action to restrict chemicals. Rather, their claims are about dollars saved by governments and industry in the process—estimated in 2019 to be over $300 million annually.57 The OECD transformed a question previously understood as the domain of science— “how do we determine if this chemical is safe?”—into a matter of economic policy. By making toxicology studies legible to and transportable among governments, OECD harmonization facilitated free-flowing trade and knowledge of toxicity for regulation. Thereby sustaining the production of chemicals and avoiding questions of responsibility. This dynamic is common in environmental regulation. Risk-based approaches that began to gain traction in the 1960s were heavily rooted in economics—in contrast to prior regulatory efforts that were rooted in biology and an assumption that a healthy environment was a human right. As historian Linda Nash put it, “the paradigm of risk imagined the world as always and unavoidably hazardous, while never asking why.”58 Who produced chemical risks? Who benefitted? Who did not? Who was harmed? Risk thinking, like the OECD, did not dwell on these questions. Quantitative risk analysis could help to maximize the payoff of regulation and optimize scarce resources. Risk thinking also pitted corporate profits against health and environmental protection, assuming that gains could only be realized if risks were accepted.59 Historian Frederick Davis has argued that the most important role of regulatory agencies like the

See OECD, Saving Costs in Chemicals Management: How the OECD Ensures Benefits to Society (OECD, 2019), https://www.oecd-ilibrary.org/environment/saving-costs-in-chemicals-management_9789264311718-en. MAD has also saved animal lives by reducing duplicative testing—according to the OECD, numbering in the “tens of thousands.” A very real benefit, to be sure, but this too has no direct implication for protective regulations. Id. 58 Linda Nash, “From Safety to Risk: The Cold War Contexts of American Environmental Policy,” Journal of Policy History 29, no. 1 (2016): 24. 59 Chemical companies played a particularly strong role in cementing this way of thinking, as they recognized the utility of risk for their bottom lines. Nash, “From Safety to Risk,” 24.

Lanier-Christensen EPA has been to set standards of permissible pollution levels.60 Similarly, at the end of the day, the OECD’s role was to coordinate allowable pollution, never questioning it. Still, the OECD’s rules have done good. They authorized regulators to trust data generated in laboratories around the world; facilitated commerce by establishing a stable regulatory environment; and saved time, money, and animal lives by reducing duplicative testing. These consequences matter. While an economic approach is not necessarily wrong, in this case it led to a focus on governance as opposed to protection, ultimately coming at the cost of knowledge, public health, and the environment.

A Tale of Two Toxicologies Regulators have prioritized knowledge claims based on studies compliant with the terms of the OECD’s Mutual Acceptance of Data. This has effectively split the field of toxicology in two. First, what I call “procedural studies”: generally done by industry, designed to satisfy regulators that a product is safe. These studies are thoroughly documented according to the GLPs and follow Test Guideline protocols. Some are published in peer-reviewed journals, but many are not published, not available to the public, and sometimes not even made available to regulators except through litigation.61 Second, “investigative studies”: usually produced in academic settings, designed to produce knowledge about toxicity, and typically published in peer-reviewed journals.62 Scientists doing this work tend to be skeptical of findings from

Frederick Rowe Davis, Banned: A History of Pesticides and the Science of Toxicology (New Haven: Yale University Press, 2014), 2. 61 Axel Mie and Christina Rudén, “Non-Disclosure of Developmental Neurotoxicity Studies Obstructs the Safety Assessment of Pesticides in the European Union,” Environmental Health 22, no. 1 (2023): 44. 62 The distinction often made is between “regulatory science” and “research science,” with differences in aims, production sites, and forms of proof. However, However, the term “regulatory science” can imply a categorical distinction between it and “normal” or “research science,” and idealize the latter in misleading ways. Sheila Jasanoff, The Fifth Branch: Science Advisers as Policymakers (Cambridge, MA: Harvard University, 1990); Alan Irwin et al., “Regulatory Science—towards a Sociological Framework,” Futures 29, no. 1 (1997): 17–31. On the

Lanier-Christensen

procedural studies, and argue that investigative ones employ more sensitive measures that are more appropriate for understanding human disease and environmental impact, and take into account the latest scientific advances and information.63 Scientists doing procedural work argue that investigative studies are not sufficiently reproducible to serve as the basis of regulatory action, which require a level of confidence that can only be guaranteed by complying with GLPs and Test Guidelines.64 Procedural and investigate studies draw upon distinct scientific practices and knowledge bases. They both produce knowledge about toxicity. But key decisions have relied primarily on one set of data: the procedural studies. Critics contend—and I agree—that investigative studies have valuable information that regulators should act on, and that by excluding it, they make less informed decisions. Take the case of bisphenol A (BPA), an endocrine-disrupting chemical widely used in epoxy resins and polycarbonate plastics. Scientists have found BPA in over 90% of Americans tested. Numerous studies show that BPA can health problems including infertility, altered

history of the term “regulatory science” and its various invocations, see A. Alan Moghissi et al., “Innovation in Regulatory Science: Evolution of a New Scientific Discipline,” Technology & Innovation 16, no. 2 (2014): 155–65; David Demortain, “Experts in the Regulation of Technology and Risk: An Ecological Perspective on Regulatory Science,” in The Oxford Handbook of Expertise and Democratic Politics, ed. Gil Eyal and Thomas Medvetz, 1st ed. (Oxford University Press, 2023), 282–313. At the most basic level, regulatory science has been invoked to refer to science either produced for or used in regulation. With chemical regulation—precisely because of the standards discussed in this paper—the two are largely synonymous. While this could make regulatory science a particularly apt term, the case of EDCs invites us to question the distinction: why can’t research science be used in regulation— why can’t research science be regulatory science? The regulatory versus research distinction is also based on where and why the science is done. But my emphasis with the terms procedural versus investigative science is how the science is done. In this sense, my terminology is most similar to that of Sarah Vogel, who used the terms “hypothesis-driven” research versus regulatory testing. Sarah A Vogel, Is It Safe?: BPA and the Struggle to Define the Safety of Chemicals (Univ of California Press, 2012). Scientists and regulators often refer to my procedural studies and Vogel’s regulatory testing as “guideline studies.” 63 John Peterson Myers et al., “Why Public Health Agencies Cannot Depend upon ‘Good Laboratory Practices’ as a Criterion for Selecting Data: The Case of Bisphenol A,” Environmental Health Perspectives 117, no. 3 (March 2009): 309–15. 64 Richard A. Becker et al., “Good Laboratory Practices and Safety Assessments,” Environmental Health Perspectives, 2009, A482–83; Christopher J. Borgert et al., “Does GLP Enhance the Quality of Toxicological Evidence for Regulatory Decisions?,” Toxicological Sciences 151, no. 2 (June 1, 2016): 206–13.

Lanier-Christensen neurological development, and cardiovascular disease.65 Several expert panels have concluded that low-dose BPA exposure during critical developmental periods alters both organ development and behaviors in children.66 Yet regulators, including the U.S. FDA, still say that BPA is safe. Why? A 2015 Newsweek headline offered an answer: the FDA concluded BPA was safe only because they “ignore[d] most studies about it.”67 Thousands of animal experiments and over a hundred human epidemiological studies have examined BPA, but the FDA has based its reviews on a subset of these studies: the procedural studies, which are generally conducted or funded by industry.68 BPA is not an isolated example. Studies of brominated flame retardants, the herbicides atrazine and glyphosate, and neonicotinoids (the insecticides linked to honeybee colony collapse disorder), using diverse and evolving methods have yielded variable results, but the evidence base as a whole suggests that current exposures have adverse health and environmental effects. In contrast, procedural studies using standardized methods consistently find no adverse effects at current exposure levels.69

See Frederick vom Saal and Laura Vandenberg’s summary of the evidence from human epidemiological and animal laboratory studies. Frederick S vom Saal and Laura N Vandenberg, “Update on the Health Effects of Bisphenol A: Overwhelming Evidence of Harm,” Endocrinology 162, no. 3 (September 23, 2020): bqaa171. A 2018 study detected BPA, BPS, and BPF in 96%, 89%, and 67% of urine samples from Americans tested. Hans-Joachim Lehmler et al., “Exposure to Bisphenol A, Bisphenol F, and Bisphenol S in US Adults and Children: The National Health and Nutrition Examination Survey 2013–2014,” ACS Omega 3, no. 6 (2018): 6523–32. 66 Organs affected include the brain, prostate, and mammary glands. Laura N. Vandenberg, Derek Luthi, and D. ’Andre Quinerly, “Plastic Bodies in a Plastic World: Multi-Disciplinary Approaches to Study Endocrine Disrupting Chemicals,” Journal of Cleaner Production 140 (2017): 373–85. 67 Douglas Main, “BPA Is Fine, If You Ignore Most Studies About It,” Newsweek, 2015, https://www.newsweek.com/2015/03/13/bpa-fine-if-you-ignore-most-studies-about-it-311203.html. 68 The FDA’s review of the literature—the latest of which was issued in 2014—included approximately 300 studies. See Jason Aungst and Steven Anderson, “Final Report for the Review of Literature and Data on BPA” (June 6, 2014), https://www.fda.gov/media/90546/download. Additional documents related to the FDA’s reviews are available at https://www.fda.gov/food/food-additives-petitions/bisphenol-bpa. See also K. Barry Delclos, “Bisphenol A: Toxicology and Pharmacokenetic Data to Inform Ongoing Safety Assessments,” https://www.fda.gov/scienceresearch/about-science-research-fda/bisphenol-toxicology-and-pharmacokenetic-data-inform-ongoing-safetyassessments. 69 On the brominated flame retardant decaBDE, Ruth E. Alcock, Brian H. MacGillivray, and Jerry S. Busby, “Understanding the Mismatch between the Demands of Risk Assessment and Practice of Scientists — The Case of Deca-BDE,” Environment International 37, no. 1 (2011): 216–25. On glyphosate, Christopher J Portier et al.,

Lanier-Christensen

Regulators have discounted investigative studies for two main reasons. Regulators and industry spokespersons say that investigative studies do not sufficiently report methods and data. In contrast, procedural studies comply with the extensive reporting requirements of GLPs, including access to raw data. Peer-reviewed journals’ reporting standards are usually less structured and formalized, and access to raw data is uncommon. But regulators have rejected investigative studies prima facie because of GLP-non-compliance. When the FDA excluded investigative studies from their review of BPA, it was not due to specific flaws or deficiencies in the studies’ methods, documentation, or reporting.70 Rather, they excluded these studies solely on the grounds that they did not comply with GLP standards. Some academic scientists have been working on ways to make their work more legible to regulators.71 But such measures are unlikely to be sufficient because of the other reason why regulators have preferred procedural studies. The second reason has to do with Test Guideline protocols. The argument goes that Test Guideline studies are more appropriate for regulatory use because they have been “validated.” Generally this means they use standardized protocols that scientists have previously tested and judged to produce uniform results, even in different laboratories. Ideally this has been done

“Differences in the Carcinogenic Evaluation of Glyphosate between the International Agency for Research on Cancer (IARC) and the European Food Safety Authority (EFSA),” Journal of Epidemiology and Community Health 70, no. 8 (2016): 741–45. On atrazine, M. D. Boone et al., “Pesticide Regulation amid the Influence of Industry,” BioScience 64, no. 10 (2014): 917–22. On neonicotinoids, Sainath Suryanarayanan and Daniel Lee Kleinman, Vanishing Bees: Science, Politics, and Honeybee Health (New Brunswick, NJ: Rutgers University Press, 2017). Alternative knowledge on colony collapse disorder has since been incorporated in European regulation and led to a ban. David Demortain, “The Science behind the Ban: The Outstanding Impact of Ecotoxicological Research on the Regulation of Neonicotinoids,” Current Opinion in Insect Science 46 (2021): 78–82. 70 E.g., it was not because the FDA decided that a study was inadequately reported, or they objected to how animals were dosed. 71 Marlene Ågerstrand et al., “An Academic Researcher’s Guide to Increased Impact on Regulatory Assessment of Chemicals,” Environmental Science: Processes & Impacts 19, no. 5 (2017): 644–55. See also Thomas Backhaus and Xenia Trier, “Empowering Academic Research in Chemical Risk Assessment and Management,” Integrated Environmental Assessment and Management 11, no. 2 (April 2015): 183–84. For an industry perspective, see Lynn S McCarty, Christopher J Borgert, and Ellen M Mihaich, “Information Quality in Regulatory Decision Making: Peer Review versus Good Laboratory Practice,” Environmental Health Perspectives 120, no. 7 (2012): 927–34.

Lanier-Christensen

through interlaboratory methods testing, or what the OECD has called “ring tests,” “round robin” or “interlaboratory calibration tests.” Since the 1980s, such studies have involved labs in several countries running the “same” study on one chemical in accordance with the draft guideline—that is, with the same materials and methods—to see if they got sufficiently similar results.72 As a result, regulators including the U.S. FDA have judged these studies to yield “reproducible results that [the FDA] can have confidence in using.”73 But the history of the OECD Test Guideline program suggests that validation may not be as meaningful as regulators and industry suggest. Few, if any, of the OECD’s original fifty-one Test Guidelines had been formally validated in interlaboratory tests at the time of adoption. Instead, they were “validat[ed] through tradition and experience.”74 That is, as the OECD also later put it, they were adopted because they were “classical ones for which experience had widely been gained.”75 Prior experience with the methods provided authorization for their use absent formal interlaboratory testing.76 Other Test Guidelines were adopted despite known

The “ring” terminology was not explained, but presumably the idea behind the term was that lab A’s results were compared with lab B’s, which were compared with C’s, and C’s with D’s, and then back to A’s. Since the late 19th century, many answers to questions about the contents of agricultural and food products and, later, environmental pollution, have been accepted as facts in the regulatory arena because they were generated by standardized methods that scientists had tested in multiple laboratories and judged to produce uniform results. By the 1960s, scientists and regulatory officials used the term “validated” to describe this status. It was a way to legitimize the methods and provide a warrant for regulators and industry to use them. In the 1970s these “validation” practices were applied to toxicity testing methods. But validation was a slippery concept. It was rarely explicitly defined and invoked in vague and diverse ways. This was a feature more than a bug: it was a social means of agreeing on when methods were good enough for regulatory use. By this, scientists generally meant that it produced sufficiently reliable and informative results to be used as a basis for public decision making, which was ultimately a matter of expert judgment. See Colleen Lanier-Christensen, “The Association of Official Agricultural Chemists and the Pursuit of Reliable Chemical Methods,” in Testing Knowledge: Validation and Regulation in the Human and Health Sciences, ed. Lara Keuck and Angela N. H. Creager, forthcoming. 73 FDA toxicologist Jason Aungst, quoted in Cornwall, “Rules of Evidence,” 566. 74 “Chemicals Group and Management Committee: Updating Porgramme: Issues of Policy and Procedure,” ENV/CHEM/CM/85.5, April 1, 1985, 7, OECD Archives. 75 “Chemicals Group and Management Committee, Extended Steering Group: Preparation of the Third High Level Meeting: Item 7: Chemicals Safety Testing and the Use of Animals,” ENV/CHEM/STG/86.4, October 10, 1986, 5, OECD Archives. 76 Other times, the OECD reasoned that they could adopt unvalidated protocols because the methods could be updated later.

Lanier-Christensen

problems. This was generally for pragmatic reasons: flawed though they might be, the methods fulfilled a regulatory need.77 Formal interlaboratory validation became increasingly important to the OECD Test Guideline program since the 1990s, and, to the OECD’s credit, it has been continuously developing new and updating existing Test Guidelines. But Test Guidelines are time-consuming to develop, standardize, and validate, and they can be just as hard to later change. Updating methods can be a cumbersome bureaucratic process, in part because of lengthy comment and revision periods as well as the fact that the OECD operates on a consensus basis, meaning that all members have to agree on any changes. This means Test Guidelines can easily become outdated, if they were not already outdated at the time of their adoption.78 Some methods have made it through the process expediently: the OECD has cited methods that have been validated in as little as one to four years.79 But in other cases, it has taken more than a decade.80 Consequently, despite efforts to update methods, the original Test Guidelines adopted in the 1980s were, as a toxicologist with the European Commission has characterized it, “frozen in time.”81 Given the time required to validate methods, following OECD Test Guidelines necessarily means conducting studies according to old standards, not state-of-the-art scientific methods.

For example, see the history of OECD Test Guideline 202, Daphnia sp., Acute Immobilization and Reproduction Test, discussed in Lanier-Christensen, Toxic Rules, Chapter 4. 78 On critiques that Test Guideline studies are outdated, see Brendan Borrell, “Toxicology: The Big Test for Bisphenol A,” Nature News 464, no. 7292 (2010): 1122–24; Warren Cornwall, “Rules of Evidence,” Science 355, no. 6325 (2017): 564–67. 79 Though the OECD ceded that the methods for eye damage which were validated in one year only got through so quickly because they were far along in the validation process when they were officially proposed. Sigman and Diderich, “OECD Defends GLP against Claims It Impedes Advancing Science.” 80 On the U.S. EPA’s and OECD’s efforts to validate guidelines for endocrine disruption, see below. Another example: It was reported in 2008 that the EU was spending €25 million per year in an effort to validate cell-based in vitro tests. The validation process by the European Centre for Validation of Alternative Methods was lengthy (taking two to eight years) and expensive (costing some $300,000 per method). And that was all before the method could be sent to the OECD for further evaluation. Trisha Gura, “Toxicity Testing Moves from the Legislature to the Petri Dish—and Back,” Cell 134, no. 4 (August 22, 2008): 557–59. 81 Thomas Hartung quoted in Gura, 558. He pointed to the Draize test from 1944 as an example—an outdated test performed with few modifications (e.g., it now uses fewer animals) that should have been replaced by now.

Lanier-Christensen

The program has been critiqued since the 1980s. Ellen Silbergeld, an environmental health scientist who worked for the Environmental Defense Fund (EDF), was a frequent participant in OECD chemicals proceedings, supported in part by a grant from the EPA to coordinate public interest input into OECD work.82 In 1989, she wrote that the “philosophy of validation” had given rise to a gap between methods in common use and formally validated methods, yielding a “validation gap” that would only increase over time. Silbergeld warned that the insistence on validation inhibited the work of testing of chemicals.83 Despite the OECD’s efforts to revisit methods through its Test Guideline updating program, over the next three decades concerns like Silbergeld’s increased. Since the 2010s, the OECD has been critiqued for not only being too slow to adopt new methods that reflect current scientific and regulatory concern, but also “imped[ing] advancing science” in toxicology.84 Even when Test Guidelines have been formally validated through interlaboratory tests, their assumed reproducibility may be questionable. Typically, their reproducibility has been largely presumed, not demonstrated. In a recent rare case when a Test Guideline study was repeated, the results were conflicting, raising the question of whether these studies are in fact reproducible.85 Nonetheless, many regulators have continued to use a method’s status as “validated” as reason enough to treat it reliable, and used this as a reason to privilege procedural

Silbergeld worked for EDF starting in 1982; she later became a professor at Johns Hopkins University. In 1993 she earned a MacArthur Foundation “genius award.” Silbergeld’s papers can be found at the Schlesinger Library, Radcliffe Institute, Harvard University. On public interest and industry participation in OECD chemical work, see Colleen Lanier-Christensen, “Creating Regulatory Harmony: The Participatory Politics of OECD Chemical Testing Standards in the Making,” Science, Technology, & Human Values 46, no. 5 (September 2021): 925–52. 83 Ellen Silbergeld, memorandum to participants of Conservation Foundation meeting on testing re “In vitro testing ideas,” March 21, 1989, 3, EKS Papers, Carton 5, Folder 144. 84 Richard Sigman and Bob Diderich, “OECD Defends GLP against Claims It Impedes Advancing Science,” Chemical Watch, May 2013, https://chemicalwatch.com/14980/oecd-defends-glp-against-claims-it-impedesadvancing-science. 85 Laura N. Vandenberg, “Chapter Two - Toxicity Testing and Endocrine Disrupting Chemicals,” in Advances in Pharmacology, ed. Laura N. Vandenberg and Judith L. Turgeon, vol. 92, Endocrine-Disrupting Chemicals (Academic Press, 2021), 35–71.

Lanier-Christensen

studies over investigative ones. They have done so even when large bodies of scientific evidence have come to contrary conclusions, suggesting that procedural studies are either wrong or at the very least produce a selective—and hence, potentially misleading—picture of toxicity. Behind both reasons for preferring procedural studies—the documentation and reporting of GLPs and the validation practices of Test Guidelines—lies a fundamental factor: the law. There is no strict legal obstacle to including investigative studies in regulatory review. That is, there is no codified legal requirement that agencies only rely on procedural studies. But regulators have argued that in practice investigative studies will not hold up to legal challenge. One former EPA official argued that the Agency generally could not use academic studies because it was “pretty hard for EPA to demonstrate the quality of the study in the absence of GLPs.”86 For example, in the 1980s the EPA wanted to use a type of semi-field tests called pond mesocosm studies as part of pesticide evaluations. Officials believed these studies were higher quality than standardized single-species tests. But the test protocols had not been standardized; they were based on only general guidelines and required a significant amount of negotiation to perform and interpret. Therefore the EPA felt it could not use the studies. In contrast, UK authorities did rely on them. As Willem Halffman has argued, this was largely due to national differences in confidentiality of proceedings and whether they were subject to legal contestation.87

Charles M. Auer, Transcript of an Interview Conducted by Colleen T. Lanier-Christensen, August 12, 2021 and February 28, 2022, 14. To my knowledge this has not been legally tested. 87 Willem Halffman, “Standardization as a Trust Device,” in The Politics of Chemical Risk : Scenarios for a Regulatory Future, ed. Roland Bal and Willem Halffman (Dordrecht ; Boston: Kluwer Academic Publishers, 1998), 276.

Lanier-Christensen

FDA officials have stressed the legal considerations involved. In the words of one FDA staffer: FDA operates behind the cutting edge. It has to because it is making decisions that have to be based on an administrative record and has to be supported by that record in a court of law possibly. And so the decisions that they make cannot be the same that would happen in a university lab where you are exploring the cutting edge of science and there is very little agreement about whether these facts are really facts or not. FDA has to operate behind that cutting edge because they need to operate in an area of science that is a little bit more cut and dry so that the force and effect of law can come into play on their decisions.88 EDCs including BPA have been a flash point in these debates over procedural and investigative studies because the science of endocrine disruption departs from the assumptions of traditional toxicology. Most regulatory safety standards in the twentieth were derived from testing that exposed animals to high doses of a chemical to determine the lowest level at which a toxic effect occurred. This was premised on the fundamental toxicological assumption “the dose makes the poison”: every chemical would be toxic at some level—and, as a corollary, every chemical would have a threshold below which it was safe (with the exception of carcinogenic effects)—the work of toxicology was to determine what that level was. Regulatory procedures also assumed that the relationship between dose and response was linear, such that testing a small range of doses would be sufficient to extrapolate the effects at lower and higher amounts. These assumptions undergirded threshold-based regulation, which has allowed certain levels of toxicants to be legally allowed in food, humans, and the environment, amounting to a permission-to-pollute regulatory regime.89

“Appendix 3 – FDA Foods Program Review of Chemical Safety Capacity and Management: Results of External Interviews,” 10, emphasis added. 89 Endocrine Disruptors Action Group, “Pollution Is Colonialism,” September 18, 2017, https://endocrinedisruptorsaction.org/2017/09/18/pollution-is-colonialism/; Reena Shadaan and Michelle Murphy, “EDC’s as Industrial Chemicals and Settler Colonial Structures:,” Catalyst: Feminism, Theory, Technoscience 6, no. 1 (2020); Max Liboiron, Pollution Is Colonialism (Duke University Press, 2021).

Lanier-Christensen

In the early 1990s, scientists began to challenge these assumptions. They found that exogenous chemicals could affect endocrine processes at low doses, with non-monotonic doseresponses (U-shaped curves where there are significant effects at very low doses and high doses, with minimal effects in between). The timing of doses also mattered. Just like endogenous hormones, exogenous EDCs produce different effects at different periods of the life cycle. Exposure to EDCs during critical windows of development, such as fetal and early life stages, can be particularly consequential, altering growth and development or resulting in disease later in life. Because this research challenged fundamental tenets of toxicology, it undermined the longstanding logic of chemical regulation.90 Regulators have worked to incorporate this new type of risk into testing standards. In 1996, the U.S. Congress passed two laws to establish programs to screen drinking water contaminants and pesticides for potential endocrine disrupting effects: The Food Quality Protection Act and the 1996 Amendments to the Safe Drinking Water Act.91 The laws empowered the EPA to screen chemicals for hormonal effects. Because the legislators expected the program to be controversial, they wanted the EPA to be “as responsible and credible as

Sarah A. Vogel, “From ‘The Dose Makes the Poison’ to ‘The Timing Makes the Poison’: Conceptualizing Risk in the Synthetic Age,” Environmental History 13, no. 4 (2008): 667–73. Other scientific fields also do not hew neatly to the procedural paradigm. Examples include toxicoepigenetics, which examines the health effects of environmental exposures due to epigenetic changes, and popular epidemiology, involving testing and measurement programs by the people affected. Anne Le Goff et al., “Toxicoepigenetics for Risk Assessment: Bridging the Gap Between Basic and Regulatory Science,” Epigenetics Insights 15 (2022): 251686572211131; Michelle Murphy, “Chemical Regimes of Living,” Environmental History 13, no. 4 (2008): 695–703. 91 Public Laws 104-170 and 104-182, respectively. The U.S. Congress had begun investigating endocrine disruption in 1993. U.S. House, Health Effects of Estrogenic Pesticides Hearing Before the Subcommittee on Health and the Environment of the Committee on Energy and Commerce, 103rd Cong, 1st Sess, October 21, 1993. The 1996 publication of Theo Colborn, Dianne Dumanoski, and John Peterson Myers’s seminal book Our Stolen Future— which chronicled the development of a growing body of scientific evidence linking synthetic chemicals with disruption of the endocrine system—drew widespread attention and galvanized governmental action. Theo Colborn, Dianne Dumanoski, and John Peterson Myers, Our Stolen Future: Are We Threatening Our Fertility, Intelligence, and Survival?: A Scientific Detective Story: With a New Epilogue by the Authors (New York: Penguin Group, 1997).

Lanier-Christensen possible.”92 To that end, they required the EPA use “validated test systems.” Legislators believed that this would shore up the Agency’s credibility and ultimately lead to regulation of EDCs.93 But there were no validated test systems for EDCs, and the work needed to develop them would take more than two decades.94 Even then, procedural and investigative studies of EDCs systematically differ in their findings.95

The Politics of Standardized Protocols Standardized protocols are a common response to concerns about bias or fraud. In principle, standardized protocols make it harder for scientists to design studies with predetermined outcomes. For this reason, some public health advocates have supported them.96

Daniel Patrick Moynihan in Congressional Record—Senate, May 18, 1994, 10692 (speaking of the earlier 1994 iteration of the bill). 93 This reflected scientists’ and policymakers’ common belief that improvements in science, including shoring up scientific credibility, would limit social disagreements. Willem Halffman, “Standardization as a Trust Device,” in The Politics of Chemical Risk : Scenarios for a Regulatory Future, ed. Roland Bal and Willem Halffman (Dordrecht ; Boston: Kluwer Academic Publishers, 1998), 265–91. 94 The EPA first created the Endocrine Disruptor Screening and Testing Advisory Committee in 1996. After public comment and additional advisory group review, the EPA reported that it began implementing the Endocrine Disruptor Screening Program (EDSP) in August 1999. U.S. EPA, “Environmental Protection Agency Endocrine Disruptor Screening Program: Report to Congress,” August 2000, 3. However, a 2002 Congressional Research Service report emphasized that while EDSP had been “‘established’ in 1998,” by 2002 it was “not yet operational.” Linda-Jo Schierow and Eugene H. Buck, “Environmental Exposure to Endocrine Disruptor: What Are the Human Health Risks?” (Congressional Research Service, February 4, 2002), 3. The first batch of test protocols for the program (for Tier 1 assays) was issued in 2009, but by 2016, the second batch (Tier 2 assays) had not been validated or approved. Environmental Protection Agency [EPA–HQ–OPP–2009–0634; FRL–8434–8], Endocrine Disruptor Screening Program; Tier 1 Screening Order Issuing Announcement, https://www.regulations.gov/document/EPAHQ-OPP-2009-0634-0001, accessed July 31, 2022. See also Kevin C. Elliott, “Standardized Study Designs, Value Judgments, and Financial Conflicts of Interest in Research,” Perspectives on Science 24, no. 5 (2016): 529–51. Assay validation status (last updated in 2015) available at https://web.archive.org/web/20150906013944/http://www.epa.gov/scipoly/oscpendo/pubs/assayvalidation/status.ht m, accessed July 31, 2022. See also the oral history of Dr. Penelope Fenner-Crisp, then the Deputy Director of the EPA’s Office of Pesticide Programs, who argued that the validation requirement was inappropriate. Penelope A. Fenner-Crisp, interview by Sarah L. Hunter-Lascoskie and Jody A. Roberts, April 3 and 4, 2012, 74-76, Philadelphia: Chemical Heritage Foundation, Oral History Transcript #0712 (Courtesy of Science History Institute). 95 In addition, the available Test Guidelines pertain to a small selection of endocrine pathways (e.g., estrogenic, androgenic, and thyroid), leaving other pathways unexplored and untested. Elise Grignard, Kelly de Jesus, and Philippe Hubert, “Regulatory Testing for Endocrine Disruptors; Need for Validated Methods and Integrated Approaches,” Frontiers in Toxicology 3 (2022): 821736. 96 For example, in the 1970s Samuel Epstein—a physician and professor of environmental health lauded by environmentalists and reviled by chemical companies—advocated for them. Samuel S. Epstein, “Public Health

Lanier-Christensen

There is also a strong argument for bureaucratic efficiency, particularly in a world where regulatory agencies like the EPA are chronically underfunded and understaffed.97 As long as regulators rely on companies for toxicity data, it is hard to imagine not prescribing detailed standards by which companies would conduct testing. But standardized protocols have distinct and predictable negative effects. Philosopher Kevin Elliott has argued that standardized protocols have four key weaknesses: incompleteness, limited applicability, ossification, and selective ignorance. Standardized protocols are incomplete because they are usually insufficient to ensure that the studies have the desired qualities. Elliott specifically points to GLPs: because they emerged in response to industry abuses, including falsified data, the procedures focus on documentation. But this does not ensure that appropriate questions are being asked, nor that studies are appropriately designed to answer those questions. Coupling GLPs with Test Guidelines, as the OECD did, may improve matters by ensuring companies employ vetted protocols, but OECD Test Guidelines usually still contain some flexibility, for example on the animal strain to use. If an investigator wanted to obtain evidence that a chemical was not harmful, they could easily choose a less sensitive strain.98 Standardized protocols can also suffer from limited applicability. The OECD recommends that “universal indicator species” (e.g., algae, daphnia, and a fish) be used to predict the effects of chemical pesticides on non-human species. But no species is actually universal.

Hazards from Chemicals in Consumer Products,” in Consumer Health and Product Hazards--Chemicals, Electronic Products, Radiation, ed. Samuel S. Epstein and Richard D. Grundy (MIT Press, 1974), 45–100. 97 On the EPA’s budget, see Jill Lindsey Harrison, Pesticide Drift and the Pursuit of Environmental Justice (MIT Press, 2011), 101–3. 98 This is a cynical take on what happened in the Consortium Linking Academic and Regulatory Insights on BPA Toxicity (CLARITY-BPA) study, a $30 million federal research program to bridge the impasse between procedural and investigative studies on BPA. The FDA insisted on using the Sprague-Dawley rat, ostensibly because it was the FDA’s “go to” animal for toxicity testing. But this rat strain was known to be insensitive to estrogenic effects; using it would increase the likelihood that the CLARITY-BPA results would back up the FDA’s existing policy stance that BPA was safe. Lynne Peeples, “Exposed: On the Edge of Research Honesty,” EHN, November 15, 2019, https://www.ehn.org/bpa-science-fda-safe-2641150483.html.

Lanier-Christensen

More often the selected species does not accurately effects on non-target species, population or ecosystem responses, or effects on other populations and ecosystems.99 The third problem Elliott identifies is ossification. The time and resources required to develop standardized protocols can be so significant as to prevent the introduction of new methods or to disincentivize methodological innovation. Sometimes, the resulting methods are no longer relevant by the time the process has finished because there are newer and better ones available.100 When GLPs were first proposed in the U.S., industry toxicologists were concerned that standards would ossify and impede the development of knowledge. Similar concerns were expressed when the OECD was designing its approach to chemical testing, but the needs of administrative bureaucracy won out. Fourth, standardized protocols can promote selective ignorance by encouraging investigators to focus their attention on some forms of risk information over others.101 By selectively directing attention, they may ignore other potentially valuable information for decision making. For example, most toxicity tests are done on individual chemicals with a small number of well-characterized model organisms, even though in real life people are exposed to mixtures of substances and the model organisms chosen likely do not provide complete information about effects on other species.102

David A. Andow and Angelika Hilbeck, “Science-Based Risk Assessment for Nontarget Effects of Transgenic Crops,” BioScience 54, no. 7 (2004): 637–49. 100 Newer methods will not always be better methods, but scientists are constantly refining testing practices to use fewer animals, produce evidence more relevant to humans and environments, and measure different endpoints of concern. Newer methods will be better matched to the prevailing values and concerns of the day (e.g., animal welfare and endocrine disruption). 101 Kevin C. Elliott, “Selective Ignorance and Agricultural Research,” Science, Technology, & Human Values 38, no. 3 (May 2013): 328–50. See also Robert Neel Proctor, “Agnotology: A Missing Term to Describe the Cultural Production of Ignorance (and Its Study),” in Agnotology: The Making and Unmaking of Ignorance, ed. Robert Neel Proctor and Londa L. Schiebinger (Stanford, Calif: Stanford Univ. Press, 2008), 1–36. 102 Rik I L Eggen, Renata Behra, and Nina Schweigert, “Challenges in Ecotoxicology,” Environmental Science & Technology 38, no. 3 (2004): 58A-64A; Andow and Hilbeck, “Science-Based Risk Assessment for Nontarget Effects of Transgenic Crops.”

Lanier-Christensen

A historical example illustrates this concern: in 1982, an aquatic toxicologist who worked for a private Massachusetts laboratory lamented the science’s lack of advancement. There had been “quantitative growth” of the science, in terms of the amount of aquatic toxicity data produced. But there had been “no real ‘qualitative growth’”—by which he meant, “No new and better questions are being asked.” There had been “an incredible increase in data but virtually no increase in knowledge.”103 He blamed regulation. Regulators, he said, “routinely … ask for the same kinds of answers (data) and can’t assimilate other types of data into the regulatory process.”104 The consequences of adopting toxicity testing standards were not only predictable, they were also predicted, and industry scientists were among the most vocal in warning of the consequences. Yet, since the 1970s, industry came to whole-heartedly embrace the standardized approach. In one light this makes sense: GLPs were imposed on them despite their protestations. Once they were coupled with Test Guidelines at the OECD, GLPs had the appeal of facilitating international trade of their products. So they complied: first reluctantly and then enthusiastically once the OECD had set up the Mutual Acceptance of Data framework. GLPs and Test Guidelines can be seen as part of the lineage of testing rules that have prompted companies to adapt research and development strategies to comply with new laws.105 In this light, industry would be invested in the standards’ continuation. However, one can hardly imagine the aboutface was entirely in good faith. Throughout the development of GLPs, industry representatives argued extensively that the standards would not ensure quality and continued to do so as companies navigated the new

103

Kj Macek, “Aquatic Toxicology: Anarchy or Democracy?,” in Aquatic Toxicology and Hazard Assessment, ed. Jg Pearson, Rb Foster, and We Bishop (West Conshohocken, PA: ASTM International, 1982), 4, emphasis added. 104 Macek, 5. 105 Carpenter, Reputation and Power.

Lanier-Christensen GLP requirements in the 1980s.106 Twenty years later, industry scientists changed their tune. For example, in 2003 a Pfizer scientist argued that GLPs “ensure the quality and integrity of nonclinical safety data” and “served to assure that high quality, scientifically sound studies are performed to demonstrate the safety of products.”107 In 2009, thirty-six academics and public health advocates penned a commentary in leading environmental health journal Environmental Health Perspectives in which they made the case “Why Public Health Agencies Cannot Depend on Good Laboratory Practices as a Criterion for Selecting Data,” illustrated by the case of BPA.108 Representatives of the American Chemistry Council, Crop Life America, the American Petroleum Institute, the Soap and Detergent Association, and the Grocery Manufacturers Association replied later that year. In their rejoinder, they argued that GLPs are “fundamental scientific principles and practices” and “indispensable for providing scientific confidence in studies conducted for chemical safety determinations.”109 If GLPs do not ensure quality studies—as everyone acknowledged upon their development—what do they do? Whose interests have GLPs and Test Guidelines served? Consider the case of “sound science,” something that sounds like something that everyone— industry and academic scientists, regulators, and everyday people—should favor. But history tells a different story. In the early 1990s the “sound science” movement emerged from collaborations between conservative business associations and major industries to undermine and

106

See, e.g., Marshall Steinberg, “The Role of the Study Director as Seen by the Sponsor,” in Managing Conduct and Data Quality of Toxicology Studies: Sharing Perspectives, Expanding Horizons : Conference Proceedings, Raleigh, North Carolina, November 18-20, 1985, ed. B. Kristine Hoover et al. (Princeton, N.J.: Princeton Scientific Pub. Co., 1986), 208. 107 Anne M. Baldeshwiler, “History of FDA Good Laboratory Practices,” The Quality Assurance Journal 7, no. 3 (2003): 157. 108 John Peterson Myers et al., “Why Public Health Agencies Cannot Depend upon ‘Good Laboratory Practices’ as a Criterion for Selecting Data: The Case of Bisphenol A,” Environmental Health Perspectives 117, no. 3 (March 2009): 309–15. 109 Richard A. Becker et al., “Good Laboratory Practices and Safety Assessments,” Environmental Health Perspectives, 2009: A482.

Lanier-Christensen

marginalize research damaging to corporations—specifically to cast doubt on scientific evidence of the adverse effects of environmental toxicants and secondhand tobacco smoke. The Advancement of Sound Science Coalition (TASSC)—established by Phillip Morris public relations firm APCO Associates and later funded by Altria (Phillip Morris’s parent company) and ExxonMobil—coordinated efforts to challenge legislation against secondhand tobacco smoke. In 1996, TASSC turned its sights on the science of endocrine disruption, including the seminal book published that year, Our Stolen Future, which chronicled the development of scientific evidence linking synthetic chemicals with endocrine disrupting effects, drew widespread attention, and galvanized governmental action.110 “Sound science”—often juxtaposed with “junk science“—has been used strategically to narrowly determine what counts as legitimate science. Pressures for regulators to use “sound science” imply that they do not use “sound” enough science. Public health scientists Roni Neff and Lynn Golman have argued that “‘sound science’ efforts have raised the bar for assurance so high as to challenge the government’s ability to protect the public.” They charge that antiregulatory advocates have used this approach to persuade the government to “disregard scientific evidence altogether in crafting regulatory standards.”111 As David Michaels wrote in Doubt is their Product, “the sound science/junk science dichotomy has worked wonders as a public relations gimmick and has gained widespread acceptance in the current debate over the use of scientific evidence as policy.”112

110

Naomi Oreskes and Erik M. Conway, Merchants of Doubt: How a Handful of Scientists Obscured the Truth on Issues from Tobacco Smoke to Global Warming (New York: Bloomsbury Press, 2010), 150-152; Vogel Is it Safe, 124-129. 111 Roni A. Neff and Lynn R. Goldman, “Regulatory Parallels to Daubert: Stakeholder Influence, ‘Sound Science,’ and the Delayed Adoption of Health-Protective Standards,” American Journal of Public Health 95, no. S1 (2005): S81–91. 112 David Michaels, Doubt Is Their Product: Manufactured Uncertainty and Public Health (New York: Oxford University Press, 2008), xii. On the impact of the “sound science” movement in science and the law, see Thomas O. McGarity, “On the Prospect of Daubertizing Judicial Review of Risk Assessment,” Law & Contemp. Probs. 66

Lanier-Christensen

The “good practice” rhetoric of Good Laboratory Practices clearly evokes the “sound science” construction and has been used similarly.113 Industry actors and regulators alike have appealed to specific studies’ compliance with “good practice” as reason enough to trust them and cited other studies’ non-compliance as reason to distrust them. In addition to having “followed procedure,” they can appeal to a normative standard. This is why many actors frame debates about procedural versus investigative studies in terms of GLPs specifically, when both GLPs and Test Guidelines are involved. The rhetorical flourish makes it hard to imagine an acceptable alternative—after all, what is the alternative to good practice if not bad practice?

Conclusions In this paper, I have argued that toxicity testing standards—namely, Good Laboratory Practice quality assurance and reporting standards and Test Guideline protocols—were adopted for two primary reasons. First, the U.S. FDA developed GLPs to make chemical testing governable. Second, the twenty-four members of the Organisation for Economic Co-operation and Development adopted GLPs and Test Guidelines as part of the Mutual Acceptance of Data framework to facilitate trade. While industrial actors played an important role in elaborating the standards, it is not a clear-cut case of regulatory capture. Standardized protocols were not created in the service of industry per se, but in large part to help regulators craft an efficient, transparent and governable system.

(2003): 155; Chris Mooney, The Republican War on Science (Basic Books, 2007); Neff and Goldman, “Regulatory Parallels to Daubert”; Carl Cranor, “Scientific Inferences in the Laboratory and the Law,” American Journal of Public Health 95, no. S1 (July 2005): S121–28; Michaels, Doubt Is Their Product. 113 Leonardo Trasande et al., “Peer-Reviewed and Unbiased Research, Rather than ‘Sound Science’, Should Be Used to Evaluate Endocrine-Disrupting Chemicals,” J Epidemiol Community Health 70, no. 11 (2016): 1051–56.

Lanier-Christensen

Nonetheless, regulators’ reliance on the standards has ultimately privileged corporate knowledge claims over others. They do so not because of who makes the knowledge claims, but rather how they are made. Investigative studies on EDCs generally make different assumptions and use different methods than traditional toxicology and therefore is left out of regulatory review.114 Governments have made purposeful decisions that regulations would not be based on comprehensive information. This stands in contrast to what one would expect from technocratic governance: gathering all available evidence and using it to make decisions. It also stands in contrast to environmental health policy discourse that has focused on “best available” science and “evidence-based” interventions.115 Regulators made these choices to make chemicals governable and to facilitate international trade, reflecting the fact that the epistemic demands of governance and market capitalism differ from those of public health. This is a problem in science in regulation writ large—a weakness in how the state knows, which is generally based on standard protocols and credentialed experts. Standardized approaches have certain benefits including bureaucratic efficiency, but, in the case examined here, it resulted in two significant negative effects. First, regulators’ reliance on this knowledge production practice created an inability to know and act on knowledge produced by other means. Second, while these standards were in part created to police how corporations produced knowledge about their products, they ultimately had the perverse effect of privileging corporate knowledge claims over public safety.

114

Frederick S vom Saal and Laura N Vandenberg, “Update on the Health Effects of Bisphenol A: Overwhelming Evidence of Harm,” Endocrinology 162, no. 3 (2020): bqaa171. 115 National Research Council, Strengthening Science at the U.S. Environmental Protection Agency: ResearchManagement and Peer-Review Practices (Washington, D.C.: National Academies Press, 2000).

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

Thank you Colleen! I think we have a lot to talk about! Before beginning I want to disclose that much of the research for this brief synopsis emerged through discovery during my participation in depositions and court cases in some mammoth efforts on behalf of schools, communities and the state in Seattle, Washington, Oregon and elsewhere. I am presently serving as a witness in states and cities around the country who are suing Monsanto, the sole producer of PCBs, the subject of this paper, for having damaged its citizens and polluted their environment. I also want to acknowledge the work of others who managed to put together extraordinary histories of part of this story decades ago with little of the data that is now emerging from these lawsuits. I am sometimes humbled by what good historians can do. Specifically, I want to acknowledge, Keith Schneider who, in the 1980s published Faking It, The Case against Industrial Bio-Test Laboratories.

The 19th-century economy depended on laying thousands of miles of railroad; digging canals; logging the nation’s forests; mining iron, coal, and lead; harvesting crops; picking cotton; and running textile mills, work that put millions of industrial workers and consumers at risk. By the beginning of the twentieth century, as the country emerged from its rural past into its new role as one of the world’s largest industrial societies a host of new chemicals had been introduced: radium, phosphorus, huge quantities of organific and inorganic lead in auto and gasoline, to name a few. And with their introduction, epidemics of radium poisonings, phossy jaw, and lead poisoning among workers and children alike took their toll. In the early years of the new century, the “death roll of industry” and the image of the industrial factory as a slaughterhouse of cattle and workers alike, combined with labor unrest and strikes among miners, steelworkers, chemical workers, lead painters, miners, and even bakers.

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

For many during the early century in what was called the Progressive Era, calls for government regulation of a variety of industries followed the publication of Upton Sinclair’s The Jungle and the numerous strikes over safety in mines and mills, while attention to the triangle fire that claimed 134 young women who jumped to their deaths, documented in Hearst and other newspapers around the country, led to massive strikes and public outcries for government regulation across the nation. Factory inspection commissions, multivolume studies of company misbehavior, and legislative acts like workmen’s compensation augured a threatening future to industry leaders seeking to remain free of the possibility of a “heavy hand of the state.” In response, industry promoted an alternative to government intervention in the management and regulation of their companies: voluntary actions to protect the public through the testing of their products, protection of the equipment, and warnings and information to the public about danger and how to protect. Organizations like the National Safety Council, the Industrial Health Foundation, and even the Manufacturing Chemists’ Association representing the chemical industry, pledged to test their products and guarantee the safety of materials introduced into the environment in exchange for limiting the reach of government regulators. If there were dangers, they promised to let users know what they were and how to protect ourselves. We could, we were assured, depend upon industry to warn us of dangers, and guarantee that their products were safe or could be used safely. The government largely accepted the idea that in a capitalist, free market economy we as a society should depend on the integrity of industries which would provide the necessary scientific data that could be used as the basis of relatively loose regulation. In exchange for this good-hearted industry role, government would stay out of industry’s lives.

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

While some called for government regulations of workplaces and consumer products, the business community proposed a set of “voluntary” alternatives to inspectors, rules, laws and oversight of industrial practices. Specifically, organizations like the National Civic Federation, made up of some of the nation’s largest corporations, along with the National Safety Council and National Association of Manufacturers led by Andrew Carnegie proposed a set of voluntary principles that would forestall government’s heavy hand. Companies promised to self-regulate and, if and when their products presented danger, to warn and protect the worker and consumer. This pattern emerged from a century of discussion about the obligations of industry and government in the protection of workers and consumers By the start of the Depression of the 1930s, as strikes over working conditions multiplied and as consumer groups moved to protect citizens from harm, the use of thousands of new chemicals led to the replacement of informal agreement to more formal standards that industries agreed to abide by. The Manufacturing Chemists Association (MCA), representing the chemical industry, was among the first industry trade associations to articulate the responsibilities of industries producing dangerous materials to the public. In the late 1930s it issued a “Statement of Legal Principles” outlining how it planned to counteract growing New Deal pressure to control the dangers of its products without government interaction. In it MCA told its members, “The manufacturer or one who holds himself out to be the manufacturer must know the qualities of his product,” and further, they needed to understand that the “manufacturer[s] cannot escape liability on the ground that he did not know it to be dangerous.” Testing and reporting was central to this obligation to know of potential dangers. A company was responsible legally and morally if they sold products that lacked instructions on how to properly use them or avoided telling consumers what they knew of its dangers. Further, the producer had to know those

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

dangers and by testing thoroughly it products before putting them in the marketplace. In the 1930s, 40s, and 50s, trade associations like the MCA, the National Safety Council along with the Industrial Health Association and the ACGIH, as well as a host of chemical and other companies all stated their fealty to the idea that they had an obligation to test their products for both short term and chronic dangers through animal studies and continual observation. If industry maintained strict vigilance over its products, there would be no reason for government to inspect, test, control or regulate it. If industry abided by these principles, there would be no need for the heavy hand of government to interfere in the lives of corporations. In 1939, trade association of the chemical, paint and other industries sent out these guidelines to their own members, and told them of their wide applicability: the “principles . . . [are] generally applicable to situations arising from the use of harmful substances.” They made clear that these principles were applicable not only to the workforce but to the general public as well. As the medical director of Union Carbide, Thomas W. Nale, later reflected, “Manufacturers could have waited for the legislative branch of government to write the [laws] for them to protect the public,” but they anticipated and were able to head off regulatory action. From the first, the degree to which industry would abide by its own established principles depended in large measure on the stakes. For many products and chemical formulations there was little conflict between profit and safety. But in the case of some chemicals, particularly the chlorinated hydrocarbons, like DDT, PCBs, VCM, emerging in the more conservative post-war years as major profit centers as pesticides (DDT), in electrical equipment, generators, plastics, paints and a host of other new consumer uses, the temptation of companies to look the other way and to avoid finding out too much about its products was overwhelming. Too much research and science was antithetical to companies’ corporate interests and limiting what you knew about

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

danger was more important than necessarily s finding out too much. And with the more conservative political environment which seemed to end the threat that the New Deal activist government posed, there was little upside to finding out how much of a material might prove toxic to rats, dogs and mice. Selective ignorance was useful in balancing safety and profits. The discovery and uses of polychlorinated biphenyls, PCBs, in the 1930s and beyond are a case in point. A unreactive, chemically inert oddity that had sat on the shelves of chemists for decades with no apparent use, suddenly emerged in the Depression to be of invaluable use as an insulator of electrical equipment and as a plasticizer in a host of new synthetic materials then emerging from. Laboratories around the nation. In short order it was used paints, plastics, carbonless copy paper, adhesives, electrical transformers and capacitors, and numerous other products. For Monsanto, the sole producer of PCBs in the nation and a dominant producer throughout thew world, the product was a goldmine. PCBs were an indestructible insulating material useful in the growing electrical power industry of the 1930s. As the New Deal set up programs to build damns, and lay powerlines throughout the South and West a huge market for PCBs in generators, as a insulating material on wire led Monsanto to profit greatly. From the first, it was apparent that workers using this material suffered a number of complaints, most specifically chloracne, a systemic skin rash and yellow atrophy of the liver, raising worries about PCBs dangers. Studies in the 1930s by Harvard’s Philip Drinker documented the potential for harm that PCBs posed. In 1966, two Swedish investigators, Jensen and Widmark, while looking at environmental samples they suspected of being tainted with DDT, and were surprised to discover instead PCBs in virtually everything they tested: salmon, pine needles, even in baby’s hair and mothers milk. In the United States, PCBs were soon discovered to have polluted the West Coast of the country, threatening peregrine falcons and vast

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

stretches of coastline. And, in the mid-1960s, it was identified in animal and human tissue, fish, waterways, and birds throughout the world, leading to demands for information as to its toxicity. By the 1960s, PCBs had become widely distributed throughout the environment and, with the environmental movement’s awakening following the publication of Rachel Carson’s Silent Spring in 1962 and the attention to DDT that followed, PCBs became the focus of journalists and environmentalists attention. By 1969 PCBs were found to have polluted the Great Lakes, the Gulf Coast, the Hudson River, and cows in the Midwest, and chickens throughout the country. Tens of thousands of pounds of chicken about to be used in Campbells Soup had to be destroyed, cattle, eggs, milk and even Purina dog chow throughout the nation had had to be destroyed. Soon, Monsanto was being asked about the information it had on its long-term testing of PCBs and their effect on humans and animals. It turned out they had never done any. Calls for information about this chlorinated hydrocarbon, DDT’s close cousin, emerged from a press and consumer advocates alarmed that this stuff seemed to be everywhere. What did Monsanto, the sole producer of PCBs in the country and of over 40 percent of all PCBs in the world know about its short- and long-term impact on human and animal health? What they learned was that Monsanto had done virtually no long term chronic studies of this indestructible material, despite decades of having produced and profited from it. In 1969, after three decades of pouring literally hundreds of millions of pounds of PCBs into the environment, Monsanto decided to test their product. “Testing” PCBs This frantic effort to make up for decades of ignorance led to one of the best-documented episodes of scientific manipulation and fraud involving Monsanto’s PCB studies and Industrial Bio-Test Laboratories (IBT), a private testing laboratory in Illinois, which in the early 1970s

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

conducted long-term studies using rats on a variety of chemicals for various corporations, including Monsanto. In 1969, faced by movements within congress to ban the production of PCBs and calls from panicked consumer groups for information, Monsanto approached IBT to conduct chronic two-year animal toxicity tests on polychlorinated biphenyls (PCBs). From the first, these studies were seen by Monsanto as part of a larger strategy to allay the concerns of the public and government— particularly the Food and Drug Administration (FDA) and the newly established EPA— to prove that PCBs “do not constitute a serious threat to the public health” and specifically were not carcinogenic, the big fear at that moment. But early tests did little to qualm the waters for according to the internal documents, they did not turn out to be “as favorable as we [Monsanto] had hoped or anticipated.” The first studies of leghorn chickens reproduction were “Particularly alarming” because of “evidence of [PCBs’] effect on hatchability and production of thin egg shells.” From the first, the objectives of Monsanto were clear: Rather than report this data, a systematic effort to first produce better results emerged. Monsanto arranged with IBT to repeat “some of the studies” in order “to arrive at better conclusions.”8 They sent IBT new samples of their PCBs that they claimed were “clean[ed] up” and told IBT they were paying them to “find a higher ‘no effect’ level,” a potential “safe” level below which the experimental animals would not show symptoms of damage. And a series of reports that emanated from the Northbrook, Illinois testing company that were sent to the EPA “found” exactly that: By 1973, they gloated that they had prolonged the sales life of PCBs through their research program: “the most important data which has led the government agencies to permit the continued but constricted

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

use of polychlorinated biphenyl are the extensive animal toxicity studies which we have completed in the last two years.”10 But the reliability of those studies was belied by two facts that were to emerge a few years later: first, the actual conditions in the IBT labs that tested PCBs for Monsanto were soon found to be compromised. Philip Smith, an assistant toxicologist in the IBT labs where PCB chronic toxicity studies were conducted, described the gross conditions under which the experimental animals were kept, which compromised the collection of reliable data: “[L]oose and wild [rats] . . . were in the rooms . . . chewing the feet off of the [experimental] animals that were in the cages.” He explained it was “difficult to tell the difference between loose laboratory animals and loose [wild] animals that have been raised outside and gotten in,” as interbreeding had occurred and technicians were not able to distinguish which rats were which. The poor professional standards maintained in the lab can be gleaned from Smith’s description that “technicians . . . were caught burning rats’ testicles with lit matches.” Dead rats were often left to decompose so badly that they “would ooze through the bottom of their cages, and all their tissues would be at a total los[s] for any pathology work.”11 Animal caretakers reported “that there were many dead animals that were stinking so bad that [the] caretaker did not want to go into the room to change the water bottles” and new, live animals were substituted for dead ones with no acknowledgment.12 Despite the obviously compromised test conditions, IBT produced seemingly scientifically rigorous reports on three of Monsanto’s PCB products (Aroclor 1254, 1260, and 1242), claiming that testing proved PCBs were not carcinogenic.13 Otis Fancher, a toxicologist at IBT, wrote to his colleagues as early as 1972 that much of the work was so shoddy that he “was ashamed to publish the work done.” He wrote that “much of the data are

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

either fudged or collected with carelessness of incompetence, particularly the data for the supplementary studies of [PCBs].”14 The second issue involved fraud: simply, Monsanto’s man at IBT made up data. In fact, data reported were inaccurate or literally invented and the language was altered by Monsanto officials themselves. In 1975, IBT’s Joseph C. Calandra sent a draft of their latest “AROCLOR 2-year Rat Feeding Studies” to George Levinskas, Monsanto’s manager of environmental assessment and toxicology, listing Aroclor 1254 as being “slightly tumorigenic.” Levinskas objected, asking that the phrasing be changed to “does not appear to be carcinogenic,” a simple but important revision that avoided raising government concerns about cancer. Calandra complied.15 What was perhaps the most egregious effort by Monsanto to shape reports that were to be used by the EPA was the use of one of its employees to supervise the ostensibly objective studies produced by IBT. Wright was employed at Monsanto beginning in 1965 as a senior research chemist and from 1968 until 1970 as a research group leader. In 1970, as IBT began its two-year chronic testing of PCBs for Monsanto, Wright moved to IBT, where he directed the toxicology lab that oversaw these studies. In late 1972, he returned to Monsanto as the toxicology manager and stayed at Monsanto until 1984, shortly before his conviction for having conspired to use the US Postal Service to defraud the government was upheld, and he was imprisoned. Philip Smith, the lab assistant in the IBT PCB studies, gave vivid descriptions of how Wright had falsified data that ended up in the report sent to the government. “The body weight data [were] non-existent,” Smith testified in one deposition he gave years later. “For intervals it was not collected.”16 He knew that, “because under Paul Wright’s instruction, I plotted out the body weight data that we had in the department and all of the data that we could find in the storage area of the department.

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

Then he [Wright] plotted out and gave me body weight numbers to put into the report for all the spaces that we had no records for.” Smith “watched him” “make up numbers” “out of his head.”17 In 1976, the FDA found inconsistent data in one of IBT’s studies, leading it to scrutinize IBT’s studies. In 1977, as questions about its studies accumulated, IBT requested a meeting with Monsanto about the chronic toxicity testing that they had conducted on several different substances. In July, Monsanto officials, including M. C. Throdahl, the company’s group vice president for environmental policy and member of the Board of Directors, and Paul Wright, now having returned to Monsanto and soon to be director of its Environmental Health Laboratory, met with officials from IBT, including A. J. Frisque, its president, and F. R. Current, IBT’s legal counsel. The “reason for the meeting,” according to an internal Monsanto memo, was the “recent actions by the FDA and the EPA (pesticides) in questioning the validity of toxicology studies performed by IBT.” The FDA had specifically questioned the studies performed on trichlorocarbanilide (TCC), an antibacterial agent that, based on IBT reports, the FDA had approved for use in soaps and lotions.18 IBT reviewed its operations and “discovered . . . major problems . . . at IBT’s Northbrook, Illinois, facility,” where their long-term PCB and other chemical rodent studies were conducted.19 At the meeting, Monsanto Vice President Throdahl “asked specifically whether ‘fraud’ was involved in the twelve” Monsanto long-term rodent studies, to which the president of IBT “replied that ‘extrapolation’ and ‘faulty interpretations’ were part of the problems . . . and that he guesses this constitutes ‘fraud.’ ” Monsanto’s representatives called this “a very damaging admission [that] was made in the presence of a [IBT] lawyer who took no exception to the question or answer.”20

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

In the late 1970s and into the early 1980s, the US government investigated the toxicological work that had been done at IBT. On May 4, 1981, a federal grand jury handed down an indictment focused on TCC, one of the 12 Monsanto chemicals then being tested in the rat toxicology labs. The indictment charged former IBT president Joseph C. Calandra, Moreno L. Keplinger, Paul L. Wright (now back at Monsanto), and James B. Plank with fraud and passing corrupted data to the federal EPA as the basis for prolonging the production of PCBs. The indictment charged that between 1970 and 1977, Wright and the others had devised and intended to devise a scheme to defraud clients and government agencies by writing and distributing false and fraudulent study reports and false and fraudulent explanations of study reports, and by concealing the fraudulent nature of the study reports and explanations of studies and study reports.21 The accusations focused on Monsanto’s Wright, and three IBT principles, who had represented that the studies had lasted 24 months when in fact the defendants “knew that the report included data from a substantial number of animals that had been on the study for significantly lesser periods of time.” Wright had removed dying and sick animals from their data base, substituting and counting newly replaced healthy animals in their stead. The defendants created inaccurate mortality tables “which the defendants then knew to be false in that it substantially under-reported . . . the number of animals that died during the study,” and thus “concealed . . . that the animal mortality . . . was substantially greater than reported in any version of the study report.”22 The indictment detailed that Monsanto’s Wright made “false, fictitious and fraudulent statements and representations . . . and concealed and covered up material facts.” Wright, who by 1976 had returned to Monsanto as the company’s “toxicology manager,” falsely predated the

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

study by two years, to March 21, 1974,23 showing that he was aware of, and continued to engage in, fraud after he had returned to Monsanto.24 In January 1978, the FDA and the EPA investigated three other long-term studies that IBT had conducted, including two that were done for Monsanto on Machete, another Monsanto herbicide, and monosodium cyanurate (ACL). The FDA concluded that in both studies there was evidence that Monsanto knew of “significant problems” at IBT “prior to submitting their [Monsanto’s] report to the US Government.” There was “strong evidence of client’s being knowledgeable of inaccuracies in the final report,” and in the other IBT study of ACL there was “strong indication of client’s knowledge of the deficiencies before they issued their report to the US Government.” The inspectors reported that “anticipated toxicity problems known to both the client [Monsanto] and test facility [IBT] were deliberately overlooked.”25 The trial of the four defendants began at the United States District Court in here in Chicago in April 1983 and continued for several months. Almost immediately, national and local newspapers picked up on the significance of the case, pointing out that it raised many questions about the integrity and honesty not only of IBT and Monsanto but of the voluntary regulatory project of industry as a whole. Monsanto’s press office denied that Wright was guilty of any fraud, arguing that “Mr. Wright is innocent and … the trial will vindicate him.”26 This was despite the fact that at a meeting of Monsanto and IBT officials as early as 1977 Monsanto had been bluntly told by IBT’s president that the IBT studies that Wright were fraudulent. Internally, Monsanto never tried to hold anyone accountable for the fraudulent work. Rather it rewarded its employees for having postponed the reckoning with the government for nearly a decade. In 1977 Monsanto promoted Wright to director of the Environmental Health Laboratory, and in 1981, when Wright was indicted, he was assigned to work on special projects,

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

including overseeing its Material Safety Data Sheets, the documents that OSHA demanded be available to warn workers about dangers of substances they were handling.27 Far from being reprimanded or fired, Wright was given merit raises in 1977, 1978, and 1980. In 1982, a year after he was indicted,28 Monsanto paid his legal defense to the tune of $1.4 million.29 And, even after he and his collaborators at IBT were sentenced to prison, he was provided back pay and financial support. In 1976, the newly passed Toxic Substances Control Act (TSCA) made PCBs the object of attention when it specifically called for Monsanto to cease its production. In 1977, the last barrel of PCBs left Monsanto’s two production plants, one in Anniston, Alabama and the other in Sauget, Illinois, right across from Monsanto’s headquarters in St Louis, ceased production and quickly sold off its remaining store of PCBs to companies around the country. Despite the duplicity that surrounded Monsanto’s studies, Monsanto continued to cite them well into the future as evidence of the safety of PCBs. In 1979, for example, a Monsanto publication cited the IBT studies of PCBs as “the most comprehensive safety tests of the time.”30 Further, in 1983 and 1985 Monsanto continued to cite the IBT studies in their Material Safety Data Sheets.31 As late as 2018, one of their experts in PCB litigation depended on these fraudulent studies only to be flustered and discredited on the stand when he was confronted by plaintiffs’ attorneys about their fraudulent nature a fact of which he had never been informed by Monsanto lawyers.32 In August 1983, Paul Wright, Moreno L. Keplinger, and James B. Plank, former assistant toxicology manager, were convicted of fraud and sentenced to jail.33 But even these convictions did not give Monsanto pause: the company gave Wright a “golden parachute,” providing him with full retirement benefits, accrued vacation time, one month’s severance, and the services of a recruitment specialist to help him find future jobs when he was released from prison.34

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

In the period following the expansion of government regulation in the early 1970s, the government continued to depend on the integrity of industries and their private laboratories to provide them with information needed to establish new standards. Hence, the EPA, OSHA, and the Consumer Product Safety Commission, along with older agencies like the FDA— government agencies with neither the resources nor the inclination to test the myriad chemicals and synthetic products yearly produced by US industry—depended on companies’ themselves for vital information. Following the revelations discussed here, “Good Laboratory Practices Regulations” were promulgated that were intended to guarantee the quality of research upon which federal regulations depend.35 But the central tension between the interests of industries and the interests of public health remained. Here, we show that the influence of industry on laboratory practices made the corruption of science more likely. With or without regulatory standards, we need to maintain vigilance over companies whose self-interest has distorted science and will continue to do so. 1. David Rosner and Gerald Markowitz, “ ‘Educate the Individual . . . to a Sane Appreciation of the Risk’: A History of Industry’s Responsibility to Warn of Job Dangers Before the Occupational Safety and Health Administration,” American Journal of Public Health 106 (January 2016): 28–35 (https://doi.org//10.2105/AJPH.2015.302912). 2. For examples of the types of historical internal documents made available through various court proceedings, see Columbia University’s and CUNY’s Web site http://www.toxicdocs.org, which provides millions of pages of internal corporate memos, private research studies, corporate statements and public relations materials, and more. 3. Robert Proctor, The Golden Holocaust: Origins of Cigarette Catastrophe and the Case for Abolition (Berkeley, CA: University of California Press, 2012); Allan Brandt, The Cigarette Century: The Rise, Fall, and Deadly Persistence of the Product that Defined America (New York, NY: Basic Books, 2009); Gerald Markowitz and David Rosner, Deceit and Denial: the Deadly Politics of Industrial Pollution, with a new epilogue (Berkeley, CA: University of California Press, 2013); Elena Conis, How to Sell a Poison: The Rise, Fall, and Toxic Return of DDT (New York, NY: Bold Type Books, 2022); Gerald Markowitz and David Rosner, Lead Wars: The Politics of Science and the Fate of America’s Children (Berkeley, CA: University of California Press, 2010); David Michaels, Doubt Is Their Product (New York, NY: Oxford University Press, 2008); David Michaels, The Triumph of Doubt: Dark Money and the

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

Science of Deception (New York, NY: Oxford University Press, 2020); Naomi Oreskes and Erik Conway, Merchants of Doubt: How a Handful of Scientists Obscured the Truth on Issues From Tobacco Smoke to Global Warming (New York, NY: Bloomsbury, 2010). 4. Eliot Marshall, “The Murky World of Toxicity Testing,” Science, new series, 220 (June 10, 1983): 1130–1132 (https://doi.org/10.1126/science. 6857237); Keith Schneider, “Faking It: The Case Against Industrial Bio-Test Laboratories,” Natural Resources Defense Council, Amicus Journal, Spring 1983, republished in “Planet Waves,” https://planetwaves.net/contents/faking_it.html; Edward Herman, “Corporate Sovereignty and (Junk) Science,” Z Magazine, November 1998 (http://www.zcommunications.org/corporatesovereigntyand-junk-science-by-edward-herman), accessed July 14, 2022; Editorial, “The Scandal in Chemical Testing,” New York Times, May 16, 1983, Section A, p. 18 (https://www.nytimes.com/1983/ 05/16/opinion/the-scandal-in-chemical-testing. html), accessed January 3, 2023. 5. See Michael Evard, Plaintiff, v Monsanto Company, et al., Defendants, Circuit Court of Cook County, Illinois, Case No. 2019-L-011574; Neal Frank v Monsanto Company, et al., Defendants, First Judicial District Court, County of Santa Fe, State of New Mexico, No. D-101-cv-2021Roundup, Monsanto’s widely used pesticide. Keith Schneider, “Lab Fraud—or Government Harassment?”, Boston Globe, April 3, 1983, p. A22; Keith Schneider, “Faking It, The Case Against Industrial Bio-Test Laboratories,” in The Amicus Brief (Washington, DC: National Resources Defense Council, 1983). IBT, at the time the nation’s largest private laboratory, was founded in 1953 by Joseph Calandra, a professor at Northwestern University, who organized three laboratories in the Chicago area to serve industry’s need for testing new products. By the early 1970s it had conducted about 22 000 toxicology studies for scores of corporations, representing between 35% and 40% of all tests conducted in private labs in the country. “Nearly half of IBT’s studies were used to support federal registrations of a mammoth array of products: insecticides, herbicides, food additives, chemicals for water treatment, cosmetics, pharmaceuticals, soaps and bleaches, even coloring for ice cream” (https://planetwaves.net/contents/faking_it.html), accessed January 4, 2023. 6. For a full discussion of the identification of PCBs as an environmental pollutant and Monsanto’s response, see Gerald Markowitz and David Rosner, “Monsanto, PCBs, and the Creation of a ‘World-Wide Ecological Problem,’ ” Journal of Public Health Policy 39 (2018): 463–540 (https://doi.org/10.1057/s41271-018-0146-8); for a detailed examination of Monsanto’s history of PCB production and pollution, see Ellen Griffith Spears, Baptized in PCBs: Race, Pollution, and Justice in an All-American Town (Chapel Hill, NC: University of North Carolina Press, 2014). 7. “Report of the Aroclor ‘Ad Hoc’ Committee (Second Draft),” October 15, 1969, “Company Confidential Attorney-Client Privilege” (https://www.toxicdocs. org/d/QXGD7Vm44Jx8qdpk742948pp7?lightbox=1), accessed January 3, 2023; Markowitz and Rosner, “Monsanto, PCBs, and the Creation of a “World- Wide Ecological Problem.’ ” 8. “The PCB-Pollution Problem,” January 21 and 22, 1970, meeting with General Electric Co. (https://cdn.toxicdocs.org/Ed/EdoyOJrmzBNJpQnGg36Oj3Y2j/EdoyOJrmzBNJpQnGg36Oj3Y2 j.pdf), accessed January3, 2023.

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

9. Wheeler to Otis Fancher, Director, IBT, May 25, 1970 (https://www.toxicdocs.org/d/YGBLJpeRa 2jDEjL7MqLRzwZgn?lightbox=1), accessed January 3, 2023. 10. Wheeler to Wilbur H. Speicher, Administrator, Industrial Hygiene, Westinghouse, February 23, 1973 (https://www.toxicdocs.org/d/3Q77xYrYKbKJaL9XdERMR6zwx?lightbox=1), accessed January3, 2023. 11. Deposition of Philip S. Smith, July 9, 1987, Cecil Scott, et al., Plaintiffs, v Monsanto Company, Defendant, in the United States District Court, Eastern District of Texas, Beaumont Division, B-84-1103-CA, p. 244, 253 (https://www.toxicdocs.org/d/Xzav0m9vEDB37Q2M6MRa0rODJ?lightbox=1), accessed January 12, 2023. 12. Deposition of Philip S. Smith, July 9, 1987, Cecil Scott, et al., Plaintiffs, v Monsanto Company, Defendant, in the United States District Court, Eastern District of Texas, BeaumontDivision, B-84-1103-CA, p. 254 (https://www.toxicdocs.org/d/r3EnY LgeRGvQVrRwOQN7ev2q?lightbox=1), accessed January 4, 2023. 13. Industrial Bio-Test, Laboratories, Inc., “Report to Monsanto Company, Two-Year Chronic Oral Toxicity With Aroclor 1254 in Albino Rats,” Nov. 12, 1971, IBT No. B7298(https://www.toxicdocs.org/d/2NNN6GKONbd0ezNzbge4O0bn7?lightbox=1), accessed February 14, 2023. 14. Otis Fancher to Don[avon] [Gordon], handwritten note, January 14, 1972, M 42627, in which Fancher also said that he was “ashamed to put my name to it” (https://www.toxicdocs.org/d/ZJ5 5yDwBnqwJOq5QRqYxRoXdL?lightbox=1), accessed February 14, 2023. 2-Year Rat Feeding Studies,” July 18, 1975(https://www.toxicdocs.org/d/NNkDddqyXv6gR9L R4wJqn7oM8?lightbox=1), accessed January 10, 2023. J. C. Calandra to George Levinskas, “Re: Aroclor 2-Year Rat Studies,” August 4, 1975(https://www.toxicdocs.org/d/dQM4xXnvnXpg 7rYaY46O7Lq96?lightbox=1), accessed January10, 2023. 16. Deposition of Philip S. Smith, July 9, 1987, “Cecil Scott et al., Plaintiffs, vs. Monsanto Company, In the United States District Court Eastern District of Texas, Beaumont Division, B84-1103-CA,MONSPCB0019134-19135, p. 183. 17. Deposition of Philip S. Smith, July 9, 1987, pp. 189–190. See also Trial Testimony of Philip Smith, October 28, 1991, WATER_PCB-00056547, pp. 1073–1074; see also ibid, pp. 1130– 1137); United States District Court, Northern District of Illinois, Eastern Division, “United States v. Joseph Calandra, Moreno Keplinger, Paul Wright and James Plank,” “Special October 1980 Grand Jury Charges,” STLCOPCB80017757. See also Keith Schneider, “Faking It: The Case Against Industrial Bio-Test Laboratories,” The Amicus Journal, Spring 1983, pp. 15–21 (https://www.toxicdocs.org/d/eybVeZMgomD8N3KRv1k41GJq?lightbox=1), accessed February 14, 2023. 18. W. Wayne Withers and G. Roush to M. C. Throdahl, July 13, 1977, PCB-ARCH075738 (https://www.toxicdocs.org/d/r3EnYLgeRGvQVrRwOQN7 ev2q?lightbox=1), accessed January 6, 2023. 19. Ibid. 20. Ibid. 21. The United States of America v Joseph C. Calandra, Moreno L. Keplinger, Paul L. Wright and James B. Plank, Filed, May 4, 1981, United States District Court, Northern District of Illinois, Eastern Division, MONSPCB0021020-MONSPCB0021037

Corrupted Science: PCBs, Roundup and Monsanto in the Early Years of the EPA David Rosner, Columbia University

(https://www.toxicdocs.org/d/pp9be89q8qkoZEzzeR39R1V1d?lightbox=1), accessed February 14, 2023. 22. The United States of America v. Joseph C. Calandra, Moreno L. Keplinger, Paul L. Wright and James B. Plank, Filed, May 4, 1981, United States District Court, Northern District of Illinois, Eastern Division, May 4, 1981, MONSPCB0021025 (https://www.toxicdocs.org/d/pp9be89q8qkoZEzzeR39 R1V1d?lightbox=1), accessed January 4, 2023. 23. Ibid. 24. Ibid. 25. FDA, EPA, Compliance Inspection, “Summary of Findings,” January 16–March 9, 1978, MONSPCB0000621. 26. Keith Schneider, “Accused of Faking Data: Trial Set for Testing Firm, Monsanto Researcher,” St. Louis Post-Dispatch, March 30, 1983, pp. 1, 10; see also Bill Richards, “Papers From Trial of Former IBT Officers Raise Many Questions on Product Safety,” Wall Street Journal, May 13, 1983, pp. 31, 37; see also FDA, EPA, Compliance Inspection, “Summary of Findings,” January 16–March 9, 1978, MONSPCB0000641. 27. January 1983 Performance Review and MONSPCB0013836-39; Monsanto, Paul Wright “Employment”

Physicists as Environmental Experts: The Case of Richard Wilson Rachel Rothschild* “In physical science a first essential step in the direction of learning any subject is to find principles of numerical reckoning and methods for practicably measuring some quality connected with it. I often say that when you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind: it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the stage of science, whatever the matter may be.”1 William Thompson (Lord Kelvin), British Physicist, 1883

Introduction

When the renowned physicist Lord Kelvin made the above remarks at a lecture to the British Institution of Civil Engineers in the late 19th century, he intended to advocate for the value of measurements in physics research.2 Despite his reference to the importance of “numbers” in science, he did not mean to imply that quantification should be the sole metric by which we judge the knowledge claims of all scientific research.3 Lord Kelvin’s comments nevertheless took on a life of their own for decades thereafter as a variety of scientific disciplines sought to assert their expert authority in American society through quantitative methods.4 The social sciences in particular drew on the achievements of physics and its mathematics to make claims of scientific *Assistant Professor, University of Michigan Law School. 1

William Thomson Kelvin, Popular Lectures and Addresses 73–74 (1889). Lord Kelvin did not specify exactly which physicists (or others) were deviating from this approach in his lecture; historians have suggested he either was critiquing the theoretical work of his contemporary, James Clerk Maxwell, or popularized displays of physics experiments. See GRAEME GOODAY, THE MORALS OF MEASUREMENT: ACCURACY, IRONY, AND TRUST IN LATE VICTORIAN ELECTRICAL PRACTICE 4 (2004); CROSBIE SMITH & M. NORTON WISE, ENERGY AND EMPIRE: A BIOGRAPHICAL STUDY OF LORD KELVIN 455 (1989). 3 Id. See also MARGARET MORRISON, RECONSTRUCTING REALITY: MODELS, MATHEMATICS, AND SIMULATIONS 201–05 (2015) (analyzing Lord Kelvin’s comments in the context of his views on the relationship between measurements and mathematical modelling in physics). On economists’ misinterpretations of Kelvin’s maxim, see G. M. PETER SWANN, ECONOMICS AS ANATOMY: RADICAL INNOVATION IN EMPIRICAL ECONOMICS 51–53 (2019). 4 THEODORE M. PORTER, TRUST IN NUMBERS: THE PURSUIT OF OBJECTIVITY IN SCIENCE AND PUBLIC LIFE 72 (2020). 2

objectivity as these disciplines became involved with an expanding administrative state.5 In an act epitomizing this quantitative fervor, the head of the University of Chicago’s social sciences division carved an abbreviated version of Kelvin’s phrase above the entrance to a campus building on the eve of World II.6 Later in the 20th century, another physicist – Richard Wilson – similarly repurposed Kelvin’s endorsement of quantification in an effort to assert his expertise in environmental science.7 Wilson spent most of his career as a professor in Harvard University’s physics department, turning to work in environmental science on behalf of the petroleum and chemical industries in the early 1970s. He would go on to play a significant role in several of the most important 20th century legal cases about scientific expertise: the 1980 Supreme Court decision Industrial Union Department, AFL-CIO v. American Petroleum Institute, known as the “benzene case,”8 the 1993 Supreme Court decision Daubert v. Merrell Dow Pharmaceuticals,9 and the 1997 Supreme Court decision General Electric Co. v. Joiner.10 Through his involvement in these decisions, Wilson helped bring about a shift in how judges evaluated the reliability of environmental research when used to justify regulations or bring tort lawsuits against polluters, leading courts to view probabilistic, quantitative risk assessment as the best method for understanding risks from pollution. Historians of science, including this author, have documented how physicists assisted regulated industries in disputing the environmental and public health dangers from pesticides, acid rain, and climate change.11 In some cases, these physicists had done no research themselves on the topic in question,12 and were simply working as paid representatives for industry.13 In other cases, physicists were very much involved in basic research about the environmental threat in question and thus could better claim to be legitimate experts in scientific disagreements on the relationship between pollution and observed harms.14 However, their work was often warped by industry interests and at times subject to outright manipulation.15 The rise of this “defensive” research

Id. at 71–73, 150–61. On the rise of economic approaches to social problems in the U.S. government, see ELIZABETH POPP BERMAN, THINKING LIKE AN ECONOMIST: HOW EFFICIENCY REPLACED EQUALITY IN U.S. PUBLIC POLICY 42–71 (2022). 6 Report of the Chicago Meeting, December 27-28, 1940, and of the New Orleans Meeting, December 27-30, 1940, 9 ECONOMETRICA 165, 178 (1941). The carving remains there today. See Chelsea Leu, Looking Up, THE UNIVERSITY OF CHICAGO MAGAZINE (May/June 2015). 7 RICHARD WILSON & EDMUND A.C. CROUCH, RISK-BENEFIT ANALYSIS 159 (1982) (Wilson references Lord Kelvin’s lectures in his first major treatise on assessing environmental risks to argue for the importance of quantification and against the use of the “precautionary principle”). 8 448 U.S. 607 (1980). 9 509 U.S. 579 (1993). 10 522 U.S. 136 (1997). 11 See RACHEL ROTHSCHILD, POISONOUS SKIES: ACID RAIN AND THE GLOBALIZATION OF POLLUTION (2019); Naomi Oreskes & Erik Conway, Merchants of Doubt (2010). See also DAVID MICHAELS, THE TRIUMPH OF DOUBT: DARK MONEY AND THE SCIENCE OF DECEPTION (2020). 12 Oreskes & Conway, supra note 11, at 8. 13 Id. at 160. 14 ROTHSCHILD, POISONOUS SKIES, supra note 11, at 58–62. 15 Id. at 77–82, 126, 132–35, 144–48.

establishment helped forestall legislative and regulatory efforts to address fossil fuel pollution, though not indefinitely.16 But while historians have charted the influence of physicists on debates over environmental harm in the halls of Congress, administrative agencies, and among the broader public, to date scholars have not explored their influence on environmental litigation. This contribution to the Clifford Symposium on The Legacy of Industrywide Deadly Misconduct aims to show that the phenomenon of physicists as hired industry experts also permeated the courts following the passage of federal environmental and public health legislation in the 1970s and the rise in toxic tort litigation in the 1980s.17 It does so by examining the work of Richard Wilson, arguably the most prominent such physicist during this period. Wilson assisted the petroleum and chemical industries in pushing back against both regulation of their products and liability for any resulting health and environmental harms.18 He eventually succeeded in shaping judicial perceptions of reliable environmental research to conform with ideals of science dominant among the postwar physics community: namely, the necessity of mathematical quantification for demonstrating real scientific knowledge. The article begins by examining Wilson’s transition to work on environmental pollution for the chemical and petroleum industries, which coincided with a broader movement of physicists into environmental science in the 1970s. On behalf of these industries, Wilson developed a new methodology for risk analysis that sought to use probabilistic calculations to quantify the chance of harm resulting from pollution exposure.19 This method was highly dependent on access to certain types of data, and during this period, many environmental health scientists believed a qualitative approach that incorporated results from many different disciplines and studies would provide a better indication of pollution risks given limitations in data availability. Nevertheless, in a series of decisions during the 1980s and 1990s, the Supreme Court adopted Wilson’s method of quantitative risk assessment as the standard by which to evaluate the reliability and strength of environmental research. Drawing on documents obtained through Freedom of Information Act requests and through discovery in litigation against tobacco, petroleum, and chemical companies,20 Part II assesses Wilson’s work for the chemical and petroleum industries as well as his influence on cases involving toxic chemical regulation and tort liability. It reveals how Wilson employed 16

Id. at 147–48, 170–88. On the relationship between the rise of toxic tort litigation and efforts to limit class action lawsuits, see DEBORAH R. HENSLER, CLASS ACTION DILEMMAS: PURSUING PUBLIC GOALS FOR PRIVATE GAIN 23–25 (2000). On the costs of toxic chemical litigation for corporations, particularly asbestos, see W. Kip Viscusi, Richard J. Zeckhauser, Patricia Born & Glenn Blackmon, The Effect of 1980s Tort Reform Legislation on General Liability and Medical Malpractice Insurance, 6 J. RISK & UNCERTAINTY 165, 166 (1993). 18 While not the sole physicist or expert to weigh in on environmental regulation and tort liability, Wilson was a dominant figure in this phenomenon – arguably the most prominent such figure when it comes to court cases. See Michael Greenberg & Karen Lowrie, Dick Wilson: Applying Uncommon Sense to Risks, 30 RISK ANALYSIS 3 (2010). 19 John Graham, Preface to the Second Edition of Risk-Benefit Analysis, in RICHARD WILSON & EDMUND A.C. CROUCH, RISK-BENEFIT ANALYSIS xi–xii (2001) (describing Wilson’s development of the field of risk analysis and influence on legal decisions like the benzene case). On Wilson’s initial decision to do this work on behalf of the chemical industry, see infra notes 43–45. 20 The two main repositories for these documents are libraries and archives at the University of California, San Francisco (UCSF Industry Documents Library) and Columbia University (Toxic Docs Archive). 17

different definitions of “good science” to serve industry interests, advocating for judges to use exacting criteria when assessing the admissibility of expert testimony in toxic tort suits when some of his own work – conducted on behalf of industry and used to oppose toxic chemical regulations – failed to meet the same standards. The ways in which Wilson influenced judicial conceptions of science without disclosing his financial ties to industry and without the requisite training in environmental health research should serve as a stark warning about the need for judges to evaluate expert qualifications and conflicts of interest more critically.

II.

High Priests of Science in an Era of Environmental Concern

The intense collaboration between the physics community and the military to develop atomic weapons paved the way for many physicists to assume unprecedented positions of authority in American society after World War II.21 Along with their prominence in the highest echelons of government, the physics community enjoyed an enormous influx of federal money for research.22 There was also an explosion in the production of trained physicists at universities in the decades after World War II, reaching an apex in the early 1970s.23 The numbers of newly minted physicists far surpassed trainees in other scientific fields, even those like engineering and chemistry that were also in high demand during the early Cold War period.24 Richard Wilson was a product of this heyday for physicists. He worked on cutting-edge atomic energy experiments as a graduate student in Britain after the end of the war, and joined Harvard’s physics department in 1955 following a few years as a postdoctoral researcher at several top U.S. laboratories.25 His professional opportunities and sources of funding during the next fifteen years were “limitless.”26 Wilson’s research focused on particle physics, which requires the use of expensive accelerators to examine fundamental atomic structures. During his first fifteen years at Harvard, Wilson attempted to make the city of Cambridge the top place for studying highenergy physics through the construction of sophisticated accelerators. Yet a series of misfortunes, including a terrible accident and fire that claimed the life of an MIT undergraduate during the construction of an experimental facility, led to the end of financial support for high-energy physics work at Harvard and MIT by the early 1970s.27 This loss deeply affected Wilson, who felt his 21

DANIEL J. KEVLES, THE PHYSICISTS 367–69; 341–347 (1995). Paul Forman, Behind Quantum Electronics: National Security as Basis for Physical Research in the United States, 1940-1960, 18 HISTORICAL STUDIES IN THE PHYSICAL & BIOLOGICAL SCIENCES 149, 152 (1987). 23 David Kaiser, Cold War Requisitions, Scientific Manpower, and the Production of American Physicists after World War II, 33 HISTORICAL STUDIES IN THE PHYSICAL & BIOLOGICAL SCIENCES 131, 135 (2002) (tracking the yearly number of physics doctorates awarded from 1890 to 1980; during the 1950s, the U.S. was producing the same number of physicists every two years as existed during the entire period from the Civil War to the Great Depression). 24 See id. at 136. 25 Interview of Richard Wilson by Katherine Sopka, Feb. 10 1977, Niels Bohr Library & Archives, American Institute of Physics, College Park, MD, www.aip.org/history-programs/niels-bohr-library/oral-histories/32461; Richard Wilson, Resume (undated) (on file with author). 26 RICHARD WILSON, PHYSICS IS FUN: MEMOIRS OF A LIFE IN PHYSICS 159 (2011). 27 Id. at 406–09. 22

physics research “seemed to be collapsing.”28 He described himself as growing “more and more frustrated in my physics career,” and sought to branch out into “other, related areas” even though, as he later acknowledged, he was “less qualified” for such work.29 The primary area into which Wilson decided to pivot was environmental and public health research – a professional choice that was not unique for physicists during this period. The collapse of Wilson’s physics career coincided with a profound shift in attention to environmental concerns across the country and within the scientific community.30 In response to the widespread public mobilization over environmental threats during the 1960s, Congress passed a raft of environmental legislation between the late 1960s and 1980.31 Most of our major environmental laws, such as the Clean Air and Water Acts, took on their modern forms during this period of legislative activity.32 Greater attention to environmental pollution, and the need for scientists to play a role in understanding and controlling its threats, significantly affected the physics community. The discipline seemingly fell overnight from its position at the top of the scientific hierarchy. As one leading physicist remarked at an annual address to his colleagues, “the ecologists have displaced the physicists as high priests in this new era of environmental concern.”33 The U.S. national laboratories, which had supported the development of nuclear weapons and power during the early years of the Cold War, swiftly created new environmental research divisions that formed the foundation for studying a variety of pollution problems.34 Given physicists’ involvement in the development of atomic technology, examining the harmful effects of nuclear technology was a logical extension of their prior work35 – and one that might allow them to maintain a degree of influence on one of the most salient issues of the time. Wilson first began contributing to these efforts as his research in high-energy physics ran into trouble, with assessments of how nuclear power might reduce other types of air pollution.36 This included expert testimony and government advising on the construction of nuclear power plants as well as opinion pieces intended to sway public opinion on their operation. However, a confluence of events soon brought many prominent physicists – including Wilson – more squarely into work on environmental pollution and its control. Concerned about 28

Id. at 431. Id. at 420. 30 The recognition that pollution could threaten public health and the environment increased in the decades after World II in large part because of research into nuclear fallout and waste. ROTHSCHILD, POISONOUS SKIES, supra note 11, at 16. For example, Rachel Carson’s famous work, Silent Spring, was heavily influenced by scientific research into the dangers of radioactive fallout and waste. See id. 31 Craig Steele, An Analysis of U.S. Federal Environmental Legislation in the Nineteenth, Twentieth and Beginning Twenty-First Centuries, with Emphasis on Presidential Party and Political Majorities in Congress, 26 J. LEGISLATIVE STUDIES 295, 303 (2020). 32 RICHARD J. LAZARUS, THE MAKING OF ENVIRONMENTAL LAW 69–70 (2008). 33 DEP’T OF ENERGY, OAK RIDGE NATIONAL LABORATORY: THE FIRST FIFTY YEARS 154 (1992). 34 Rachel Rothschild, Environmental Awareness in the Atomic Age: Radioecologists and Nuclear Technology, 43 HISTORICAL STUDIES IN THE NATURAL SCIENCES 492, 528 (2013). 35 Id. 36 WILSON, PHYSICS IS FUN, supra note 26, at 437. Wilson also became involved in testifying in support of nuclear power plants, including the infamous Vermont Yankee power plant, subject to one of the earliest National Environmental Policy Act cases. See Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519 (1978). 29

new environmental regulations and energy security in the wake of the 1973 Yom Kippur War, the fossil fuel industry launched multiple new research efforts to bolster the domestic power sector and ensure they had experts capable of representing their interests in scientific debates on environmental pollution.37 To oversee these industry research efforts, they turned to the former “high priests” of science: physicists.38 One prominent, early effort to forge ties between the fossil fuel industry and physicists was launched by Edward Teller, a former Manhattan project physicist who became a top government advisor during the Cold War.39 With the help of businessmen in the fossil fuel industry, Teller established the Edward Teller Center for Science, Technology and Political Thought in 1973 and began hosting events on energy, the environment, and national security with participants from universities, government research laboratories, and the business sector.40 While the Center dissolved by 1975, it existed for long enough to bring physicists like Richard Wilson into the orbit of industry representatives who were hoping to recruit experts to combat environmental regulations.41 Many physicists involved with the Center during its brief period of existence became prominent critics of research connecting fossil fuel pollution to environmental harms.42 At the invitation of several friends in the physics community, Wilson spent the summer of 1974 at Teller’s Center, dubbing it his “right wing summer.”43 It was there that he met the vice president of a chemical company, “the man who started me on my interests in chemical risks.”44 The vice president told Wilson that the company hoped to convince regulators that they were doing enough to control the risk of cancer from exposure to a toxic substance known as vinyl chloride, which was one of the first chemicals government agencies flagged for regulation in the early 1970s due to its harmful health effects.45 Wilson agreed to look into the topic, and in the course of his work for the company, he began to develop a new approach to environmental regulation that would come to be known as quantitative risk assessment. 37 Nilo Lindgren, The First Five Years: Chauncey Starr and the Building of EPRI, 3 EPRI J. 4, 8 (Jan./Feb. 1978) (discussing physicist Chauncey Starr’s vision for developing EPRI environmental research as a way to influence federal policies, which was pivotal in his selection to lead the organization). 38 ROTHSCHILD, POISONOUS SKIES, supra note 11, at 70–72. 39 On Teller’s work on developing the hydrogen bomb, see HERBERT FRANK YORK, THE ADVISORS: OPPENHEIMER, TELLER, AND THE SUPERBOMB 62–66, 75–86 (1989). 40 EDWARD TELLER & JUDITH SCHOOLERY, MEMOIRS: A TWENTIETH CENTURY JOURNEY IN SCIENCE AND POLITICS 519–20 (2009). 41 See National Center for Atmospheric Research, Staff Notes No. 416 (June 14, 1974), “Energy Symposium Conference to be Given” at 4 (on file with author) (describing several courses offered at the Center during the summer of 1974, including one by Wilson on “Energy and the Environment” and one by Teller on “The Problems of Energy— Domestic and Foreign”). 42 One prominent example is Fred Seitz, who participated in the Center and went on to become a leading critic of the scientific consensus on human-induced climate change. For Seitz’s involvement in climate change debates, see Oreskes & Conway, supra note 11, at 170–208. Other examples include physicist and astronaut Walter Cunningham, who served on the Center’s board while it existed. See Russell Lewis, NASA Apollo Astronaut Walt Cunningham Has Died at Age 90, NPR, Jan. 3, 2023, https://www.npr.org/2023/01/03/517135016/nasa-apollo-astronaut-waltcunningham-died (“Cunningham was a physicist and later became known for his skeptical views of climate change, disagreeing with overwhelming scientific belief that humans are to blame for increasing global temperatures.”). 43 WILSON, PHYSICS IS FUN, supra note 26, at 463. 44 Id. at 455, 467. 45 Id. at 491.

The concept of weighing how serious a risk might be, and whether it was worth taking action against it, has a long history. But the notion that one should use a particular methodological approach to quantify risk probabilistically did not emerge until the 1970s,46 and Wilson – working as a consultant to industry – became a key pioneer. Quantitative risk analysis for environmental pollutants grew out of a methodology called systems analysis that employed mathematical techniques to aid complex weapons systems, such as those used for intercontinental ballistic missiles and nuclear weapons.47 The method’s emphasis on quantifying possible malfunctions in a system served to minimize the risk of failures like a bomb not reaching its target.48 Wilson’s initial approach to risk analysis mirrored the method’s prior military applications and his understanding of the challenges that nuclear technology posed for society. As he wrote in an early paper on risk analysis for toxic chemicals, physicists had long grappled with balancing the dangers of radiation exposure with the beneficial uses of x-rays and other peaceful applications of atomic physics.49 The principal task of the risk analyst, therefore, would be to “express numerically, insofar as possible, the risks and benefits which are likely to result from project outcomes.”50 Calculating the results required “scientific procedures or simulation models to estimate the likelihood of an accident and its probable consequences.”51 The value of these methods was ultimately to allow a decision maker to engage in a cost-benefit assessment of various alternative options52 and to “identify potential weak spots in technological systems.”53 When applied to toxic chemicals, however, this approach to quantification and risk modeling ran up against a number of problems that were not as fatal to risk analysis of weapons systems or radiation exposure, particularly the absence of adequate data on exposure and carcinogenic potential.54 The “first step” in Wilson’s method of risk assessment for environmental pollution required one to determine the “exposure to the carcinogen” and “carcinogenic effect at this exposure.”55 However, exposure data for the vast majority of environmental pollutants was 46 See William Boyd, Genealogies of Risk: Searching for Safety, 1930s-1970s, 39 ECOLOGY L.Q. 895, 900, 942 (2012) (describing the development of quantitative risk assessment during the 1970s and noting that at least until the 1970s the government’s decision-making process in environmental regulation was “up for grabs”). 47 Linda Nash, From Safety to Risk: The Cold War Contexts of American Environmental Policy, 29 J. POL’Y HIST. 1, 9 (2017); Boyd, supra note 46, at 900 n.8, 940–41. 48 See David Hounshell, The Medium is the Message, or How Context Matters: The RAND Corporation Builds an Economics of Innovation, 1946-1962, in SYSTEMS, EXPERTS, AND COMPUTERS: THE SYSTEMS APPROACH IN MANAGEMENT AND ENGINEERING, WORLD WAR II AND AFTER, 256 (Agatha C. Hughes & Thomas Hughes eds., 2011). For an example of the type of error analysis and other techniques from this period, see, e.g., KENNETH R. BRITTING, INERTIAL NAVIGATION SYSTEMS ANALYSIS 114–18 (1971). Note that such techniques were also useful for U.S. space exploration and civil aviation. Id. at vii. 49 Richard Wilson, Risk-Benefit Analysis for Toxic Chemicals, 4 ECOTOXICOLOGY & ENVIRONMENTAL SAFETY 370, 371 (1980). 50 ANDREW J. VAN HORN & RICHARD WILSON, THE STATUS OF RISK-BENEFIT ANALYSIS, INFORMAL REPORT: BROOKHAVEN NATIONAL LABORATORY 2 (Dec. 1976) (on file with author). 51 Id. 52 Id. at 5. 53 Id. at 9. 54 The challenge of adequate data collection for environmental modeling is not unique to toxic chemicals or risk benefit. See generally PAUL N. EDWARDS, A VAST MACHINE: COMPUTER MODELS, CLIMATE DATA, AND THE POLITICS OF GLOBAL WARMING (2010). 55 Van Horn & Wilson, supra note 50, at 11.

extremely scant during this period, which meant anyone using this method had to make her own assumptions or rely on data from industry, with all the attendant conflict of interest problems and uncertainties that posed. After calculating the risk of harm per exposure level, Wilson’s approach to risk analysis then involved comparing the risk to other contemporary threats, like cigarette smoking.56 Only if an environmental risk was higher than things like living in the same home as a smoker would regulation be necessary. These comparisons would inculcate industry against pollution regulations so long as other societal risks were higher, without considering whether those other risks should also be reduced through public health measures.57 In contrast to Wilson’s method for assessing risk, other scientists in this period offered a different approach that placed much less emphasis on articulating risk of harm as discrete numbers in a probabilistic sense. Those trained in biology, medicine, public health, and the environmental sciences relied on a more interdisciplinary assessment process that drew on a variety of research studies and analyses. Focusing on the process of carcinogenesis – the development of cancer – these scientists argued that epidemiological studies of populations, animal testing, studies of chemical structure, and molecular research on DNA damage and cellular changes could all provide indications that a substance posed a risk of cancer.58 To determine whether a chemical was carcinogenic would thus require a process that incorporated multiple ways of observing harmful biologic changes.59 For instance, toxicologists might expose animals to high levels of a chemical; if there was a “positive oncogenic effect” on the tested animals, this could provide sufficient evidence to conclude that the substance posed a health risk and should be regulated.60 During the 1970s, the more holistic method of assessing and understanding risk gained ground among top government scientists and regulators responsible for controlling environmental pollutants at the Environmental Protection Agency (EPA) and Occupational Safety and Health Administration (OSHA), among other agencies.61 These experts and agency officials were not opposed to quantifying risk probabilistically if there was sufficient data from epidemiological studies to develop a dose-response model.62 For instance, in 1978 the Department of Health, Education and Welfare conducted a large-scale study to estimate the relationship between toxic chemicals and cancer risk, with the aim of determining the probability that someone diagnosed with cancer developed the disease from an occupational exposure.63 However, such analyses could 56

Id. See also WILSON & CROUCH, RISK-BENEFIT ANALYSIS, supra note 7, at 211–22. Van Horn & Wilson, supra note 50, at 59 (describing an example of this calculation for vinyl chloride using data from the Society of the Plastic Industry). 58 INTERAGENCY REGULATORY LIAISON GROUP, SCIENTIFIC BASES FOR IDENTIFICATION OF POTENTIAL CARCINOGENS AND ESTIMATION OF RISKS 246, 258 (1979) (“The judgment that a substance poses a carcinogenic hazard derives from the evaluation of the total evidence provided by all of the sources. Different data sources may not contribute equally to the cumulative evaluation, depending on the specific nature and extent of the data, the scientific quality of the studies, and the adequacy of their documentation.”). 59 Id. at 258. 60 See, e.g., EDF, INC. v. EPA, 548 F.2d 998 (D.C. Cir. 1976) (describing the EPA’s reliance on animal testing to assess carcinogenicity for certain pesticides). 61 Interagency Regulatory Liaison Group, supra note 58, at 258. 62 They were, however, cautious about over reliance on them in the regulatory process. Id. at 259. 63 See generally NAT’L INST. OF OCCUPATIONAL SAFETY AND HEALTH, ESTIMATES OF THE FRACTION OF CANCER IN THE UNITED STATES RELATED TO OCCUPATIONAL FACTORS (1978). See also 44 Fed. Reg. 8210–11 (1979) (describing one of the first attempts to use probabilistic risk assessment to evaluate alternative ozone standards under 57

not provide a discrete numerical probability of developing cancer from exposure to a particular carcinogen. Nor could a holistic approach to assessing cause and effect between pollution exposure and health or environmental damage utilize quantification in the way that Wilson and other physicists desired.64 Given the challenges with limited data, leading environmental health experts argued that “[q]uantitative risk assessment is a premature science fostered by pressures to express public health hazards in economically simplistic terms.”65 Changes in certain assumptions could lead to wildly different estimates, and the “narrowness” of the approach – looking a single chemical in isolation – did not reflect the way that exposure usually occurred, with multiple other toxic substances acting in concert on the body.66 And even when sufficient data existed to develop dose-response models, probabilistic assessments of risk required extrapolating from these studies in ways that could be problematic because of differences between the studied group and the rest of the population. Age, sex, preexisting diseases, and variations in genetic susceptibility and life stage complicated simplistic comparisons along the lines Wilson and others suggested.67 Over the coming decades, environmental health scientists would continue developing methods to assess risk that included research less susceptible to quantification via probabilistic calculations.68 However, the approach favored by physicists like Wilson would come to dominate understandings of environmental risk within the judiciary. By advancing arguments that quantification was the only way to provide sufficient evidence to regulate pollution or hold industry liable for its harms, Wilson helped shape legal conceptions of adequate scientific evidence to conform to industry preferences.

III.

The Influence of Physicists on Judicial Understandings of Environmental Risk

As industries sought to contest environmental regulations and liability for pollution, physicists like Wilson and his methodology of risk analysis offered a patina of scientific credibility to their legal claims. Wilson became a highly sought after industry expert on these issues, and over the Clean Air Act and acknowledging that such an approach was incomplete and still in the early stages of development). 64 On the problems this holistic approach to causation poses for probabilistic risk assessment, see David Egilman & Samantha Howe, Against Anti-Health Epidemiology: Corporate Obstruction of Public Health via Manipulation of Epidemiology, 13 INTERNATIONAL J. OCCUPATIONAL & ENVIRONMENTAL HEALTH 118 (2007). 65 See, e.g., Samuel Epstein, Cancer Inflation and the Failure to Regulate, MIT TECHNOLOGY REVIEW 47–49 (Dec./Jan. 1980) (discussing the significant problems posed by the fact that industry controlled much of the information necessary to do probabilistic risk assessments, whether exposure data or the types of chemicals in certain facilities). Epstein was a professor of occupational and environmental medicine at School of Public Health, University of Illinois Medical Center; he had decades of experience working toxic chemical pollution. 66 Mary H. O’Brien, Alternatives to Risk Assessment: The Example of Dioxin, 3 NEW SOLUT 39, 39 (1993). 67 See Interagency Regulatory Liaison Group, supra note 58, at 263–64. 68 See Sheldon Krimsky, The Weight of Scientific Evidence in Policy and Law, 95 AM J. PUBLIC HEALTH S129 (2005) (describing the development of this approach, eventually called a “weight of the evidence” method).

the next forty years he appeared on their behalf at numerous Congressional hearings and agency rulemakings concerning pollution regulations.69 He also submitted amicus briefs in litigation in support of their positions, served as a spokesperson at media events,70 and published numerous papers that supported industry views on the minimal hazards from current levels of environmental pollution.71 The most important, lasting victories from these efforts would come through altering how judges assessed scientific evidence when reviewing agency pollution regulations and admitting expert testimony in tort suits. Before examining the ways that Wilson sought to influence environmental law on behalf of polluters, it is worth pausing to grapple with how much of he may have deliberately bent his work in risk analysis to serve their interests. This question is a complicated one, as Wilson’s scientific positions did not always map neatly onto industry’s priorities. For instance, contrary to other industry-funded scientists,72 he accepted that there was a linear relationship between toxic exposures and harms at low doses.73 He also appeared to believe that risk analysis could empower workers, privately suggesting to the petroleum and chemical industries that labor unions would be able to bargain for restrictive controls if they received more information about the relative harms associated with workplace exposures.74 At times, he characterized his political views as leaning left, calling himself the “token liberal” in one right wing organization.75 And his methodology was drawn from accepted approaches within fields like systems analysis. While it is therefore difficult to know how much Wilson was consciously engaged in a form of scientific subterfuge, a few things are evident from examining his work and advocacy. One is that Wilson never conducted any direct research relevant to the question of how pollution might harm human health and the environment.76 His primary scientific output continued to be in physics, with his publications on environmental risk analysis almost always appearing in popular outlets, conference proceedings, and journals specifically created for the emerging field, some of which were later accused of lacking any credible peer review process and agreeing to publish 69 For instance, utility companies hired Wilson to serve as a spokesperson for nuclear power during the mid1970s, and he travelled the country providing testimony about the safety of fission, helping to defeat at least one ballot initiative that would have placed a moratorium on nuclear power plants. Scott A. Kripe, Professor Advocates Nuclear Power: Wilson Promotes Reactors for Utilities, THE HARVARD CRIMSON (Dec. 2, 1976) (on file with author). 70 See, e.g., PVC Safety Group Communications Committee: Minutes (Mar. 9, 1978) (UCSF Industry Documents Library) (on file with author) (listing Wilson as one of two experts at a press conference to combat vinyl chloride regulations and develop “PR support” for the industry’s position). 71 Wilson published hundreds of papers in his life; his memoir lists over 900. While an exhaustive review of every publication is beyond the scope of this paper, none of the works this author read, nor any cited herein, disclosed Wilson’s funding and connections with industry. 72 See, e.g., Edward J. Calabrese, Hormesis: A Revolution in Toxicology, Risk Assessment and Medicine, 5 EMBO REPORTS S37 (2004). 73 See Richard Wilson, Risks Caused by Low Levels of Pollution, 51 YALE J BIOL MED 37, 42–43 (1978) (supporting the use of a linear dose-response model for chemical carcinogenesis given data on radiation carcinogenesis). 74 This suggestion could, of course, have been a cynical ploy to stave off regulations, but it appears to have originated with Wilson rather than with industry. Richard Wilson, Notes on OSHA Hearing on Occupation Carcinogens 1–3 (1978) (on file with author). 75 WILSON, PHYSICS IS FUN, supra note 21, at 516. 76 Id. at 551–96 (listing Wilson’s research publications along with opinion pieces, conference proceedings, and other essays).

articles favorable to industries in exchange for payment.77 Despite his lack of training in relevant fields, Wilson positioned himself as an authority on how to translate environmental science into legal principles for pollution regulations and industry liability – a claim to expertise that is at odds with how the vast majority of scientists understand appropriate qualifications for scientific advising.78 Indeed, Wilson’s scientific representations in several major legal controversies over pollution were at best willfully ignorant of research in the environmental health sciences and at worst deliberately obfuscated many issues. Perhaps most egregiously, Wilson espoused wildly different views about what metrics courts should use to judge the adequacy of scientific evidence for regulations versus tort suits. He advocated for raising the evidentiary burden on government agencies that issued pollution regulations while simultaneously making it harder for tort plaintiffs to hold industries liable, and in both instances, he successfully persuaded the courts to incorporate his views of what constituted reliable science. Wilson’s attempt to influence judicial assessments of environmental and public health research first took shape in response to federal regulations concerning toxic chemical exposures. The potential for probabilistic risk analysis to stem pollution regulations was immediately apparent to physicists employed by coal, petroleum, and chemical companies.79 As agencies like EPA and OSHA began the process of issuing toxic chemical controls in the 1970s, industry representatives asked Wilson to conduct quantitative, probabilistic risk assessments to demonstrate that it was unnecessary to lower exposure amounts to chemicals like asbestos, vinyl chloride, and benzene.80 Quantification of pollution risks allowed for seemingly straightforward comparisons between harms from pollution and other risks that citizens appeared to accept as unworthy of government regulation. For example, in Wilson’s earliest work on behalf of the chemical industry, he testified against further regulation of vinyl chloride by analogizing between the supposed yearly cancer risk from exposure to the chemical and exposure to secondhand smoke for 100 hours, eating

See Special Contributions: Correspondence about Publication Ethics and Regulatory Toxicology and Pharmacology, 9 INTERNATIONAL J. OCCUPATIONAL & ENVIRONMENTAL HEALTH 386, 386–91 (2003). 78 See Torbjørn Gundersen, Scientists as Experts: A Distinct Role?, 69 STUDIES IN HISTORY & PHILOSOPHY OF SCIENCE 52, 57–58 (2018) (“Scientists who are selected to contribute as experts should be able to decide whether the task lies within the scope of their area of expertise. This requires a kind of meta-expertise of being able to recognize the limitations of one’s expertise . . . The role of experts requires that scientists understand their contribution both in regard to the rest of the scientific community and in regard to how their knowledge will be understood and interpreted.”). 79 See, e.g., Chauncey Starr, Richard Rudman & Chris Whipple, Philosophical Basis for Risk Analysis, 1 ANNUAL REVIEW OF ENVIRONMENT & RESOURCES 629 (1976) (Starr and the other scientists worked for the Electric Power Research Institute, the main scientific arm of the coal industry at the time); see also Chauncey Starr, Social Benefits vs. Technological Risk, 165 SCIENCE 1232 (1969). 80 See, e.g., Richard Wilson, Risk-Benefit Analysis for Vinyl Chloride (1976) (on file with author) (stating that it was “written for the Society of the Plastics Industry and read in an abbreviated version at the centennial meeting of the American Chemical Society”); Letter from Richard Fleming, Vice President, Air Products and Chemicals, Inc., to Donald R. Goodwin, Emissions Standards and Engineering Division, EPA (Sep. 22, 1977), Attachment: Supplementary Comments, at 1 (on file with author) (describing Wilson’s assistance in opposing EPA’s proposed reductions in vinyl chloride emissions).

“two tablespoons of peanut butter” and “drinking one half pint of municipal drinking water.”81 Through these comparisons to other “ordinary” risks, Wilson argued that further restrictions on vinyl chloride would eliminate “de minimis” exposures that posed no real threat.82 He characterized environmental and public health researchers as wanting to “eliminate all known risks and force this by law,”83 and positioned his methodology in opposition to the “technology forcing” approach Congress adopted in statutes like the Clean Air Act.84 Probabilistic risk assessment, in his view, would provide a more rational, balanced approach to pollution regulations.85 Environmental health scientists viewed these efforts to make probabilistic risk assessment the primary decision-making tool for regulators as inconsistent with current research on toxic chemicals and an attempt to undemocratically impose industry preferences on the population. As one epidemiologist put it when debating Wilson at a conference in 1980, “[t]hose who . . . lost the struggle for the congressional conscience in the beginning of this decade, now offer such analyses as risk/benefit analysis as a last-ditch effort, labeling it scientific — ergo, good — and they directly or indirectly state that the earlier judgments were not a decision-making process at all or were not scientific — ergo, bad.”86 Wilson’s scientific contemporaries were also critical of the way he and other physicists would minimize the harms from pollution and overstate the costs of regulating without giving any “hint of their often close and pre-existing relationship with the client industry.”87 Initially, many courts resisted arguments that agencies needed to quantify risk of harm in probabilistic terms in order to justify pollution regulations. Some judges were openly wary of reifying “one technique of risk assessment” as acceptable given the rapidly changing pace of scientific investigations into pollution hazards.88 In decisions throughout the 1970s, courts found a variety of scientific studies could provide grounds for regulating a hazardous substance, such as 81

The Society of the Plastics Industry, VCM-PVC Management Summary: On the Legislative Front 1 (undated, likely 1977) (on file with author) (describing Wilson’s work as an “industry consultant” testifying against restrictions on vinyl chloride). 82 Id. at 2–7. 83 Richard Wilson, Analyzing the Risks of Daily Life, MIT TECHNOLOGY REVIEW (Feb. 1979) at 43. 84 Richard Wilson, Traps and Errors in Risk Analysis, Lecture at George Washington University Mar. 31, 1982 at 11 (on file with author) (“In setting the air quality standards for moving sources (automobiles), Congress demanded a technology forcing approach . . . Is this technology forcing procedure the best way for technology improvement? Congress, after this success [of the 1970s], thinks so. If we feel otherwise we must find a better way . . .”). 85 Wilson was not alone in this attempt to paint environmental health research as “unscientific.” See PAUL F. DEISLER, JR. & RICHARD C. SCHWING, HISTORY OF THE SOCIETY FOR RISK ANALYSIS THROUGH THE YEAR 2000 11 (2020) (describing OSHA’s “unscientific” cancer research as one of the motivating factors for creating the Society for Risk analysis). 86 See Transcript of Excerpts from Seminar on Risk Assessment, Acceptability and Management held Oct. 22– 23, 1980, Enclosed in Letter from Gilbert Gude, Congressional Research Service, to Mr. Sheldon Samuels, Industrial Union Department, AFL-CIO (June 3, 1981), at CRS-30 (on file with author) (UCSF Industry Documents Library). 87 See Epstein, supra note 65, at 46. 88 See, e.g., EDF v. EPA, 598 F.2d 62, 84 (D.C. Cir. 1978) (“We are currently in a period of rapid change in assessing and regulating toxic substances. TSCA has established new regulatory mechanisms that may well lead to wholesale estimation of the risks of toxic substances. A holding from this court that only one technique of risk assessment (such as "extensive studies on comparative mammalian toxicity", Brief of Industry Petitioners at 53) is acceptable to the detriment of others would unnecessarily inhibit EPA from evaluating new approaches and formulating appropriate policy.”).

reports from laboratory experiments,89 or medical reports of disease resulting from exposure to the hazard.90 This judicial resistance to demanding quantification was not exclusive to suits involving industry challenges to pollution regulations. Courts also deferred to agencies when they weighed a variety of scientific evidence but did not quantify harms even when environmental groups challenged the agency decision.91 Judicial doctrine on this question changed after litigation over OSHA’s 1978 benzene regulations, which provided Wilson and his industry backers a huge legal coup in defining what type of evidence agencies would need to issue toxic chemical controls. While OSHA had developed quantitative risk assessments for other toxic chemical regulations issued that decade, the agency was unable to do so for benzene because of a lack of good exposure data from epidemiological studies and the failure to induce cancer from benzene exposure in animal studies.92 OSHA instead relied on what was known as a comparative cohort approach for assessing risk, which involved examining rates of cancer among workers exposed to benzene (and no other suspected carcinogens) with other industrial workers and the general population.93 Wilson developed a quantitative risk assessment on behalf of petroleum and chemical companies opposed to the 1978 benzene standards, which purported to show that reductions in benzene exposure would save very few lives. His calculations were not submitted into the rulemaking record, peer reviewed, or published.94 Wilson also made a number of contestable assumptions but did not provide any range of uncertainties in his calculations, creating “an entirely spurious impression of precision.”95 Despite these problems with his analysis, the Supreme Court approvingly cited Wilson’s risk assessment when striking down OSHA’s regulation in the 1980 benzene case.96 Justice Stevens, writing for a plurality of the court, concluded that OSHA needed to use probabilistic risk assessment to avoid a potential delegation problem with OSHA’s authorizing statute.97 He was sharply critical of OSHA’s decision to reject Wilson’s methodological approach, finding that 89

See, e.g., Ethyl Corp. v. EPA, 541 F.2d 1, 17 n.30 (D.C. Cir. 1976); EDF, INC. v. EPA, 548 F.2d 998, 1007– 08 (D.C. Cir. 1976). 90 See, e.g., Res. Mining Co. v. Envtl. Prot. Agency, 514 F.2d 492, 508 n. 26, 512 (8th Cir. 1975). 91 See, e.g., Sierra Club v. Morton, 510 F.2d 813, 821 (5th Cir. 1975) (“Interior's decision to project possible environmental damage from all tracts by the matrix approach, as opposed to the use of a more detailed analysis for a few select sensitive points, certainly does not evince a lack of good faith effort to afford the decisionmaker with the necessary quantitative information concerning the potential impact of oil spillage. Because no exact data exists until a spill occurs at a given location, any analysis of future oil spillage involves a degree of speculation. Therefore, every attempt to select quantitative values will be to some extent arbitrary.”). 92 In fact, benzene is a rare example of a toxic chemical for which there was solid epidemiological evidence of carcinogenicity but no useful animal studies; it was only decades later that scientists were able to demonstrate benzene’s leukemogenic potential through experiments on animals. 93 Rachel Rothschild, Origins of the Major Question Doctrine, 100 IND. L. J. (forthcoming 2024). 94 Id. 95 Reply Brief for the Federal Parties at 2a, Indus. Union Dep't, AFL-CIO v. Am. Petroleum Inst., 448 U.S. 607 (1980) (Nos. 78-911, 78-1036). See also Daniel Wartenberg & Caron Chess, The Risk Wars: Assessing Risk Assessment, 3 NEW SOLUTIONS 16 (1993) (describing how the use of specific numbers provided the illusion of certainty and accuracy). 96 Indus. Union Dep't, AFL-CIO v. API, 448 U.S. 607, n. 38. (1980) 97 Id. at. 642–46.

Wilson had successfully demonstrated its usefulness for assessing the rationality of a regulation.98 Nothing in the opinion seemed to leave open the possibility of using alternative scientific evidence or methods for determining whether there was serious harm; the only leeway Justice Stevens appeared to give the agency was an acknowledgement that the calculations need not serve “as a mathematical straightjacket.”99 The Supreme Court’s benzene decision was a decisive turning point in legitimizing probabilistic, quantitative risk assessment along the lines Wilson had outlined.100 While there have been many subsequent critiques of the methodology and efforts to replace it, to this day environmental and public health agencies are loathe to issue toxic chemical regulations without quantifying the probability of harm for fear that the rules will be struck down in the courts.101 Wilson achieved a similar victory for industry by raising the bar for admission of expert evidence in tort litigation. He became involved in this work during the 1980s at the invitation of physicist Fred Seitz, who spent a large part of his career casting doubt on scientific research connecting greenhouse gas emissions to climate change while serving as a paid consultant to fossil fuel companies.102 Wilson first met Seitz through an organization called Scientists and Engineers for Secure Energy, which sought to promote nuclear power.103 In the course of looking for legal representation to contest the shuttering of nuclear power plants, Seitz connected with the Atlantic Legal Foundation, a conservative litigation group founded in the late 1970s to advance corporate interests, including opposition to environmental regulation.104 Seitz subsequently invited Wilson to join the Foundation’s Science Advisory Board in 1987 with the aim of reducing the use of “junk science” in tort suits.105

Id. at 654. Id. at 655. 100 Rothschild, Origins of the Major Questions Doctrine, supra note 93. Wilson felt the benzene case, along with several other U.S. Court of Appeals decisions from the late 1970s and early 1980s that required other agencies to use “risk analysis” before regulating, had vindicated his probabilistic, quantitative approach. See Wilson, Traps and Errors in Risk Analysis, supra note 84, at 12–14. 101 Rothschild, Origins of the Major Questions Doctrine, supra note 93. 102 WILSON, PHYSICS IS FUN, supra note 21, at 516. Seitz also played a prominent role in casting doubt on the connection between cigarette smoke and lung cancer while serving as a paid scientific advisor for tobacco companies. On Seitz’s work for the tobacco and fossil fuel industries, see Oreskes and Conway, supra note 11, at 5–8, 33, 57–64, 170–208. 103 Id. at 515. 104 The organization was originally called the “Mid-Atlantic Legal Foundation” but later changed its name. On its connection to corporate interests, see Oliver Houk, With Charity for All, 93 YALE L. J. 1415, 1456 (1984). On its creation along with other right-oriented legal groups, see ANN SOUTHWORTH, LAWYERS OF THE RIGHT: PROFESSIONALIZING THE CONSERVATIVE COALITION 15, 29, 197 n. 17 (2009). Some of the foundation’s earliest cases involved litigation over the promotion and operation of nuclear power plants. See Constructors Asso. of W. Pa. v. Kreps, 573 F.2d 811 (3d Cir. 1978); Nat. Res. Def. Council, Inc. v. United States Nuclear Regulatory Com., 582 F.2d 166 (2d Cir. 1978); Westinghouse Elec. Corp. v. United States Nuclear Regulatory Com., 598 F.2d 759 (3d Cir. 1979). They also took up other conservative causes like the use of race in hiring decisions. See, e.g., Chmill v. Pittsburgh, 488 Pa. 470, 412 A.2d 860 (1980). 105 WILSON, PHYSICS IS FUN, supra note 21, at 516. The Foundation had begun intervening in products liability suits during the 1980s as business groups pushed to limit mass tort litigation. On the Atlantic Foundation’s earliest such work in suits over exposure to the drug diethylstilbestrol (DES), which causes birth defects, see Houk, supra note 104, at 1497. 99

The case of Daubert v. Merrell Dow Pharmaceuticals offered Wilson and the Atlantic Legal Foundation its first major opportunity to shape how courts understood scientific evidence to better serve industry interests. The litigation arose over the drug bendectin, which many women took throughout the 1950s and 1960s to treat nausea during their pregnancies.106 Merrell Dow Pharmaceuticals, which manufactured bendectin, came under scrutiny in the 1970s after it pulled another one of its drugs, thalidomide, from the market following reports that it caused horrific birth defects.107 The doctor who discovered the connection between thalidomide and birth defects became a sought-after plaintiffs’ expert in the years thereafter, and began looking into a possible similar issue with bendectin on behalf of families whose children suffered birth defects.108 While Merrell Dow stopped selling bendectin in 1983, it faced an onslaught of litigation in the coming years alleging the drug had caused birth defects just like thalidomide.109 In the bendectin lawsuit that reached the Supreme Court, the plaintiffs’ experts sought to use animal studies, chemical analyses and a “reanalysis” of epidemiological studies to demonstrate that bendectin could lead to birth defects.110 However, the U.S. Court of Appeals for the 9th Circuit excluded the latter evidence because the reanalysis had not been peer reviewed.111 They then granted summary judgment to Merrell Dow, reasoning that the animal studies and chemical analyses were insufficient to show that ingesting bendectin caused the birth defects.112 The plaintiffs appealed to the Supreme Court in 1991, arguing that the 9th Circuit had incorrectly excluded the reanalysis in light of amendments to the federal rules of evidence the previous decade.113 The lawsuit attracted the attention of major business groups concerned that the Supreme Court could make it too easy for plaintiffs to introduce expert testimony.114 It was the first time the Court had ever addressed what standards judges should use in admitting scientific evidence,115 and the importance of the case prompted the Atlantic Legal Foundation’s President to ask Wilson to draft an amicus brief laying out how judges should decide whether to admit expert testimony.116 With an eye towards limiting the admissibility of plaintiffs’ experts, the brief argued that judges 106

JENNIFER VANDERBES, WONDER DRUG: THE SECRET HISTORY OF THALIDOMIDE IN AMERICA AND ITS HIDDEN VICTIMS 76–78 (2023). 107 Joseph Sanders, The Bendectin Litigation: A Case Study in the Life Cycle of Mass Torts, 43 HASTINGS L. J. 303, 316 (1992). 108 TRENT STEPHENS & ROCK BRYNNER, DARK REMEDY: THE IMPACT OF THALIDOMIDE AND ITS REVIVAL AS A VITAL MEDICINE 197–99 (2009). 109 Sanders, supra note 107, at 319. 110 Daubert v. Merrell Dow Pharms., 951 F.2d 1128, 1129 (9th Cir. 1991). 111 Id. at 1131. 112 Id. 113 Id. 114 See David G. Savage, Justices to Rule on “Junk Science” Case: High Court: Ruling Will Address Whether Evidence Must Be Generally Accepted by Scientists before It Can Be Presented to a Jury, LOS ANGELES TIMES, Oct. 14, 1992, at SDB3 (noting that President Bush, Vice President Dan Quayle, and attorneys for business interests had denounced more lenient admission of expert testimony). 115 Savage, supra note 114. The Association of Trial Attorneys in America had only identified toxic torts as a distinct type of legal claim in the late 1970s. SHEILA JASANOFF, SCIENCE AT THE BAR: LAW, SCIENCE, AND TECHNOLOGY IN AMERICA 118 (1995). 116 According to Wilson, he and two attorneys with the Atlantic Legal Foundation drafted the brief in a few hours in the LaGuardia airport lounge. See WILSON, PHYSICS IS FUN, supra note 21, at 518.

should only allow expert testimony if it utilized a methodology that was widely accepted in the scientific community, with peer review and publication of the research serving as necessary (though not sufficient) grounds for admissibility.117 The brief also heavily criticized expert assessments that were only developed in the course of litigation, like the reanalysis at issue in Daubert, and asserted that courts should never admit such evidence.118 Wilson subsequently recruited a group of famous scientists to sign-on to the brief, most of whom specialized in the hard sciences and several of whom had a record of taking pro-industry positions on environmental issues, such as the dangers of asbestos exposure.119 It is striking to compare Wilson’s position on scientific evidence in Daubert with his use of probabilistic risk assessment methodology in the benzene case. As noted, he developed his analysis of benzene’s risks at the request of industry in order to oppose government standards; the analysis had not been peer reviewed, nor had it been published. Most of his scientific contemporaries at the time rejected the methodology. And yet, the Supreme Court ultimately imposed the method on OSHA, largely based on Wilson’s expert testimony. On first glance, it is difficult to reconcile Wilson’s stance on the kind of scientific evidence that industry could marshal to combat government regulation with his position on the evidence plaintiffs should be able to use in tort litigation. There is, of course, the possibility that Wilson was simply arguing for whatever benefitted his industry backers. But Wilson’s conceptions of what constituted “good science” might also explain the contradiction. As he later wrote, the closer a 117

Brief for Nicolaas Bloembergen, Erminio Costa, Dudley Herschbach, Jerome Karle, Arthur Langer, Wassily Leontief, Richard S. Lindzen, William N. Lipscomb, Donald B. Louria, John B. Little, A. Alan Moghissi, Brooke T. Mossman, Robert Nolan, Arno A. Penzias, Frederick Seitz, A. Frederick Spilhaus, Dimitrios Trichopoulos & Richard Wilson, as Amicus Curiae in Support of Respondents at 10–13, Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 590 (1993) (No. 92-102). 118 Id. at 24. 119 The majority of the scientists who signed on to the brief specialized in physics and chemistry; a number of the scientists had confirmed ties to industry, which they did not disclose, and/or adopted positions at odds with the scientific consensus on environmental pollution problems. Among the physicists, signatories included Fred Seitz, discussed supra, as well as Richard Lindzen, who has long denied the scientific consensus on human-induced climate change. See John J. Fialka, A Climate Change Dissenter Who Has Left His Mark on U.S. Policy, CLIMATEWIRE, July 6, 2011, https://subscriber.politicopro.com/article/eenews/2011/07/06/a-climate-change-dissenter-who-has-left-hismark-on-us-policy-159761. One of the scientists with an environmental health background, Dimitrious Trichopoulos at Harvard’s School of Public Health, later came under fire for not disclosing his ties to the chemical industry in work that claimed the toxic chemical dioxin was not linked to cancer in humans. See Lennart Hardell et al., Secret Ties to Industry and Conflicting Interests in Cancer Research, 50 AMERICAN J. INDUSTRIAL MEDICINE 227 (2007). Another signatory, environmental health professor Brooke Mossman, had recently faced severe criticism from her contemporaries for publishing an article in the New England Journal of Medicine that downplayed the harms from asbestos exposure without disclosing that she had worked as a consultant to asbestos companies. See Jock McCulloch & Geoffrey Tweedale, Shooting the Messenger: The Vilification of Irving J. Selikoff, 37 INTERNATIONAL J. HEALTH SERVICES 619 (2007) (describing the controversy and criticism of the New England Journal Medicine, which subsequently reviewed its policy on publication of authors who had conflicts of interest). A third professor with a background in environmental health, Alan Moghissi, also had ties to industry funding, including serving as the chair of a science advisory committee for a non-profit founded in 1993 whose stated purpose was to be a “new ally against ill-conceived environmental regulation”; it was created by a group of industry trade organizations, such as the Independent Petroleum Association of America. See Center for Media & Democracy, The Junkyard Dogs of Science, 5 PR WATCH 1, 3 (1998) (on file with author); FY 1996 DOE, EPA, and NOAA R&D Budget Authorizations, H. Comm. Sci., Sub. Comm. Energy and Environment, 104th Congress, 1st Sess., 30 (1995).

discipline was to an “art” than a “science,” the more hesitant judges should be about admitting such expert testimony.120 20th century physics presumably fell in the latter camp – if not representing a model of pure “science.”121 In that light, it may have appeared perfectly consistent to Wilson for judges to impose probabilistic, quantitative risk assessments on government agencies even if the method was not generally accepted and the risk assessments had not been peer reviewed, while rejecting other types of evidence in tort suits that failed to meet those criteria. Wilson’s views about what constituted good science ended up having a demonstrable influence on the Supreme Court’s decision in Daubert and subsequent cases on admissibility of expert evidence, with lasting ramifications for tort suits against polluters. When the Supreme Court issued its ruling in Daubert, it cited Wilson’s brief first in its discussion of the standard judges should use when making decisions on expert testimony.122 While not requiring peer review and publication, the Justices listed these criteria as extremely important factors in determining admissibility.123 They also added several other considerations, such as the testability of an expert’s method, rate of error, and, crucially, the “general acceptance” of a technique.124 Though the Supreme Court did not follow the brief’s recommendations to the letter, Wilson felt the decision was far better than he had anticipated, seemingly because the Justices introduced other potential avenues to disqualify plaintiff’s experts.125 After the Daubert case, Wilson and the Atlantic Legal Foundation sought to build on the decision through filing amicus briefs in other cases concerning admissibility of expert testimony. As he did in the Daubert case, Wilson organized and drafted an amicus brief for the 1997 Supreme Court case General Electric Co. v. Joiner,126 which concerned whether exposure to a toxic chemical known as polychlorinated biphenyl (PCB) had caused the plaintiff’s small cell lung cancer.127 While the authors of the amicus brief described themselves as independent experts with no financial stake in the litigation, several of the signatories had received money from the tobacco, petroleum, and/or chemical industries since the 1980s.128 The brief criticized the “reliability” of research connecting PCB exposure to lung cancer, particularly the use of animal studies, and argued that “the methodology used by the Plaintiffs’ experts” – known as a “weight of the

120

See Richard Wilson, Public Comment upon the OMB Peer-Review Standard for Range of Studies (Dec. 15, 2003) (on file with author) (stating, in discussing the Daubert case, that “the more the subject issue is closer to an art than a science, the more a judge must be cautious about accepting expert testimony”). 121 On the history of distinctions between pure and applied science in the American context, see Paul Lucier, The Origins of Pure and Applied Science in Gilded Age America, 103 ISIS 527 (2012). 122 The court began this discussion after finding that the federal rules of evidence superseded the Supreme Court’s prior Frye decision. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 590 (1993). 123 Id. at 593. 124 Id. at 594. 125 WILSON, PHYSICS IS FUN, supra note 21, at 518. 126 522 U.S. 136 (1997). 127 WILSON, PHYSICS IS FUN, supra note 21, at 518. 128 Co-signatories included Fred Seitz, discussed supra; Arthur Langer, who appears to have received money from tobacco and asbestos companies at least as far back in the 1980s, see, e.g., Contract Dr. Arthur Langer Contract.; Memorandum Concerning Contractual Matters Prepared By RJR [R.J. Reynolds] (on file with author) (UCSF Industry Documents Library), as well as Michael Gough, a biologist by training who became the director of the Cato Institute’s Director for Science and Risk Studies during the 1990s.

evidence” approach – was “not scientific.”129 Yet at this time, agencies like the EPA had concluded that these animal studies, along with other epidemiological research, indicated a probable connection between PCB exposure and lung cancer.130 In addition to ruling on the standard of review for overturning a district court judge’s admission of expert evidence131 – the main issue in Joiner – the Supreme Court majority opinion opted to directly address the use of a qualitative “weight of the evidence” approach and its scientific validity.132 The Justices affirmed the district court judge’s conclusion that “isolated studies of laboratory animals” were insufficient to demonstrate that PCB exposure caused cancer.133 While the majority opinion did not directly cite to Wilson’s brief, none of the others submitted in the case made a similar argument about the underlying scientific research, nor was this question necessary to address in order to dispose of the case. As Justice Stevens criticized in a concurrence, the majority’s decision to reject the “weight of the evidence” approach outright rather than remand the matter to the Court of Appeals removed a methodology for assessing environmental hazards that many scientists and government agencies deemed reliable.134 Wilson would later state that the legal briefs he drafted in Daubert and Joiner were his “most influential work because it helped to change the way science is looked at in the courtroom.”135 That does not appear to be an exaggeration. In the years after these cases, several empirical studies found that Daubert had an exclusionary effect on admission of expert evidence,136 which disproportionately affected plaintiffs’ experts.137 Rather than serving as a general guidepost, the opinion’s criteria became more of a checklist, prohibiting the admission of evidence that failed any one item. In almost all cases this worked to the advantage of corporate defendants138 given that plaintiffs bear the burden of establishing causation.139 And following the 129

Brief for Bruce N. Ames, Michael Gough, Arthur M. Langer, Rodney Nichols, Frederick Seitz, Arthur Canfield Upton & Richard Wilson in Support of Petitioners at 20, 522 U.S. 136 (1997) (No. 96-188). 130 See EPA, PCBs: Cancer Dose-Response Assessment and Application to Environmental Mixtures 44 (1996), https://www.epa.gov/sites/default/files/2015-10/documents/pcbs_cancer_dose-response_assessment_and_ application_to_environmental_mixtures.pdf. 131 Gen. Elec. Co. v. Joiner, 522 U.S. 136, 143 (1997). 132 Id. at 143–47. 133 CARL F. CRANOR, TOXIC TORTS: SCIENCE, LAW, AND THE POSSIBILITY OF JUSTICE 54 (2016). 134 Justice Stevens agreed that an “abuse of discretion” standard was the correct standard on appeal for decisions about admitting expert evidence, but dissented from the rest of the majority opinion. For his discussion of the majority’s assessment of the underlying scientific issues, see Joiner, 522 U.S. at 153 (“It is not intrinsically “unscientific” for experienced professionals to arrive at a conclusion by weighing all available scientific evidence— this is not the sort of “junk science” with which Daubert was concerned. After all, as Joiner points out, the Environmental Protection Agency (EPA) uses the same methodology to assess risks, albeit using a somewhat different threshold than that required in a trial.”). 135 Michael Greenberg & Karen Lowrie, Dick Wilson: Applying Uncommon Sense to Risks, 30 RISK ANALYSIS 3, 5 (2010). 136 Gary Edmond & David Mercer, Daubert and the Exclusionary Ethos: The Convergence of Corporate and Judicial Attitudes towards the Admissibility of Expert Evidence in Tort Litigation, 26 LAW & POLICY 231, 243–44 (2004). 137 Cranor, supra note 133, at 5, 24 (2016). 138 David Mercer, Weighty Knowledge: Hyper Expertise and the Vertical Integration of Expertise (HEVIE), in YEAR BOOK 2006 OF THE INSTITUTE FOR ADVANCED STUDIES ON SCIENCE, TECHNOLOGY & SOCIETY (Arno Bammé, Günter Getzinger, and Bernhard Wieser, eds., 2007) at 341. 139 Cranor, supra note 133, at 38–40.

Supreme Court’s decision in Joiner, the vast majority of lower court judges eschewed the “weight of the evidence” method for assessing causation and excluded all but a select few animal studies of toxic chemicals and many other types of toxicological research.140 In the ensuing years, Wilson never seemed to develop an awareness of the potential misuses or limits of his approach to risk analysis. He continued efforts to define what constituted reliable scientific research alongside the Atlantic Legal Foundation, pioneering the concept of needing “sound science” in both regulatory efforts to control pollution and in tort lawsuits against industry during George W. Bush’s presidency.141 Together, Wilson and the Foundation sought to discredit research connecting a variety of pollutants to harmful environmental and public health effects, with little regard for the state of scientific consensus on various topics.142 For example, they argued against any connection between asbestos exposure and cancer and between greenhouse gas emissions and climate change – issues on which the vast majority of experts disagreed.143 In what appears to be his last Congressional testimony in 2007, Wilson spoke in opposition to a bill that would ban asbestos without disclosing his relationship with industry, arguing that the proposal was not “based on a modern understanding of the cancer risk from various asbestos fiber-types” and that current risks from asbestos were negligible compared to others of modern life.144 As public health researchers pointed out, however, Wilson’s “arguments do not comport to the scientific opinions of the world’s major scientific bodies, but rely on a small selected group of papers by authors, only a few of which have conducted research or contributed to the valid scientific research on the health effects of asbestos.”145 Yet as late as 2013, when Wilson was then in his 80s, he

140

David E. Bernstein, The Misbegotten Judicial Resistance to the Daubert Revolution, 89 NOTRE DAME L. REV. 27, 46–48, 57 (2013) (describing the shift away from a “weight of the evidence” approach after Joiner and assessing a few outlier opinions, notably the First Circuit’s opinion in Milward v. Acuity Specialty Products Group, Inc.). 141 See Atlantic Legal Foundation, Science and the Law: Sound Science in the Courtroom (working webpage created in 2002 and last updated in 2008, retrieved through web archive and on file with author). Wilson created and maintained the website for the Atlantic Legal Foundation. 142 For example, Wilson helped craft an amicus brief in the Massachusetts v. EPA case on EPA regulation of greenhouse gas emissions that argued it was inappropriate to regulate emissions from motor vehicles because of the negligible effects on the problem. Brief Amicus Curiae of Ernest L. Daman, Tom A. Hendrickson Nathan H. Hurt, Klaus S. Lackner, Dennis K. McBride, A. Alan Moghissi, Harold W. Olsen, Paolo F. Ricci Peter P. Rogers and Richard Wilson In Support of Respondent, Massachusetts v. Environmental Protection Agency, 549 U.S. 497 (2007) (No. 051120). 143 Atlantic Legal Foundation, Science and the Law, supra note 141. On the issue of electromagnetic exposure and health harms, see World Health Organization, Radiation: Electromagnetic Fields, https://www.who.int/newsroom/questions-and-answers/item/radiation-electromagnetic-fields (last visited May 11, 2024); on the scientific consensus about asbestos and industry attempts to cast doubt on this research using scientists who had previously worked for the tobacco industry, see Xaver Baur & Arthur L. Frank, Ongoing Downplaying of the Carcinogenicity of Chrysotile Asbestos by Vested Interests, 16 J. OCCUPATIONAL MEDICINE & TOXICOLOGY 6 (2021); on a similar phenomenon with climate change, see Oreskes and Conway, supra note 11. 144 Asbestos: Still Lethal/Still Legal: The Need To Better Protect The Health Of American Workers And Their Families, Hearing Before S. Comm. on Health, Education, Labor, and Pensions, 110th Cong. 19 (2007). 145 Id. at 46 (responding to Wilson, Dr. Richard Lemen – an environmental health science, former Assistant Surgeon General of the Public Health Service and Acting Director of NIOSH, noted that “[s]ome of Wilson’s cited publications appear as commentaries in publications of dubious distinction, i.e., The Apocalyptics, some are not even scientifically peer reviewed publications, and some represent his own commentaries, while the rest are selected because they support Wilson’s own arguments”).

assisted the Atlantic Legal Foundation in filing amicus briefs supporting asbestos manufacturers in tort suits.146 Nor does it appear that any judges ever questioned Wilson’s credentials in environmental science and risk assessment,147 even in litigation where he offered testimony explicitly on behalf of industry, such as the benzene case. The judiciary’s failure to consider the possibility that university scientists could have financial conflicts of interest represents a serious lapse in evaluating the credibility of experts in litigation against agency regulations and tort suits against polluters.148 It allowed physicists like Wilson to impose their vision of what constituted reliable environmental science on regulatory agencies and tort plaintiffs without appropriate scrutiny.

IV.

Conclusion

The phenomenon of physicists serving as experts for industry in debates over environmental research exerted a significant influence over judicial conceptions of reliable science after the passage of environmental laws in the 1970s and the rise of toxic tort litigation in the 1980s. Harvard physicist Richard Wilson, who pioneered the method of probabilistic, quantitative risk assessments for environmental pollution, is a prime example of how the petroleum and chemical industries employed physicists to combat regulation of toxic chemicals and tort suits against polluters. By emphasizing the need to quantify harms to the exclusion of other types of research and even other criteria for assessing scientific reliability, Wilson and his industry backers contributed to a legal landscape where it is both difficult to restrict toxic chemical exposures and hold industry accountable for any resulting environmental and public health harms.

146

Brief of Richard Wilson, Ronald Gots, Arthur M. Langer, Robert P. Nolan, Emanuel Rubin, and James Watson, Strickland v. Union Carbide Corporation, et al., California Supreme Court (No. S212424) (Aug. 15, 2013). 147 Wilson and the Atlantic Legal Foundation repeatedly saw their briefs cited favorably by the Supreme Court and other judges. See, e.g., Kumho Tire Co. v. Carmichael, 526 U.S. 137, 150 (1999) (citing favorably to their brief on the issue of reliability in scientific testimony). 148 In each of the Atlantic Legal Foundation’s briefs, for instance, the scientists attest to the fact that they have financial stake in the litigation, but make no disclosures regarding any consulting or other financial ties to companies that may be affected by the decision.

SESSION II: OTHER INDUSTRIES? Farm Until It’s Gone: Industrial Animal Agriculture and the Limits of Law Douglas Kysar, Yale University Public Health Nuisances Catherine Sharkey, New York University Tragic Exposure, Mass Litigation and Regulatory Failure: Does It Have to Be This Way? Wendy Wagner, University of Texas Steve Gold, Rutgers University Thomas McGarity, University of Texas Evidence of Compliance Maggie Wittlin, Fordham University Discussant: Gregory Mark, DePaul University

(Back to main agenda)

THE 30 TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

FARM UNTIL IT’S GONE Industrial Animal Agriculture and the Limits of Law

Douglas A. Kysar*

Draft: May 29, 2024 Introduction ..............................................................................................................1 I. Chains of Supply ..................................................................................................4 II. A Planet Transformed .....................................................................................999 III. Corporate Protective Equipment .................. 27Error! Bookmark not defined. IV. “The Nature of the Farm” and “The Problem of Socialized Cost”..................40 Conclusion .............................................................................................................53

INTRODUCTION Marie Wood lived her entire 95 years in Shelby County, Indiana. Much of that time was spent on a seventeen-acre farm now owned by Marie’s daughter. Covered with loamy, high-lime soil, the property was dotted by a small log cabin, later updated with siding, indoor plumbing, and electricity. Farmland was devoted primarily to soy and corn production, but space was reserved for growing fruit and vegetables, and a barn yard housed a variety of animals, some raised for companionship and some for consumption. Despite having worked many other jobs during her life to supplement the family’s income, Marie’s obituary stressed that “[s]he loved the small farm where she raised milk cows, turkeys and chickens and wore many hats throughout her life, all requiring long hours of hard work to support two children; yet she never complained and met each hardship with the strength and determination of a survivor, always managing to find the joy in life and sharing that happiness with all who met her.”1 In addition to loving her small farm, Marie also loved to tell a certain joke about farming. In this joke, a reporter interviews a farmer who just experienced a tremendous windfall by winning

* I am grateful to Jennifer Bass, Vermont Law School ’24, for outstanding research assistance and to participants of the 30th Annual Clifford Symposium on Tort Law and Social Policy for helpful comments and conversations. 1

Marie Wood Sprong, Shelbynews.com, July 7, 2009.

the state lottery. The reporter asks the farmer what they will do with all this new wealth, to which the farmer replies, “Farm until it’s gone.” The joke lands because it reveals both the farmer’s devotion to their craft and a wry awareness that this commitment will not be rewarded by the market adequately to sustain it. In the terminology of economists, the farmer’s labor preference is “sticky,” but this term hardly captures the cultural and psychological nuances at work when someone is so committed to a way of life that they are willing to knowingly risk their economic survival to maintain it. A similar joke asks, “How do you make a million dollars farming? Start with two million.” Both jokes upend the logic that is supposed to govern human economic behavior. Rather than seeing farming as determined by economic incentives, the paramount value is stewardship of land and animals, whether or not the market supports it. Farmland in the United States is no longer stewarded by the Marie Woods of the world. When she was born in the early twentieth century, agriculture took place on small, diversified farms in rural areas. It was labor intensive, employing nearly half of the U.S. workforce, not to mention millions of service animals.2 Over the course of the century, though, “[t]he number of farms across the nation fell from nearly 7 million at its peak during the Great Depression to just over 2 million at the end of the century.”3 With this large decline in the number of farms came an accompanying growth in average farm acreage, even as the total land area devoted to agriculture in the United States remained relatively stable (Figure 1). Many factors drove these changes, including the rise of highly mechanized and specialized production practices, technological developments in animal and crop genetics, changes in fertilizer and pesticide manufacturing, growth in off-farm labor opportunities, and relaxation of agricultural antitrust and foreign trade policies. Stark as it is, the historical decline in farm numbers understates the full economic transformation that has taken place since the largest 3 percent of farms today account for nearly 52 percent of total farm income. Small family operations, which make up 88 percent of U.S. farms overall, receive only 19 percent of total farm income, requiring them to rely overwhelmingly on off-farm revenue sources to make ends meet,4 just as Marie Wood did to live the life she loved.

Carolyn Dimitri, Anne Effland, and Neilson Conklin, The 20th Century Transformation of Agriculture and Farm Policy, USDA Economic Research Service, Economic Information Bulletin No. 3, June 2005. 3

Dimitri C, Effland A (2020). From farming to food systems: the evolution of US agricultural production and policy into the 21st century. Renewable Agriculture and Food Systems 35, 391–406. https://doi.org/10.1017/S1742170518000522. 4

USDA, Economic Research Service, Farming and Farm Income, February 29, 2024, available at https://www.ers.usda.gov/data-products/ag-and-food-statistics-charting-the-essentials/farming-and-farm-income/.

Figure 1

The transformation of U.S. agriculture into a highly industrialized and concentrated system reverses the logic implied by the two farming jokes: The manner in which land is used and animals are raised now is dictated by the profit motive. Farming bends to market discipline, rather than trying to endure in spite of it. One important feature of the economic logic driving industrial agriculture today is the ability to externalize numerous social and environmental costs of production, including the longterm costs associated with using natural resources in unsustainable ways. Whereas the farmers depicted in the two jokes were willing to countenance economic failure in order to preserve indefinitely their working relationship with land and animals, the industrial agriculture system today is willing to sacrifice both natural and human capital for short-term economic gain, even at the system’s own long-term peril. “Farm until it’s gone” today does not refer to the loss of a farmer’s monetary windfall, but to the loss of soil health, water quality, aquifer and groundwater reserves, local air quality, global atmospheric stability, worker safety, consumer health, animal welfare, biodiversity, pathogen resistance, and a host of other social and environmental “goods,” all of which are harmed or endangered by the practices of modern industrial farming. This Article contends that no sector better illustrates the reversal of logic in contemporary farming than industrial animal agriculture. Part I provides a brief overview of transformations that have taken place in the livestock sector over the past several decades, leading to a system of meat production that would be unrecognizable to anyone who participated in animal farming during the 3

previous 11,000 years of livestock domestication.5 Part II summarizes a plethora of ways in which intensive meat production imposes harms onto consumers, workers, neighboring communities, animals, ecosystems, and the global atmosphere, among other recipients of industrial animal agriculture’s “negative externalities.”6 Part III offers a sampling of avenues by which the industry has largely avoided legal accountability for these many harms. Taken together, Parts I through III paint a picture of an industry dominated by a few multinational corporate behemoths that has become a destructive, planet-altering force in less than a century, yet has managed to escape all but the least consequential regulatory controls. Part IV turns to theory, using industrial animal agriculture and its impacts to revisit the teachings of Ronald Coase’s two landmark papers, The Nature of the Firm and The Problem of Social Cost. For Coase, firms arise when transaction costs in the market make internal production of some good cheaper than contracting for it externally. Likewise, transaction costs explain why some actors might suffer negative externalities even though they may be harmed more than the externalizer benefits. Part IV takes these insights further, arguing that firms can impose costs externally with the same deliberateness that they produce benefits internally. Indeed, the handful of companies that dominate global meat sales today appear to be organized with as much attention to externalizing risk and harm as to internalizing beneficial production. Moreover, firms do not face a given fixed landscape of “entitlements” and “transaction costs” that determines whether internal production is cheaper than contracting or whether victims of cost externalization are able to bargain for relief. Rather, the landscape of “entitlements” and “transaction costs” is something that can be managed by firms with the same profit-maximizing logic that drives other decisions. If the insights of Part IV are accurate, then it suggests that meaningful reform of a sector as powerful and pervasive as industrial animal agriculture will not happen through private arrangements, litigation, or regulation alone. What will be required instead is a moment of democratic self-understanding in which people perceive their powers of popular sovereignty to include constructing and shaping markets, rather than merely tinkering with the ill impacts of markets after-the-fact. Modern meat conglomerates already understand this possibility and have used it to construct a system of production that is as detrimental to society, nature, and the climate as it is insulated from accountability. Left unchecked, they will farm until we’re gone. I. CHAINS OF SUPPLY When several U.S. slaughterhouses closed or operated at reduced capacity during the COVID-19 pandemic, the meat industry was forced to “depopulate” millions of livestock animals that could no longer be processed and brought to market.7 In many instances, pigs and chickens 5

See J. Hartung, A Short History of Livestock Production, in LIVESTOCK HOUSING: MODERN MANAGEMENT TO ENSURE OPTIMAL HEALTH AND WELFARE OF FARM ANIMALS pp. 81–146 (Wageningen Academic Publishers: Wageningen, The Netherlands, 2013). 6

See Donald J. Boudreaux and Roger Meiners, Externality: Origins and Classifications, 59 Natural Resources Journal 1 (2019) (tracing the intellectual history of the economic concept of externalities). 7

“Depopulation” is defined as “a method by which large numbers of animals must be destroyed quickly and efficiently with as much consideration given to the welfare of the animals as practicable.” USDA’s Animal and

were killed by shutting down ventilation systems inside confined animal feeding operations (CAFOs), often with the addition of heat, until most animals eventually died from hyperthermia or suffocation. Workers later combed through facilities to identify still-living animals for targeted euthanasia. Those involved in these campaigns reported a non-trivial degree of psychological distress from the experience.8 While some cattle producers were able to let their stock pasture longer and eventually go to market at a heavier weight, swine and poultry producers had no such leeway. With animals strategically bred to have short growth cycles that take place within massive indoor facilities that house animals by the tens or even hundreds of thousands, pig and chicken farms simply could not cope with a processing backlog. More time alive for these animals would have meant more feed costs, more waste to contain, and even more intense and dangerous confinement conditions as new animals continued to arrive.9 The throughput of animal lives in the global meat supply chain is staggering. In the United States alone, nearly 10 billion land animals are slaughtered for food every year.10 Globally, an estimated 69 billion chickens, 1.5 billion pigs, 656 million turkeys, 574 million sheep, 479 million goats, and 302 million cattle were slaughtered for food in 2018.11 These numbers do not include the billions of additional animals whose lives are devoted to egg or dairy production, nor do they begin to account for the astonishing number of fish lives currently used within aquaculture.12 This expansion of animal agriculture has happened in rapid order. As shown in Figure 2, meat production worldwide has grown exponentially over the past half century (e.g., the number of chickens in global production increased from 5 to 20 billion between 1970 and 201013). No sign of slowing appears, with global meat consumption projected to rise a further 14 percent over the Plant Health Inspection Service, Foreign Animal Disease Preparedness and Response Plan/National Animal Health Emergency Management System Guidelines 1 (Aug. 2015), https://www.aphis.usda.gov/animal_health/emergency_management/downloads/nahems_guidelines/mass_depop_eu than.pdf. 8 9

See Matthew Scully, The Human Cost of ‘Culling’ Livestock and ‘Depopulating’ Farms, NAT. REV. (May 7, 2020), https://www.nationalreview.com/2020/05/coronavirus-pandemic-human-cost-of-culling-livestockdepopulating-farms/ (“The system has its own unbending schedules and logic. No sheltering in place for factoryfarmed pigs, cows, chickens, and other creatures when yet more troubles appear. When they can’t die on a kill line, because a slaughterhouse has closed, that just means they have to die somewhere else to get out of the way – even if, as in this case, they’re all bound for landfills, blast furnaces, or burial pits. . . . [W]ith the merest pause the meat system convulses with ‘backlog,’ requiring travails for which producers expect our sympathy.”). 10

Econ. Rsch. Serv., U.S. Dep’t of Agric., Livestock and Meat Domestic Data, Meat Statistics Tables, Historical, https://www.ers.usda.gov/data-products/livestock-and-meat-domestic-data/livestock-and-meat-domestic-data (Sept. 27, 2023) (summing total 2022 slaughter counts per animal). 11

Andrea J Garmyn, ‘More than meat: contributions of livestock systems beyond meat productions’ (2021) 11 Animal Frontiers 1, 1. 12

Becca Franks et al. ,Animal welfare risks of global aquaculture. Sci. Adv.7, eabg0677(2021). DOI:10.1126/sciadv.abg0677 (estimating 250 to 408 billion individual animals killed within the global aquaculture industry in 2018). 13

https://ourworldindata.org/meat-production

decade ending in 2030. Enhanced productivity may satisfy some of this increased demand, but overall “stocks of farmed animals for meat will increase during the next decade, rising 11%, 9%, 2% and 18% for poultry, pigs, beef cattle and sheep respectively.”14 Figure 2

In addition to its vastness, the animal agriculture industry also is remarkable for its consolidation. In the United States, the four largest processors of beef, pork, and chicken control 85, 70, and 54 percent of their markets, respectively.15 These percentages represent significant increases in concentration from just a few decades prior.16 Globally, major firms such as JBS, Smithfield, Tyson, and Cargill are pursuing similar control over meat and dairy markets worldwide.17 Such concentration gives animal agriculture firms significant influence over the supply chain and its policy context.

See The Organisation for Economic Cooperation and Development and the Food and Agriculture Organization, OECD-FAO Agricultural Outlook 2021-2030, (OECD Publishing 2021). Looking farther ahead to 2050, levelsof beef, pork, and chicken production are expected to rise by 39, 55, and 58%, respectively. FAO, “The future of food and agriculture—Alternative pathways to 2050” (Rome, 2018); www.fao.org/3/I8429EN/i8429en.pdf. 15

Claire Kelloway and Sarah Miller, ‘Food and Power: Addressing Monopolization in America’s Food System’, 27 March 2019, Open Markets Institute 1, 3-4. 16

Claire Kelloway and Sarah Miller, ‘Food and Power: Addressing Monopolization in America’s Food System’, 27 March 2019, Open Markets Institute 1, 3-4. 17

Heinrich Böll Stiftlung, Friends of the Earth Europe, and Bund, Meat Atlas: Facts and figures about the animals we eat (2021) 26-27.

Poultry. The poultry sector is illustrative. Nearly nine out of ten chickens slaughtered in the United States today are grown under a system called “contract farming,” in which poultry processors offer detailed contractual arrangements to farmers who commit to raising the companies’ chickens.18 Under these contracts, poultry processors specify a plethora of terms, including what chicks, feed, medicine, and additives farmers must use, the precise design and construction characteristics of CAFO facilities, extensive operational parameters, and the method for compensating farmers once birds reach slaughter weight. In most cases, price is not specified in advance but rather is subject to a controversial “tournament system” in which payment is determined by a comparison of any individual farmer’s output against the average of certain metrics of performance achieved by other growers during a given period.19 Capital costs associated with building and upgrading CAFO facilities are borne by growers, as are largely the risks of disease, supply chain disruption, chick population failure due to poor breeding stock, and many other foreseeable and unforeseeable losses during the growing period. The system just described is a form of vertical integration, but a distinctive one: Farmers provide the labor and own the land and facilities (generally subject to significant debt), while processors (or “integrators” as they are often called) provide chicks, feed, transportation, veterinary consultation, and other materials and services. The allocation of economic and other forms of risk is carefully calibrated, subject to the bargaining dynamics that determine who benefits from the calibrating. On that score, half of chicken farmers work today in areas dominated by only one or two processors,20 leaving them with little bargaining power to obtain better contracts. Processors also have been found to engage in punitive and anticompetitive behavior when farmers raise complaints about contract terms or mistreatment.21 As concluded by the Small Business Administration Inspector General, the market power of processers in the poultry sector is now so strong and the level of micromanagement of farmers so intricate that the latter should no longer be considered independent business entities.22 Pork. Swine production also illustrates the dramatic technological, social, and economic shifts that have taken place in animal agriculture. Once supplied by thousands of small farmers across the country, the number of pork producers in the United States fell by more than seventy 18

“Vertical Integration: What it is – and why it’s good for the chicken industry... and you,” National Chicken Council, accessed April 15, 2019, https://www.nationalchickencouncil.org/industry-issues/vertical-integration/. 19

“Over many years, the U.S. Department of Agriculture (USDA) has received numerous complaints from poultry growers about the use of tournament systems and many have suggested that USDA should ban, restrict, or condition the use of tournament systems or particular aspects of those systems.” https://www.federalregister.gov/documents/2022/06/08/2022-11998/poultry-growing-tournament-systems-fairnessand-related-concerns. 20

James M. MacDonald, “Technology, Organization, and Financial Performance in U.S. Broiler Production” (U.S. Department of Agriculture Economic Research Service, 2014), https://www.ers.usda.gov/publications/pubdetails/?pubid=43872. 21

Christopher Leonard, The Meat Racket: The Secret Takeover of America’s Food Business (New York: Simon & Schuster, 2014). 22

U.S. Small Business Administration Office of Inspector General, “Evaluation Report: Evaluation of SBA 7(A) Loans Made to Poultry Farmers” (2018), https://www.sba.gov/sites/default/files/oig/SBA-OIG-Report-18-13_0.pdf (Processors exercise “comprehensive control” over contract growers via “contractual mandates and restrictions, management agreements, operating procedures, oversight, inspections, and market controls.”).

percent between 1990 and 2021.23 Despite this decline, overall production continues to grow both in terms of the number of pigs slaughtered annually24 and the average weight of individual animals at death.25 Pork meat today overwhelmingly originates from large CAFOs located in a few key states,26 with the top forty producers owning about two-thirds of the total pork breeding stock.27 Pigs are confined indoors in high density conditions,28 typically with slatted floors over a pit structure to enable capture of the up to ten pounds of manure per day that a fully grown pig can generate. Use of production contracts similar to the poultry industry expanded significantly in the pork sector during the 2000s, such that by 2020 almost 75 percent of U.S. pork production occurred under such contracts.29 Beef and dairy. Like the poultry and swine industries, the beef and dairy sectors also have undergone dramatic changes over the past century. Today, the United States is the world’s largest producer and consumer of beef, with approximately 87 million cows in production at a given moment.30 The United States is also a major dairy producer, generating around 11 percent of the world’s milk output each year.31 The vast majority of U.S. beef, veal, and dairy cows are housed in CAFOs and other intensive facilities colloquially known as “factory farms.” In the “finishing” stage of beef production, these facilities increasingly take the form of large, open-air feedlots that can house more than 100,000 cows at a time. Although the beef supply chain is less dominated by contract farming than poultry and swine,32 the industry has become highly concentrated at the feedlot and processing stages. Today, the 5 percent of feedlots that house more than 1,000 animals finish 80 to 85 percent of all beef 23

Hogs & Pork: Sector at a Glance, Econ. Rsch. Serv., U.S. Dep’t of Agric., https://www.ers.usda.gov/topics/animal-products/hogs-pork/sector-at-a-glance/ (last visited May 10, 2024). 24

Nat’l Agric. Stat. Serv., U.S. Dep’t of Agric., 2017 Census of Agric. Summary and State Data 23 (2019), https://www.nass.usda.gov/Publications/AgCensus/2017/Full_Report/Volume_1,_Chapter_1_US/us v1.pdf. 25

Hannah Price, As Hogs Get Heavier, Might Pork Quality Suffer?, Nat’l Provisioner (Jan. 9, 2020), https://www.provisioneronline.com/articles/109382-as-hogs-get-heavier-might-pork-quality-suffer. 26

Hannah Ritchie, How Many Animals are Factory Farmed?, Our World in Data (Sept. 25, 2023), https://ourworldindata.org/how-many-animals-are-factory-farmed (over 98 percent of U.S. pork comes from pigs raised in factory farm conditions). 27

Terri Queck-Matzie, An Overview of Pork Production in the U.S., Successful Farming (Oct. 2, 2019), https://www.agriculture.com/livestock/hogs/an-overview-of-pork-production-in-the-us. 28

Christopher G. Davis et al., U.S. Dep’t of Agric., ERR-308, U.S. Hog Production: Rising Output and Changing Trends in Productivity Growth (2022), https://www.ers.usda.gov/webdocs/publications/104437/err-308.pdf (noting that average capacity per farm ranges from 474 pigs in the breeding phase to 2,800 pigs in the finishing phase). 29

Farm Structure and Contracting, Econ. Rsch. Serv., U.S. Dep’t of Agric., https://www.ers.usda.gov/topics/farmeconomy/farm-structure-and-organization/farm-structure-and-contracting/ (last visited May 10, 2024). 30

Nat’l Agric. Stat. Serv., U.S. Dep’t of Agric., Cattle 1 (2024), https://downloads.usda.library.cornell.edu/usdaesmis/files/h702q636h/6108x003v/kk91h696g/catl0124.pdf. 31

U.N. Food & Agric. Org., World Food and Agriculture: Statistical Yearbook tbl. 23 (2023), https://www.fao.org/3/cc8166en/cc8166en.pdf. 32

Farm Structure and Contracting, Econ. Rsch. Serv., U.S. Dep’t of Agric., https://www.ers.usda.gov/topics/farmeconomy/farm-structure-and-organization/farm-structure-and-contracting/ (Mar. 8, 2022).

cattle processed in the United States.33 The slaughtering stage is similarly concentrated: As early as 1997, 80 percent of U.S. cattle slaughter was handled by large plants that process more than 500,000 cattle annually, as compared to only 16 percent 20 years earlier. By 2018, just four companies – Cargill, Tyson Foods, JBS SA, and National Beef Packing – processed 85 percent of U.S. grain-fed cattle.34 This is a 60 percent increase in market share for the four largest beef packers between 1977 to 2018.35 Within dairy, consolidation has occurred at a rate that exceeds most other sectors of U.S. agriculture.36 Across three decades, the number of dairy farms in the country dropped by 75 percent even as total milk production increased by 50 percent.37 By the year 2017, more than half of all dairy cows were housed in large operations with more than 1,000 head, compared to only 9 percent just 25 years earlier.38 The largest dairy operation in the United States holds over 100,000 cows.39 Although dairy farmers historically organized into marketing cooperatives in order to effectively bargain with concentrated buyers, today the cooperatives themselves exhibit signs of concentration, with 83 percent of U.S. milk being sold through just three cooperatives.40 II. A PLANET TRANSFORMED The rise of industrial animal agriculture has reshaped the planet in ways that defy comprehension. Researchers estimate that 96 percent of mammalian biomass now consists of humans and animals devoted to human use and consumption. Put differently, of all mammals now alive on Earth, only 4 percent by weight are wild, while 36 percent are human and 60 percent are livestock, pets, and other instrumentalized forms of life (Figure 3). Among avian species, 70 percent of Earth’s biomass today consists of chickens and other poultry, while only 30 percent

Sector at a Glance: Cattle & Beef, Econ. Rsch. Serv., U.S. Dep’t of Agric., https://www.ers.usda.gov/topics/animal-products/cattle-beef/sector-at-a-glance (last visited July 27, 2023). 34

Explainer: How Four Big Companies Control the U.S. Beef Industry, Reuters (June 17, 2021), https://www.reuters.com/business/how-four-big-companies-control-us-beef-industry-2021-06-17. 35

Claire Kelloway & Sarah Miller, Open Markets Inst., Food and Power: Addressing Monopolization in America’s Food System 4 (2019), https://static1.squarespace.com/static/5e449c8c3ef68d752f3e70dc/t/614a2ebebf7d510debfd53f3/163225158 3273/200921_MonopolyFoodReport_endnote_v3.pdf. 36

James M. MacDonald et al., U.S. Dep’t of Agric., ERR-274, Consolidation in U.S. Dairy Farming 11 (2020), https://www.ers.usda.gov/webdocs/publications/98901/err-274.pdf. 37

Id.; U.S. Dep’t of Agriculture, Milk Production 18 (2023), https://downloads.usda.library.cornell.edu/usdaesmis/files/h989r321c/jh345531b/n8711359j/mkpr0223.pdf. 38

James M. MacDonald et al., U.S. Dep’t of Agric., ERR-274, Consolidation in U.S. Dairy Farming 1, 12 (2020), https://www.ers.usda.gov/webdocs/publications/98901/err-274.pdf. 39

https://ruminants.ceva.pro/dairy-production (listing RockView Dairy in Southern California as the third largest dairy farm in the world based on herd size). 40

Food & Water Watch, The Economic Cost of Food Monopolies: The Dirty Dairy Racket 5 (2023), https://www.foodandwaterwatch.org/wp-content/uploads/2023/01/RPT2_2301_EconomicCostofDairyWEB.pdf.

represents wild birds.41 These distributions signify massive declines in wild animal life from the time before intensive human agriculture and development.42 One researcher calculates that 85 percent of wild mammal terrestrial biomass has disappeared since the rise of humans, with much of the decline occurring after 1900 as the global human population increased by six billion and farming transformed into its current highly industrialized form and scale.43 In short, humanity has re-engineered the tree of life to suit its purposes. The billions of cattle, sheep, pigs, poultry, and other livestock animals that humanity brings into and out of existence each year hold jarring consequences for the planet. In substantial part to grow livestock feed crops, humans have pumped and redistributed enough groundwater from beneath the earth’s surface to have literally shifted the planet’s rotational pole by 31.5 inches between 1993 and 2010 alone.44 Our re-engineering of the tree of life has tilted the very planet it grows upon. As this Part summarizes, while altering the Earth’s axis may not have obvious or immediate welfare consequences, numerous other impacts of industrial animal agriculture do. Taken together, the negative externalities summarized in this Part suggest that, for industrial animal agriculture, the designations “external” and “internal” are misleading. That which is “external” may be more constitutive of the industry than that which is “internal.”

See Bar-On, Y.M., Phillips, R. and Milo, R., 2018. The biomass distribution on Earth. Proceedings of the National Academy of Sciences, 115(25), pp.6506-6511. 42

See Ceballos, G., Ehrlich, P.R. and Dirzo, R., 2017. Biological annihilation via the ongoing sixth mass extinction signaled by vertebrate population losses and declines. Proceedings of the national academy of sciences, 114(30), pp. E6089-E6096. 43

See “Wild mammals have declined by 85% since the rise of humans, but there is a possible future where they flourish” Published online at OurWorldInData.org. Retrieved from: 'https://ourworldindata.org/wild-mammaldecline' [Online Resource] 44

See Seo, K.W., Ryu, D., Eom, J., Jeon, T., Kim, J.S., Youm, K., Chen, J. and Wilson, C.R., 2023. Drift of Earth's pole confirms groundwater depletion as a significant contributor to global sea level rise 1993–2010. Geophysical Research Letters, 50(12), p.e2023GL103509.

Figure 3

Consumers. Worldwide, meat consumption per capita has grown significantly over the past six decades (Figure 4). A global average figure, however, obscures vast differences in national per capita meat consumption. The United States, for instance, consumes an average of 149 kilograms of meat per person per year, whereas in India the average is only 14 kilograms.45 Even a single country’s average consumption level can obscure important details: An estimated one half of all beef consumed in the United States on a given day is eaten by just 12 percent of the country’s

Food and Agriculture Organization of the United Nations, Food Balances (2023).

population.46 Remarkable as it is, that statistic may be even more notable for its consonance with the equivalent breakdown for U.S. alcohol consumption.47 Whether by design or happenstance, these two industries have managed to capture a loyal minority who consume in prodigious amounts, despite widespread awareness that excessive consumption of beef and alcohol carries numerous health risks. Figure 4

One need not eat meat at an alcoholic’s equivalent level of consumption to risk a variety of serious health concerns. Intake of meat above standard nutritional guidelines has been linked to an increased risk of cardiovascular diseases, type 2 diabetes, and certain cancers.48 Meat contains high levels of saturated fats and cholesterol which can contribute to plaque formation in the arteries. A diet high in meat and low in fruit, vegetables, and whole grains also has been linked to an increased risk of stroke. High levels of red and processed meat consumption in particular are

Willits-Smith, Amelia, Harmonii Odinga, Keelia O’Malley, and Donald Rose. 2023. “Demographic and Socioeconomic Correlates of Disproportionate Beef Consumption among US Adults in an Age of Global Warming” Nutrients 15, no. 17: 3795. https://doi.org/10.3390/nu15173795. 47

Philip J. Cook, Paying The Tab: The Costs and Benefits of Alcohol Control (the top 10 percent of U.S. alcohol consumers drink over half of all alcohol sold and consumed in the country). 48

Walker P, Rhubart-Berg P, McKenzie S, Kelling K, Lawrence RS. Public health implications of meat production and consumption. Public Health Nutr. 2005 Jun;8(4):348-56. doi: 10.1079/phn2005727. PMID: 15975179.

associated with increased mortality, including from colorectal and stomach cancer.49 Evidence even suggests that a diet high in processed meat leads to an increased risk of dementia.50 All told these and other health risks of high consumption of meat contribute to a noteworthy increase in premature death.51 Conversely, researchers estimate that a diet rich in plants conforming to standard nutritional guidelines would avoid an estimated five to eight million deaths per year worldwide.52 Alongside prematurely lost lives are the additional health care costs of treating those who are afflicted with cardiovascular diseases, diabetes, and other maladies associated with high meat consumption. The global health care costs associated with red and processed meat consumption alone were estimated at $285 billion for 2020.53 Public Health. The production and consumption of meat negatively impacts public health in ways other than diet and nutrition. Most notably, the emergence and spread of infectious diseases is made more likely by the practices of industrial animal agriculture. As environmental scientist Matthew Hayek explains, humanity now faces an “infectious disease trap” due to the way it produces meat.54 With more and more land being cleared for livestock and feed crop production, the likelihood of disease spread from wild to domesticated animal populations and ultimately to humans increases. Indeed, since 1940, an estimated half of all zoonotic disease emergence has been associated with agriculture.55 Significantly, much of the expansion of livestock and feed production is occurring in “tropical, developing countries, where 75% of deaths are attributable to infectious diseases, where the risk of disease emergence might be greatest, and where disease surveillance and access to health care, particularly for those infections that accompany extreme

Rashmi Sinha, et al., Meat Intake and Mortality: A Prospective Study of Over Half a Million People, 169 Arch Intern Med. 562 (2010); Wenming Shi, Red Meat Consumption, Cardiovascular Diseases, and Diabetes: A Systematic Review and Meta-Analysis, 44 Eur Heart J. 28 (2023); Mengying Wang, et al., Red Meat Consumption and all-Cause and Cardiovascular Mortality: Results From the UK Biobank Study, 61 Eur J Nutr. 5 (2022). 50

Tian-Shin Yeh, Deborah Blacker & Alberto Ascherio, To Meat or not to Meat? Processed Meat and Risk of Dementia, 114 Am J Clin Nutr 7 (2021) (discussing “a hazard ratio of 1.44 for each 25 g per day increase in processed meat intake, meaning that eating 1 more slice of ham, 2 more slices of bacon, or 1 more sausage link each day over an average 8 years of follow-up would increase the risk of dementia by more than 40%”) (citing Huifeng Zhang, et al., Meat Consumption and Risk of Incident Dementia: Cohort Study Of 493,888 UK Biobank Participants, 114 Am J Clin Nutr 1 (2021)). 51

Fahimeh Haghighatdoost, et al., Differences in All-Cause Mortality Risk Associated with Animal and Plant Dietary Protein Sources Consumption, 13 Sci. Reports 3396 (2023) (“[O]ur results suggest an independent positive association for animal protein and all-cause mortality.”). 52

Marco Springmann et al., ‘Analysis and valuation of the health and climate change cobenefits of dietary change’ (2016) 113 Proceedings of the National Academy of Sciences of the United States of America 4146, 4147. 53

Springmann M, Mason-D'Croz D, Robinson S, Wiebe K, Godfray HCJ, Rayner M, Scarborough P. Healthmotivated taxes on red and processed meat: A modelling study on optimal tax levels and associated health impacts. PLoS One. 2018 Nov 6;13(11):e0204139. doi: 10.1371/journal.pone.0204139. PMID: 30399152; PMCID: PMC6219766. 54

Matthew N. Hayek, The Infectious Disease Trap of Animal Agriculture, Sci Adv. (Nov. 2, 2022), https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9629715/. 55

Loh Zambrana-Torrelio, et Al., Targeting Transmission Pathways for Emerging Zoonotic Disease Surveillance and Control, 15 Vector Borne Zoonotic Dis. 15, 432–437 (2015).

poverty, are most limited.”56 Overall, the rate of new emerging infectious diseases, most of them zoonotic in origin, increased significantly over the latter half of the twentieth century,57 a period of time that coincides with the dramatic expansion and transformation of meat production. On the other hand, if land clearing is to be reduced through intensification of animal agriculture production, as is typically recommended, then other disease risks will be enhanced. This is because intensification of animal agriculture “entails the concentration and confinement of animal bodies and their wastes, trading off deforestation for other multiple well-documented and potentially cascading risks for zoonotic disease emergence.”58 One important such risk is the spread of antimicrobial resistance,59 as intense confinement necessitates the heavy use of antibiotics and other antimicrobial agents to prevent and treat disease. Approximately 66 percent of medically important antibiotics in the United States are used for animal agriculture.60 The World Health Organization considers antimicrobial resistance to be one of the greatest threats to global health and development, linked to millions of deaths per year.61 According to the World Bank, antimicrobial resistance could result in $1 trillion in additional healthcare costs and as much as 3.8 percent loss of global economic output by 2050.62 It is no coincidence that the meat sectors most reliant on intensive confinement for production – poultry and pork – are also sectors heavily associated with infectious disease concerns. The 2009 swine flu pandemic, which was caused by a novel reassortment of multiple influenza strains circulating in pigs, spread to 74 countries and is estimated to have caused between 151,700 and 575,400 human deaths in the first year of circulation.63 In the United States alone,

Jason R. Rohr, et. Al., Nat Sustain., Emerging Human Infectious Diseases and the Links to Global Food Production (June 11, 2019), https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7091874/. 57

Jones K. E., Patel N. G., Levy M. A., Storeygard A., Balk D., Gittleman J. L., Daszak P., Global trends in emerging infectious diseases. Nature 451, 990–993 (2008). 58

Matthew N. Hayek, The Infectious Disease Trap of Animal Agriculture, Sci Adv. (Nov. 2, 2022), https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9629715/. 59

E.g., ORG. FOR ECON. COOP. & DEV., ANTI-MICROBIAL RESISTANCE IS A GLOBAL CHALLENGE FOR FOOD SYSTEMS AND PUBLIC HEALTH, https://www.oecd.org/agriculture/topics/antimicrobial-resistance-and-agriculture. 60

David Wallinga et al., A Review of the Effectiveness of Current US Policies on Antimicrobial Use in Meat and Poultry Production, 9 Current Env’t Health Reports 339 (2022). 61

https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance (citing Antimicrobial Resistance Collaborators. (2022). Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. The Lancet; 399(10325): P629-655. DOI: https://doi.org/10.1016/S0140-6736(21)02724-0). 62

Jonas,Olga B.; Irwin, Alec; Berthe,Franck Cesar Jean; Le Gall,Francois G.; Marquez,Patricio V. Drug-resistant infections : a threat to our economic future (Vol. 2) : final report (English). HNP/Agriculture Global Antimicrobial Resistance Initiative Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/323311493396993758/final-repor 63

CDC. Past Pandemics. Influenza (flu). https://www.cdc.gov/flu/pandemic-resources/basics/past-pandemics.html. Accessed 10 Nov 2018. Pigs are especially concerning given their susceptibility to both avian and human influenza, and their resulting ability to serve as a “mixing vessel” for novel diseases. Rebecca K. McLean & Simon P. Graham, The Pig as an Amplifying Host for New and Emerging Zoonotic Viruses, 14 One Health (2022); Wenjun Ma et al., The Pig as a Mixing Vessel for Influenza Viruses: Human and Veterinary Implications, 3 J. Molecular & Genetic Med. 158 (2008).

swine flu was linked to almost 275,000 hospitalizations and 12,500 deaths.64 At the time of this writing, the Centers for Disease Control (CDC) is closely monitoring the spread of the H5N1 bird flu virus through U.S. poultry and dairy cow populations, which has leaped to at least two farmworkers.65 Avian influenzas such as H5N1 are especially concerning to public health officials because they can have extremely high mortality rates in humans. In 2013, the novel avian influenza virus H7N9 spread across poultry in China, resulting in over 1500 human cases with a 40 percent mortality rate. The morality rate of H5N1 has been estimated to be 60 percent.66 Workers. In recent decades, the meatpacking industry has shifted from having thousands of smaller, often unionized facilities to having only a few extremely large facilities placed strategically in areas with attractive regulatory conditions and little union presence, as well as proximity to growers, feedlots, and transportation hubs. Today, around 500,000 workers are employed by U.S. slaughterhouses, with disproportionate numbers of them racial minorities or foreign-born.67 Most plants have hundreds or more workers standing shoulder to shoulder at a given time, each undertaking a single step in the process of breaking down an animal into parts. Depending on the stage of the process, that might mean making a cut in a carcass as it passes by hanging from a conveyer belt overhead, or it might mean working at a long metal table as animal portions move past workers to be further broken down. Workers hold knives and scissors in cold refrigerated conditions and make the same cut thousands and thousands of times a day. Meals and breaks are strictly limited, leading some workers to wear adult diapers to avoid being penalized for taking unscheduled bathroom breaks.68 Worker dignity is not the only value placed at risk by modern slaughterhouse conditions. A 2016 study by the U.S. Government Accounting Office noted that “meat and poultry workers continue to face worksite hazards that put them at risk of severe and lasting injury.”69 According to Human Rights Watch, [Slaughterhouse] workers have some of the highest rates of occupational injury and illness in the United States. They labor in environments full of potentially lifethreatening dangers. Moving machine parts can cause traumatic injuries by crushing, amputating, burning, and slicing. The tools of the trade – knives, hooks, scissors, and saws, among others – can cut, stab, and infect. The cumulative trauma of repeating the same, forceful motions, tens of thousands of times each day can cause severe and disabling injuries. Such hazards are compounded by the fact that workers earn low wages, and most are not given paid sick leave. Even for those with paid sick leave, performance point systems work to penalize employees for absences and incentivize them to continue working through illness and injury. These 64

McKiver v. Murphy-Brown, LLC, 980 F.3d 937, 980 (4th Cir. 2020).

https://www.statnews.com/2024/05/22/bird-flu-in-humans-michigan-reports-h5n1-infection-in-dairy-farm-worker/

Kain T, Fowler R. Preparing intensive care for the next pandemic influenza. Crit Care. 2019 Oct 30;23(1):337. doi: 10.1186/s13054-019-2616-1. PMID: 31665057; PMCID: PMC6819413. 67

https://www.epi.org/blog/meat-and-poultry-worker-demographics/

Oxfam No Relief report

conditions understandably lead to extremely high turnover rates among workers. In turn, the socioeconomic composition of the slaughterhouse workforce has shifted to encompass high numbers of immigrant workers from Central and South America or Southeast Asia, many of whom do not have legal authorization to work. The difficult labor conditions prevalent in U.S. slaughterhouses were put on dramatic display during the COVID-19 crisis.70 Transmission risk was exacerbated not only by the densely crowded work conditions and the multiple hours of exposure over a shift, but also by the loud and frigid work environment, which forces workers to yell to communicate and which permits viruses to stay viable outside the body for extended periods of time. In 2020, the Occupational Safety and Health Administration (OSHA) found the COVID-19 risk to workers at one large South Dakota plant was more than 70 times higher than the state’s general population risk.71 According to the CDC, nearly 90 percent of meat processing workers who contracted COVID-19 in April and May of 2020 were members of a racial or ethnic minority group. A 2021 congressional investigation determined that, among just the five largest meatpacking companies, at least 59,000 workers tested positive for COVID-19 and at least 269 died during the initial months of the pandemic.72 Not surprisingly, these high transmission levels within slaughterhouses also impacted COVID-19 prevalence in surrounding communities.73 Indeed, the entire country was affected: A study in late 2020 estimated that up to eight percent of all COVID-19 cases and four percent of deaths in the United States were linked to meat processing plants.74 Farmers. Because of the concentrated market power enjoyed by meat processing firms,75 contract farmers in animal production generally exist in a precarious financial state. They typically take on substantial debt to purchase and construct CAFOs but remain vulnerable to meat companies’ demands for additional capital enhancements, operational changes, or unilateral termination of contracts – a system that journalist Christopher Leonard likened to modern-day sharecropping.76 In 2022, the median household income from poultry farming was negative

Fred Gerr, Meatpacking Plant Workers: A Case Study of a Precarious Workforce, 18 J. OCCUPATIONAL & ENV’T HYGIENE 154 (2021). 71

Government Accountability Office, Meat and Poultry Worker Safety: OSHA Should Determine How to Address Persistent Hazards Exacerbated by COVID-19, GAO-23-105104, Jun 20, 2023, at p. 13. 72

Government Accountability Office, Meat and Poultry Worker Safety: OSHA Should Determine How to Address Persistent Hazards Exacerbated by COVID-19, GAO-23-105104, Jun 20, 2023, at p. 14. 73

Tina L. Saitone et al., COVID-19 Morbidity and Mortality in U.S. Meatpacking Counties, 101 Food Pol’y (2021) (estimating more than 300,000 COVID cases attributable to slaughterhouses, including approximately 10,700 deaths attributable to pork plants alone, during the first year of the pandemic) 74

Taylor CA, Boulos C, and Almond D, 2020. “Livestock plants and COVID-19 transmission.” Proceedings of the National Academy of Sciences, Dec. 15, 2020, 117(50), 31706-31715. https://doi.org/10.1073/pnas.2010115117. 75

E.g., Shefali Sharma, Mighty Giants: Leaders of the Global Meat Complex, INSTITUTE FOR AGRICULTURE & TRADE POLICY (Apr. 10, 2018), https://www.iatp.org/blog/leaders-global-meat-complex; Dong Li & Dennis L. Weisman, Ruffled Feathers: The Chicken Cartel in the United States, 68 ANTITRUST BULLETIN 47 (2023); Explainer: How Four Big Companies Control the U.S. Beef Industry, REUTERS (June 17, 2021). 76

Christopher Leonard, The Meat Racket: The Secret Takeover of America’s Food Business (2014). https://www.ers.usda.gov/webdocs/charts/90796/commodityspecializationnovember2023_d.html?v=5762.2

$4,069, while hog farming produced negative $166.77 Given the remote and economically depressed location of many CAFO facilities, less opportunity exists for contract farmers to earn off-farm income in order to make ends meet. Like small farmers more broadly in the United States,78 they experience disproportionately high rates of depression and suicide. Many unique aspects of contract animal farming – such as the prospect of catastrophic losses due to disease outbreaks or natural disasters, and the need to participate in “depopulating” of facilities and disposing of mass carcasses during times of crisis – can contribute to these high levels of mental distress.79 Neighboring communities. A 2008 Pew Commission study on the impact of industrial animal agriculture on rural communities concluded that “industrialization draws wealth and life away from the very rural communities it purports to benefit and which once thrived as a result of diverse, and more sustainable, forms of livestock production.”80 While promising economic investment and growth, CAFOs can leave rural communities financially worse off by decreasing local spending, lowering property values, reducing employment opportunities, and outcompeting smaller enterprises.81 Social cohesion also can suffer as various qualify of life harms associated with CAFOs (e.g., noxious odors, constant truck traffic, piling of downed hogs in open-air “dead boxes” which attract flies, gnats, and buzzards82) spark conflict among neighbors and reduce opportunities for outdoor connection. In addition to economic and social impacts, living in a rural community near a CAFO facility is associated with a variety of environmental, health, and safety threats. CAFOs require ventilation systems to constantly expel toxic gases which would otherwise accumulate and cause animal fatalities. Those same gases expose workers, their families, and others living nearby. As summarized by Judge Harvie Wilkinson in his widely noted concurrence in a North Carolina hog nuisance dispute, “Given that these gases could kill pigs, it is entirely unsurprising that 77

Christopher Leonard, The Meat Racket: The Secret Takeover of America’s Food Business (2014). https://www.ers.usda.gov/webdocs/charts/90796/commodityspecializationnovember2023_d.html?v=5762.2 78

Cora Peterson Et Al., Suicide Rates By Industry And Occupation — National Violent Death Reporting System, 32 States, 2016, Centers For Disease Control And Prevention (Jan. 24, 2020), https://www.cdc.gov/mmwr/volumes/69/wr/mm6903a1.htm?s_cid=mm6903a1_e&deliveryName=USCDC_921DM17737. Thompson, J.M., R.M. and A.D. Hagerman. 2023. "Stress and Resiliency among Confined Animal Producers" Choices. Quarter 1. Available online: https://www.choicesmagazine.org/choices-magazine/themearticles/farm-stress/stress-and-resiliency-among-confined-animal-producers 79

Thompson, J.M., R.M. and A.D. Hagerman. 2023. "Stress and Resiliency among Confined Animal Producers" Choices. Quarter 1. Available online: https://www.choicesmagazine.org/choices-magazine/theme-articles/farmstress/stress-and-resiliency-among-confined-animal-producers 80

BROTHER DAVID ANDREWS & TIMOTHY J. KAUTZA, IMPACT OF INDUSTRIAL FARM ANIMAL PRODUCTION ON RURAL COMMUNITIES, PEW COMM’N ON INDUS. FARM ANIMAL PROD. 35 (2008). 81

Donham KJ, Wing S, Osterberg D, Flora JL, Hodne C, Thu KM, Thorne PS. Community health and socioeconomic issues surrounding concentrated animal feeding operations. Environ Health Perspect. 2007 Feb;115(2):317-20. doi: 10.1289/ehp.8836. Epub 2006 Nov 14. PMID: 17384786; PMCID: PMC1817697. Despite housing tens of thousands of animals, a swine or poultry CAFO might employee fewer than ten workers given the high levels of automation deployed. See McKiver v. Murphy-Brown, LLC, 980 F.3d 937, 983 (4th Cir. 2020) (observing that defendant’s hog CAFO employed just one worker to manage 14,000 pigs across 12 sheds). 82

McKiver v. Murphy-Brown, LLC, 980 F.3d 937, 981 (4th Cir. 2020).

approximately 50 percent of [CAFO] workers experience one or more of the following health outcomes: bronchitis, toxic organic dust syndrome, hyper-reactive airway disease, chronic mucous membrane irritation, occupational asthma and hydrogen sulfide intoxication.”83 The bulk of environmental, health, and safety threats from CAFOs are attributable to the massive amount of animal manure produced at the facilities, although heavy use of antibiotics and agrochemicals are also implicated.84 A single CAFO can produce as much as 1.6 million tons of manure per year, more than the amount generated by the human populations of some large U.S. cities.85 The waste is typically collected in huge open-air lagoons where it decomposes and is eventually sprayed on nearby fields, but where it also can contaminate land, surface water, and groundwater through leaching, overfilling, or flooding. Even under normal conditions, waste collection and spraying give rise to airborne emissions like particulate matter, volatile organic compounds, hydrogen sulfide, endotoxins, and methane, as well as soil and water contaminants like nitrates.86 Often, pollution impacts are exacerbated by overapplication of manure since the sheer amount of animal waste collected is more than could possibly be used as fertilizer. These various environmental releases impact human health and well-being through multiple pathways. Scientists estimate, for instance, that animal agriculture is responsible for 12,720 annual deaths from airborne particulate matter exposure.87 More broadly, neighbors of CAFO facilities experience higher levels of respiratory tract, skin, and eye irritations, chronic lung disease, inflammation, immune function impairment, high blood pressure, infant mortality, nausea, headaches, sleep disturbances, anxiety, and depression. One epidemiological study concluded that living in proximity to a North Carolina hog CAFO is significantly associated with a higher risk of mortality, especially cardiovascular mortality but also anemia and kidney disease morality.88 Given the many negative externalities of CAFOs, one might expect them to be concentrated near populations holding less power and resources to object to their presence. In North Carolina and elsewhere, studies consistently show that CAFOs are disproportionately located in communities that are low-income and non-white, and near schools that are low-income and non-

McKiver v. Murphy-Brown, LLC, 980 F.3d 937, 979 (4th Cir. 2020).

E.g., Amanda D. Emert et al., Atmospheric Transport of Particulate Matter and Particulate-Bound Agrochemicals from Beef Cattle Feedlots: Human Health Implications for Downwind Agricultural Communities, 894 SCIENCE TOTAL ENV’T (2023) (finding elevated levels of particulate matter and pesticides in communities downwind of cattle feedlots). 85

https://www.gao.gov/assets/gao-08-944.pdf

E.g., Laima Cesoniene, The Impact of Livestock Farming Activity on the Quality of Surface Water, 26 Env’t Sci. & Pollution Rsch. 32678, 32684 (2019) (“One of the most dangerous pollution sources is intensive livestock facilities, whose production waste has a negative impact on soil and water quality.”). 87

https://www.pnas.org/doi/10.1073/pnas.2013637118

Son JY, Miranda ML, Bell ML. Exposure to concentrated animal feeding operations (CAFOs) and risk of mortality in North Carolina, USA. Sci Total Environ. 2021 Dec 10;799:149407. doi: 10.1016/j.scitotenv.2021.149407. Epub 2021 Aug 2. PMID: 34365264; PMCID: PMC8530906.

white.89 To make matters worse, the vulnerability of these populations to the adverse impacts of CAFO exposure is higher than the general population due to compounding factors like pre-existing medical ailments, housing stock condition, lack of air conditioning, and inadequate health care. As aptly summarized by Judge Wilkinson, “At the end of all this wreckage lies an uncomfortable truth: these nuisance conditions were unlikely to have persisted for long – or even to have arisen at all – had the neighbors of [CAFOs] been wealthier or more politically powerful.”90 Ecosystems, biodiversity, and natural resources. In addition to local environmental, health, and safety risks imposed on neighboring communities, industrial animal agriculture also contributes to larger-scale and longer-term environmental harms. As is well known, animal agriculture is a leading cause of habitat destruction91 and biodiversity loss,92 both directly through the clearing and use of land for animal grazing and indirectly through the devotion of land to feed crop production. Animals convert crop calories to meat calories at less than a one-to-one ratio, turning the majority of feed input calories into feces, heat, or inedible body parts.93 Thus, human diets rich in meat consumption require more agricultural land to sustain. Worldwide, nearly onehalf of habitable land (i.e., ice- and barren-free) is devoted to agriculture, with crop production taking up one-third and animal grazing two-thirds.94 Because large portions of crop production is devoted to livestock feed, the overall share of agricultural land devoted to livestock is approximately 80 percent.95 In addition to direct habitat destruction, livestock production also contributes to biodiversity loss through habitat alterations involving vegetation, pathogens, introduced poisons and medicines, soil structure and composition, water quality, disruption of predator-prey dynamics, spread of invasive species, and overuse of resources needed by other flora and fauna.96

McKiver v. Murphy-Brown, LLC, 980 F.3d 937, 982 (4th Cir. 2020) (“It is well-established – almost to the point of judicial notice – that environmental harms are visited disproportionately upon the dispossessed – here on minority populations and poor communities.”). 90

McKiver v. Murphy-Brown, LLC, 980 F.3d 937, 982 (4th Cir. 2020).

E.g., TIM G. BENTON ET AL., CHATHAM HOUSE, FOOD SYSTEM IMPACTS ON BIODIVERSITY LOSS 6–7 (Feb. 2021) (“Over the past 50 years, the biggest driver of habitat loss has been the conversion of natural ecosystems for crop production or pasture. . . . The rapid expansion of animal farming has been behind much of this land expansion.”). 92

E.g., U.N. FOOD & AGRIC. ORG., THE IMPACT OF LIVESTOCK ON BIODIVERSITY 1 (2019), https://www.fao.org/3/ca4960en/ca4960en.pdf (“Livestock is among the sectors with highest impacts on biodiversity.”). 93

For example, the feed conversion ratio for cattle may be up to 24 kilograms of feed for a single kilogram of edible beef. Mesfin Mekonnen et al., ‘Water productivity in meat and milk production in the US from 1960 to 2016’ (2019) 132 Environmental International 10 tbl. 5. 94

https://ourworldindata.org/global-land-for-agriculture. See also Ellis, E. C., Klein Goldewijk, K., Siebert, S., Lightman, D., & Ramankutty, N. (2010). Anthropogenic transformation of the biomes, 1700 to 2000. Global Ecology and Biogeography, 19(5), 589-606 (estimating that by 2000, 55% of Earth’s ice-free land had been converted into cropland, pasture, and urban areas). 95

https://ourworldindata.org/global-land-for-agriculture.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10873014/#:~:text=Specifically%2C%20on%20a%20global%20sc ale,2005%3B%20Wazna%2C%202016).

Life’s most essential resource is water. As noted at the outset of this Part, excess groundwater usage is a major sustainability challenge associated with agriculture.97 A recent investigation determined that 45 percent of over 80,000 groundwater wells surveyed in the United States showed a statistically significant decline since 1940, as water is being pumped at rates that far exceed annual recharge levels. In the past decade, four of every ten sites analyzed hit their lowest levels on record. Climate change exacerbates the problem as higher levels of evaporation, declining snowpack, and less predictable precipitation patterns diminish the natural rate of recharge. Even left untouched and in the absence of climate impacts, groundwater and aquifer sources would take centuries or even millennia to replenish. As the associate director for water resources at the U.S. Geological Survey noted, “There is no way to get that [water] back. There’s almost no way to convey how important it is.”98 Like the unsustainable depletion of groundwater, industrial animal agriculture also depends on the heavy use of chemical fertilizers, pesticides, and herbicides. Global use of such chemicals in agriculture has increased exponentially since World War II with dramatic ecological implications. For instance, each summer a vast “dead zone” threatening the survival of marine life emerges in the Gulf of Mexico due in substantial part to nutrient pollution runoff from agricultural land devoted to feed crop production. Along with hundreds of other such coastal “dead zones” worldwide, the Gulf of Mexico’s annual decline in oxygen availability due to fertilizer-induced algal blooms is just one manifestation of humanity’ alteration of the planet’s nitrogen cycle. Nitrogen is essential to living organisms, and the supply of nitrogen in ecosystems is a critical factor effecting the nature and diversity of plant and animal life. Until human interventions over the past century, new nitrogen was delivered to ecosystems from the atmosphere only through limited processes involving lightning and nitrogen-fixing microbes. The resulting low availability of nitrogen caused ecosystems to evolve toward low productivity and high biodiversity. Humanity has reversed that equation, developing instead monocultural landscapes that are high in productivity but low in biodiversity. Some of these monocultural crops – including soybeans, peas, and alfalfa – are themselves producers of nitrogen since they support nitrogenfixing microbes, a further way in which humanity has transformed the planet’s nitrogen cycle. Most critical to this transformation was the early twentieth century discovery of the Haber-Bosch process which combines nitrogen from the air with hydrogen to produce ammonia. By enabling industrial-scale production of ammonia, the Haber-Bosch process gave rise to the modern fertilizer industry which in turn powered the dramatic increase in crop yields that supported the dramatically expanding size of both the human population and the livestock population it feeds upon. For this reason, the historian and geographer Vaclav Smil has referred to the Haber-Bosch process as the “detonator of the population explosion.”99

See also Brian D. Richter et al., Water Scarcity and Fish Imperilment Driven by Beef Production, 3 NATURE SUSTAINABILITY 319, 320 (2020) (“[I]rrigation of cattle-feed crops . . . is the single largest consumptive [water] user at both regional and national scales [in the United States], accounting for 23% of all water consumption nationally, 32% in the western US and 55% in the Colorado River basin.”). 98

https://www.nytimes.com/interactive/2023/08/28/climate/groundwater-drying-climate-change.html

Smil, V. Detonator of the population explosion. Nature 400, 415 (1999). https://doi.org/10.1038/22672.

This massive infusion of artificially produced nitrogen into ecosystems – approximately 176 million tons of ammonia are produced each year100 – comes with significant implications for the planet. Although introduced nitrogen eventually returns to the atmosphere, it first cascades through ecosystems and contributes to a variety of harmful impacts, including eutrophication of coastal areas (e.g., “dead zones”), soil and water acidification, increased concentrations of lifethreatening particulate matter and ozone in the atmosphere, depletion of the protective ozone layer in the stratosphere, and biodiversity loss in both terrestrial and aquatic ecosystems. In short, the technological bargain that supported feeding a vastly larger human population may come at a longterm ecological cost. Researchers who catalogued available evidence regarding the detrimental impacts of human fixation of nitrogen concluded that going forward, despite the productivity benefits of artificial fertilizer use, “the negative health consequences on ecosystems and people may become more diverse, and might in total increase more rapidly than the benefits.”101 Climate stability. Globally, ammonia production accounts for 1.8 percent of carbon dioxide emissions,102 on par with both the aviation and maritime sectors. Yet fertilizer production is just one of many ways in which industrial animal production contributes to climate change. According to an influential study by the United Nations Food and Agriculture Organization (FAO), approximately 14.5 percent of global anthropogenic greenhouse gas emissions – some 7.1 gigatons (Gt) of annual carbon-dioxide equivalent (CO2e) emissions – are attributable to livestock.103 Subsequent FAO estimates have ranged between 11.2 percent and 18 percent, while a significant peer-reviewed study put the total as high as 20 percent.104 Regardless of the exact number, animal agriculture is a significant contributor to climate change. For perspective, analysts calculate that the five largest global meat and dairy companies collectively are responsible for more greenhouse gas emissions than the fossil fuel giants Exxon, Shell, or BP; twenty animal agriculture companies contribute more emissions together than Germany, Canada, Australia, the United Kingdom, or France.105 Researchers also have concluded that the Paris Agreement’s temperature goals cannot be achieved without substantial reform of the global food system, even if all other sources of

100

https://royalsociety.org/-/media/policy/projects/green-ammonia/green-ammonia-policy-briefing.pdf

101

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3682738/#:~:text=Overall%2C%20as%20human%20fixation%20o f,more%20rapidly%20than%20the%20benefits 102

https://royalsociety.org/-/media/policy/projects/green-ammonia/green-ammonia-policy-briefing.pdf

103

See Gerber, P.J., et al., Tackling climate change through livestock – A global assessment of emissions and mitigation opportunities, (FAO 2003). 104

See Bray, D., & Poston, T. M. (2024). The Methane Majors: Climate Change and Animal Agriculture in U.S. Courts. Columbia Journal of Environmental Law, 49(S), 145–248. https://doi.org/10.52214/cjel.v49iS.12548 105

See Institute for Agriculture and Trade Policy, ‘Emissions Impossible: How big meat and daily are hearing up the planet’, 18 July 2018

emissions are curtailed.106 Globally, animal-based products account for as much as 57 percent of greenhouse gas emissions related to food production.107 Beef and dairy dominate livestock sector emissions, producing roughly 62 percent of livestock emissions globally or 5.0 Gt CO2e per year. Pigs, poultry, buffalo, and small ruminants such as sheep and goats have lower emissions, with each category representing between 7 and 11 percent of livestock sector emissions. For pigs and poultry, much of this difference occurs because the animals produce fewer GHG emissions per unit of product than cattle or other ruminant animals.108 The smaller shares of total sector emissions represented by non-cattle livestock, however, should not be taken to indicate that they are benign from a climate perspective, as the absolute scale of emissions remains globally significant. For instance, poultry production generates approximately 0.8 Gt CO2e per year, an amount that would make poultry the sixth largest emitter worldwide if viewed as a separate country.109 Dramatic though these figures may be, they likely understate the climate impact of livestock agriculture.110 Under current measurement approaches, the sector has two main sources of GHG emissions, both of which are primarily composed of methane. The first source is “enteric fermentation,” whereby methane produced in an animal’s stomach during digestion is released from its body.111 Enteric fermentation – primarily from ruminant animals such as cows which can release 250 to 500 litres of methane per individual per day112 – is the largest anthropogenic source of methane in the United States.113 The second source is animal excrement, which emits methane and nitrous oxide as it decomposes. Nitrous oxide is 265-298 times more powerful than CO2 on a 100-year timeframe.114 GHG emissions from animal waste vary depending on handling techniques.

106

See Clark, M.A., Domingo, N.G., Colgan, K., Thakrar, S.K., Tilman, D., Lynch, J., Azevedo, I.L. and Hill, J.D., 2020. Global food system emissions could preclude achieving the 1.5 and 2 C climate change targets. Science, 370(6517), pp.705-708. 107

See Xu, Xiaoming, et al. "Global greenhouse gas emissions from animal-based foods are twice those of plantbased foods." Nature Food 2.9 (2021): 724-726. See also Michael B. Eisen and Patrick O. Brown, Rapid Global Phaseout of Animal Agriculture Has the Potential to Stabilize Greenhouse Gas Levels for 30 Years and Offset 68 Percent of CO2 Emissions This Century, PLOS CLIMATE 2–3, 7 (Feb. 1, 2022) (accounting for both “emission reduction and biomass recovery” impacts, a “global phaseout” of animal agriculture over the next fifteen years “would have the same effect, through the end of the century, as a 68% reduction of CO2 emissions”). 108

Food and Agricultural Organization of the United Nations, ‘Global Livestock Environmental Assessment Model’, 2017. 109

Union of Concerned Scientists, ‘Each Country’s Share of CO2 Emissions’, 14 January 2022.

110

See Kysar, Douglas A., and Jonathan Lovvorn. 2024. “Animal Production” in Oxford Handbook of Global Animal Law (Anne Peters, Kristen Stilt, and Saskia Stucki, eds.). London: Oxford University Press. 111

Henning Steinfeld et al., Livestock's Long Shadow: Environmental Issues and Options (FAO 2006) 113.

112

K.A. Johnson and D.E. Johnson, ‘Methane emissions from cattle’ (1995) 73 Journal of Animal Science 2483, 2483. 113

Environmental Protection Agency, ‘Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2019’, 2021, 7 (hereinafter EPA Inventory). 114

G. Myhre, et al., ‘Anthropogenic and Natural Radiative Forcing’ in: Climate Change 2013: The Physical Science Basis (Cambridge University Press 2014) 659, 714.

Wet storage, in which waste is stored in a large pit or lagoon as occurs at most industrial-scale CAFOs, produces more methane than dry systems.115 Some scientists also argue that the conventional reliance on a 100-year timeframe to calculate the global warming potential (GWP) of non-CO2 greenhouse gases understates the warming significance of methane emissions, which would have a GWP 84-86 times that of CO2 if considered over a 20-year time horizon that arguably is more appropriate for near term policymaking.116 Currently, animal agriculture contributes 32 percent of annual anthropogenic methane emissions and is a primary reason why atmospheric concentrations of methane are at their highest level in at least 800,000 years.117 Recognizing that methane sources are coming under increasing political pressure, industrial meat producers have latched onto a controversial alternative scientific technique for expressing global warming potential. Known as GWP*, this technique focuses on annual changes in emissions levels rather than absolute contributions, on the theory that short-lived gases dissipate more quickly than CO2 and therefore should not be considered “stock” pollutants in the same manner. This approach, however, has the convenient effect of grandfathering in the entire pre-existing global livestock herd as if it were somehow benign from a climate perspective, a feature that industry (and the academic researchers it supports) have promoted with gusto.118 Given such ongoing debate over how best to assess the contribution of methane emissions to the climate problem, recent researchers not aligned with the meat industry have sought instead to model the share of actual warming that could be attributed to direct livestock emissions in a given year. Using this approach, the direct non-CO2 emissions of livestock were estimated to represent 19 percent of all climate warming from all anthropogenic sources in the year 2010,119 far too much to simply grandfather in as somehow part of the “natural” state of the planet. In addition to emissions from enteric fermentation and manure waste, animal agriculture also generates significant emissions through deforestation, manufacturing of fertilizers and pesticides for feed crops, application of fertilizers on feed croplands, feed processing and transportation, on-farm energy use, and energy use for processing, packing, and transporting livestock commodities. Feed production merits special focus, as the FAO estimates that emissions associated with feed production constitute 41 percent of the overall GHG emissions of the

115

EPA Inventory, § 5, 5-12.

116

See Environmental Protection Agency, ‘Understanding Global Warming Potentials’, 18 October 2021.

117

See United Nations Environment Programme, Global Methane Assessment: Benefits and Costs of Mitigating Methane Emissions (2021). Estimates of animal methane emissions rely on models that are uncertain and may have a conservative bias. For instance, a recent study concludes that ‘bottom-up’ methods of estimating livestock emissions – which depend on herd inventories and biological assumptions about average levels of animal emissions – may underreport emissions from CAFOs by 39-90% as compared to ‘top-down’ methods that utilize actual atmospheric measurements taken in proximity to such facilities. Mathew N Hayek and Scott M Miller, ‘Underestimates of methane from intensively raised animals could undermined goals of sustainable development’ (2021) 16 Environmental Research Letters 1, 7. 118

Morris, V., Jacquet, J. The animal agriculture industry, US universities, and the obstruction of climate understanding and policy. Climatic Change 177, 41 (2024). https://doi.org/10.1007/s10584-024-03690-w. 119

Andy Reisinger and Harry Clark, ‘How much do direct livestock emissions actually contribute to global warming?’ (2017) 24 Global Change Biology 1749, 1754.

livestock sector, totaling some 3.3 Gt CO2e per year.120 When estimating the role of feed production in livestock emissions, analysts face important choices over how to consider associated land use impacts. Often, researchers only attribute land use emissions to animal agriculture in the year a change occurs; thus, past land conversion for feed production or grazing is not counted, although the land’s ability to sequester carbon continues to be impaired by animal production. More recently, researchers have sought to quantify the ongoing “carbon opportunity cost” of using land to grow feed for animals.121 A particularly notable study attempted to estimate the carbon storage potential of native ecosystem land restoration opportunities that would be created through a shift in global food production to plant-based diets by the year 2050. The study showed that 332– 547 Gt of CO2 sequestration could be achieved in this manner, an astonishing amount that might well enable the world to stay below 1.5 °C of warming, if only humans could give up eating meat.122 Animals. At the time of COVID-related livestock depopulation efforts, members of the American Veterinary Medical Association (AVMA) signed on open letter to protest the use of ventilation shutdown techniques, noting that “less cruel options exist but apparently were deemed too expensive and inconvenient.”123 Such disregard of animal well-being in favor of economic productivity is routine for industrial animal producers, who oversee 99 percent of meat and dairy animals in the United States.124 Because efficiency gains are monetized and animal welfare costs are not, inhumane production practices appear to be more beneficial than they would from an accounting perspective that included nonhuman animal impacts. Consider confinement, which is perhaps the defining characteristic of industrial animal agriculture. In swine and poultry facilities, animals are confined in tight spaces that severely restrict their movement and ability to engage in normal behaviors. To breed new pigs, the pork industry prefers to house pregnant sows within gestation crates rather than group housing or a freerange environment. Gestation crates lower costs by allowing for a larger number of animals, greater control over feed and water intake, ease of inspection and access for veterinary treatments, 120 121

See Mathew Hayek, et al., ‘The carbon opportunity cost of animal-sourced food production on land’ (2021) 4 Nature Sustainability 21-24; Timothy Searchinger et al., ‘Assessing the efficiency of changes in land use for mitigating climate change’ (2010) 565 Nature 249-253; Paul West et al., ‘Trading carbon for food: Global comparison of carbon stocks vs. crop yields on agricultural land’ (2010) 107 Proceedings of the National Academy of Sciences of the United States of America 19645-19648. 122

See Mathew Hayek, et al., ‘The carbon opportunity cost of animal-sourced food production on land’ (2021) 4 Nature Sustainability 21, 21. 123

Open Letter from AVMA Member Veterinarians 1 (AVMA Members Letter), https://news.vin.com/apputil/image/handler.ashx?docid=9705967 (last visited May 10, 2024). The AVMA’s Depopulation Guidelines state that ventilation shutdown is not recommended for chickens, turkeys, ducks, or pigs, and that it should only be used as a “last resort” in constrained circumstances because the process is neither humane nor sufficiently effective. Am. Veterinary Med. Ass’n, AVMA Guidelines for the Depopulation of Animals: 2019 Edition 45, 61-62 (2019), https://www.avma.org/sites/default/files/resources/AVMA-Guidelines-for-theDepopulation-of-Animals.pdf. 124

Hannah Ritchie, How Many Animals Are Factory Farmed?, OUR WORLD IN DATA (Sep. 25, 2023), https://ourworldindata.org/how-many-animals-are-factory-farmed (citing estimates based on USDA data that 99% of U.S. livestock animals are raised on industrial-scale CAFOs, including 70% of cows, 98% of pigs and egg-laying hens, 99.9% of turkeys, and 99.97% of broiler (meat) chickens).

and reduced opportunities for aggression between animals. For the animals, however, the tight conditions of gestation crates only allow them to stand or lie down – even turning around to face the opposite direction is prohibited. As a result, confined sows suffer injuries such as sores, ulcers, and abrasions, as well as engage in higher levels of stereotypic behaviors such as repeated biting, chewing, licking and rubbing.125 Following a four-month gestation period and approximately three weeks of nursing, their offspring are taken away even though pigs in the wild spend several months together with their mothers nursing, grooming, and foraging. Sows will be artificially inseminated and put through this process repeatedly for the short period in which they are capable of reproduction under confinement conditions.126 Their offspring, in turn, will be quickly brought from a birth weight of 2-3 pounds to a market weight of 280 pounds, resulting in their slaughter only 6-7 months from birth despite having a lifespan of 10-15 years.127 Battery hens also spend the majority of their lives in severely confined spaces. Although state sales restrictions and consumer demand are helping to shift the market toward cage-free production, the majority of eggs sold in the United States continue to come from laying hens confined in battery cages.128 Developed in the 1930s, the battery cage system is ingenious: Multiple hens are confined in a cage with a slanted floor that allows eggs to roll out of the cage onto a conveyor belt for collection. To prevent aggressive pecking due to the crowded conditions, hens are debeaked, often with a hot blade trimmer. The floor of the cage is made of wire to allow waste and feathers to fall onto a separate conveyor belt for removal. Feed and water, supplemented with antibiotics and other additives to suppress infection and boost productivity, are delivered via automated systems. Temperature and light are controlled to maximize egg laying. A chicken in the wild only lays eggs seasonally in response to a certain number of hours of daylight exposure, producing around 10-15 eggs per year for as long as 5-10 years depending on the breed. Within a CAFO environment, on the other hand, chickens are made to produce 250-300 eggs per year. Because of the resulting physiological toll on their bodies, factory farmed hens experience declining production after only a period of months. As productivity drops, battery hens in the United States typically are forced to molt by being denied food for one or more weeks, a process that rejuvenates egg-laying for an additional stretch of time. Even with forced molting, most battery hens are killed at about 70 weeks of age. Their lives will have passed without the 125

https://www.avma.org/sites/default/files/resources/WelfareImplicationsOfGestationSowHousing.pdf

126

Estimating the life span of a factory farmed sow is complicated by the fact that within the industry, “[l]ongevity of sows is commonly measured as the number of parity at removal,” where “parity” refers to the number of times a sow has given birth and “removal” means being sent to slaughter or euthanized. Koketsu, Y., Iida, R. Farm data analysis for lifetime performance components of sows and their predictors in breeding herds. Porc Health Manag 6, 24 (2020). https://doi.org/10.1186/s40813-020-00163-1. In the United States, large facilities with more than 6,000 sows reportedly removed 46% of their breeding herd after no more than two litters. https://www.nationalhogfarmer.com/hog-reproduction/part-3-what-is-going-on-with-female-death-loss-and-culling#. One of the primary reasons for these short lifespans is lameness and other debilitating hoof and leg conditions associated with confinement. https://www.nationalhogfarmer.com/hog-welfare/structural-conformation-impacts-onsow-longevity. 127

https://www.avma.org/sites/default/files/resources/WelfareImplicationsOfGestationSowHousing.pdf

128

https://porkcheckoff.org/pork-branding/facts-statistics/life-cycle-of-a-market-pig/

possibility of expressing a variety of innate behaviors – such as foraging, wing-flapping, preening, dustbathing, scratching, perching, roosting, and building of secluded nests for egg laying – that are prevented by the battery cage conditions. One might think that male chicks at a battery hen facility would be sold to poultry meat producers, but instead they are culled at birth by the billions worldwide. This is because poultry meat producers rely on different breeds that have been selected to maximize the speed and quantity of meat production, particularly white breast meat. Today’s broiler chickens are as much as 5 times the size of unselected breeds, yet they require only 5 to 9 weeks to mature to slaughter weight. This rapid growth leads to leg and skeletal disorders, muscle myopathies, and heart and lung development issues. The short life span is spent entirely indoors in a grow-out barn with thousands of other broilers where they may suffer ammonia burns from standing and lying in accumulated waste. After transport to a slaughterhouse, they are shackled upside down by their legs and passed through an electrified bath which is meant to stun them before they are killed. In addition to difficult confinement conditions, industrial livestock animals are subjected to a variety of husbandry practices that can negatively impact their well-being. For instance, while beef cattle typically start out in comparatively higher welfare grazing conditions, they are “finished” in densely stocked feedlots where they are given a quick-fattening diet of corn that is poorly matched to their ruminant digestive system and can cause painful conditions like acidosis. Whereas cows used to require 4-5 years to reach a slaughter weight of 1200 pounds, they now reach market after only 13-15 months. Similarly, dairy cows are regularly injected with an artificial growth hormone that boosts production but can result in lameness and mastitis, a painful infection of the udders. The modern dairy herd has been genetically selected to produce as much as 12 times the amount of milk needed to feed a calf, a physically stressful level of production that reduces the useful lifespan of a dairy cow to about 3 years rather than the 15-20 years it could otherwise live. In addition to chicken debeaking, discussed above, other regularly practiced physical alterations of livestock include tail docking, dehorning, teeth grinding, branding, ear notching, and castration. Typically, these procedures are undertaken without anesthesia or pain relief such that the animal experiences a period of acute pain. Some procedures such as tail docking may also result in untreated chronic pain. One response to concerns about the animal welfare implications of industrialized agriculture is to double down on industrialization: Bullish technologists imagine a future in which advanced genome editing techniques alter livestock animals such that their capacity for pain, sentience, cognition, or any other indicia of moral considerability has been conveniently deleted. That prospect may sound dystopian and far-fetched but consider the fact that industrial animal researchers already have experimented with selectively breeding blind chickens because they exhibit less of the aggressive behaviours (e.g., feather-pecking, cannibalism) that complicate confined production. Still, make no mistake: Humanity’s unabashed instrumentalization of nonhuman lives will raise moral questions even if somehow those lives play out free from pain. When a being is brought into and out of existence solely for the purpose of manufacturing protein and its life is maximized toward that end alone, the being suffers the dignitary opportunity cost of not living the kind of life it is capable of having. Intentionally diminishing livestock animals’ capacities may lower the perceived welfare costs of industrial production, but it does not reduce the less tangible harms that come from unchecked domination of other lives.

III. CORPORATE PROTECTIVE EQUIPMENT When slaughterhouse workers requested personal protective equipment from their employers during the early stages of the COVID-19 crisis, they reportedly were told at some facilities to pull their hair nets down over their nose and mouth. At Tyson’s Waterloo, Iowa facility, managers formed a betting pool to predict how many line workers would test positive for the virus.129 At the same time, the meat industry lobbied successfully at federal and state levels for immunity from tort liability for COVID-related losses. While personal protective equipment was in shockingly scarce supply even among health care workers on the front lines of the crisis, corporate protective equipment was amply provided through law. As this Part explains, such liability shields during the pandemic are representative of the kinds of legal protections that industrial animal agriculture more widely enjoys. For an industry causally connected to a variety of severe, even planet-tilting, consequences, meat and dairy production remains remarkably unregulated. Labor and employment. The Fair Labor Standards Act of 1938 provides for a national minimum wage and a forty-hour work week with time-and-a-half compensation guaranteed for overtime work.130 Agricultural workers are largely excluded from the Act’s benefits.131 The National Labor Relations Act of 1935 protects the right of employees to form trade unions, engage in collective bargaining for better terms of employment, and take certain collective labor actions such as strikes.132 The Act does not extend these rights to agricultural workers.133 These two Acts form the bedrock of federal labor and employment law. Why were farmworkers excluded from their protections? Because at the time the Acts were passed, two sectors of employment– farmworkers and domestic workers – were heavily filled by African American workers. At the behest of Southern Democrat legislators who did not want to see the status and power of those workers elevated, President Roosevelt agreed to exempt the two sectors in exchange for the Southern Democrats’ support for worker rights legislation overall. Farmworkers today remain disproportionately non-white and substantially less protected by laws than other employment sectors. In addition, an estimated 40 percent of agricultural workers today are undocumented,134 a status that constricts those workers’ willingness to assert and bargain over legal protections. While the injury rate for agriculture workers is among the highest of any employment sector in the country, OSHA has issued only a small handful of rules 129

Winders DJ, Abrell E. Slaughterhouse Workers, Animals, and the Environment: The Need for a Rights-Centered Regulatory Framework in the United States That Recognizes Interconnected Interests. Health Hum Rights. 2021 Dec;23(2):21-33. PMID: 34966222; PMCID: PMC8694297. 130

Fair Labor Standards Act, 29 U.S.C. §§ 201–219.

131

29 U.S.C. § 213(a)(6). In 1966, Congress extended the federal minimum wage to farm workers, but they still do not enjoy the benefit of overtime pay. 132

National Labor Relations Act, 29 U.S.C. §§ 151–169.

133

National Labor Relations Act, 29 U.S.C. §§ 152(3). https://www.osha.gov/lawsregs/regulations/standardnumber/1928 134

https://www.ers.usda.gov/topics/farm-economy/farm-labor/#:~:text=In%202018– 20%2C%2030%20percent,percent%20held%20no%20work%20authorization.

designed to address safety in agriculture.135 Moreover, a 1976 appropriations measure by Congress prohibits OSHA from using federal funds to conduct inspections, investigate injuries or deaths, or respond to employee complaints on farms with 10 or fewer non-family workers.136 Given the intense automation prevalent in the meat and dairy sector, even massive CAFOs can operate with 10 or fewer employees. Likewise, because such CAFOs are formally seen as independent from the vertical integrators that dictate their operations through contract farming, they remain below the OSHA coverage threshold even though in economic reality the few consolidated meat processors are in control of most farmers and workers in the industry. The result is a striking gap in federal workplace safety law: According to a recent journalistic analysis, “workers are unprotected by federal OSHA labor laws at 96 percent of the operations that hire people to produce pork, eggs, beef, poultry, and milk in America.”137 Without the threat of OSHA inspections or investigations, injuries and deaths become an expected cost of business that is hidden from public scrutiny: “Over the decade between 2011 and 2020 . . . 85 percent of the deaths related to animal agriculture were not reported to the agency.”138 As occupational health and safety expert Deborah Berkowitz summarized, “[a]griculture is dangerous, animals are dangerous, and really, the government’s hands are tied to help workers. And a lot of these are immigrant workers who are very scared to complain and speak up.” Such power dynamics affect even portions of the meat supply chain, such as slaughterhouses, that are not exempted from labor and employment laws to the degree that workers in the field or on the farm are. As noted above, the crowded and dangerous conditions within contemporary U.S. slaughterhouses were brought to public attention during the COVID-19 crisis, as many facilities became transmission hotspots. But the story of COVID and slaughterhouses is a longer and more complicated one that illustrates the meat industry’s special talent for securing corporate protective equipment from lawmakers. The Food Safety Inspection Service (FSIS) within the U.S. Department of Agriculture is responsible for regulating slaughterhouses. In 1997, the FSIS initiated faster line speeds and other deregulatory measures in a limited number of chicken and pig slaughterhouses through a pilot program. This program was adopted under President Clinton who, although a Democrat, was a Democrat from Arkansas. The FSIS pilot program authorized certain slaughterhouses to reduce the number of inspectors on the line and to increase the chicken slaughter rate from 140 to 175 birds per minute (bpm), and the pig slaughter rate to any speed at all. 175 birds per minute is almost 3 chickens per second, necessitating slaughterhouse workers to repeat the same cut at that same rate for the duration of their shifts. Awkwardly, the program was phased in during the same period the Clinton Administration was fighting for an ergonomics safety standard to be issued through OSHA, a goal that was eventually achieved in the waning days of the presidency, but which was swiftly nullified by a Republican Congress in March 2001. To increase the line speed of 135

https://www.osha.gov/laws-regs/regulations/standardnumber/1928

136

https://www.osha.gov/enforcement/directives/cpl-02-00-163/chapter-10

137

https://civileats.com/2022/11/14/injured-and-invisible-1-few-protections-animal-agriculture-workers-cafos-dairymigrants-injuries/ 138

https://civileats.com/2022/11/14/injured-and-invisible-1-few-protections-animal-agriculture-workers-cafos-dairymigrants-injuries/. States do not close the gap as less than half have adopted their own plans to govern health and safety in private workplaces.

slaughterhouses at the same time one commits to protecting workers from repetitive stress injuries requires a certain tolerance for cognitive dissonance. In 2012, President Obama’s head of the Department of Agriculture, farm lobbyist Tom Vilsac, saw FSIS propose a rule that would have increased line speeds for chicken slaughterhouses industry wide. After receiving substantial opposition from worker, consumer safety, immigrant rights, environmental, and other organizations, FSIS instead allowed 20 chicken slaughterhouses from the pilot program to increase their line speeds even further. In 2017, with Trump in the White House and agricultural trading company founder Sonny Perdue as head of the Department of Agriculture, the National Chicken Council submitted to FSIS a petition for rulemaking requesting that the agency implement a waiver system to allow chicken slaughterhouses to operate without line speed limits, emphasizing that doing so would result in increased “production volume” and remove “arbitrary production limitations.” FSIS formally denied the petition, but less than a month later, the agency announced that it would begin granting waivers to additional chicken slaughterhouses, authorizing them to increase their speeds to 175 bpm.139 In 2019, FSIS finalized a rule (which was bluntly called the “Pig Slaughter Deregulation Rule”) that allows all pig slaughterhouses to opt out of line speed limits completely. In addition, the rule reduced the number of federal inspectors on the slaughter floor and allowed slaughterhouse personnel to assume inspection responsibilities historically assigned to agency officials. Summarizing the benefits of the rule, the FSIS emphasized that it would increase producer profits annually by “roughly $87.64 million”, increasing production by 12.49 percent – or roughly 11.5 million more pigs per year.140 Around the same time period, the FSIS issued the first-ever line speed waiver for a cattle slaughterhouse and speed waivers were granted to an additional 35 chicken slaughterhouses that were not part of the pilot program, with a record number issued in spring 2020 at the same time as COVID-19 outbreaks began occurring at slaughterhouses across the country. Thoughout the pandemic, the agency tried to advance a proposed rule that would have allowed all chicken slaughterhouses to increase line speeds to 175 bpm. Although that rule was not finalized before the end of Trump’s presidency, President Biden’s Secretary of Agriculture – Tom Vilsack, again – has stated that the item remains on the USDA’s “long term action” agenda. During the COVID-19 period when FSIS was furthering its campaign of line speed deregulation, OSHA was receiving thousands of complaints from slaughterhouse workers who felt 139

Agriculture frequently secures such workarounds following regulatory defeat before its oversight agencies. In 2017, for instance, the Food and Drug Administration (FDA) restricted use of certain antibiotics for food animal growth promotion. At least thirteen medically important antibiotics, however, remain FDA-approved for disease prevention with no durational limits, essentially allowing constant animal application. Pig and cattle antibiotic sales initially declined with the 2017 rule change, but then increased 12 and 5 percent respectively by 2020 as industry adjusted to the new regulatory environment. David Wallinga et al., A Review of the Effectiveness of Current US Policies on Antimicrobial Use in Meat and Poultry Production, 9 Current Env’t Health Reports 339 (2022). 140

A federal court in March 2021 overturned the agency’s revocation of line speed limits after a challenge was brought by a coalition of labor, environmental, animal welfare, and consumer advocacy groups. In November 2021, FSIS in consultation with OSHA created a “time-limited trial” program in which facilities were granted line speed waivers in exchange for studying and disclosing data on worker impacts. The program’s deadline has been extended twice and will now last at least through January 15, 2025. https://www.porkbusiness.com/news/ag-policy/usdaextends-new-swine-inspection-system-line-speed-trialsagain#:~:text=As%20a%20result%2C%20all%20NSIS,as%20of%20June%2030%2C%202021.

they were being forced to work while sick or when they would be exposed to others who were sick. At that point, OSHA could have issued an Emergency Temporary Standard (ETS) which takes immediate effect upon publication and is valid for six months. It is only available to OSHA when “employees are exposed to grave danger from exposure to substances or agents determined to be toxic or physically harmful or from new hazards, and that such emergency standard is necessary to protect employees from such danger.” The COVID-19 pandemic would seem to be exactly the kind of situation that would justify an ETS. Nevertheless, OSHA declined to issue an ETS for meatpacking plants, choosing instead only to issue voluntary recommendations together with the CDC for the protection of meat processing workers from COVID-19. Even then, the Administration went to great lengths to reassure the industry that “good faith attempts” at compliance were enough to avoid sanctions. The press statement accompanying the guidance asserted that state authorities could not order a plant to close or comply with additional protections. And the Department of Labor declared it would participate in litigation and advocate for employers in any future lawsuits arising out of employees’ illnesses or deaths. What explains these government responses? When COVID hit, the meat industry took the position that continued food supply was a national priority and that essentially no changes should be undertaken to business as usual. In March of 2020, the CDC announced that social distancing – staying six feet apart – was among the most important measures one could take to prevent the spread of the disease. But for meatpacking facilities, maintaining six feet of distance between workers would have required a reconfiguration of the entire production line. So, the industry resisted. One notorious example was a letter from the CEO of Smithfield Foods, Kenneth Sullivan, to Nebraska Governor Pete Ricketts, in which Sullivan opined that “[s]ocial distancing is a nicety that makes sense only for people with laptops.” He complained that public health messaging was impacting his business operations because Smithfield employees “work in close proximity to each other and are increasingly asking one question: ‘Why are we here?’ This is a direct result of the government continually reiterating the importance of social distancing.” Federal and state actors cooperated with the meat industry in resisting any alterations to their usual operations. In Nebraska, for instance, after plant outbreaks began, Governor Ricketts stopped local health departments from announcing the numbers of cases tied to meatpacking plants. He also worked to prevent any temporary plant shutdowns, refused requests to implement state-level industry requirements to protect workers, and asserted that civil unrest stemming from meat shortages would result if processing plants were temporarily shut down as an infection control measure. At the federal level, when the CDC came out with quarantine guidance, the agency advised that anyone who came in close contact with an infected individual for 15 minutes should quarantine for 14 days. That guidance initially applied to everyone, including workers. But the meatpacking industry feared that enforcing the quarantine guidance would lead to massive labor shortages in the plants. After pressure from the Trump Administration, the CDC softened its guidance for essential workers. The new guidance said that employers could require essential workers to come back to work even after an exposure and before completing a full quarantine period. In statements to the press, the Administration indicated that workers who did not comply could be fired and would not be eligible for unemployment.

Throughout this period, federal and state governments were not gathering data on COVID cases in a way that would allow identification of high-risk factors by occupation. A nonprofit organization called the Food and Environment Reporting Network assigned a researcher to start building a database which, by mid-April of 2020, revealed that many meatpacking facilities were becoming significant hotspots. As this information received publicity, local health departments and hospitals started to protest and demand closures of the plants. Most of the facilities are in rural areas and nearby hospitals would easily be overwhelmed by an influx of sick workers and community members. Inconveniently for the industry, data also seemed to show that the plants with line speed waivers posed a greater risk of transmission than other plants. The industry again turned to the President and offered a draft executive order to protect them. The resulting order referenced the Defense Production Act, a statute designed to let the federal government take control of industrial facilities during wartime or similar moments of national crisis. A number of representatives in Congress had been pressing the President to invoke the Act in order to force greater production of personal protective equipment and materials for COVID testing and tracking. Instead, the Act was invoked to ensure continued full-scale production of meat. The executive order was strangely written, and it actually did not confer any power to any agency to keep plants open, but it did contain enough language to give the industry a platform for pushing back against local health departments that were demanding shutdowns. Of the hundreds of meat processing plants that had serious outbreaks, OSHA sanctioned only two. A Smithfield plant in South Dakota was fined $13,494 after 1294 workers contracted coronavirus and four employees died – roughly $10 per infected employee. OSHA also fined a JBS plant in Colorado less than $16,000. Meager as they were, these fines did not mark a departure from OSHA’s longstanding practice with regard to safety incidents in animal agriculture. Environment and climate change. Twenty-five years ago, J.B. Ruhl identified a glaring regulatory gap enjoyed by agriculture when it comes to environmental law, calling it the “vast ‘anti-law’ of farms and the environment.”141 The gap today is even wider as knowledge of the environmental harms caused by agriculture has strengthened and expanded, while the legal exemptions enjoyed by the industry have remained in place. The fact that animal agriculture’s environmental harms have become more widely recognized and documented is somewhat surprising, given that the industry enjoys an “administrative exemption” from federal reporting requirements for hazardous air releases under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), as well as the Emergency Planning and Community Right-to-Know Act (EPCRA). “Animal waste at farms” is simply deemed a hazard the EPA wishes to ignore.142 Thus, although both ammonia and 141

J.B. Ruhl, Farms, Their Environmental Harms, and Environmental Laws, 27 Ecology L. Q. 263, 267 (2000).

142

CERCLA/EPCRA Administrative Reporting Exemption for Air Releases of Hazardous Substances from Animal Waste at Farms, 73 Fed. Reg. 76948 (Dec. 18, 2008). A 2017 D.C. Circuit ruling vacated the two exemptions, Waterkeeper All. v. EPA, 853 F.3d 527 (D.C. Cir. 2017), but Congress responded one year later by codifying the CERCLA exemption through an act known as the Fair Agricultural Reporting Method Act. See Pub. L. No. 115141, §§1101-1103, 132 Stat. 1147 (2018). EPA subsequently used the Act’s CERCLA exemption as a basis for also exempting CAFOs under EPCRA. Amendment to Emergency Release Notification Regulations on Reporting Exemption for Air Emissions from Animal Waste at Farms; Emergency Planning and Community Right-to-Know

hydrogen sulfide are hazardous air pollutants subject to reporting mandates under CERCLA and EPCRA, releases of these gases from CAFOs are exempt. More broadly, both federal and state regulators lack basic information regarding the location, size, and pollution impacts of CAFOs.143 Following a 2013 agreement with EPA, regulators in Iowa conducted a three-year comprehensive manual satellite survey and were surprised to find 5,000 more CAFOs in Iowa than they were previously aware of, with approximately 25 percent of the newly discovered CAFOs considered medium or large in size.144 An ingenious machine learning approach devised by Daniel Ho and researchers at Stanford automates the process of CAFO identification within satellite images, dramatically reducing the cost and labor burden of detection. A 2019 study proved the technique by identifying 580 additional poultry CAFOs in North Carolina, an increase of 15 percent over a manual enumeration.145 A subsequent study adapted the technique further to create the first national opensource dataset of poultry CAFOs in the United States, identifying in the process a number of counties containing many CAFOs that had been reported to the U.S. Census of Agriculture as containing none.146 Even if there were perfect information on the location, size, and impacts of CAFOs, regulators would struggle to apply substantive environmental controls to them given the variety of exemptions agriculture enjoys under key laws. For instance, although the Clean Water Act sets ambitious national goals to control water pollution and improve water quality, the EPA has struggled mightily to apply the Act to CAFOs or cropland. The latter are considered “non-point sources” and are formally exempt from most regulatory controls. The former are actually defined as “point sources” in the Act and thus should be subjected to the National Pollution Elimination Discharge System (NPEDS). Under this system, point sources of water pollution must apply for permits and submit an adequate waste management program which is subject to agency review and citizen comment and appeal. In addition, if a facility is found to be in violation of its approved permit, the Act provides for citizen suit enforcement. A series of unfortunate events unfolded in the early 2000s, however, as the EPA attempted to promulgate a CAFO regulation under the Act. The final rule compromised substantially in industry’s favor from a proposed rule, but was still challenged in court by both industry and Act, 84 Fed. Reg. 27,533, 27,535 (June 13, 2019). The Biden administration has considered revoking the 2019 extension but has faced considerable opposition from industry and trade groups. https://www.nationalhogfarmer.com/livestock-management/nppc-no-legitimate-reason-for-farm-emissionsreporting-rule-. 143

Concentrated Animal Feeding Operations: EPA Needs More Information and A Clearly Defined Strategy to Protect Air and Water Quality from Pollutants of Concern Technical Report No. GAO-08-944 (US Government Accountability Office, 2008). 144

Ryan Levandowski, Polluting ‘til the Cows Come Home: How Agricultural Exceptionalism Allows CAFOs Free Range for Climate Harm, 33 Geo. Env’t L. Rev. 151, 159 (2020). 145

Handan-Nader, C., Ho, D.E. Deep learning to map concentrated animal feeding operations. Nat Sustain 2, 298– 306 (2019). https://doi.org/10.1038/s41893-019-0246-x 146

C. Robinson, B. Chugg, B. Anderson, J. M. L. Ferres and D. E. Ho, "Mapping Industrial Poultry Operations at Scale With Deep Learning and Aerial Imagery," in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 15, pp. 7458-7471, 2022, doi: 10.1109/JSTARS.2022.3191544

environmental groups. The resulting ruling narrowed EPA’s authority such that only CAFOs with known discharges, rather than a potential to discharge, could be required to obtain an NPEDS permit.147 A subsequent attempt by EPA to overcome this narrowing was challenged by the National Pork Producers Council, resulting in a court opinion affirming that “there must be an actual discharge into navigable waters to trigger the CWA’s requirements and the EPA’s authority.”148 This “actual discharge” requirement for CAFOs coupled with an exemption EPA provided for “agricultural stormwater” runoff has meant in practice that few CAFO facilities face meaningful CWA oversight. As one group of researchers concluded, “only after a number of catastrophic pollution events or repeated documented pollution problems will EPA or an authorized state require a CAFO to obtain NPDES permit coverage.”149 Like the CWA, the EPA technically has authority under the Clean Air Act (CAA) to regulate air pollutants from CAFOs, but the agency has never managed the gumption to use it. In 2005, the agency entered into a consent agreement with certain CAFOs offering them immunity from prosecution for violating CAA standards if they agreed to participate in a study designed to help EPA devise new CAFO air emissions modelling factors. Although the study was completed long ago, the EPA has not finalized its CAFO emissions model and thus very few CAFOs in the United States are regulated by the CAA despite the many known toxic and hazardous air pollutants released by such facilities.150 Of growing controversy is the EPA’s refusal to use the CAA or other authority to regulate greenhouse gas emissions relating to industrial animal agriculture.151 The agency’s 2009 reporting rule for greenhouse gas emissions exempted all but 100-110 of the very largest U.S. CAFO operators from reporting obligations. Even the CAFOs that are covered by the rule have been given de facto exemptions because Congress annually passes an appropriations rider that precludes EPA from using federal funds to enforce greenhouse gas emissions reporting requirements against CAFOs.152 Beyond reporting obligations, the EPA also could have begun developing substantive regulatory requirements for CAFOs following Massachusetts v. EPA and the agency’s endangerment finding regarding greenhouse gases. However, both the agency and Congress have

147

Waterkeeper Alliance, Inc. v. Environmental Protection Agency, 399 F.3d 486 (2d Cir. 2005).

148

National Pork Producers Council v. Environmental Prot. Agency, 635 F.3d 738, (5th Cir. 2011).

149

Diamond, D., L. Ashwood, A. Franco, A. Imlay, L.Kuehn, and C. Boutwell. 2022. “Farm Fiction: Agricultural Exceptionalism, Environmental Injustice and U.S. Right-to-Farm Laws.” Environmental Law Reporter, 52: 10727, 10741. 150

U.S. Envtl. Prot. Agency, Off. of Inspector Gen., Eleven Years After Agreement, EPA Has Not Developed Reliable Emission Estimation Methods to Determine Whether Animal Feeding Operations Comply with Clean Air Act and Other Statutes 11 (2017), https://perma. cc/6CKK-BHKR. 151

Ryan Levandowski, Polluting ‘til the Cows Come Home: How Agricultural Exceptionalism Allows CAFOs Free Range for Climate Harm, 33 Geo. Env’t L. Rev. 151, 159 (2020). 152

See Greenhouse Gas Reporting Program (GHGRP): Subpart JJ – Manure Management, U.S. Envtl. Prot. Agency, https://perma.cc/TZ7H-8TPB (“EPA is not implementing subpart JJ of 40 CFR Part 98 using funds provided in its appropriations due to a Congressional restriction prohibiting the expenditure of funds for this purpose.”). For an excellent history of the reporting rule saga with regard to animal agriculture, see Claire Regenstreif, Animal Agricultural Exceptionalism in the 21st Century, 37 J. Envtl. L. & Litig. 249, 268-283 (2022).

taken steps to mitigate this outcome,153 leaving the agricultural sector essentially unregulated with regard to its contributions to climate change. It should be noted that such agricultural climate exceptionalism is commonplace throughout the world. With its sixth assessment report on the state of climate change science, the Intergovernmental Panel on Climate Change firmly placed methane on the global policy agenda in advance of the twenty-sixth Conference of the Parties held in 2021 under the United Nations Framework Convention on Climate Change.154 In light of the blunt conclusions of the report, methane was highlighted as an area in which swift and significant emissions reductions realistically could be achieved,155 essentially buying humanity time as it struggles to tackle the more challenging transitions needed to reduce carbon dioxide and other GHG emissions.156 World leaders responded with the Global Methane Pledge, a voluntary initiative by over 100 countries to reduce global methane emissions at least 30 percent from 2020 levels by 2030.157 Animal agriculture was largely absent from these discussions, a significant omission given that experts believe the 2030 methane targets are unattainable without significant reductions in the agricultural sector. In the United States, legislative proposals to address climate change included a “methane pollution fee” that would be levied only on fossil fuel firms, not livestock producers.158 Likewise, the Biden Administration announced plans to impose new regulations to curb methane emissions from the oil and gas industry,159 but offered only “voluntary adoption of climate-smart agriculture practices”160 for the meat sector. The Global Methane Pledge echoed this stark differentiation by calling for regulatory “standards to achieve all feasible reductions in the energy and waste sectors” while only “seeking abatement of agricultural emissions through technology innovation as well as incentives and partnerships with farmers.”161 At its annual meeting following

153

For instance, Congress has used the appropriations process to prevent EPA from requiring Title V permits under the CAA for CAFO greenhouse gas emissions. The EPA has likewise denied petitions from environmental groups to develop new source performance standards for those emission, citing a lack of necessary data regarding all sources of CAFO air pollution and the need to develop a comprehensive regulatory approach. Letter from E. Scott Pruitt, Adm’r, Envtl. Prot. Agency, to Jonathan Lovvorn & Daniel Lutz, The Humane Soc’y of the United States, at 3-4 (Dec. 15, 2017), https://perma.cc/ZLZ4-ZGNV. 154

Intergovernmental Panel on Climate Change, ‘Sixth Assessment Report’, 7 August 2021.

155

Ibid SPM-36.

156

Ilissa B Ocko et al, ‘Acting rapidly to deploy readily available methane mitigation measures by sector can immediately slow global warming’ (2021) 16 Environ. Res. Lett. 054042. 157

European Commission, ‘Global Methane Pledge’, 2021.

158

Kellie Lunney, ‘House Energy Panel Advances Oil Industry Methane Pollution Fee’, 13 September 2021, Bloomberg Law. 159

Environmental Protection Agency, ‘Controlling Air Pollution from Oil and Natural Gas Industry’, 14 December 2021. 160

The White House, ‘Joint US-EU Press Release on the Global Methane Pledge’, 18 September 2021. In practice, the promised “climate-smart agricultural practices” became a government-subsidized greenwashing effort to allow meat producers to market their products as somehow being climate-friendly. https://newrepublic.com/article/175337/bs-behind-usdas-new-climate-friendly-beef-label. 161

the climate conference, the head of the U.S. beef lobby was quoted as saying, “Any way you slice it, that outcome from COP26 was a win for us.”162 Animal welfare. For an industry premised on the intense manipulation, confinement, and control of animals, surprisingly little animal welfare law applies. Indeed, David Wolfson and Marianne Sullivan famously concluded that “it is not unfair to say that, as a practical matter, farmed animals have no legal protection at all. As far as the law is concerned, they simply do not exist.”163 The federal Animal Welfare Act expressly exempts farmed animals from its coverage by defining “animals” not to include “farm animals, such as, but not limited to livestock or poultry, used or intended for use as food or fiber, or livestock or poultry used or intended for use for improving animal nutrition, breeding, management, or production efficiency, or for improving the quality of food or fiber.”164 No other federal law applies to the raising of farmed animals, as opposed to their transport or slaughter, and thus neither the Department of Agriculture nor any other federal agency has authority to promulgate regulations pertaining to the treatment of farmed animals during most of the time they are alive. The federal Humane Methods of Slaughter Act was added to the Meat Inspection Act in 1967 to declare that “[n]o method of slaughtering or handling in connection with slaughtering shall be deemed to comply with the public policy of the United States unless it is humane.”165 The Act expressly defines humane slaughter methods as those in which “all animals are rendered insensible to pain by a single blow or gunshot or an electrical, chemical or other means that is rapid and effective, before being shackled, hoisted, thrown, cast, or cut.”166 The Secretary of Agriculture is granted authority to issue regulations further designating slaughter methods that are humane and practical. The Act was aptly described by Senator Hubert Humphrey as “a mild and modest beginning in the field of humane slaughter.”167 One major shortcoming is that the Act’s general enforcement provision – which prevented the federal government from purchasing livestock products made from animals that were inhumanely slaughtered – was repealed in 1978.168 Even more surprising is the fact that the Act has been deemed by the Department of Agriculture not to apply to poultry, despite covering slaughter of “cattle, calves, horses, mules, sheep, swine, and other livestock.”169 In defending this remarkable exclusion of billions of livestock animals from protection,170 the agency pointed to the Poultry Products Inspection Act and argued that the latter statute mandates poultry slaughter “in a manner that is consistent with 162

Zach Boren, ‘Beef lobbyists celebrate methane ‘win’ at COP26,’ 7 March 2021, Unearthed.

163

Foxes in the Henhouse.

164

7 U.S.C.A. § 2132(g).

165

7 U.S.C.A. § 1901.

166

7 U.S.C.A. § 1902.

167

Levine v. Conner, 540 F.Supp.2d 1113, 1119 (2008), vacated and remanded Levine v. Vilsack, 587 F.3d 986 (2009). 168

Levine v. Vilsack, 587 F.3d 986, 992 (2009). The only remaining enforcement provision in the Act pertains to the treatment of nonambulatory livestock. 169

7 U.S.C.A. § 1902(a) (emphasis added).

170

“Treatment of Live Poultry Before Slaughter.” 70 Fed. Reg. 56,624 (Sept. 28, 2005).

good commercial practices, which means they should be treated humanely.”171 As animal law scholar and advocate Jonathan Lovvorn notes, however, this rather thin reed of protection likely means that chickens, ducks, geese, and other birds can be slaughtered without first being rendered unconscious or otherwise insensible to pain.172 This de facto exemption is especially convenient for the poultry industry since stun baths not infrequently fail to work. Other than very limited federal laws regarding livestock treatment during interstate transport,173 the primary method by which cruel and inhumane treatment of farmed animals might be challenged is through state anti-cruelty statutes. All fifty states have statutes that criminalize acts of animal cruelty. However, the use of a criminal statute to achieve a regulatory goal like the reduction of animal suffering throughout an industry is limited by design, as Wolfson and Sullivan noted: Contrary to regulatory schemes generally set up by legislatures to govern industry conduct, criminal anticruelty statutes which govern the farming industry’s treatment of animals do not provide for the promulgation of specific regulations to govern animal welfare, and the farming industry is not subject to any sort of regulatory enforcement of farmed-animal welfare standards, does not undergo any inspections to determine whether farmed animals are being afforded appropriate treatment, and is not answerable to any governmental administrative agency (federal or state) on the subject of farmed-animal welfare. In addition, the burden of proof on the prosecution is very high, that is, beyond a reasonable doubt. Moreover, unlike regulatory statutes, criminal anticruelty statutes necessarily require that the prosecution demonstrate a mental state on the part of the defendant that may be hard to prove. To make matters worse, many state anti-cruelty statutes exempt animal agriculture entirely from coverage, while those statutes that do apply to farmed animals still exempt any act that is considered a “normal,” “common,” or “customary” agricultural practice. For instance, Colorado’s statute excludes from coverage “accepted animal husbandry practices utilized by any person in the care of companion or livestock animals.”174 Likewise, the anti-cruelty statute of Illinois provides that “[n]othing in this Act affects normal, good husbandry practices utilized by any person in the production of food.”175 In Nevada, “[t]he provisions of [the anti-cruelty statute] do not . . .

171

“Treatment of Live Poultry Before Slaughter.” 70 Fed. Reg. 56,624 (Sept. 28, 2005). A subsequent FSIS directive clarified that enforcement against poultry facilities should only occur in cases of an “ongoing pattern or trend.” FSIS Directive 6110.1, Verification of Poultry Good Commercial Practices (U.S.D.A. 2018). 172

See Jonathan R. Lovvorn, Animal Law in Action: The Law, Public Perception, and the Limits of Animal Rights as a Basis for Legal Reform, 12 Animal L. 133, 137-38 (2005). 173

Twenty-Eight Hour Law, 49 U.S.C. § 80502(a)–(b) (2016). First passed in 1873, this Act prohibits livestock animals from being transported across-state lines for more than 28 hours without “feeding, water, and rest.” Advocates have long criticized the Twenty-Eight Hour Law as out-of-date and underenforced; between 2006 and 2016, there were only 10 USDA enforcement inquiries and, in any event, violations are subject to a maximum $500 civil penalty per violation. 49 U.S.C. § 80502(d). 174

Colo. Rev. Stat. § 18-9-201.5(1) (2017).

175

510 Ill. Comp. Stat. 70/13 (2015).

[p]rohibit or interfere with established methods of animal husbandry, including the raising, handling, feeding, housing, and transporting of live-stock or farm animals.”176 The animal welfare concerns described above in Part II all result from practices that are considered customary within industrial animal agriculture. It is customary, for instance, to starve laying hens for one week or more as their productivity declines in order to force molting and a reset of their reproductive cycles. If this act was done to a pet chicken in someone’s backyard, it would constitute a crime, in many states a felony. When done to thousands of chickens within a CAFO, however, it becomes immunized as a standard industry practice. Tort. A bedrock principle of the American legal system is that individuals retain a right to seek an official determination when they believe they have been wronged. Thoughtful commentators would go farther and argue that individuals have a right of access to a substantive body of law that provides meaningful redress for wrongs. Since its founding, the U.S. legal system has largely fulfilled these government responsibilities through the common law of tort.177 With respect to animal agriculture facilities, disputes with neighbors regarding the noxious or harmful impacts of production long have been overseen by the centuries-old framework of nuisance and trespass law.178 Beginning in the 1970s and 80s, agricultural producers decided they wanted to change this framework. As suburban sprawl increasingly placed long-standing farmland in danger of being challenged in tort, the farm lobby secured through state legislatures so-called “right to farm” laws. Most modestly, these laws sought to codify the common law “coming to the nuisance” doctrine as a defense to protect established farms from complaints by newcomers. But over time, as the agriculture sector increasingly consolidated and industrialized, the laws were expanded to provide an ever-widening sphere of protection for agriculture, including for CAFOs.179 As agriculture law expert Neil Hamilton explained, “the reality is that many states took the [original] idea and applied it much broadly to protect any livestock facility meeting regulatory requirements, even those in existence after the neighboring homes. This action, in effect, turned the laws into a ‘right to commit nuisance’ rather than a right to farm.”180 The American Legislative Exchange Council’s model “right to farm” law provides a laundry list of the kinds of provisions states have adopted: (a) a defense for “generally accepted agriculture and management practices”; (b) an expansion of the “coming to the nuisance” defense to cover pre-existing farms even if they change ownership or size, adopt new technologies, or switch to entirely different modes of production; and (c) fee-shifting to defendants in the case of 176

Nev. Rev. Stat. § 574.200(1)(f) (2015).

177 178 179

For an excellent study of the history, design, and litigation impact of right to farm laws, see Diamond, D., L. Ashwood, A. Franco, A. Imlay, L.Kuehn, and C. Boutwell. 2022. “Farm Fiction: Agricultural Exceptionalism, Environmental Injustice and U.S. Right-to-Farm Laws.” Environmental Law Reporter, 52: 10727, 10741. 180

Neil D. Hamilton, Harvesting the Law: Personal Reflections on Thirty Years of Change in Agricultural Legislation, 46 Creighton L. Rev. 563, 577–78 (2013). See also Jonathan Morris, One Ought Not Have So Delicate a Nose: CAFOs, Agricultural Nuisance, and the Rise of the Right to Farm, 47 Envtl. L. 261 (2017).

losing complaints.181 Building on these examples, states have adopted provisions that exempt all farming operations that have been in operation for a certain period of time, usually one or two years, irrespective of whether the operation pre-dated the complaining neighbor’s arrival. The combination of this exemption with the exemption for changes in size or manner of production can have dramatic consequences: “[I]f a parcel of land had been used for row crops for many years and then transitions into a 20,000-head cattle feedlot, surrounding neighbors would be barred from bringing a nuisance suit against the feedlot, even if they attempted to do so within the first year of the feedlot’s existence.”182 Pennsylvania’s right to farm law offers even more creative protection by allowing for substantial operational changes to be immunized so long as they continue for one year before suit or are subject to a waste management program approved by the state agency. Such waste management programs, however, are notoriously inadequate in design and enforcement. Finally, in order to consolidate regulatory authority further into federal and state agencies long criticized as captured, many right to farm laws expressly preempt the power of local governments to declare farming operations nuisances. Still others prohibit local governments from using zoning ordinances, construction permits, or other laws to regulate agriculture more broadly. Such laws are especially pertinent in light of numerous municipal and county governments throughout the United States that have attempted to ban or more thoughtfully consider the appropriate scale and mode of industrial agricultural land use within their jurisdictions. The combination of agricultural exceptionalism in state and federal regulation, robust protection from tort law, and preemption of local governmental power creates on open invitation to externalize that CAFOs find difficult to resist: “If an agricultural facility is compliant with state and federal laws and is effectively insulated from nuisance liability, what reason does it have to consider the harm suffered by neighbors as a result of its operations?”183 Political accountability. One reason a CAFO might have to consider the harms it externalizes would be a shift in the political winds sufficient to undo the many legal exemptions and protections that the industry enjoys. In order to help forestall such dynamics, the animal agriculture industry has invested in the procurement of so-called “ag-gag” laws which impose civil and criminal penalties on actions that would otherwise be standard practices used by journalists, activists, or employee whistleblowers.184 Specifically, the laws prohibit photographing and recording without the consent of a facility owner, and accessing facilities under false pretenses, through deception, or without consent. The latter prohibition is designed to target undercover investigators, including academics,185 who obtain access to facilities by getting hired as workers.

181

https://alec.org/model-policy/right-to-farm-act/

182

Harvard study by Diamond et al. As one court put it, this protection means that the traditional “coming to the nuisance doctrine” now includes coming to “the potential future nuisance.” Himsel v. Himsel, 122 N.E.3d 935 (Ind. Ct. App. 2019) (holding that Indiana right to farm law immunized farm from nuisance liability after it converted from row crops to an 8,000 head hog CAFO). 183

Jonathan Morris, One Ought Not Have So Delicate a Nose: CAFOs, Agricultural Nuisance, and the Rise of the Right to Farm, 47 Envtl. L. 261, 286 (2017). 184

Justin Marceau

185

Timothy Pachirat, Every Twelve Seconds: Industrialized Slaughter and the Politics of Sight (2011) (ethnographic analysis by a political scientist who worked undercover at a Great Plains slaughterhouse for five months).

Whereas the common law would regard the employer’s consent as a valid defense to trespass in such cases, ag-gag laws vitiate it. The most ingenious provisions within ag-gag laws are mandatory reporting periods for documented abuses. Missouri’s law, for instance, states that if a “farm animal professional videotapes or otherwise makes a digital recording of what he or she believes to depict a farm animal subjected to abuse or neglect,” the employee must provide the recording to law enforcement within 24 hours of recording or face criminal charges.186 While this measure might seem to be a well-intended effort to alert authorities to instances of farmed animal mistreatment, its effect in application is quite the opposite. Non-undercover employees who might otherwise have been motivated to document what they viewed as troubling practices by their employers now have a strong reason not to do so, since the law only requires alerting law enforcement if a recording is made. More important, undercover employees no longer have the ability to stay in place and build a case for sustained or pattern-and-practice mistreatment at a facility. If they stay undercover, they face later criminal prosecution for not timely disclosing their initial findings. If they disclose, they enable the facility to cover up what might be a more systemic problem by blaming a rogue employee or otherwise characterizing the incident as isolated. Ag-gag laws are representative of a broader class of laws championed by agriculture that are designed to quell criticism, stifle competition, and otherwise preserve the industry’s vast political power. A number of states, for instance, have adopted agricultural disparagement laws which provide a cause of action to agricultural producers that is more advantageous than common law libel and defamation torts. The laws typically attempt to lower the standard of liability below the constitutional “actual malice” standard. Many states also provide for punitive damages and fee-shifting to prevailing plaintiffs. Similarly, most states with agricultural disparagement laws shift the burden of proof onto defendants to demonstrate that a critical statement concerning an agricultural product was true, where “truth” is defined with reference to “reasonable” or “reliable” scientific inquiry. Such definitions can make it difficult for a defendant to prevail in the case of statements premised on emerging science and, in any event, necessitate hiring expert witnesses and other costly efforts to build an evidentiary record. The chilling effect potential of ag-gag laws is illustrated by the laws’ most famous defendant, Oprah Winfrey. In 1996, the talk show host spoke with a guest Howard Lyman, a farmer and animal rights activist, about safety concerns of beef consumption in light of the mad cow disease outbreak happening in Europe and the United Kingdom. Winfrey and Lyman were subsequently sued for over $10 million in damages by several cattle producers who alleged not only defamation and negligence causes of action, but also a violation of Texas’s False Disparagement of Perishable Food Products Act.187 Although Winfrey prevailed in a directed verdict, she first had to mount a costly defense and endure a lengthy trial which prompted her to remark: “To be on trial, may I just say, is one of the worst experiences of anybody’s life.”188

186

Mo. Rev. Stat § 578.013(1).

187

Texas Beef Grp. v. Winfrey, 11 F. Supp. 2d 858 (N.D. Tex. 1998), aff'd, 201 F.3d 680 (5th Cir. 2000).

188

https://www.keranews.org/texas-news/2018-01-10/the-time-oprah-winfrey-beefed-with-the-texas-cattle-industry

Winfrey has never spoken publicly on mad cow disease again nor has she made the original episode available for rebroadcast or other use. The latest legislative trend in this category is the adoption of state laws that attempt to protect the meat industry from competition by alternative protein suppliers. Florida, for instance, recently adopted a law banning the sale of cultivated or lab-grown meat, with the state’s governor proclaiming “We will save our beef” in a press release following his signing of the bill.189 Several states have adopted or are considering laws that restrict the use of terms like “meat,” “milk,” or “dairy” in conjunction with products not derived directly from live animals. At the federal level, several U.S. Senators in 2023 proposed a bill that would prohibit food manufacturers from using “milk,” “cheese,” “yogurt,” or any other “market name for a dairy product” to describe their products unless they are derived from the “lacteal secretions” of “hooved animals.” As seems to be required in today’s Congress, the bill bears a convoluted title which generates a pithy acronym: “‘Defending Against Imitations and Replacements of Yogurt, Milk, and Cheese To Promote Regular Intake of Dairy Everyday Act’ or the ‘DAIRY PRIDE Act.’”190 IV. “THE NATURE OF THE FARM” AND “THE PROBLEM OF SOCIALIZED COST” “Clean air is a God-given right, but it’s a right that my family and I are being denied.” – Elsie Herring, neighbor of an industrial North Carolina hog facility “We should get down on our knees and thank our heavenly Father. People put up with production to enjoy the benefits of consumption.” – Jimmy Dixon, hog farmer and North Carolina state legislator191 The analysis thus far raises a puzzle. Industrial animal agriculture appears to impose harms on consumers, public health, workers, farmers, neighbors, ecosystems, animals, natural resources, and climate stability. In some cases, these harms are of immediate and devastating human consequence. In other cases, they raise questions that go to the heart of the moral character of our species. In the particular case of climate change, the threatened harm runs even deeper to the very habitability of the planet. Yet somehow, despite this catalogue of consequences, the industry has managed to not only largely avoid legal and political accountability, but also to continue to benefit

189

https://www.flgov.com/2024/05/01/governor-desantis-signs-legislation-to-keep-lab-grown-meat-out-of-florida/ (“‘Today, Florida is fighting back against the global elite’s plan to force the world to eat meat grown in a petri dish or bugs to achieve their authoritarian goals,’ said Governor Ron DeSantis. ‘Our administration will continue to focus on investing in our local farmers and ranchers, and we will save our beef.’”). Alabama also has banned the manufacture, sale, and distribution of cultivated meat food products in the state, punishable by criminal and civil penalties for violations. https://perma.cc/6T6Y-D3QV 190

https://www.congress.gov/bill/118th-congress/senate-bill/549/text?s=1&r=30

191

https://ncnewsline.com/2017/11/15/neutering-nuisance-laws-north-carolina/.

from massive levels of public subsidization across the globe.192 What aspects of our thinking about law, politics, and economics help this situation to persist? To address that question, this Part turns to familiar but still fruitful theoretical terrain. With two deceptively simple papers, one written while still an undergraduate student, Ronald Coase exerted a profound influence over how we think about markets, law, and policy. Together these papers offer a view of social ordering that depends heavily on private ownership and contracting, with a relatively narrow – or at least hesitant – role for government regulation. The analytical engine driving both papers is the concept of transaction costs which influences whether and when actors will contract to exchange valuable entitlements. Though these ideas now seem omnipresent, Coase’s 1991 Nobel Memorial Prize in Economic Sciences was awarded “for his discovery and clarification of the significance of transaction costs and property rights for the institutional structure and functioning of the economy.”193 This Part uses Coase’s pioneering logic to explain features of the contemporary meat sector that otherwise appear difficult to explain. As will be seen, the explanatory power of Coase’s framework may be greater than even he perceived. The Coasean Framework. In The Nature of the Firm, Coase addressed a tension within economic theory between societal allocation of goods and services in a market economy and the allocative decision making that occurs within an organization. As Coase noted, at the aggregate level, “[a]n economist thinks of the economic system as being co-ordinated by the price mechanism and society becomes not an organisation but an organism.”194 Yet within a firm, managers routinely make decisions based on a variety of factors, only some of which are related to the price mechanism. Indeed, as Coase more strongly put it, “the distinguishing mark of the firm is the supersession of the price mechanism.”195 What could explain the prevalence of administrative or managerial decision making within firms when those organizations’ external fortunes are dictated by the iron laws of supply and demand? For Coase, this question was tied to a larger mystery regarding why firms emerge within market economies at all. Why should any production occur through organizations rather than through elaborate webs of contracts among individual producers? The answer lies in the fact that “there is a cost of using the price mechanism.”196 While his contemporaries within economics often assumed a “frictionless” economy in which all goods and services could be exchanged without cost, Coase recognized that in the real world a variety of transaction costs exist that can impede otherwise valuable exchanges. Thus, when deciding how to obtain a factor of production, the entrepreneur faces a choice between bearing some portion of the transaction costs of contracting 192

According to analysis by the Environmental Working Group, the U.S. Department of Agriculture spent an estimated $59 billion in subsidies for livestock and seafood producers between 1995 and 2021. See https://www.ewg.org/news-insights/news/2023/08/usda-livestock-subsidies-top-59billion#:~:text=The%20Department%20of%20Agriculture%20has,other%20alternatives%20to%20animal%20protei ns. Globally, almost one-third of agricultural subsidies support the meat and dairy industries. Marco Springmann & Fabian Freund, Options for Reforming Agricultural Subsidies from Health, Climate, and Economic Perspectives, 13 Nature Commc’ns 1, 2 (2022). 193

https://www.nobelprize.org/prizes/economic-sciences/1991/coase/facts/ (emphasis added).

194

Coase, The Nature of the Firm, at p. 387

195

Coase, The Nature of the Firm, at p. 389.

196

Coase, The Nature of the Firm, at p. 390.

with an external producer or investing in the capacity to produce the needed factor internally within the context of an organization. In the latter case, the entrepreneur saves on transaction costs but takes on the costs and risks associated with management and production. In Coase’s elegant distillation, “[t]he question always is, will it pay to bring an extra exchange transaction under the organising authority?” It is this core insight that Coase uses to explain the emergence and the optimal size of the modern firm. Transaction costs also play a critical role in Coase’s monumental article, The Problem of Social Cost. In this work, Coase grapples with the concept of the negative externality, first introduced by Alfred Marshall but brought to prominence in the work of Alfred Pigou. Prior economists like Pigou had viewed the “actions of business firms which have harmful effects on others” as representing “a divergence between the private and social product of the [business].”197 The standard policy response in such cases was to make the firm liable in tort law for the harm caused, to tax the firm at an amount representing the monetary equivalent of the harm caused, or to engage in more direct control of the firm’s activities through an injunction or regulation. Coase disrupted this standard analysis by first highlighting what he termed “the reciprocal nature of the problem.”198 Whereas Pigou had seen the issue of negative externalities through moralized terms like polluter and victim, Coase sees only two actors who both want to utilize a finite resource. In the case of pollution, the factory wants to use the air as a repository for waste while the affected neighbors want to preserve it for health and well-being. To Pigou, the factory is a wrongdoer that needs to be disciplined through a pollution tax to ensure that its activities are a net benefit to society. To Coase, neither party has the moral high ground. The question instead is, how can society ensure that the higher-valued user gains access to the contested resource?199 Coase reinforced this framing of the question through a second example closer to the heart of this Article: [Consider] the problem of straying cattle which destroy crops on neighbouring land. If it is inevitable that some cattle will stray, an increase in the supply of meat can only be obtained at the expense of a decrease in the supply of crops. The nature of the choice is clear: meat or crops. What answer should be given is, of course, not clear unless we know the value of what is obtained as well as the value of what is sacrificed to obtain it.200 The genius in Coase’s contribution lies in his recognition that within market societies the “victims” of pollution (or trespass) can contract with a factory (or cattle farmer) to “purchase” unpolluted air (or untrammeled cropland). Conversely, if a court orders an injunction against harm-causing activities, an affected firm can respond by contracting with plaintiffs to surrender their right to enforce the injunction in exchange for monetary payment. In either case, Coase’s brilliance lies in 197

Coase, The Problem of Social Cost, at p. 1.

198

Coase, The Problem of Social Cost, at p. 2.

199

Coase, The Problem of Social Cost, at p. 15 (“The economic problem in all cases of harmful effects is how to maximise the value of production.”). 200

Coase, The Problem of Social Cost, at p. 2.

extending the logic of market exchange to the domain of legal entitlements, viewing such entitlements both as factors of production and as subjects of contractual exchange.201 This insight leads him to perceive that, in a world free of transaction costs, resources will end up devoted to their highest valued use, irrespective of whether an entitlement is initially allocated to a polluter or to its victims.202 The hitch once again, though, is that the real world is decidedly not without friction. When sizable enough, transaction costs will prevent otherwise welfare-enhancing exchanges from taking place. Thus, when a court or a regulator considers whether to order abatement or damages in the case of pollution, the decision matters very much to the ultimate allocation of a legal entitlement. If the entitlement is given to a lesser valued user when transaction costs exceed the gains to trade, then the optimal use of the resource will not prevail. For damages awards or taxes, the government’s task is complicated further by the need to estimate in economic terms the value of the harm being sanctioned. Because Coase thought that government decision makers were unlikely to have good information about either transaction costs or the economic values at stake in resource conflicts, he feared that the interventions of courts and agencies often would be worse than if the government were “to do nothing about the problem at all.”203 Put succinctly, “[a]ll solutions have costs and there is no reason to suppose that government regulation is called for simply because the problem is not well handled by the market or the firm.”204 The Framework Furthered. It is fitting that meat production featured in Coase’s landmark work, albeit in the form of a stylized conflict between a rancher and a crop farmer reflecting traditional modes of agriculture that were already disappearing by 1960 when The Problem of Social Cost was published. A close examination of the contemporary meat industry reveals additional implications of the Coasean framework, ones that complicate further the task of promoting overall social welfare. To begin with, there is little reason to suppose that firms engage in decisions regarding internal and external formation only when it comes to value-creating factors of production. In The Nature of the Firm, Coase elegantly demonstrates that firms bring positive factors of production in-house when the management costs of doing so are outweighed by the transaction costs of obtaining those factors externally through the market. But what about the value-destroying consequences of production? Whether in the form of worker injuries, local air and water pollution, infectious disease, climate change, or otherwise, firms have obvious economic incentives to avoid taking financial or legal responsibility for the harmful impacts of their activities. As much would have been obvious to Pigou and Coase, but those and other economists have tended to view negative externalities as a relatively narrow category of unintentional side-effects of production. 201

Coase, The Problem of Social Cost, at p. 44 ((“If factors of production are thought of as rights, it becomes easier to understand that the right to do something which has a harmful effect (such as the creation of smoke, noise, smells, etc.) is also a factor of production.”). 202

Coase, The Problem of Social Cost, at p. 10 (“The judges’ view that they were settling how the land was to be used would be true only in the case in which the costs of carrying out the necessary market transactions exceeded the gain which might be achieved by any rearrangement of rights.”). 203

Coase, The Problem of Social Cost, at p. 18.

204

Coase, The Problem of Social Cost, at p. 18.

The modern meat sector instead seems to have placed cost externalization at the core of its business model, viewing the relegation of negative consequences of production to others as a central organizational objective.205 This is understandable given just how vast and varied those negative consequences are. An effort by the Rockefeller Foundation to rigorously quantify and monetize the negative externalities of the U.S. food system concluded that, while consumers spent $1.1 trillion on food in 2019, that same year the U.S. food system generated $2.1 trillion worth of social and environmental costs.206 In other words, the economic value of what was destroyed by the U.S. food system in 2019 was roughly twice the amount paid for what was created. It is an open question whether consumers would have been willing to pay $3.2 trillion for the food they purchased that year, such that nothing about the system of food production would have changed even in the presence of Pigouvian-style cost internalization. What seems likely, though, is that the organization and output of the agriculture sector would have been vastly different if the price mechanism had incorporated the full social costs of production.207 This line of analysis runs head-on into to the critique that Coase levied against Pigou: Merely identifying a divergence between the “private” and “social” product of an industry does not, in Coase’s view, by itself raise concerns or compel a government response.208 For one thing, the government’s proposed cure could be worse than the market disease.209 It also could be that the various interests impacted by modern agriculture have already attempted, and failed, to alleviate their harms through private bargaining. In that case, the persistence of the harms would merely indicate that the costs for firms of avoiding harm outweighed the amount that harmsufferers were willing to pay to induce a rearrangement of production. As Coase stated in response to economists in the Pigouvian tradition who instinctively sought cost-internalization when

205

Ashwood, L., Diamond, D. and Thu, K. (2014), Where's the Farmer?. Rural Sociol, 79: 227. https://doi.org/10.1111/ruso.12026 206

The Rockefeller Foundation, True Cost of Food: Measuring What Matters to Transform the U.S. Food System (July 2021). It bears stressing that the Rockefeller report analyzed the U.S. food system as a whole, rather than only the industrial animal agriculture sector. Evidence specific to the meat sector, however, reinforces the report’s concerns. See, e.g., Pieper, M., Michalke, A. & Gaugler, T. Calculation of external climate costs for food highlights inadequate pricing of animal products. Nat. Commun. 11, 6117 (2020). https://doi.org/10.1038/s41467-020-19474-6 (finding that if greenhouse gas emissions impacts were incorporated into the price, meat would cost 146 percent more than at present). 207

The Food and Land Use Coalition conducted a similar externality analysis for the global food system, concluding that the system’s environmental and social costs total nearly $12 trillion a year, a number that exceeds the market value of the system’s global output. See Food and Land Use Coalition, Growing Better: Ten Critical Transitions to Transform Food and Land Use 37 (2019), available at https://www.foodandlandusecoalition.org/wpcontent/uploads/2019/09/FOLU-GrowingBetter-GlobalReport.pdf. 208

In Coase’s more strident phrasing, “[n]othing could be more ‘anti-social’ than to oppose any action which causes any harm to anyone.” Coase, The Problem of Social Cost, at p. 35. 209

Coase, The Problem of Social Cost, at p. 44 (“[I]n choosing between social arrangements within the context of which individual decisions are made, we have to bear in mind that a change in the existing system which will lead to an improvement in some decisions may well lead to a worsening of others.”).

confronted with social harms, “the activities which they would like to see stopped or curtailed may well be socially justified.”210 Is modern meat production “socially justified”? The remainder of this Part argues “no.” The ability of victims of harm to bargain for relief depends on the framework of legal entitlements and transaction costs in a different and even more vexing way than Coase appreciated. The allocation of legal entitlements and the incidence and significance of transaction costs are not exogenous factors to firm decision making. Rather, the landscape of entitlements and transaction costs that confounds the problem of social cost is one that has been and will be intentionally shaped by economic actors. As a result, when victims suffer more than the productive value lost in order to abate a harm, the harm-causing firm still might seek to maximize its private position by altering the bargaining landscape rather than agreeing upon an exchange. Valuation and Indeterminacy. At first blush, this claim seems ridiculous. When would it ever make sense for a rational economic actor to raise the cost of transferring resources to a highervalued use? The key lies in the diverse ways value can be expressed. In Coase’s stylized example, the farmer and the cattle-raiser bargain over land access by expressing their willingness to pay for the resource. In Coase’s rendering, both parties have the means to adequately express their valuation in these terms, such that no question is raised regarding the appropriate method of measuring social value. Likewise, because the stakes at issue are not believed to fundamentally alter the actors’ real or perceived economic position, the values expressed through bargaining are not seen to depend on which actor is first granted the legal entitlement. Valuation in the real world is a much messier and ethically fraught exercise. For most of the world’s population, willingness to pay to obtain an entitlement is critically dependent on budget constraints. Thus, whether one is initially granted a legal entitlement in many cases strongly influences the valuation one is able to attach to it.211 When an individual is granted a legal entitlement, they can demand a price equal to their willingness-to-accept valuation before surrendering it, rather than only offer what they are willing (or able) to pay to acquire the entitlement. In textbook economic conditions, this gap between willingness-to-accept and willingness-to-pay valuations disappears, but in the real world it remains as a yawning chasm. Indeed, one significant reason for the persistence of valuation gaps in the real world is that many goods – such as human health and environmental quality – have a low substitution elasticity, such that one’s expressed valuation is affected very much by whether one is first endowed with the good at issue.212 Given the right to clean air and water, one is likely to demand far more to surrender it than one would be willing and able to pay to purchase it. Consider a simple illustration. As noted above, living near a CAFO is associated with higher mortality risk from various causes including cardiovascular disease, anemia, and kidney disease.213 Assume that exposed neighbors of a particular CAFO would be willing to pay $100 210

Coase, The Problem of Social Cost, at p. 26.

211

See generally Elizabeth Hoffman and Matthew L. Spitzer, Willingness to Pay vs. Willingness to Accept: Legal and Economic Implications, 71 WASH. U. L. Q. 59 (1993). 212

See id. at __.

213

collectively per year to abate this risk. Assume further that the CAFO earns $1000 per year from its operations and could install pollution controls that would eliminate the neighbors’ risk with an amortized cost of $500 per year. Based on the neighbors’ expressed willingness-to-pay valuation, the harm caused by the pollution is outweighed by the cost of avoiding it. Thus, per Coase, the problem of social cost turns out not to be a problem at all. The CAFO is the higher-valued user of the local environment and society is apparently better off not spending $500 to abate a nuisance that only causes $100 of harm. But what if the neighbors are first endowed with a right to clean air and water in the local environment, such that the CAFO has to purchase this right from them in order to continue operations without installing pollution controls? Suppose the neighbors would only be willing to give up their entitlement to be free of excess mortality risk for a payment of $750 per year. At that point, the CAFO would spend $500 annually for pollution controls rather than make the payment of $750 per year. Net profits would shrink to $500 annually for the CAFO, and society would once again be better off since the harm valued at $750 would be avoided through a $500 investment. In both cases society will appear to have maximized overall welfare, but the identity of the “highervalued user” has completely reversed. Legal Entitlements and Transaction Costs as Factors of Production. This latent circularity or entitlement-dependency is well understood in relation to the Coase Theorem.214 What is less widely appreciated is that the gap between valuations also opens opportunities for actors to use law more affirmatively to shape outcomes. For instance, assume again that neighbors are endowed with the right to be free from excess mortality risk, but that the CAFO could persuade lawmakers to reverse the initial entitlement allocation by investing $200 in annual lobbying efforts. As that point, rather than spend $500 per year on pollution controls, the CAFO would spend $200 on legislative influence, altering the legal backdrop such that the harm imposed on the neighbors shrinks from $750 to $100 in apparent value. The CAFO’s annual profits would be reduced from $1000 to $800, but the farm would much prefer that result to the $500 it would net by installing and operating pollution controls (Table 1).

214

Table 1 Initial entitlement to CAFO (WTP)

Initial entitlement to neighbors (WTA)

Initial entitlement to neighbors but CAFO lobbies to change law

CAFO initial profits

$1000

Valuation of harm

$100 (harm suffered)

$750 (harm avoided)

$100 (harm suffered)

Cost of control

$500 (not adopted)

$500 (adopted)

$500 (not adopted)

Cost of lobbying

n/a

$200

CAFO net outome

$1000

$500

$800

Neighbors net outcome

-$100

Social net outcome

$900

$500

$700

A striking feature of this thought exercise is that the net social outcome appears to be maximized by not allowing the neighbors to express their valuation of harm through a willingnessto-accept frame. That is, both the farm’s private outcome and the overall social product appear to be higher when the victims of industrial activity are placed in an initial position of legal disempowerment. This observation raises profound questions about the proper calculation of social welfare, the just distribution of entitlements, and the appropriate scope of moral and political considerations in legal economic analysis. The focus for now, though, remains on the narrower finding that – at least on the example’s assumptions – firms can avoid social welfare-maximizing outcomes by investing in legal change. Indeed, they can shift the very definition of what outcomes are social welfare maximizing by altering the law’s initial allocation of entitlements. By taking property rights as given and by focusing only on impediments to voluntary exchange, Coase did not explore this conceptual possibility. Still, even if conceptually possible, is this scenario realistic? Arguably, the aftermath of a series of successful hog farm nuisance suits in North Carolina is directly on point. Beginning in 2013, claims were brought by North Carolina trial lawyer Mona Lisa Wallace on behalf of 26 communities against the Smithfield Foods subsidiary, Murphy-Brown, alleging that the company’s pigs were being housed in facilities that posed environmental, health, and safety risks to

neighboring households.215 The industry first sought to block the suits by obtaining a legislative bill that narrowed the definition of nuisance in agricultural contexts, but a federal court held that the law did not apply retroactively.216 After years of procedural wrangling, plaintiffs proceeded to win unanimous and sizable jury awards in five separate bellwether trials, including a $473 million punitive damages award in one case.217 Once plaintiffs survived an appeal to the Fourth Circuit, garnering in the process scathing criticism of the defendant in the concurring opinion by Judge Wilkinson quoted above, the industry turned back to the state legislature. There the industry obtained, over a gubernatorial veto, further amendments to North Carolina’s right-to-farm laws that drastically curtailed the ability of future plaintiffs to seek relief against harmful agricultural operations. The most potent piece of corporate protective equipment adopted was a limitation on damages to the diminution in fair market value of the impacted property that is attributable to the nuisance.218 Because North Carolina’s CAFOs are disproportionately located in poor and minority communities where property values are low,219 this amendment had the effect of making future hog farm nuisance litigation infeasible for plaintiffs’ lawyers to finance. Essentially, the amendment had the effect of moving neighbors of North Carolina hog farms from the second to the third column of Table 1. Other methods exist by which a firm can profitably invest in changing the landscape of bargaining rather than undertake socially welfare maximizing behavior (Table 2). Assume, for instance, that the affected neighbors are actually well-off and have no difficulty expressing their valuation of the harm through a willingness-to-pay frame at the higher level of $750. In that case, when the CAFO is allocated the right to cause harm and transaction costs are not an impediment to exchange, the neighbors can band together to offer the CAFO $500 to install pollution controls 215

For an engaging journalistic account of these lawsuits – complete with a foreword by John Grisham, natch – see Corban Addison, Wastelands: The True Story of Farm Country on Trial (Penguin Random House, 2022). 216 217

See McKiver v. Murphy-Brown, LLC , E.D.N.C., No. 14-CV-180-BR (Verdict: $51 million [reduced to $3 million] for 10 plaintiffs); Williams vs. Murphy-Brown, LLC, E.D.N.C., 7:14-CV-180-BR (Verdict: $25 million [reduced to $630,000]; Artis v. Murphy-Brown, LLC, No. 7:14-CV-237-BR (Verdict: $473.5 million [reduced by statute to $94 million]; Gillis v. Murphy-Brown, LLC, E.D.N.C., No. 14-cv-185-BR (Verdict: compensatory damages from $100-$75,000 per plaintiff); McGowan v. Murphy-Brown, LLC, No. 7:14-CV-182-BR (Verdict: damages totaling $420,000). 218

An Act to Clarify the Remedies Available in Private Nuisance Actions Against Agricultural and Forestry Operations 2017 N.C. Sess. Laws 11 (“H.B. 467”), codified as N.C. Gen. Stat. § 106-702. Additional restrictions included immunity for existing operations even if they undergo a change in size, technology, or type of agriculture practiced, a strict one-year limitations period for the narrow category of claims not immunized, and a requirement that plaintiffs live within one-half mile of the source of the nuisance and hold a formal legal possessory interest in the affected household. An Act to Make Various Changes to the Agricultural Laws, 2018 N.C. Sess. Laws 113, § 10(a) (“S.B. 711”), codified as N.C. Gen. Stat. § 106-701. The latter restriction was a direct response to the industry’s failed argument in court that plaintiffs who live in an affected household but who do not hold a formal property interest should not have legal standing to sue for nuisance. See In re NC Swine Farm Nuisance Litig., 2017 WL 5178038, at *4-5 (E.D.N.C. Nov. 8, 2017) (unpublished). For a thorough accounting of the industry’s legislative maneuvering, see Andie D’Angelo, Note, Taking the Whole Hog: How North Carolina's Right-to-Farm Act Strips Access to Nuisance Suits for Vulnerable Communities, 34 Georgetown Evntl. L. Rev. 141-165 (2021). 219

In Duplin County, North Carolina – the largest pork-producing county in the nation – nearly half of residents are Black or Latino, and nearly a quarter live at or below the federal poverty line. https://ncnewsline.com/2017/11/15/neutering-nuisance-laws-north-carolina/

plus $125 as a sweetener – an even split of the gains to trade between them. When the neighbors are allocated the entitlement to be free from harm, however, the polluting farm might find it advantageous to invest in impeding the neighbors’ ability to coordinate and enforce their rights, rather than spending $500 to install and operate pollution controls. For example, suppose it costs only $200 per year for the farm to invest in polluting the community’s information environment, such that the neighbors are unable to trace their elevated health risks to the farm’s operations. In that case, the farm would find it cheaper to invest in disinformation efforts rather than pollution controls. A similar result would follow if the farm could spend $200 per year in successful efforts to undermine the neighbors’ ability to cooperate together, perhaps by sowing dissent and discord within the community. In both cases, the farm’s obstructionist actions would leave society significantly worse off but the farm better off.220

220

On the assumed facts of Table 2, a particularly enterprising farm in column 3 might try to persuade the neighbors that the firm holds the entitlement to pollute, despite being in a situation where the law actually favors the neighbors. If successful, the firm could then further convince the neighbors to pay for installation of the pollution controls on the terms of column 1. Depending on the cost of this deception campaign, the farm could end up better off than simply trying to block enforcement of the neighbors’ rights. In this (unrealistic?) scenario, the higher-valued user would end up with the contested resource and the Coase Theorem would (sort of?) hold, albeit with deadweight social loss in the form of deception expenses.

Table 2 Initial entitlement to CAFO (WTP=WTA)

Initial entitlement to neighbors (WTA=WTP)

Initial entitlement to neighbors but CAFO impedes enforcement

CAFO initial profits

$1000

Valuation of harm

$750 (harm avoided)

$750 (harm suffered)

Cost of control

$500 (adopted)

$500 (not adopted)

Cost of impeding enforcement

n/a

$200

CAFO net outcome

$1125

$500

$800

Neighbors net outcome

-$625

-$750

Social net outcome

$500

$50

Again, the analytic point seems plain, but does it find expression in the real world? A case study of a decades-long nuisance dispute between Oregon cherry growers and a polluting aluminum smelter documented numerous instances of a defendant seeking to impede legal enforcement by raising transaction and coordination costs, including through obfuscation of scientific information regarding the plaintiffs’ harms, fabrication of information regarding defendants’ economic costs, and deliberate undermining of social cohesion within the affected community.221 The most striking feature of the case study is that when the smelter and its lawyers were caught submitting false documentation to the court and finally felt compelled to engage in earnest settlement negotiations, the parties structured a consent decree and an arbitration system that effectively governed their conflicts for the next three decades until the smelter was shut down

221

See Con Reynolds and Douglas A. Kysar, Tort as Regulation, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3449770.

for reasons unrelated to the pollution dispute.222 Coasean co-existence attained but only after an exasperated judge shut down the defendant’s years of deliberate attempts to impede the plaintiffs’ enforcement of their entitlement. During the North Carolina hog nuisance litigation, the pork industry similarly obtained multiple legislative changes that sought to raise the costs of enforcement facing plaintiffs in the very cases they were litigating. As recounted by the plaintiffs’ lawyer: “There were numerous laws that were changed during the period these cases were pending. A law was passed that made farm records private and confidential so that attorneys and others could not get access to them. There were [ag]-gag laws that would make it a misdemeanor for employees at processing plants to provide certain information. There was a law that affected our ability to use drones. There was a law passed that made it much harder to bring a lawsuit for a family member who had lived there, moved away and came back.”223 Similarly, researchers have begun to document a complicated web of entities including meat conglomerates and their trade associations that seek to obfuscate climate change science through now familiar, yet still effective, tactics honed by manufacturers of products like asbestos, tobacco, lead paint, and opioids.224 From the industry’s perspective, the return on investment from such efforts can be dramatic: If a few million dollars invested in a university program can help establish the legitimacy of GWP* as an accepted measure of animal agriculture’s climate impact, the industry could reap untold billions in the form of avoided regulatory costs. In short, circumstances can exist in which it profits a firm to invest in obstructing the enforcement of legal entitlements, rather than merely bargaining over the optimal allocation of those entitlements. Taking the Example to Its Extreme. Actors might invest not only in changing the allocation of a legal entitlement or the costs of enforcing it. They might also invest in more blunt efforts to block the legal capacity of impacted interests to hold entitlements or bargain over those entitlements at all. For instance, it might seem curious at first glance that an agribusiness firm would expend significant resources to mount a court challenge to a seemingly quixotic effort by the citizens of Toledo, Ohio to grant legal rights to Lake Erie.225 It might appear even more odd that the main opposition to the initially successful Lake Erie Bill of Rights ballot measure in Toledo was funded by the Houston office of the international fossil fuel company, BP.226

222

In the words of the lawyer for the plaintiffs, “Although the claimants started out as Daniel in the lion’s den, eventually they made common cause, and the lion lay down with the lamb, with none to make them afraid.” Con Reynolds and Douglas A. Kysar, Tort as Regulation, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3449770. 223

https://www.corporatecrimereporter.com/news/200/mona-lisa-wallace-on-the-case-against-smithfield-foods/. See also https://ncnewsline.com/2017/11/15/neutering-nuisance-laws-north-carolina/ (describing legislative changes to North Carolina right-to-farm laws and attributing them to “the enormous political power wielded by the NC Pork Council,” of which Murphy-Brown is a key member). 224

See, e.g., Oliver Lazarus, Sonali McDermid, and Jennifer Jacquet, ‘The climate responsibilities of industrial meat and dairy producers’ (2021) 165 Climatic Change 1-21; Morris, V., Jacquet, J. The animal agriculture industry, US universities, and the obstruction of climate understanding and policy. Climatic Change 177, 41 (2024). https://doi.org/10.1007/s10584-024-03690-w. 225

https://aldf.org/article/federal-judge-strikes-down-lake-erie-bill-of-rights/.

226

https://www.greatlakesnow.org/2020/10/legal-fees-toledo-residents-lebor-drewes-farms/.

What both the farm and the fossil fuel major recognized, however, is that raising the transaction costs of Coasean bargaining – in this case, by ensuring that Lake Erie remains an unrecognized legal entity unable to bargain or sue at all on its own behalf and dependent for its well-being on altruistic gestures by human decision makers and their political representatives – can be a profit-maximizing strategy. Indeed, with strong lobbying support from agriculture and energy companies, bills blocking legal personhood for animals and nature have proliferated during recent state legislative sessions in direct response to advocacy campaigns to promote the kinds of legal rights for animals and nature that have gained salience and momentum domestically and abroad.227 Rather than just invest in shaping the entitlements and valuations represented in Tables 1 and 2, these companies recognize that it is better to keep some impacted interests off the Tables entirely. This possibility was stumbled over by Richard Epstein in a brief passage from his influential article, The Path to ‘The T.J. Hooper’.228 Epstein lays out a libertarian vision of private norm development that is well aligned with Coase’s economic analysis. Specifically, Epstein distinguishes between stranger and non-stranger cases in the development of customary norms of behavior. He argues that tort law should defer to customary norms in non-stranger cases for reasons that accord with Coase’s reasoning: In such cases, the transaction costs of bargaining are presumed to be low and the resulting norms to reliably reflect the interests of all affected parties. In stranger cases, on the other hand, Epstein suggests that harmed individuals should be entitled to strict liability protection from tort law since their interests and autonomy presumably were not reflected in the development of the customary norm. The most striking feature of Epstein’s influential article is how few situations seem, in his view, to rise to the level of a “stranger” case in which law should not defer to the results of private exchange. Indeed, the primary concrete example in which Epstein concedes that tort law could play a salutary role by imposing liability rather than deferring to private ordering is one that, by comparison, casts doubt on most any other claim of unjust harm: Once the plaintiff is not part of the group that sets the custom, the evident risk is that the custom will evolve to meet the standards of the members of the group and not the outsiders who are harmed by it. Taken to its extreme, the uniform practice of slavery does not offer a customary justification relative to the slaves whose interests are suppressed by the custom in whose formation they played no part.229 Lurking behind Epstein’s “extreme” example is his normatively appropriate assumption that all humans have legal standing to own property, engage in contracts, and contribute to the 227

E.g., Utah legal personhood law

228

Richard A. Epstein, The Path to “The T.J. Hooper”: The Theory and History of Custom in the Law of Tort, 21 J. Legal Stud. 1 (1992). To the extent that this Part’s response to Epstein implies a connection between legal disempowerment of human slaves and nonhuman animals, it is a deeply fraught and reluctant one, prompted by ideologies all too quick to use the “extreme” case of human slavery to make analytical points. For an intellectually powerful and morally astute engagement with the difficulties of analyzing overlapping logics of oppression among living beings, see Aph & Syl Ko, Aphro-ism: Essays on Pop Culture, Feminism, and Black Veganism from Two Sisters (2017). 229

Id. at 19.

development of social customs through private ordering. Yet, at the time of slavery, dominant customs reflected quite the opposite view. More important, dominating slaveholders at that time enshrined their preferred view of custom within law, despite Epstein’s hope that law somehow overwhelms custom in only those few moments when law is needed. Slaves were not “strangers” at that time in the eyes of the law, they were “property”. Law’s role in constructing the status, power, and capacity of actors is treated as exogenous in Epstein’s theory, just as Coase seems to treat the allocation of legal entitlements and the existence of transaction costs as natural givens, rather than as objects of influence and control by self-interested actors. The argument of this Part has been that these efforts to contain the bounds of legal analysis give rise to conceptual externalities that must be internalized. CONCLUSION Coase rightly criticized the laissez faire notion that one should build policy on an ideal vision of freely bargaining individuals and firms interacting within an imagined pre-governmental, state of nature.230 Industrial animal agriculture underscores the fruitlessness of this exercise and Coase’s wisdom in rejecting it. The meat industry of today exists because governments have chosen to subsidize its operations through voluminous and lucrative non-market programs. The industry also critically depends on the ability of producers to use law and policy affirmatively as factors of negative production, enabling cost containment through externalization of harms onto disempowered communities and deliberate efforts to keep those communities disempowered. From undocumented slaughterhouse workers to climate change victims in Pakistan to future generations without groundwater access to billions of voiceless and voteless livestock animals, the harms of industrial meat production are concentrated among interests that hold little or no power. That distribution is no coincidence: Meat producers have an obvious economic interest in preserving the status quo distribution of power and they understandably view investment in law and policy protections to safeguard that distribution as a prudent exercise of business judgment. Economists sometimes defend the profit motive as socially desirable in terms that presume adherence by firms to baseline norms of law and ethics, as if those norms were somehow fixed and exogenous. Milton Friedman’s classic statement from 1970 is representative: “[Corporate agents’] responsibility is to conduct the business in accordance with their desires, which generally will be to make as much money as possible while conforming to the basic rules of the society, both those embodied in law and those embodied in ethical custom.”231 In his 1962 magnum opus, Friedman similarly qualified self-interested corporate maximization by reference to law and ethical norms: “There is one and only one social responsibility of business – to use its resources and engage in activities designed to increase its profits so long as it stays within the rules of the game.”232 230

Coase, The Problem of Social Cost, at p. 43 (“In a state of laissez faire, is there a monetary, a legal or a political system and if so, what are they? . . . The answers to all these questions are shrouded in mystery and every man is free to draw whatever conclusions he likes.”).

231

Milton Friedman, The Social Responsibility of Business is to Increase Its Profits, N.Y. TIMES, Sept. 13, 1970.

232

MILTON FRIEDMAN, CAPITALISM AND FREEDOM.

What happens, though, when “the basic rules of the society” – “the rules of the game” – themselves become a subject of profit-maximization? What happens to the legal economic intellectual framework when we recognize that actors can intentionally invest in acquiring legal protections to alter baseline bargaining positions and ensuing cost-benefit valuations which, in turn, are supposed determine what is social welfare maximizing? Where do we go from there? Close examination of the industrial meat sector suggests that the power of law is even more limited than Coase wished for it to be. Market actors today see investment in legal baselines and bargaining conditions as a profit center like any other factor of production. Bargaining in the shadow of the law233 includes the possibility of altering law itself and therefore the shadow circumstances of bargaining. The “basic rules of the society” and “the rules of the game” are not matters that can be taken for granted as Friedman, Coase, and Epstein each in their own way hoped. The market instead is a creation of law, and law cannot simply wait to be invoked only when the market “fails,” is overburdened by “transaction costs,” or encounters a “stranger” according to academic theory. The shape and character of the market, including its tendency to “fail,” to be weighed down by “transaction costs,” and to acknowledge some as “strangers” in need of protection, depends first and always on law itself. The modern meat sector understands this dynamic well. Accordingly, reform of a sector as powerful, pervasive, and shrewdly led as industrial animal agriculture will not happen through market incentives, litigation, or regulation, either alone or in combination. Instead, until democracy is revived and until people recognize that their popular sovereignty includes the power to construct and shape markets, rather than merely to fidget with them after the fact, real change will remain elusive. To paraphrase the titles of Coase’s two momentous articles, the nature of the farm today is to socialize costs. That is the problem.

233 Robert Mnookin & Lewis Kornhauser, Bargaining in the Shadow of the Law: The Case of Divorce, 88 YALE L.J. 950 (1979).

Very early draft – Not fair to cite/circulate

Public Health and Safety Nuisances Catherine M. Sharkey*

Table of Contents Introduction ..................................................................................................................................... 1 I. Public Health and Safety Nuisances ............................................................................................ 3 A. Historical Conception of “Public Health and Safety” as “Public Right” ............................... 3 1. Not limited to “land-based” interests................................................................................... 4 2. Includes “public health and safety” ..................................................................................... 6 B. Public Nuisance Per Se ........................................................................................................... 8 II. Enforcement: Risk Regulation in the 21st Century................................................................... 10 A. A Common Law for the Age of Regulations ....................................................................... 10 B. Public Nuisance .................................................................................................................... 11 1. Private ................................................................................................................................ 12 2. Public ................................................................................................................................. 14 III. Conclusion: Beyond Opioids .................................................................................................. 16

Introduction In 2022, in City & County of San Francisco—a bellwether bench trial in a multi-district litigation (MDL), which likely constitutes the largest civil action in American history—a federal district court judge determined that “the opioid epidemic, defined by high rates of opioid abuse, addiction, and overdoses, constitutes a public nuisance.” A bellwether case is a representative case chosen from the larger set of claims in the MDL based on how similar they are to the rest of the lawsuits, which serves as a litmus test for the larger group of MDL plaintiffs. “Bellwether” derives from leading sheep (or wether), which were adorned with bells to direct shepherds to their flocks. City & County of San Francisco is a

Segal Family Professor of Regulatory Law and Policy, New York University School of Law.

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

bellwether within this original meaning—it leads the way to recognition of a distinct category of public nuisance, the “public health and safety nuisance.” Within tort law, the law of nuisance implicates both ordinary private disputes and larger societal harms. Perhaps not surprisingly, in the past several years, the law of private nuisance has not changed much, but there has been explosive growth in the law of public nuisance.1 A public nuisance is “an unreasonable interference with a right common to the general public.”2 Public nuisance raises two fundamental questions: (1) what is a “public right” (or “right common to the general public”)?; and (2) who can enforce violations of such a public right? Our bellwether case, City & County of San Francisco, tackles these questions in an especially fraught context, namely whether public nuisance claims should be recognized against opioid manufacturers, distributors, and retailers for marketing, advertising, and sales practices that fueled the enduring opioid abuse public health crisis.3 Judge Charles Breyer of the Northern District of California had no difficulty concluding that “the ongoing opioid epidemic” was a “public nuisance”: Widespread opioid use has strained the city’s hospitals. It has forced streets, parks, and public spaces to close. It has exacerbated crime and homelessness. Every year, San

In recent years, public nuisance claims have been brought against those who advertised lead paint to pay for lead paint abatement in buildings (see, e.g., People v. ConAgra Grocery Prod. Co., 227 Cal. Rptr. 3d 499 (Ct. App. 2017); against oil companies to provide relief against sea-level rise attributable to global warming (see, e.g., City of New York v. Chevron Corp., 993 F.3d 81 (2d Cir. 2021). 2 RST § 821B. 3 In a set of coordinate bellwether cases in Ohio, a jury found CVS, Walmart, and Walgreens defendants liable on a public nuisance theory for contributing to the opioid crisis. In a subsequent bench trial, the defendants were ordered to pay $650.6 million to an abatement fund. In the spring of 2022, Johnson & Johnson and distributors AmerisourceBergen, Cardinal Health, and McKesson reached a settlement of $26 billion with states and municipalities.

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

Francisco devotes significant resources to a multiprong fight against the opioid epidemic. That fight includes this case. But other courts have expressed grave concerns about extending public nuisance in their jurisdictions to the sale or distribution of opioids. While much ink has been spilled by scholars wrestling with how to conceptualize public nuisance, thus far unexplored is the correspondence between negligence per se and public nuisance. City and County of San Francisco invites this inquiry; Walgreens violated myriad federal health and safety regulations. According to Judge Breyer, these regulatory violations form the basis of his finding of “unreasonable conduct” on the part of Walgreens. But a much deeper link could be forged between statutory and regulatory violations and public nuisance. It goes to the heart of what constitutes a “public right” enforceable in tort. Moreover, it provides a principled way of cabining the extent of societal harms cognizable as a public nuisance tort.

I. Public Health and Safety Nuisances A. Historical Conception of “Public Health and Safety” as “Public Right” The notion of a “public health and safety” nuisance comports with the historical conception of “public right” in public nuisance. To begin, notwithstanding some vociferous cries to the contrary, historical recognition of public nuisance was by no means limited to the land-based interests that are essential to private nuisance. Moreover, public health and safety was specifically recognized as a sub-category of actionable public rights.

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

City and County of San Francisco is our lodestar. In recognizing the opioid epidemic as a public nuisance in City and County of San Francisco, Judge Breyer insisted that public nuisance is not limited to land-based interests and can include public health and safety. First, he dispensed with the objection that all nuisances—whether private or public—must implicate a land-based interest: Unlike nuisance law in other states, “California law has never imposed such a requirement that there be some form of injury to land or property[.]” See Ileto v. Glock Inc., 349 F.3d 1191, 1213 (9th Cir. 2003) (“[T]he district court’s determination that a nuisance must be associated with property is contrary to clearly established California law.”). By both statute and common law, California has “consistently has defined nuisance in broad terms that encompass injuries to health, or acts that are ‘indecent or offensive to the senses’ or obstructions to the free use of property in any manner that might interference with the ‘comfortable enjoyment of life or property.’”4 Nor is California is not an outlier. A rich corpus of historical cases—going back centuries, but especially prominent in the 19th and early 20th century—backs the conception of public right as encompassing public health and safety. 1. Not limited to “land-based” interests In an amicus brief to the Oklahoma Supreme Court, Professor Richard Epstein argued that “the interests in lands, air, or water protected by the law of nuisance are identical for both private

Id. (quoting Acuna, 60 Cal. Rptr. 2d 277, 285 (1997)). See Cal. Civ. Code § 3479 (“Anything which is injurious to health, including, but not limited to, the illegal sale of controlled substances, or is indecent or offensive to the senses, or an obstruction to the free use of property, so as to interfere with the comfortable enjoyment of life or property, or unlawfully obstructs the free passage or use, in the customary manner, of any navigable lake, or river, bay, stream, canal, or basin, or any public park, square, street, or highway, is a nuisance.”) (emphasis added).

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

and public nuisance,” and, moreover, that neither private nor public nuisance “has ever covered financial losses . . .”5 Several courts, including the Oklahoma Supreme Court, which overturned a $465 million trial verdict for public nuisance against opioid manufacturers,6 have been persuaded that public nuisance traditionally applies only “in the context of conduct that interferes with public property or resources.”7 But a survey of historical cases belies the position that “public right” was limited either to its origin in blocking public highways and waterways or to protecting only land-based interests. Indeed, even scholars skeptical (or outright hostile) to modern public nuisance cases concede the existence of a long history of public nuisance for non-land-based purposes, including to enforce conventional morality or to protect against harms from faulty products.8 William Blackstone, writing in 1765, identified eight categories of public nuisance.9 To be sure, the first category is “annoyances in highways, bridges, and public rivers, by rendering the same inconvenient or dangerous to pass: either positively, by actual obstructions; or negatively, by want of reparations . . . .” But the categories widen from there to include “[a]ll those kind of nuisances, (such as offensive trades and manufactures) which when injurious to a private man are actionable, . . . particularly the keeping of hogs in any city or market or town . . .” as well as “[a]ll

Amicus Curiae Brief of Competitive Enterprise Institute at 8, Oklahoma ex rel. Hunter v. Johnson & Johnson, Appeal No. 118,474 (Okla. Oct. 22, 2020); see also Epstein, The Private Law Connections to Public Nuisance Law: Some Realism About Today’s Intellectual Nominalism, 17 J.L. Econ. & Pol’y 282, 285 (2022) (calling for a return to a landbased view of public nuisance, grounded in private nuisance). 6 State ex rel. Hunter v. Johnson & Johnson, 499 P.3d 719, 727 (Okla. 2021) (overturning a $465 million trial verdict). 7 City of Huntington v. Amerisource Bergen Drug Corp., 609 F. Supp. 3d 408, 472 (S.D. W. Va. 2022). 8 Professor Thomas Merrill, for example, disagrees, as a historical matter, that public nuisance must be land-based, but argues, more radically, that public nuisance is not a tort but rather a public action to regulate prospective harms. Now that legislative and regulatory bodies can define proscribed conduct in advance, Merrill argues the use of public nuisance for tort should be curtailed. Merrill, Public Nuisance as Risk Regulation, 17 J. L. Econ. & Pol’y 347 (2022). See also Don Gifford & Richard Boldt, Interference with the Democratic Process as a Public Nuisance (working paper). 9 4 William Blackstone, Commentaries.

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

disorderly inns or ale-houses, bawdy-houses, gaming-houses, stage-plays unlicensed, booths and stages for rope-dancers, mountebanks, and the like . . . .”10 Fast forward to the 19th century. A private plaintiff successfully sued New York City for public nuisance for erecting a bridge across a public street that obstructed access to his store, alleging loss of business resulting from the delay caused by the obstruction—and injury that bears no direct relation to land-based interests and, moreover, is for purely financial losses. Myriad air and water pollution cases proliferated in the late 19th and early 20th centuries. An 1874 Wisconsin case involved a defendant who released animal waste and other offensive liquids into a stream from a distillery and farm.11 An 1888 Pennsylvania case pinned public nuisance liability where defendant’s factory released lead and arsenic gas, sickening a neighbor.12 2. Includes “public health and safety” Not only does the historical record support the extension of public nuisance claims beyond blockages of highways and waterways and land-based interests, it specifically included public health and safety as a recognized public right. The best evidence for this can be found in a plethora of public nuisance actions over the course of the 19th century relating to mosquito-borne illnesses arising from stagnant ponds. Such cases followed a familiar pattern; they generally arose after someone dammed a waterway, creating a reservoir of standing water that attracted mosquitos. People nearby who suffered from malaria or other mosquito-borne illnesses as a result sued the dam builders for public nuisance, claiming

Blackstone’s additional categories include lotteries; cottages; fireworks; eavesdroppers; and “scolds”—some of which have attracted attention in their own right. See, e.g., Maureen E. Brady, Cottages as Public Nuisances: The Long History of Land Use Regulation of the Poor, DePaul Law Review (forthcoming 2024); Julia Keller, Eavesdropping: The Forgotten Public Nuisance in the Age of Alexa, 77 Vand. L. Rev. 169 (2024). 11 Greene v. Nunnemacher, 36 Wis. 50 (Wis. 1874). 12 Price v. Grantz, 118 Pa. 402 (Penn. 1888).

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

that by creating the conditions that attracted the mosquitos to the entire area, they infringed upon public health. Consider Mills v. Hall & Richards,13 an 1832 New York case, where defendant had built a dam that created a pool of water that attracted malarial mosquitoes, and “after the rebuilding of the dam, the fever and ague was more common in that section of the country than it had been before.” Plaintiff and his family suffered health effects. The court, recognizing the action for public nuisance, opined: “a man may overflow his own land; but if such overflow spread disease and death through the neighborhood, it may be abated, and he must respond in damages for the special injury which any individual may have sustained from it.” Similar 19th and early 20th century public health-based public nuisances were recognized in jurisdictions throughout the United States.14 In one noteworthy 1902 Alabama case, where the plaintiff lived adjacent to malarial ponds forged by defendant’s dam for a gristmill, the court deemed the ponds a public nuisance because they violated the right of the plaintiff to “dwell in his homestead freed from the peril of disease and death.15 Threats to public health were likewise recognized in the context of sewage and spoiled milk. A 1933 Kentucky court addressed accumulated filth as a result of flooding that “imperiled the health of the people of the neighborhood,” and concluded that “no rule of law is better settled than that whatever endangers the public health is a public nuisance.”16 And a 1906 Ohio case deemed spoiled milk harmful to health and thus a public nuisance “by nature” such that it could

9 Wend. 315-16 (N.Y. Sup. Ct. 1832). See, e.g., Story v. Hammond, 4 Ohio 376 (1832) (plaintiff brought a nuisance suit alleging that the defendant’s mill ponds made him and his family sick). 15 Richards v. Daugherty, 133 Ala. 569, 573-74 (1902). 16 Ludlow v. Commonwealth, 247 Ky. 166, 168 (Ky. App. 1933). In that case, the City of Ludlow was held criminally liable for public nuisance after it failed to maintain its sewers, which could not sufficiently carry water, leading to flooding and odor. 14

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

be destroyed for abatement purposes.17 And finally fireworks were the common target of safety-related public nuisance actions. An 1881 New Jersey case designated it a per se public nuisance to fire explosives in a public highway.18

B. Public Nuisance Per Se City and County of San Francisco provides a jumping off point for consideration of a novel conception of public nuisance per se. At trial, the plaintiff established that Walgreens pharmacies, over a period of fifteen years, violated federal statutes and regulations. Federal regulations specify that, separate and apart from the prescribing doctor’s responsibilities, “a corresponding responsibility rests with the pharmacist who fills the prescription.”19 Specifically, pharmacists and pharmacies are required “to identify and resolve objective signs ‘arising during the presentation of a prescription’ that create ‘a reasonable suspicion that the prescription is not, on its face, legitimate.’” Judge Breyer detailed Walgreens’ systematic regulatory violations over a fifteen-year period to substantiate the requisite “unreasonable conduct” on the part of Walgreens warranting public nuisance liability. Judge Breyer’s approach is supported by the Restatement (Second) of Torts which stipulates that among the “[c]ircumstances that may sustain a holding that an interference with a public right is unreasonable” is “whether the conduct is proscribed by a statute, ordinance or administrative regulation.”20

Kaiser v. Walsh, 17 Ohio 324. In that case, the plaintiff sought an injunction against the milk inspector who, per local ordinance, destroyed milk above 50 degrees. See also Deems v. Baltimore, 30 A. 648 (Md. 1894) (city was authorized, under the police power, to inspect and destroy adulterated milk because such milk represented an imminent danger to life and health). 18 Jenne v. Sutton, 43 N.J.L. 257 (1881). CITE additional cases. 19 21 C.F.R. § 1306.04(a). 20 RST § 821B(2)(b). The other two are:

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

But I want to suggest a more foundational role for the violation of statutes, ordinances, or administrative regulations in public nuisance, namely at the very core of the conception of a public right. Whereas Judge Breyer asserted that the opioid epidemic interfered with five distinct classes of public rights, “the public health, the public safety, the public peace, the public comfort or the public convenience.” He missed an opportunity to showcase a distinct—much more narrowly defined—subcategory, namely one that arises by repeated violations of statutory and regulatory duties: public nuisance per se. It is standard tort fare that negligence per se stipulates that the violation of a statute or regulation amounts to breach of a duty of care. In this manner, negligence per se imports a “public law” standard into private tort litigation. The statute or regulation must meet the criteria of “class of persons” and “class of risks”—namely the plaintiff falls within the class of individuals protected by the statute or regulation and has been harmed by the class of risks likewise falling within the ambit of protection. Moreover, to prevail on a negligence claim, a private plaintiff must still prove causation and damages. Let us consider the tort of public nuisance against this background of traditional negligence per se actions. “Public nuisance per se” might stipulate that the violation of a statute or violation is tantamount to violation of a “public right” (or “right common to the public”). Two fundamental features would nonetheless distinguish public nuisance per se from negligence per se. First, there is no “class of persons”/ “class of risks” requirement. Indeed, this is precisely why a public

(a) Whether the conduct involves a significant interference with the public health, the public safety, the public peace, the public comfort or the public convenience, or . . . (c) whether the conduct is of a continuing nature or has produced a permanent or long-lasting effect, and, as the actor knows or has reason to know, has a significant effect upon the public right. Id. §§ 821B(a)&(c).

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

nuisance per se action might be needed (as is not duplicative of negligence per se actions). Second, public nuisance per se can be enforced by the state as well as by private individuals.

II. Enforcement: Risk Regulation in the 21st Century A. A Common Law for the Age of Regulations Much of my prior work engages the deterrence or regulatory role of tort law and the interplay of tort law and the regulatory state (i.e., reams of articles on federal preemption of state tort law), as well as how tort has—or should—respond to issues of widespread societal harms (e.g,. punitive damages), including the proliferation of purely financial losses due to negligence (e.g., the economic loss rule). Public nuisance certainly fits right here. It is a regulatory tort that addresses widespread harm; moreover, often it’s aimed at financial losses (e.g. abatement expenses). Thus, when Professor Thomas Merrill argues that public nuisance is risk regulation and, as such, not a tort. He has it half right.21 Public nuisance is about risk regulation, but this by no means distinguishes it from other torts. Nowhere perhaps is this more salient than in the context of negligence per se actions in highly regulated field. Take, for example, negligence per se actions in the realm of medical devices, which are highly regulated by federal regulations, enforced by the Food and Drug Administration. Notwithstanding this form of public regulation, the tort of negligence per se allows for private enforcement of breaches of regulatory duties. Moreover, such torts persist even in the face of a statutory scheme with an express preemption provision that ousts most design defect and failure-to-warn claims, while still allowing the more narrowly formulated negligence per se actions. If Merrill were right, once we have federal health and safety regulations that can be

Merrill, Public Nuisance as Risk Regulation, 17 J. L. Econ. & Pol’y 347 (2022).

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

publicly enforced, there is no further need for a regulatory role for torts (such as negligence per se). But, just as we do have negligence per se actions, so public health and safety nuisances per se can play a necessary regulatory enforcement role.

B. Public Nuisance Public nuisance can be enforced privately (private public nuisance) or publicly (public public nuisance). Direct public action—typically administrative regulation or criminal prosecution—regulates general widespread damages arising from public nuisances. Where such general damages are of low intensity and are widely diffused across an extended population, private actions for admitted grievances would simply be too costly to maintain.22 The enforcement function is thus centralized to preserve the deterrent and control objectives of the tort law, even though direct compensation to aggrieved parties is necessarily abandoned. Private actions, however, have been maintainable (harkening all the way back to the 14th century Anonymous case) for “special,” “peculiar,” or “disproportionate” harm to the individual plaintiff arising from a public nuisance. In a prior Clifford symposium article, I elaborate the enforcement rationale, whereby the special injury rule deputizes a subclass of impacted individuals or entities to sue to force the tortfeasor to internalize the social costs of its activities.23

Note here that the Restatement (Second) of Torts raises the specter of the class action. See RST 821C(2)(c) (“have standing to sue as a representative of the general public, as a citizen in a citizen’s action or as a member of a class in a class action”). 23 Sharkey, Public Nuisance as Modern Business Tort: A New Unified Framework for Liability for Economic Harms, 70 DePaul L. Rev. 431, 448–50 (2020). In that article, I contend that this “channeling” rationale explains why the “‘special injury’ requirement has emerged as the core doctrinal distinction” between cognizable public nuisance claims and barred negligent interference with economic relations claims.

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

1. Private Private public nuisance actions are the most “tort-like” in that a private individual brings suit for private harms (albeit arising from the violation of a public right). Per the Restatement (Second) of Torts, “[i]n order to recover damages in an individual action for a public nuisance, one must have suffered harm of a kind different from that suffered by other members of the public exercising the right common to the general public that was the subject of interference.”24 Courts have struggled to articulate what amounts to “special injury,” namely private harms that are “different in kind” from those of the general public. By way of illustration, consider two contrasting malarial mosquito cases. As discussed above, courts typically have recognized as requisite violations of “public right” the public health impact, including fever, malaria and other mosquito-borne illnesses arising from stagnant waters created by defendants’ construction of dams. But they have divided sharply on the question of when private individuals can enforce violation of such a public health right. Thus, in Mills, the New York Supreme Court allowed plaintiff and his family to recover in damages “for the special injury which any individual may have sustained from it” even as “disease and death [spread] though the neighbourhood.” Whereas, in Belton v. Wateree Power, the South Carolina Supreme Court—likewise recognizing invasion of a public right where a company constructed a dam that led to stagnant pools where mosquitoes thrived—refused to allow plaintiff who had suffered illnesses (as well as a decrease in the value of their land) on the ground that because the entire community was affected by the malaria-carrying mosquitoes the plaintiff did not adequately allege special injury.25

24 25

RST §821C(1). 115 S.E. 587 (S.C. 1922).

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

Perhaps the strongest rationale for the “special injury” rule is the channeling enforcement rationale, which aims to effectuate optimal societal deterrence. Indeed, the historical cases evince an interest in balancing under- and over-deterrence concerns. Blackstone raised the concern regarding multiplicity of actions, when he noted that a public nuisance is typically not privately actionable because “it would be unreasonable to multiply suits by giving every many a separate right of action.”26 There is, nonetheless, a robust category of private public nuisance cases, which seem justified on deterrence or channeling grounds. In the modern MDL opioid context, the West Boca Medical Center case is illustrative. West Boca Medical Center survived a motion to dismiss on its private nuisance claims.27 The court noted that West Boca’s injuries were not “merely derivative” of the “public” injury because, as a hospital, West Boca incurred increased costs stemming from the purchase of excess opioid pills—an injury sufficiently distinct in kind to survive the pleading stage. In my prior Clifford Symposium article, I argued that the court’s reasoning was justified on deterrence or channeling grounds; concentrated losses fell onto the hospital, which stood in the best position to enforce. On the question of public right, the court referred to earlier orders in the MDL, which cited the Restatement (Second) of Torts § 821B in determining that public health qualifies as a public right.28 The hospital could not have brought a negligence per se action for violation of various statutes and regulations, as it did not fall within the “class of person”/ “class of risks” protected by such statutes. With regard to negligence per se, the “class of person,”/ “class of risks” designation

4 Blackstone, Commentaries 167. W. Boca Med. Ctr., Inc. v. AmerisourceBergen Drug Corp. (In re Nat’l Prescription Opiate Litig.), 452 F. Supp. 3d 745, 774-75 (N.D. Ohio 2020). 28 Id. at 774. 27

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

accords standing to a private individual (or entity) to proceed to enforce against violation of the public right. The “special injury” rule serves as a standing surrogate—allowing the private individual to enforce against the violation of the public right. 2. Public Recall that we began with an assumption that, as a general matter, diffuse widespread harms should be enforced via administrative (or criminal) action. Where private individuals, however, suffered “special injury,” they emerge as necessary enforcers. What, then, is the rationale for public enforcement of public nuisance actions? With regard to abatement of the public nuisance, the Restatement (Second) of Torts vests “authority as a public official or public agency to represent the state or a political subdivision in the matter.” But what about with regard to pursuit of what I have termed per se public health and safety nuisances. For, after all, if there has been repeated statutory and regulatory violations, the state presumably is well positioned to enforce administratively. In what circumstances, then, should it be authorized to pursue tort actions? Critics of public public nuisance actions—none perhaps more ardent on this front than Professor Thomas Merrill—would have us believe this is an inherently modern issue, one begat of the unholy alliance between states’ attorney generals and plaintiffs’ attorneys. But this overlap has a distinct historical pedigree. When Blackstone invoked the “keeping of hogs in any city or market or town” as a subcategory of public nuisance, he cited Dominus Regina v. Wigg. It is a peculiarly interesting case for my purposes here. The defendant had been indicted for keeping hogs in the street of London. His defense was that he could not be indicted 14

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

because he had violated a statute prohibiting this activity, which laid out its own remedy. In other words, for a statutory violation, the state must exclusively pursue statutory enforcement and remedy. The court emphatically rejected this position, deciding that the state could choose to pursue either remedy. And, since at common law it was a nuisance to keep hogs in the city, the defendant could be indicted. Return now to the City and County of San Francisco opioids case. Walgreens violated— repeatedly over fifteen years—statutory and regulatory directives with regard to the dispensing of opioids. But in order to bring a negligence per se action, a plaintiff would have to demonstrate that she were within the “class of persons”/ “class of risks” protected by the statute or regulation. Here, the statutes and regulations are intended to protect individual members of the public. Thus, neither public entities nor private entities such as the West Boca Medical Center could enforce against such violations.29 The state (and other public entities) have at their disposal statutory or regulatory enforcement. But—as Blackstone’s hogs case reminds us—that need not be there exclusive option. With respect to health and safety regulations, the public nuisance per se option should be on the table. It could be pursued by a private individual (or entity) who has suffered “special damages” or alternatively by the state or other public entity.

Indeed, in the West Boca case, the negligence per se claim was dismissed because the private medical center did not fall into the class of persons intended to be protected by the Controlled Substances Act or by the Florida Drug and Cosmetic Act, both of which are intended to protect individual members of the public. Likewise, the negligence per se claims brought by the Blackfeet Tribe and Muscogee Nation were dismissed because plaintiff tribes are not part of the intended class of individual beneficiaries either of the C.S.A. or state statutes in Montana or Oklahoma. See also Cleveland Bakers and Teamsters Health & Welfare Fund et al. v. Purdue Pharma, L.P. et al. (plaintiffs, union insurance funds, are not individual members of the public and are therefore not within the intended beneficiaries of either the federal C.S.A. or the Ohio Pharmacy Board regulations).

Sharkey, Public Health and Safety Nuisances Very early draft – Not fair to cite/circulate

III. Conclusion: Beyond Opioids Throughout this article, City and County of San Francisco has been our bellwether, pointing the way toward recognition of public health and safety per se nuisances. But of course opioids is only (significant) application. JUUL also springs to mind. On April 12, 2023, JUUL reached a $462 million settlement in lawsuits brought by New York, California, Colorado, Illinois, Massachusetts, New Mexico, and Washington, D.C.30 The New York complaint alleged that JUUL interfered with both the public right to health and safety and the public right not to be injured or defrauded by wrongful conduct. Separately, the New York complaint alleges that JUUL violated myriad state and federal statutes. And while the complaint alleges in detail the ways in which JUUL violated state and federal statutes, none of these violations grounds the public nuisance claim. My article suggests a framing for JUUL of public health and safety nuisance per se. Such a framing would not occupy the field of public nuisance claims—for example, the public nuisance claims against social media companies would fall outside—but it would solidify the basis for courts’ recognition of public health and safety nuisances where there have been blatant and egregious unenforced violations of health and safety statutes and regulations.

This multistate effort was co-led by the Attorneys General of California and New York.

Warning to readers: This is a very early and incomplete draft! Regulatory Failure, Tragic Exposures, followed by Mass Litigation: Does it Have to be this Way? Steve Gold (Rutgers), Tom McGarity (U of Texas) & Wendy Wagner (U of Texas) – 5/29/2024

Table of Contents I. THE GAP BETWEEN REGULATORY ACTION, LITIGATION, AND RESPONSE ......................................... 4 A. Overview ............................................................................................................................... 5 B. Case Studies .......................................................................................................................... 6 II. REGULATORY DESIGN: WHY AGENCIES ARE FAILING TO PROVIDE EARLY WARNINGS, OFTEN DESPITE PROTECTIVE LEGISLATION .................................................................................................................. 8 A. Early Warning Detection: on Paper....................................................................................... 9 B. Early Warning Detection: In Practice. ................................................................................. 13 1. Ability to generate research ............................................................................................. 14 2. Contribution of industry in risk detection and assessment. ............................................. 15 3. Acquiring Information from Industry Files. ...................................................................... 17 4. Sharing Information Underlying Risk Assessments. ......................................................... 19 5. Defending against Industry Opposition. ........................................................................... 20 6. Executive Branch Engagement. ........................................................................................ 21 C. Acting on Early Warnings. .................................................................................................. 22 1. Congressionally Imposed Shackles. ................................................................................. 22 2. Presidentially Imposed Shackles. ..................................................................................... 25 III. REFORM ................................................................................................................................... 28 A. The Sentinel Idea ................................................................................................................. 29 B. Enumerated Functions and Added Guardrails on the Sentinel’s Operation ........................ 31 C. Anticipating Skepticism ...................................................................................................... 34 1. Won’t Politicians Simply Ignore the Sentinel’s Warnings? .............................................. 34 2. Will the Sentinel Actually do the Work? ........................................................................... 36 3. If the Sentinel Does the Work and it Does Change the Scientific Conversation, Will That Lead to Its Political Demise?................................................................................................ 36 3. Isn’t the Creation of the Sentinel Politically Impossible? ................................................. 38 IV. CONCLUSION ............................................................................................................................ 39 APPENDIX 1 ................................................................................................................................ 41 APPENDIX 2 ................................................................................................................................ 44 1. Asbestos ................................................................................................................................ 44 The Industry Prepares for Regulation................................................................................... 46 EPA Regulation of Asbestos. ................................................................................................. 47 2. MTBE ................................................................................................................................... 60 3. “Forever chemicals”: PFOA, PFOS, and other PFAS .......................................................... 65

A central role for health and environmental agencies like the EPA is to identify and protect the public from emerging hazards that threaten widespread social and economic harm.1 Many environmental and public health laws explicitly identify early response as one of the key missions of the implementing agency.2 These statutes direct the protective agencies to be on the alert to locate hazards before they become costly disasters. Every law student learns that tort law provides redress for harm after it happens, while government regulation aims to prevent harm that would otherwise occur. So, when we hear about a plaintiff’s attorney discovering a prominent company’s widespread use of a highly toxic chemical for decades only after reading through a storeroom of discovery documents,3 we can’t help ask “where were the regulators?” And unfortunately, this is not an isolated incident. Regulators have taken a hands-off approach to many other hazards that tort lawyers only later discover have been causing chronic health and environmental harms. In this essay we ask why regulators seem to fall behind tort litigation with some frequency, even though the regulators are equipped with far better tools, authorities, and directives to prevent harm ex ante. The conventional reason given for this regulatory failure in the face of mass toxic torts places the blame on the often flagrant misconduct of the defendant industries: we learn that a defendant-corporation, or more than one, strategically lied to regulators, manipulated the scientific record, assiduously resisted testing, and suppressed what little internal information that did exist about suspected hazards.4 It seems that nothing short of “corporate-death-penalty” type sanctions can deter the otherwise highly profitable corporate behavior.5 But in many of these cases, industry duplicity doesn’t provide the whole story. Even as a tsunami of asbestos litigation began to engulf the courts, early efforts to regulate asbestos were watered down and delayed by political interference and ultimately killed by a judicial interpretation of statutory provisions tailored to make regulation difficult.6 In regulating underground storage tanks, EPA chose to satisfy industry rather than maximize protection of the environment, leading to widespread contamination of groundwater with the fuel additive MTBE7 and enormous cleanup liability.8 Feeble statutory information-gathering authority meant that toxic tort litigation, rather than government regulatory efforts, exposed the danger of PFOA.9 But

See infra Section IIA See infra Table 2 and and accompanying text. 3 See infra {PFAS case study}; see generally {Bilott book}. 4 See, e.g., Tess Legg et al., The Science for Profit Model—How and why corporations influence science and the use of science in policy and practice, 16 PLOS ONE (2021), available at https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0253272 . 5 See Roy Shapira & Luigi Zingales, Is Pollution Value-Maximizing? The DuPont Case (2017). 6 See {Asbestos Case Study}, infra App. 2. 7 Methyl tert-butyl ether. 8 See {MTBE Case Study}, infra App. 2. 9 Perfluorooctanoic acid. 2

PFOA, it turned out, had been in the chemical substance inventory EPA maintains under TSCA10, along with some 600 “active” chemicals in the same PFAS11 family listed in the same database -often for decades – with little evidence of meaningful regulatory oversight.12 And sadly, these regulatory disappointments are not outliers. In many toxic tort cases, regulatory oversight actually precedes the toxic tort litigation, and only in retrospect do we discover that the hazard occurred partly because regulators failed to protect the public, despite strong-seeming statutory directives. In fact, in some cases, even after tort litigation is successful, regulatory responses may still be nonexistent or slow in coming. In endeavoring to connect the dots to explain why the regulatory agencies sometimes fail to do their assigned sentinel work, we spotlight the ways that the agencies’ broad mandates, which promise protective action, can be undermined by politically-imposed shackles that constrain the agency from gathering information, processing that information, sharing that information, and acting on it to provide protective responses.13 We expose how the two political branches, like puppeteers operating marionettes, work behind the scenes to rein in agency action, while still carefully preserving the misleading façade of a protective regulatory program. This pattern, although by no means universal – agencies often do manage to take preventive action – is familiar.14 Our essay begins in part I by describing several situations in which successful toxic tort litigation antedated effective regulation even though agencies appeared to have had the authority to prevent or limit the harm. In the second section, we investigate from high altitude some of the reasons agencies fail at their protective task. Although for a variety of reasons we cannot always connect the dots in detail back to a specific regulatory failure such as those identified in Section I, our investigation exposes a wide variety of politically-imposed shackles by both Congress and the Executive branch that help explain the resultant lack of preventive action despite statutory authorities that promise public protection. The essay concludes with a very preliminary proposal to circumvent what otherwise seems like an intractable problem in the foundational design of the administrative state. We believe any solution must overcome the challenge of ever-present political control over regulatory agencies by creating a new, independent scientific body free from political control and tasked with the sole job of identifying hazards in need of regulatory

See {Forever Chemicals Case Study}, infra App. 2. Per- and polyfluoroalkyl substances. 12 US EPA, EPA’s Per- and Polyfluoroalkyl Substances (PFAS) Action Plan 11-12 (Feb. 2019) (stating that the TSCA Chemical Substances Inventory included 1,223 compounds, 602 of which were commercially active, meaning that they were in U.S. commerce at some point between June 2006 and June 2016). 13 This shackles research is part of a larger book project between McGarity and Wagner under contract with Columbia University Press. 14 In some cases, these political constraints consist of direct interventions on agency action to protect an economically powerful actor from regulatory oversight. But in most cases, the shackles involve more general, incremental tinkering with agency authorities – often in the fine print of the authorizing statutes, new legislation, or in executive oversight. 11

attention. Appreciating that some might view this proposal as naive, we endeavor to be realistic and troubleshoot some of the more obvious concerns. I. THE GAP BETWEEN REGULATORY ACTION, LITIGATION, AND RESPONSE Over the last seventy years, tort litigation has proven a formidable force in holding industries accountable for some of the widespread harms their products and activities inflict on society. Indeed, tort law is often the first on the scene – serving to catalyze regulatory and legislative action rather than follow it.15 But today, even though we have now tasked multiple expert agencies, bolstered by dozens of protective statutes with the specific purpose of locating and protecting us again emerging hazards, the same sequence – tort litigation followed by more targeted regulatory or statutory response -- remains a common pattern. In this section we explore whether – now with a fairly robust regulatory infrastructure in place to address emerging hazards – the influence of tort litigation is receding. In theory at least, expert agencies should be busy identifying and restricting preventable hazards before they injure large numbers of people. But a preliminary look at the state of toxic tort litigation suggests that this is not the case. In the first, brief subsection, we search for examples of tort litigation that follow this backwards pattern – arising despite the existence of protective regulatory statutes governing those same hazards. We then present three case studies that engage more deeply with the interplay between tort litigation and regulatory response in the cases of asbestos, PFAS, and MTBE. These case studies not only reveal regulatory inaction preceding tort litigation – even when protective mandates and agency authority are in place to anticipate and prevent the harms – but significant delays and inaction even after the tort victims began winning in court. The disturbing pattern of regulatory action lagging rather than leading in protecting public health, even in cases where regulators are charged with anticipating emerging hazards, is not always present of course. But the examples recounted here suggest that the pattern occurs with sufficient frequency to warrant further study; indeed, it is tort litigation that brings these regulatory disappointments into view since we may have never learned of the hazards without the ingenuity and perseverance of tort attorneys.

See, e.g., Nora F. Engstrom & Robert L. Rabin, Pursuing Public Health Through Litigation, 73 STAN. L. REV. 285, 35-61 (2021) (positing the “catalyst” theory for the role of tort litigation in sparking greater regulatory oversight); Mark Nevitt and Robert Percival, Can Environmental Law Solve the “Forever Chemical” Problem?, 57 Wake Forest L Rev 239 (2022) (discussing one attorney’s pivotal role in discovering and exposing the hazards of PFOS and related PFAS chemicals); Robert Bilott, Exposure: Poisoned Water, Corporate Greed, and One Lawyer’s Twenty-Year Battle against DuPont (1st ed., Atria Books 2019) (describing the attorney’s discovery of C8 used in nonstick coatings, a hazardous PFAS chemical); David Michaels and Celeste Monforton, ‘Scientific Evidence in the Regulatory System: Manufacturing Uncertainty and the Demise of the Formal Regulatory System’ (2005) 13 JL and Pol 17, 18–24 (discussing the role of the plaintiff’s attorney in bringing the risks of diacetyl to light for OSHA).

A. Overview A substantial collection of books and articles document the leading role of tort litigation in holding companies accountable for latent harms in settings in which existing regulation was nonexistent or inadequate. In accounts of tort litigation involving widely used hazards such as asbestos, tobacco, opioids, DES, and Dalkon Shield, tort litigation preceded meaningful regulation. But in each of these cases, at least at some point in the litigation’s history, regulatory controls were put in place to protect against the preventable harms, and yet victims claimed these regulatory controls were inadequate. Perhaps these cases are outliers -- relics of an embryonic regulatory state? Our case studies, detailed below,16 suggest not. But to gain a more systematic view, we searched for examples of other cases and litigation that appears to follow this same pattern of tort litigation occurring despite the fact that protective regulatory authority, ostensibly to prevent the hazards, is already in place. With the help of a database of illustrative toxic tort settlements involving latent hazards available on Lexis, coupled with our own encounters with toxic tort litigation, Table 1 assembles a preliminary list of toxic tort cases that occur despite pre-existing regulatory oversight. Although this list is incomplete, it nevertheless reveals that tort litigation is thriving despite the presence of a regulatory system intended to prevent these same harms. Some of the litigation involves isolated hazards – for example a side effect of a drug used by a small set of the public. But much of this litigation is large and even massive. All of the cases in Table 1 involve tort litigation that exposes lapses in our regulatory response ex ante, ex post, or both. See Appendix 1 [Note to readers: this table is extremely preliminary; our inventory and classification of cases requires significant development]. A subset of cases involves injuries that result from more discrete failure in regulator’s oversight of specific drugs, pesticides, and consumer products at the licensing stage. Most cases, however, involve claims that allege harms for hazards that not only arose before the advent of meaningful regulation but that continued because subsequent regulatory controls were inadequate. For example, tort litigation against sterilizing facilities allege the plaintiffs’ cancers and related harms were caused by excessive levels of ethylene oxide emissions that occurred primarily after passage of a protective mandate in the Clean Air Act in 1970 followed by major amendments in 1990.17 At

See infra Appendix 2. See, e.g., Ethylene Oxide Litigation, available https://www.levylaw.com/ethylene-oxide-lawsuits/; David A. Fusco, Litigation Minute: Ethylene Oxide—What it is and why Should Care, available at https://www.klgates.com/Litigation-Minute-Ethylene-OxideWhat-it-Is-and-Why-You-Should-Care-2-72023 (observing that “[l]aw is typically downstream of science, and EO litigation is no exception”); Brenda Goodman & Andy Millier, Sterilization facilities hit with a wave of lawsuits over ethylene oxide, Georgia Health News (Sept. 8, 2020), available at 17

least four years after this litigation was filed, in 2024, EPA formally recognized the insufficiency of its existing standard for regulating ethylene oxide emissions and promulgated a rule lowering the emissions standard.18 We also inventoried the cases to assess whether, even after tort litigation makes the inadequacy of regulation salient, our regulatory system re-calibrates to provide more rigorous regulatory protections in the future. Certainly tort litigation does catalyze political discussions as a result of the information it generates about hazards and industry misconduct.19 But even here we see preliminary evidence that regulatory responses are not always swift or complete. Instead, in some settings the deterrent function of tort litigation remains the primary and sometimes only institutional vehicle for minimizing future harm. Industries may change their practices because tort liability makes them unprofitable, but in these cases regulators sit on the sidelines, contributing little to alter the corporate calculations. [We will add one or more examples here.] This anecdotal tallying of cases does not tell us how often or the reasons that the regulatory system sometimes trails tort law in discovering and acting on public hazards. However, in the abstract, the relatively extensive list of cases does suggest that there are more than a few instances in which meaningful protective regulation lagged – rather than preceded – this common law litigation. B. Case Studies To gain a bit more purchase on what is actually occurring under the surface of these numerous examples, we engage in several case studies. In the case studies we endeavor to better understand why regulators are sometimes sitting on the sidelines, following the litigation rather than leading it. To that end, we present, in roughly chronological order, three case studies – on asbestos, MTBE, and PFOS -- that illustrate the sequence of our title: regulatory failure, tragic exposures, and mass litigation. Even given the detailed nature of these case studies, the causes of these regulatory failures are still not always easy to identify. Collectively, however, they demonstrate a number of reasons why tort liability so often precedes protective regulation even though timely regulation, often already authorized by statute, could have prevented many of the exposures that eventually resulted in tort liability.

https://www.georgiahealthnews.com/2020/09/sterilization-companies-hit-wave-ethylene-oxide-lawsuits/ (summarizing the 2020 lawsuit and its relationship to EPA’s research). 18 The final 2024 EPA rule cuts permissible emissions of ethylene oxide by ninety percent. See, e.g., Fact Sheet: Biden-Harris Administration Takes Action to Protect Communities from Dangers of Ethylene Oxide Pollution and Secure America’s Medical Supply Chains, available at https://www.whitehouse.gov/briefing-room/statements-releases/2024/03/14/fact-sheet-biden-harrisadministration-takes-action-to-protect-communities-from-dangers-of-ethylene-oxide-pollution-and/ 19 For an example, see the case study of PFAS in Appendix 2.

NOTE TO CONFERENCE PARTICIPANTS: To ensure we located all of the relevant information occurring at the intersection of regulation and tort litigation, our case studies are currently way too long and need to be trimmed back for the text. The full case studies are provided in Appendix 2. To facilitate reading in the interim, however, we attempt below to extract some of the more relevant findings in bullet point form. 1. Asbestos • Protective legislation did not exist when the first successful litigation against an asbestos manufacturer for asbestotic disease was filed in 1969. • EPA promulgated relatively weak regulation under the Clean Air Act in 1973. • After enactment of the Toxic Substances Control Act (TSCA) in 1976, which directed EPA to oversee the chemical market precisely in order to prevent unreasonable risks posed by substances like asbestos, regulatory efforts stalled and ultimately failed. • More than 700,000 plaintiffs have sued thousands of defendants, seeking compensation for illness allegedly caused by exposure to asbestos. • Exposures, illness, and lawsuits continued after TSCA’s enactment. • Only in 2024 did EPA effectively ban asbestos. 2. MTBE • Notwithstanding TSCA, MTBE became widely used as a fuel additive with limited toxicity information. • Rules promulgated by EPA were insufficient to prevent release of MTBE from underground storage tanks with resultant widespread groundwater contamination, including drinking water supplies. • As groundwater contamination by MTBE spread and became known, hundreds of claims were filed across the country by states, municipalities and other public water systems, private water suppliers, and homeowners. 3. PFAS • Environmental discharges of PFOS, PFOA and other PFAS preceded protective legislation. • TSCA failed to provide EPA with timely knowledge of harmful characteristics of these 600+ PFAS compounds registered in its inventory. • Before EPA gave any consideration to regulating PFOA, tort litigation exposed evidence that PFOA is persistent, bioaccumulative, and toxic. • EPA initially pursued voluntary agreements to phase out use of PFOA, PFOS, and other long-chain PFAS, but took no action to ensure the safety of replacements. • More than two decades elapsed between settlement of the first PFAS lawsuit and promulgation of regulations setting binding drinking water standards for some PFAS and designating PFOA and PFOS as CERCLA hazardous substances. • Significant portions of EPA’s plan for regulating PFAS remain incomplete.

II. REGULATORY DESIGN: WHY AGENCIES ARE FAILING TO PROVIDE EARLY WARNINGS, OFTEN DESPITE PROTECTIVE LEGISLATION

When it comes to detecting and protecting the public against latent hazards, we expect regulators to come to the rescue and be the first on the scene, identifying and responding to early warnings of hazards before they become public health and litigation crises. But in far too many cases, our regulators are missing in action. No early warning is sounded. Or, even after it becomes clear that there are significant health risks associated with an underregulated substance, still little to no regulatory action is taken. Instead, tort litigation must pick up the slack. As one of many examples, a century after we learned of the latent hazards of asbestos20, and over fifty years after the first plaintiff convinced a jury of factual causation between asbestos and lung cancer21, EPA finally banned almost all uses and imports of chrysotile asbestos, the last form of asbestos known to be imported, processed or distributed within the United States.22 The fact that our regulatory system routinely disappoints expectations in anticipating and addressing latent hazards is well-established beyond our effort to collect the evidence in Section I.23 In a stunning body of work, sociologists, historians, and lawyers have carefully traced out why regulators sometimes neglect to act on early warnings of risk and/or failed take action even after widespread health risks are clear.24 We learn that in some cases, industry was able to insinuate itself so deeply into regulatory and legislative processes that it gained effective control over the pace and nature of the agencies’ regulatory response.25 In other cases, we learn that part

See infra {asbestos case study} (noting recognition of asbestosis in 1920’s). Borel v. Fibreboard Paper Prods. Corp., 493 F.2d 1076 (5th Cir. 1973) (affirming verdict for plaintiff on products liability theory). 22 89 Fed. Reg. 21,970 (Mar. 28, 2024). In two months the rule drew three challenges from industry and one from a non-profit organization claiming that the rule is not strong enough. Texas Chem. Counc. v. US EPA, No. 24-60193 (5th Cir. Apr. 19, 2024) (Petition for Review); Ohio Chem. Tech. Counc. v. US EPA, No. 24-3325 (6th Cir. Apr. 19, 2024) (Petition for Review); Am. Chem. Counc. v. US EPA, No. 24-11206 (11th Cir. Apr. 18, 2024) (Petition for Review); Asbestos Disease Awareness Org. v. US EPA, No. 241090 (D.C. Cir. Apr. 19, 2024) (Petition for Review). 23 Although the case studies recounted in two sequential European Environment Agency reports do not explicitly identify failed regulatory response as a recurring theme, it is a familiar pattern among the chapters. See, e.g., European Environment Agency, “Late lessons from early warnings: science, precaution, innovation”, Report No. 1/2013 (2013), available at https://www.eea.europa.eu/publications/late-lessons-2; European Environment Agency, “Late lessons from early warnings: the precautionary principle 1896–2000”, Report No 22/2002 (2002), available at https://www.eea.europa.eu/publications/environmental_issue_report_2001_22. 24 See, e.g., infra note == and accompanying text (books on mass tort litigation). 25 EPA’s regulation of underground storage tanks (USTs), which failed to protect the public from releases of MTBE, is an example. See infra {MTBE case study} (quoting an EPA official who explained that EPA “worked closely with representatives of the major oil companies” when drafting UST rules. See also, e.g., Valerie Brown and Elizabeth Grossman, How Monsanto Captured EPA (and Twisted the Science) to Keep 21

of the regulatory lapses can be chalked up to weak controls that were insufficient to counteract the predictable inclination of regulated actors to avoid learning or sharing bad news about their activities.26 In other settings, the agency’s statutes have been interpreted to demand far too heavy an evidentiary burden on agencies to justify regulatory action.27 Recommended correctives have targeted each of these regulatory deficiencies in isolation.28 In this section, we search for a more systemic explanation for the recurring regulatory lapses. To carry out the investigation, we compare the general goals and promises of our regulatory programs, on paper, against practical impediments that arise in practice to impede regulators from locating and acting on emerging hazards. As we discuss below,29 while we can identify the general existence of numerous impediments to protective regulation, whether and when they might explain a specific regulatory failure is generally non-transparent. Deliberative process privileges, limited transparency requirements imposed by the APA, and a host of other inevitable features of black-box decision-making mean that whether one or more “shackle” is constraining an agency from taking protective action may be unknowable unless a courageous whistleblower chooses to come forward. Yet the diverse array of shackles, considered against the recurring failure of the regulatory system relative to tort litigation, suggests there is a larger story to tell. At the risk of a spoiler alert, in our view the best explanation for the recurring regulatory failure is both patently obvious and profoundly pessimistic. Despite its position as the fourth branch and its oft-villainized position as a “deep state” filled with “unaccountable” bureaucrats,30 in reality, the regulatory state has become so embedded in, infused by, and directed by political controls from Congress and the Executive that agencies have vastly less practical discretion than is ordinarily assumed. This includes far fewer opportunities to issue and address early warnings of emerging hazards, at least when the industries affected enjoy significant political power. A. Early Warning Detection: on Paper We begin by documenting that it is in fact the responsibility of agencies to sound early warnings and engage in responses that are proportionate to the public risks. We then collect the reasons Glyphosate on the Market, In these Times (November 2017), available at https://inthesetimes.com/features/monsanto_epa_glyphosate_roundup_investigation.html . 26 The “forever chemicals” PFOA and PFOS are an example. See infra {Forever chemicals case study}; Robert Bilott, Exposure 95 (2019) (Noting that under the Toxic Substances Control Act, “a kind of honor system” governed manufacturers’ disclosure of toxicity information about existing chemicals). 27 EPA’s first effort to ban asbestos is an example. See Corrosion Proof Fittings v. EPA, 947 F.2d 1201, 1215 (5th Cir. 1991) (invalidating rule because EPA did not “consider all necessary evidence” and did not “give adequate weight to statutory language requiring it to promulgate the least burdensome, reasonable regulation required to protect the environment adequately”). 28 [citations to scholarly or advocacy work to be added]. 29 See infra on deliberative process, etc. 30 See, e.g., David Bernhardt, The Deep State is All Too Real, Wall Street Journal, May 9, 2023 (using these terms).

why this does not occur in many high-profile cases, despite widespread and costly public harms. We close by exploring why, even after the risks become evident as a result of private tort litigation, the regulatory system may still remain inattentive. This cumulative, birds’ eye view of regulatory failure provides us with a launching pad to consider more systematic reform proposed in Part III. Congress has created a number of expert agencies with the specific purpose of protecting the public from emerging hazards. The establishment of the EPA exemplifies these public aspirations for anticipatory regulation. EPA’s short founding charter clearly lays out the central problem the agency was created to address, namely that “[o]ur national government today is not structured to make a coordinated attack on the pollutants which debase the air we breathe, the water we drink, and the land that grows our food.”31 The nation therefore needed an agency to provide this protective work. Among EPA’s four enumerated functions in its charter is the expectation that it will conduct “research on the adverse effects of pollution and on methods and equipment for controlling it, . . . [and also] gather[] . . . information on pollution [to] . . . strengthen[] environmental protection programs and recommend[] policy changes.”32 A similar story can be told for other agencies with regard to the protective goals motivating their creation.33 After establishing a protective agency like EPA, Congress then passes one or more laws providing detailed directions for that agency to engage in early identification and response work with respect to emerging hazards. The Toxic Substances and Control Act (TSCA), for example, includes among its purposes that “adequate information should be developed with respect to the effect of chemical substances and mixtures on health and the environment and that the development of such information should be the responsibility of those who manufacture and those who process such chemical substances and mixtures.”34 In the Occupational Safety and Health Act, OSHA and the National Institute of Occupational Safety and Health (NIOSH) are expected to “explor[e] ways to discover latent diseases, establish[] causal connections between diseases and work in environmental conditions, and conduct[] other research relating to health problems.”35 To ensure that the agencies have the necessary tools to do this early detection work, Congress typically provides multiple early detection and response authorities. An illustrative but not exhaustive list of these authorities with respect to EPA is tallied up in Table 2.

See Reorganization Plan No. 3 of 1970, 35 Fed. Reg. 15,623 (Oct. 6, 1970), reprinted as amended in 5 U.S.C. app. at 202–07 (2006) 32 Id. 33 For example, the Occupational Safety and Health Act is intended “to assure so far as possible every working man and woman in the Nation safe and healthful working conditions.” 29 U.S.C. § 651(b). {Research ongoing;other e.g.s CPSC, FDA}. 34 15 USC 2601(b). 35 29 USC 651(b)(6).

TABLE 2 Statute TSCA

CWA

Authorization to use new information to broaden scope of regulation § 4(a) test rules [listed here because order must be based on “may present unreasonable risk” finding] § 5(f) rules

§ 304 water quality criteria & effluent limitation guidelines

OPA SDWA

RCRA CAA

CERCLA

OSH Act

§ 1412(b)(1)(B)(i) authorizing listing of previously unregulated contaminants that “may have an adverse [health] effect”` § 3001(a),(b) authorizing revision of haz waste criteria, listed hw, characteristic hw § 108(a) criteria documents, list of criteria pollutants, NAAQS subject to revision § 112(b)(2) EPA authorized to add HAPs “which present, or may present . . . a threat of adverse human health effects” § 102(a) authorizing designation of additional haz substs which when released may present substantial danger to public health

§ 6(b)(5) may set new standards when needed for statutory purposes to assure no material impairment of health if feasible § 20(a)(5) HHS may require employers to measure, report on exposures (listed here because can be used to support new standard)

Authorization to develop or collect new information on health effects § 4(a) test rules [listed here because a testing rule would develop new information] § 8(a)(2)(E), 8(d), 8(e) requiring reporting of envtl / health effects data (but see § 14 cbi) § 10 EPA & HHS “shall conduct such research, developing & monitoring . . . .” § 27 HHS “may conduct . . . projects for the development . . . of . . . methods” for determining health / envtl effects § 104(a),(b),(c) EPA shall establish programs for research into water pollution effects, collect and make results available, and conduct research on health effects w/HHS § 104(n) EPA shall research effects of pollution on estuaries § 7001(c)(4) interagency committee shall establish research program to monitor / evaluate envtl effects § 1442(a)(1)(B) EPA may conduct research & studies re: causes, diagnosis, treatment of disease or impairments, including improved methods to identify / measure health effects § 8001(a) EPA w/w/o DOE shall conduct research / studies re adverse health / welfare effects of release of hw § 103(a),(b),(i) EPA shall establish national R&D program on health / welfare effects & collect and make info available, and coordinate w/other fed agencies § 103(h) NIEHS may conduct program of basic research on health risks § 104(i) ATSDR shall establish & maintain inventory of research on toxic substance effects, develop tox profiles for priority substances, and if info inadequate to determine health effects, “shall assure the initiation of a program of research designed to determine the health effects” including lab & other studies § 20(a)(1),(4) HHS shall conduct research & experiments on OSH, including “to explore new problems” [but we know what happened….] § 20(a)(5) HHS may require employers to measure, report on exposures USC § 699a NIOSH shall expand research on bioterrorist threats USC § 671a requires study of take-home exposures (since completed)

Congress’ grand legislative plan is unmistakable. In the US, the protective agencies are established in order to learn of emerging hazards and assemble rigorous scientific evidence to locate and respond to public risks.36 Indeed, the agencies’ expert role in doing this work assessing risks and “finding facts” is lauded as the central (and perhaps the only legitimate) contribution of regulators, even by those who are otherwise skeptical of the breadth of the agencies’ delegated authorities.37 So . . . . what happened? B. Early Warning Detection: In Practice. As any lawyer knows, the devils are in the details. Once we trace out the regulatory reality of these broad-brush, aspirational mandates and tools to see what is practically possible from the perspective of a regulator operating in the trenches, reality looks quite different. Despite numerous statutory authorities that direct agencies to investigate and respond to emerging public hazards, multiple constraints impede regulators from doing this important work.38 We emphasize in this account only the constraints on regulators since the goal is to offer a critical check on the conventional wisdom that regulators generally enjoy “broad regulatory discretion.” However, even accepting that our analysis offers a worst case view, it is evident that at least with respect to the agencies’ ability to provide and act on early warnings, there are many more “shackles” undermining those regulatory efforts than is ordinarily assumed, at least when viewing regulatory actions from high altitude. 36

Although these statutory mandates direct the agencies to engage in early detection work, the courts have concluded that any resulting agency inaction in carrying out their assignments (if no deadlines are specified in the statute) is effectively not judicially reviewable. The courts do not want to become involved in micro-managing agency priorities and hence leave it to agency discretion whether and how to carry out their discretionary duties. See, e.g., Sidney A. Shapiro, Rulemaking Inaction and the Failure of Administrative Law, 68 Duke Law Journal 1805-1842 (2019); Daniel E. Walters, Symmetry's Mandate: Constraining the Politicization of American Administrative Law Symmetry's Mandate: Constraining the Politicization of American Administrative Law 119 MICH. L. REV. 455 (2020). And how the agency manages its duties and priorities is also typically non-transparent. All internal decision-making can be classified as deliberative in nature and protected from FOIA. As a result, the public may not know what is going on inside the agency black box regarding early hazard detection unless the agency decides to volunteer this information. See, e.g., United States Fish & Wildlife Service v. Sierra Club, Inc., 141 S. Ct. 777 (2021) (providing an expansive interpretation of the scope of deliberative process protections). 37 See, e.g., Gundy v. United States, 139 S. Ct. 2116, 2139 (2019) (Gorsuch, J., dissenting) (emphasizing that under the nondelegation, the agencies’ constitutional role is primarily to fill in the details and find facts); Dep’t of Transp. v. Ass’n of Am. R.Rs., 575 U.S. 43, 78-97 (2015) (Thomas, J., concurring in the judgment) (emphasizing that the agencies’ constitutional responsibilities are to make “factual determinations” rather than broad policy). 38 For more complete discussions of these shackles, see, e.g., Thomas McGarity, Wendy Wagner, U.S. Agency Experts in Shackles: The Quest for Information, 35 Journal of Environmental Law 65–86 (March 2023).

This section first maps out some of these general “shackles” limiting the agencies’ ability – in practice – to provide early warnings despite mandates that suggest otherwise. The second subsection then identifies a parallel set of reasons why, even after tort litigation documents widespread public harms from a toxic hazard, the agencies are often constrained from responding in a protective way. Readers will recognize some of these shackles from the Section I case studies, but others tend to be obscured to view and yet may also help explain why the regulatory system sometimes lags tort litigation despite the construction of institutions intended to do the opposite. 1. Ability to generate research Ideally, an expert agency tasked with responsibility for identifying emerging hazards before they cause widespread public harms would be equipped with substantial scientific capabilities to do this research in-house. Agencies, in theory, would be able and indeed expected to use their resources and authority to initiate scientific research on toxicity, fate, and exposures in ways that inform the assessment of potential public risks. Indeed, this is among the primary reasons Congress created agencies like EPA.39 However, it may come as no surprise that these aspirations are only as good as the budgets Congress and the Executive ultimately bestow on the agencies.40 And, in the case of identifying emerging hazards, those budgets are often far too limited to allow the agencies to conduct serious investigative work.41 The CPSC provides a stark example. Among the limited statutory purposes enumerated in CPSC’s authorizing statute was the goal of “promot[ing] research and investigation into the causes and prevention of product-related deaths, illnesses, and injuries.”42 This role was viewed by Congress as vital to enable CPSC to then protect the public “against unreasonable risks of injury associated with consumer products,” “an unacceptable number of

See supra note 31 - 33. Jody Freeman and Sharon Jacobs, Structural Deregulation, 135 Harv. L. Rev. 585 (2021) (discussing how presidents control agency budgets in political ways to manage agency priorities and workload); Eloise Pasachoff, The President’s Budget as a Source of Agency Policy Control, 125 Yale Law Journal 2182 (2017) (describing the statutory mandate under which the President controls the budget that is proposed to Congress each year). 41 See generally McGarity and Wagner, supra note, 38, at 7-8. While falling outside the ambit of public hazards, a considerable body of research traces how Congress has endeavored to use budget controls deliberately to slow and even halt the listing of endangered species by the U.S. Fish and Wildlife Service. See, e.g., Noah Greenwald and others, Shortchanged: Funding Needed to Save America’s Most Endangered Species 3 (Research report for the Center for Biological Diversity 2016) (discussing how budget cuts exacerbated the Service’s inability to complete recovery plans for 22% of its listed species); see also Consolidated Appropriations Act, Pub. L. 115–141, § 120 (2018) (forbidding any federal funds from being used to fulfill protections or potential protections for the sage grouse). 40

[which] . . . [we]re distributed in commerce [at the time the legislation was passed].”43 Yet, despite these noble goals, CPSC has been chronically underfunded for decades; it has only a very modest budget and staff that lacks the ability to learn of latent toxic harms in consumer products before they cause injuries.44 By contrast, EPA’s premier scientific department, the Office of Research and Development (ORD), is equipped with state-of-the-art research facilities and staffed with impressively-credentialed scientists who provide a much more robust research role for the agency. But even here – when set against breadth of the Executive and Congressional ambitions regarding EPA’s role in anticipating emerging risks – the agency’s resources are not even close to the early warner expectations set out for the agency in legislation.45 2. Contribution of industry in risk detection and assessment. The pharmaceutical companies, under the FDCA, and the pesticide companies, under FIFRA, are legally required to submit a battery of tests and research to support their applications to market products.46 These statutory requirements place the burden on industry to assess at least some health and environmental risks before the products can be introduced into commerce. Indeed, under the FDCA, industry is now also required to collect and share with FDA all relevant adverse event reports concerning approved drugs post-market.47

15 USC 2051(a)(1) & (3) See, eg, Rachel E. Barkow, ‘Insulating Agencies: Avoiding Capture Through Institutional Design’ (2010) 89 Tex L Rev 15, 67 (discussing CPSC’s limited budget and staff); Justine Fuga, Trading Public Nuisance for Product Safety: Reviving the Office of Technology Assessment, 13 Drexel L Rev 489, 505– 07 (2021) (describing the resource-intensive work needed for the CPSC research lab to provide ex ante protection for emerging technologies and its failure to keep up with such protections). 45 [Research ongoing.] Moreover, since some statutes assign heavy research responsibilities on EPA as a precondition to deadline-driven regulation (e.g., the required five-year review of the seven ambient air pollutants), a significant portion of the ORD’s research budget must necessarily be dedicated to these tasks, leaving fewer opportunities for exploratory work. Congressional budgetary priorities are also starkly obvious with respect to the National Institute for Environmental Health Sciences (NIEHS), which was created for “the conduct and support of research, training, health information dissemination, and other programs with respect to the factors in the environment that affect human health,” 42 U.S.C. § 285I, and which has a specific mandate for toxicological research, see 107 Stat. 122 § 463A. In fiscal year 2023, Congress appropriated about $1 billion to NIEHS, accounting for slightly more than two percent of the total budget of the National Institutes of Health. NIEHS, Report to the National Advisory Environmental Health Sciences Council 1 (Feb. 21, 2023). 46 See, e.g., FIFRA, 7 USC § 136a (governing pesticide approval); FDA, 21 USC § 355(b), (c). 47 See, e.g., 21 U.S.C. 356b. This requirement eventually exposed that the anti-inflammatory drug Vioxx caused heart attacks, causing the popular drug to be removed from the market and spurring about 47,000 lawsuits that the manufacturer settled for nearly $5 billion. See McDarby v. Merck & Co., 949 A.2d 223, 229–50 (N.J. Super. Ct. App. Div. 2008) (summarizing the marketing, research, and regulatory history of Vioxx); Snigda Prakash & Vikki Valentine, Timeline: The Rise and Fall of Vioxx (Nov. 10, 2007), available at https://www.npr.org/2007/11/10/5470430/timeline-the-rise-and-fall-of-vioxx (describing the settlement). 44

However, under all other environmental and health-protective statutes, responsibility for assessing the all aspects of risky industrial activities (not simply the latent hazards, but also the economic benefits of the activity, the potential exposure pathways, and the conditions of use) rest almost exclusively with the agency.48 Industry bears little to no responsibility for providing much, if any, evidence regarding the risks of their products or activities. For example, under TSCA, the manufacturers of the 50,000 existing chemicals sold in commerce are not required to conduct any research or even a literature search on the toxicity of their chemicals, provide any evidence of the economic and social benefits of their chemicals, or shoulder any responsibility for tracing out the potential fate and transport of their chemicals after manufacture.49 The same heavy informational burden is placed on the regulatory agencies overseeing the manufacture of dietary supplements,50 cosmetics,51 and virtually all consumer products that contain hazardous substances.52 The same is also true for regulatory programs governing industries that emit hazardous pollutants into the air, water, or onto land,53 or that generate them in the process of production in ways that present risks to workers or neighbors.54 Even in the “prior approval” regulatory regimes of FIFRA and FDCA, however, fine print tucked into parts of the authorizing statutes serve to limit industry’s responsibilities for information generation in important ways. For example, under FIFRA EPA is barred from requiring industry to provide information on the social need for a proposed pesticide.55 Thus, even though the statutory approval process is conditioned on EPA’s risk-benefit analysis that a given pesticide does not present an “unreasonable risk” to health and the environment,56 Congress has statutorily barred agencies from soliciting from industry information that is central to this calculation. In a related context under FIFRA, after EPA regulators devised a way to require

See generally Wendy E. Wagner, Commons Ignorance: The Failure of Environmental Law to Produce Needed Information on Health and the Environment, 53 Duke Law Journal 1619-1745 (2004) (making this argument); McGarity & Wagner, supra note 38, 8-12 (same). 49 15 U.S.C. § 2604(d)(1)(A) (industry is required only to disclose the risks that are “known” or “reasonably ascertainable”). 50 See, e.g., Michael A. McCann, Dietary Supplement Labelling: Cognitive Biases, Market Manipulation and Consumer Choice 31 Am. JL and Med 215, 251 (2005) (observing that “[t]he lack of mandatory [Adverse Event Reporting] reflects perhaps the greatest defect of DSHEA.... [T]he inspector general for the Department of Health and Human Services estimates that the [current] system reveals less than 1% of actual adverse reactions to dietary supplements”). 51 {Research ongoing.} 52 See, e.g., Gulf South Insulation v. U.S. CPSC, 701 F.2d 1137 (5th Cir. 1983) (placing a heavy evidentiary burden on CPSC as a precondition to regulate consumer products). 53 See, e.g., Wagner, Commons Ignorance, supra note ==. 54 {Research ongoing. Cite OSHA/Residual risk.} 55 See 7 USC 136a(3)(c)(5) (prohibiting EPA from making the “lack of essentiality” a basis for denying a pesticide registration). 56 See 7 USC 136a(a).

industry to produce more exculpatory research to rebut a presumption of toxicity, Congress swiftly amended the statute to raise the information burden on EPA.57 3. Acquiring Information from Industry Files. Even if the agencies cannot enlist industry directly in providing data and research on potential risks, the agencies are still equipped – through their authorizing statutes – with various tools to extract information from regulated parties deemed relevant to the agencies’ assessments of emerging hazards.58 Since the regulated industries create the processes and substances, they are generally privy to a great deal of information that might not otherwise be available to regulators.59 Even without a requirement that industrial actors conduct their own inhouse research, then, the information access tools available to agency regulators are important. And facially the agencies seem well-equipped to access whatever information they need and in some cases even to order the industry to conduct additional in-house information collection. Most protective statutes provide agencies with authorities to request relevant in-house information through information requests60 and also impose adverse effects reporting requirements on industry to learn of continuing risks.61 Both the chemical regulatory statute, TSCA, and the pesticide law, FIFRA, even equip EPA with the power to order testing on particular chemicals.62 TSCA also authorizes EPA to order industry to begin systematically collecting in-house information on a range of topics, including existing evidence of environmental effects and exposure.63 Agencies can also enter facilities (albeit by administrative warrant) to collect information and/or assess potential violations.64

See 7 USC 136a(c)(8) (requiring “validated test or other significant evidence raising prudent concerns of unreasonable adverse risk to man or to the environment” as a precondition to regulatory restrictions on pesticides). 58 See, e.g., CAA Section § 114 (broad information-collection authority for existing information in industry files under the Clean Air Act); CERCLA § 104(e) (same broad authority under Superfund statute). 59 See, e.g., Cary Coglianese, Richard Zeckhauser, and Edward Parson, ‘Seeking Truth for Power: Informational Strategy and Regulatory Policymaking’ (2004) 89 Minn L Rev 277, 289 (describing the manufacturers’ superior control of information and how it provides advantages over regulators in practice). 60 See, e.g., 42 USC 9604(e) (authority to submit information requests to industry under CERCLA). 61 See, e.g., 15 USC § 2607(c), (e) (under TSCA); 7 USC § 136d(a)(2) (under FIFRA);15 USC § 2064 (b) (CPSC). 62 See, e.g., 15 U.S.C. 2603(a) (test rules under TSCA); 7 U.S.C. 136a(c)(2)(B) (data call in authority under FIFRA) 63 See 15 U.S.C. 2607(a). 64 When the agency does not wish to give the industry a “heads up” but conduct a surprise inspection, then the agency will need to go through the work of obtaining an ex parte warrant. See, e.g., Marshall v. Barlow's, Inc., 436 U.S. 307, 324 (1978) (holding that warrantless searches under the Occupational Safety and Health Act, 29 U.S.C. §§ 651-678 (2000), are unconstitutional).

Yet, again, the outward façade of vigorous information collection authorities is more limited upon closer inspection.65 For example, under the original TSCA, EPA was statutorily required to present evidence that a chemical caused a non-trivial risk to health and the environment as a prerequisite to imposing a testing order; an obvious Catch-22.66 Even in the much-lauded 2016 amendments to TSCA that purport to address this impediment, EPA still carries the burden of establishing some need for the testing and must specify in detail how the testing must be done.67 EPA must also follow a series of procedural steps before it can order industry to conduct testing, a process that can be (and regularly is) challenged by the affected industries.68 But beyond idiosyncratic impediments tucked within specific statutes, all agency information collection authorities come with an additional, significant string attached; namely, for each piece of solicited information that an agency intends to collect from more than nine separate entities, the agency must first justify imposition of the information demand on industry under the 1980 Paperwork Reduction Act.69 The mandated PRA clearance process – administered by the White House Office of Management and Budget – necessitates that before the information request can be sent, the agency must first promulgate a rulemaking for its information collection proposal.70 In that proposal, the agency must document not only its need for the information but assess the burden it places on industry.71 OMB oversees these rulemakings and regularly imposes further conditions narrowing an agency’s proposed request.72 Observers note that the resources and time needed for an agency to comply with the PRA (a rulemaking takes on average a year to complete) sometimes cause agencies to “forgo useful information collection, delay actions that would support useful rulemaking, and inhibit program improvements that would meaningfully improve government functioning.”73

Wagner, Commons Ignorance, supra note =, at 1670–77 (describing the many ways that agencies lack statutory authority to access information vital to their regulatory missions). 66 See, e.g., Chem. Mfrs. Ass'n v. EPA, 859 F.2d 977, 984 (D.C. Cir. 1988) (holding that the EPA must establish a "more-than-theoretical" probability of an unreasonable risk in order to require additional testing). 67 See 15 USC 2603(a) (setting forth multiple conditions the EPA must meet to order testing). 68 See, eg, Lynn L. Bergeson, ‘Section 4—Testing of Chemicals and Mixtures’ in Lynn L. Bergeson and Charles M. Auer (eds), New TSCA: A Guide to the Lautenberg Chemical Safety Act and Its Implementation (1st ed., American Bar Association 2017) at 42–48. EPA can also issue orders to require this information, but in doing so, it must first explain why a rulemaking was not possible, See, e.g., id. at 47. 69 44 USC § 3501–3520. 70 44 USC § 3503(3)(A)(i). 71 See, eg, Jeffrey S Lubbers, Paperwork Redux: The (Stronger) Paperwork Reduction Act of 1995, 49 Admin L Rev 111, 114 (1997); Joseph D. Condon, Stop Regulating Government Paperwork With More Government Paperwork, 9 MICH. J. ENVTL. & ADMIN. L. 213, 226–28 (2020). 72 Condon, supra note =, at 215, 231–32; see also McGarity & Wagner, supra Shackles, at 17. 73 Condon, supra note =, at 215.

4. Sharing Information Underlying Risk Assessments. Central to the public accountability of agencies is the ability of the broader public and stakeholders to access and ultimately oversee the agency’s work and analyses. Access to the agency’s own records (including what it acquired from industry) is essential to this public participation. Yet agencies are even shackled from sharing some of this key internal information with stakeholders. Perhaps most significant, information with trade secret protections claimed by industry is exempted from public disclosure.74 The benefits of remaining out of the public eye, coupled with the expansive reach of trade secret privileges have led to significant overclaiming by industry in a number of regulatory programs.75 As a result, considerable information that the public needs to check the agency’s work identification and response to emerging hazards is simply not publicly available. For example, even with amendments that purport to limit industry overclaiming, about 20% of the chemicals sold on the open market are “trade secret” protected in all respects; we simply do not know that the chemicals are sold, much less what is known about health risks, exposures, or other features.76 But even for the remaining 30,000+ chemicals sold in commerce, roughly 90% of those chemicals involve at least one (and generally many more) trade secret claims that can conceal the identity of the manufacture, location of the facilities, nature of the uses of the product, process and waste streams, etc.77 Other statutes provide other types of impediments to public access to vital information available to agencies. In the approval of pesticides, for example, Congress prohibited the public from accessing the underlying industry health studies that support an application until after the EPA’s approval is final.78 Even at that point, the public can only “view” the research in EPA offices in DC after signing a form shared with the manufacturer about that person’s identity and reason for viewing the research.79

See 5 USC 552(b)(4). See, e.g., Wagner and Michaels, Equal Treatment for Regulatory Science: Extending the Controls governing the Quality of Public Research to Private Research, 30 AMERICAN JOURNAL OF LAW & MEDICINE 119, 129-35 (2004) (recounting the evidence of overclaiming of trade secrets by industry). 76 See, e.g., Wendy Wagner & Steve Gold, Legal Obstacles to Toxic Chemical Research, 375 SCIENCE 138, 139 (2022) (making this point); {as of May 2024, chemical identity is CBI for 6,038 out of 42,377 (14.2%) of “active” chemicals in commerce on the TSCA inventory}. 77 Id. The initial decision about whether, and how expansively, to assert trade secret protection is, of course, left to the entity submitting information. Thus, in annual reports voluntarily submitted to EPA under the PFOA Stewardship Program, see infra {Forever Chemicals Case Study, App. 1}, some reporting companies claimed trade secret protection for virtually all the information submitted; others claimed protection for both production data and emission data; and others claimed protection for both production data. {add citations}. 78 See 7 USC § 136h(g)(1). 79 Id. 75

As discussed further below, Congress also provided an exception to FOIA for internal deliberative processes.80 This exception blocks public access to key information stakeholders would normally need to assess the veracity of the agency’s technical analyses, including in some cases the technical assessments prepared by scientific staff to inform the agency’s decision process.81 These various impediments do not mean of course that an interested member of the public cannot use FOIA to access at least some information regarding the agencies’ efforts to identify, trace, or restrict an emerging hazard (if a proposed rule has not yet issued). But of course the member of the public would have to learn of the agency’s efforts to know what to request. 5. Defending against Industry Opposition. Once an agency publicly signals that it is investigating an emerging hazard, opposition from those using and producing it will likely follow.82 This opposition will be particularly strong if the agency proposes some kind of action – such as an information request, requiring testing, or preparing rules that propose to list a substance as hazardous. Undertaking any kind of early warning analysis by the agency thus can be expected to draw down resources far beyond the preliminary costs of accessing and analyzing the scientific information. Indeed, a significant portion of the agency’s resources slotted for early hazard detection could well go to defending the work against opposition rather than conducting the analyses themselves. As one example, during the first forty years of its chemical regulatory program, EPA promulgated rules requiring industry to conduct further testing on a select few chemicals.83 For each of these test rules, industry’s involvement began early in the agency’s work; on average, industry engaged with the agency 28 times per rule before a proposed test rule was ever published.84 During notice and comment, industry was typically the only group filing comments – averaging eight formal comments per rule – which were then followed by an average of 15 more communications between industry and the EPA before a rule was finalized. Despite its extensive involvement throughout the development of each the test rules, industry still appealed several in court; ultimately EPA revised every single test rule multiple times, presumably to stave off further litigation.85 As a result in part of this vigorous industry opposition, in over forty years

See 5 USC 552(b)(5). See, e.g., United States Fish & Wildlife Service v. Sierra Club, Inc., 141 S. Ct. 777 (2021) (allowing deliberative process privilege for agency’s scientific-technical analyses underlying an endangered species listing). 82 Wagner, Commons Ignorance, supra note =, at 1649-59 (recounting some of this literature). 83 See, e.g., Deliberative Rulemaking, at 626-35 (describing these test rules in detail). 84 See id. at 627-32 (providing this data and discussing industry’s extensive involvement in all of the test rules). 85 Dynamic Rulemaking at 212, 217–18, 222–26. 81

EPA promulgated test rules for only 200 chemicals out of an inventory of chemicals (most untested) that exceeded 50,000.86 6. Executive Branch Engagement. Thus far, the only shackles we have considered are those imposed by Congress or the Executive Branch to impede or constrain the agency’s ability to engage in early detection work. However, under the increasingly expansive unitary executive, the concept of agency independence free of political control is a relic of the past. The Chief Executive – typically working through delegates appointed within the agencies – enjoys absolute power to control all aspects of the agencies’ policy work, including early hazard detection.87 And, down in the trenches, there is evidence (thanks to whistleblowers) that some political officials have taken full advantage and overtly manipulated the scientific information produced by agency staff.88 For example, on some occasions when staff scientists produced analyses antithetical to the President’s campaign promises or ideological views, those analyses have been terminated or suppressed.89 Appointees have also sometimes worked with staff – at varying levels of collegiality – to influence and sometimes alter their data and findings.90 Appointees can and do sometimes prohibit staff from communicating their findings with the scientific community and the public at large.91 Appointees can and do determine staffing, redistribute recalcitrant staff experts, and of course decide on the appropriate resource allocations and budgets.92 And appointees can and do determine whether and how agency assessments go through external peer review (including selecting the reviewers), even when this external review is required by statute.93 This Executive Branch engagement in the expert work of agency staff, we might hope, can at least be documented to provide the public with a sense of the extent of political interference in 86

See U.S. GOV'T ACCOUNTABILITY OFFICE, REPORT No. GAO-05-458, CHEMICAL REGULATION: OPTIONS EXIST TO IMPROVE EPA's ABILITY TO ASSESS HEALTH RISKS AND MANAGE ITS CHEMICAL REVIEW PROGRAM 18 (2005), available at http://www.gao.gov/new.items/d05458.pdf. 87 See generally Wendy Wagner, A Place for Agency Expertise: Reconciling Agency Expertise with Presidential Power, 115 COLUMBIA LAW REVIEW 2019 (2015) (documenting this power). 88 See, e.g., Thomas McGarity & Wendy Wagner, Deregulation using Stealth Science Strategies, 68 DUKE L.J. 1719, 724-46 (2019) (providing examples). 89 Id. 90 Id.; {we may add discussion of EPA’s toxicity assessment for PFBS, the PFAS 3M developed to replace the forever chemical PFOS. The first version of the document, issued just before the Trump Administration left office, was changed at the behest of a political appointee. A few months after the Biden Administration took office, the document was revised to reverse the changes. EPA’s Office of Inspector General later concluded that the political interference “left the public vulnerable to potential negative impacts on human health.” US EPA, The EPA’s January 2021 PFBS Toxicity Assessment Did Not Uphold the Agency’s Commitments to Scientific Integrity and Information Quality 1 (Mar. 7, 2023).} 91 Id. 92 Id. at 1747-56. 93 Id. at 1756-67.

the agency’s protective work. But once again our institutional design disappoints. All interactions among staff or between staff and others in the agency (whether appointees, their own delegatees, or others), can (and routinely are) privileged from public disclosure as “Executive Process.”94 According to the Supreme Court, even the agency’s technical analyses used to assess whether a species is endangered can be sequestered from public view when the agency claims it is inherently political in nature.95 The prospect of political control over scientific staff analyses does not doom the agencies’ ability to provide early warnings; agencies can and do assess many emerging hazards despite political control. Indeed, some administrations are credited with encouraging the agency to do more protective work and may even allow the analyses to be done free of political interference. However, because of the deliberative process privilege, those on the outside the agency cannot determine when staff analyses represent the work of the agency’s experts and when it has been influenced by political considerations. The only way to truly assess the reliability of an agency’s hazard assessment is to replicate it. C. Acting on Early Warnings. When agencies do attempt to respond to early warnings (often resulting from information produced during tort litigation), they find themselves shackled by limitations in their statutes, mandates imposed by presidential executive orders and constraints on executive branch action imposed by reviewing courts. These shackles afford stakeholders and political operatives multiple opportunities to delay or forestall regulations designed to address the risks identified by the early warnings. 1. Congressionally Imposed Shackles. a. Procedural Shackles. Having empowered agencies to respond to early warnings, Congress has imposed procedural restrictions on agency action that make it difficult for agencies to do their jobs. For example, the Occupational Safety and Health Act empowers the Occupational Safety and Health Administration to promulgate occupational safety and health standards, but the statute provides for judicial review under the “substantial evidence” test that the Administrative Procedure Act reserves for formal rulemaking.96 This feature and suggestions in the statute’s legislative history convinced OSHA to provide trial-type hearings with experts presenting testimony subject to 94

See generally Shilpa Narayan, Note, Proper Assertion of the Deliberative Process Privilege: The Agency Head Requirement, 77 Fordham L Rev 1183, 1187–202 (2009). 95 See Sierra Club, supra note 36. 96 29 U.S.C. § 655(f).

cross-examination by multiple parties to the rulemaking.97 In addition to requiring a huge resource commitment, the hearings lengthen the decision-making process.98 Congress has also imposed generic procedural requirements that make it difficult for agencies to act rapidly in response to early warnings. For example, the Small Business Regulatory Enforcement Fairness Act (SBREFA) required EPA and OSHA (but not other agencies) to convene a “review panel” of government officials and small business representatives and consider their views prior to publishing major regulations.99 This first bite at the apple, enacted during the anti-government Gingrich Congress in 1996, quite obviously advantaged business interests over the interests of the beneficiaries of environmental and occupational safety and health legislation. b. Analytical Shackles. In 1980, Congress passed the Regulatory Flexibility Act, which requires agencies to prepare “regulatory flexibility analyses” detailing the impact on small businesses and small governmental entities of rules that have a substantial impact on a significant number of small businesses or small governmental entities.100 Since most important health and safety regulations affect significant numbers of small entities in significant ways, the regulatory flexibility analysis requirement is a hurdle that the agencies must clear before they can react to early warnings. The SBREFA allowed small businesses to seek judicial review of the content of the impact statements that agencies were obliged to prepare under the Regulatory Flexibility Act of 1980.101 This attachment point for judicial review ensured that the agencies could not ignore the regulatory flexibility analysis requirement. To avoid remand and possible vacatur, agencies became very careful not to omit any possible impact that the regulated industry might seize upon to persuade a reviewing court to remand the regulation to the agency for further analysis.102 The same Congress that passed SBREFA enacted the Unfunded Mandates Reform Act of 1995, which requires executive branch federal agencies to prepare written statements for all proposed and final regulations that would “impose an enforceable duty upon the private sector” 103

29 C.F.R. pt. 18, subpart A; See also, 15 U.S.C. § 57a(b), (c), (d) (1988) (specifying procedures for rulemaking by the Federal Trade Commission that go beyond the requirements of the APA); 42 U.S.C. § 7607(d) (1988) (specifying procedures for certain rulemakings under the Clean Air Act that go beyond the requirements of the APA). 98 When compared to the other shackles discussed here, however, trial-type hearings are at best a modest hindrance. Thomas O. McGarity, Some Thoughts on “Deossifying” the Rulemaking Process, 41 Duke L. J. 1385, 1399 (1992). 99 Small Business Regulatory Enforcement Fairness Act, Pub. L. 104-121 (1996). 100 Regulatory Flexibility Act, Pub. L. 96-354 (1980). 101 Small Business Regulatory Enforcement Fairness Act, supra. 102 {research ongoing}. 103 Pub. L. No. 104-4, 104th Cong., 1st Sess. (1995), 2 U.S.C. 1501.

of more than $100,000,000 (adjusted for inflation) in any given year, with certain exceptions.104 The statement must contain “an identification of the law under which the regulation is being promulgated,” a “qualitative and quantitative assessment of the anticipated costs and benefits” of the mandate on the private sector “as well as the effect of the Federal mandate on health, safety, and the natural environment.”105 This statutory requirement did not add greatly to the inability of agencies to react to early warnings because (as discussed below) the agencies have since 1981 been required to prepare similar analyses by executive orders issued by every president since Ronald Reagan. Should a president ever rescind the executive orders, however, the agencies would still be subject to the considerable analytical hurdle imposed by the Unfunded Mandates Reform Act. In addition, the agencies must “identify and consider a reasonable number of regulatory alternatives and from those alternatives select the least costly, most cost-effective or least burdensome alternative that achieves the objectives of the rule.”106 This substantive requirement, which goes beyond the requirements of the current executive orders, has the potential to confine discretion considerably to the extent that it is enforced by the Office of Information and Regulatory Affairs in the White House Office of Management and Budget, as discussed in more detail below.107 c. Deregulatory Riders. Riders to appropriations acts and other “must-pass” legislation are a source of virtually invisible congressionally imposed shackles on agency discretion.108 A rider is a provision attached to an unrelated bill that Congress enacts and the President signs without going through the normal legislative process (with the possibility of a filibuster) and without the threat of a presidential veto.109 Since the 1990s, allies of regulated industries in Congress have employed such riders to “stall, modify, or eliminate an ongoing regulatory program that is being implemented by a regulatory agency pursuant to duly enacted authorizing legislation.”110 For example, after two lengthy government shutdowns due to Congress’ failure to pass an appropriations bill, President Clinton in April 1996 signed an appropriations bill that contained riders prohibiting the Occupational Safety and Health Administration from completing its ergonomics standard and limiting EPA’s ability to issue certain drinking water standards.111

104

2 U.S.C. §§ 658(7); 1532. Id. at § 1532(a). 106 Id. at § 1535(a). 107 {x-ref} 108 Thomas O. McGarity, Deregulatory Riders Redux, 1 Mich. J. Env. & Admin. L. 33 (2012). 109 Id. at 36. 110 Id. at 26. 111 Id. at 50. Another example, in the Endangered Species Act context, is the recurring rider preventing the Fish and Wildlife Service from listing the sage grouse as a threatened or endangered species. See supra note 41. 105

Shackles imposed by riders lack transparency; they can be hidden in 1,000-page appropriations bills and therefore receive little or no public attention.112 Because they do not come from congressional authorizing committees with expertise in the subject matter and a sense of ownership over the relevant issues, the shackles are imposed with virtually no deliberation or debate.113 They are quietly enacted at the behest of industries (or even individual companies) that were unable to persuade the regulatory agencies or Congress to impose the shackles through the normal regulatory and legislative processes. Some beneficiaries of regulations that are delayed or never implemented because of riders may have criticized the agencies charged with issuing those regulations, but the real blame lies with Congress for tolerating deregulatory riders. 2. Presidentially Imposed Shackles. Over the past 45 years, presidents have imposed shackles on agency rulemaking that have limited their ability to respond to early warnings. a. Impact Analysis Requirements. Presidents have issued a number of executive orders requiring agencies to prepare a wide variety of analyses of various aspects of major regulations.114 Agencies must prepare regulatory impact analyses setting out the costs and benefits of proposed and final major regulations as well as significant alternatives.115 A separate executive order requires agencies to include in the regulatory impact analysis “an evaluation of the environmental health or safety effects of the planned regulation on children,” and “an explanation of why the planned regulation is preferable to other potentially effective and reasonably feasible alternatives considered by the agency.”116 While most, if not all, of these analyses may be desirable aids to sound agency decision-making, the requirements add greatly to the agency’s rulemaking burden, and they have great potential to slow down or halt (in cases where OIRA and the agency cannot agree on the content117) the agency’s progress. b. Additional Considerations.

112

McGarity [Deregulatory Riders], supra note =, at 72. Id. at 73-76. 114 Administrative Conference of the United States, Rulemaking Requirements from the Executive Office of the President (August 15, 2023) 115 Executive Order 12,866, 58 Fed. Reg. 51,735 (October 4, 1993). 116 Executive Order 13,045, 62 Fed. Reg. 19,885, 19,887 (April 21, 1997) 117 See infra {OIRA discussion, subpart d below}. 113

In addition to the factors that authorizing legislation requires agencies to consider, executive orders also require agencies to take into account other considerations in writing their rules to the extent consistent with agency statutes. For example, President Clinton’s “environmental justice” executive order requires executive branch agencies to identify and address disproportionately high and adverse human health or environmental effects of its programs, policies, and activities on minority populations and low-income populations.”118 Another executive order signed by President Clinton requires agencies to adhere, to the extent permitted by law, to prescribed criteria designed to advance federalism.119 Although these additional considerations cannot require the agencies to achieve a result inconsistent with their statutes, they require the agencies to devote resources to ascertaining those impacts and to documenting their consideration of those impacts in the preambles to the regulations they promulgate. c. Centralized OIRA Review. The Office of Management and Budget has played a powerful role in agency rulemaking since the Reagan Administration through its obscure Office of Information and Regulatory Affairs (OIRA), which Congress created in the Paperwork Reduction Act of 1980 to oversee agency paperwork requirements. Very early in his tenure, President Reagan signed Executive Order 12291, which required agencies to prepare for every major rule a comprehensive analysis of the costs and benefits of multiple alternatives. When not prohibited by statute, agencies were obliged to choose the alternative with the greatest net benefits. The Office of Information and Regulatory Affairs (OIRA) in OMB was to review all regulatory analysis documents for compliance.120 During the Reagan Administration, OIRA became adept at using its procedural power to delay the review process to secure substantive changes in the regulations under review.121 It soon became clear that political as well as objective criteria entered the mix and that OMB was also serving as a gatekeeper in advancing the administration’s deregulatory agenda.122 It was also 118

Executive Order 12,898, 59 Fed. Reg. 7629 (February 16, 1994). President Biden reaffirmed President Clinton’s order and more fully specified the ways in which agencies are required to consider environmental justice in all their activities, including rulemaking. Executive Order 14,096, 88 Fed. Reg. 25,251 (Apr. 21, 2023). 119 Executive Order 13,132, 64 Fed. Reg. 43,255, 43,256 (August 10, 1999). 120 Executive Order No. 12291, 3 C. F. R. 127 (1982). 121 George C. Eads & Michael Fix, Relief or Reform? Reagan's Regulatory Dilemma (1984) at 111; Graham K. Wilson, The Politics of Safety and Health (1985), at 80; American Bar Association, Section of Administrative Law and Regulatory Practice, Report to the House of Delegates, in 1 ABA Reports with Recommendations to the House of Delegates, 1990 mid-year meetings 101A (February 12-13, 1990) at 7. 122 Oliver A. Houck, President-ex and the new (approved) decision making, 36 Am. U. L. Rev. 535 (1987) at 540-41; William F. West and Joseph Cooper, The presidency and public policy making 192 (George C. Edwards, Steven A. Shull and Norman C. Thomas, eds, 1986) at 201.

clear that OIRA was performing its review function behind closed doors, thereby hiding from public view the extent to which OMB served as a powerful “conduit” for industry views.123 Over time, mid-level managers in the agencies got the message, and they began to craft regulations in a more business-friendly fashion to enhance the likelihood that they would survive OIRA review.124 The nature and extent of OIRA’s influence over agency rules waxed and waned in subsequent Administrations, not necessarily in ways that might be predicted based on the President’s political party.125 Over the years, various reforms have somewhat increased the transparency of OIRA’s role.126 In general, however, OIRA has throughout its history used its power to review agency regulations to shackle agencies that have tried to implement protective statutes by promulgating major regulations. As illustrated by the long delay that OIRA imposed on EPA’s 1989 asbestos regulations, OIRA has delayed the regulatory process, sometimes for years. It has given regulated entities and their trade associations a powerful second bite at the apple as OIRA considers agencies’ proposed and final regulations. And it has imposed its skeptical view of the value of health, safety and environmental regulations on the agencies, sometimes in direct conflict with the agency’s authorizing statute. This, despite the fact that no statute assigns this shackling role to that office. 123

Joan Claybrook, Retreat from Safety (1984), at xix; George C. Eads & Michael Fix, Relief or Reform? Reagan's Regulatory Dilemma (1984) at 108, 111; Harold H. Bruff, Presidential Management of Agency Rulemaking, 57 Geo. Wash. L. Rev. 533 (1989) at 580. 124 Rochelle L. Stanfield, EPA Administrator Lee Thomas is more a Manager than a Policy Maker, [volume] MAT. J. 391 (February 15, 1986), at 393. 125 During the George H.W. Bush administration, OIRA became such a political lightning rod that the Senate never approved a permanent head of the office. Charles Kolb, White House Daze (1994), at 78; Kirk Victor, Tale of the Red Tape, 24 Nat. J. 684 (March 21, 1992) at 687. As a result, the regulatory oversight role was played primarily by the Council on Competitiveness, headed by Vice President Dan Quayle. During the Obama administration, OIRA played a significant deregulatory role under the leadership of Cass Sunstein. For example, in March 2012, Sunstein wrote a memorandum to the executive branch agencies commanding them to be sensitive to the costs and cumulative effects of the rules that they promulgate and to “simplify requirements on the public and private sectors.” Cheryl Bolen, FEDERAL AGENCIES TOLD TO TAKE INTO ACCOUNT CUMULATIVE COSTS OF RULES ON BUSINESS, 43 Env't Rep. Cur. Dev. (BNA) 764 (March 23, 2012). During the Trump Administration, agency heads were so committed to a deregulatory agenda that OIRA had little additional role to play in delaying or interfering with regulation. See James Goodwin, The Progressive Case Against OIRA (Center for Progressive Reform, August 6, 2020). 126 The legislation that reauthorized OIRA in 1984 required that all documents exchanged between OIRA and the agencies during the review process were to be made public after the final rule was promulgated, and required OIRA to provide the relevant agency with copies of all correspondence with outsiders, advise the agency of all oral communications, and invite an agency representative to all meeting with outsiders. Harold H. Bruff, Presidential Management of Agency Rulemaking, 57 Geo. Wash. L. Rev. 533 (1989) at 582-83. During the Clinton Administration, OIRA maintained a public log setting out the status of all regulatory actions it reviewed and a notation of all written communications that it received from outside sources. Robert Duffy, Regulatory Oversight in the Clinton Administration, 27 Presidential Studies Quarterly 71 (1997); OMB Files Disclose Regulatory Meetings Between Officials, Private Interest Groups, [volume] Occupational Safety and Health Reporter 816 (1993).

III. REFORM As long as agencies are controlled by the political branches (which seems an immutable and perhaps constitutionally required feature of administrative design), the prospect for agencies providing comprehensive early warnings of impending hazards seem grim.127 In settings in which significant economic consequences can follow agency action, we can expect the multiple mechanisms for political interference to stealthily muzzle and delay the agencies’ capacity to identify and respond to emerging toxic risks. In this concluding section we propose an institutional reboot – the creation of a new body designed to fulfill this sentinel (early warning) role that the federal regulatory regime so desperately needs. This new, “Sentinel” agency would be an arm of the federal government, but it would be located completely outside of the political process. The Sentinel’s task would be strictly informational -- to identify and assess emerging latent hazards in need of attention. The Sentinel itself would be comprised of a small blue-ribbon panel of carefully selected, highly regarded scientists equipped with the needed resources and authorities to conduct original research, prepare literature syntheses, access private information, and establish monitoring programs designed to track potential future hazards. A legislative charter establishing and authorizing the new scientific agency would provide it with sustained funding, basic directions, and information collection authorities. But how the Sentinel proceeds in identifying and assessing emerging hazards would be entirely determined by it. The Sentinel, in this way, will be well-positioned to get ahead of the tort litigation curve with respect to at least some emerging hazards in ways not currently provided by the efforts of our current agencies or the academic community at large. Indeed, the Sentinel would be wellpositioned to provide the judiciary with valuable guidance on particularly challenging mass causation questions that arise with some regularity in tort litigation and have, in the past, damaged the courts’ reputation.128 By drawing attention to largely ignored, but potentially deleterious hazards, the Sentinel will also jumpstart scientific interest among the research

127

See, e.g., EU, Late Lessons, supra note =, at 168 (discussing how, based on individual case studies, many of the “late lessons” from early warnings can be ascribed to political resistance to regulate those hazards). 128 One of the greatest threats to the legitimacy of the courts is their capacity to assess general causation, and this has become a central theme not only of academic work, see, e.g., see MICHAEL D. GREEN, BENDECTIN AND BIRTH DEFECTS: THE CHALLENGES OF MASS TOXIC SUBSTANCES LITIGATION 20–22 (1998); MARCIA ANGELL, SCIENCE ON TRIAL: THE CLASH OF MEDICAL EVIDENCE AND THE LAW IN THE BREAST IMPLANT CASE (1996), but also has driven significant (and partly successful) efforts to reform the tort system. See, e.g., PETER HUBER, GALILEO’S REVENGE: JUNK SCIENCE IN THE COURTROOM (1991); STEVEN J. MILLOY, JUNK SCIENCE JUDO: SELF-DEFENSE AGAINST HEALTH SCARES AND SCAMS (2001).

community in these new topics.129 Finally, in providing this early warning service, the Sentinel will provide an important check on regulatory lapses that otherwise fly under the radar by creating greater accountability for political actors that endeavor to impede the agencies’ efforts. After fleshing out the basic concept for this new agency, we then discuss logistical issues related to the Sentinel’s legislative charge, its powers, and other more specific implementation issues. We close the section by anticipating challenges. A. The Sentinel Idea At base, the Sentinel would be a scientific institution whose only responsibility would be to provide scientific information on emerging risks. To that end, it would be structured to represent the best minds in the scientific community, encompassing a broad interdisciplinary mix of the relevant sciences. It would be required to follow the highest scientific standards of integrity, including for avoiding conflicts of interest, engaging rigorous peer review, and maintaining complete independence from political influence and control. While scientific work is clearly not value-free (as decades of literature make clear), the Sentinel’s work nevertheless will be governed by professional and not political oversight. To ensure the work remains responsive to the broader public, the Sentinel’s work priorities will be partly independently determined and partly guided by the general public (working through public interest nonprofits) and by regulated industry, as discussed below. The most critical design feature of the Sentinel is ensuring its complete independence from political control and influence with regard to how or what it does with its early warning research responsibilities.130 Congress, in creating the Sentinel, would need to abdicate its subpoena power and capacity to conduct oversight hearings with respect to the Sentinel. And the Executive would be stripped of all authority to recommend or guide the Sentinel’s work. Instead all communications between Executive agencies (as well as Congress) and the Sentinel directors and staff would need to be logged on a public, online record that provides details of the nature of the communications. The chair and about twenty scientific directors that operate the Sentinel (perhaps with a larger staff of several dozen or more scientists) would be selected by the NAS, with no constraints on the NAS other than to require full disclosure of conflicts of interest in the viable candidates, notice and comment on the slate of candidates, and a requirement that NAS follow its own 129

This service is certainly performed by tort litigation ex post, see supra Section I, but engaging the Sentinel will provide a much more robust set of priorities for future early hazard research as well as triggering this research sooner before they cause mass harm. The scientific reputation of the Sentinel will likely nudge the research priorities of scientists and funding agencies alike. 130 See, e.g., Late Lessons at 178 (emphasizing the need for independent scientific analysis that is free of stakeholder and political controls and influence).

existing guidelines on conflicts of interest and selection process.131 But again, the NAS would have final say on how to weight those considerations and the selection of scientists would be non-reviewable by the courts, Congress and the Executive Branch. The NAS would also determine the appropriate disciplinary mix of scientists to fill out the board of directors. Although delegating the task of staffing the Sentinel will rest considerable, unchecked powers with the NAS, the NAS seems to be the best institution among the alternatives to set up the fledgling agency. As the preeminent independent scientific institution serving the US for more than 180 years, the NAS enjoys both trust and experience, as well as enforcing the highest scientific standards for its members.132 This positions it to be able to make basic staff selections in a rigorous and relatively objective way. Indeed, studies of politically insulated agencies find that precisely because their institutional survival depends on their perceived neutrality, they will be meticulous about following high standards of scientific integrity, including political independence.133 The NAS’s track record on avoiding appearing to favor one political party over another generally tracks those predictions.134 And, while the NAS is certainly not above reproach or criticism,135 even skeptics concede that it is highly sensitive to perceptions of bias or political favor and tends to, if anything, go overboard in trying to steer an apolitical course in the

131

The legislation governing the Sentinel’s creation could reference NAS conflict standards or other principles, see, e.g., NAS, Composition and Balance, Conflicts of Interest, and Independence: Policies and Procedures, available at https://www.nationalacademies.org/about/institutional-policies-andprocedures/conflict-of-interest-policies-and-procedures, and demand basic transparency on all aspects of the scientists’ dossiers with regard to past work and conflict of interest. Notice and comment on the slate of candidates will also be vitally important to open up the NAS’ process to public view and solicit information from the public on the candidates that might not otherwise be available to the NAS. See, e.g., U.S. General Accounting Office, Federal Advisory Committees: Additional Guidance Could Help Agencies Better Ensure Independence and Balance 43 (Washington, D.C.: GAO, 2004) (underscoring the value of the notice and comment step). Additional requirements could also be added, such as requiring the NAS to interview candidates to ensure they are suitable for this kind of highly exploratory and knowledge-expanding work. 132 See, e.g., NAS Code of Conduct, available at https://www.nasonline.org/about-nas/code-ofconduct/nas-code-of-conduct.pdf. RA: Cite on value of NAS. 133 See, e.g., Bruce Bimber, The Politics of Expertise in Congress: The Rise and Fall of the Office of Technology Assessment (1996) (discussing this dynamic as partly explaining the OTA’s success in remaining above the political fray); Lawrence McCray, Doing Believable Knowledge Assessment for Policymaking: How Six Prominent Organizations Go about It, draft (February 11, 2004), https://www.files.ethz.ch/isn/19524/McCray-DoingBelieveableKnowledgeAssessment.pdf (describing “high end” independent scientific organizations and exploring how they have managed to maintain their credibility on contested issues of regulatory science). 134 Research ongoing. 135 The harshest source of criticism is the tendency of the NAS to over-reach its charge and opine about policy-relevant issues. See, e.g., Ian Fein, Reassessing the Role of the National Research Council: Peer Review, Political Tool, or Science Court, 99 CAL. L. REV. 465 (2011). Stephen Hilgartner, Science on Stage: Expert Advice as Public Drama (Stanford, Calif.: Stanford University Press, 2000); Center for the Science in the Public Interest, Ensuring Independence and Objectivity at the National Academies (Washington, D.C.: CSPI, 2006).

selection of its own committee members.136 It therefore follows that while the NAS may not welcome the onerous responsibility of establishing and selecting the members of the Sentinel, they will nevertheless do the job well because their future credibility and hence existence depends on it. Once formed, the Sentinel will enjoy similar independence from all political influence and control. However, while the Sentinel’s independence would not be constrained in conducting its work, the Sentinel would be required to document all ex parte communications with any external parties, including government agencies, in a public log that each director and the staff maintain online. Except for that critical transparency requirement, there would be no federal controls on how the Sentinel approaches its work, except for a few relatively minor legal guardrails discussed below. Thus the Sentinel would need to be explicitly exempted from statutes like FACA and FOIA, although several (perhaps three) years after a file is open on an emerging hazard, certain specified internal documents could perhaps be subject to public release. B. Enumerated Functions and Added Guardrails on the Sentinel’s Operation Imbuing a group of experts with both complete independence in their work (provided they follow high scientific standard) and control over what and how they do their work is a largely unprecedented and admittedly risky move; even the NAS issues reports primarily in response to political requests. As a result, the Sentinel must be gently guided towards producing consensus reports and risk assessments on the most pressing public hazards that have received less study and public attention.137 Drawing in part from recommendations from the European Environmental Agency’s “Late Lessons from early warnings” reports, we recommend three, roughly equal functions for the Sentinel to keep its focus on investigating emerging hazards to be explicated in its legislative charter. First, to address a longstanding data deficit that has stymied the scientific community for decades, the Sentinel would be responsible for developing and coordinating existing monitoring programs and research to better inform scientific understanding of emerging hazards.138 Second, the Sentinel would be expected to identify and assess emerging risks that

136

Research ongoing The institutional niche the sentinel agency fills, in other words, is the role of pioneer in locating widespread risks for which scattered scientific research suggests hazards or in tracing out preliminary evidence or theories that largely uninvestigated classes of chemicals – like was the case with PFAS – demand fuller scientific attention and research given their widely dispersed and potentially toxic nature. 138 See, e.g., EU Late Lessons, supra note =, at 171-72 (underscoring the need for more systematic monitoring to assist in the detection of emerging hazards); EU Late Lessons Report #2, supra note = , at 675 (underscoring the need for “more long term monitoring of biological and ecological systems, focusing on ‘surprise sensitive’ parameters such as bees, amphibians, invertebrates, foetuses, etc.”). There are many monitoring programs and granting opportunities in place, but they appear completely 137

have the potential for widespread public harm using the available data and literature, supplemented as needed by original research.139 The Sentinel should also be encouraged to give priority to those emerging hazards that it believes are most likely to be under-addressed by agencies precisely because they involve significant economic disruptions and/or are shrouded in information held by industry or locked in inaccessible databases.140 (Obviously this charge will be politically controversial). Third and finally, the Sentinel would be required to institute an open nominations portal by which any person or institution (including the judiciary and members of Congress) can recommend study of a particular emerging risk or offer information. To ensure there is some direct engagement with the broader affected groups, the Sentinel would be required by its charter to produce a limited number of reports in response to formal nominations that are made annually by two government-organized panels: one representing the general public and nonprofit communities and another representing regulated industry.141 Industry might, for example, request the Sentinel to assess hazards that it believes are being subjected to causally-unjustified tort liability. The process for these parallel nomination tracks would need to be spelled out in the authorizing charter.142 With regard to the second and third functions, the Sentinel would be expected to locate existing datasets, research, and literature – including from industry -- and synthesize them to assess the extent and dangers of various emerging risks. Ideally this assessment would include not only predicted harms to the US public but global populations.143 The Sentinel would also be expected to engage in exploratory work (including using information authorities discussed below) on suspected potential hazards for which little to nothing is known. The Sentinel’s synthesis work, once in draft, would be subject to rigorous external peer review (reviewers might even be selected by a separate scientific board), potentially multiple times, and the peer review disconnected from one another and from policymaking more generally. See, e.g., Annie Brett, ‘Information as Power: Democratizing Environmental Data’ (2022) 2022 Utah L Rev 127 (2022). 139 With regard to this particular function the Sentinel would be expected to prioritize cutting edge, interdisciplinary synthesis work that is unlikely to arise within the scientific community based on publication and related pressures. See, e.g., EU Late Lessons #2, supra note, at 627-28 & 635 (discussing the tendency of scientists to examine only better-studied hazards and how their research priorities are influenced by publication biases); see also Late Lessons, supra note =, at 174-75 (emphasizing the need for practically-grounded assessments of potential hazards that goes well beyond the scientific literature). 140 See, e.g., Late Lessons Report #2, supra note =, at 695 (emphasizing “[t]he need for research to focus more on the potential hazards of emerging technologies in addition to research on product applications” as well as a greater focus on “emerging hazards rather than on well-known risk”). 141 See, e.g., Late Lessons, supra note =, at 177, 183 (emphasizing the critical role that lay publics should play both in prioritizing and guiding assessments of emerging hazards as well as informing the inquiries with local knowledge). 142 For example, the nomination committees for these two distinct groups would need to be exempt from the “balance” requirement of FACA because the panelists are specifically selected to represent distinct sectors and views and hence be “unbalanced”. 143 The US’s industrial leadership, coupled with deficient oversight of emerging risks of these manufacturers, may be responsible for an extraordinary amount of damage to the global commons more generally. Research ongoing.

assessments would be made public.144 Since there are so many existing Executive agencies already engaged in aspects of this early warning work – all of which are under political management – the Sentinel would also be expected to coordinate their efforts with these agencies to avoid duplication.145 However, what and how the Sentinel approaches its work will ultimately be left to the discretion of its scientific board of directors. In terms of ensuring a steady output, the Sentinel may also need to be encouraged in its authorizing legislation to produce a minimal number of “recommended” outputs or reports. In addition to reports that respond to the public and industry nominations, the Sentinel might be expected to release a minimum number of (five?) additional reports annually on emerging hazards of concern. The Sentinel might also be expected to publish two versions of each report - one for the scientific community and one for lay audiences.146 In its charge, the Sentinel must also be reminded explicitly of its limits, the most important being that it must not issue policy recommendations or engage in broader value-laden judgments about the costs and benefits of emerging risks. Rather, the Sentinel’s assignment would be limited to the preliminary scientific task of identifying hazards with scientific techniques and data and synthesizing that literature into comprehensible findings. While some of this pioneering work will require novel scientific methodologies, the Sentinel must be reminded in its charge to be particularly diligent about identifying the uncertainties inherent in this work.147 Peer reviewers should also be alerted – explicitly – to pay careful attention to assertions in Sentinel draft reports

144

This is how NAAQS process works. See, e.g., Wendy Wagner, “Science in the Administrative Process” at 29-40 (2013) (commissioned by ACUS), available at http://www.acus.gov/researchprojects/science-administrative-process (summarizing the NAAQS process that utilizes these same processes) 145 The agencies that would need to be consulted include IARC; NTSB; NIOSH; NAS; Great Lakes Science committee; NIEHS and NTP; and EPA’s IRIS. 146 Other more specific features of the Sentinel’s establishment could require that reports include some basic, minimal features. Canvassing the literature, it appears there are at least three features that are considered important. First are acknowledgements of uncertainty and a willingness to identify emergent risks in ways that do not fit the old statistical models. Second, is outreach to impacted communities for information. Third is an expectation that the consensus reports include concrete conclusions, however qualified by the existing uncertainties. See EU, Late Lessons, supra note =, at 169 and 185. The Sentinel could also be encouraged to engage in certain types of assessments, such as life cycle analyses. Id. This might include an expectation that the Sentinel rate hazards or provide other comparative benchmarks so that readers can understand the societal implications of its work. For example, the Sentinel could be expected to assign a rating to the risk such as (1) the available information supports ending inquiry into the studied risk [hazard?]; (2) the available information supports continued consideration of the studied risk by the Sentinel; (3) the available information supports urgent additional research on the studied risk; (4) the available information supports urgent additional research on the studied risk and taking precautionary action pending the results of that research; or (5) the available information supports urgent action to reduce or eliminate the studied risk. 147 It is on this point that the NAS has perhaps most often drifted off course and entered a buzz saw of criticism for overreaching into policy recommendations. See supra note == (with Fein).

that overstep the Sentinel’s scientific role by providing insufficient explication of significant underlying assumptions or uncertainties. Finally, the legislation establishing the Sentinel must provide the new agency with a full set of legal authorities to access any and all information needed to assess emerging risks. These information tools include authorities to request information from industry – similar to the agencies’ information collection authorities -- only without the requirements of the Paperwork Reduction Act. Congress would need to provide the Sentinel with a generous research budget, and a dedicated source of funding independent of the annual appropriations process, to fund studies that are expected to inform its work.148 Finally, the Sentinel will require clearance (perhaps through nondisclosure contracts) that allow it to access all confidential information currently held by industry as well as by the agencies and courts, including trade secrets, sealed court records, homeland security, and deliberative process information. C. Anticipating Skepticism The creation of a new scientific agency, even one with a relatively small staff and budget, may seem not only radical, but politically out of reach. The unconventional nature of the agency with respect to independence also raises both practical and theoretical concerns. We endeavor to anticipate some of the most significant challenges here, but acknowledge that there is a lot more work to do in refining such a proposal. 1. Won’t Politicians Simply Ignore the Sentinel’s Warnings? Skeptics will question whether rigorous scientific warnings will ultimately make a difference to regulatory action. Given the existing political impediments to regulation, there is a risk that the Sentinel’s hopefully highly regarded reports will simply line library shelves and congressional offices without altering the current course of action. Political inaction is likely, at least for some of the hazards. However, the Sentinel – if it is established to be independent and guided by an interdisciplinary blue-ribbon board of scientists – will produce scientific signals that are more likely to inform and guide public understanding of emerging risks in ways that lead to policy responses. As opposed to the adversarial science emerging from court rooms, journals, and in the media – which involves often heavily-financed and carefully planned strategies to produce ends-oriented research, advance false scientific claims, and throw “doubt” on relatively robust scientific conclusions149 – the Sentinel’s consensual, highly regarded work will serve as a kind of scientific referee on the state of the 148

See, e.g., EU Late Lessons, supra note =, at 172 (emphasizing the need for more systematic research dedicated to scoping out potential emerging hazards). 149 See Tess Legg et al., supra note ==.

science at the time.150 Indeed, we need only reference the Science Panel established in the PFOA class action as an example of what rigorous, blue-ribbon scientific panels can accomplish in terms of focusing and advancing scientific understandings of emerging risks in ways that generate desirable policy outcomes.151 Again we concede that it is not always the case that even rigorous consensual science triggers a policy response, but collective scientific engagement with the challenges of identifying emerging risks is arguably a vital first step if we are to take the goal of preventative protection seriously. Moreover, a formal Sentinel agency comprised of the very best scientists working interdisciplinarily to identify potential widespread health hazards creates precisely the kind of institution that is not likely to arise on its own.152 The EU’s Late Lessons reports underscore why the academic scientific community, while certainly invaluable, has not been more helpful in identifying and assessing emerging hazards.153 Publication and professional biases, as well as the strictures of obtaining grant funding, lead scientists to study the more familiar toxic agents while neglecting investigations into emerging risks.154 The lack of interdisciplinary study and limited access to “local” knowledge (including information in industry’s files) also causes the relatively limited academic work that is done on emerging hazards to be afflicted with blind

150

Both NAS and the Health Effects Institute (HEI) (an independent scientific blue-ribbon organization discussed at infra notes == and accompanying text) provide exactly the same referee function but only upon request. One could argue IPCC does the same service as the Sentinel with respect to climate change, although there is more controversy regarding their operation and management. Research ongoing 151 See {Bilott at 243 (check p)}; In re E. I. du Pont de Nemours & Co. C-8 Personal Injury Litig., 337 F. Supp. 3d 728, 732, 735 (S.D. Ohio 2015) (emphasizing that Science Panel consisted of “three completely independent, mutually-agreeable, appropriately credentialed epidemiologists” and followed “specific instructions on how to choose the objective criteria and protocols it would utilize in its studies”); see also infra {Forever ChemicalsCase Study}. 152 Cf. EU, Late Lessons, supra note =, at 168-69 (listing the need for government-funded investigatory work that, among other things, provides long-term monitoring and research into early warnings; identifies and works to reduce blind spots and gaps in knowledge; identifies and reduces interdisciplinary obstacles to learning; and identifies and reduces institutional obstacles to learning and action); EU, Late Lessons #2, supra note =, at 639 (emphasizing the need to consider societal needs in the prioritization of research; to conduct innovating research aimed at extending current knowledge; communicating findings clearly; and making the research openly available and independent of vested interests). Relying on the academic community to somehow act as private sentinels through the publication of their individual works is unrealistic in other ways as well, including the individualized (and nonconsensual nature) of individual research projects (which again are vulnerable to ends-oriented attack by affected parties); the scientists’ own reluctance to engage in politically-charged topics (particularly given the industry harassments that have transpired for those unlucky scientists who have published damaging work); and the spirit of science which not only seeks out the truth but does so in a decidedly skeptical way that is essentially antithetical to efforts to generate consensus. Indeed, the IPCC remains controversial precisely because it endeavors to collect scientists to opine collectively on politically-salient risks; something that scientists are programmed to avoid doing voluntarily. Research ongoing. 153 See, e.g., supra note =and accompanying text (discussing tunnel vision and publication bias). 154 EU Late Lessons #2, supra note =, at 635.

spots and to under-state critical uncertainties.155 Indeed, the Sentinel will not only provide a critical public service in producing consensual reports for the public, but will trigger more independent scientific inquiry as a result of its discoveries. 2. Will the Sentinel Actually do the Work? Without deadlines and specific directives, there is a justifiable worry that the Sentinel will engage in a great deal of synthesis “work” but ultimately be unable to reach consensus on anything, thus suffering from a kind of paralysis by analysis. Particularly since the Sentinel’s assignment is both open-ended and demands scientific creativity and unconventional methods, inaction by the Sentinel seems possible. And there is also a risk, hopefully less significant, that the experts appointed to the Sentinel will engage in their own pet projects in ways that miss the action or even distract from more important problems.156 The legislative charge establishing the Sentinel might be able to head off some of these risks. First, the experts appointed to the Sentinel could be required to undergo a rigorous interview process with the NAS that specifically seeks to ensure their willingness (and ideally past academic track record) to venture into scientific areas that are poorly conceptualized and that might not lead to publishable findings.157 Second, NAS could be tasked with not only setting up the Sentinel, but reviewing or even “auditing” its performance annually. (Or some other agency like GAO or perhaps the NAS might be better suited to conduct this annual audit work in public reports). 3. If the Sentinel Does the Work and it Does Change the Scientific Conversation, Will That Lead to Its Political Demise?

155

See, e.g., EU Late Lessons, supra note =, at 169, 171, 173-74 (discussing blind spots and insufficient attention to various uncertainties). 156 Some of these challenges occurred in past governmental efforts to enlist the scientific community to better understand emerging risks, although these efforts did not involve the creation of a consensus panel. For example, a large grant process to support the scientific study acid rain produced what policymakers regarded as a great deal of expensive, unhelpful research that essentially advanced scientists’ own pet projects. See, e.g., Daniel Sarewitz, Frontiers of Illusion: Science, Technology, and the Politics of Progress 62-63 (1996) (describing the National Acid Precipitation Assessment Program and the U.S. Global Change Research Program as examples of “the divergence between research trends and social needs”). 157 See, e.g., EU Late Lessons, supra note =, at 169 (emphasizing the need for scientists involved in assessing emerging hazards to be “alert to—and humble about—the potential gaps in those bodies of knowledge” currently available to assess hazards). The scientists would need to be entrepreneurs capable of thinking outside conventions to the extent it means proceeding with high levels of unknowns, uncertainties, and questions and yet still make progress.

On the flip side, there is a very real risk that if the Sentinel is too successful in its work, the political process will be quick to bring its activities and even its existence to an abrupt halt. (Or at least equally worrisome, the Sentinel may become increasingly risk averse in its selection of projects over time to minimize the political fallout). This is another undeniable possibility. With respect to the risks of complete dismantling, there is at least one reason to be cautiously optimistic that the Sentinel might survive such political maneuvers. Experience has shown that overt – highly salient -- dismantling of public health and environmental protections tends to be politically unpopular. To completely eliminate a successful Sentinel will likely be an uphill task for opponents. The prospect of incremental and largely invisible shackling through political channels, however, seems much likely and even inevitable. The best antidote against this risk of more incremental meddling is to anticipate it in a bullet-proof charter that guards against more predictable lines of attack on the Sentinel’s budget, staffing, investigative powers, and functions. The more explicit and clear the charter, the more impediments to opponents in recalibrating its powers in largely invisible ways.158 However, the risk that the scientific staff in the Sentinel will grow more cautious and less entrepreneurial over time is a more challenging possibility to protect against. Perhaps rotating different chair-persons on a regular basis (two years), keeping the scientific board itself on threeyear terms, ensuring annual audits by a GAO-like institution (to include audits of the NAS’ own process for selecting the board), and perhaps adding legal protections to the scientific board, such as immunity not only from congressional but judicial subpoena powers, misconduct proceedings, etc., will stave off familiar means of harassing and chilling the work of individual scientists.159 Additionally, if there is the prospect of large professional awards for engaging in the sentinel work rigorously (e.g., Pulitzer Prizes, Nobel Prizes, NAS Awards, etc.) this might also offset the professional risks of sticking one’s neck out. Although it would be more challenging to establish, the Sentinel could also be created as a nongovernmental entity, thus dodging all of these political control worries.160 The Health Effects 158

A great deal of legislative mischief arises precisely because there is no “complexity” cap on legislation, allowing for members to insert many features that go unnoticed and can even be passed without a comprehensible understanding of the implications. See, e.g., Wendy Wagner with Will Walker, INCOMPREHENSIBLE! A STUDY OF HOW OUR LEGAL SYSTEM ENCOURAGES INCOMPREHENSIBILITY, WHY IT MATTERS, AND WHAT WE CAN DO ABOUT IT chpt. 7 (2019). 159 Perhaps anticipatorily the charter could also designate a trusted referee (again the NAS) to investigate credible charges against the Sentinel or staff. Although it is a stretch, perhaps Sentinel members might be given expansive legislative defamation-like protections enforced by the DOJ to protect the members against non-credible attacks against their scientific integrity. 160 See, e.g., EU Late Lessons, supra note =, at 187 (suggesting that the nongovernmental institutions can provide early warning services if structured properly). A privately funded creation of the Sentinel is a potential alternative path. {We are considering whether to suggest also Executive Branch creation of the sentinel as a possibility. Executive creation would necessarily be temporary and limited and serve only to get the ball rolling with a variety of challenges involved as a political control matter.} Added study is needed to lay out the pros and cons of these alternatives.

Institute (HEI), which provides highly valued and respected “impartial” scientific reviews on “the health effects of air pollution”, is a private, scientific organization by EPA and the motor vehicle industry.161 One of HEI’s roles is to serve as scientific referee for disputes over high profile policy relevant science.162 Despite its nongovernmental creation, HEI’s work is very highly regarded and appears to be trusted by all sides.163 This suggests that a privately funded Sentinel is also not implausible, although the Sentinel’s task -- which reaches industries that might find they have more to lose than gain from the added scientific attention than was the case for the auto industry (e.g., the chemical industry) -- might be difficult to finance and/or to establish, particularly in ways that are viewed as reliable and unbiased. 3. Isn’t the Creation of the Sentinel Politically Impossible? Both the legal and political viability of the Sentinel proposal – particularly in light of loudly voiced concerns about the “deep state” – also present serious challenges. On the legal side, creating a government institution with great independence moves in precisely the opposite direction from the Supreme Court’s pressing concerns about unaccountable agencies and its tendency to accept arguments based on the theory of the unitary executive. However, since the Sentinel’s only tasks would be informational in nature, its work should not trigger any Separation of Powers concerns regarding unconstitutional delegations of “lawmaking” power. The political impediments to the establishment of a Sentinel agency are vastly more challenging, however. While the Sentinel could in theory be created by an enterprising President using his inherent authorities [?], it will still need a budget that transcends election cycles and hence would likely require legislative enactment. Such a step will encounter partisan and polarized opposition that could sink its prospects. However, again there may be several ways to design the Sentinel’s charter to anticipate and defuse at least some of this opposition. First, as already mentioned, the Sentinel’s legislative charge could include an expectation that the Sentinel annually conducts assessments on several (two) topics that industry identifies as important. Since industry has voiced concerns over phantom risks and false positives leading to costly and unnecessary regulation in the past,164 and

161

See “What is the Health Effects Institute” at https://www.healtheffects.org/about. See generally Thomas McGarity & Wendy Wagner, BENDING SCIENCE: HOW SPECIAL INTERESTS CORRUPT PUBLIC HEALTH RESEARCH 262-63 (co-authored with Tom McGarity) (2008) (providing an overview of the HEI’s role in mediating politically charged scientific disputes). This role included refereeing challenges to the Six Cities, a controversial study done at Harvard finding spikes in mortality connected with exposure to fine particulates in the air. Id. 163 See, e.g., Sheila Jasanoff, The Fifth Branch: Science Advisers as Policymakers 209-16 (1990). Research ongoing for more recent literature on HEI. 164 See, e.g., Steffen Foss Hansen and Joel A. Tickner, The precautionary principle and false alarms – lessons learned, in EU, Late Lessons #2, supra note =, at 17. 162

has been indisputably subject to tort liability that scientists believe was causally mistaken,165 this nomination process may be an attractive feature that counterbalances the Sentinel’s risks at least to some industries (particularly the pharmaceutical and medical device companies).166 Second, issues related to climate change should be explicitly excluded from the sentinel’s areas of potential study. Climate change is not only politically divisive, but a well-organized lobby exists to oppose greenhouse gas restrictions that seem likely to oppose the creation of such an agency. By taking climate off the table – which also avoids conflict and possible duplication with the prominent role of the IPCC – some industry resistance to creation of the Sentinel agency might be defused. Third, the diffuse and unpredictable nature of the emerging risks that the Sentinel investigates may ultimately make the Sentinel’s work less threatening to individual industries ex ante. With unspecified targets and unclear ramifications of the resulting information, some industries may find that mustering up vigorous opposition is not worth the candle. It could be unpalatable to be seen as being against more information. By contrast, though, it does seem likely that the very well organized and economically and politically powerful chemical industry will find itself in the sight lines of the Sentinel and invest considerable effort to oppose it. Countervailing broad support to overcome industry’s potentially vigorous opposition may not be sufficient, particularly given the diffuse benefits of scientific information on emerging risks. Incremental steps towards the Sentinel idea – perhaps established tentatively by a committed and enterprising President,167 by the Federal Judicial Center to assist with high-profile general causation cases, or as a nongovernmental institution – could at least get the camel’s nose under the tent. Alternatively, the charge of the Sentinel itself could sidestep the chemical industry initially by focusing on Sentinel on a more limited set of potential hazards -- like consumer products – that are not backed by as powerful and well-organized lobbies and for which the chemical industry is less directly implicated. These incremental steps still make progress and may begin to establish a more robust concept for a Sentinel service that could gradually be expanded over time. IV. CONCLUSION Litigation precedes the regulation of many catastrophic public hazards for a number of reasons, but one underlying explanation is that there is currently no centralized regulatory body – acting 165

See supra note = and accompanying text (Angell). Even if judgments based on jury verdicts are unanimously reversed, such as in the Bendectin cases, the litigation imposes substantial transaction and market costs. {Green book, cited supra; Sanders book on Bendectin}. 166 It would be more trustworthy than a court-appointed panel, and even in the latter case the scientific consensus helped bring the litigation to a swifter conclusion. See, e.g., Laural L. Hooper, Joe S. Cecil & Thomas E. Willging, Assessing Causation in Breast Implant Litigation: The Role of Science Panels, 64 L. & CONTEMP. PROBS. 139 (2001). 167 Executive Orders were an important force in the growth and expansion (as well as the creation) of the NAS. See, e.g., NAS History at https://www.nasonline.org/about-nas/history/.

free of politics – capable of serving a critical sentinel function. As a result, in the US we must rely on private actors, primarily in the tort system and to a lesser extent in academia, to do this early detection, sentinel work. This delayed and incomplete approach to early warnings is expensive, the scientific burden is extraordinarily high and replete with delay, and public harm becomes a prerequisite to taking action. We have tried to pull the camera back to better understand the recurring failure of agencies to address emerging threats. Expert agencies may have been created to be on the lookout for hazards in ways that raise early warnings for the political process, but in the current institutional and constitutional system this protective role faces significant headwinds -- the basic design the of bureaucracy itself. Because of the numerous political points of entry to control the work of regulators, expert agencies are sometimes blocked in efforts to identify and assess emerging hazards in a rigorous way. In hindsight, it is perhaps not surprising that the political branches attach many invisible strings and guardrails on the ability of agencies to document and restrict hazards, particularly in settings that threaten economic harm to influential industries. But our current design of the administrative state means that going forward, tort litigation will continue to be indispensable in locating emerging hazards. That is, until we can design a better institutional approach.168 In this paper we have proposed a new institutional structure to provide an early warning service in ways that bypass existing structural impediments faced by the agencies. Without this new institutional structure – that breaks out of the current Catch 22 – we will remain largely dependent on the slow and costly litigation process to learn about emerging public risks caused by powerful economic actors.

168

In cases like asbestos, tobacco, and opioids (and a host of other hazards, see Table 1), the continuing “benefits” to those products to their producers lead to a predictable political brake of sort on agency warning initiatives and agency action.

APPENDIX 1 Table 1: Illustrative Table based primarily on Lexis-Compiled Litigation Settlements involving Tort Claims for Latent Hazards169 Tort litigation targeted legacy hazards occurring before 1970’s

Tort litigation targets hazards post 1970-80’s, despite the existence of the regulatory state [this needs to be rephrased]*

Some, albeit incomplete, regulatory response followed tort litigation

The hazard is still partly/largely unregulated

Yes

(this is the key column for our paper) 1. Environmental Contamination from Industrial Facilities/Landfills/Air Pollution [1B] Asbestos [2] Benzene [3] Chromium [3A] Electronic Waste (E-Waste) [3B] Ethylene Oxide (EtO) [4] Lead [5] Mercury [6] Nitric Oxide [6A] PCBs [6B] Perfluorooctanoic Acid (C8) [7] Radiation [9] Trichloroethylene (TCE) and 1,2,3-Trichloropropane (TCP) [10] Miscellaneous Chemicals AppD.02A Hydrofracking AppD.03 Chemical Release from Trains/Trucks AppD.05 Underground Storage Tanks (USTs) [1] Gasoline/MTBE 2. Residential Exposure to Toxins

Yes

169

Drawn from Appendix D Illustrative Awards and Settlements, 5 Toxic Torts Guide Appendix D.syn (2023), available at https://plus.lexis.com/api/permalink/d7aee1a8-0c3d-4c34-bc5b09fbcbd1e031/?context=1530671

[1] Asbestos (for AsbestosContaining Talcum Powder Cases) [2] Lead Paint [4] Pesticides/Herbicides [4A] Polychlorinated biphenyls (PCBs) [5] Spray-Polyurethane Foam Insulation [6] Miscellaneous Toxins 3. Occupational Exposure to Toxins [1] Asbestos for AsbestosContaining Diesel Exhaust Cases, see AppD.07[6A] [2] Benzene [5] Cleaning Solvents [5A] Coal Dust [6] Diacetyl [7] Ethyl Acrylate [8] Gasoline/Hydrocarbons [9] Hexavalent Chromium/Sodium Dichromate [10] Hydrogen Sulfide Gas [11] Lead [12] Manganese [12A] Mercury [13] Metal Working Fluids [16] Paint Fumes [17] Pesticides [17A] Polychlorinated biphenyls (PCBs) [18] Radiation [19] Silica [20] Wood Dust [21] Miscellaneous Chemicals 4. Prescription Drugs

[1A] Actos

Virtually all of the drug claims involve failure to warn (or fraud) claims save one; it is unclear where responsibility should be placed for these alleged lapses and it likely varies Yes

Research ongoing

[2] Bisphosphonates (Fosamax/Zometa) [3] Botox

Yes

[5] Hormone Replacement Therapy [5A] Humira/Remicade/Other TNF Blockers [5B] Methylene Blue

Yes

[5D] Pradaxa

Yes

[6A] Risperdal

Yes

[7] Sulindac

[7A] Testosterone Replacement Therapy (TRT) [9] Xarelto

Yes [Design Defect claim] Yes

Yes

Research ongoing Research ongoing Research ongoing Research ongoing Research ongoing Research ongoing Research ongoing Research ongoing Research ongoing Research ongoing

ADDITIONAL DRUGS/SUPPLEMENTS Phen-fen Vioxx Many more … 5. Consumer Products [1A] Cigarettes [1B] Vaping Devices [3] Over-the-Counter Drugs [4] Talc/AsbestosContaminated Talcum Powder [5] Toys 6. Pesticides Round-Up Research ongoing

APPENDIX 2 THREE DETAILED CASE STUDIES AT THE INTERSECTION OF LITIGATION AND REGULATION 1. Asbestos Production of asbestos cement in the United States began in 1903.170 Within two years U.S. companies were manufacturing flat asbestos-cement board and asbestos brake linings.171 After companies developed a technique for spraying asbestos for insulation in 1931, production increased dramatically.172 In 1945, companies began to manufacture multiflex material consisting of asbestos fabrics with a coating of resin and introducing it into air ducts.173 By the late 1950s, asbestos was being used in the construction industry in asbestos cement products, building insulation, and fire-proofing materials.174 In other industries, it was used valve packings, gaskets, boiler lagging and pipe covering, protective clothing, shielding materials, and as automotive brake linings.”175 The first medical reports of an industrial disease called “asbestosis” appeared in the medical literature in the late 1920s.176 By the end of 1930, it was well-accepted in the medical community that exposure to asbestos caused pulmonary asbestosis and that the incidence of the disease and the duration of the onset period was dose-related.177 By 1933, the fact that asbestosis was a fatal progressive disease was well known in the medical community.178 Most of these reports involved chrysotile asbestos imported from Canada.179 By the mid-1930s, asbestosis was scheduled as an occupational disease eligible for workers compensation benefits.180 While the

170

Douglas H. K. Lee & Irving J. Selikoff, Historical Background to the Asbestos Problem, 18 Env. Res. 300-14 (1979) at Table 1. 171 Douglas H. K. Lee & Irving J. Selikoff, Historical Background to the Asbestos Problem, 18 Env. Res. 300-14 (1979) at Table 1. 172 Douglas H. K. Lee & Irving J. Selikoff, Historical Background to the Asbestos Problem, 18 Env. Res. 300-14 (1979) at 304. 173 Douglas H. K. Lee & Irving J. Selikoff, Historical Background to the Asbestos Problem, 18 Env. Res. 300-14 (1979) at Table 1. 174 Asbestos, 19 Ind. Hygiene J. 161 (1958) 175 Asbestos, 19 Ind. Hygiene J. 161 (1958) 176 Barry I. Castleman, Asbestos: Medical and Legal Aspects (5th ed. 2005), at 5-7. See also Douglas H.K. Lee & Irving J. Selikoff, Historical Background to the Asbestos Problem, 18 Env. Research 300-14 (1979), at 308. 177 Barry I. Castleman, Asbestos: Medical and Legal Aspects (5th ed. 2005), at 14-15. 178 Barry I. Castleman, Asbestos: Medical and Legal Aspects (5th ed. 2005), at 23-25. 179 Morris Greenberg, The Art of Perpetuating a Public Health Hazard, 47 J. Occupational and Environmental Medicine 137 (2005), at 137. 180 Irving J. Selikoff, A Landmark Case in Asbestosis, 266 JAMA 898 (1991) at 900.

asbestos industry could not deny the existence of asbestosis in workers exposed to asbestos, it downplayed the virulence of asbestosis.181 Although the medical literature contained a number of reports of lung cancers associated with exposure to asbestos during the 1930s and 1940s, the industries that manufactured and used asbestos strongly resisted any government intervention.182 An article published by Groff Conklin in Scientific American in 1948 characterized asbestos as an environmental carcinogen and noted that “[t]he reluctance of industrial management to publicize the existence of a carcinogenic hazard is understandable, for it could easily become difficult for plants so labeled to recruit workers, or even to sell their commodities.”183 A paper published by the eminent English statistician Richard Doll in 1955 concluding that there was an elevated risk of lung cancer over the general population settled the matter.184 The asbestos industry’s denial that asbestos exposure could cause lung cancer suffered a huge blow with a study of insulation workers that Dr. Irving Selikoff and two co-authors presented at a symposium put on by the New York Academy of Sciences in 1964 and published by in the Annals of the New York Academy of Sciences (NYAS) in 1965.185 The careful epidemiological study of 307 consecutive deaths of insulation workers who had been exposed for at least 20 years concluded that “[l]ung cancer was found to be at least seven times as common as expected and cancer of the gastrointestinal tract three times as common as expected,” and “[t]here were ten instances of mesothelioma of the pleura or peritoneum.”186 And the bulk of these workers’ exposures were to asbestos levels at or below the 5 mppcf that the industry had deemed to be safe.187

181

Jock McCulloch & Geoffrey Tweedale, Defending the Indefensible (2009), at 51 Barry I. Castleman, Asbestos: Medical and Legal Aspects (5th ed. 2005), at 42-43; Peter Bartrip, Beyond the Factory Gates (2006) at 64; Philip E. Enterline, Changing Attitudes and Opinions Regarding Asbestos and Cancer, 1934-1965, 20 Am. J. Ind. Med. 685 (1991) at 691; Irving J. Selikoff & Douglas H. K. Lee, Asbestos and Disease (1978) at 26 183 Groff Conklin, Cancer and the Environment, 180 Scientific American 11 (January, 1949), at 14 184 R. Doll, Mortality from Lung Cancer in Asbestos Workers, 12 British J. Industrial Medicine 81-86 (1955); David S. Egilman, Public Health and Epistemology, 22 Am. J. Ind. Med. 457 (1992) at 457 185 I. J. Selikoff, J. Churg & E. C. Hammond, The Occurrence of Asbestosis Among Insulation Workers in the United States, 132 Annals N.Y. Acad. Sci. 9 (1965). See Morris Greenberg, Biological Effects of Asbestos: New York Academy of Sciences 1964, 43 Am. J. Ind. Med. (2003) at 543 (“The sounding of the international alarm that was to initiate the slow retreat from the prodigious use of asbestos, can be dated from a conference hosted by the New York Academy of Sciences in 1964, and in particular its publicizing of the previously reported causal association between asbestos exposure and lung cancer and malignant mesotheliomas.”) 186 I. J. Selikoff, J. Churg & E. C. Hammond, The Occurrence of Asbestosis Among Insulation Workers in the United States, 132 Annals N.Y. Acad. Sci. 9 (1965) at 152. 187 I. J. Selikoff, J. Churg & E. C. Hammond, The Occurrence of Asbestosis Among Insulation Workers in the United States, 132 Annals N.Y. Acad. Sci. 9 (1965) at 142. 182

The Industry Prepares for Regulation. Soon after the Environmental Protection Agency and the Occupational Safety and Health Administration were created in 1970, the asbestos industry created the Asbestos Information Association of North America to “deal with the media, work with the government to regulate workplace conditions, and eventually set up a legal arm to assist the whole industry in defending itself against liability.”188 Matthew M. Swetonic, the Manager of Environmental Affairs at Johns-Manville Corporation, wrote the original prospectus for AIA/NA and became its first Executive Secretary.189 According to Swetonic, the “deciding factor” in inducing the companies to join the association “was our warning that the two new regulatory agencies created by Congress that year – OSHA and EPA – were both eyeing asbestos as their first problem to tackle.”190 The original concept for AIA was that it “would limit its activities to providing accurate, unbiased information on asbestos and health to the press, to the public and to interested politicians and government officials.”191 It soon broadened its focus, however, to “assume whatever activities and responsibilities it deem[ed] necessary to protect the interests of the asbestos manufacturing industry in the United States vis-à-vis asbestos-health.”192 The bulk of its attention was devoted to public relations and government relations.193 It created “a regulatory information arm to work with government agencies on the development of reasonable workplace and environmental standards” and “a legal arm to assist the industry as a whole in defending itself against liability claims.”194 To assist in its efforts, AIA/NA retained Hill & Knowlton, the largest public relations agency in the world at the time.195 Hill & Knowlton trained industry executives and physicians to argue that “disease associated with asbestos is rare, the public is not at risk, asbestos saves lives, and

188

Jock McCulloch & Geoffrey Tweedale, Defending the Indefensible (2009), at 99. About the Speaker, Attachment to Asbestos Textile Institute, Minutes of General Meeting, June 7, 1973 190 Matthew M. Swetonic, Death of the Asbestos Industry, In Crisis Response (J. Gottschalk, Ed. 289-308 1993) at 300. 191 Matthew M Swetonic, Why Asbestos: A Presentation Before the Asbestos Textile Institute, June 7, 1973, at 2-6, 8, 10; Jock McCulloch & Geoffrey Tweedale, Defending the Indefensible (2009), at 110. 192 Matthew M Swetonic, Why Asbestos: A Presentation Before the Asbestos Textile Institute, June 7, 1973, at 2-3. 193 Matthew M Swetonic, Why Asbestos: A Presentation Before the Asbestos Textile Institute, June 7, 1973, at 2-6, 8, 10; Jock McCulloch & Geoffrey Tweedale, Defending the Indefensible (2009), at 110. 194 Matthew M. Swetonic, Death of the Asbestos Industry, In Crisis Response (J. Gottschalk, Ed. 289-308 1993) at 308. 195 Prospectus, Asbestos Information Association/North America, attachment to Letter from Eight Corporate Vice Presidents to Raymond J. Sachs, dated September 1, 1971, at 6. 189

proper controls had been introduced.”196 Reports of cancer in asbestos workers were explained away as a remnant of the distant past when the industry was reasonably unaware of the risks.197 Speaking at a meeting of the Asbestos Textile Institute in June 1973, Swetonic said he had “no doubts that the insulation workers are in fact dying as rapidly as Dr. Selikoff says they are.”198 He urged the audience not to forget that “5,000 of our workmen are still going to die of asbestosrelated disease and that, all things considered, there is absolutely nothing that we can do to prevent it.”199 It was, in fact, the job of AIA/NA to prevent the government from attempting to prevent those deaths. Because of space limitations, the following description of federal agency attempts to regulate asbestos in light of its known potential to kill people will be limited to EPA’s regulatory efforts over the years.

EPA Regulation of Asbestos. The Asbestos NESHAP. Asbestos was at the top of the newly created EPA’s priority list when Congress enacted the Clean Air Act in 1970. In December 1970, EPA initiated a series of public hearings in anticipation of writing the very first national emissions standard for hazardous air pollutants (NESHAP) to control emissions of asbestos.200 For decades, construction companies had sprayed asbestos as inexpensive insulation material in buildings like the World Trade Center. Emissions from building construction and demolition activities could affect both workers and people located near those operations.201 The asbestos industry bitterly opposed EPA’s efforts to establish a NESHAP for asbestos. A raft of telegrams to EPA Administrator Bill Ruckelshaus orchestrated by the AIA/NA stressed that there was “no evidence that anyone in the general public has ever contracted disease from exposure to ‘slight amounts’ of asbestos.”202 The AIA/NA played a major role in the “game 196

Jock McCulloch & Geoffrey Tweedale, Defending the Indefensible (2009), at 110 Jock McCulloch & Geoffrey Tweedale, Defending the Indefensible (2009), at 111 (quoting JM 1978 annual report). 198 Matthew M Swetonic, Why Asbestos: A Presentation Before the Asbestos Textile Institute, June 7, 1973, at 9. 199 Matthew M Swetonic, Why Asbestos: A Presentation Before the Asbestos Textile Institute, June 7, 1973, at 10. 200 Matthew M. Swetonic, Death of the Asbestos Industry, In Crisis Response (J. Gottschalk, Ed. 289-308 1993) at 303. 201 Jock McCulloch & Geoffrey Tweedale, Defending the Indefensible (2009), at 203 202 Telegram from R. Power Fraser, Vice President to William Ruckelshaus, dated April 13, 1971. See Memorandum from William P. Raines to Board of Directors, AIA/NA, dated April 6, 1971 (orchestrated campaign). 197

plan” that the industry devised for the three industry representatives to the advisory committee that EPA assembled to offer suggestions for the content of the standard.203 Representatives of AIA/NA member companies also met privately with EPA officials to urge their view of the issues on the agency.204 Having learned of the contents of the EPA proposal before it became public, the industry was generally happy with EPA’s approach,205 which provided for easily met control techniques, rather than specifying numerical emissions standards.206 The companies agreed that they would not oppose an outright ban on spraying asbestos on buildings for fireproofing.207 The industry, however, had a “major problem” with the proposed requirement for stabilizing, wetting or otherwise treating asbestos mill tailings dumps to prevent dust emissions.208 The AIA/NA prepared a 9-page document commenting on the proposal and offering suggestions for changing it prior to the August 11 meeting.209 At the oral hearings that EPA held on the proposal, AIA/NA representative F.A. Zimmerman testified that “the general public is not in danger from the minute amounts of asbestos fiber present in the ambient air,” but “it is prudent and feasible to establish reasonable emission standards and controls that will protect the public and ensure that hazardous levels of asbestos will not develop in the community air in the future.”210 The industry therefore supported the proposal with the exception of the requirements for controlling emissions from tailings piles.211 The final regulations that EPA issued on April 6, 1973 did not ban the use of asbestos,212 but they did effectively ban spray-on asbestos materials containing more than one percent asbestos and 203

Memorandum from W.P. Raines to AIA/NA Board of Directors and AIA/NA Administrative SubCommittee, dated August 9, 1971. 204 Prospectus, Asbestos Information Association/North America, attachment to Letter from Eight Corporate Vice Presidents to Raymond J. Sachs, dated September 1, 1971. 205 Memorandum from W.P. Raines to AIA/NA Board of Directors and AIA/NA Administrative SubCommittee, dated August 9, 1971, at 1-2. 206 Memorandum from Charles F. Bien to R.P. Fraser, re: Special meeting of Asbestos Information Association/North America Board of Directors, dated August 2, 1971 207 Memorandum from W.H. Page to Numerous Recipients re: Report on Meeting of A.I.A. Subcommittee on E.P.A. Proposed Regulations on Emission of Asbestos, December 22, 1971 208 Memorandum from W.P. Raines to AIA/NA Board of Directors and AIA/NA Administrative SubCommittee, dated August 9, 1971. 209 Asbestos Information Association/North America, Proposed National Emission Standards for Asbestos, Comments and Suggestions, August 11, 1971. 210 PF.H. Zimmerman, Statement of F.H. Zimmerman on Behalf of the Asbestos Information Association/North America, EPA Hearings on Proposed Emission Control Standards for Asbestos – Los Angeles, California, February 16, 1972, at 1. 211 PF.H. Zimmerman, Statement of F.H. Zimmerman on Behalf of the Asbestos Information Association/North America, EPA Hearings on Proposed Emission Control Standards for Asbestos – Los Angeles, California, February 16, 1972, at 1. 212 Environmental Protection Agency, National Emission Standards for Hazardous Air Pollutants: Asbestos, Beryllium, and Mercury, 38 Fed. Reg. 8820 (1973), at 8820.

surfacing roadways with asbestos.213 The agency agreed with the industry that it was “not practicable, at this time, to establish allowable numerical concentrations or mass emission limits for asbestos.”214 The regulations prohibited “visible emissions” in mills, but they excluded dumps and open storage operations.215 They also prohibited visible emissions from the nine manufacturing operations that were found to be “major sources” of asbestos.216 Asbestos mines were prohibited from exposing miners to levels in excess of those recommended by the American Conference of Governmental Industrial Hygienists (ACGIH).217 Johns-Manville’s Director of Environmental Affairs, Matthew Swetonic proudly announced to the Asbestos Textile Institute that there was not a single major industry recommendation made to the EPA at the public hearings on the proposed asbestos emission regulations that was not accepted either in toto or in principle in the final standards.”218 On October 14, 1975, EPA amended the asbestos regulations to add two more industries to the “manufacturing facilities” subject to the regulations and to prohibit attempts to circumvent the no visible emissions requirements by diluting emissions with additional air, and to specify extensive work practice requirements for demolition and disposal of asbestos.219 TSCA Regulation. Introduction. Under the original Toxic Substances Control Act of 1976 (TSCA), when EPA had a reasonable basis to conclude that exposure to a toxic substance would present an “unreasonable risk” of injury to health or the environment, the agency was required apply the least burdensome of eight listed regulatory options, one of which was a requirement prohibiting the use of the substances in commerce.220 Recognizing that asbestos was a prime candidate for regulation under the new statute, EPA devoted considerable staff resources to gathering and analyzing scientific data on 213

Environmental Protection Agency, National Emission Standards for Hazardous Air Pollutants: Asbestos, Beryllium, and Mercury, 38 Fed. Reg. 8820 (1973), at 8821-22. 214 Environmental Protection Agency, National Emission Standards for Hazardous Air Pollutants: Asbestos, Beryllium, and Mercury, 38 Fed. Reg. 8820 (1973), at 8820. 215 Environmental Protection Agency, National Emission Standards for Hazardous Air Pollutants: Asbestos, Beryllium, and Mercury, 38 Fed. Reg. 8820 (1973), at 8821. 216 Environmental Protection Agency, National Emission Standards for Hazardous Air Pollutants: Asbestos, Beryllium, and Mercury, 38 Fed. Reg. 8820 (1973), at 8821. 217 Environmental Protection Agency, National Emission Standards for Hazardous Air Pollutants: Asbestos, Beryllium, and Mercury, 38 Fed. Reg. 8820 (1973), at 8821. 218 Matthew M. Swetonic, Death of the Asbestos Industry, In Crisis Response (J. Gottschalk, Ed. 289-308 1993) at 307 219 Environmental Protection Agency, National Emission Standards for Hazardous Air Pollutants: Amendments to Standards for Asbestos and Mercury, 40 Fed. Reg. 48,292 (1975). 220 Toxic Substances Control Act of 1976, Pub. L. 94-469 (October 11, 1976) § 6(a).

the risks posed by asbestos and various substitutes in a number of exposure contexts.221 As part of that data-gathering process, EPA issued an advance notice of proposed rulemaking toward the end of the Carter Administration in 1979.222 The 1980 election, however, brought in an administration that was committed to deregulation and not at all enthusiastic about imposing any limits on marketing potentially dangerous products. Upper-level political appointees in EPA during the first few years of the Reagan administration did nothing to advance a regulatory agenda while they urged various asbestosusing industries to adopt voluntary reductions on exposure to asbestos.223 As regulatory action at the federal level stalled, plaintiffs were receiving large damage verdicts from juries in courtrooms across the country.224 The landmark case of Borel v. Fibreboard Paper Products Corp., in which the Fifth Circuit Court of Appeals held in September 1973 that plaintiffs suffering injuries due to exposure to asbestos in the workplace and elsewhere could file products liability actions against asbestos manufacturers under section 402A of the Restatement (Second) of Torts, “triggered the greatest avalanche of toxic-tort litigation in the history of American jurisprudence” at the time.225 Discovery in these lawsuits produced millions of documents many of which documented the efforts of asbestos companies to fund research concluding that asbestos presented no risk to workers and their families, to attack scientists whose research suggested otherwise, and generally to manufacture uncertainty about the science about asbestos and human health. After losing a large number of these lawsuits, the nation’s largest asbestos producer, Johns Manville, filed for bankruptcy protection in 1982.226 As the Reagan Administration dallied, the National Research Council of the National Academies of Sciences (NAS) issued a report on non-occupational risks posed by asbestos.227 The NAS report, which came out in 1984, concluded that “[n]on-occupational exposure to asbestiform fibres in air present[ed] a risk to human health,” but “[t]he extent of this risk [was] highly uncertain, depending on the nature and amount of exposure and other factors.228 The NAS panel disagreed with the industry position that the commonly used chrysotile form of asbestos was not 221

Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 551. 222 Environmental Protection Agency, Commercial and Industrial Use of Asbestos Fibers and Consumer Products Containing Asbestos: Statement of Policy on Coordination of Regulatory Activities, 44 Fed. Reg. 60056 (October 17, 1979). 223 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 552. 224 See, e.g., Borel v. Fibreboard Paper Products Corp., 493 F.2d 1076 (5th Cir. 1973). 225 Paul Brodeur, Outrageous Misconduct (1985( 226 Johns Manville, JM Historical Timeline, https://www.jm.com/en/ourcompany/HistoryandHeritage/company-history/ 227 National Academies of Sciences/National Research Counsel, Asbestiform fibres: Non Occupational Health Risks (1984) 228 National Academies of Sciences/National Research Counsel, Asbestiform fibres: Non Occupational Health Risks (1984), at 4.

likely to be carcinogenic when it concluded that ““[a]ll major commercial types of asbestos fibres used in the United States have been associated with lung cancer, mesothelioma, and asbestosis in humans.”229 The report concluded that “[s]ome deaths from mesothelioma and lung cancer will probably result from current and past levels of exposure to asbestos in ambient air.”230 The risks non-workers made asbestos a prime candidate for regulation under TSCA. The 1984 Proposal. After Bill Ruckelshaus replaced Anne Gorsuch as EPA Administrator in 1983 following a series of scandals at the agency, EPA was more receptive to the idea of regulating asbestos under TSCA. In September 1984, the Natural Resources Defense Council (NRDC) filed a petition under section 21 of TSCA asking EPA to prohibit the further uses of asbestos in automobile and truck brakes, which clearly posed risks to the general public.231 The agency granted the petition in December 1984 based upon its conclusion that “the use of asbestos in brakes does present risks to human health,” because it was “a demonstrated human carcinogen that causes lung cancer and mesothelioma . . . as well as other lung disorders.”232 The agency therefore initiated a proceeding “to gather information on the desirability and feasibility of reducing risks associated with the use of asbestos in brakes.”233 With surprising speed, the agency staff drafted two regulations that went far beyond regulating asbestos in brakes in late 1984, but the agency ran into an impenetrable roadblock in the Office of Management and Budget’s Office of Information and Regulatory Affairs (OIRA). EPA spent almost two years in a great debate with the Office of Information and Regulatory Affairs over whether EPA should defer asbestos regulation to the Occupational Safety and Health Administration, which had been establishing limits on asbestos fibers in the air at levels that OSHA deemed economically and technologically feasible for the regulated companies.234 OIRA economists also insisted that EPA prepare a formal regulatory analysis to OIRA’s satisfaction

229

National Academies of Sciences/National Research Counsel, Asbestiform fibres: Non Occupational Health Risks (1984), at 4. 230 National Academies of Sciences/National Research Counsel, Asbestiform fibres: Non Occupational Health Risks (1984), at 3. 231 Environmental Protection Agency, Asbestos; Response to Citizen’s Petition, 49 Fed. Reg. 49,311 (December 19, 1984), at 49,311. 232 Environmental Protection Agency, Asbestos; Response to Citizen’s Petition, 49 Fed. Reg. 49,311 (December 19, 1984), at 49,311. 233 Environmental Protection Agency, Asbestos; Response to Citizen’s Petition, 49 Fed. Reg. 49,311 (December 19, 1984), at 49,311. 234 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 554; Phillip Shabecoff, "E.P.A. Pulls Back on Asbestos Rules," New York Times, 9 Mar. 1985, 7, col. 5; Robert S. Taylor, "EPA's Plan to Drop Asbestos Restriction is Reconsidered in Response to Criticism," Wall Street Journal, 11 Mar. 1985, 8, col. 3.

quantifying and monetizing the costs and benefits of EPA’s preferred option and other statutory options.235 AIA/NA and the government of Canada, the location of the world’s largest asbestos mine, had many meetings with EPA and OIRA officials in an all-out effort to send asbestos regulation to OSHA and to convince them that the risks posed by asbestos to non-workers were trivial.236 Over many months, OIRA demanded that EPA base its decision on formal cost-benefit analysis as a condition to allowing the rulemaking process to continue.237 After Congressman John Dingell, the powerful chairman of the House Energy and Commerce Committee, intervened with an infamous “Dingellgram” demanding documentation of OIRA’s interactions with EPA and the industry over asbestos, OIRA finally relented.238 EPA issued a notice of proposed rulemaking on January 29, 1986 that went far beyond asbestos use in brakes to propose a complete ban on that use and most other uses of asbestos.239 The Gold Book. With the rulemaking process underway, EPA decided to address the issue of automobile brakes with a guidance document. While some U.S. auto manufacturers had by then ceased using asbestos in their brakes, millions of older cars and replacement brakes still contained the material. In addition, asbestos materials for brakes and replacement brakes were imported into the United States.240 During repair and replacement of brakes containing asbestos, auto mechanics and shop workers were at risk for heavy exposure to the material.241 The same was true for people who elected to replace their own brakes. While OSHA had promulgated regulations setting forth work practice requirements to protect workers during auto repair, OSHA inspections were rare and testing in auto shops for asbestos fibers was nearly nonexistent.242 And EPA’s NESHAP for asbestos did not cover the interiors of garages and repair shops. 235

Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 560-67. 236 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 557. 237 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 562-65. 238 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 565-67. 239 Environmental Protection Agency, Asbestos: Proposed Mining and Import Restrictions and Proposed Manufacturing, Importation, and Processing Prohibitions, 51 Fed. Reg. 3738 (January 29, 1986). 240 See Andrew Schneider, U.S. Imports of Asbestos Break Material Are on the Rise, St. Louis PostDispatch, at A-14 (October 26, 2003); Andrew Schneider, EPA Warning on Asbestos is Under Attack; Lawyers Target Materials Detailing Dangers in Brakes, St. Louis Post-Dispatch, at A1 (October 26, 2003). 241 David Michaels & Celeste Monforton, How Litigation Shapes the Scientific Literature: Asbestos and Disease Among Automobile Mechanics, 5 Brooklyn J. L. & Policy 1137 (2007), at 1149-52. 242 See Andrew Schneider, EPA Warning on Asbestos is Under Attack; Lawyers Target Materials Detailing Dangers in Brakes, St. Louis Post-Dispatch, at A1 (October 26, 2003).

In June 1986 EPA issued a 16-page booklet for use by automobile garages to protect workers from asbestos-related diseases called the “Gold Book.”243 The booklet described in laymen’s terms the health effects of asbestos and the large exposures that people changing brakes and clutches would encounter.244 The booklet detailed the possible adverse consequences of exposure to asbestos in the absence of adequate prevention, including asbestosis, mesothelioma, lung cancer, and other cancers, all of which had long latency periods.245 It then suggested how exposures could be controlled in brake cleaning, clutch repair, filter change and waste disposal.246 And it warned that if workers planned to wear respirators for protection, they had to be worn all day.247 People involved in brake and clutch replacement relied on the Gold Book for more than 15 years when attorneys for auto mechanics and shop workers began to cite it for the proposition that EPA believed that asbestos caused lung cancer and mesothelioma. Worried that juries would trust EPA, rather than their expert witnesses, defendants attempted, often unsuccessfully, to exclude the Gold Book from the evidence used at trials.248 On August 19, 2003, Dino Privitera of Morgan, Lewis & Bokius filed a petition with EPA under the Information Quality Act demanding that it cease disseminating the Gold Book.249 The law firm did not disclose who was paying for the petition. But the firm did represent the insurance industry and many corporate defendants in asbestos litigation.250 The information Quality Act of 2001 required the Office of Management and Budget to write “policy and procedural guidance to Federal agencies for ensuring and maximizing the quality, objectivity, utility, and integrity of information . . . disseminated by Federal agencies.” It then 243

Environmental Protection Agency, Guidance for Preventing Asbestos Disease Among Auto Mechanics (1986). 244 Environmental Protection Agency, Guidance for Preventing Asbestos Disease Among Auto Mechanics (1986), at 1-5. 245 Environmental Protection Agency, Guidance for Preventing Asbestos Disease Among Auto Mechanics (1986), at 1-3. 246 Environmental Protection Agency, Guidance for Preventing Asbestos Disease Among Auto Mechanics (1986), at 6-7. 247 Environmental Protection Agency, Guidance for Preventing Asbestos Disease Among Auto Mechanics (1986), at 9. 248 See, e.g., Memorandum of Law in Support of Defendant, Ford Motor Company’s Motion in Limine to Preclude the Environmental Protection Agency’s “Guidance for Preventing Asbestos Disease Among Auto Mechanics (The “Gold Book”), Jeanne C. Locker v. Ford Motor Co., Civ. No. 07C-11-014 ASB, Superior Court for New Castle County, Delaware, July 13, 2009; Memorandum of Law in Support of Defendant, General Motors Corporation’s Motion in Limine to Preclude the Environmental Protection Agency’s “Guidance for Preventing Asbestos Disease Among Auto Mechanics (The “Gold Book”), Civ. Nos. 08C-02-177, et al., Superior Court of Delaware, April 3, 2009. 249 Letter from Dino Privitera to Information Quality Act Staff, dated August 19, 2003. 250 Barry I. Castleman, Asbestos: Medical and Legal Aspects (5th ed. 2005), at 455.

required federal agencies to prepare their own guidelines and establish procedures under which affected persons could “seek and obtain correction of information . . . that does not comply with the guidelines.”251 The statute was quietly inserted into an appropriations bill, between a property acquisition appropriation for the Gerald R. Ford Museum and a provision relating to the nonforeign area cost-of-living allowances, at the behest of a tobacco industry consultant without any hearings or even any mention of its existence.252 The law firm challenged the guidance document on grounds that the statements made in the Gold Book “were based on inadequate and inappropriate scientific data and literature” and have been used to convey misperceptions about the hazards of brake mechanic work and the risks of contracting asbestos-related illnesses.253 The firm requested that EPA “discontinue disseminating the Gold Book and post a caveat on EPA’s website to the effect that the 1986 Gold Book is no longer current from a scientific perspective,” or in the alternative, to update the guidance document.254 EPA pondered the petition for three years before writing to Mr. Privitera to tell him that EPA was releasing a draft brochure entitled “Current Best Practices for Preventing Asbestos Exposure Among Brake and Clutch Repair Workers” that would “supersede” the Gold Book.255 The new version of Gold Book consisted of two pages of do’s and don’ts for mechanics and home car repair mechanics.256 The new brochure did mention the fact that asbestos “can cause serious health problems when inhaled into the lungs,” because “[f]ibers embedded in the lung tissue overtime may result in lung diseases such as asbestosis, lung cancer, or mesothelioma.”257 Following EPA’s replacement of the Gold book with the two-page summary of recommendations, companies cited that substitution as proof that EPA recognized that “the scientific evidence was at odds with statements made in the Gold Book.”258

251

Information Quality Act of 2000, Section 515 of the FY 2001 Appropriations Act, P.L. 106-554, 114 Stat. 2763A-153-154 (December 21, 2000). 252 Thomas O. McGarity, et. al., Truth and Science Betrayed: The Case Against the Information Quality Act (Center for Progressive Regulation, February 2005), at 2-3. 253 Letter from Dino Privitera, Morgan, Lewis & Bockius, LLP to Information Quality Guidelines Staff, U.S. EPA, August 19, 2003 at 1-2, 254 Letter from Dino Privitera, Morgan, Lewis & Bockius, LLP to Information Quality Guidelines Staff, U.S. EPA, August 19, 2003 at 2. 255 Letter from James B. Gulliford, Assistant Administrator for Environmental Information, to Dino Privitera, dated August 17, 2006. 256 Environmental Protection Agency, Current Best Practices for Preventing Asbestos Exposure Among Brake and Clutch Repair Workers (March 2007). 257 Environmental Protection Agency, Current Best Practices for Preventing Asbestos Exposure Among Brake and Clutch Repair Workers (March 2007). 258 Memorandum of Law in Support of Defendant, Ford Motor Company’s Motion in Limine to Preclude the Environmental Protection Agency’s “Guidance for Preventing Asbestos Disease Among Auto Mechanics (The “Gold Book”), Jeanne C. Locker v. Ford Motor Co., Civ. No. 07C-11-014 ASB, Superior Court for New Castle County, Delaware, July 13, 2009, at 13.

Hearings on the Proposed Ban. The 1976 TSCA departed from the Administrative Procedure Act’s notice-and-comment rulemaking procedures in that it provided for an “informal” in-person hearing at which any person could present relevant evidence. In addition, if the presiding officer concluded that there were “disputed issues of material fact,” participants were entitled to “cross examine” witnesses to assist in resolving those issues.259 Consequently, EPA held lengthy hearings on the 1986 proposal in which “25 groups gave oral testimony and EPA officials made themselves available for nine days of cross examination.”260 The hearings and EPA’s efforts to assimilate the written submissions and oral testimony consumed nearly three years.261 The 1989 EPA Asbestos Ban. On July 12, 1989, EPA published a final rule phasing out, at staged intervals, the future manufacture, importation, processing and distribution of almost all products containing asbestos.262 During the 8-year phase-out period, products containing asbestos would have to be labeled.263 The final regulation also provided for petitioning EPA for exemptions.264 In support of this action, EPA found that all forms of asbestos, including chrysotile, were capable of causing lung cancer and mesothelioma and that lung cancers had been documented in a wide variety of situations in which people had been exposed to asbestos.265 The agency found the industry’s argument that chrysotile was far less carcinogenic than other types of asbestos to be “inconclusive and inconsistent.”266 In fact, some of the highest risks in the epidemiological studies were attributable to exposure to chrysotile, suggesting that “chrysotile exposures can confer an extremely high risk of lung cancer.”267 In any event, EPA believed that it was “prudent and in the public interest to consider all fiber types as having comparable carcinogenic potency

259

Toxic Substances Control Act of 1976 § 6(c)(2). Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 556. 261 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 556. 262 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989) 263 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,461. 264 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,462. 265 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,469. 266 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,470. 267 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,470. 260

in its quantitative assessment of mesothelioma risk.”268 In reaching these conclusions, the agency undertook an extensive analysis of the benefits and costs of banning asbestos,269 and it examined, albeit in a somewhat more cursory fashion, the costs and benefits of other alternatives, including the "do nothing" option.270

Based on its analysis of the costs and benefits of prohibition and various alternatives, EPA concluded that the continued manufacture, importation, processing, and distribution in commerce of most asbestos-containing products poses an unreasonable risk to human health.271 In addition, the prohibitions in the regulation were, in the agency’s judgment, “the least burdensome means of controlling the exposure risks posed throughout the life cycle of asbestos-containing products.”272 At the same time, the $458.89 million cost of the rule would be spread out over millions of people and over a 13-year period so that the cost to any individual would be negligible.273

The 1991 Fifth Circuit Remand. Several asbestos companies challenged the asbestos ban in the Fifth Circuit Court of Appeals.274 The court held that “EPA has presented insufficient evidence to justify its asbestos ban.”275 It based this conclusion upon two grounds: “the failure of the EPA to consider all necessary evidence and its failure to give adequate weight to statutory language requiring it to promulgate the least burdensome, reasonable regulation required to protect the environment adequately.”276 The court found that EPA had not adequately considered less burdensome options short of doing nothing. The court noted that "[w]hile the EPA may have shown that a world with a complete ban of asbestos might be preferable to one in which there is only the current amount of 268

Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,470.s 269 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,468-87. 270 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,487-89. 271 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,467. 272 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,467. 273 Environmental Protection Agency, Asbestos; Manufacture, Importation, Processing and Distribution in Commerce Prohibitions, 54 Fed. Reg. 29,460 (1989), at 29,468. 274 Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991) 275 Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1215. 276 Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1215.

regulation, the EPA has failed to show that there is not some intermediate state of regulation that would be superior to both the currently-regulated and the completely-banned world."277 The court interpreted the statute bizarrely to say that “EPA’s regulation cannot stand if there is any other regulation that would achieve an acceptable level of risk.”278 The court then explained how EPA could make this showing: Upon an initial showing of product danger, the proper course for the EPA to follow is to consider each regulatory option, beginning with the least burdensome, and the costs and benefits of regulations under each option. The EPA cannot simply skip several rungs, as it did in this case, for in doing so, it may skip a less-burdensome alternative mandated by TSCA.279 The court went on to prescribe in detail the nature of the cost-benefit analysis that the agency was required to employ. Without even a nod to the extensive economic and philosophical literature on the propriety of discounting future health benefits to present value, the court held that EPA was obliged to make the attempt.280 The court referred the agency to an article in The Economist that, according to the court, explained the use of discount rates for non-monetary goods.281 The court also criticized the agency for projecting benefits too far out into the future, for relying too heavily upon the concept of population exposure, and for its evaluation of the availability of substitutes for asbestos.282 To the astonishment of many, the Department of Justice declined to ask the Supreme Court to review the Fifth Circuit’s extremely aggressive substitution of its interpretation of the statute and its view of the proper way to go about regulatory analysis for EPA’s approach.283 EPA’s Response to Corrosion Proof Fittings. The Corrosion Proof Fittings defeat was deeply demoralizing to the agency’s toxic substances staff. It had spent a decade assembling a solid record of scientific support for banning asbestos, a universally acknowledged killer, all to no avail because an overly aggressive court had interpreted the statute to require the agency to analyze the costs and benefits of eight regulatory approaches and select the least burdensome of those approaches. But the agency leadership toward the end of the George H.W. Bush administration was not inclined to launch another major regulatory initiative under section 6 of TSCA. Instead, the agency returned to the same kind of

277

Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1217. Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1216. 279 Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1217. 280 Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1218. 281 Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1218. 282 Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th Cir. 1991), at 1218-19. 283 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 583. 278

voluntary initiatives that it undertook at the first of the Reagan administration.284 And it appeared that many companies were, in fact, willing to move away from asbestos so long as reasonably expensive substitutes were available.285 The end result was that EPA did not attempt to ban another toxic substance under section 6 until after the statute was amended 25 years later.286 The Frank R. Lautenberg Chemical Safety for the 21st Century Act. On June 22, 2016, President Obama signed the Frank R. Lautenberg Chemical Safety for the 21st Century Act, which contained comprehensive amendments to the Toxic Substances Control Act. In a grand compromise between the chemical industry and some environmental groups, the new law allowed EPA regulation under section 6 to preempt more stringent state action while amending section 6 in several ways designed to facilitate EPA regulatory activity in an effort to breathe new life into the moribund statute. The threshold for section 6 regulations remains “unreasonable risk of injury to health or the environment,” but new language specifies that EPA is to make the unreasonable risk finding “without consideration of costs or other nonrisk factors.”287 When EPA determines that the manufacture, processing, distribution in commerce, use, or disposal of a chemical substance or mixture presents an unreasonable risk, it must choose among the same listed options, which included a complete ban, but it no longer is required to select the “least burdensome” action.288 And the agency may conduct a “paper hearing” under section 553 of the APA without allowing an opportunity for oral presentation and cross-examination.289 In promulgating a requirement, EPA must consider “the effects of the chemical substance or mixture on health and the magnitude of the exposure of human beings to the” substance or mixture, the “benefits of the chemical substance or mixture for various uses,” and the “reasonably ascertainable, economic consequences of the rule,” including “the likely effect of the rule on the national economy,” the “costs and benefits” of the regulation and one or more primary alternative regulatory actions, and “the cost effectiveness” of the rule and one or more primary alternatives.290 The amended statute then preempts any state regulation of a chemical or 284

Barry I. Castleman, Asbestos: Medical and Legal Aspects (5th ed. 2005), at 510. For example, General Motors GM wrote to EPA that it expected that all GM vehicles would be asbestos-free within the time frame contemplated by the 1989 Final Rule. Barry I. Castleman, Asbestos: Medical and Legal Aspects (5th ed. 2005), at 510-11. 285 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 586. 286 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 534. 287 15 U.S.C. §§ 2605(a), 2605(b)(4). 288 15 U.S.C. §§ 2605(a)(1)-(7); Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 545. 289 15 U.S.C. § 2605(c)(3). 290 15 U.S.C. §§ 2605(c)(2).

mixture for which EPA has concluded that the chemical or mixture does not present an unreasonable risk.291 EPA Regulation of Asbestos Under the Lautenberg Amendments. The timing of the enactment of the Lautenberg amendments was unfortunate. Before EPA could begin implementing the new law, the 2016 elections put Donald Trump in the White House, and he appointed devoted deregulators to fill the political positions at the top of EPA.292 Given that asbestos was at the top of any reasonable list of candidates for regulation under the new regime, the Trump EPA conducted a “risk evaluation” for asbestos. With some artful manipulation of the data, however, it concluded that 16 of the 32 asbestos uses did not pose any unreasonable to human health.293 The Trump administration, however, did not base any regulatory requirements on that risk evaluation. In April 2022, the Biden administration reversed course and proposed to ban the ongoing use of chrysotile asbestos, the only form of asbestos known to be imported, processed or distributed within the United States at the time.294 The Biden OIRA did not get in the way, and the agency expeditiously analyzed and responded to the written public comments in the rulemaking docket. The resulting final asbestos regulation was the first rule to be finalized under the 2016 Lautenberg amendments.295 Forty years after it first proposed to ban asbestos in brake linings, EPA finally banned nearly all uses and imports of chrysotile asbestos in March 2024.296 The regulation prohibits the import or manufacture of any chrysotile asbestos or chrysotile asbestos-containing products for asbestos diaphragms used in the chlor-alkali industry.297 The ban extends to the processing, distribution and commercial use of such products after a 5 year phase-out period.298 A similar ban with a two-year phase-out applies to the manufacture, import, processing, distribution and commercial use of chrysotile asbestos used for sheet gaskets in chemical production. Another ban with a 291

15 U.S.C. § 2617(a)(1)(B). Thomas O. McGarity, Demolition Agenda (2022) 293 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 594-95. 294 Environmental Protection Agency, Asbestos Part 1; Chrysotile Asbestos, 87 Fed. Reg. 21,706 (April 12, 2022); Environmental Protection Agency, Press Release, Biden-Harris Administration Finalizes Ban on Ongoing Uses of Asbestos to Protect People from Cancer, March 18, 2024. 295 Environmental Protection Agency, Press Release, Biden-Harris Administration Finalizes Ban on Ongoing Uses of Asbestos to Protect People from Cancer, March 18, 2024. 296 Environmental Protection Agency, Part 1; Chrysotile Asbestos; Regulation of Certain Conditions of Use Under the Toxic Substances Control Act (TSCA), 89 Fed. Reg. 21,970 (March 28, 2024). 297 Environmental Protection Agency, Part 1; Chrysotile Asbestos; Regulation of Certain Conditions of Use Under the Toxic Substances Control Act (TSCA), 89 Fed. Reg. 21,970 (March 28, 2024), at 21,985. 298 Environmental Protection Agency, Part 1; Chrysotile Asbestos; Regulation of Certain Conditions of Use Under the Toxic Substances Control Act (TSCA), 89 Fed. Reg. 21,970 (March 28, 2024), at 21,985. 292

shorter 180-day phase-out applies to the manufacture, import, processing, distribution and commercial use of chrysotile asbestos-containing products used for oilfield brake blocks, aftermarket automotive brakes and linings, other vehicle friction products, and various gaskets. The ban extends to the manufacture, import, processing, and distribution of consumer-use products, but it does not extend to use by consumers of those products.299 While this action represents a major step forward toward eliminating future damage caused by asbestos products, it does nothing to address risks to human health posed by legacy uses, the subject of a second EPA rulemaking under the Lautenberg amendments.300 By one estimate, asbestos still results in more than 40,000 deaths in the United States every year.301 Multiplied by the 35 years since the original ban would have gone into effect, that represents 1,400,000 lives that might have been saved had EPA not been shackled by a statute that, as interpreted by an unsympathetic court, imposed so many procedural and analytical requirements that it was virtually designed to fail. 2. MTBE EPA’s regulation of the gasoline additive MTBE represents a failure of two of its regulatory programs – its chemical testing program under the Toxic Substances Control Act and its underground storage tank program under the Hazardous and Solid Waste Amendments to the Resource Conservation and Recovery Act.302 Methyl Tertiary-butyl Ether ("MTBE") was a widely used gasoline additive between the late 1970s and the early 2000s. Gasoline refining companies added MTBE to their products for three major purposes. First, some refining companies beginning the late 1970s blended MTBE into mid- and high-grade gasoline instead of tetra-ethyl lead to enhance octane and thereby prevent engine knocking and resulting performance loss. Second, in the late 1980s, petroleum companies began to blend MTBE into gasoline to meet state and federal winter oxygenate requirements for a limited number of areas that had not attained the National Ambient Air Quality Standards for carbon monoxide. Third, petroleum refiners used MTBE to meet the requirements of the 1990 Clean Air Act Amendments for reformulated gasoline ("RFG") in the

299

Environmental Protection Agency, Part 1; Chrysotile Asbestos; Regulation of Certain Conditions of Use Under the Toxic Substances Control Act (TSCA), 89 Fed. Reg. 21,970 (March 28, 2024), at 21,987. 300 Madeline Lyskawa, EPA Says Legacy Asbestos Poses Unreasonable Risk, Law 360, April 15, 2024); Bob Egelko, Court Tells EPA It Must Regulate Toxic Materials, San Francisco Chronicle, November 15, 2019. 301 Rachel Rothschild, Unreasonable Risk: The Failure to Ban Asbestos and the Future of Toxic Substances Regulation, 47 Harv. Env. L. Rev. 529 (2023), at 531-32. 302 Except where otherwise indicated, the MTBE description and analysis is drawn from Thomas O. McGarity, MTBE: A Precautionary Tale, 28 Harv. Env’l L. Rev. 281 (2004). Citations to the relevant EPA and industry documents may be found in the footnotes to that article.

ten large metropolitan areas with the most severe summertime photochemical oxidant (ozone) levels. Because MTBE was very soluble in water, it traveled much more rapidly in groundwater than the other components of gasoline. In addition, it was more resistant to biodegradation. MTBE was much more likely than the other components of gasoline to contaminate drinking water, and it was much more difficult to remove from contaminated groundwater than those other components. Because MTBE had an extremely unpleasant taste and odor and was detectable in drinking water at very low concentrations, its presence rendered entire aquifers unusable as drinking water supplies. The expanded use of MTBE as a gasoline additive in the 1990s and the deteriorating state of America's sixty-year-old infrastructure of underground storage tanks for gasoline combined to yield an MTBE groundwater crisis by the end of the 1990s when studies documented MTBE in many of the nation's drinking water supplies. When ARCO first began producing MTBE in December 1979, very little was known about its chronic toxicity. ARCO and other petroleum companies agreed in late 1980 to an industry-wide effort, supervised by the American Petroleum Institute ("API"), to conduct core toxicity testing on MTBE. EPA had authority under section 4 of the Toxic Substances Control Act (TSCA) to order the manufacturer of a chemical substance to conduct specific health and environmental toxicity testing if EPA determined that the sufficient testing was not already available and the chemical might present an unreasonable risk to the environment or would be produced in substantial quantities and enter the environment in substantial quantities or give rise to significant human exposure. The statute also created an Interagency Testing Committee (ITC), composed of representatives of several federal agencies, to nominate for testing chemicals that met this multifaceted threshold test. Once a chemical appeared on the ITC "priority list" of 50 chemicals, EPA had to decide within one year whether to issue a rule ordering further testing. By the time that the ITC began its investigation of MTBE in 1985, the industry had already learned a lot about MTBE. First, the industry knew that MTBE was beginning to contaminate groundwater in many states as a result of leaking USTS. Second, the industry had learned that MTBE migrated faster in groundwater than other gasoline constituents and that its pungent odor and unpleasant taste were detectable by the human senses at very low concentrations. The industry was also in the process of learning more about MTBE's toxicology. API's Toxicology Committee had in October 1980 approved several core toxicology tests on MTBE as Phase I of a larger project. The tests were, however, devoted exclusively to inhalation exposure to MTBE by refinery workers and people in the vicinity of gasoline pumps, and the industry consistently rejected suggestions by the API Ad Hoc MTBE Toxicology Group to test MTBE in

drinking water. When the Phase I studies were completed in mid-1984, the industry concluded that the results were "rather unremarkable in terms of causing harm." Instead of moving to Phase II (long-term carcinogenicity studies), the API group hoped that the industry's proactive effort would "preclude . . . an unnecessary test rule by EPA under TSCA." The ITC in 1985 entered into a contract with CRCS, Inc. to conduct an "information review" of the existing published and unpublished literature on the health and environmental effects of MTBE. A March 1986 "Working Draft" reported that although about 1.37 billion pounds of MTBE had been used in 1984 as an octane-enhancing blending component in gasoline, the "toxicological information" on MTBE was limited. Genotoxicity studies "gave conflicting results," and no chronic toxicity or carcinogenicity studies were located for MTBE. The survey located no empirical information at all on consumer exposure, which it assumed would come exclusively via the inhalation and dermal routes. Information on accidental releases of MTBE came exclusively from an unpublished 1985 API report concluding that MTBE "could be released to the environment in fugitive [air] emissions or accidental spills." The report did not mention the large body of information that the industry possessed on accidental releases of MTBE into the environment via leaking USTS. The CRCS working draft listed leaking USTS as one of three minor sources of releases. With the distribution of the CRCS report, the industry launched a major effort to avoid an EPA rule requiring "time consuming and expensive" testing. Arco submitted to the ITC a critique of the working draft of the CRCS report that explained its "differences in interpretation of the data" and provided "new information that is expected to be beneficial to the ITC in any further analysis of this material." Arco complained that “erroneous assumptions have been made that cause the potential hazards of MTBE to be seriously overestimated." The Arco submission acknowledged that human exposure "from accidental spills of MTBE could occur," but they "should be regarded as a minimal possibility," and it assured the ITC that "MTBE losses would be extremely small from" leaking USTS. On October 31, 1986, the ITC recommended that MTBE be tested for chronic inhalation toxicity including neurotoxic, hematologic, and oncogenic (carcinogenic) effects and that additional monitoring be undertaken of worker and customer exposure in the "breathing zone" during gasoline transfer operations. The report made no reference at all to MTBE-contaminated groundwater due to leaking USTS. In response, Arco argued that testing for chronic inhalation health effects was "not necessary," because "worst case" exposures to MTBE from gasoline vapors were "well below the 'no observable adverse effect level' even when very conservative safety factors are applied." This did not, of course, answer the question whether testing for oncogenicity, for which there were not necessarily any no-effect levels, would be desirable. Arco did not mention the possibility that chronic exposure might occur via ingestion of contaminated groundwater.

On December 17, 1986, EPA hosted a "focus meeting" on MTBE at which most of the major industrial players gathered to discuss the possibility of arriving at an agreed upon consent order for performing additional testing on MTBE. No representatives of environmental groups attended this or any of the subsequent focus meetings. At the meeting, EPA's project manager noted that "an additional concern" identified by EPA's Test Rules Development Branch was "contamination of ground water supplies by MTBE." She related that more than 700,000 USTS were used for petroleum products and "about 30% of these tanks leak." The industry representatives, however, insisted that there was no reason to require any additional testing of MTBE because there should be "very little cause for concern of health hazards with MTBE." After more than a year of additional negotiations, EPA published notice of a Consent Order to which EPA and five major oil companies had agreed. The companies agreed to conduct several mutagenicity tests, several pharmacokinetics tests to determine oral, dermal, and inhalation routes of exposure, three neurotoxicity tests, an inhalation oncogenicity test in two species and an inhalation two-generation reproduction and fertility effects study. The companies did not agree to do any environmental testing, and they agreed to very little in the way of toxicity testing via the ingestion route through which humans would be exposed to MTBE in drinking water. Soon after Congress enacted the Hazardous and Solid Waste Amendments, EPA in early 1985 created a new Office of Underground Storage Tanks (OUST) to undertake the necessary implementation efforts. Fearing that it lacked sufficient time and resources to promulgate detailed requirements, the office decided to explore "quicker, less detailed approaches to regulation." A "Development Plan" proposed to develop "broad, performance-oriented tank technical standards rather than detailed, technology-based operating and design requirements." In other words, the agency was prepared to specify broad goals and let the regulated industry fill in the details. According to an EPA official who was heavily involved in drafting the regulations, the agency "worked closely with representatives of the major oil companies," regularly contacting major oil company representatives "for information and input regarding UST issues." On September 23, 1988, EPA published final regulations establishing technical requirements for USTS. The preamble noted somewhat defensively that the statute specifically authorized EPA to "consider industry practices and consensus codes in developing appropriate UST regulations," and it observed that several important industry codes had been updated in the year since the proposal came out. This increased activity "support[ed] EPA's proposed reliance on these codes as providing the most up-to-date consensus practices and expertise concerning what constitutes proper UST system management." Since "the successful implementation of" the final regulations depended "a great deal on the regulated community's voluntary compliance," the final regulations relied "as much as possible" on "familiar industry codes." For new and replacement tanks, the final regulations required only protected single-walled tanks with release detection. Although the agency agreed with commenters who suggested that "there will probably be more releases to the environment" from single-walled tanks than from double-

walled tanks with interstitial monitoring, the more protective option was "not believed to be necessary to protect human health and the environment." In addition, the "widely available" technologies for cleaning up "petroleum products" provided "the means to ensure that adverse impacts from such releases (when they occur) can be managed and remediated." Finally, the preamble mentioned that double-walled systems entailed "greater capital and installation costs" that did not "justify" the environmental benefits and that the "current trends in industry" were not in the direction of double-walled tanks. The fact that MTBE was not easily remediated through "widely available" cleanup technologies was apparently lost on the agency. The regulations required a "gradual" upgrade or replacement of existing tanks over a period of 10 years. The agency agreed with commenters who argued that a more rapid upgrade schedule that would "prevent a significant number of future product releases," but it rejected their argument that industry resources were capable of meeting a more rapid upgrade requirement. A 10-year upgrade program, on the other hand, would "complement current industry trends towards upgrading or replacing voluntarily." Instead of requiring upgrades to the new tank standards, the final rule allowed owners to meet the upgrade requirements by lining the interiors of existing steel tanks in accordance with industry standards so long as the tanks were tested within the next ten years and at five-year intervals thereafter. The final rule allowed owners to choose between six broad leak detection technologies, each of which were required to comply with "method-specific" performance standards. In particular, owners were allowed to rely upon semi-annual tightness testing and inventory control as a leak detection mechanism. The agency reached this conclusion despite its own previously expressed "serious reservations" about inventory control as a leak detection technique. The final rule also relaxed the performance requirements for inventory control. An EPA enforcement official later observed that the regulations achieved "flexibility" for the industry at the "expense of stringency and enforceability." The petroleum industry was generally "comfortable with" the technical requirements, and it expressed relief that the regulations turned out to be so flexible. Environmental groups, by contrast, were quite unhappy with the new rules. They argued that EPA should have required new and replacement tanks to be double-walled. They also maintained that EPA gave the industry far too long to replace or upgrade steel tanks. A spokesperson for the Environmental Defense Fund complained that the regulations were designed to minimize the economic impact on the industry, not to protect human health and the environment. Perhaps because they did not want to delay EPA's implementation of the regulations, however, the environmental groups declined to challenge them in court. The agency failure in the case of MTBE was not simply a matter of the right hand not knowing what the left hand was doing. EPA's OTS staff was fully aware of what its OUST staff knew about the presence of MTBE in groundwater near many USTS at the time that EPA negotiated a testing agreement that did not require testing of the health effects of MTBE in groundwater.

More importantly, EPA's OUST staff knew that MTBE was in gasoline when it wrote lenient USTS technical standards that declined to require double-walled tanks and deferred to industrypromulgated national consensus standards. The problem was not lack of knowledge, but a lack of the political will to address the special problems of MTBE during a deregulatory period of EPA's existence. The consequences of EPA’s regulatory failure became manifest a decade later. On March 25, 1999, California Governor Gray Davis issued an Executive Order banning MTBE from California gasoline by the earliest possible date. Soon thereafter, a finding by a California jury that several oil companies were liable for contaminating the water supply of Lake Tahoe led to a $67 million settlement, and several oil companies tentatively agreed to pay the City of Santa Monica $30 million in damages and spend more than $200 million for a new water treatment plant. As similar lawsuits proceeded throughout the country, 15 more states banned MTBE from gasoline. On July 22, 2003, the Houston Chronicle reported that one of the largest manufacturers of MTBE had filed for bankruptcy protection. The Energy Policy Act of 2005 removed the oxygenate requirement for reformulated gasoline and added a renewable fuel requirement. In response, refiners stopped using MTBE and began blending fuel with ethanol.303 3. “Forever chemicals”: PFOA, PFOS, and other PFAS “Forever chemicals”304 have become so notorious, it is possible that almost as many people have heard of them as have been exposed to them.305 And lately they have received a great deal of regulatory attention.306 But the PFAS “forever chemicals” may be the type specimen of a hazard that prompted regulation only after being exposed by private tort litigation. Attorney Robert Bilott told that story from his perspective in his book Exposure.307 Bilott recounted how, through a family connection, he agreed to represent Earl Tennant, a West Virginia

303

Environmental Protection Agency, Methyl Tertiary Butyl Ether (MTBE), https://archive.epa.gov/mtbe/web/html/gas.html. 304 The sobriquet describes at least some extremely persistent chemicals within the very broad class of per- and poly-fluorinated alkyl substances (PFAS). [cite]. These compounds’ characteristic carbonfluoride bonds resist biological and chemical degradation. Hubertus Brunn et al., PFAS: Forever Chemicals – Persistent, Bioaccumulative and Mobile. Reviewing the Status and the Need for Their Phase Out and Remediation of Contaminated Sites, 35(1):20 Env’t Scis. Eur. at 2 (2023), https://doi.org/10.1186/s12302-023-00721-8. 305 “[I]f you are reading this during the first quarter of the twenty-first century, you already have PFOA [perfluorooctanoic acid] in your blood.” NATHANIEL RICH, SECOND NATURE 49 (2021). 306 See, e.g., 89 Fed. Reg. 39,124 (May 8, 2024) (finalizing listing of PFOA and PFOS, closely-related eight-carbon PFAS, as hazardous substances under CERCLA); 89 Fed. Reg. 32,532 (Apr. 26, 2024) (setting Maximum Contaminant Level for PFOA, PFOS, and several other PFAS under SDWA). 307 ROBERT BILOTT, EXPOSURE (2019). For shorter retellings of the history of Bilott’s litigation against PFAS manufacturers, see Rich, supra note Error! Bookmark not defined., at 17-51; Mark P. Nevitt & Robert V. Percival, Can Environmental Law Solve the “Forever Chemical” Problem, 57 WAKE FOREST

cattle farmer who believed his animals were being poisoned by a chemical in runoff from an E.I. Du Pont de Nemours landfill.308 Bilott’s unexpected turn as a plaintiff’s lawyer was nearly stymied when he could not identify a “listed and regulated” substance in the landfill that could explain the runoff’s apparent toxicity.309 Bilott’s problem was that the unknown cause of disease and death in Tennant’s cattle, perfluorooctanoic acid (PFOA), was not a “listed and regulated” chemical when Bilott filed Tennant’s lawsuit against Du Pont in June of 1999. Nor did it become one, at the federal level, until at least 2020.310 The pertinent question for this essay is: why? Part of the answer is that the use and discharge of PFOA, its chemical sibling PFOS, and other long-chain PFAS began before the era of environmental regulation. Du Pont began using PFOA in the manufacture of Teflon™ in 1951311; researchers at 3M Corporation discovered PFOS in 1953.312

L. REV. 239, 251-55 (2022) (briefly recounting the history of three lawsuits Bilott filed against PFAS manufacturers). 308 Bilott, supra note Error! Bookmark not defined., at {pin}. 309 Id. at 44, 60 (describing Du Pont’s refusal to provide discovery about chemicals in the landfill or the manufacturing plant that used the landfill unless the chemicals were listed and regulated or specifically identified in the discovery request); id. at 60, 149, 194, 203, 213 (describing successful motions to compel or for sanctions and documents received in response). 310 In 2020, EPA required Toxics Release Inventory reporting for 172 PFAS including PFOA. 85 Fed. Reg. 37,354, 37,357–64 (June 22, 2020) (listing perfluorooctanoic acid and numerous PFOA salts as chemicals subject to reporting); see generally 42 U.S.C. § 11023 (establishing toxic release inventory). The 2020 National Defense Authorization Act required EPA to take this action. P.L. 116-92 § 7321 (Dec. 20, 2019). EPA issued its first binding regulation limiting the discharge or presence of PFOA in environmental media, as opposed to merely requiring reporting, only in 2024. 89 Fed. Reg. 32,532, 32,532 (Apr. 26, 2024) (National Primary Drinking Water Regulation setting 4 parts per trillion Maximum Contaminant Level for PFOA, PFOS, and other PFAS). This regulation followed EPA’s establishment of much higher drinking water advisory levels for PFOA in 2009 and 2016. See US EPA, Provisional Health Advisories for Perfluorooctanoic Acid (PFOA) and Perfluorooctane Sulfonate (PFOS) 4-5 (calculating provisional health advisory levels of 400 ppt (4 μg/L ) for PFOA and 200 ppt for PFOS); US EPA, Drinking Water Health Advisory for Perfluorooctanoic Acid (PFOA) 9 (May 2016) (“issuing a lifetime drinking water Health Advisory (HA) for PFOA of 0.07 micrograms per liter (μg/L) [70 ppt]”). The earliest binding federal regulation we found of any kind that affected any fluorinated alkyl chemical was a significant new use rule (SNUR) promulgated by EPA in 2002. This rule effectively prohibited renewed manufacture or import of thirteen derivatives of perfluorooctanesulfonic acid (PFOS) that 3M had already discontinued. 67 Fed. Reg. 11,008, 11,008–09 (Mar. 11, 2002). EPA and the Department of Defense also banned the discharge of fluorinated aqueous film-forming foam from military vessels in 2017. 82 Fed. Reg. 3,173, 3,178 (Jan. 11 2017). 311 Bilott, supra note Error! Bookmark not defined., at 63. 312 Sara Samora, The History of PFAS: From World War II to Your Teflon Pan, MANUFACTURING DIVE (Dec. 6, 2023), www.manufacturingdive.com/news/the-history-behind-forever-chemicals-pfas-3mdupont-pfte-pfoa-pfos/698254 (last visited May 26, 2024).

Red flags of these compounds’ toxicity began to emerge from animal studies as early as the 1960’s.313 Nevertheless, they escaped regulation under the slew of environmental statutes enacted in the 1970’s. Plants using or manufacturing PFAS emitted large quantities of PFAS into the atmosphere,314 but these compounds were not and still are not regulated as hazardous air pollutants.315 Plants discharged PFAS into waters of the United States, unconstrained by effluent limitations in discharge permits, 316 but these compounds were never regulated as toxic water pollutants.317 PFAS were disposed of in landfills318, but neither discarded PFAS nor PFASbearing industrial wastes were regulated as hazardous waste.319

313

See Bilott, supra note Error! Bookmark not defined., at 169 (describing efforts by Du Pont to limit employees’ exposure to PFOA during the 1950’s, rat study in 1962 showing effects on liver and other organs, and dog study in 1964 showing liver damage); see also Environmental Working Group, For 50 Years, Polluters Knew PFAS Chemicals Were Dangerous But Hid Risks From Public, https://static.ewg.org/reports/2019/pfa-timeline/3M-DuPont-Timeline_sm.pdf (last visited May 26, 2024) (presenting documents from 3M and Du Pont relating to toxicity of PFOA, PFOS and other PFAS) (hereafter, EWG 50 Years); id. at 65 (copy of 1965 rat study sponsored by Du Pont). 314 For example, Du Pont reported that in 1999, its Washington Works in Parkersburg, West Virginia, where PFOA was used in the manufacture of Teflon™, emitted about 24,000 pounds of PFOA to the air. Bilott, supra note Error! Bookmark not defined., at 51. Befitting the “forever chemicals” moniker, PFOA and PFOS are still present in ambient air. J. Zhou et al., PFOS Dominates PFAS Composition in Ambient Fine Particulate Matter (PM2.5) Collected Across North Carolina Nearly 20 Years After the End of Its US Production, 23 ENVT. SCI. PROCESSES IMPACTS 580, 582 (“PFOA and PFOS were present at the highest concentrations of the 34 PFSs targeted in this analysis”). 315 See 42 U.S.C. § 7412. PFOA and PFOS were, of course, already in use at the enactment of the modern Clean Air Act in 1970. Congress did not include these compounds in the statutory list of hazardous air pollutant established by the 1990 amendments to the Clean Air Act, id. § 7412(b)(1) and EPA has not added them to the list by regulation, see 40 C.F.R. § 61.01 (listing a number of hazardous air pollutants). 316 For example, Du Pont reported that in 1999, Washington Works discharged 55,000 pounds per year of PFOA into the Ohio River. Bilott, supra note Error! Bookmark not defined., at 51. See also, e.g., Congressional Res. Svce., Regulating PFAS Under the Clean Water Act 1 (June 28, 2022) (“In some instances, EPA has used NPDES permit authorities to address PFAS and has taken steps to encourage states to use such authorities.”); Letter from David W. Boothe to US EPA OPPT Document Control Office 31-34 (Oct. 29, 2010) (attachment reporting PFOA discharges from a Du Pont manufacturing plant in Pascagoula, Mississippi to the Mississippi Department of Environmental Quality), available at Regulations.gov, EPA docket # OPPT-2006-0621; State of Mississippi, Draft Water Pollution Control Permit No. MSP090360 (Mar. 29, 2023) (requiring reporting of PFAS emissions from the Pascagoula plant but not specifying a discharge limitation) (copy on file with authors). [add NJ data re: emissions from Solvay] 317 See 42 U.S.C. § 1317(a) (establishing list of toxic pollutants). 318 See Bilott, supra note Error! Bookmark not defined., at 80-83 (describing disposal of PFOAcontaminated sludge in unlined digestion pits at Du Pont’s Washington Works, excavation of 7,100 tons for re-disposal at non-hazardous-waste Dry Run Landfill, and subsequent disposal of PFOA-contaminated biocake at Dry Run Landfill). 319 See [CFR] (PFAS not listed as hazardous waste); [CFR] (PFAS not listed as constituents tested in Toxicity Characteristic Leaching Procedure to identify characteristic hazardous waste). {note that at least one authorized state has regulated long-chain PFAS waste as hazardous waste}

Each of the media-based pollution statutes, of course, required a factual predicate for regulation. Hazardous air pollutants are those “which present, or may present . . . a threat of adverse human health effects. . . .”320 To list a toxic water pollutant, EPA must “take into account toxicity of the pollutant, its persistence, degradability,” and the presence of, importance of, and effect on affected organisms.321 Solid waste is hazardous waste only if it may “[c]ause . . . an increase in mortality or an increase in serious irreversible, or incapacitating reversible, illness”322 or “[p]ose a substantial present or potential hazard to human health or the environment . . . .”323 Until about the year 2000, not only was EPA unable to establish any of these predicates with respect to PFOA, PFOS or other PFAS, but these chemicals were not even on EPA’s radar screen – even though, by the mid-1970s, researchers had found high levels of PFAS in production workers, had discovered that PFAS are biopersistent, and had found toxic effects of PFAS in animals.324 As Bilott observed during the Tennant litigation, “[i]t appeared that the only existing PFOA research had been conducted by industrial scientists who worked for companies that manufactured or used it.”325 By coincidence, much of that research occurred shortly before and shortly after the 1976 enactment of the Toxic Substances Control Act (TSCA).326 TSCA might have brought the growing knowledge of PFAS hazards into the light, but it did not. TSCA’s failure to reveal, let alone protect against, these hazards resulted from inherent flaws in the statutory design. To the limited extent that TSCA required collection (not generation) of toxicity data, the requirement applied only to new chemicals327 – but PFOA and PFOS were not new. Of course, the original TSCA also required chemical manufacturers to alert EPA to “information which reasonably supports the conclusion that [a chemical] presents a substantial risk”328 – but that requirement proved toothless both in general329 and in the specific cases of 320

42 U.S.C. § 7412(b)(2). 33 U.S.C. § 1317(a)(1). 322 40 C.F.R. § 261.10(a)(1)(i). 323 Id. § 261.10(a)(ii); see also id. § 261.11 (specifying criteria for listing hazardous waste). 324 EWG 50 Years, supra note Error! Bookmark not defined., at 275–331 (providing copies of studies and internal corporate documents created between 1977 and 1979 summarizing research on blood levels and health effects of PFOS, PFOA and other PFAS). 325 Bilott, supra note Error! Bookmark not defined., at 59. 326 P.L. 94-469. 327 John S. Applegate, Synthesizing TSCA and REACH: Practical Principles for Chemical Regulation Reform, 37 ECOL. L.Q. 721, 732 (2008) (“TSCA’s screening process applies only to new chemicals (or significant new uses of existing ones)”). 328 P.L. 94-469 § 8(e). 329 Applegate, supra note Error! Bookmark not defined., at 732 (“reporting [under section 8] is inconsistent”); id. at 736 (noting that in practice reporting was left “largely to the manufacturer’s own judgment); Bilott, supra note Error! Bookmark not defined., at 95 (“For the regulation of [existing] chemicals–such as PFOA and PFOS–[under TSCA,] everyone was basically left to operate under a kind of honor system.”) 321

both PFOA and PFOS (at least until after the fact).330 Although 3M eventually told EPA enough about PFOS331 to lead to the compound’s “voluntary” phase-out,332 the power of civil discovery in a simple tort action revealed vast amounts of information that might otherwise have remained hidden – or that EPA itself might have had to sue to obtain. Indeed, when EPA finally took administrative enforcement action against Du Pont for violating TSCA’s notification requirement, much of the evidence EPA used to support the complaint came from Bilott.333 “Much of what I was sharing [w/ EPA] was hard-won internal DuPont data obtained only through all my discovery battles and would not turn up in a search of publicly available documents.”334 Even armed with information about PFOA and PFOS from Bilott, 3M, and elsewhere, EPA did not swing a regulatory hammer on the use and release of these chemicals. Du Pont, for its part, began manufacturing its own PFOA in 2002, after 3M announced its phase-out of PFOS.335 Six years after 3M’s agreement to stop using PFOS, EPA finally achieved an agreement with eight manufacturers to phase out the use and discharge of PFOS, PFOA, and a handful of other longchain PFAS.336 This “PFOA Stewardship Program,” on the one hand, merely committed the industry to continue a trend that started once the handwriting was on the wall about long-chain

330

See US EPA, EPA Settles PFOA Case Against DuPont for Largest Environmental Administrative Penalty in Agency History, press release (Dec. 14, 2005) (describing settlement for $10.25 million penalty and $6.25 million in supplemental environmental projects, resolving a complaint that included seven counts under section 8(e) of TSCA for failure to notify EPA); US EPA, EPA Settles Case Involving 3M Voluntary Disclosures of Toxic Substances Violations, press release (Apr. 25, 2006) (describing settlement for a $1.5 million penalty for 244 TSCA violations, including failures to notify, and noting that “3M produced valuable, previously unreported information that will help the scientific community”). 331 US EPA, EPA and 3M Announce Phase Out of PFOS, press release (May 16, 2000) (stating that “3M data supplied to EPA indicated that [PFOS and related] chemicals are very persistent in the environment, have a strong tendency to accumulate in human and animal tissues and could potentially pose a risk to human health and the environment over the long term.”). 332 Id. (“Following negotiations between EPA and 3M, the company today announced that it will voluntarily phase out” PFOS). 333 Complainant’s Initial Prehearing Exchange for Hearing on Count 1, In re E. I. du Pont de Nemours and Company (Feb. 17, 2005) (listing various letters from Bilott to EPA, attachments thereto, and depositions from the class action Bilott filed against Du Pont after the Tennant litigation settled). 334 Bilott, supra note Error! Bookmark not defined., {somewhere between 190 and 250}. 335 [Dupont 2009 report p 36 “Fayetteville Works’ APFO Manufacturing Facility commenced production in late 2002.”]; Bilott, supra note Error! Bookmark not defined., at 227. 336 Steve C. Gold & Wendy E. Wagner, Filling Gaps in Science Exposes Gaps in Chemical Regulation, 368 SCIENCE 1066, 1066 (2020).

PFAS337 while, on the other hand, its languid pace allowed continued use of these compounds for a decade.338 When the data dam finally broke, it unleashed a torrent. The number of scientific articles published on PFAS increased exponentially in the first two decades of the twenty-first century.339 Bilott generated some of the new data through a settlement of class action litigation that provided for detailed epidemiologic study of PFOA exposure and health effects by a neutral “Science Panel.”340 The Science Panel ultimately found “probable links” between PFOA exposure and six human diseases or conditions.341 The companies that signed up for the PFOA Stewardship Program committed to eliminating long-chain PFAS, but the agreement was conspicuously silent about the chemicals these companies might use as replacements. Some of these replacements also appear to be toxic (although not necessarily as toxic, persistent, or bioaccumulative as PFOA and PFOS).342 Some of them were long obscured by TSCA’s liberal treatment of manufacturers’ claims of confidential business information343, to the point that even EPA scientists had to resort to novel methods to

337

Id. at 1066 (“By the time of this agreement . . . the phase-out of PFOA was already well under way informally”). Some of the agreeing companies told EPA in 2006 that they had already met the program’s goal for 2010. Letter from Michael Lacy to Stephen L. Johnson, March 1, 2006 (stating that “Solexis has already achieved a greater than 95% reduction from the 2000 baseline year”), available at https://www.regulations.gov/document?D=EPA-HQ-OPPT-2006-0621-0005 (last visited May 22, 2020). [see also Bilott 258, Du Pont says it will phase out PFOA by ‘06] 338 {cite to 2010/2015 PFOA Stewardship Program with “work toward” complete elimination of longchain PFAS by 2015; find in the docket the latest report documenting releases/product content]. [Note that most of the 8 companies eliminated long-chain PFAS before the 2015 target}. 339 Majeti Narasimha Vara Prasad & Sailaja V. Elchuri, Environmental Contaminants of Emerging Concern: Occurrence and Remediation, 28 Chem. Didactics Ecology Metrology 57, 59 (2023) (graph showing number of articles per year increasing from 501 in 2000 to 7,074 in 2022); see also Bilott, supra note Error! Bookmark not defined., at 281 (“At that point [2007-2010], the volume of PFOA research around the globe was exploding.”). 340 {Bilott at 243 (check p); cite one of the Sargus opinions.} 341 {Bilott at 306}. 342 Du Pont developed GenX (hexafluoropropylene oxide dimer acid, a six-carbon PFAS also known as HFPO-DA), as a substitute for PFOA, and 3M developed PFBS (perfluorobutane sulfonic acid, a four carbon PFAS) as a substitute for PFOS. Evidence exists that both these compounds are toxic, see Ashley Ahearn, A Regrettable Substitute: The Story of GenX, 127 ENV'T HEALTH PERSP. (Supp.) 1 (2019) (GenX); Human Health Toxicity Values for Perfluorobutane Sulfonic Acid (CASRN 375-73-5) and Related Compound Potassium Perfluorobutane Sulfonate (CASRN 29420-49-3) (Apr. 2021). GenX and PFBS were included in the 2024 regulation setting maximum contaminant levels under the Safe Drinking Water Act. {cite}. 343 Wendy E. Wagner & Steve C. Gold, Legal Obstacles to Toxic Chemical Research, 375 Science 138, 140 (2022) (describing difficulty of identifying chemical that are listed confidentially in EPA’s TSCA database).

identify the compounds found in environmental media344 – and publicly available toxicity data were nonexistent.345 Also conspicuously absent from the PFOA Stewardship Program was any regulatory response to the problems presented by the widespread environmental dissemination of persistent and bioaccumulative PFAS. “After the PFOA Stewardship Program fanfare, EPA’s engagement in all PFOA matters essentially came to a screeching halt . . . . EPA, in fact, would take no significant additional action on PFOA for another decade.”346 In 2019, EPA published a multi-pronged PFAS action plan347, which EPA updated in 2021.348 The 2024 rules that established enforceable limits under the Safe Drinking Water Act and listed PFOA and PFOS as CERCLA hazardous substances are the action plan’s most concrete regulatory accomplishments.349 {Need to add something about TSCA 8a7 reporting rule.} On other fronts, however, regulatory goals still seem distant. The action plan calls for EPA to build a technical foundation to address PFAS air emissions350; to date EPA’s visible progress is a rule allowing data collection that was proposed in July 2023 and has not been finalized.351 EPA has announced plans to develop effluent limitation guidelines for water discharges of PFAS352 but has not yet completed the task, although EPA has issued guidance for addressing PFAS in water pollution discharge permits.353 Small wonder that in 2021, as EPA inaugurated its PFAS strategic road map, two scholars concluded that “[t]he regulatory response to PFAS in the United States has been slow and fragmented. . . .” 354 The decades in which PFOA and PFOS spread in environmental media around the world, their danger undetected and their release unregulated, later resulted in a cataclysmic amount of litigation and liability. After the Science Panel report, individuals in the plaintiff class who suffered from the diseases linked to PFOA exposure sought compensation, as the settlement

344

Id.; {Washington et al., Science 2020}. Gold & Wagner, supra note Error! Bookmark not defined., at 1067. 346 Bilott, supra note Error! Bookmark not defined., at 273. 347 {cite}. 348 {cite to Strategic Road Map}. 349 {x-ref see supra}. 350 {cite action plan}. 351 88 Fed. Reg. 54,118 (July 24, 2023). 352 88 Fed. Reg. 6,258 (Jan. 31, 2023). 353 {cite to 2022 guidance documents for EPA regions and state/federal permitting authorities, respectively}. 354 Daniel Renfrew & Thomas W. Pearson, The Social Life of the “Forever Chemical,” 12 ENV’T & SOC’Y.: ADV. RES. 146, 150 (2021). 345

permitted them to do.355 After three jury trials resulted in increasing damage awards,356 Du Pont settled all the claims for $670.7 million.357 Thousands of other PFAS lawsuits have been filed, mostly involving the legacy long-chain compounds like PFOA and PFOS. Some of these suits are based on industrial releases, while others are based on intentional releases of PFOS and other PFAS used in aqueous film-forming firefighting foams. The plaintiffs range from individuals, including firefighters and consumers of drinking water, to cities and other water utilities whose drinking water sources have been contaminated, to states that have sought response costs or damages for injury to natural resources from PFAS manufacturers. Publicly-reported settlements and judgments exceed $13 billion, with many cases pending and more cases being filed.358 But the legal liabilities associated with PFAS pale beside the environmental and public health costs. One study suggests that entirely removing PFAS from the global environment – an obviously unrealistic goal – would cost more than the global GDP.359 Another, relying on a broad swath of epidemiologic literature, estimates that long-chain PFAS are responsible for a disease burden ranging $5.52 to $62.6 billion in annual costs in the United States alone.360 In a self-laudatory corporate biography published two years after it decided to discontinue PFOS, 3M praised its highly creative scientists for “the fuzzy front-end experimentation” that led to the development of Scotchgard, the wildly successful PFOS-based fabric protector.361 If only there had been a way to ensure fuzzy front-end experimentation about not just PFOS’s valuable properties but also about its environmental and health effects.

355

{cite a Sargus opinion} Bilott, supra note Error! Bookmark not defined., at 352, 359, 363 (describing verdicts awarding $1.6 million, $5.6 million including $0.5 million in punitive damages, and $12.5 million including $10.5 million in punitive damages). {probably told more efficiently in Rich}. 357 Id. at 364. 358 The largest single settlement was between 3M and a class of some 11,000 public water systems, under which 3M will pay between $10.5 and $12.5 billion to the class over a period of thirteen years. In re Aqueous Film-Forming Foams Prods. Liab. Litig., 2024 WL 1341122 (Mar. 29, 2024) (approving the settlement). Other significant settlements and judgments include {list}. 359 Alison L. Ling, Estimated Scale of Costs to Remove PFAS from the Environment at Current Emission Rates, 918 SCI. TOTAL ENV’T 170647 (2024), 1 https://doi.org/10.1016/j.scitotenv.2024.170647 . 360 Vladislav Obsekov et al., Leveraging Systematic Reviews to Explore Disease Burden and Costs of Perand Polyfluoroalkyl Substance Exposures in the United States, 15 EXPOSURE HEALTH 373, 390 (2023). 361 3M Company, A CENTURY OF INNOVATION: THE 3M STORY 54 (2002). 356

Devastatingly early draft – please do not cite or share

EVIDENCE OF COMPLIANCE Maggie Wittlin* I.

Introduction

Unlike tobacco or asbestos,1 opioids came to market with the approval of a major federal regulator charged with protecting public health: the Food and Drug Administration. And opioid manufacturers have capitalized on their FDA approval both to advertise their product and to succeed in litigation. Famously, in 1995, the FDA allowed Purdue to claim that OxyContin’s delayed-absorption mechanism was “believed to reduce the abuse liability” of the drug relative to faster-acting painkillers.2 Purdue used this decision to pitch OxyContin to doctors as a less addictive alternative.3 In lawsuits against them, opioid manufacturers have argued that the federal statutory framework regulating drugs and drug labels preempts state-law claims,4 and they have sometimes succeeded in getting claims dismissed on this basis.5 And they have successfully argued for denial of class certification using an FDA representative’s testimony that, when used according to approved labeling, the benefits of the drug outweigh the risks.6 The court denied certification based on lack of commonality, in part because the proposed class included those who had been prescribed the drug in accordance with its labeling.7 The advantages opioid manufacturers have obtained from FDA approval have exceeded the reliability of the agency’s decisions. Purdue obtained permission to include the OxyContin insert suggesting reduced abuse liability based not on clinical studies, but on the theory that drug abusers would prefer a drug whose effects could be felt immediately.8 Indeed, some users wanted a quicker hit—which they obtained by crushing or dissolving OxyContin.9 And in tension with the FDA representative’s Congressional testimony, a Purdue spokesman later acknowledged, “[p]atients who take OxyContin in accordance with its F.D.A.-approved labeling instructions will likely develop physical dependence.”10 In part due to misrepresentations from manufacturers,11 and in part due to its own failures, the agency failed to prevent the opioid crisis.12 As the case of opioids demonstrates, regulatory approval—and, more broadly, regulatory compliance—provides product manufacturers advantages in marketing and in litigation, some of * Associate Professor, Fordham University School of Law.

1 Robert L. Rabin, Keynote Paper: Reassessing Regulatory Compliance, 88 Geo. L.J. 2049, 2069 (2000).

2 Patrick Radden Keefe, The Family That Built an Empire of Pain, The New Yorker, Oct. 30, 2017; Barry Meier, Origins of an Epidemic: Purdue Pharma Knew Its Opioids Were Widely Abused, N.Y. Times, May 29, 2018. 3 Meier, supra. Its claims went beyond those approved by the FDA. Id. 4 See, e.g., State ex rel. Jennings v. Purdue Pharma L.P., No. CVN18C01223MMJCCLD, 2019 WL 446382 (Del. Super. Ct. Feb. 4, 2019). 5 E.g. North Dakota ex rel. Stenehjem v. Purdue Pharma L.P., No. 08-2018-cv-001300, 2019 WL 2245743, at *8 (N.D. Dist. Ct. May 10, 2019); Caltagirone v. Cephalon, Inc., 190 A.3d 596 (Pa. Super. 2018); see Nora Freeman Engstrom & Robert L. Rabin, Pursuing Public Health Through Litigation: Lessons from Tobacco and Opioids, 73 Stan. L. Rev. 285, 336 (2021). 6 Wethington v. Purdue Pharma LP, 218 F.R.D. 577, 588 (S.D. Ohio 2003). 7 Id. at 590. 8 Meier, supra. 9 Keefe, supra. 10 Patrick Radden Keefe, Empire of Pain 563 (2021). 11 Keefe, supra (“They duped the F.D.A., saying it lasted twelve hours.”). 12 Freeman Engstrom & Radin, supra, at 337.

Devastatingly early draft – please do not cite or share

which may be well deserved, but some of which may obscure manufacturer misrepresentations, regulator politics, or other aspects of regulatory decision-making that could limit its relevance to a particular question. When new products pose a serious threat to public health or safety, they might well be in highly regulated industries. And manufacturers in those industries will surely use compliance with regulations however they can do avoid liability and bad publicity. It matters, then, how courts use and permit parties to use regulatory approval and compliance. This short article examines one small piece of that puzzle: in a products liability case where the plaintiff’s claim is not preempted by regulation, and where the governing state law provides neither a regulatory compliance defense nor a presumption in favor of a defendant who complied with regulatory standards, should evidence of compliance be admissible to show absence of a design defect or adequate warning? The longstanding, traditional answer to the question is “yes,”13 and the Third Restatement agrees: compliance with a safety statute or regulation is generally admissible and probative, but not dispositive, on the question of design defect.14 But late last year, the Pennsylvania Supreme Court went the other way, broadly declaring compliance evidence inadmissible in products liability cases.15 And while the state is in a decided minority, it is not the only one to do so.16 Cases that address the issue often fail to explain satisfactorily the way in which compliance evidence may be relevant to the question of design defect. I suggest that compliance evidence may be broadly relevant in three ways: first, as evidence of the reasonableness of the defendant’s conduct (not typically at issue in a “strict” products liability case), second, as expert-like evidence on the issue of whether the product is unreasonably dangerous, and third, as indirect evidence of a specific fact underlying the agency’s decision. The second possibility, which I think is the best reading of the dominant rationale for admitting the evidence, raises several problems. I suggest that under that theory of relevance, the admissibility of compliance should depend on the answer to two questions. The first is a question of state policy: is the ultimate determination of design defect a question where the jury should necessarily exercise its own judgment, as opposed to deferring to someone else’s weighting? The second is a question of evidence: has the proponent given the jury sufficient ability to ascertain the probative value of the compliance evidence? Some states, then, may reasonably wish to exclude compliance evidence entirely. Others should simply take care when admitting it. II.

Opting Against Uniform Deference

Commentators have vigorously and rigorously debated whether state tort law should adopt regulatory standards as its own, either via a regulatory compliance defense, a presumption in favor of a compliant defendant, or by other means. Several powerful arguments counsel in favor of deference to regulatory standards. First and foremost, government agencies overwhelmingly benefit from significantly greater expertise, time, and resources than a jury has in an individual trial. Therefore, when an agency decides essentially the same question as the jury would be asked, the agency has greater institutional competence than 13 Teresa Moran Schwartz, The Role of Federal Safety Regulations in Products Liability Actions, 41 Vand. L. Rev. 1121,

1136 (1988); 1 Owen & Davis on Prod. Liab. § 2:25 (4th ed.); 63A Am. Jur. 2d Products Liability § 899-903. 14 Restatement (Third) of Torts: Prod. Liab. § 4(b) (1998). Failure to comply is, however, dispositive of defect under the Restatement. Id. § 4(a). 15 Sullivan v. Werner Co., 306 A.3d 846 (Pa. 2023). 16 See Malcolm v. Evenflo Co., 217 P.3d 514 (Mont. 2009).

Devastatingly early draft – please do not cite or share

juries, and courts should defer to its determination.17 This argument is pressed with particular strength with respect to agencies like the FDA that have a rigorous, product-specific process that draws on scientific expertise, demands significant evidence, and prioritizes public safety, particularly where the agency sets an optimal, as opposed to minimal, standard.18 Allowing juries to deviate upward from these standards could lead to over-deterrence and inhibit the development of valuable products, including drugs,19 in part because juries might be swayed by the plights of individual plaintiffs.20 Further, in some areas, clear and uniform standards are valuable to producers and consumers alike.21 For example, the California Supreme Court, citing the FDA, emphasized the need for “uniformity in presentation and clarity of message” in nonprescription drug warnings.22 Other arguments counsel in favor of allowing jurors to impose liability despite regulatory compliance. Many of these rely on anticipated or observed problems with agency standard-setting. For one, producers have a much higher per-capita stake in obtaining favorable regulations than consumers do, and they therefore may push agencies to underregulate.23 In the extreme, industries may capture regulatory agencies. In addition to vulnerability to industry influence, agencies may also be politicized, too dependent on manufacturer-provided data, too slow to respond to new information, or insufficiently resourced.24 Juror findings of liability can partially compensate for agency under-regulation, pushing producers to make safer products. And they can allow injured plaintiffs to obtain compensation for their injuries.25 There are several ways to implement deference to agencies. The first, federal preemption, takes the case away from the tort system entirely. Preemption is not a matter of state tort law adopting regulatory standards but rather a Congressional decision (at least in theory) to, within a domain, regulate exclusively via federal law, to the exclusion of the police powers of the states.26 Preemption may be express in the language of a statute or implied by the scheme it creates; particularly in the case of “implied preemption,” courts need to interpret the applicable statute to determine whether it preempts state tort suits. In cases where the Supreme Court has deemed 17 See Carl Tobias, FDA Regulatory Compliance Reconsidered, 93 Cornell L. Rev. 1003, 1033 (2008) (noting these

benefits in the context of the FDA); Wilson v. Piper Aircraft Corp., 577 P.2d 1322, 1334-35 (Or. 1978) (Linde, J., concurring) (Federal Aviation Administration approval should preclude contrary decisions absent a difference in standards or deficient regulator performance); Jeff Webb, Note, Tire Design Defect Claims: Why Compliance with the Recently Promulgated Federal Motor Vehicle Safety Standard 139 Should Constitute a Defense as a Matter of Law, 37 U. Tol. L. Rev. 551, 568 (2006). 18 See Rabin, supra, at 2074; Lars Noah, Rewarding Regulatory Compliance: The Pursuit of Symmetry in Products Liability, 88 Geo. L.J. 2147 (2000); Michael D. Green, Prescription Drugs, Alternative Designs, and the Restatement (Third): Preliminary Reflections, 30 Seton Hall L. Rev. 207 (1999). 19 Tobias, supra, at 1029 (citing Perez v. Wyeth Labs. Inc., 734 A.2d 1245, 1259 (N.J. 1999)). 20 Edward Correia, The Relevance of FDA Regulation in Medical Device Product Defect Cases, 24 DePaul J. Health Care L. 1, 4 (2023). 21 I note uniformity may be less important when aggregate litigation, including class actions and multi-district litigation, is the primary vehicle for suing the company. Fewer suits means fewer decisions, even for a widely used product. 22 Ramirez v. Plough, Inc., 863 P.2d 167, 177 (Cal. 1993) (quoting 50 Fed.Reg. 51400, 51402 (Dec. 17, 1985)). 23 Neil Komesar, The Logic of the Law and the Essence of Economics: Reflections on Forty Years in the Wilderness, 2013 Wis. L. Rev. 265, 320 (2013) 24 Tobias, supra, at 1021; Schwartz, supra, at 1153-60. Other political, structural, and statutory reasons may cause agencies to err on the side of over-protection. See David E. Bernstein, What to Do About Federal Agency Science: Some Doubts About Regulatory Daubert, 22 Geo. Mason L. Rev. 549, 560 (2015) (“[agencies] are often explicitly required by statute to utilize a conservative standard for assessing that risk”). 25 Tobias, supra, at 1031-32. 26 See Cipollone v. Liggett Grp., Inc., 505 U.S. 504, 516 (1992).

Devastatingly early draft – please do not cite or share

products liability claims preempted, it has often emphasized the regulatory role of tort law and how it can set standards different from those in a federally mandated regulatory scheme.27 States—either legislatures or courts—may choose to adopt a regulatory compliance defense that allows manufacturer defendants to escape liability if their product complied with regulatory standards. In a sense, then, the state adopts the regulatory standard as the standard for nondefectiveness in its products liability law. The Third Restatement suggests this possibility when the regulation is recent, the standard addresses the exact issue presented, and when the standard was established through a process that was “full, fair, and thorough and reflected substantial expertise.”28 Although commentators have frequently advocated for a regulatory compliance defense,29 few states have opted for one.30 More states have instead opted for a rebuttable presumption that a product compliant with applicable federal safety standards is not defective. These laws are typically statutory,31 and they reflect a substantive policy decision to defer to regulatory standards (unless there is a good reason not to).32 Some statutes do not specify what evidence will rebut the presumption, while others require proof either that the regulation is substantively inadequate or that the manufacturer withheld or misrepresented information to the relevant agency.33 There is much to recommend the rebuttable presumption.34 At its best, it creates a products liability scheme that relies on agency expertise where it applies but rejects it where the agency’s process was compromised, or where at the applicable time, its decision was demonstrably wrong. If Congress, courts, or a state legislature opts for one of the above options, regulatory standards will be at the heart of the case—they might kill the case entirely. But what if they don’t? What role, if any, should evidence of regulatory compliance play? III.

Compliance as Evidence of No Design Defect

27 See Catherine M. Sharkey, Products Liability Preemption: An Institutional Approach, 76 Geo. Wash. L. Rev. 449, 459-

66 (2008) (citing opinions). Sharkey recommends an “agency reference model” of preemption, in which courts look to agencies to supply the data to make an informed decision about whether there should be a uniform federal policy. Id. at 452-53. 28 Restatement (Third) of Torts: Prod. Liab. § 4 cmt. e (1998). 29 See Noah, supra; Webb, supra; Ausness, The Case for a “Strong” Regulatory Compliance Defense, 55 Md. L. Rev. 1210 (1996) 30 Tobias, supra. But see Ramirez v. Plough, Inc., 863 P.2d 167, 176 (Cal. 1993) (“Lacking the procedure and the resources to conduct the relevant inquiries, we conclude that the prudent course is to adopt for tort purposes the existing legislative and administrative standard of care on this issue.”). 31 See, e.g., Wright v. Ford Motor Co., 508 F.3d 263 (5th Cir. 2007) (citing Tex. Civ. Prac. & Rem. Code Ann. § 82.008); Schultz v. Ford Motor Co., 857 N.E.2d 977 (Ind. 2006) (citing Ind. Code § 34–20–5–1 (1998)); Egbert v. Nissan N. Am., Inc., 2007 UT 64, 167 P.3d 1058 (citing Utah Code § 78–15–6(3)). 32 But see Tuggle v. Raymond Corp., 868 S.W.2d 621, 625 (Tenn. Ct. App. 1992) (characterizing the rebuttable presumption as a “refuge” for manufacturers acting in “good faith”). 33 E.g., Tex. Civ. Prac. & Rem. Code Ann. § 82.008(b). Michigan’s provision saying the presumption did not apply if the manufacturer intentionally withheld or misrepresented information to the FDA was at issue in Warner-Lambert Co., LLC v. Kent, 552 U.S. 440 (2008). The state has since eliminated the exception from its statute. See 2023 Mich. Legis. Serv. P.A. 285 (S.B. 410); Mich. Comp. Laws Ann. § 600.2946. 34 See generally Christopher Scott D'Angelo, Effect of Compliance or Noncompliance with Applicable Governmental Product Safety Regulations on a Determination of Product Defect, 36 S. Tex. L. Rev. 453 (1995).

Devastatingly early draft – please do not cite or share

If we have reached this point, then the state has rejected a policy of deference to regulatory standards. Federal preemption has not swooped in to divest the state of private regulation in this area, and the state has neither borrowed its products liability standard from federal regulations nor determined that its existing product defect standard is best analyzed via a regulatory process. It has not embraced uniformity as a goal. Perhaps it has decided to accept the regulatory standard as a minimum, holding that products out-of-compliance with regulatory standards are per se defective, as the Third Restatement suggests.35 But it does not defer to regulatory approval. Instead, it has decided that juries will determine whether the defendant’s product is defective under the state’s products liability standard. Assuming that is all the case—that the state has tasked the jury with applying its standard—the overarching evidentiary question is: how best to enable the jury to make the decision with which it is tasked? The relevant decision is whether the product has a design defect under state law. Products liability generally provides that a product is defective in design if its design renders it “unreasonably dangerous”36 or “not reasonably safe.”37 The state may use one or more of several tests for this determination. Many states employ a “risk-utility” test, in which the factfinder balances the product’s risks against the costs of improving product safety.38 The determination is often guided by factors enumerated in a classic article by John Wade: “(1) the usefulness and desirability of the product; (2) probability and magnitude of potential injury; (3) availability of substitutes; (4) manufacturer’s ability to eliminate the unsafe character; (5) the user’s ability to avoid danger; [and] (6) the user’s probable awareness of the danger . . . .”39 The test has been analogized to the Hand Formula, and the riskutility inquiry is sometimes described as a type of negligence inquiry.40 The Third Restatement’s formulation of this inquiry requires the plaintiff to show that the harm could have been reduced by a “reasonable alternative design.”41 In addition or alternative to the “risk-utility” test, states may use a “consumer expectations” test, where a product is defective if it is “in a condition not contemplated by the ultimate consumer, which will be unreasonably dangerous to him.”42 Some states have a combined test, where a plaintiff may choose which theory to invoke.43 And some states may give the jury free rein to decide whether a product design poses “unreasonable danger.”44 So, does evidence of compliance help a jury decide whether the defendant’s product is “not reasonably safe”? Under the Federal Rules of Evidence, which are mirrored in substantial part in 35 Restatement (Third) of Torts: Prod. Liab. § 4(a) (1998). 36 See Restatement (Second) of Torts § 402A (1965).

37 Restatement (Third) of Torts: Prod. Liab. § 2 (1998).

38 See Dan B. Dobbs, Paul T. Hayden and Ellen M. Bublick, The Law of Torts § 456 (2d ed.).

39 Id. (citing John W. Wade, On the Nature of Strict Tort Liability for Products, 44 Miss. L. Rev. 825, 837 (1973)). The

article also lists “(7) the manufacturer’s ability to spread the loss,” but this is less often used in negligence determinations. Id. California has developed a similar set of factors in Barker v. Lull Engineering Co., 573 P.2d 443 (Cal. 1978): “the gravity of the danger posed by the challenged design, the likelihood that such danger would occur, the mechanical feasibility of a safer alternative design, the financial cost of an improved design, and the adverse consequences to the product and to the consumer that would result from an alternative design.” 40 See Restatement (Third) of Torts: Prod. Liab. § 2 cmt. d (1998). 41 Restatement (Third) of Torts: Prod. Liab. § 2 (1998). 42 Restatement (Second) of Torts § 402A cmt. g (1965). 43 See, e.g., Falk v. Keene Corp., 782 P.2d 974 (Wash. 1989) (citing RCW 7.72.030). 44 See, e.g., Nesselrode v. Exec. Beechcraft, Inc., 707 S.W.2d 371, 378 (Mo. 1986) (“Under our model of strict tort liability the concept of unreasonable danger, which is determinative of whether a product is defective in a design case, is presented to the jury as an ultimate issue without further definition.”) (noted as current in Hopfer v. Neenah Foundry Co., 477 S.W.3d 116, 128-29 (Mo. Ct. App. 2015)).

Devastatingly early draft – please do not cite or share

most states, if evidence is relevant, it is admissible unless excluded by some other law,45 including Rule 403, which allows a court to exclude relevant evidence if its harms—including unfair prejudice, confusing the issues, and misleading the jury—substantially outweigh its probative value.46 The first step to determining admissibility, then, is to look at the relevance of compliance evidence. Does it make a fact of consequence more or less probable?47 For many people, the intuitive answer of “yes” is so obvious that it hardly seems worth investigating. But in light of court decisions finding this evidence inadmissible, it’s worth attempting to parse out the paths by which compliance might prove lack of defect and then examining potential problems with those uses. A. Evidence of Conduct First, although it is not directly pertinent to a “strict” products liability analysis, courts and commentators largely agree that compliance evidence is relevant to the reasonableness and good faith of the defendant’s conduct.48 I see a few ways this can be so. First, if the manufacturer could have chosen not to comply, the act of complying with federal regulatory requirements is itself a reasonable one and tends to show the defendant acted with care. They took at least one appropriate step toward making a safe product. (And even if compliance was required, if the manufacturer had the option to skimp and instead complied meticulously, that shows care.49) A jury could decide that this is somewhat probative of due care or decide that compliance with regulatory requirements is, itself, due care.50 Second, the fact that the manufacturer took one careful step tends to show that they were approaching the process of product design and production carefully and are more likely to have taken other careful steps.51 Third, the regulatory standard provides a credible signal to the manufacturer that its product is sufficiently safe. Because this signal provides a reason to believe that the product is adequate—or at least in the ballpark of adequacy—a reasonable manufacturer might take fewer additional precautions than it would without this signal. This might explain to the jury why a manufacturer did not conduct other tests or add more safety features: it had reason to believe its product passed muster.52 To the extent that state law atypically53 depends on the reasonableness of the defendant’s conduct—as opposed to the reasonableness of the risks presented by the product—evidence of purposeful compliance is almost certainly relevant. And this would be so even if the regulatory standard were seriously flawed—the effort to comply and the signal from the seal of regulatory

45 Fed. R. Evid. 402. If states have statutory authority making this evidence admissible, see, e.g., Ark. Code Ann. § 16-

116-205, the usual evidentiary rules will not keep it out. But analysis of those rules is still helpful to consider how states can best structure their laws. 46 Fed. R. Evid. 403. 47 Fed. R. Evid. 401. 48 See Correia, supra, at 7. 49 See id. 50 Cf. Tuggle v. Raymond Corp., 868 S.W.2d 621, 625 (Tenn. Ct. App. 1992) (describing Tennessee’s rebuttable presumption as a “refuge” for compliant manufacturers). 51 I do not believe this raises character evidence issues, as it is limited to the defendant’s approach to this process. Using evidence of compliance with respect to an unrelated product would probably rely on impermissible character inferences. See Fed. R. Evid. 404. 52 See Malcolm v. Evenflo Co., 217 P.3d 514 (Mont. 2009). 53 The negligence-like principles used for risk-utility analysis in “strict” product liability speak to “the condition of the product,” not “the reasonableness of the manufacturer’s conduct.”

Devastatingly early draft – please do not cite or share

approval do not depend on the standard actually reflecting reasonable safety.54 This use of compliance evidence more frequently arises in the context of punitive damages, where the plaintiff must show some bad faith, or similar, on the part of the defendant.55 Compliance tends to negate bad faith, and it should be admissible for that purpose.56 Some who have argued against the use of compliance evidence in products liability cases suggest it is only relevant to the manufacturer’s level of care and does not help a jury evaluate the safety of the product itself.57 But more courts indicate that compliance evidence can also illuminate whether the product itself is defective. How might that be? B. Compliance as Pseudo Expert Opinion Evidence Compliance evidence ultimately takes the form of a conclusion: the defendant complied with regulatory standards. Either the defendant’s product was approved by regulators, or the regulator set standards for products in that category, and the defendant’s product satisfied them. This means, given the information the regulatory agency had, and given the process it used to make its decision, the regulatory agency concluded that the defendant’s product—or products that share its features— satisfies the standard under which the agency makes safety decisions. The most straightforward way in which this determination is relevant to the jurors’ decision is that it gives something like a neutral, qualified expert opinion on an issue that is very close to the ultimate issue that the jury must decide: is this product’s design unreasonably dangerous? Although not typically framed as “expert opinion evidence,” this is, as I understand it, the dominant rationale for admitting this evidence: an agency’s approval indicates that a weighing of risks and benefits works out in favor of the product, and the product is, in fact, not defective.58 However, I submit that the proximity of this evidence to expert opinion testimony raises several potential concerns. First, as the majority in the recent Pennsylvania decision, Sullivan v. Werner Co.,59 appeared to suggest, if the state commits to the jury the task of weighing risks and utility, a bare ultimate

54 The probative value of this would be undermined—or at least countered—by proof that the regulatory standard

resulted from industry capture or, worse, misrepresentation by the defendant. But that proof could be introduced separately. 55 See Malcolm, 217 P.3d at 530 (citing Montana statutory law for the requirement of “actual fraud or actual malice”). 56 If specific circumstances undermine the probative value of compliance, the plaintiff may introduce them. See Gryc v. Dayton-Hudson Corp., 297 N.W.2d 727, 734 (Minn. 1980) (reasoning that compliance did not preclude punitive damages, in part because the defendant knew the regulatory standard was inadequate). In addition, if compliance evidence is deemed irrelevant to liability but relevant to punitive damages, bifurcation may be advisable. See Malcolm, 217 P.3d at 533 (McGrath, C.J., concurring). 57 See Malcolm, 217 P.3d at 521-22; Sullivan v. Werner Co., 306 A.3d 846, 856-58 (Pa. 2023) (reviewing the plaintiff’s argument); Rucker v. Norfolk & W. Ry. Co., 396 N.E.2d 534, 538 (Ill. 1979) (Goldenhersh, C.J., dissenting). 58 See, e.g., Wagner v. Clark Equip. Co., 700 A.2d 38, 50 (Conn. 1997) (“The permissible inference to be drawn from this evidence is that a product that meets the minimum safety requirements contained in the OSHA regulation is not defective.”); Rucker v. Norfolk & W. Ry. Co., 396 N.E.2d 534, 537 (Ill. 1979) (“If the product is in compliance with Federal standards, the finder of fact may well conclude that the product is not defective, thus ending the inquiry into strict liability. If a finding is entered that the product is defective . . . [t]he fact of compliance may indicate to the finder of fact that the defect is not unreasonably dangerous.”); Appellant’s Brief at 31, Sullivan v. Werner Co., 306 A.3d 846 (Pa. 2023) (18 EAP 2022) (citing Rucker); Correia, supra, at 13 (“A determination by experts at the FDA that a product is safe clearly meets this test [of relevance] if the question is whether a product is, in fact, safe.”). 59 306 A.3d 846 (Pa. 2023).

Devastatingly early draft – please do not cite or share

conclusion does not assist them with that process.60 The conclusion, in itself, does not illuminate particular attributes of the product,61 including risks—their existence, probability, or severity— utility, or infeasibility of alternative designs. The implication, here, is that evidence is helpful to the jury if it presents concrete, underlying facts, and it is unhelpful if it simply tells them how to come out. This mirrors an issue that arises frequently in the context of expert testimony: the helpfulness of witness opinions on ultimate issues. Federal Rule of Evidence 704 and its state counterparts eliminate an older rule prohibiting witnesses from testifying to their opinion on an issue that the jury will need to decide. However, Rule 704 does not create a basis for admitting testimony that simply tells jurors what decision to reach or those that are phrased in terms of “inadequately explored legal criteria”—such opinions are considered unhelpful to the trier of fact.62 In design defect cases, the line between a helpful expert opinion on an ultimate issue and an impermissible one can be fine. Often, courts have excluded expert testimony that a product is “unreasonably dangerous” or “defective.”63 But expert opinions that do not contain terms with legal significance are often admitted,64 so an expert might testify that “in his opinion, the risks outweighed the benefits” without running afoul of the “helpfulness” requirement.65 Further, at least one federal appellate court was untroubled by testimony that a product was “unreasonably dangerous beyond the expectation of the average user” when the expert explained the basis for his opinion in sufficient detail for the jury to independently evaluate its dangerousness.66 So, in the context of experts, courts may permit opinion testimony that does not simply go to the factors that underly risk-utility analysis (or, if applicable, allow a jury to evaluate consumer expectations). A state might choose to take a stricter view. It might decide that if a jury is tasked with weighing facts to make a legal judgment, evidence is appropriate only as to those facts, and the task of weighing is completely within the province of the jury. Any testimony as to how a jury should weigh the evidence—or any conclusory evidence that does not separate its factual basis from its weighing—would be unhelpful. There is some appeal to this idea. If a state has concluded that the lay judgment of the jury—the jury’s normative view on when a product’s risks outweigh its utility— is an important aspect of a products liability case, in the same way that jurors bring lay, normative judgment to a regular negligence claim,67 then it may be appropriate to exclude expert opinions on the ultimate balance. In this way, evidentiary determinations can effect a policy decision of the state by concluding that an external opinion on how a jury should balance risks and benefits is unhelpful to their task. 60 Id. at 861-62 (“Compliance evidence does not prove any characteristic of the product; rather, it diverts attention from

the product’s attributes to both the manufacturer’s conduct and whether a standards-issuing organization would consider the product to be free from defects. Neither of these considerations are pertinent to a risk-utility analysis.”). 61 Id. 62 Fed. R. Evid. 704 advisory committee notes (citing Fed. R. Evid. 701, 702, 403). 63 See, e.g., Walker v. Conagra Brands, Inc., No. 820CV01679FWSKES, 2023 WL 8885148 (C.D. Cal. Sept. 21, 2023); In re 3M Combat Arms Earplug Prod. Liab. Litig., No. 3:19MD2885, 2021 WL 765019 (N.D. Fla. Feb. 28, 2021); In re C. R. Bard, Inc., Pelvic Repair Sys. Prod. Liab. Litig., No. MDL 2187, 2018 WL 4212409 (S.D.W. Va. Sept. 4, 2018); Strong v. E. I. DuPont de Nemours Co., 667 F.2d 682 (8th Cir. 1981). 64 See In re C.R. Bard, Inc., 948 F. Supp. 2d 589 (S.D.W. Va. 2013). 65 Dalbotten v. C. R. Bard, Inc., No. 1:20-CV-00034-SPW, 2023 WL 1068771 (D. Mont. Jan. 27, 2023). 66 See Karns v. Emerson Elec. Co., 817 F.2d 1452 (10th Cir. 1987). 67 See Mark P. Gergen, The Jury’s Role in Deciding Normative Issues in the American Common Law, 68 Fordham L. Rev. 407, 424-25 (1999).

Devastatingly early draft – please do not cite or share

Or it might conclude, in line with the Pennsylvania Supreme Court’s suggestion, that the proper role of evidence is not to increase the probability that a jury will come to the “correct” answer by whatever means necessary, but rather to help the jury come to the answer the right way. To the extent a reliable-but-naked expert opinion points jurors in the correct direction but does not aid the reasoning process that they have been instructed to undertake (other than by prompting motivated reasoning to come to a preordained conclusion), it could be deemed unhelpful to the jury. Assuming the state permits expert testimony that does not simply elucidate strengths and flaws of the product but in fact goes to balancing, should it admit compliance evidence? Although regulatory standards will likely be presented by an expert witness, the standards themselves are, of course, not an expert witness. Yet, they carry many of the same benefits and risks—as well as some additional ones. As for benefits, both experts and agencies come with extensive, relevant expertise, which they apply to the facts of the case—or facts closely related, in the case of generally applicable regulatory standards. Agencies also have the benefit of not having been hired by one of the parties for purposes of the litigation. And, depending on the agency and the issue, they may have put significantly more time and resources into the decision. As for risks, both experts and agencies obscure at least part of their decision-making from the factfinder. Experts may rely on inadmissible facts or data, and they do not necessarily need to disclose the basis for their opinion.68 When evidence of compliance is admitted, it does not necessarily come with a full record of the agency’s decision-making.69 The evidence may be little more than the regulatory standard itself.70 And for both experts and agencies, juries face a problem of epistemic incompetence: evaluation of the problem is sufficiently difficult that juries may have trouble choosing whom to believe and may resolve the problem based on signals other than the quality of the testimony or evidence.71 They may defer, even where they are capable of making their own judgment. These problems seem to me to be heightened in the context of compliance evidence. For one, the agency appears neutral and has the imprimatur of the government, and I would expect juries to defer to its decisions readily.72 Indeed, in Sullivan itself, the deliberating jury asked whether OSHA inspects every product put on the market.73 The juror may well have almost no understanding of the basis for the agency’s decision, whereas experts typically discuss the basis for their opinions and may be required to do so on cross-examination.74 The bulk of the expert’s testimony will likely consist of the risks and utility of the product or alternatives, with the ultimateissue-type determination somewhat summative of the rest of the testimony. The expert has been qualified in an area, and their methodology has been evaluated and deemed reliable or generally 68 Fed. R. Evid. 703, 705.

69 See, e.g., Sullivan v. Werner Co., 306 A.3d 846, 851 (2023) (noting that the defendant’s expert “submitted a report

suggesting the scaffold met federal Occupational Safety and Health Administration (OSHA) regulations”). 70 See, e.g., Wagner v. Clark Equip. Co., 700 A.2d 38, 48 (Conn. 1997). 71 See David H. Kaye. et al., The New Wigmore: A Treatise on Evidence: Expert Evidence § 1.4.2 (3d ed. 2021). 72 See Wagner v. Clark Equip. Co., 700 A.2d 38, 51 (Conn. 1997) (“compliance with a federal regulation may carry more weight with a jury than compliance with an industry standard, because a federal regulation has the imprimatur of the federal government”). 73 Sullivan v. Werner Co., 306 A.3d 846, 852 (Pa. 2023). 74 See Fed. R. Evid. 705.

Devastatingly early draft – please do not cite or share

accepted.75 Further, although “litigation science” may frequently be biased, it is at least designed to answer the specific question in the litigation, whereas the agency’s determination may answer only an adjacent question about product safety. And unless someone with knowledge of the agency’s specific process is on the stand, the plaintiff’s lawyers cannot cross-examine a witness to uncover bias or improper influence or flaws in the agency’s factual basis or its methods. These issues could leave the jury fundamentally unable to evaluate the probative value and reliability of the standards. For example, even if the agency’s resources and mandate make its evaluation of scientific facts reliable and worthy of jury deference, its ultimate weighing of those facts to produce decisions inevitably include political considerations that are appropriate in agency decision-making but likely should not factor into a jury’s decision under the products liability standard.76 If the jury is unable to examine the agency decision, it will get the factual, normative, and political all lumped into one, with little hope of disaggregating them. If the court-approved theory of relevance is this one—the compliance evidence is admitted as a sort of expert opinion to which the jury can choose to defer or agree—then I suggest that courts require additional steps to improve the quality of the evidence. First, for the opinion to be of use to the jury, the proponent should introduce sufficient evidence of the process by which the agency made its decision for the jury to evaluate that decision’s probative value. This would include, at a minimum, the standard under which the agency made its determination. Was the agency performing essentially the same inquiry as the jury, engaging in a risk-utility analysis? Or was it setting a minimum standard, below that of the applicable products-liability regime? Was it evaluating product safety for the purposes of consumers, or was it setting a different standard—potentially higher, potentially lower—for the workplace?77 The agency rule may not be useless if it was formulated under a different standard, but the jury should at least know what the opinion means. Further, enunciating the standard, as opposed to providing evidence that the product “complied with the standards” or that the agency “determined it was reasonably safe,” minimizes the problem of the evidence simply telling the jury what conclusion to reach, like an unhelpful expert opinion. The proponent should also introduce information—at some level of generality—about the process itself: what evidence the agency considered and how it obtained that evidence, the identities of the decisionmakers, and so on. And, if it is available, the proponent should provide the jury with information about the agency’s intermediate conclusions, separating out the agency’s scientific determinations from its policy decisions.78 This splitting, which would give the jury the benefit of an agency’s scientific expertise while allowing it to choose whether or not to give weight to its policyinfused conclusion, would be ideal for enabling rational jury decision-making. (Although if all of the scientific results were independently available, the agency’s ultimate decision might have little value). The plaintiff could then introduce evidence casting doubt on the value of the agency’s process or make arguments for why, given the elicited information, the jury should not give the evidence much weight.

75 See Fed. R. Evid. 702; Frye.

76 See Noah, supra, at 2153 (arguing for a regulatory compliance defense despite political considerations that go into the

decision). 77 This question, unsurprisingly, arises when defendants seek to introduce evidence of compliance with OSHA regulations. See 63A Am. Jur. 2d Products Liability § 900. 78 See Bernstein, supra, at 564-65.

Devastatingly early draft – please do not cite or share

I hesitate, however, to treat agency determinations like expert evidence in the sense of requiring a Daubert-type hearing.79 Although some scholars have suggested this in the context of challenges to agency action,80 the idea has not gained much traction.81 Agencies have little incentive to rely on the junk science that Daubert was intended to exclude, and judicial gatekeeping may add significant cost with little benefit to the reliability of the proceeding. If the proponent of the compliance evidence cannot produce sufficient agency-process evidence to enable the jury to use it rationally, the court should exclude the evidence. The signal is too hard to interpret, and the temptation for deference is too great. C. Inferring Underlying Facts This “expert opinion on an ultimate issue” idea may not be the only way compliance evidence can shed light on product defect. Some courts suggest that this evidence may indeed assist the jurors with their balancing task by illuminating Wade factors or how the jury should balance them.82 These might include facts such as the probability of injury given the manufacturer’s design or the probable severity of the injury, the feasibility of an alternative design, or the safety benefits of the manufacturer’s design. Certainly, studies from the administrative record could directly shed light on these questions. This sort of data, introduced through an appropriate expert, could be exceedingly valuable to a jury. And if the administrative record, or testimony from an agency representative, concerns a conclusion on some underlying fact of interest, that, too, could be helpful to a jury’s analysis, as supported expert evidence. But more interestingly, if the defendant demonstrates to the jury that the agency considered a factual question,83 its ultimate decision on appropriate standards or approval of a product may provide evidence of how it resolved that question. For example, say the plaintiff’s expert has proposed an alternate design, and there is a dispute about whether the cost of the alternate design is high or low. If the agency considered that design but decided not to require it, that provides some evidence—although a noisy signal—that it concluded the cost is high. And if an informed, neutral entity concluded the cost is high, that, in turn, provides evidence that the cost is, indeed high. Importantly, for standards evidence to be relevant for this purpose, the proponent would need to submit evidence of the agency’s underlying data and process.84 Alternatively, if it is clear an alternative design would have been preferable, but the parties dispute whether it was available at the time, the fact that the agency did not require it would speak to that question.

79 Or Frye-type hearing, if the state employs that standard.

80 E.g., Paul S. Miller & Bert W. Rein, “Gatekeeping” Agency Reliance on Scientific and Technical Materials After

Daubert: Ensuring Relevance and Reliability in the Administrative Process, 17 Touro L. Rev. 297 (2000). 81 David E. Bernstein, What to Do About Federal Agency Science: Some Doubts About Regulatory Daubert, 22 Geo. Mason L. Rev. 549, 552 (2015). 82 Compare Kim v. Toyota Motor Corp., 424 P.3d 290, 298 (Cal. 2018) (“Depending on the circumstances, evidence of other manufacturers’ design decisions may aid the jury’s understanding of these complexities and trade-offs, and thus may provide some assistance in determining whether the manufacturer has balanced the relevant considerations correctly.”). 83 See Fed. R. Evid. 104(b). 84 Kim v. Toyota Motor Corp., 424 P.3d 290, 304 (Cal. 2018) (Cato, J. concurring) (“It is the basis for the design decision that the jury must analyze. To admit the conclusion (the industry custom-and-practice evidence) before the foundational evidence establishing that the appropriate risks and benefits were balanced would be pure speculation.”).

Devastatingly early draft – please do not cite or share

Again, one danger in this evidence is the noise surrounding the signal. In some instances, even if the agency did not create a record of its scientific conclusions, they may be inferable from the ultimate decision. But I would imagine that in most cases, the decision would rest on so many factors that it would be difficult for the jury to gain significant insight into particular questions. Again, if the evidence is sufficiently unreliable and confusing, it should be excluded. D. Other Reasons Not to Exclude Entirely Although plaintiffs have, unsurprisingly, generally advocated excluding evidence of compliance to show lack of design defect, there are several reasons why a process that permits this evidence, with proper foundation, might ultimately benefit plaintiffs. For one, the process of examining the origin of regulatory standards and the process of regulatory decision-making may uncover problems in either the agency’s decision-making or the conduct of the defendant or the defendant’s industry, such as fraud on the agency. This could support the educative or “catalytic” function of tort litigation: litigation can sometimes reveal problems that spur public concern or corrective government action.85 For another, if courts exclude this evidence entirely, manufacturers might pressure other state actors to take a more heavy-handed approach. A state legislature might enact a statutory rebuttable presumption that a defendant who has complied with regulations is not liable. Less plausibly, Congress might amend a regulatory statute to preempt other claims. In any event, a rule that allows optional jury deference when the jury understands what it is deferring to might be the most plaintiff-friendly one that is politically palatable. IV.

Conclusion

Industrywide misconduct has caused massive suffering and death in the cases of the key examples in this symposium: opioids, asbestos, and especially tobacco. But improper industry influence can lead to smaller regulatory disasters.86 In 1953, a federal statute—not an administrative regulation—specified a test to determine whether fabric was “so highly flammable . . . as to be dangerous when worn by individuals,”87 and clothing that passed the test could be sold in interstate commerce. In 1969, 4-year-old Lee Ann Gryc was wearing compliant pajamas when she reached across an electric stove, her pajamas ignited instantly, and she suffered severe burns.88 In her family’s suit against the manufacturer, evidence was presented that “the test was adopted as a result of industry influence and, therefore, served to protect the textile industry rather than the public.”89 The test was, in fact, wholly inadequate: “newspaper passed the CS 191-53 test with a 48-percent margin of safety.”90 The manufacturer introduced evidence that its pajamas complied with the act to defend against both liability and punitive damages.91 And yet, with appropriate information about the defendant’s knowledge and with expert testimony, the jury awarded the plaintiffs $750,000 in

85 See Rabin, supra, at 2068-69; Freeman Engstrom & Rabin, supra, at 292.

86 In 1980, 311 Americans died in fires where clothing was the first item ignited.

https://www.federalregister.gov/documents/2008/03/25/E8-5569/standard-for-the-flammability-of-clothing-textiles.

87 Flammable Fabrics Act of 1953, Pub. L. No. 83-88, § 4(a), 67 Stat. 111, 112 (1953). 88 Gryc v. Dayton-Hudson Corp., 297 N.W.2d 727, 729 (Minn. 1980). 89 Id. at 734. 90 Id.

91 Id. at 730.

Devastatingly early draft – please do not cite or share

compensatory damages and $1,000,000 in punitive damages.92 This classic case of industry influence shows how naked compliance evidence can obscure industry-caused problems, but it also shows that a sufficiently well-informed jury may not simply defer to federal standards.

92 Id. at 729.

SESSION III: ENABLERS Discovery Gamesmanship in Mass Torts Seth Endo, Seattle University New Empirics on Litigation Secrecy (Virtual Presentation) David Freeman Engstrom, Stanford University* Toxic Torts and the Shadowland of Responsibility John Goldberg, Harvard University Benjamin Zipursky, Fordham University Effects of Sponsorship on Research Results Michael Saks, Arizona State University Discussant: Stephan Landsman * Collaborators concerning the reported upon research include Nora Freeman Engstrom, Stanford University; Jonah Gelbach, University of California, Berkeley; Austin Peters, Stanford University

(Back to main agenda)

THE 30 TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

Draft: Subject to Change Prior to Publication Forthcoming 74 DUKE L.J. __ (2024).

SECRECY BY STIPULATION Nora Freeman Engstrom, David Freeman Engstrom, Jonah B. Gelbach, Austin Peters, & Aaron Schaffer-Neitz † GM Ignition Switch. Dalkon Shield. Oxycontin. For decades, protective orders—court orders that require parties to maintain the confidentiality of information unearthed during discovery—have hid deadly defects and pervasive abuse from the public, perpetuating unnecessary harm. But how worrisome are these protective orders, really? Under Rule 26(c)’s plain language, protective orders are to be granted only upon a showing of “good cause.” Doesn’t that adequately cabin the orders’ entry? Prominent judges and scholars have long insisted it does and that, under Rule 26(c), the day-to-day grant of protective orders is careful, not cavalier. Critics disagree. They charge that parties frequently agree to sidestep Rule 26(c)’s “good cause” requirement and that judges, although formally duty-bound to protect the public interest, uncritically acquiesce to their demands. Worried about judicial rubber-stamping, some, in fact, have spent decades pushing to tighten Rule 26(c)’s standards—while others have, just as vigorously, opposed these efforts, insisting that the status quo works well enough. This debate has raged since the late 1980s. But until now, it’s mostly run aground on the shoals of basic, but unanswered, factual questions: Are stipulated protective orders really de rigeur? Are they becoming more prevalent? †

Nora Freeman Engstrom is the Ernest W. McFarland Professor of Law at Stanford Law School (SLS) and Co-Director of the Deborah L. Rhode Center on the Legal Profession (Rhode Center). David Freeman Engstrom is the LSVF Professor of Law at SLS and Co-Director of the Rhode Center. Jonah B. Gelbach is the Herman F. Selvin Professor of Law at Berkeley Law School and a Non-Resident Fellow at the Rhode Center. Austin Peters is a recent graduate of SLS and holds a Ph.D. in Political Science from Stanford. He is also a Non-Resident Fellow at the Rhode Center. Aaron Schaffer-Neitz is in the SLS Class of 2024 and a Civil Justice Fellow at the Rhode Center. We benefited from astute comments supplied by Stephen B. Burbank, William Eskridge Jr., C. Scott Hemphill, Deborah R. Hensler, Timothy D. Lytton, Larry C. Marshall, Robert L. Rabin, and Judith Resnik. We are grateful to Ari Berman, Devin Flynn, Sreya Guha, Vrushank Gunjur, and Jessica Seigel for excellent research assistance and to SLS, the Rhode Center, and the Arthur & Charlotte Zitrin Foundation for supporting this research. For a copy of the methodological Appendix, please contact Nora Freeman Engstrom at nora.engstrom@law.stanford.edu.

Electronic copy available at: https://ssrn.com/abstract=4811151

And are joint motions for protective orders actually meticulously scrutinized? Using state-of-the-art machine-learning techniques, this Article analyzes an original dataset of over 2.2 million federal cases to answer these persistent and profoundly important questions. Along the way, we find that stipulated protective orders are surprisingly prevalent. Grant rates for stipulated protective orders are sky high. And even though many insist that judges are scrupulous in the entry of such orders, over our entire study period, a majority of federal judges never ever rejected a joint protective order request. We offer the first comprehensive accounting of stipulated protective orders in federal litigation. In so doing, we aim not only to revitalize—and discipline—the perennial and consequential debate surrounding Rule 26(c). We also offer a fortified empirical foundation on which to ground inquiry into broader questions, including the role of transparency and privacy in a system ostensibly committed to “open courts,” tort law’s vital information-forcing function, adversarialism as a procedural cornerstone of American litigation, and trial-court discretion and fidelity to higher law.

Table of Contents INTRODUCTION .....................................................................3 I. THE FORMAL LAW REGARDING PROTECTIVE ORDERS ...... 11 A. Protective Orders 101 ............................................... 11 B. Protective Orders, Generally..................................... 14 C. Stipulated Protective Orders ..................................... 16 II. REFORM HISTORY: PERENNIAL PROPOSALS AND DEFICIENT DATA ................................................................ 22 A. Past Efforts to Limit Expansive Protective Orders..... 22 B. Political and Policy Debates Over Protective Orders . 26 1. Reformers’ Arguments in Favor of Greater Transparency .............................................................. 27 2. Critics’ Arguments Opposing Reform Efforts ........ 32 C. Why Debate Has Stalled: Dueling Anecdotes and Anemic Empirics............................................................ 37 III. DATA: NEW EVIDENCE ON STIPULATED PROTECTIVE ORDERS .............................................................................. 42

2 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

A. A Note About Methodology...................................... 43 B. Motions for Stipulated Protective Orders: Time Trends and Prevalence ............................................................... 47 C. Rule 26(c) Motions: Grants and Denials................... 52 D. Qualitative Assessments ........................................... 56 1. Qualitative Review of 300 Granted POs................. 57 2. Qualitative Review of 100 Denied POs .................. 63 IV. IMPLICATIONS .............................................................. 65 A. Rebooting the Rule 26(c) Debate .............................. 66 B. Stunting the Information-Forcing Function of Tort Law ....................................................................................... 69 C. Rule Fidelity, Judicial Hierarchy, and Judge Discretion ....................................................................................... 71 D. Open Courts, Adversarialism, and Access to Justice . 77 CONCLUSION ...................................................................... 82 INTRODUCTION What if you found out that one judge’s order could have saved the life of your loved one? Thousands of loved ones? Protective orders—court orders issued under the auspices of Rule 26(c), which require parties to maintain the confidentiality of information unearthed during discovery— have made these hypotheticals a grim reality. Consider the 2014 General Motors ignition switch debacle.1 As unusual accidents piled up, injured parties predictably sued. In one case, plaintiffs, the parents of Brooke Melton, a twenty-nine-year-old pediatric nurse who had died when her Chevy Cobalt stalled on a Georgia highway, uncovered smoking-gun evidence: Brooke’s Cobalt had had a defective ignition switch. Further digging revealed that some

In 2014, the ignition switch problem led GM to recall some 2.6 million vehicles, and the defect was eventually implicated in 275 injuries and 124 deaths. See Mike Spector, Jaimi Dowell & Benjamin Lesser, How Secrecy in U.S. Courts Hobbles the Regulators Meant to Protect the Public, REUTERS (Jan. 16, 2020), https://www.reuters.com/investigates/special-report/usa-courtssecrecy-regulators/.

3 Electronic copy available at: https://ssrn.com/abstract=4811151

in GM knew and had long known the switch was defective.2 Indeed, in new Cobalt models, the faulty ignition switch, responsible for Brooke’s accident, had been quietly replaced. Brooke’s parents wanted to share this explosive evidence. “We thought that people needed to know. There were still people out there driving those cars,” Brooke’s mother explained.3 But a 2011 stipulated protective order, entered by a Georgia state court, barred such disclosures.4 While the Meltons were muzzled, more motorists died.5 The Melton case is tragic. But it isn’t anomalous. Protective orders (sometimes called “POs”)—often used in conjunction with secret settlements, non-disclosure agreements (NDAs), attorney lockout provisions, and orders to seal judicial records—ensure that many documents surfaced in litigation never see the light of day.6 And while it can be 2

Spector et al., supra note 1. For more on the GM ignition switch debacle and its aftermath, see generally Nora Freeman Engstrom, When Cars Crash: The Automobile’s Tort Law Legacy, 53 WAKE FOREST L. REV. 293, 328–35 (2018); ANTON R. VALUKAS, REPORT TO BOARD OF DIRECTORS OF GENERAL MOTORS COMPANY REGARDING IGNITION SWITCH RECALLS (2014); LANCE A. COOPER WITH MARK TABB, COBALT COVER-UP: THE INSIDE STORY OF A DEADLY CONSPIRACY AT THE LARGEST CAR MANUFACTURER IN THE WORLD (2020). For the roots of the Meltons’ extraordinary discovery, see Bill Vlasic, An Engineer’s Eureka Moment with a G.M. Flaw, N.Y. TIMES, Mar. 29, 2014, at A1, https://www.nytimes.com/2014/03/29/business/a-floridaengineer-unlocked-the-mystery-of-gms-ignition-flaw.html. 3

Spector et al., supra note 1 (quoting Beth Melton).

Id. The judge who issued the protective order justified her decision on the logic that “[t]he role of litigation is not to regulate GM.” Id. For the fact that the protective order was stipulated, see Consent Protective Order of Confidentiality, Melton v. General Motors, No. 2011-A-2652 (Cobb Cnty. Ct. Dec. 13, 2011), available at https://www.autosafety.org/wpcontent/uploads/import/Melton%20consent%20protective%20order% 20Dec%20%202011.pdf. 5

See Spector et al., supra note 1. According to Kevin Vincent, previously Chief Counsel for NHTSA, NHTSA was “stymied” by the protective order in the Melton litigation. If NHTSA had that evidence, Vincent said, “[w]e could have acted sooner.” Id. Likewise, Lance Cooper, the Meltons’ attorney, believes that if there had not been a PO the defective design would have been more quickly revealed. Telephone Interview with Lance Cooper, Founding Partner, The Cooper Firm (Feb. 29, 2024). 6

For further discussion of orders to seal, which are akin to POs in some respects, see infra notes 41–46 and accompanying text. For a discussion of secret settlements, see David Freeman Engstrom, Nora

4 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

difficult to decouple the effects of POs, as against other secrecy mechanisms (since they often exist in tandem), evidence indicates that overly broad POs have led to the suppression of a wide range of information that, if revealed sooner, could have avoided significant harms, even death.7 The list is long: Oxycontin,8 Zyprexa (a drug used to treat schizophrenia and bipolar disorder),9 Zomax (a prescription painkiller manufactured by McNeil Pharmaceutical, a subsidiary of Johnson & Johnson),10 the Dalkon Shield intrauterine contraceptive device,11 Prempro (an estrogen

Freeman Engstrom, Jonah B. Gelbach, Austin Peters & Garret Wen, Shedding Light on Secret Settlements: An Empirical Study of California’s STAND Act, 91 CHI. L. REV. (forthcoming 2024). For a discussion of attorney lockout (sometimes called “buyout”) provisions, which are formally banned by ABA Model Rule 5.6(b), see Lynn A. Baker, Mass Torts and the Pursuit of Ethical Finality, 85 FORDHAM L. REV. 1943, 1957–65 (2017). 7

While all the harms of litigation secrecy cannot be laid at the feet of POs, reformers see POs as the keystone of a broader secrecy architecture. As one scholar puts it: “Almost all facets of the broader court-confidentiality problem stem from a common root—protective orders.” Dustin B. Benham, Proportionality, Pretrial Confidentiality, and Discovery Sharing, 71 WASH & LEE L. REV. 2181, 2190 (2014). 8

Elizabeth Chamblee Burch & Alexandra D. Lahav, Information for the Common Good in Mass Torts, 70 DEPAUL L. REV. 345, 345–47 (2021) (explaining that a 2015 deposition of Richard Sackler contained damning information concerning Purdue’s activities but that the video of Dr. Sackler’s deposition was kept under wraps for years, by virtue of a stipulated PO). 9

Sunshine in Litigation Act of 2008: Hearings on H.R. 5884 Before the H. Comm. on the Judiciary, Subcomm. on Commercial and Administrative Law, 110th Cong. 51 (2008) (statement of Sen. Herb Kohl); Alexander C. Egilman, Aaron S. Kesselheim, Harlan M. Krumholz, Joseph S. Ross, Jeanie Kim & Amy Kapczynski, Confidentiality Orders and Public Interest in Drug and Medical Device Litigation, 180 JAMA INTERNAL MED. 292, 295 (explaining that information, leaked by a medical expert in violation of a protective order, clarified the drug’s risks). 10

Benjamin Weiser & Elsa Walsh, Drug Firm’s Strategy Avoid Trial, Ask Secrecy, WASH. POST, Oct. 25, 1988. 11

Daniel J. Givelber & Anthony Robbins, Sequestered Science: The Consequences of Undisclosed Knowledge, 69 L. & CONTEMP. PROBS. 131, 133–34 (2006); S. REP. NO. 110-439, 110th Cong., 2d Sess. 4 (2008).

5 Electronic copy available at: https://ssrn.com/abstract=4811151

hormone therapy),12 tampons,13 and cigarettes.14 POs have been used to hide deadly defects in Remington rifles,15 playground equipment,16 ATVs,17 cars,18 trucks,19 and tires.20 They have concealed child sexual abuse.21 And they are often Egilman et al., supra note 9, at 293, 295 (reporting that information covered by protective order showed that the manufacturer “downplayed risks of hormone-associated breast cancer”). 12

See Tom Riley & Mary K. Hoefer, Protective Orders: Machiavelli Would Be Pleased, TRIAL, Nov. 1984, at 30 (describing litigation involving Rely tampons). 14

See infra note 115 and accompanying text.

See Editorial, Amid Court Secrecy, Guns Continued to Kill, USA TODAY, Dec. 28, 2014, https://www.usatoday.com/story/opinion/2014/12/28/sealedsettlements-remington-rifles-model-700-editorials-debates/20979945/ (discussing how sealing and protective orders concealed a defect in Remington’s Model 700 rifle); Scott Cohn, Huge Trove of Remington Rifle Documents Is Made Public, MSNBC, Nov. 15, 2016, https://www.cnbc.com/2016/11/15/huge-trove-of-remington-rifledocuments-is-made-public.html#:~:text=The%20documents%20%E2%80%94%20more%20tha n%20130%2C000,to%20make%20the%20documents%20public. (same). 16

For a discussion of how POs thwarted inquiry into defective playground equipment, see S. REP. NO. 110-439 110th Cong., 2d Sess. 7 (2008). 17

For how POs stymied regulators’ investigation into defective ATVs, see Spector et al., supra note 1. 18

Arlin R. Thrush, Note, Public Health and Safety Hazards Versus Confidentiality: Expanding the Mediation Door of the Multi-Door Courthouse, 1994 J. DISP. RESOL. 235, 249 (describing how GM used POs to silence scores of plaintiffs injured by negligently designed fuel tanks in the 1970s and 1980s). 19

Benjamin Lesser, Dan Levine, Lisa Girion & Jaimi Dowell, How Judges Added to the Grim Toll of Opioids, REUTERS (June 25, 2019), https://www.reuters.com/investigates/special-report/usa-courtssecrecyjudges/#:~:text=The%20trail%20of%20hidden%20evidence,health%2 0and%20safety%20under%20wraps. 20

For how POs insulated Cooper Tire from scrutiny, see S. REP. NO. 110-439, 110th Cong., 2d Sess. 6. (2008). 21

See Annysa Johnson & Ellen Gabler, Then-Archbishop Timothy Dolan Tried to Protect Money from Claims, Records Show, MILWAUKEE J. SENTINEL (July 2, 2013), https://archive.jsonline.com/news/religion/Cardinal-Dolan-sought-toprotect-money-from-claims-struggled-with-Vatican-to-defrockabusers-b9943953z1-213832541.html (explaining that information

6 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

used to shield the disciplinary records of police, frustrating civil rights litigation aimed at reforming department policies.22 How are POs implicated in these scandals? The text of Federal Rule of Civil Procedure 26(c), which governs POs, seems to guard against these precise situations. Per Rule 26(c), POs, which bar dissemination of discovery materials beyond the litigants themselves, are to be issued only for “good cause.”23 And, many appellate courts are quite emphatic that the willy-nilly issuance of POs violates Rule 26(c)’s clear command. As the First Circuit has explained: “[I]f good cause is not shown, the discovery materials in question should not receive judicial protection and therefore would be open to the public for inspection. Any other conclusion effectively would negate the good cause requirement of Rule 26(c).”24 Clear enough. And, even though appellate court authority gets a little murkier when the PO is filed jointly by the parties, most appellate courts hold that, even then, the trial court must independently conduct a good cause analysis. 25 As the Seventh Circuit has explained: “In deciding whether to issue a stipulated protective order, the district court must

about abusive Catholic priests was hidden behind a “broad protective order”). 22

Chelsea Hanlock, Note, Settling for Silence: How Police Exploit Protective Orders, 109 CALIF. L. REV. 1507, 1512, 1558–59 (2021) (concluding that POs prevent plaintiffs’ firms from building “a database of problem officers or collect[ing] misconduct and training materials that show systemic disciplinary failures”). POs may also prolong problematic officers’ tenures by preventing public pressure campaigns. See id. at 1522–24. 23 FED. R. CIV. P. 26(c). Fortifying this view, in the early 1990s, the Advisory Committee considered, and rejected, a proposed amendment to Rule 26(c) that would have diluted this good cause requirement. For discussion of that failed attempt, see infra notes 78–80 and accompanying text. 24

Pub. Citizen v. Liggett Grp., Inc., 858 F.2d 775, 789 (1st Cir. 1988).

As explained in detail in Part I.C., we uncover a surprising circuit split when it comes to “stipulated” POs. Two circuits (the Eleventh and Ninth) hold that the parties say-so essentially substitutes for a judicial good cause analysis, while three circuits (the Seventh, Sixth, and Third) take the opposite view, holding that, even when a motion for a PO is jointly filed, the court must still conduct an independent Rule 26(c) review.

7 Electronic copy available at: https://ssrn.com/abstract=4811151

independently determine if ‘good cause’ exists.”26 The Sixth insists that a court “cannot abdicate its responsibility to oversee the discovery process and to determine whether filing should be made available to the public.” 27 In fact, says the Sixth Circuit: “A district court abuses its discretion where it makes neither factual findings nor legal arguments supporting the need for the order.”28 The Third Circuit likewise holds that even when both parties consent to a PO, Rule 26(c) still demands that the court make “an independent determination of ‘good cause.’”29 Many have long suspected, however, that the on-theground reality is somewhat different. There have long been hints, in fact, that, particularly when the motion for a PO is jointly filed, many busy and burdened trial court judges simply acquiesce to the parties’ request.30 Indeed, fueled by that concern, beginning in the 1980s, reformers repeatedly tried to beef up Rule 26(c), believing that without fortification the provision wasn’t adequately protective of the public interest. Yet, as we’ll see, reformers’ efforts ultimately came to naught, in no small part because reformers were never able to amass convincing evidence that, in day-to-day practice, Rule 26(c)’s dictates were not being scrupulously followed. Absent such evidence, the reformers’ campaign was overcome by inertia— and by opponents’ dogged insistence that nothing was amiss. Exemplifying this position, then-District Court Judge Mark Kravitz registered opposition to reforms on behalf of the Judicial Conference before the House in 2009, demanding: “what I want to hear is evidence of Federal courts . . . not doing 26

Jepson, Inc. v. Makita Elec. Works, Ltd., 30 F.3d 854, 858 (7th Cir. 1994). 27

Procter & Gamble Co. v. Bankers Tr. Co., 78 F.3d 219, 227 (6th Cir. 1996). 28 In re Nat’l Prescription Opiate Litig., 927 F.3d 919, 929 (6th Cir. 2019) (quotation marks, citation, and alteration omitted). 29

Littlejohn v. Bic Corp., 851 F.2d 673, 680 n.15 (3d Cir. 1988).

In fact, in an admittedly dated opinion, a district court went so far as to say that it was “unaware of any case in the past half-dozen years of even a modicum of complexity where an umbrella protective order has not been agreed to by parties and approved by the court.” Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 529 F. Supp. 866, 889 (E.D. Pa. 1981). For further discussion, see Gustavo Ribeiro, (Marked Confidential): Negative Externalities of Discovery Secrecy, 100 DENV. L. REV. 171, 190 (2022) (observing that “courts regularly enter stipulated proposed protective orders with little to no . . . inquiry into whether good cause exists”).

8 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

what the rule says it should do, which is only grant protective orders for good cause shown.”31 Unless reformers could muster that evidence, Judge Kravitz (and so many others) successfully argued, hands off. 32 Reformers never could. And the almost total lack of evidence about POs’ entry and effect has remained, to date, the entire subject’s Achilles heel. As one scholar recently observed: “[D]espite the sizable amount of existing scholarship about confidential discovery, scholars have not deeply considered how confidential discovery is implemented in practice.”33 Thus, while some who advocate for changes to Rule 26(c) insist that POs are too easily and too frequently granted, we have never known whether that is actually true. In fact, even basic empirical questions have so far resisted resolution. We haven’t known how common stipulated POs are. We haven’t known how closely judges scrutinize them. Nor have we known the kinds of cases in which stipulated POs actually appear. This Article tackles these persistent and consequential questions. Drawing on a novel dataset consisting of more than 2.2 million federal court docket reports from 2005 through 2014, we show that approximately 45% of PO motions are the product of a jointly-filed Rule 26(c) motion, a much larger proportion than previous estimates suggest. We show that the percentage of stipulated POs steadily grew over the study period among cases in which an answer was filed. And, most provocatively, we show that most judges grant all, or almost all, of the stipulated motions for POs they consider. We estimate that judges grant 95% to 97% of all joint requests for POs, depending on the year. And we find that over half of the judges who considered at least 25 stipulated POs in our dataset never denied a single stipulated PO request.

Sunshine in Litigation Act of 2009: Hearing Before the Subcomm. on Commercial and Admin. Law of the H. Comm. on the Judiciary, 111th Cong. 59 (2009) (statement of Hon. Mark R. Kravitz for the Judicial Conference of the United States). 32

In the words of Arthur Miller: “Because proponents of reform have not demonstrated that significant modification of the present framework is necessary, the existing pragmatic and discretionary balancing technique should be retained.” Arthur R. Miller, Confidentiality, Protective Orders, and Public Access to the Courts, 105 HARV. L. REV. 427, 491 (1991). 33

Seth Katsuya Endo, Contracting for Confidential Discovery, 53 U.C. DAVIS L. REV. 1249, 1253 (2020).

9 Electronic copy available at: https://ssrn.com/abstract=4811151

We also supplemented our quantitative review with a qualitative analysis of 400 stipulated POs, 300 of which granted the parties’ joint request for a PO, and 100 of which consisted of (relatively rare) judicial denials. This analysis lends still further support to those who suggest that POs are routinely rubber-stamped. In all, 68% of stipulated POs in the “grant” dataset did not even pay lip service to Rule 26(c)’s good cause standard, and 83% lacked a particularized assessment of the parties’ need for secrecy. In short, stipulated POs are much more common than previously thought. They are granted at extremely high rates. They are often entered without so much as mentioning Rule 26(c)’s good cause standard. And their prevalence has grown over time. The remainder of this Article proceeds in four Parts. Part I offers a primer on POs. This Part explains the role of POs, their close cousin sealing orders, and the legal standards for entering both. Then, it considers the peculiar case of stipulated POs. What are judges to do when litigants jointly ask for a PO? That depends. Some circuits permit judges to grant stipulated POs solely on the parties’ say-so, but most others require a more searching inquiry. Part II turns to policy and politics. It catalogs past reform efforts aimed at tightening Rule 26(c), identifies and assesses the arguments in favor of and against those reforms, and, lastly, points out that these efforts have sputtered because of something like empirical exhaustion. Reform opponents have been able to hang their hats on the fact that there is, as Robert Weiner put it, “[n]o academic study, no Rand Corp. analysis, no state-by-state survey” that “has suggested any problem with protective orders in our courts.”34 Absent such a study—and absent evidence that day-to-day PO practice departs from Rule 26(c)’s command—opponents have been able to insist that nothing is amiss. The heart of this piece, Part III, offers that overdue academic study. Drawing on millions of federal court dockets, we present original evidence quantifying how often stipulated POs are granted, and we also bolster that quantitative 34

Robert N. Weiner, Protective Orders and Nest-Feathering: Plaintiffs Lawyers Around the Nation are Pressing for Laws that Would Open Discovery Files in Personal-Injury Cases to Public Scrutiny. Their Motivations are Not Nearly as Pure as Many Suppose, LEGAL TIMES , Sept. 23, 1991.

10 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

assessment with a robust qualitative review. We find that stipulated POs are much more common than previously thought, are granted at extremely high rates, and often fail to so much as mention Rule 26(c)’s good cause standard. If there is, as Arthur Miller says, a problem when a judge’s entry of a PO is “automatic or cavalier,”35 then our evidence suggests we may have a widespread problem in our courts. Finally, Part IV steps back to assess the implications of our findings and to situate these findings in larger debates. Most concretely, our evidence torpedoes long-articulated reassurances concerning the exercise of fine-tuned, individualized judicial discretion in the entry of POs. Accordingly, this Article ought to reorient policymaking regarding Rule 26(c), and it ought to reinvigorate reform activity. But, just as clearly, what we uncover has implications far beyond Rule 26(c). Indeed, our findings touch upon, and contribute to, broad and enduring debates, including the role of private civil litigation in general and tort litigation in particular, our wavering commitment to the adversarial (rather than inquisitorial) resolution of disputes, the gulf between the law on the books and the law in action, and lower courts’ sometimes surprising disobedience to official commands. I. THE FORMAL LAW REGARDING PROTECTIVE ORDERS This Part offers a primer on POs. It unfolds in three steps. First, Subpart A provides a brief overview of POs, including how they differ from their close cousins, orders to seal. Subpart B explains how Rule 26(c) addresses POs, generally. Then Subpart C zeros in on stipulated POs. This last Subpart explains why some plaintiffs might agree to keep certain information under wraps, and it also canvasses how courts interpret Rule 26(c)’s good cause standard when the PO motion is the product of party consent.

A. Protective Orders 101 The discovery default is disclosure. A party that obtains information from her adversary through the civil discovery process is presumptively entitled to share that information with others, at her discretion. In the words of the Ninth Circuit: “It is well-established that the fruits of pretrial

Miller, supra note 32, at 491.

11 Electronic copy available at: https://ssrn.com/abstract=4811151

discovery are, in the absence of a court order to the contrary, presumptively public.”36 However, as any litigator well knows, some of the information that passes through parties’ hands in litigation realistically shouldn’t be shared. That is where POs come in. POs allow a party to override the pro-disclosure presumption by prohibiting parties from sharing the information that they unearth.37 In the federal system, Federal Rule of Civil Procedure 26(c) governs this “override” mechanism. It provides that any party or person “from whom discovery is sought may move for a protective order,” and, upon such motion, “[t]he court may, for good cause, issue an order to protect a party or person from annoyance, embarrassment, oppression, or undue burden or expense.”38 The resulting order may, among other things, prohibit discovery, condition its scope, prescribe special procedures for its dissemination, or limit who may view discovered or discoverable materials.39 Furthermore, just as a Rule 26(c) order may vary in how the information may be used, an order granted pursuant to Rule 26(c) may be narrowgauge or capacious; it may restrict the disclosure of just a targeted set of documents, or instead may offer “blanket” (also known as “umbrella”) protection.40

San Jose Mercury News, Inc. v. U.S. Dist. Ct.—N. Dist. (San Jose), 187 F.3d 1096, 1103 (9th Cir. 1999); see Jepson, Inc. v. Makita Elec. Works, Ltd., 30 F.3d 854, 858 (7th Cir. 1994) (“Absent a protective order, parties to a law suit may disseminate materials obtained during discovery as they see fit.”). 37

See Seattle Times Co. v. Rhinehart, 467 U.S. 20, 34–35 (1984) (explaining that “the liberality of pretrial discovery” creates “significant potential for abuse” and that POs vindicate a substantial interest in preserving the secrecy of information that “could be damaging to reputation and privacy”). 38

FED. R. CIV. P. 26(c).

See id. 26(c)(1)(A)–(H).

Blanket POs “provide that all assertedly confidential material disclosed (and appropriately identified, usually by stamp) is presumptively protected unless challenged.” DAVID F. HERR, ANNOTATED MANUAL FOR COMPLEX LITIGATION § 11.432 (4th ed. 2022). They are often entered “without a particularized showing [of good cause] to support the claim for protection.” Id. Thus, blanket POs empower litigants to designate, for themselves, any document or deposition as confidential or sometimes highly confidential. Documents are subject to the terms of the PO unless (or until) the

12 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

One additional introductory note about POs relates to their close cousin: orders to seal.41 POs and orders to seal are very frequently confused (including, as we will see, by litigants and judges).42 But the two are different, and they are subject to different standards. An order to seal comes in once the document or testimony, unearthed during discovery, becomes a “judicial record,” typically when it is “filed with the court, or otherwise incorporated or integrated into a district court’s adjudicatory proceedings.”43 The paradigm instance is a document, produced by an opponent during discovery, that is attached as an exhibit to a summary judgment motion. Prior to the attachment process, the document is held privately by litigants and governed only by Rule 26(c); after the attachment process, the document is a “judicial record,” for which restrictions on public disclosure are subject to a higher standard. Special protections for judicial records kick in because, unlike mere discovery material (for which there is no common law right of access and, correspondingly, no First Amendment protection), there is a common law and constitutionallyprotected right to access judicial records—in part because these records necessarily play a role in the adjudicatory process.44 These records inform the court’s determination, and, without the ability to scrutinize the ground for the court’s decision, the public cannot assess the decision itself. Given all this, motions to seal face a higher bar.45 Indeed, in some receiving party alleges that the document is not confidential, at which point the court will determine whether there is good cause to protect the file. Blanket orders are contrasted with “particularized protective orders,” which, as the name suggests, cover only the materials specifically listed in the PO. Id. 41

For cogent discussions of these mechanisms’ differences and similarities, see June Med. Servs., L.L.C. v. Phillips, 22 F.4th 512, 521 (5th Cir. 2022) and Shane Grp., Inc. v. Blue Cross Blue Shield of Mich., 825 F.3d 299, 305–07 (6th Cir. 2016). 42

For the fact that courts and litigants sometimes confuse the two standards, see Endo, supra note 33, at 1254 & n.15. 43

In re Cendant Corp., 260 F.3d 183, 192 (3d Cir. 2001) (internal citations omitted). 44

See Fair Lab’y Pracs. Assocs. v. Riedel, 666 F. App’x 209, 211–12 (3d Cir. 2016) (“A strong presumption in favor of public accessibility attaches to judicial records. . . . The presumption of public access is a common law doctrine that predates the Constitution.”). 45

E.g., Pintos v. Pac. Creditors Ass’n, 605 F.3d 665, 678 (9th Cir. 2010) (“[A] party seeking to seal judicial records must show that compelling

13 Electronic copy available at: https://ssrn.com/abstract=4811151

circuits, sealing decisions are governed by a kind of strict scrutiny. As the Fourth Circuit puts it, when it comes to judicial records, “the denial of access must be necessitated by a compelling government interest and narrowly tailored to serve that interest.”46 Compared to that strict scrutiny, as Subparts B and C explain, Rule 26(c)’s “good cause” requirement is markedly less demanding.

B. Protective Orders, Generally As noted above, Rule 26(c) states “[t]he court may, for good cause, issue an order to protect a party or person from annoyance, embarrassment, oppression, or undue burden or expense.” Yet, although the standard is clear enough, when it comes to the formal interpretation of this standard, there is some surprising inter-circuit variation.47 Four circuits—the First, Fifth, Sixth, and Eighth— maintain that “[a] finding of good cause must be based on a particular factual demonstration of potential harm” that would occur if the particular information were to be divulged, “not on conclusory statements.”48 In these circuits, mere assertions—that keeping the information under wraps is reasons supported by specific factual findings outweigh the general history of access and the public policies favoring disclosure.” (quotation marks and alterations omitted)); In re Neal, 461 F.3d 1048, 1053 (8th Cir. 2006) (declaring that “only the most compelling reasons can justify non-disclosure of judicial records” (quotation marks omitted)). 46

Rushford v. New Yorker Mag., Inc., 846 F.2d 249, 253 (4th Cir. 1988). 47

For a thorough analysis of this conflicting circuit precedent, see generally Austin Peters, Jonah Gelbach, David Freeman Engstrom, Nora Freeman Engstrom, Devin Flynn & Aaron Schaffer-Neitz, Secrecy Roulette (working paper, 2024). 48

Anderson v. Cryovac, 805 F.2d 1, 7 (1st Cir. 1986); see also EEOC v. BDO USA, L.L.P., 876 F.3d 690, 698 (5th Cir. 2017) (holding that Rule 26(c) “contemplates a particular and specific demonstration of fact” (quotation marks omitted)); Serrano v. Cintas Corp., 699 F.3d 884, 901 (6th Cir. 2012) (“This Circuit has endorsed the view that to justify a protective order, one of Rule 26(c)(1)’s enumerated harms ‘must be illustrated with a particular and specific demonstration of fact, as distinguished from stereotyped and conclusory statements.’” (quotation marks omitted)); accord 8A RICHARD L. MARCUS, FEDERAL PRACTICE & PROCEDURE (WRIGHT & MILLER) § 2035 (2023 update) (“The courts have insisted on a particular and specific demonstration of fact, as distinguished from stereotyped and conclusory statements, in order to establish good cause.”).

14 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

necessary, helpful, or valuable—do not cut it.49 Two other circuits take a slightly different approach, explicitly weighing public versus private interests. The Seventh Circuit, for example, holds that “good cause” for secrecy exists when “the property and privacy interests of the litigants” predominate over the public’s interest in the publication of relevant information.50 Likewise, the D.C. Circuit has ruled that district courts should “take[] into account,” and balance “all relevant interests,” including those of the litigants, third parties, relevant statutes, and even those protected by the First Amendment.51 And, in three other circuits—the Third, Ninth, and Eleventh—there is yet another standard. These circuits require something like particularized interest plus, where a movant seeking a PO must make a particularized showing of potential harm and demonstrate that the weight of interests favors secrecy.52 The upshot: Although there is disagreements as to the particulars, in all circuits—at least when the PO is contested—the law on the books requires some substantial level of scrutiny.

E.g., P.R. Med. Emergency Grp., Inc. v. Iglesia Episcopal Puertorriqueña, Inc., 318 F.R.D. 224, 233 (D.P.R. 2016) (rejecting a motion for a PO where the movant offered only broad allegations of harm that would befall it absent the order, “devoid of substantiation through facts”). 50

Citizens First Nat’l Bank of Princeton v. Cincinnati Ins. Co., 178 F.3d 943, 944 (7th Cir. 1999). 51

United States v. Microsoft Corp., 165 F.3d 952, 960 (D.C. Cir. 1999).

See, e.g., Shingara v. Skiles, 420 F.3d 301, 306 (3d Cir. 2005); Phillips ex rel. Estates of Byrd v. General Motors Corp., 307 F.3d 1206, 1210–11 (9th Cir. 2002); Kleiner v. First Nat’l Bank of Atlanta, 751 F.2d 1193, 1205-06 (11th Cir. 1985). The law is less certain in the remaining three circuits (the Second, Fourth, and Tenth). For the Second, see Haidon v. Town of Bloomfield, 552 F. Supp. 3d 265, 269 (D. Conn. 2021) (describing various approaches to assessing good cause and noting “the Second Circuit has not yet weighed in on the issue”); Schoolcraft v. City of New York, 2013 WL 4534913, at *3 (S.D.N.Y. 2013) (“There is a split within the district courts of this Circuit as to the showing necessary to establish that good cause exists.”). For the Fourth, see Ardrey v. United Parcel Serv., 798 F.2d 679, 682 (4th Cir. 1986) (noting that district courts enjoy a “wide latitude in controlling discovery”). And, for the Tenth, see Rohrbough v. Harris, 549 F.3d 1313, 1321 (10th Cir. 2019) (asserting that the 26(c) good cause standard is “highly flexible” (quoting United States v. Microsoft Corp., 165 F.3d 952, 959 (D.C. Cir. 1999)).

15 Electronic copy available at: https://ssrn.com/abstract=4811151

C. Stipulated Protective Orders The above assumes that the motion for a PO is, and will be, contested: that one litigant will seek the motion for a PO and the other litigant will oppose that motion. Yet, it turns out, motions for POs often are not contested. As we explain in Part III.B, our data reveal that motions for POs are commonly stipulated (sometimes called “unopposed,” “joint,” or “jointly filed”). In fact, we find that motions for POs are stipulated nearly half of the time. Given that it is typically defendants who are eager to keep unflattering information out of the public eye, it might, initially, seem odd that so many PO motions are jointly filed. But, anecdotal evidence indicates that plaintiffs’ lawyers often make the strategic decision, essentially, to go along to get along. 53 Plaintiffs’ lawyers, it is said, recognize that, without a PO, the defendant is apt to put the brakes on discovery and to be less forthcoming in depositions, when responding to interrogatories, and when divulging documents—and, in the scheme of things, the PO battle isn’t the battle the lawyer wants to fight.54 As one plaintiffs’ lawyer explained, the lawyer “may be so concerned with gaining access to the key documents she needs to present her client’s case that she . . . may decide it isn’t worth slowing down the litigation to fight.”55 Or, as Lance Cooper, the lawyer for Ken and Beth Melton put it: “plaintiffs’ lawyers want to try to get the documents as soon as possible to prosecute their case.”56 A

S. REP. NO. 110-439, at 9 (2008) (Conf. Rep) (testimony of Leslie Bailey). 54

As Professor Howard Erichson has observed: “With a protective order in place, a responding party is more willing to turn over information rather than asserting and litigating every plausible relevance objection and privilege objection.” Howard M. Erichson, Court-Ordered Confidentiality in Discovery, 81 CHI-KENT L. REV 357, 359 (2006); see also FRANCIS H. HARE, JAMES L. GILBERT & WILLIAM H. REMINE, CONFIDENTIALITY ORDERS 2 (1988) (“It is commonplace for defense counsel to offer to furnish the requested documents without opposition, if the plaintiff will stipulate to an order.”); Miller, supra note 35, at 492 n.322 (suggesting that plaintiffs agree to POs “to facilitate [their] own access to discovery materials”). 55

S. REP. NO. 110-439, at 9 (2008) (Conf. Rep) (testimony of Leslie Bailey). There is, then, yet another question which is how broad the PO is to be. For a discussion of various options see supra note 40. 56

Cooper Interview, supra note 5.

16 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

PO is thought to promote the prompt disclosure of documents—and so motions for POs are often jointly filed. Even when the parties agree to the terms of a PO, however, that agreement is not judicially enforceable on its own. Even if the motion for a PO is jointly filed, the court must still enter the PO under Rule 26(c). But pursuant to what standard? Must a trial court still rigorously assess good cause, as it would if the Rule 26(c) motion were contested? In answering this question, appellate courts again differ on the particulars.57 In two circuits, party agreement temporarily obviates the good cause showing.58 Taking this tack, the Ninth Circuit has explained: “While courts generally make a finding of good cause before issuing a protective order, a court need not do so where (as here) the parties stipulate to such an order.”59 Similarly, the Eleventh Circuit has held that stipulated POs may be entered based on agreement alone.60 However, both Circuits hedge, explaining that, because the parties “never established good cause for protection in the first place,” any “party seeking the stipulated order’s protection must satisfy Rule 26(c)’s good cause standard” if secrecy is later challenged.61 57

For a cogent summary of this debate, see generally Brief for Amici Curiae Civil Procedure Law Professors in Support of Petitioners, Doe 7 v. Chiquita Brands Int’l, Inc., 142 S. Ct. 312 (2021) (No. 20-1599) [hereinafter Civil Procedure Scholar Br.]. 58

The Fifth Circuit may well agree that stipulation eliminates the need for a good cause showing but has never said so explicitly. Cf. Binh Hoa Le v. Exeter Finance Corp., 990 F.3d 410, 418–20 (5th Cir. 2021) (faulting a district court for entering a sealing order, but not a PO, based solely on the parties’ stipulation). 59 In re Roman Catholic Archbishop of Portland in Or., 661 F.3d 417, 424 (9th Cir. 2011); see also Beckman Indus., Inc. v. Int’l Ins. Co., 966 F.2d 470, 476 (9th Cir. 1992) (explaining that parties did not have “to make a ‘good cause’ showing” when entering “a stipulated blanket [protective] order”). 60

In re Chiquita Brands Int’l, Inc., 965 F.3d 1238, 1249–50 (11th Cir. 2020) (accepting the district court’s practice of accepting stipulated POs without finding good cause); see also Chicago Trib. Co. v. Bridgestone/Firestone, Inc., 263 F.3d 1304, 1307 (11th Cir. 2001) (observing that the stipulation process “postpones the necessary showing of ‘good cause’ required for entry of a protective order until the confidential designation is challenged”). 61

In re Chiquita, 661 F.3d at 1249–50.

17 Electronic copy available at: https://ssrn.com/abstract=4811151

Operationalizing this standard, one district court from Florida (answerable to the Eleventh Circuit) has explained: “Parties have the freedom and flexibility to agree on the terms of stipulated protective orders designed to protect ‘confidential’ and ‘highly confidential’ material. . . . [C]ourts typically enter the proposed stipulated protective orders jointly submitted by the parties.”62 Likewise, a district court in California (hailing from the Ninth Circuit) has recently declared: “Although courts generally make a finding of good cause prior to issuing a protective order, a court need not do so if the parties stipulate to entry of a protective order.”63 The Eleventh and Ninth Circuits, however, appear to be in the minority. In three circuits, the law is clear that trial court judges are duty-bound to subject even stipulated POs to quite careful scrutiny.64 Thus, the Seventh Circuit has explained that, just as the trial court has a nondelegable duty to conduct an “independent inquiry” when assessing the fairness and adequacy of a class action settlement pursuant to Rule 23(e) (even when the parties agree to settle), “[i]n deciding whether to issue a stipulated protective order, the district court must independently determine if ‘good cause’ exists.”65 Party agreement, the Seventh Circuit has emphasized, simply does not obviate the need for careful review. Similar to the Seventh, the Sixth Circuit has observed that, even when the parties agree to a PO’s terms, trial courts’ “discretion [to enter that order] is limited by the careful

Malibu Media, LLC v. Doe, 2014 WL 1292692, at *3 (M.D. Fla. 2014) (quotation marks and alterations omitted). 63

In re Facebook, Inc. Consumer Priv. User Profile Litig., 2021 WL 3209711, at *3 (N.D. Cal. 2021); see also Ledford v. Idaho Dep’t of Juv. Corr., 2013 WL 5798682, at *1 (D. Idaho 2013) (“If the parties stipulate to a protective order—as they did here—the district court may enter a protective order without first finding good cause.”). 64

In some circuits, no case clearly establishes a governing standard for stipulated POs. See generally Peters et al., supra note 47 (expounding on relevant standards); cf. Minter v. Wells Fargo Bank, N.A., 2010 WL 5418910, at *2 (D. Md. 2010) (“Neither this Court nor the Fourth Circuit has explicitly defined the parameters of the initial ‘good faith’ review required pursuant to a stipulated confidentiality order . . . .”). 65

Jepson, Inc. v. Makita Elec. Works, Ltd., 30 F.3d 854, 858 (7th Cir. 1994). Arthur Miller has also analogized the situation to the situation a court faces when reviewing a class action settlement. See infra note 75 and accompanying text.

18 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

dictates of Fed. R. Civ. P. 26.”66 A trial court, the Sixth Circuit insists, simply “cannot abdicate its responsibility to oversee the discovery process and to determine whether filings should be made available to the public.”67 Thus, says the Sixth Circuit: “A district court abuses its discretion,” where it grants an uncontested motion for a PO and “makes neither factual findings nor legal arguments supporting the need for the order.”68 Likewise, the Third Circuit has indicated that, even when both parties consent to a PO’s entry, Rule 26(c) still demands “an independent determination of ‘good cause.’”69 Employing this stricter standard, a district court in the Seventh Circuit has explained: “[E]ven if the parties stipulate to the terms of a protective order . . . the parties must satisfy the good cause requirement contained in Rule 26(c)(1) . . . . Since the Court is the primary representative of the public interest in judicial proceedings, it must review requests for stipulated POs for good cause without acting as a rubberstamp.”70 Similarly, a district court in Pennsylvania (answerable to the Third Circuit) has explained: “Stipulated protective orders must still meet the requirements of Rule 26(c), which requires demonstrating the existence of confidential information and good cause as to why such information should not be disclosed.”71 Other influential voices have also come down in favor of this harder-edged position. Take the Federal Judicial Center. It has clearly stated that, even when the parties file a joint motion for a PO, that motion can only be granted upon 66

Procter & Gamble Co. v. Bankers Tr. Co., 78 F.3d 219, 227 (6th Cir. 1996). 67

Id.

In re Nat’l Prescription Opiate Litig., 927 F.3d 919, 929 (6th Cir. 2019). 69

Littlejohn v. Bic Corp., 851 F.2d 673, 680 n.15 (3d Cir. 1988); accord Pansy v. Borough of Stroudsburg, 23 F.3d 772, 785 (3d Cir. 1994) (declaring it “[d]isturbing[]” that “some courts routinely sign orders which contain confidentiality clauses without considering the propriety of such orders, or the countervailing public interests which are sacrificed by the orders” and further advising “whether an order of confidentiality is granted at the discovery stage or any other stage of litigation . . . good cause must be demonstrated to justify the order”). 70

Elder v. Bimbo Bakeries USA, Inc., 2022 WL 3443766, at *2 (S.D. Ill. 2022). 71

Sprinturf, Inc. v. Sw. Recreational Indus., Inc., 216 F.R.D. 320, 323 (E.D. Pa. 2003).

19 Electronic copy available at: https://ssrn.com/abstract=4811151

an independent judicial determination “that the order is supported by good cause.”72 The canonical Wright & Miller treatise likewise advises that, even as the product of party stipulation, POs “are not authorized simply on the requesting parties’ say-so. . . . Even when the parties consent, the court may not enter an order unless Rule 26(c) is satisfied.”73 Leading scholars, too, resist the notion that party agreement somehow substitutes for good cause under Rule 26(c). “Judges,” Arthur Miller explains, “must guard against any notion that the issuance of protective orders is routine, let alone automatic, even when the application is supported by all the parties.”74 Miller elaborates: “When all the parties support the protective order . . . the court is faced with an essentially non-adversarial situation” similar to the fiduciary burden that federal judges bear when independently “evaluating a proposed class action settlement under Federal Rule 23(e).”75 Professor Howard Erichson concurs. According to Erichson, “the parties’ say-so alone, without some showing of a need for confidentiality, does not constitute good cause for the granting of a protective order.”76 Or, as Laurie Doré says: Rule 26(c) authorizes a district court to issue a protective order only for “good cause shown.” It does not carve out any exception, temporary or otherwise, for stipulated orders. Even if the parties agree to the terms of a protective order, then, they must still demonstrate good cause to justify its issuance.77

ROBERT TIMOTHY REAGAN, FEDERAL JUDICIAL CENTER, CONFIDENTIAL DISCOVERY: A POCKET GUIDE ON PROTECTIVE ORDERS 6 (2012). 73

MARCUS, supra note 48, at § 2035.

Miller, supra note 35, at 492. Miller further advised that this “careful[]” review must be “tailor[ed]” to the facts of each case and “should take account of a kaleidoscope of factors, including the likely outcome on the merits, the value or importance of commercial or personal data, the identity of the parties and any apparent outside interests, [and] the existence of any threat to health and safety.” Id. at 492–93. 75

Id. at 492, n.322.

Erichson, supra note 54, at 373.

Laurie Kratky Doré, Secrecy by Consent: The Use and Limits of Confidentiality in the Pursuit of Settlement, 74 NOTRE DAME L. REV.

20 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

Lastly, there is something like a legislative history argument that further bolsters the position of the many courts, institutions, and academics who insist that party agreement does not constitute good cause under Rule 26(c). Namely, in the mid-1990s, rulemakers considered whether to amend Rule 26(c) to say that a PO could be issued “for good cause shown or on stipulation of the parties.”78 The amendment was controversial, in part because it was understood at the time that, if the amendment passed, it would have, “worsened the court secrecy problem”—and, ultimately, the amendment was defeated.79 Rejecting the amendment by voice vote, the Judicial Conference “express[ed] concern that the proposed rule would change existing practice by allowing entry of protective orders without a showing of good cause.”80 Under the canon of statutory construction known as the “rejected proposal rule,” non-adopted amendments can reflect what a statute is not.81 So, the fact that the Judicial Conference 283, 342 (1999); see also Civil Procedure Scholar Br., supra note 57, at 14 (“There is no basis in the Federal Rules for a court to approve a stipulated protective order without conducting an independent assessment of good cause. To the contrary, though parties often will stipulate to the entry of a discovery protective order, it is the court that ultimately must enter the order, and the court may do so only in compliance with Rule 26(c)’s good cause requirement.” (quotation marks omitted)); Endo, supra note 33, at 1259 (“After the parties propose a stipulated protective order, the court must determine if there is good cause to issue the order.”). 78

Memorandum of Comm. on Rules of Practice and Procedure to Comm. on Rules of Practice and Procedure 18 (June 2, 1995), https://www.uscourts.gov/sites/default/files/fr_import/CV6-1995.pdf [hereinafter Letter of June 2, 1995] (emphasis added). 79

Saundra Torry, Judges Reject Record-Secrecy Rule, WASH. POST, Mar. 15, 1995, at A8, https://www.washingtonpost.com/archive/politics/1995/03/15/judgesreject-record-secrecy-rule/2ec968fc-4f37-40ba-b430-2fc1054a0d4e/. 80

Letter of June 2, 1995, supra note 78, at 1–2. See also Linda Greenhouse, Judicial Conference Rejects More Secrecy in Civil Court, N.Y. TIMES, Mar. 15, 1995, at B9, https://www.nytimes.com/1995/03/15/us/judicial-conference-rejectsmore-secrecy-in-civil-court.html. 81

See FRANK B. CROSS, THE THEORY AND PRACTICE OF STATUTORY INTERPRETATION 65 (2009) (explaining that a rejected amendment “provides fairly persuasive evidence that the content of the amendment was not the legislative intent”); FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 144 (2000) (finding that Congress’s decision to reject “bills that would have granted the FDA [] jurisdiction” over

21 Electronic copy available at: https://ssrn.com/abstract=4811151

expressly considered and rejected an amendment to Rule 26(c) that would have established that party agreement could stand in for good cause supplies some ammunition to those who believe that, when it comes to POs, the parties’ say-so doesn’t suffice.82 II. REFORM HISTORY: PERENNIAL PROPOSALS AND DEFICIENT DATA This Part canvasses past and present debates regarding POs—as well as the meager empirical evidence that has informed the discussion. Subpart A explains that POs have long been the subject of reform efforts, while Subpart B catalogs prominent pro and con arguments. Then, Subpart C traces prior empirical studies, which have been wellintentioned but limited. Given this deficient data, it is no surprise that, to this point, debates addressing possible reforms to Rule 26(c) have backstopped on anecdote and hunches, rather than fact.

A. Past Efforts to Limit Expansive Protective Orders The modern debate over litigation secrecy—and over the value of, and standards for issuing, POs—dates back to the late 1980s, when the premier plaintiff-side organization, the American Tort Law Association (ATLA), now renamed the American Association for Justice (AAJ), put litigation secrecy high on its organizational agenda.83 In this initial push for tobacco clarifies that the FDA does not have that jurisdiction). Like many statutory canons, this theory of interpretation is not iron-clad. See William N. Eskridge, Jr., Interpreting Legislative Inaction, 87 MICH. L. REV. 67, 71, 94–108 (1988) (describing this rule and critiquing it); NLRB v. C & C Plywood Corp., 385 U.S. 421, 427 (1967) (refusing to draw an inference as to intent from a rejected amendment). 82

Interestingly, in Glenmede Tr. Co. v. Thompson, 56 F.3d 476, 485 & n.15 (3d Cir. 1995), the court rejected a reading of Rule 26(c) that would have been “tantamount to permitting the parties to control the use of protective orders,” and, in so doing, drew on this rejected amendment. 83

ATLA Fights Secrecy in Litigation, ATLA ADVOC., Sept. 1989 at 1 (discussing a resolution that “discourage[d] attorneys from entering into secrecy agreements and . . . encourage[d] courts not to enter into or enforce secrecy agreements without good cause”). For a discussion of ATLA’s early efforts, see Miller, supra note 32, at 442–43; Gail Diane Cox, Yearly Meeting: Sunshine in San Diego for ATLA, NAT’L L.J., July 30, 1990, at 3; Bill Wagner, Secrecy Betrays Justice: ATLA’s

22 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

greater transparency, ATLA—pointing to a few high-profile scandals—insisted that the information uncovered in litigation sometimes pertained to serious public health hazards and that, in such cases, courts that rubber-stamped expansive POs were complicit in perpetuating consumer harm.84 Responding to ATLA’s concerns, in 1989, Congress called to order its first hearing on the matter, followed soon after by the introduction of a federal “Sunshine in Litigation” bill on the House floor.85 That initial bill sought to allow parties to share evidence pertaining to public health and safety with regulators even if that information was subject to a PO.86 Yet, even though the bill was quite narrow, opposition to it was broad. Rallied, in part, by Alfred Cortese, a hard boiled corporate lobbyist, and buoyed by the prolific writing of Professor Arthur Miller, who was then at Harvard, corporate interests parried back “pro-sunshine” arguments. Cortese, in all this, took quite a crabbed view of court transparency: “What is the public right to observe the legal system? What is it? . . . The right of the public to observe the court system means that they have a right to show up in court when there is President’s Opinion, NAT’L L.J., July 24, 1989, at 1, 4 (recounting examples of POs harming public health and describing ATLA’s efforts to “encourage[] attorneys to resist” POs). 84

Miller, supra note 32, at 442 (“According to ATLA, protective orders . . . are being used with increasing frequency to hide deadly product defects or other ‘public hazards’ from the public.”); Lawyers for Civil Justice and the National Chamber Litigation Center, Comments to the Committee on Rules of Practice and Procedure on the Need for Amendment of Federal Rule of Civil Procedure 26(c) (Apr. 18, 1994), at 3, available at https://www.uscourts.gov/sites/default/files/fr_import/CV1994-04.pdf [hereinafter Comments for LCJ] (explaining that early efforts were based on the claim “that information produced in litigation revealed serious defects in consumer products or public exposures to toxic materials” and that the courts that issued these expansive POs were “unwitting co-conspirators” in perpetuating harm); see also Russ M. Herman, Secrecy, Discovery Abuse Breed Unethical Conduct, NAT’L L.J. Aug. 1, 1988, at 18-21 (voicing early opposition to widespread use of POs). Others, however, charged that the PO reform movement wasn’t really motivated by the public interest and was, instead, “a camouflaged effort to get marketable information.” See Tripp Baltz, Shhhh Confidentiality in the Courts, CHI. LAW., Jan. 1991, at 50–51. 85

Comments for LCJ, supra note 84, at 5.

Spector et al., supra note 1.

23 Electronic copy available at: https://ssrn.com/abstract=4811151

a public trial. That is the extent of the public’s right to observe.”87 In the ensuing decades, although a few in Congress (most notably, the indefatigable Senator Herb Kohl of Wisconsin) continued to champion reform, the tug-of-war between sunlight and secrecy reached something of a stalemate.88 Thus, in a slow-motion Groundhog Day, between the late 1980s and the late 2010s, every few years, a version of a Sunshine in Litigation Act popped up in the House or Senate. Every few years, the Act kicked off heated rounds of discussion. And, every few years, the Act was defeated.89 State efforts have mostly followed the same trajectory. At roughly the same time as the first Sunshine in Litigation Act was introduced in Congress, a wave of similar reforms made their way through state legislatures. In fact, between 1990 and 1994, state legislators introduced some ninety-three discrete proposals to limit POs.90 Most of these bills died quick deaths.91 Yet, a smattering squeaked through. In one of the first reforms, promulgated in 1990, the Texas Rules Committee revised Rule 76a to make it far more 87

Michelle Conlin, Dan Levine & Lisa Girion, Why Big Business Can Count on Courts to Keep its Deadly Secrets, REUTERS (Dec. 19, 2019), https://www.reuters.com/investigates/special-report/usa-courtssecrecy-lobbyist/. For more on the defense bar’s efforts, see Baltz, supra note 84, at 51 (noting that the president of the International Association of Defense Counsel “declared that one goal during his term would be to preserve civil defendants’ rights to protective orders”). 88

Corporate lawyers referred to the Sunshine in Litigation Act as the “perennial Kohl bill.” Id. 89

Robert Timothy Reagan, The Hunt for Sealed Settlement Agreements, 81 CHI.-KENT L. REV 439, 441 (2006). More recently, in 2019, during a congressional hearing on court transparency, Rep. Jerrold Nadler indicated he planned to reintroduce the Sunshine in Litigation Act, although it does not appear that such an Act was introduced. See The Federal Judiciary in the 21st Century: Ensuring the Public’s Right of Access to the Courts, Hearing Before the H. Subcomm. on Courts, Intellectual Property, and the Internet of the Comm. on the Judiciary, 116th Cong. 5 (2019) (Rep. Jerrold Nadler). 90

Comments for LCJ, supra note 84, at 6. For a comprehensive compilation, see Miller, supra note 35, 429–31, n.7. 91

Comments for LCJ, supra note 84, at 6 (“[N]otwithstanding the broad media coverage and legislative fervor used to promote such legislation and rules, only three such restrictive proposals out of 93 were ever adopted.”).

24 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

difficult for courts to issue POs or to seal documents pertaining to “public health and safety, or the administration of public office or the operation of government.”92 The same year, Florida enacted the Sunshine State’s version of the Sunshine in Litigation Act. It forbids POs that conceal “a public hazard or any information concerning a public hazard” or “information which may be useful to . . . the public in protecting themselves from injury which may result from public hazard.”93 Around the same time, Virginia enacted a statute specifying that, in lawsuits involving personal injury or wrongful death, no PO shall “prohibit an attorney from voluntarily sharing such materials or information with an attorney involved in a similar or related matter,” subject to a few restrictions.94 In 1994, Washington enacted a provision subjecting POs to a balancing test that considers the risk of public hazards.95 And in 1995, Louisiana joined the budding movement, restricting the entry of POs and orders to seal “if 92

See TEX. R. CIV. P. 76a (allowing courts to issue protective and sealing orders covering documents pertaining to “general public health and safety” only after determining that private interests outweigh public health risks and that the order is the least restrictive means of protecting private interests). For further discussion, see generally Lloyd Doggett & Michael J. Mucchetti, Public Access to Public Courts: Discouraging Secrecy in the Public Interest, 69 TEX. L. REV. 643 (1991). In enacting the Rule, reformers overcame stiff opposition from business interests. E.g., David E. Chamberlain, Proposed Rule 76(a): An Elaborate TimeConsuming, Cumbersome Procedure, 53 TEX. B.J. 348 (1990) (insisting that 76(a)’s enactment would impair settlement, delay the resolution of conflict, and imperil private information); Letter from Jack C. Goldstein to Charles Herring, Jr. (Dec. 26, 1989), in PROPOSED RULE 76A AND COMPANION AMENDMENTS TO RULE 166B(5) 170, 170 (1990) (threatening that, if Texas courts made it harder to access POs, “legitimate businesses . . . [would] locate outside Texas” and that “thieves and pirates [would] look to Texas for ‘political asylum’ from traditional principles of business ethics and morals”). 93

FLA. STAT. ANN. § 69.081. A “public hazard” is defined as “an instrumentality, including but not limited to any device, instrument, person, procedure, product, or a condition of a device, instrument, person, procedure or product, that has caused and is likely to cause injury.” Id. § 69.081(2). For more on the Act’s contemporary application, see Goodyear Tire & Rubber Co. v. Schalmo, 987 So. 2d 142, 145 (Fla. Dist. Ct. App. 2008). 94

VA. CODE ANN. § 8.01-420(A). For more on the law’s 1989 enactment, see Alan B. Morrison, Protective Orders, Plaintiffs, Defendants and the Public Interest in Disclosure; Where Does the Balance Lie?, 24 U. RICH. L. REV. 109, 122–23 (1989). 95

WASH. REV. CODE ANN. § 4.24.611.

25 Electronic copy available at: https://ssrn.com/abstract=4811151

the information or material sought to be protected relates to a public hazard.”96 After a very long hiatus—and catalyzed in part by the #MeToo movement—transparency efforts may be seeing a tentative resurgence. In 2022, Washington, D.C. Councilmember Mary Cheh introduced the Sunshine in Litigation Act. The legislation would require D.C. judges to consider public health and safety before granting a PO or sealing court records. Cheh cited the opioid epidemic as an impetus for her bill, noting, “[c]ourt-sanctioned secrecy in such cases can be a matter of life and death.”97 Also in 2022, California state Senator Connie Leyva introduced the “Public Right to Know Act.”98 The bill would create a presumption against court orders concealing information about defective products or environmental hazards unless the court finds that the public interest in disclosure is outweighed by a substantial need for secrecy.99 Yet, like so many predecessors, both bills were defeated.100

B. Political and Policy Debates Over Protective Orders As suggested above, thirty years’ worth of reform efforts have generated heated arguments for and against change. This is so despite the modesty of many of the proposals. Reformers have never sought (and, indeed, virtually no one has ever advocated) an outright ban on POs. 96

LA C.C.P. ART. 1426(C). The provision also prevents courts from issuing POs where the information may be useful to members of the public in protecting themselves from injury that might result from a public hazard. Id. The provision, however, has less bite than it may first appear. For discussion, see Dustin B. Benham, Tangled Incentives: Proportionality and the Market for Reputation Harm, 90 TEMP. L. REV. 427, 450–52 (2018); Roma Perez, Two Steps Forward, Two Steps Back: Lessons to Be Learned from How Florida’s Initiatives to Curtail Confidentiality in Litigation Have Missed Their Mark, 10 FLA. COASTAL L. REV. 163, 218 (2009). 97

Sally Greenberg, Sunshine in Litigation Act introduced in the District of Columbia, NAT’L CONSUMERS LEAGUE (Nov. 2, 2022), https://nclnet.org/dc-sunshine-in-litigation-act/. 98

SB-1149, 2021-2022., Cal. Leg. Reg. Sess. (2022).

Id.

100

The California Senate passed the legislation, but the effort died in the Assembly. Id. A public hearing was held on the D.C. legislation in January 2023, but the bill progressed no further. Greenberg, supra note 97.

26 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

Everyone seems to agree that trade secrets and the protection of personal privacy justify POs in some circumstances.101 Instead, the fight over POs has been protracted and bitter. But it’s been waged on markedly narrow terrain. Reformers have tended to focus on POs (regardless of whether they are the product of party contestation or consent) that conceal health and safety hazards. And reformers have sought, not to outlaw such POs, but merely to subject those—and only those—to exacting judicial scrutiny.102 Opponents have tended to resist these reforms. Below, we rehearse the two sides’ now-familiar positions. 1. Reformers’ Arguments in Favor of Greater Transparency Arguments in favor of greater transparency tend to focus on harm, cost, and institutional legitimacy. First, reformers highlight harm—and, in particular, POs’ welldocumented role in concealing and thus perpetuating catastrophic injury. POs, in reformers’ telling, prevent consumers and regulators from learning of defective products and allow those products to circulate unabated.103 They cause people to associate with abusers. 104 And they make it harder for litigants to hold bad actors, including violent police 101

Richard L. Marcus, The Discovery Confidentiality Controversy, 1991 U. ILL. L. REV. 457, 487 (1991) (explaining that reformers “accept broad confidentiality interests in commercial litigation and to protect personal privacy”). 102

E.g., S. REP. NO. 110-439 110th Cong., 2d Sess. 10 (2008) (explaining that the Sunshine in Litigation Act “merely requires an additional step—consideration of public health and safety”). Sunshine in Litigation Act of 2008: Hearings on H.R. 5884 Before the H. Comm. on the Judiciary, Subcomm. On Commercial and Administrative Law, 110th Cong. 2 (2008) (statement of Rep. Linda T. Sanchez) (“H.R. 5884 is modest in its scope. Its key provision would require courts to do what some Federal Judges already do: consider the public’s interest in health and safety before entering certain confidentiality orders that would conceal information from the public . . . .”). 103

Dustin B. Benham, Dirty Secrets: The First Amendment in Protective-Order Litigation, 35 CARDOZO L. REV. 1781, 1785–86 (2014); Richard Zitrin, The Judicial Function: Justice between the Parties, or a Broader Public Interest, 32 HOFSTRA L. REV. 1565, 1565– 66 (2004); see also Egilman et al., supra note 9, at 295 (collecting pharmaceutical examples). 104

See Johnson & Gabler, supra note 21 (explaining that information about abusive catholic priests was hidden behind a “broad protective order”).

27 Electronic copy available at: https://ssrn.com/abstract=4811151

departments105 and discriminatory employers,106 to account. As one reformer summarized in an early Senate hearing: “Men, women, and children in this country are unwittingly buying and using dangerous products, inhaling and drinking toxic pollutants, [and] being treated by incompetent doctors . . . because of unnecessary secrecy in the courts.”107 Second, and relatedly, some argue that reform is needed because the status quo is not working. Left to their own devices, judges are not adequately safeguarding the public interest. POs, reformers insist, have become commonplace; they are “routinely requested in virtually every product liability, automobile design, toxic tort, environmental, medical malpractice, pharmaceutical, and consumer fraud case in the country.”108 And, rather than scrutinizing POs for good cause, judges too often simply offer a stamp of approval on the parties’ say-so. As one commentator put it: “[S]tipulated protective orders are often approved pro-forma by overburdened courts anxious to avoid time consuming inquiries into discovery disputes.”109 As another explained in 105

See Hanlock, supra note 22, at 1558–59 (concluding that POs prevent plaintiffs’ firms from building “a database of problem officers or collect[ing] misconduct and training materials that show systemic disciplinary failures”). 106

See Jamillah Bowman Williams, Diversity as a Trade Secret, 107 GEO. L.J. 1685, 1727 (2019) (arguing that POs may soon be used to safeguard companies’ diversity data, “making it impossible for other potential litigants to determine whether they also have a related [employment discrimination] claim” and derailing “systemic discrimination claims”). 107

Examining the Use of Secrecy and Confidentiality of Documents by Courts in Civil Litigation, Hearing Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice, 101st Cong. 66 (1990) (statement of Arthur H. Bryant of Trial Lawyers for Public Justice) [hereinafter Bryant Testimony]. 108

Id.; see also Lori E. Andrus, Fighting Protective and Secrecy Orders: Sunshine is the Best Disinfectant, PLAINTIFF MAG. (2014), at 1, https://plaintiffmagazine.com/recent-issues/item/fightingprotective-and-secrecy-orders-2 (“Protective orders have become routine, particularly in complex cases.”). 109

Seymour Moskowitz, Discovering Discovery: Non-Party Access to Pretrial Information in the Federal Courts 1938-2006, 78 U. Colo. L. Rev. 817, 826 (2007); see also Andrus, supra note 108, at 1 (asserting that “with courts overburdened and understaffed, judges are all too often content to sign off on blanket protective orders without any showing of good cause”); Dustin B. Benham, Foundational and Contemporary Court Confidentiality, 86 MO. L. REV. 211, 222 (2021)

28 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

Senate testimony when asked what factors judges consider when deciding whether to issue a PO: “I think the truth is that judges rarely weigh these orders at all.”110 Instead, “what happens is that the parties agree to them and the judges, faced with the parties agreeing and no one objecting, simply sign off.”111 Too often, adds Judge Joe Anderson of South Carolina, he and his colleagues merely “rubber-stamp confidentiality orders presented to them, sometimes altogether ignoring or merely giving lip service to the body of law and existing court rules that are supposed to apply.”112 Third, unbridled POs, reformers allege, raise the cost of litigation. POs stymie plaintiff cooperation across actions involving overlapping issues or defendants, requiring each plaintiffs’ lawyer to build cases in isolation and from scratch.113 To quote a lawyer for Public Citizen’s Litigation Group, “[e]ssentially [a PO] makes every plaintiff’s lawyer (“The parties often agree to protective orders supported by thin or nonexistent proof of good cause. Courts enter them because the parties have agreed.”); Conlin et al., supra note 87 (stating that the entry of a PO is a “pro forma exercise[]”). 110

Bryant Testimony, supra note 107, at 162.

111

Id.; see also Federal Sunshine in Litigation Act and Federal Court Settlements Sunshine Act, Hearings on H.R. 2017 and H.R. 3803 Before the H. Comm. on the Judiciary, Subcomm. on Intellectual Property and Judicial Administration, 102nd Cong. 85 (1992) (statement of Rep. William J. Hughes) (“Unfortunately, I have the perception, as do a number of my colleagues, that the courts are, just as a matter of course, approving secrecy orders when submitted without making the independent determination as to whether they serve the public good.”). 112

Joseph F. Anderson Jr., Hidden from the Public by Order of the Court: The Case against Government-Enforced Secrecy, 55 S.C. L. REV. 711, 715 (2004). In the article, Judge Anderson quoted another judge who confessed, in a moment of candor, that she “would sign an order that stipulated that the moon was made out of cheese if the lawyers came in and asked me to sign it.” Id. at 729 (quoting Judge Judith McConnell of the San Diego, California, Superior Court). 113

See, e.g., Francis H. Hare Jr., James L. Gilbert & Matthew S. Ellenberger, Confidentiality Orders in Products Liability Cases, 13 AM. J. TRIAL ADVOC. 597, 602 (1989) (“By prohibiting the disclosure of information gleaned from discovery, a confidentiality order forces each attorney to develop his client’s case in a vacuum.”). Plaintiffs can always attempt to gain access to prior discovery by intervening to modify POs. But there is no guarantee that courts will grant their modification request, and, even if the request is granted, there is no guarantee that discovery will be forthcoming. For discussion, see Benham, supra note 7, at 2211–12.

29 Electronic copy available at: https://ssrn.com/abstract=4811151

reinvent the wheel in every case.”114 This silo-ing can, in turn, impair litigation efficiency and may systematically slant (or, some suggest, further slant) the litigation playing field toward well-heeled players.115 Not merely theoretical, these dynamics were on vivid display in early tobacco litigation. In early battles, cigarette company disclosures were subject to expansive POs, and such orders prevented plaintiffs’ counsel (who were typically cash strapped “lone wolf” solo practitioners) from sharing discovery with one another.116 Unable to divulge what they learned, plaintiffs’ lawyers battling the cigarette companies were condemned to build each case alone and from scratch, which dramatically—but asymmetrically—increased the cost and burden of litigation.117 Fourth, reformers argue that overbroad POs inhibit transparent judicial decision-making—and that transparency 114

Paul M. Barrett, Protective Orders Come Under Attack—Plaintiffs Get Judges to Open Court Files, WALL ST. J., Aug. 31, 1988 (quoting C.F. Hitchcock). 115

See Erichson, supra note 54, at 367 (explaining that, when it is not stymied by a PO, “[c]oordination among counsel in related cases not only promotes litigation efficiency, but also enhances the quality of legal work and tends to level the field in asymmetrical multiparty litigation”); see also HARE, ET AL., supra note 54, at 15–19 (similar). For the classic account of why litigation may be slanted toward certain “repeat” players, see generally Marc Galanter, Why the “Haves” Come out Ahead: Speculations on the Limits of Legal Change, 9 L. & SOC’Y REV. 95 (1974). 116

Nora Freeman Engstrom & Robert L. Rabin, Pursuing the Public Health Through Litigation: Lessons from Tobacco and Opioids, 73 STAN. L. REV. 285, 292, 298 (2021). 117

See Marc Z. Edell, Cigarette Litigation: The Second Wave, 22 TORT & INS. L.J. 90, 91 (1986) (“[P]rotective orders, obtained by the defendants in almost every instance, prohibited the dissemination of discovery to either the public or to other lawyers who were involved in similar litigation. This required plaintiffs’ lawyers to initiate discovery anew in each case.”); Robert L. Rabin, A Sociolegal History of the Tobacco Tort Litigation, 44 STAN. L. REV. 853, 860 (1992) (explaining that, in the early tobacco litigation, POs prevented plaintiffs’ lawyers from “collaborat[ing] or realiz[ing] economies of work-product”); Karen E. Meade, Commentary, Breaking Through the Tobacco Industry’s Smoke Screen, 17 J. LEGAL MED. 113, 119 (1996) (explaining that, in the first wave of the tobacco litigation, cigarette companies imposed massive costs on their under-resourced adversaries by obtaining POs “to prevent the release of information to other potential plaintiffs”—and asserting that, due to these POs, in each new case, “new discovery had to be undertaken”).

30 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

is a good unto itself, important to the promotion of judicial accountability and essential to democratic processes. 118 As Senator Herb Kohl of Wisconsin once put it, “the courts are charged with doing the public’s business and pursuing the public interest—not just the interests of individual litigants before the courts.”119 As vindicators of the public interest, courts should meaningfully interrogate all efforts to restrict public access—including not just judicial procedures but also the fruits of discovery processes. 120 Fifth and finally, in what is perhaps less an argument than an anti-argument, reformers cast doubt on critics’ claims that the reforms they champion, if enacted, would cause the sky to fall. An example: As we detail below, critics warn that, if reforms were to be enacted, limits on POs would intensify acrimony, complicate and prolong the discovery process, and ultimately clog courts.121 Reformers do not buy these or other dark predictions. Illustrating: Legendary jurist Abner Mikva 118

See Judith Resnik, The Privatization of Process: Requiem for and Celebration of the Federal Rules of Civil Procedure at 75, 162 U. PA. L. REV. 1793, 1835–36 (2014) (explaining that courts serve—and must serve—“as a site of democratic practices”). 119

The Sunshine in Litigation Act, Hearing on S.1404 Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice, 103rd Cong. 5 (1994) (statement of Sen. Herbert Kohl); see also Bryant Testimony, supra note 107, at 56 (“I think the bottom-line question here is whether the courts are designed simply to resolve private disputes without regard for their effect on the public, or whether the public is the one that not only funds the system, but should determine what happens in it.”); The Sunshine in Litigation Act, Hearing on S.1404 Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice, 103rd Cong. 40–41 (1994) (statement of Hon. Abner J. Mikva) [hereinafter Mikva Testimony] (“I think that many scholars, and lawyers, and even judges forget that the courts are public institutions. They talk about privacy interests and the importance of respecting consensual positions as if the only two parties in interest in the court system are the parties to the lawsuit. I have never been able to understand how we could justify the heavy expenditure of public funds and resources on the courts if the only interest to be served is that of the litigants.”). 120

See David Luban, Settlements and the Erosion of the Public Realm, 83 GEO. L.J. 2619, 2657 (1995) (articulating a “public-life conception” of the courts and describing information-generation as “a fundamental public interest” of litigation, not a side effect); Owen Fiss, The Forms of Justice, 93 HARV. L. REV. 1, 29 (1979) (“[C]ourts exist to give meaning to our public values, not to resolve disputes”). 121

See infra notes 132–133 and accompanying text.

31 Electronic copy available at: https://ssrn.com/abstract=4811151

has argued in Senate testimony that restrictions on POs could accelerate (rather than complicate) litigation because, if a defendant cannot count on an expansive PO, the defendant may be more likely to proffer a generous settlement right out of the gates, even prior to the start of discovery.122 2. Critics’ Arguments Opposing Reform Efforts Numerous scholars, practitioners, and policymakers see matters differently—and vehemently oppose state and federal reform activity. Critics’ arguments tend to fall into one of five buckets. First, critics have consistently and powerfully voiced an “if it ain’t broke, don’t fix it,” argument.123 This argument has a few flavors. In its first guise, opponents challenge reformers’ claim that overbroad POs have concealed health hazards from public attention. Seizing this baton, Arthur Miller, for instance, wrote in a classic 1991 Harvard Law Review piece: “The allegation that protective orders are concealing information important to public health and safety obviously should arouse concern, but its validity is doubtful,” and, later in the piece doubled down, concluding that “no evidence has been presented that the current [PO] practice has created significant risks to public health or safety.”124 Likewise, in 1994, in a Senate Hearing to assess a Sunshine in Litigation Act, a witness from Lawyers for Civil Justice reassured the Committee: “[P]rotective orders are not preventing the public from obtaining information needed to protect[] public health or safety, as has been alleged.”125 122

See Mikva Testimony, supra note 119, at 76. For similar analyses, see Scott A. Moss, Illuminating Secrecy: A New Economic Analysis of Confidential Settlements, 105 MICH. L. REV. 867, 874 (2007). 123

Robert Weiner, Being Sued Doesn’t Mean Being Stripped of Privacy: Some Plaintiffs’ Lawyers Want the Right not only to Delver into Defendants’ Files, But Also to Publicize Whatever they Find, THE RECORDER (Feb. 8, 1990). 124

Miller, supra note 35, at 478, 501. See also Arthur R. Miller, Private Lives or Public Access: The Debate Over Courthouse Confidentiality, 77 A.B.A. J. 65, 67 (1991) (“There is simply no reason to believe that current court rules and practices create any risks to public health or safety. Indeed, all indications are that the current system works rather well.”). 125

The Sunshine in Litigation Act, Hearing on S.1404 Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice, 103rd Cong. 92 (1994) (statement of Alfred W. Cortese on Behalf of Lawyers for Civil Justice); see also Richard J. Vangelisti,

32 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

Piling on, defense lawyer Robert N. Weiner, a longtime Sunshine in Litigation Act critic, has likewise gone on record dismissing reformers’ argument that POs have “conceal[ed] information critical to public safety” as “contrived.”126 Second, in another spin on the “if it ain’t broke” theme, critics maintain that reformers’ concerns about judicial rubberstamping are also overstated—even chimerical. In critics’ view, the PO status quo is working as intended: POs are rarely issued, and, when they are issued, they reflect carefully exercised judicial discretion. On the former, Rep. Chris Cannon of Utah argued when opposing the Sunshine in Litigation Act of 2008: “This bill is unnecessary because discovery protective orders are rare.”127 On the latter, Professor Richard Marcus has argued that, under the law, judges must apply Rule 26(c)’s “principles with care” and that “[j]udges generally seem to be doing just that.”128 Arthur Miller has likewise insisted: “The ‘good cause’ requirement is [already] strict. Federal courts have interpreted the rule to mean that the party seeking confidentiality must make a particularized factual showing of the harm that would be sustained if the court did not grant a protective order.”129 Proposed Amendment to Federal Rule of Civil Procedure 26(c) Concerning Protective Orders: A Critical Analysis of What It Means and How It Operates, 48 BAYLOR L. REV. 163, 175–76 (1996) (“[E]mpirical data does not support the conclusion that protective orders are a threat to public health and safety.”); Marcus, supra note 101, at 464 (“Despite the widely publicized instances of supposed cover-ups of hazards, hard data is generally lacking and the critics’ broader assertions about widespread harm may be validly questioned.”). 126

Weiner, supra note 123.

127

Sunshine in Litigation Act of 2008: Hearings on H.R. 5884 Before the H. Comm. on the Judiciary, Subcomm. on Commercial and Administrative Law, 110th Cong. 35 (2008) (statement of Rep. Chris Cannon). 128

Marcus, supra note 101, at 506.

129

Miller, supra note 35, at 433. See also, e.g., LAWYERS FOR CIVIL JUSTICE, SEALING FATE: THE PROPOSAL TO RESTRICT JUDICIAL DISCRETION OVER SEALING CONFIDENTIAL INFORMATION WOULD IMPOSE UNWORKABLE STANDARDS ON THE COURTS, CONFLICT WITH STATUTORY PRIVACY RIGHTS, AND STOKE UNPRECEDENTED SATELLITE LITIGATION 1, 4 (2021) (insisting that judges do not rubberstamp motions for POs); Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 529 F. Supp. 866, 889 n.40 (E.D. Pa. 1981) (“[W]e doubt that any judge would approve a consent order not demonstrably rooted in Rule 26(c) . . . .”); Doré, supra note 113, at 302 (questioning “claims that

33 Electronic copy available at: https://ssrn.com/abstract=4811151

Later, Miller asserted that, “[u]nder existing law, the courts have discretion to accept or reject” POs, and, in operation, “they exercise” this discretion responsibly “on a case-by-case basis.”130 Third, those who oppose reform predict that limiting POs would impose high litigation costs on parties and the public. In particular, critics argue that broad protections on the disclosure of discovery material facilitate the free flow of information—and, in so doing, reduce inter-party acrimony and promote the efficient resolution of disputes.131 The converse is also true; without POs, critics reason, discovery would grind to a halt.132 In the words of one defense lawyer federal district courts have perfunctorily acceded to a plethora of stipulated requests for discovery protective orders”). 130

Miller, supra note 35, at 436. asserted:

In another publication, Miller

When information possibly implicating public health and safety surfaces in documents produced in litigation, the decision about whether it should be released to the public should rest where it always has—within the sound discretion of the court. Only the trial judge has no axe to grind and no prospect of pecuniary gain. Existing rules and procedures are more than adequate to accomplish this end. Arthur R. Miller, Private Lives or Public Access: The Debate Over Courthouse Confidentiality, 77 A.B.A. J. 65, 67 (1991). 131

See, e.g., Richard L. Marcus, Myth and Reality in Protective Order Litigation, 69 CORNELL L. REV. 1, 21–23 (1983) (suggesting that greater access to POs reduces resistance to discovery); Benham, supra note 96, at 430 (“The reasoning goes that without some confidentiality, litigants would zealously resist producing relevant information and settle fewer cases.”). 132

Vangelisti, supra note 125, at 178 (“[R]ather than facilitating the plaintiffs’ cases, the lack of protective orders would grind discovery to a halt and increase the costs of litigation.”); Miller, supra note 35, at 483 (“Limiting the availability of protective orders makes the discovery process more contentious, protracted, and expensive.”); Sunshine in Litigation Act of 2008: Hearings on H.R. 5884 Before the H. Comm. on the Judiciary, Subcomm. on Commercial and Administrative Law, 110th Cong. 38 (2008) (statement of Prof. Arthur Miller) (“Confidentiality is of paramount importance during discovery because the willingness of the parties to produce information voluntarily often hinges on a guarantee that it will be preserved. Remove this guarantee and discovery will become more contentious, requiring frequent court intervention.”); Examining the Use of Secrecy and Confidentiality of Documents by Courts in Civil Litigation, Hearing Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice,

34 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

at a Senate Hearing: “[I]f the defendant is not sure that its secret formula is going to be protected in discovery, then what’s going to happen? It’s going to fight. It’s going to fight producing it, and that takes time and resources of the court and the parties.”133 Fourth, some critics argue that any effort to cut down on litigation secrecy (via beefed-up restrictions on POs) will predictably backfire. There are, again, two flavors to this critique. First, in an age when arbitration is seemingly ascendant, some fret that, without blanket POs, litigants will simply opt out; they will privatize their disputes, avoiding courts altogether. Giving voice to this concern, Professor Richard Marcus has written: “[O]pening up the discovery process could have the ironic effect of deterring claimants from seeking relief in court in order to avoid the resulting publicity.”134 On this thinking, some secrecy (in the form of a PO) is perhaps less than ideal, but it’s better than a sharper turn away from judicial administration. Relatedly, some others argue that, particularly in the current age of Federal Rule of Civil Procedure 26(b)(1)’s proportionality review, without 101st Cong. 186 (1990) (written responses from The Hon. Joseph F. Weis Jr.) (“Without the availability of protective orders, the time for disposing of litigation would unquestionably increase, and the public would be poorly served by the additional delay.”); Examining the Use of Secrecy and Confidentiality of Documents by Courts in Civil Litigation, Hearing Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice, 101st Cong. 190–91 (1990) (statement of Prof. Arthur R. Miller) (“If people couldn’t voluntarily agree on confidentiality . . . . litigants would then be given an incentive to engage in trench warfare not to reveal the propriety, the important, the private.”); Marcus, supra note 128, at 484–85 (contending that “presumptive public access would disrupt orderly pretrial preparation by fomenting opposition to broad discovery” and “disrupt the cooperative exchange of information between the parties”). 133

See, e.g., The Sunshine in Litigation Act: Does Court Secrecy Undermine Public Health and Safety, Hearing Before the S. Comm. of the Judiciary, Subcomm. on Antitrust, Competition Policy and Consumer Rights, 110th Cong. 10 (2007) (statement of Robert N. Weiner). 134

See Marcus, supra note 101, at 486; see also The Sunshine in Litigation Act, Hearing on S.1404 Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice, 103rd Cong. 29 (1994) (statement of Hon. Patrick E. Higginbotham) (explaining that, if POs were sharply cabined, “the increased discovery contests would . . . add to the pressures that encourage some parties to pursue nonpublic means of dispute resolution”).

35 Electronic copy available at: https://ssrn.com/abstract=4811151

POs, some courts might be more inclined to short-circuit discovery and deny a party the right to information, full stop.135 If that’s right—and the choice is between more discovery (but subject to a PO) versus less discovery (but not subject to a PO)—it’s not obvious, if you value transparency, that the latter is a better bet.136 Fifth and finally, critics argue that any effort to restrict POs in order to further the public interest is fundamentally misguided because courts simply do not exist to advance the public interest; they exist exclusively to resolve private interparty disputes.137 And any extra review of POs, particularly when those orders are jointly requested, diverts courts’ attention from this fundamental purpose.138 On this 135

In re Halkin, 598 F.2d 176, 195 (D.C. Cir. 1979) (“The only plausible alternative to a protective order may be the denial of discovery altogether.”); Miller, supra note 35, at 476 (“[I]f judges’ discretion to issue protective orders is undercut, the courts’ only means of maintaining privacy might be to deny discovery altogether.”); id. at 484 (“[C]ontrary to the hopes of the proponents of public access, the net effect of banning protective orders might well be a constriction in the flow of litigation information, not an expansion.”); Jack H. Friedenthal, Secrecy and Civil Litigation: Discovery and Party Agreements, 9 J.L. & POL’Y 67, 97 (2000) (predicting that if courts “cannot avoid problems through the use of protective orders, they are more likely to do so by curtailing discovery at the outset”); cf. Seattle Times Co. v. Rhinehart, 467 U.S. 20, 34 (1984) (noting that “it is necessary for the trial court to have the authority to issue protective orders conferred by Rule 26(c)” to offset the “liberality of pretrial discovery”). As an aside but an important one: We are skeptical of the assumption that underlies this argument. If the Rule 26(b)(1) proportionality standard would warrant the declination of discovery due to disclosure concerns, surely that fact alone establishes good cause for a PO to prevent disclosure—meaning that, even if Rule 26(c) were fortified in line with reformers’ desires, the “problem” opponents worry about still wouldn’t materialize. For a discussion of the proportionality standard and the many normative judgments that its application embeds, see generally Jonah B. Gelbach & Bruce H. Kobayashi, The Law and Economics of Proportionality in Discovery, 50 GA. L. REV. 1093 (2016). 136

Again, as we explain at note 135, we believe that this argument relies on a misapplication of Rules 26(b)(1) and 26(c). 137

E.g., Weiner, supra note 34 (“The courts’ job in civil cases is to resolve disputes between private parties.”). 138

Miller, supra note 35, at 431 (noting that a focus on public access would divert courts “from their primary mission” of resolving disputes among litigants); Marcus, supra note 101, at 470 (“The primary purpose for which courts were created . . . is to decide cases according

36 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

view, any discussion of POs and the public interest simply misses the point. If there is no dispute over POs, “then courts should readily accede to the parties’ mutual desire for confidentiality.”139

C. Why Debate Has Stalled: Dueling Anecdotes and Anemic Empirics As noted, the debate about POs has raged for more than thirty years, although, in that time, the arguments themselves have remained static. Reformers and traditionalists are covering essentially the same territory now as they covered in the early 1990s. One reason the debate has fizzled is entirely explicable. Reformers and critics sharply disagree about first principles— and about the proper role of civil litigation. Does private civil litigation exist merely to solve parties’ discrete disputes?140 If so, POs, which advance and perhaps streamline private dispute resolution, are entirely unobjectionable. Or should private litigation advance broader, public-regarding aims?141 If so, POs subvert those aims and should be curtailed. That debate, which strikes right at the heart of what litigation does and is, won’t be advanced by empirics. Another big reason the PO debate has stalled is also fairly explicable. The debate is running aground on the shoals of warring—but unresolved—predictions about how a restriction on POs would ripple through the broader litigation landscape. Reformers and opponents disagree concerning whether tighter limits on POs’ entry would lead to greater

to the substantive law. The collateral effects of litigation should not be allowed to supplant this primary purpose.”). In addition to all the above, some add a procedural critique: that any change to Rule 26 ought to come from the Rulemaking process, rather than from Congress. For a discussion, see Vangelisti, supra note 125, at 172–73. 139

Doré, supra note 113, at 289; cf. Erichson, supra note 54, at 360 (“A light standard of good cause for discovery confidentiality reduces the burden on the court . . . .”). 140

See supra notes 137–139 and accompanying text (advancing this conception). 141

See supra notes 118–120 and accompanying text (advancing this conception).

37 Electronic copy available at: https://ssrn.com/abstract=4811151

efficiency and more disclosure or, instead, would backfire and have (on balance) the opposite effect.142 A third fundamental reason the debate has stalled, however, is harder to explain. The debate has sputtered because there is not only a dispute about the fundamental role of civil courts, or about what would happen if Rule 26(c) were to be fortified. There is also a dispute about what is happening, in the courts, today. As the above accounting demonstrates, there exists a sharp and persistent disagreement about the status quo—and it is very hard to know whether a system should be reformed, or how reforms should be designed or implemented, without first understanding how that system currently operates. On this score, there are, and there have long been, striking disagreements about basic facts that go right to the heart of the PO debate. These include: (1) how often POs are issued, (2) whether POs have actually shielded public health hazards from scrutiny, (3) the prevalence of stipulated (rather than opposed or contested) POs, and (4) arguably most importantly, whether courts currently subject stipulated POs to any meaningful scrutiny. Fueling this uncertainty is a striking lack of empirical research assessing these basic how, when, and why questions. Thus, as Professor Seth Endo recently observed, “[n]either the academic literature nor the jurisprudence grapples much with the on-the-ground practice of stipulated protective orders.”143 Indeed, we know of only three studies that address these matters. None completely answers the critical question of how Rule 26(c) is (or more often, isn’t) appropriately applied. And each is limited by data hurdles that we are able to surmount. We describe each below. To start, in 2020, Seth Katsuya Endo of the Seattle University School of Law published a piece in which he evaluated 100 proposed stipulated POs presented to district courts across the country in the single month of January 2018. Endo calculated how many of the proposed orders were granted, and among granted orders, he analyzed whether the judge modified the proposed order before its entry. Ultimately, Endo found that only 5 of the 100 proposed POs in his dataset were denied, and, studying the text of the 95 approved orders, that only 32 “described specific types of 142

Compare supra notes 114–117 and accompanying text, with supra notes 131–54 and accompanying text. 143

See Endo, supra note 33, at 1275.

38 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

information to be protected or harms that would follow from public disclosure.”144 Many of the others merely copied model orders or parroted the list of confidential information from Rule 26(c).145 Surprisingly too, Endo found that 15 of the 95 motions for POs that were ultimately granted contained a significant mistake of law—including that the parties confused the legal burden for obtaining a PO and for filing material under seal—and that judges only rarely corrected this (elementary) mistake.146 And, “in none of the cases did the docket reflect that any entity actually intervened to . . . challenge a confidentiality designation.”147 Further, while it is not the core contribution of his paper, Endo’s analysis implicitly estimated that stipulated POs were requested in, at most, 1.15% of all civil cases, or fewer than 4,000 cases per year.148 Endo’s analysis offers a significant contribution and suggests that courts too readily rubber-stamp half-baked stipulated POs. But his limited sample size (of only 100 motions for POs) limits the force of his findings. And, as we explain below, we believe that estimating the number of stipulated POs based on searching Bloomberg for a single month leads to a substantial undercount of their general incidence.149 More recently, in a student note, Chelsea Hanlock analyzed 595 cases brought against New York City Police officers in the Eastern and Southern Districts of New York between 2014 and 2018.150 In her sample, 139 cases had stipulated POs.151 Evaluating the text of proposed and granted stipulated POs and the thoroughness of judges’ analysis, Hanlock, like Endo, found that denials were rare; she found a 144

Id. at 1277, 1286.

145

Id.

146

Id. at 1288–89. For how the two mechanisms differ, see notes 36 through 46 and accompanying text. 147

See Endo, supra note 33, at 1279.

148

See id. at 1276 n.161 (inferring that there are roughly 330 stipulated POs each month (or approximately 4,000 stipulated POs each year) because that is how often relevant search terms appear in available dockets). 149

See infra Subpart III.B and accompanying text.

150

Hanlock, supra note 22, at 1532–34.

151

Id. at 1538.

39 Electronic copy available at: https://ssrn.com/abstract=4811151

denied-in-full rate of less than 1% and a denied-in-part rate of 2.9%.152 Like Endo, she also found that most judges’ review was cursory. In her words, most judges did not “engage in a rigorous good cause analysis.”153 For example, most orders justified confidentiality, at least in part, because “the City . . . would not produce requested documents” unless a PO was entered; only one submission made a particularized showing.154 Thus, Hanlock’s study, like Endo’s, supports the rubber-stamping idea. But, given its small sample size, its findings are suggestive, rather than firm—and, although Hanlock zooms in on the important issue of police misconduct in New York, she cannot speak to other types of claims, defendants, or locations. A third and final study, conducted for the Advisory Committee on Civil Rules, is bigger, but its data are also more than thirty years old.155 In this study, Elizabeth Wiggins and co-authors evaluated a random assortment of roughly 40,000 civil cases involving POs in three federal district courts from 1990 to 1992.156 The prevalence of civil cases with PO activity varied from 5% to 10% across the three districts and years studied.157 Of cases with PO activity, a strong majority of motions for POs were contested; only 17% to 26% of cases with POs had motions for stipulated POs (with variation across districts).158 Crunching Wiggins et al.’s numbers yields the conclusion that 1.3% of federal cases involve motions for stipulated POs—i.e., that stipulated POs are very rare.159

152

Id. at 1542.

153

Id. at 1540–42.

154

Id. at 1540–41.

155

ELIZABETH C. WIGGINS, MELISSA J. PECHERSKI, & GEORGE CORT, PROTECTIVE ORDER ACTIVITY IN THREE FEDERAL JUDICIAL DISTRICTS: REPORT TO THE ADVISORY COMMITTEE ON CIVIL RULES 3 (1996). 156

Id. at 1, 3.

157

Id. at 3.

158

Id. at 4–5. About 75% came by opposed motions, with the residual initiated sua sponte by the court. Id. 159

Their Table 1 reports the total number of cases studied for each district and year in the study, as well as the percentage of these that had PO activity; from these data we were able to calculate that 6% of all cases had PO activity. Id. at 3. Numbers reported in their Table 3 indicate that of 927 cases the authors studied more closely, there were 207 stipulated agreements by the parties, which is 22%. Together, the

40 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

Wiggins et al. further found that courts denied 30% to 45% of POs, although courts only rarely denied stipulated POs.160 Soon after this study was issued, its core findings—that courts frequently deny POs and that parties rarely file joint motions for POs—were enthusiastically trumpeted by critics of PO reform.161 While the above efforts contribute to the debate, their limitations soften their empirical punch. The academic studies each evaluate only a small sample of POs in a sample restricted by court or year, gaining analytic granularity at the expense of generalizability. Similarly, the Wiggins study covers only three courts—and is, bluntly, old.162 6% and 22% figures imply an estimate that 1.3% of cases considered by Wiggins et al. had stipulated protected orders. 160

Id. at 6. Courts rejected, on the record, only 1% of the stipulated POs in the study. Id. at 4, 6. The researchers proffered an explanation for these infrequent denials: “[P]arties [may] discuss with the court whether a protective order is warranted and what provisions should be included before a formal agreement is presented.” Id. at 6. Only as an “alternate explanation” did the researchers consider that “judges are reluctant to reject an agreement between opposing parties.” Id. 161

E.g., Doré, supra note 113, at 302 (stating that the Wiggins “study does not support claims that federal district courts have perfunctorily acceded to a plethora of stipulated requests for discovery protective orders or that such orders create significant hazards to public health and safety”); Sunshine in Litigation Act of 2009: Hearing Before the Subcomm. on Commercial and Admin. Law of the Comm. on the Judiciary, 111th Cong. 59 (2009) (statement of Hon. Mark R. Kravitz for the Judicial Conference of the United States) (pointing to the Wiggins study to support the notion that reform is not necessary because, as it is, “judges den[y] or modif[y] a substantial proportion of motions for protective orders”). Indeed, those opposed to reform continue to cite the Wiggins study, despite its advanced age. E.g., Lawyers for Civil Justice Comment to the Advisory Comm. on Civil Rules, Mar. 24, 2021, at 4 [hereinafter LCJ 2021 Comment] (citing the Wiggins study and declaring that “[t]he FJC’s report also shows that the number of orders protecting . . . documents . . . is also small”). 162

Also, potentially of note is a recent study involving orders to seal (POs’ close cousin). In particular, Reuters recently analyzed Westlaw data from 3.2 million civil suits filed in federal courts between 2006 and 2016 and also conducted a more detailed review of the 115 largest “mass product-liability actions from the past 20 years.” REUTERS, How We Did the Data Analysis, https://www.reuters.com/investigates/special-report/usa-courtssecrecy-how/ (last visited Jan. 5, 2023). The authors found that “judges allowed litigants to seal material in at least 65 percent of productliability actions.” Id. This study is useful but sheds little light on POs

41 Electronic copy available at: https://ssrn.com/abstract=4811151

In sum: For more than thirty years, the propriety of POs has been hotly debated. But the debate has stalled—in large part, because even basic questions concerning the entry of POs have defied resolution. Without these facts, the debate has devolved into warring rhetoric, punctuated by confident but unsupported claims concerning, among other things, the prevalence of POs, the prevalence of stipulated POs, and the rigor of judicial review. Below, drawing on docket reports from over 2.2 million cases, we present robust empirical evidence to reset the empirical terrain. III. DATA: NEW EVIDENCE ON STIPULATED PROTECTIVE ORDERS The above discussion shows that PO practice is frequently debated but rarely studied. In this Part, we apply sophisticated text-based machine learning tools to analyze an original dataset of more than 2 million federal dockets in civil cases filed from 2005 to 2012 in order to bridge persistent empirical gaps.163 Subpart A begins with a brief note on methodology. It discusses our dataset and outlines how we used machine learning and natural language processing tools to identify POs within federal dockets. For those interested, our Appendix offers additional methodological detail. Those uninterested in any discussion of methodology may choose to skip directly to the meat of our analysis. Subparts B, C, and D present our results. In particular, Subpart B provides information on trends and prevalence. It reveals that the fraction of cases with joint motions for POs more than doubles previous estimates. By our study period’s end in 2012, joint motions for POs were present in more than 8% of cases in which an answer was filed. Subpart C then considers judicial decisionmaking and estimates how often and also focuses on judicial decisionmaking in the largest product liability MDLs. 163

With respect to text-based tools, we extensively use natural language processing, the branch of machine learning that performs text analytics. While no data extraction technique is perfect, our strategies yield high accuracy rates, particularly when compared to the standard keyword approaches used in existing literature. For further details concerning both our data and methodology, see the Appendix.

42 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

judges deny litigants’ joint PO requests. Here, consistent with Endo and Hanlock, we find that judicial denials are rare. But, more than that, our judge-level statistics reveal something astonishing: A majority of federal district court judges in our sample—all of whom considered at least twenty-five stipulated motions for POs—never rejected a stipulated PO during the period of inquiry. This fact undermines the oftrepeated claim that federal judges individually scrutinize each PO that they enter and suggests that the entry of stipulated POs is not careful but, rather, cavalier.164 Finally, in Subpart D we supplement our quantitative investigation with a qualitative assessment of 300 granted and 100 denied stipulated POs. Analyzing these orders, we find that the high rate of denials unearthed in our big data analysis appears to reflect rubberstamping on the ground, lending still further support to reformers’ concerns.

A. A Note About Methodology Before turning to our analyses, we briefly describe our dataset and how we leveraged machine learning and natural language processing tools to sift through millions of docket sheets. Our dataset includes docket-sheet information on case activity between January 1, 2005 and December 31, 2014, and includes more than 2.2 million cases filed between 2005 and 2012. These data were acquired from Thomson Reuters roughly a decade ago and have been used in several other studies.165 The dataset contains docket-level information on virtually all civil actions filed in the federal district courts between January 1, 2005 and December 31, 2014. To our knowledge, it is the most up-to-date, and easily the most comprehensive, federal district court dataset available for research currently in existence.166

164

See supra note 128–130 (repeating that claim).

165

See, e.g., Jonah B. Gelbach, Material Facts in the Debate over Twombly and Iqbal, 68 STAN. L. REV. 369, 389–93 (2016); Jonah B. Gelbach, Rethinking Summary Judgment Empirics: The Life of the Parties, 162 U. PA. L. REV. 1663 (2014); Jonah B. Gelbach, Beyond Transsubstantivity (working paper). More details about the data appear in the Appendix. 166

Comprehensive district court dockets are notoriously hard to amass. This difficulty stems, in part, from the fact that most dockets sit behind a paywall. See David Freeman Engstrom & Jonah Gelbach, Legal

43 Electronic copy available at: https://ssrn.com/abstract=4811151

The sheer size of our docket dataset allowed us to overcome the challenges that other researchers have faced; our data is more recent and has nationwide geographic coverage.167 Our dataset is also the only one that contains virtually all civil cases filed during the period. That said, one limitation of our data is that we study only those cases that were filed between 2005 and 2012. While we are unaware of any work suggesting there was a dynamic shift in PO activity in the past twelve years, our data would not shed light on such patterns. Yet the dataset’s enormity also presents its own difficulty. Given the size of our dataset, when locating motions for contested or stipulated POs, we faced a massive “needle in a haystack” problem: Among nearly 80 million docket entries in our possession, only a small fraction addressed POs, creating the question of how to find what we were looking for. We overcame that challenge using a fourstep procedure that included both simple text processing and machine learning methods. First, we retrieved the full text of all docket entries that contained the string “PROTECTIVE ORDER” (text was all capitalized in our underlying data); we refer to the full set of all such entries as the “matching set.” We suspect that virtually all POs in our dataset contain this phrase. However, this query produced a high number of false positives—docket entries in our matching set that included “PROTECTIVE ORDER” but did not directly involve the docketing or requesting of a PO.168

Tech, Civil Procedure, and the Future of Adversarialism, 169 U. PA. L. REV. 1001, 1063 (recounting difficulties in creating large datasets of federal dockets because of PACER fees); Jonah B. Gelbach, Free PACER in LEGAL TECH AND THE FUTURE OF CIVIL JUSTICE (David Freeman Engstrom ed., 2023) (similar). 167

As described elsewhere, the (CM/ECF) software used by district courts went into use at different times, and entries docketed in a district before that time may be less reliable. See Jonah B. Gelbach, Beyond Transsubstantivity, forthcoming in NYU Journal of Legislation & Public Policy. In our dataset, for instance, 4.08% of cases occur before the relevant district adopted the CM/ECF system. We treated these cases as missing data and used imputation methods to predict whether they have a stipulated protective order or not. Such methods leverage non-missing data to make educated guesses about missing data, and are conventional in social science literature. See Appendix at 5–6. 168

For example, a scheduling order could use the phrase “PROTECTIVE ORDER” by directing the parties that any requests for such orders are due by a certain date.

44 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

We then coded a random sample of 2,250 docket entries from the first step’s matching set. A docket entry was relevant to this coding if it was either (a) created by one or more parties to docket a motion for a PO, or (b) created to docket a court order resolving a motion for a PO. We labeled each docket entry text string to designate whether it concerned a PO. We categorized an order as a stipulated protective order (SPO) if either the motion seeking the PO or the order granting it had docket entry text calling it “stipulated,” “consented,” “agreed,” or “joint.” We coded all other POs as general protective orders (GPOs). Third, we trained a series of supervised machine learning models to predict which entries in our matching set involved SPOs or GPOs. This approach is typical of supervised machine leaning, with humans correctly coding a random sample of text and a computer algorithm then applied to this coded sample to “learn” which textual elements predict that a textual string represents a PO request. We used our supervised learning models to identify docket entries that involved (1) motions for stipulated or general POs, and (2) judicial orders about stipulated and general POs.169 Compared to benchmarks in the law and data science literature, our models performed with high accuracy.170 Fourth, we used textual pattern matching to identify when judges deny requests for stipulated or general POs.171 169

For details, including the pre-processing steps we took to clean the data, see Appendix at 3. 170

For instance, our classifier predicting stipulated PO achieves 98% accuracy and has an F1 score = .94. When compared to the common F1 = .7 referenced in existing legal research using similar methods, our method achieves a high level of accuracy. See also Briand D. Feinstein & Jennifer Nou, Strategic Subdelegation, 20 J. EMPIRICAL LEGAL STUD. 746 (2023) (discussing classifier performance); Julian Nyarko, Stickiness and Incomplete Contracts, 88 U. CHI. L. REV. 1, 32 n.10 (2021) (discussing F1 = .7 benchmark). For more details about classifier performance, see Appendix at 4. 171

Our approach to pattern matching used regular expressions, which involves matching any combination of a chosen string of text and other text. For example, one regular expression is “DECLIN+*”, which matches all of “DECLINE,” “DECLINES,” and “DECLINED” (the plus sign means “one or more characters,” and the asterisk character means “any character,” so “+*” means “at least one of any character,” and “DECLIN+*” means "the string ‘DECLIN’ followed by at least one more character). Regular expression searches are a common form of searching in text-as-data research. JUSTIN GRIMMER, MARGARET E.

45 Electronic copy available at: https://ssrn.com/abstract=4811151

This approach looked for sentences that contained a variant of the term “protective order” and some version of either “declines,” “denied” or “unable to adopt.” Taken together, these steps produced a comprehensive dataset of stipulated and non-stipulated motions for POs—and the judicial grant or (far less frequently) denial of each. A key advantage of the process just described is that can be done using only the text of each case’s docket entries, which allows large-scale quantitative analysis. To conduct our qualitative review of granted and denied SPOs, we needed to review the resolving orders filed in cases’ dockets. We took a random sample of 300 granted SPOs and 100 denied SPOs. And, after cleaning the data,172 a team of trained researchers retrieved the text of the underlying district court orders from PACER, manually reviewed each order against a set of agreedon criteria and tracked results in a shared spreadsheet. After data collection, a trained researcher performed random spot checks on 10% of the sample (30 grants and 10 denials) to ensure that orders were tagged consistently. Finally, for denied stipulated POs, we manually searched docket reports for other references to POs to determine whether a PO was ever entered in the case. As a final methodological note, we considered two phenomena that would undermine our conclusions. First, we assessed whether our results are likely driven by the use of model POs created by district courts. If so, one might think our results would capture district-level, rather than case- and judge-level, features of stipulated PO practice.173 But we

ROBERTS & BRANDON STEWART, TEXT AS DATA: A NEW FRAMEWORK FOR MACHINE LEARNING AND THE SOCIAL SCIENCES 178–83 (2022) (discussing dictionary methods). 172

We excluded, for example, HIPAA protective orders which are granted pursuant to a statutory scheme. See 45 C.F.R. § 164.512(e). We also excluded dockets that were outside of our observation period (e.g., orders entered before 2005), those that did not have an attached order, or (in the rare circumstance) those that were false positives (e.g., non-stipulated). 173

Some district courts publish model POs for parties appearing before the court. See, e.g., U.S. Dist. Ct., N. Dist. Cal., Model Stipulated Protective Order 6, https://www.cand.uscourts.gov/wpcontent/uploads/forms/model-protectiveorders/CAND_StandardProtOrd.Feb2022.pdf. See also infra Subpart III.D (surveying model POs).

46 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

found little evidence to support such a proposition.174 Second, we considered the possibility that judicial analysis related to the good cause standard might be present in judicial opinions filed alongside the granted orders themselves, because our main empirical approach does not address such opinions. We searched for the word “opinion” in docket entries that mentioned stipulated POs and found that the fraction of entries that included this word was only a small fraction of 1%. We conclude that neither the presence of model orders nor the possibility that judicial analysis appears in separate opinions is a plausible threat to our conclusions. We now present the novel descriptive insights generated by this massive empirical effort.

B. Motions for Stipulated Protective Orders: Time Trends and Prevalence We first consider a key, but heretofore unanswered question: Are motions for stipulated POs a common feature of contemporary federal litigation? As discussed in Part II, no prior study answers this question using comprehensive data. With some back-of-the-envelope calculations, Endo estimated that stipulated POs featured in roughly 1% of federal civil litigation, fewer than 4,000 federal cases a year.175 The 174

This question is difficult to test systematically because: (1) model orders might not describe themselves as such, and (2) districts’ model orders, where they exist, might have changed over our study period. We used two parallel approaches to test model orders’ prevalence. First, we searched all text-searchable (roughly 80% of the orders were searchable) stipulated PO grants collected in our qualitative review for the term “model.” No orders used that word in connection with a model PO, although a handful of cases used the phrase “standing protective order” or “standing order” in a context that suggested they were model orders. All in all, our search of the text-searchable orders showed little reason to think model orders are prevalent enough to play an important role in our results. Second, we selected nine district-year combinations that appeared in our set of 300 granted stipulated POs (there were 2-4 orders in each of the nine district combinations). We looked for indicia of model orders, such as identical terminology, consistent formatting, or similar internal structure but did not find such evidence. Although it is difficult to draw definitive conclusions given the nature of the orders at issue, the results of these investigations strongly suggest that model orders are unlikely to constitute a substantial portion of the POs in our dataset. 175

Endo, supra note 33, at 1277 n.161. To arrive at this estimate, Endo used Bloomberg Law searches to track the number of times that the

47 Electronic copy available at: https://ssrn.com/abstract=4811151

Wiggins et al. research, performed under the auspices of the Federal Judicial Center and often touted by critics of reform, offered a similar estimate.176 Ultimately, as discussed below and shown in Figures 1, 2, and 3, we find that joint motions for POs, though relatively uncommon overall, are considerably more common than these earlier studies indicate. Figure 1 plots our estimates for the number of POs requested within two years of filing, by year.177 Figure 1 is notable in three respects. First, our results suggest that there are an average of 9,000 stipulated PO requests in the federal civil courts per year—with stipulated POs requested in 2.8% of all civil cases over our full study period. That figure may seem small, but it is more than double what others have previously suggested.178 Equally notable: The trend lines term “stipulated protective orders” was mentioned in federal dockets and found it was mentioned in 0.8% to 2.3% of cases (depending on the year). Id. at 1273. Because dockets typically will reflect both the filing and the court’s resolution of a motion for a PO, a count of all instances in which text related to a PO appears in a case’s docket will overstate the prevalence of PO requests—Endo suggests by a factor of 100 to 200%. Id. at 1273 n.150. This all means that stipulated POs appear, in Endo’s telling, in at most, 1.15% of civil cases in federal courts. 176

See supra notes 161–165. In those footnotes we explain that the Wiggins research suggested that stipulated POs were present in approximately 1.3% of civil cases. 177

We limit all our analyses to two years after filing date to account for the fact that our access to docket entry text extends only through December 31, 2014. For cases that terminate by that date, we observe all docketed activity. But for all other cases, we see only whatever docket activity happens before the end of 2014. To understand the problem, take, for example, a case filed on January 1, 2014. For that case, we observe exactly one year of activity, so if a stipulated PO were requested in the case’s second year, we would miss it. Comparing this 2014 case with only one year of docket activity to cases filed in earlier years would introduce measurement error due to the difference in the observable portions of the two cases’ life cycles. (Statisticians call this problem “right-censoring.”) To correct for this source of bias, we (1) analyze only POs requested within two years of filing and (2) therefore limit our analysis window to cases filed during calendar years 2005 and 2012. By doing so, we can make an “apples to apples” comparison across the years. Finally, all of our plots below discussing “requests” count only those instances in which judges, rather than parties, create a docket entry relating to a determination about parties’ request for a stipulated PO. This measure is methodologically conservative, because it excludes requests for POs that are mooted because of case termination or for other reasons (such as settlement). 178

See supra notes 175–176 and accompanying text.

48 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

show a steady increase over time. For cases filed in 2005, we estimate that roughly 8,500 stipulated motions for POs were filed. By 2012, this number had grown to nearly 12,500.179 Overall, this represents a roughly 50% increase during this period. Figure 1: Total Number of Requests for POs, by Filing Year of Case (2005-2012) Note: This figure counts requests, not cases with requests, e.g., if two requests are made in the same case, then each is counted in this figure.

One explanation for this observed uptick in joint Rule 26(c) motions is a possible uptick in the number of cases filed per year: if more cases were filed per year during this period, we would expect the number of joint Rule 26(c) motions to increase in rough lockstep. To a large extent, that is what we find. Figure 2 plots the percentage of civil cases with a joint motion for a PO filed within two years of the case’s initiation.180 The percentage bounces around over time, with a slight upward trend over the seven-year period between the filing of 2005-vintage and 2012-vintage cases in the percentage of cases with a joint motion for a PO. We regard this evidence as being at most weakly suggestive of a slight

179

Although it’s not the core contribution of this Article, we also find that general (non-stipulated) PO activity increased during this period. 180

We note that the unit of analysis in Figure 2 is the case, rather than the order. Thus, case A with one stipulated PO and case B with two stipulated POs both add one to the numerator of the data plot in Figure 2.

49 Electronic copy available at: https://ssrn.com/abstract=4811151

increase in the overall prevalence of SPOs.181 Still, even the lowest plotted percentage in Figure 2 more than doubles previous estimates.182 Figure 2: Percentage of Civil Cases with at Least One Stipulated PO Requested, by Filing Year of Case (20052012)

While the estimates offered thus far suggest that motions for stipulated POs are more prevalent in contemporary litigation than previously understood, our findings still might not fully capture the force of stipulated POs, nor the key trends. Why? The figures presented above look at all civil cases. But many civil cases in federal court— by some estimates, one-third or more—never reach discovery.183 And, discovery, of course, is the critical stage of litigation where concerns about POs kick in. To get a sense of how prevalent stipulated POs are in cases that reach the discovery phase (i.e., cases where POs possibly matter), Figure 3 plots the percentage of cases with a 181

The slope of a trend line fit using ordinary least squares is 0.0008, reflecting that the fitted value of the SPO percentage increased roughly half a percentage point (0.0008 times 7 is 0.0056—or 0.56 percentage points). The trend line is not very precisely estimated, though, which suggests that the evidence of an increase in the rate of stipulated POs during our study period is relatively weak. 182

See supra notes 175–176 and accompanying text.

183

See Jay Tidmarsh, Opting Out of Discovery, 71 VAND. L. REV. 1801, 1803 (2018) (“In a third or more of federal cases, no discovery occurs . . . .”).

50 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

joint Rule 26(c) motion (within two years of filing) in cases where a defendant has answered a plaintiff’s complaint. To be sure, discovery can start before the defendant answers the complaint. But in most cases, discovery battles do not commence until after the answer is filed.184 Figure 3 indicates, not surprisingly, that joint motions for POs are even more common in cases where a defendant answers the plaintiff’s complaint. And the share is growing over the study period. In 2005, 5.5% of such “post-answer” cases featured a joint motion for a PO. By 2012, that figure grew to 8.5%. This growth represents over a 60% increase— and several times the overall rate among all civil cases. Figure 3: Percentage of Cases with a Stipulated PO Request, Among Civil Cases with an Answered Complaint, by Filing Year of Case (2005-2012)

Finally, Figure 4 shows the proportion of POs requested that were stipulated POs. These data show that stipulated PO requests, once thought to account for 17 to 26% of protection motions,185 account for a much larger share. On average, we estimate stipulated POs accounted for a whopping 45% of all PO motions filed per year—a two to three times larger proportion than previously thought.

184

See Alexandra D. Lahav, Procedural Design, 71 VAND. L. REV. 821, 872 (2018) (describing the “textbook order of civil procedure” as proceeding sequentially from “filing, [to] motion to dismiss or answer, [to] discovery”); see also FED. R. CIV. P. 26(d)(1) (clarifying that “[a] party may not seek discovery from any source before the parties have conferred as required by Rule 26(f),” subject to certain requirements). 185

See supra note 158.

51 Electronic copy available at: https://ssrn.com/abstract=4811151

Figure 4: Percentage of Requested POs that Involved a Stipulated Request, by Filing Year of Case (2005-2012)

C. Rule 26(c) Motions: Grants and Denials We now assess how judges respond to these Rule 26(c) motions. As noted, reformers have long argued that stipulated POs are granted, pro forma, without adequate judicial scrutiny.186 Critics, meanwhile, have insisted the opposite. As the Judicial Conference of the United States put it when opposing reform efforts: When motions for POs are filed, “courts review such motions carefully and often deny or modify them to grant only the protective orders needed, recognizing the importance of public access to court filings.”187 Our data permit us to adjudicate this debate. Figure 5 reports the average denial rate for stipulated PO requests. In this plot (and all analyses that follow), we take a conservative view. We classify a motion for a stipulated PO as “denied”: (i) even if the motion is partly denied (i.e., even if the motion is granted in part), or (ii) even if the judge denies the parties’ motion not on the merits, but rather as moot. In so doing, we err on the side of reporting PO denials. Particularly given our generous definition of “denials,” the primary fact that stands out in Figure 5 is just how low the 186

See supra notes 108–111 and accompanying text.

187

Sunshine in Litigation Act of 2008: Hearings on H.R. 5884 Before the H. Comm. on the Judiciary, Subcomm. on Commercial and Administrative Law, 110th Cong. 74 (2008) (statement of Hon. Mark R. Kravitz for the Judicial Conference of the United States). For similar assurances, see supra notes 128–130 and accompanying text.

52 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

denial rate for stipulated POs is. For our entire sample, we find that judges deny, on average, only 4.1% of stipulated PO requests. In other words, roughly 96% of the time, when the parties file a joint motion for a PO, such a PO is granted. Figure 5: Share of Stipulated PO Requests Denied Per Year

What case types feature higher and lower denial rates? Parsing PO denials by case type, Figure 6 reveals modest variation. The highest denial rates (north of 6%) are in forfeiture cases.188 IP cases—where many believe POs are justified—have a slightly higher than average denial rate.189 Intriguingly, civil rights cases and tort cases produce some of the lower denial rates (on the order of 3.4 and 3.6%, respectively). This result is notable because both cases can involve facts and disputes with high levels of public interest.

188

Because forfeiture cases make up a small share of the set of cases with a stipulated PO, it is possible that their especially high denial rate may be attributable to statistical variation—essentially noise. 189

We caution that these analyses speak only to denial rates for stipulated POs. They are not necessarily a proxy for a judge’s willingness to permit secrecy writ large.

53 Electronic copy available at: https://ssrn.com/abstract=4811151

Figure 6: Percent of Stipulated POs Denied by Case Type

Next, we evaluate judicial behavior, and here we find something truly remarkable. We zeroed in on individual judges—but, to avoid judges who have considered too few stipulated POs which would create unreliable (i.e., “noisy”) estimates, we restricted our focus only to judges with substantial Rule 26(c) activity, which we defined as those who had ruled upon at least twenty-five motions for stipulated POs during the period of study.190 Evaluating only that subset of 727 district and magistrate judges, we found that fully 392 judges—54%—never once denied a stipulated PO request.191 Figure 8 makes the point graphically.

Because both district court judges and magistrate judges render decisions about stipulated POs, we couldn’t simply assign a given order to the district court judge presiding over the entire case. Instead, we built a nuanced regular expression to extract the judge’s name from the individual docket entry. Evaluating our extraction method via random samples of stipulated POs, we found that the method correctly identified 98 to 99% of judges issuing the stipulated POs. 190

191

Also remarkable, it appears that, in their PO-granting and -denial behavior, district court and magistrate judges don’t stay on script. Trial court opinions display sharp (and seemingly inexplicable) intra-circuit variation, and the opinions do not resemble what one might expect if trial courts were following circuit court precedent with fidelity. That is, some circuit court precedent is strict when it comes to POs, and some circuit court precedent is lax when it comes to POs, but trial judges in “strict” circuits do not appear to scrutinize POs more carefully than do trial judges in “lax” circuits. For further discussion of this dynamic, which raises serious questions about decisional consistency, adherence to precedent, and judges’ basic fidelity to law, see generally Peters et al., supra note 47.

54 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

Figure 7: Distribution of Judge-Level Stipulated PO Denial Rate

Notes: The bin size for the x-axis is 2.5 percent. The analysis is limited to judges who have decided at least twenty-five stipulated POs. The fact that over half of the judges in this analysis never denied parties’ joint request for a PO is remarkable. Opponents of PO reform have long argued that trial judges are in the best position to exercise ground-level discretion in determining whether a PO is warranted and, if so, what shape that PO should take—and they have long assured policymakers that, on a day-to-day basis, district court judges exercise this discretion with care and fidelity.192 To have so many judges with a zero percent denial rate raises questions about that assertion and implies that searching case-by-case scrutiny is not the norm.193 192

See supra notes 128–130 and accompanying text.

193

To be sure, although grant rates—in isolation—are probative of rubber-stamping, they aren’t necessarily proof of it. Caution is warranted, in part, because strategic litigation effects make it notoriously difficult for researchers to map the grant rate to how judges apply the law. See, e.g., Gelbach, Material Facts, supra note 165, at 389–93. Further, as Wiggins et al. notes, it is possible that part of what might drive the results is that “parties [may] discuss with the court whether a protective order is warranted and what provisions should be included before a formal agreement is presented.” WIGGINS ET AL., supra note 155, at 6. If judges and parties are regularly engaged in this kind of informal, invisible, undocketed exchange, then our review of the docket would show rubber stamping where there has been more consideration.

55 Electronic copy available at: https://ssrn.com/abstract=4811151

D. Qualitative Assessments To fortify our analysis, we supplemented our quantitative assessment with a qualitative evaluation of 300 granted and 100 denied stipulated POs randomly drawn from our docket dataset. Although these 400 cases represent a relatively small sample of orders, the sample is large enough to give us a helpful peek at what, if any, standards judges actually apply when ruling on requests for stipulated POs. The sample, from a range of district courts and civil case types, contains stipulated PO rulings entered between 2005 and 2014 in cases that were filed between 2005 and 2012.194 In performing this review, we were inspired by earlier work, including by Endo.195

Although we cannot rule out either effect, there are two reasons we are skeptical that informal channels explain our findings. First, our detailed qualitative review did not surface substantial evidence of such iterative collaboration. See infra Subpart III.D. Second, if informal discussions were an important factor in our results, we would expect to find sharply different denial rates in cases where there was only one docketed entry reflecting a stipulated PO request than in cases with multiple such docketed entries. Our reasoning is that, after parties heard from chambers that their initial joint PO request did not pass muster, and how to fix it, they would be expected to file a superseding, second request that would pass muster. So, if the informal-discussion hypothesis is importantly correct, we ought to see an especially low denial rate in cases where we identify two docketed motions for stipulated POs and one docketed order adjudicating those requests. In such cases, we found a denial rate of 7.7%—higher, not lower, than the overall average. Of course, there might be differences in cases with one versus multiple joint PO requests, so we do not wish to push the statistics in this footnote too far. Nevertheless, this pattern is inconsistent with the simple informaldiscussion story detailed above. 194

One other prefatory note: Eighteen orders in our random sample (6.0% of the draw) happened to be identical orders entered on the same day and by the same judge as part of an MDL. Given our commitment to the random draw, we included all eighteen orders in our analysis. 195

See generally Endo, supra note 33 (conducting a qualitative review of 100 granted stipulated POs).

56 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

Figure 8: Count of Tagged Stipulated PO Grants, by Year Filed and Circuit196

Figure 9: Count of Tagged Stipulated PO Denials, by Year Filed and Circuit

1. Qualitative Review of 300 Granted POs Evaluating each of these 300 stipulated POs, we first assessed how the PO applied Rule 26(c)’s good cause standard—and, as an initial matter, whether the PO actually used the term “good cause” when granting the order. In other words: Did the court’s order make any mention of the formal 196

The random pull did not include any cases from the D.C. Circuit because the D.C. district court (1) has a low caseload relative to other districts and (2) adjudicates very few stipulated POs.

57 Electronic copy available at: https://ssrn.com/abstract=4811151

requirement for granting a PO?197 A mere 32% of stipulated POs referred to “good cause” explicitly.198 Roughly twothirds of the orders failed to so much as mention the relevant legal standard.199 However, these figures do not on their face establish that judges abandoned their legal duties. For one, recall that two circuits explicitly permit judges to grant stipulated POs without finding good cause.200 In addition, when a judge signs a stipulated PO that makes a particularized good cause inquiry, that provides good evidence that the order satisfied 26(c) obligations in even the strictest of circuits. But the absence of discussion of the good cause standard, on the order itself, is merely probative, rather than conclusive, as to whether Rule 26(c) was followed. It remains possible that the judge carefully engaged with the Rule 26(c) standard in a way that our empirical analysis fails to discern. Perhaps, for example, the judge did so in a hearing, where no transcript was ever made, or via informal communication via clerks. Or parties might have justified their PO in other briefing to the court which would not show up in the requested or approved PO. Yet, with that caveat, it is highly curious that, in the three circuits that require judges to make a good cause finding even when the PO request is stipulated, only 24% of PO grants so

197

FED. R. CIV. P. 26(c).

198

See, e.g., Stipulation and Ord. Protecting Confidential Information; Granting Motion for Protective Order at 1, United States v. Arg Co., Civil Action No. 10-cv-311 (N.D. Ind. Apr. 1, 2013), ECF No. 47 (“In view of this stipulation, the Court finds that good cause exists . . . .”). 199

See, e.g., Stipulated PO at 1, Whitfield v. PEP Boys – Manny, Moe, & Jack, No. 13-cv-11070 (E.D. Mich. Oct. 23, 2013), ECF No. 23 (offering no justification for issuance of PO). Note, however, that a small number of orders (5.7%) mentioned possible injury stemming from disclosure without explicitly referencing “good cause.” E.g., Order Granting Motion to Enter Proposed Agreed PO, at 1, Humphrey v. Burgos, No. 2:06-cv-00045-APR (N.D. Ind. July 12, 2007), ECF No. 58 (specifying that “[party] contends that said documents . . . contain sensitive and proprietary information relating to its operations as a local government” and, if distributed, would disrupt “its ability to effectively carry out its police powers and activities”). 200

See Subpart I.C. (describing case law).

58 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

much as mentioned good cause.201 This is an even lower proportion than our full sample. Next, we explored whether the Rule 26(c) order justified the grant based on individualized facts or, in contrast, merely by recycling boilerplate language. Recall that, in at least some circuits, a “district court abuses its discretion where it makes neither factual findings nor legal arguments supporting the need for” a stipulated PO.202 Here, out of 300 stipulated POs, only 17% provided a 26(c) good cause determination tailored to the facts of the case.203 The remaining orders offered no analysis to support a good cause finding or included boilerplate language.204 This finding is

201

Id. This figure is based on analysis of 83 stipulated PO grants. An additional 9.6% of the orders had a particularized good cause statement but did not include the term “good cause.” In re Nat’l Prescription Opiate Litig., 927 F.3d 919, 929 (6th Cir. 2019) (quotation marks, citation, and alteration omitted). 202

203

For example, one order noted that a PO was justified because litigants would be discovering employees’ personnel files, and that discovery of those files would constitute an “unnecessary invasion of their privacy interests.” Agreed PO/As Modified at 1, Socite v. Steak ‘N Shake Operations, Inc., No. 1:06-cv-03888 (N.D. Ill. Jan. 23, 2007), ECF No. 41. Another noted that parties sought marketing and financial strategy documents that derive value from “not being generally known to other parties,” and would cause damage to the parties if revealed. Amended Stipulated PO at 2–3, Sparknet Commc’ns, L.P. v. Bonneville Int’l Corp., No. 1:05-cv-02677 (N.D. Ill. July 5, 2005), ECF No. 95. 204

For example, one order noted that “[d]isclosure and discovery in this action are likely to involve production of confidential, proprietary, or private information for which special protection from public disclosure and from use for any purpose other than prosecuting this litigation would be warranted.” Ord. by Magistrate Judge Granting Motion for Stipulated PO (as Modified) at 1, Navarrette v. TD Banknorth, N.A., No. 5:07-cv-02767-JW (N.D. Cal. Dec. 12, 2007), ECF No. 19. And at least one order, after offering a boilerplate good cause statement, emphasized that “[t]he Parties believe that good cause exists.” Order Pursuant to Stipulation: Discovery PO at 1–2, First Serv. Networks, Inc. v. First Serv. Maint. Grp., Inc., No. 2:11-cv-01897-LOA (D. Ariz. May 21, 2012), ECF No. 28 (emphasis added). This example hearkens back to the analogy comparing stipulated POs to proposed class action settlements that have support from all named parties—wherein both the Seventh Circuit and Arthur Miller agree that, as in the Rule 23(e) context, when it comes to Rule 26(c), party consent does not suffice. See supra notes 65 and 75 and accompanying text.

59 Electronic copy available at: https://ssrn.com/abstract=4811151

consistent with others’ research.205 And, as above, an even lower proportion of POs (12%) included an individualized good cause justification in the three circuits that explicitly require good cause findings when granting stipulated POs.206 Third, we considered the extent to which courts modified the POs proposed by the parties. Litigants seeking stipulated POs generally submit detailed proposed orders. Upon receipt, a judge may sign a proposed order as written, manually modify the proposal, or draft and sign her own version. For our purposes, this is relevant because if a judge modifies a PO, there is at least some clear evidence that the judge has scrutinized the order. In assessing whether judges modify, we used a very generous definition of “modification.” To be sure, we did not credit as modifications any situations like the one displayed in Figure 10, in which judges simply (1) added prefatory language noting that the court was granting the proposed order, (2) removed the word “proposed” from the order, (3) dated the document, or (4) signed the document. But we counted all other changes to the text of a proposed PO, even minor changes to grammar, as a modification.207 Still, even under this capacious definition, courts modified the text of less than 22% of all proposed stipulated POs.208 Judges did not modify any element of the party’s 205

Seth Endo, for example, found that 34% of sampled orders granting a stipulated PO described either (i) “specific types of information to be protected or [(ii)] harms that would follow from public disclosure” while the remaining 66% used “generic language.” Endo, supra note 33, at 1286. Our results differ from Endo’s, likely because we only counted the second of Endo’s two categories of orders: orders that explained what harms would befall litigants if documents were made public. For similar results, see Hanlock, supra note 22, at 1540 (noting that orders in her study “recognized good cause as the standard” but generally “did not engage in a rigorous good cause analysis”). 206

In other words, 12% of our 83 sampled stipulated PO grants from the Third, Sixth, and Seventh Circuits had an individualized justification. 207

To determine if a PO was modified, we looked for handwritten notes, text in a different font, and other indicia of modification in the entered order. Where the order was ambiguous on its face, we downloaded the original PO request from PACER and compared the two orders (requested and granted) side-by-side. 208

More specifically, we found evidence of judicial modification in 14.7% of orders. But we could not tag a further 6.7% of all orders

60 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

proposed stipulated PO in at least 78% of cases. In many cases, the court merely crossed out the word “proposed,” signed, and dated the order proposed by litigants. And, where courts did modify a PO, they often did so to clarify an unrelated issue: the standard for sealing court records.209 Figure 10: Example Unmodified PO210

To recap, our qualitative assessment of 300 stipulated POs reveals that most stipulated POs fail to mention Rule 26(c)’s “good cause” standard. Fewer offer an individualized assessment of the facts justifying the judge’s (implicit or explicit) finding that the standard has been satisfied. And, generally, as exemplified by the proposed-cum-actual order depicted in Figure 10, when judges grant the parties’ joint motion, the judge merely strikes out the word “proposed” and signs his or her name. Finally, we surveyed model POs promulgated by district courts—and, in so doing, found yet more evidence of rote and relaxed Rule 26(c) review. We initially identified eighteen federal district courts (out of the ninety-four districts) that have adopted district-wide model POs.211 Of those (often because the party’s initial proposal was not available on PACER), meaning that the modification rate could be as high as 21.3%. 209

See supra note 41–46 and accompanying text. In the sample, 59% of all modified POs were modified to, at least in part, clarify the standards for sealing documents. 210

The judge made no other changes to the proposed order. Amended Stipulated PO, Sparknet Commc’ns, L.P. v. Bonneville Int’l Corp., No. 1:05-cv-02677 (N.D. Ill. July 5, 2005), ECF No. 95. At least 79% of all stipulated PO grants were functionally identical to Figure 10. 211 Those courts are the Northern District of California, the Southern District of California, the District of Idaho, the Southern District of Indiana, the Northern District of Iowa, the District of Kansas, the District of Minnesota, the District of Nebraska, the District of New Hampshire, the Northern District of Ohio, the Southern District of Ohio, the Northern District of Oklahoma, the District of Oregon, the Northern District of Texas, the District of Utah, the Western District of Washington, the Northern District of West Virginia, and the Southern District of West Virginia. Citations follow in the ensuing footnotes.

61 Electronic copy available at: https://ssrn.com/abstract=4811151

eighteen models, only two facially require a judicial finding of good cause based on the circumstances of the case.212 Another two mention that the court found “good cause” in entering the order but do not require particularized justifications.213 The remaining fourteen do not even purport to make a good cause determination—and, indeed, seven do not mention the standard at all.214 Once again, these results are not dispositive.

212

Both model orders ask litigants to explain why the litigation calls for a PO. U.S. Dist. Ct., Dist. Kan., Protective Order 3, https://www.ksd.uscourts.gov/civil-forms; U.S. Dist. Ct., Dist. Ore., Model Two-Tier Protective Order 1 (2018), https://ord.uscourts.gov/index.php/164-attorneys/forms. 213

U.S. Dist. Ct., Dist. Idaho., Model Protective Order 8, https://id.uscourts.gov/district/forms_fees_rules/Civil_Forms.cfm (to locate, scroll down to “Litigation and Discovery Plan”); U.S. Dist. Ct., Dist. Utah, Standard Protective Order for Cases Filed Before December 1, 2023, at 1, https://www.utd.uscourts.gov/usdc-forms. 214

U.S. Dist. Ct., S. Dist. Ind., Uniform Protective Order (“Good cause” appears in instructions accompanying the model order but not in the order itself), https://www.insd.uscourts.gov/forms/uniform-protective-order; U.S. Dist. Ct., N. Dist. Iowa, Stipulated Protective Order, https://www.iand.uscourts.gov/forms-0 (to locate, scroll down to “Civil”); U.S. Dist. Ct., Dist. Minn, Stipulation for Protective Order Form, https://www.mnd.uscourts.gov/sites/mnd/files/forms/Stipulation-forProtective-Order-Form.pdf; U.S. Dist. Ct., Dist. Neb., Protective Order: No AEO, no HIPPA , https://www.ned.uscourts.gov/internetDocs/forms/Protective%20Order-no%20AEO%20no%20HIPAA.pdf; U.S. Dist. Ct., S. Dist. Ohio, One-Tier Protective Order, https://www.cand.uscourts.gov/wpcontent/uploads/forms/model-protectiveorders/CAND_StandardProtOrd.Feb2022.pdf; U.S. Dist. Ct., W. Dist. Wash., Model Stipulated Protective Order, https://www.wawd.uscourts.gov/sites/wawd/files/ModelStipulatedProtecti veOrder%20CLEAN_2.1.23.pdf; U.S. Dist. Ct., S. Dist. W.V., Agreed Protective Order, https://www.wvsd.uscourts.gov/forms/agreedprotective-order. Five use the term “good cause” only when discussing other procedures, including procedures for subsequent challenges to individual confidentiality designations. U.S. Dist. Ct., N. Dist. Cal., Model Stipulated Protective Order 6, https://www.cand.uscourts.gov/wpcontent/uploads/forms/model-protectiveorders/CAND_StandardProtOrd.Feb2022.pdf; U.S. Dist. Ct., S. Dist. Cal., Model Protective Order, https://www.casd.uscourts.gov/forms.aspx?list=miscellaneousforms; U.S. Dist. Ct., N. Dist. Okla., CV-29 Form (Confidential Protective Order) 8, https://www.oknd.uscourts.gov/protective-orders; U.S. Dist. Ct., N. Dist. Tex., Protective Order 4, 9, https://www.txnd.uscourts.gov/sites/default/files/forms/MO62AProtectiveOrder.pdf; U.S. Dist. Ct., N. Dist. W.V., Sample Protective Order 3, 6,

62 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

But they nonetheless bolster our findings and suggest that district courts do not routinely evaluate parties’ good cause justifications. 2. Qualitative Review of 100 Denied POs To round out our empirical assessment, we evaluated 100 denials of stipulated POs, roughly 3.5% of all stipulated PO denials in our study.215 Of these, 91 orders fully denied the parties’ joint motion, while 9 denied it only in part.216 Notably, despite our random sampling, forty-four of our stipulated PO denials were from the Central District of California, a product of the district’s large caseload and comparatively high denial rate. Just three judges (district and magistrate) from the Central District of California accounted for 23% of the PO denials in our entire random sample. So, why did courts reject stipulated POs? We found that there were several reasons—and, in many instances, judges’ denials highlighted numerous deficiencies. As Figure 11 demonstrates, over half (54%) of all stipulated PO meritsbased denials were traceable to the fact that the parties tried to smuggle in provisions that committed the court to automatically seal court filings. (As explained previously, the sealing determination and the entry of a PO are often confused, but the two concepts should be kept distinct.217) A similar share (50%) of rejected stipulated POs centered on good cause. Often, parties either made no good cause statement at all, or made it with insufficient particularity. The last large category of rejections, accounting for 45% of denials, were proposed POs that sought protection for an overly broad, or poorly https://www.wvnd.uscourts.gov/sites/wvnd/files/Proposed%20Draft%20P rotective%20Order%20For%20Form.pdf. Two mention the good cause standard for entering a PO but do not make a good cause determination. U.S. Dist. Ct., Dist. N.H., Civil Form 5: Protective Order Form 1, 9, https://www.nhd.uscourts.gov/pdf/Civil%20Form%205.pdf (stating that the order applies only to “materials which may be subject to restrictions on disclosure for good cause”); see also U.S. Dist. Ct., N. Dist. Ohio, Form Protective Order 1, 9, https://www.ohnd.uscourts.gov/sites/ohnd/files/CivilRules_AppendixL.p df (same) 215

More precisely, we analyzed 113 stipulated POs but excluded thirteen denials because they did not explain the reason for denial. 216

As noted above, in our analyses, we counted a PO as “denied” if it was denied in any part. 217

See supra notes 41–46 and accompanying text.

63 Electronic copy available at: https://ssrn.com/abstract=4811151

described, category of documents. Together, at least one of these three errors was mentioned in 84% of all meritdenials.218 Beyond that, around one-quarter (23%) of denials included improper provisions for resolving disputes between parties, such as provisions that did not adequately describe mechanisms for contesting the confidentiality of a document. And a further 16% of orders were denied because they implied that the court had jurisdiction after the conclusion of the case. All other reasons for rejection accounted for 33% of denials.219 Figure 11: Stated Reasons for Denying a Stipulated PO

Second, and with more normative force, we consider whether POs were ever entered after an initial denial of a request for a stipulated PO. If a PO is not entered following a denial, the court’s initial denial ensured that discovery occurred without judicially imposed privacy, and parties were free to reveal information to others. We might also reason that, where a PO is never entered, (i) the PO was of little value to parties to begin with, (ii) satisfying the court’s objection(s) would have devalued the PO, or (iii) the parties could not meet the good cause requirement. Where a PO is later entered, we might reason that the PO was of comparatively higher value to litigants, who must have done additional work to convince the court to grant the order. Ultimately, we found that following

218

Notably, we identified many of these same errors in granted POs.

219

Most frequently, POs in this category were rejected because they: (1) failed to say that members of the public had a right to challenge confidential designations (9%), or (2) purported to bind the court, court employees, or the jury (8%).

64 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

denial in full of stipulated PO requests,220 40% of cases had no subsequent entry of a PO. IV. IMPLICATIONS In this final Part, we step back and assess the implications of our work. Starting at ground level, Subpart A argues that our findings should, at a minimum, discipline future debate surrounding Rule 26(c). During decades of heated back-and-forth, many reform proposals have been short-circuited by claims that stipulated POs are (1) very rare and (2) when filed, are carefully scrutinized by courts.221 Our data powerfully undercut both those claims, especially the second. Moving back slightly, Subpart B addresses the implications of our findings for tort law in particular. As Subpart B explains, there is a dawning recognition among many tort law scholars that a significant—and perhaps the most significant—function of tort litigation is to generate information. Tort law’s raison d’être is not to provide compensation, corrective justice, or even old-style deterrence, these scholars argue; it is, rather, to bring “stubborn information to light.”222 Through this lens, the Article’s implications are clear. If tort law’s principal function is to bring “stubborn information to light,” when it comes to tort law, POs, which prevent the public disclosure of surfaced information, have a significant—and heretofore radically underestimated—antisocial effect. Subparts C and D assess our results through a widerangle lens. Subpart C addresses the implications of our findings when it comes to the relative role of rulemaking versus discretion in American procedure. Subpart D then considers how our analysis contributes to a brewing debate about the future of the American civil justice system. Indeed, the debate about POs both reflects and sheds light on a slew of vitally important questions, including the nature of adversarialism and whether and how to adapt the “open 220

When a court partially grants a PO request, it enters the PO with some modification, and litigants need not file additional requests to enter a PO. Partial grants are therefore excluded from this analysis. 221

See supra notes 123–130 and accompanying text.

222

Wendy Wagner, Stubborn Information Problems & the Regulatory Benefits of Gun Litigation, in SUING THE GUN INDUSTRY: A BATTLE AT THE CROSSROADS OF GUN CONTROL & MASS TORTS 271, 271, 275– 76 (Timothy D. Lytton ed., 2006).

65 Electronic copy available at: https://ssrn.com/abstract=4811151

courts” principle, as the American civil justice system enters a new and challenging chapter.

A. Rebooting the Rule 26(c) Debate First, our empirical discoveries can and should reboot the perennial debate surrounding Rule 26(c). As noted previously, opposition to Rule 26(c) reform proposals has long anchored on two arguments. First, that POs are rare; and second, that trial judges exercise careful, considered judgment when entering them. These core claims, moreover, have issued from prestigious quarters, lending heft to their frequent airing. For instance, District Judge Mark R. Kravitz, testifying in Congress in 2008 on behalf of the Judicial Conference in opposition to the Sunshine in Litigation Act, directly asserted that “there is no empirical evidence to suggest that protective orders . . . are substantially abused in the Federal courts.”223 Or, as he put it at the same hearing: “[C]ourts review [PO] motions carefully and often deny or modify them to grant only the protective orders needed, recognizing the importance of public access to court filings.”224 Some twenty years earlier, another defender of the Rule 26(c) status quo, Professor Arthur Miller—himself a former Reporter and Advisory Committee member—reassured readers in an influential Harvard Law Review article: “No one is advocating the automatic or cavalier issuance of protective orders . . . let alone that they be granted without regard for substantially deleterious effects on public health and safety.”225 “The key,” Miller continued, “is retaining judicial discretion.”226 Because judges already “take their Rule 26(c) obligations very seriously,”227 Miller argued 223

Sunshine in Litigation Act of 2008: Hearings on H.R. 5884 Before the H. Comm. on the Judiciary, Subcomm. on Commercial and Administrative Law, 110th Cong. 74 (2008) (testimony of The Hon. Mark R. Kravitz for the Judicial Conference of the United States). 224

Id. at 81 (statement of The Hon. Mark R. Kravitz for the Judicial Conference of the United States). 225

Miller, supra note 35, at 491. Extending the thought, Miller advised: “Judges must guard against any notion that the issuance of protective orders is routine, let alone automatic, even when the application is supported by all the parties.” Id. at 492. 226

Id. at 491.

227

Federal Sunshine in Litigation Act and Federal Court Settlements Sunshine Act, Hearings on H.R. 2017 and H.R. 3803 Before the H. Comm. on the Judiciary, Subcomm. on Intellectual Property and Judicial Administration, 102nd Cong. 15 (1992) (statement of Prof. Arthur R. Miller).

66 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

to Congress soon after, and because “proponents of reform have not demonstrated that significant modification of the present framework is necessary, the existing pragmatic and discretionary balancing technique should be retained.”228 And, Judge Edward Becker has also publicly expressed “doubt [that] any judge would approve a consent order not demonstrably rooted in Rule 26(c).”229 Finally, consider a colloquy between Rep. William Hughes—a supporter of the 1990s-era Sunshine in Litigation Act—and Stephen C. Bransdorfer, Deputy Assistant Attorney General, who opposed reform: Mr. Hughes: Mr. Bransdorfer, let me ask you a question—just by way of a threshold question. Is it important for judges to make an independent determination that serves the public interest to sign a protective order? Mr. Bransdorfer: Yes, it is. Mr. Hughes: Is that policy being followed throughout our system today? Mr. Bransdorfer: I think it is. . . . I think, by and large, a fair study of the system would indicate that . . . the use of protective orders does involve a showing of cause.230 Our findings, drawn from a “fair” study of the system—indeed, a comprehensive study of more than 2.2 million dockets spanning nearly a decade of filings—run directly contrary to these claims. Stipulated POs are quite common, perhaps increasingly so, and most judges do not appear to be carefully exercising discretion in entering them. Rather, the majority of judges in our data apparently never— not even once—saw a stipulated PO they wouldn’t sign. Approval of POs appears to be “automatic or cavalier”— 228

Miller, supra note 35, at 491. Zenith Radio Corp. v. Matsushita Elec. Indus. Co., 529 F. Supp. 866, 889 n.40 (E.D. Pa. 1981). 229

230

67 Electronic copy available at: https://ssrn.com/abstract=4811151

exactly the situation Professor Miller, alongside many others, have long said doesn’t exist, should not be countenanced, and, if proven, would impel serious reform consideration.231 Of course, our results do not necessarily mean that arguments about the prevalence of POs or the level of judicial care in entering them are the only ones to be considered in a renewed debate about whether or how to revise Rule 26(c). It could be that policymakers could still decide that, particularly in an increasingly digital age, Rule 26(c)’s anemic implementation strikes the right balance between transparency and privacy and that the rule’s text should merely be brought into alignment with the rule’s on-the-ground 232 implementation. What is clear, however, is that, so far, the decades-long debate about POs has been deeply impoverished. Defenders of the Rule 26(c) status quo have fended off reform attempts by relying on dusty empirics and reassuring but, it turns out, inaccurate claims that the system is functioning as intended. By sweeping away these claims around stipulated POs’ prevalence and judicial care in entering them, this Article can and should prompt renewed debate. And, just as critically, it ought to reorient the debate to more squarely focus on substantive questions, including the role of private civil litigation in a well-ordered society and the relative importance

231

Miller, supra note 35, at 491.

232

For instance, reform champions and opponents, we noted previously, have also long clashed over first principles, such as whether the primary purpose of adjudication is the resolution of private disputes or the public elaboration of legal norms—a fundamental tension we return to below. See infra Subpart IV.D. A robust, revived debate about the future of Rule 26(c) should also consider the ways the world, including civil litigation, has changed in recent decades. As just one example, those who oppose Rule 26(c) reforms could plausibly argue that tectonic shifts in the political economy of information—for instance, the increasing velocity, virality, and permanence of information in a digital world—justifies a more privacy-protective approach, particularly early in litigation, in order to balance transparency values against the need to protect parties, and particularly defendants, from undue reputation-ruining publicity before a court has fairly and finally adjudicated the allegations made against them. For seminal works on virality, see KARINE NAHON & JEFF HEMSLEY, GOING VIRAL (2013); JONAH BERGER, CONTAGIOUS: HOW TO BUILD WORD OF MOUTH IN THE DIGITAL AGE (2013). For more discussion of privacy concerns and court data, see infra note 285 and accompanying text.

68 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

of transparency and secrecy as against competing values in the resolution of disputes.

B. Stunting the Information-Forcing Function of Tort Law Second, our findings have deep implications for the aims—and limits—of the tort system. Here, an increasingly popular view among tort law scholars is that an important benefit—perhaps the primary benefit—of the tort system is its “information-forcing function.” This information-forcing theory posits that tort law indeed promotes safety. But it does so not exclusively, and perhaps not even mostly, via the muchdiscussed path of cost internalization that law-and-economics scholars such as Richard Posner, Mitch Polinsky, and Steven Shavell have long championed.233 Rather, according to this new view, tort law operates in large part through its informational effect.234 The idea is that plucky plaintiffs’ lawyers—decentralized and well incentivized by the contingency fee—are uniquely positioned to connect dots, follow leads, depose insiders, and pry damning documents out of company vaults.235 And these lawyers’ persistent efforts yield vital information that promotes sensible regulatory activity and also ensures that reputations reflect reality and steer consumers toward safer goods, services, and

233

See Nora Freeman Engstrom, Facilitating the Information-Forcing Function of Tort Law, JOTWELL (Mar. 22, 2022), https://torts.jotwell.com/facilitating-the-information-forcing-functionof-tort-law/. 234

See Nora Freeman Engstrom & Michael D. Green, Tort Theory and Restatements: Of Immanence and Lizard Lips, 14 J. TORT L. 333, 372 n.5 (2021). 235

See Engstrom, supra note 2, at 333–35 (describing these dynamics); Wendy Wagner, When All Else Fails: Regulating Risky Products Through Tort Litigation, 95 GEO. L.J. 693, 700 (2007) (explaining that, compared “to their agency counterparts, litigants in tort cases are generally both more eager and more able to access asymmetric information held by manufacturers and industrial polluters”).

69 Electronic copy available at: https://ssrn.com/abstract=4811151

workplaces.236 Information, these scholars believe, is not ancillary. It is tort law’s beating heart.237 The notion isn’t merely theoretical. Over recent decades, examples of instances where the tort system has generated critical information concerning public safety have accumulated.238 Litigation against gun manufacturers and sellers has generated significant policy-relevant information, including that gun executives have long known how guns are diverted into illegal secondary markets.239 Litigation addressing clergy sexual abuse has surfaced a mountain of information, including how Church officials sought to conceal the misconduct of various priests.240 Litigation involving polychlorinated biphenyls (PCBs) has uncovered startling new evidence concerning the danger and prevalence of these industrial chemicals.241 Litigation against cigarette companies 236

Burch & Lahav, supra note 8, at 356 (“When made public, information produced in discovery allows the press and researchers to study the documents, connect the dots, unmask health risks, shed light on regulatory failures, and pressure companies to make safer products.”); Jacob Asaf & Roy Shapira, An Information-Production Theory of Liability Rules, 89 U. CHI. L. REV. 1113, 1148 (2022) (arguing that “the process of litigation propels regulators to act. . . because it provides them with valuable information on the regulated industry to which they were not privy” and also because it “makes available information more salient, thereby facilitating more intense media scrutiny of the regulator”). 237

Wagner, supra note 235, at 696 (highlighting the “indispensable role of tort litigation in lowering information-related barriers to regulating risky products”); Benjamin Ewing & Douglas A. Kysar, Prods and Pleas: Limited Government in an Era of Unlimited Harm, 121 YALE L.J. 350, 375 (2011) (explaining that “the tort system is a vital source of information gathering and intragovernmental feedback”). 238

Wagner, supra note 235, at 711–17 (collecting examples); see also Abbe R. Gluck, Elizabeth Chamblee Burch & Adam S. Zimmerman, Against Bankruptcy: Public Litigation Values Versus the Endless Quest for Global Peace in Mass Litigation, 133 YALE L.J. FORUM 525, 556– 57 (2024) (same); Joanna C. Schwartz, The Cost of Suing Business, 65 DEPAUL. L. REV. 655, 680–83 (2016) (same). 239

Timothy D. Lytton, Using Tort Litigation to Enhance Regulatory Policy Making: Evaluating Climate Change Litigation in Light of Lessons from Gun-Industry and Clergy-Sexual-Abuse Lawsuits, 86 TEX. L. REV. 1837, 1845–46 (2008). 240 Id. at 1852–53. See also Wagner, supra note 235, at 725–27 (describing various revelations). 241 Gerald Markowitz & David Rosner, Monsanto, PCBs, and the Creation of a World-Wide Ecological Problem, 39 J. PUB. HEALTH POL’Y 463, 463 (2018) (explaining that “an enormous trove of previously private Monsanto

70 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

has uncovered information to show who knew what when about the danger of smoking.242 And, litigation against opioid manufacturers, sellers, and distributors has—likewise— generated stunning information concerning the breadth, depth, causes, and character of that continuing epidemic.243 Seen through this conceptual lens, the problem of POs is plain. If it is true, as some scholars claim, that tort law exists chiefly to bring “stubborn information to light,” then the automatic and unreflective entry of POs stunts this vital informational exchange and stymies tort law’s core utility.244

C. Rule Fidelity, Judicial Hierarchy, and Judge Discretion A third set of implications sounds in judicial decisionmaking: Our findings raise deep concerns about fidelity to procedural rules and the complex interplay of rulemaking, judicial hierarchy, and trial judge discretion in the creation of American procedure. Viewed from one angle, widespread rubber-stamping of POs despite Rule 26(c)’s clear command, as supported by the stunning fact that more than half the federal judges in our data never denied a joint request for a PO during the study period, indicates an especially sharp disjunction between the law on the books and the law in action. Rule 26(c) states plainly that POs are to be granted only upon a showing of good cause. As the Eleventh Circuit has explained: “The plain text of [Rule 26(c)] suggests that a district court must find good cause to issue a protective order.”245 Or, to quote the Third: “[T]he burden of justifying the confidentiality of each and every document sought to be covered by a protective order

reports, papers, memos, letters, and studies” concerning PCBs has been made available to researchers, as a result of litigation). 242

Engstrom & Rabin, supra note 117, at 355–57.

243

Id.

244

Engstrom, supra note 2, at 333–35 (quotation marks omitted); cf. Egilman et al., supra note 9, at 295 (collecting pharmaceutical examples of POs shielding information relevant to public health from disclosure). 245

In re Chiquita Brands Int’l, Inc., 965 F.3d 1238, 1249 (11th Cir. 2020) (per curiam).

71 Electronic copy available at: https://ssrn.com/abstract=4811151

remains on the party seeking the protective order; any other conclusion would turn Rule 26(c) on its head.”246 Supporting this authority, as noted previously, the Federal Judicial Center has unambiguously stated: “It is common for parties to present to the court a stipulated protective order for the court to sign. . . . It is only proper for the court to issue the order upon the court’s finding that the order is supported by good cause.”247 The Wright and Miller treatise concurs.248 And leading scholars, too, have resisted the notion that party agreement suffices, under Rule 26(c). Steve Gensler, for example, cautions that, even when parties stipulate “it is the court that ultimately must enter the [protective] order, and the court may do so only in compliance with Rule 26(c)’s good cause requirement.”249 Gustavo Ribeiro concurs. “Even if the parties agree to proposed stipulated protective orders,” cautions Ribeiro, “the court must still determine whether good cause for issuing the order exists.”250 Yet, even before our study, there were hints that this command was often honored in the breach. As one district court recently acknowledged: “It has become common practice for district courts to enter broad, stipulated protective orders without strict compliance with Rule 26(c).”251 And, more than a quarter-century ago, the influential Defense Research Institute also seemed to own this fact, instructing its members to “routinely seek a protective order limiting dissemination of documents, even though defense counsel can make no special claim of confidentiality.”252

246

Cipollone v. Liggett Group, Inc., 785 F.2d 1108, 1122 (3d Cir. 1986). 247

REAGAN, supra note 72, at 6.

248

Recall, it provides that “[e]ven when the parties consent, the court may not enter an order unless Rule 26(c) is satisfied.” MARCUS, supra note 48, at § 2035. 249

GENSLER, FEDERAL RULES OF CIVIL PROCEDURE, RULES AND COMMENTARY, Rule 26 (2021 ed.). 250

Ribeiro, supra note 30, at 190.

251

Fant v. City of Ferguson, 2019 WL 4221515, at *3 (E.D. Mo. 2019).

252

Kerry A. Kearney & Tracy G. Benson, Preventing Non-Party Access to Discovery Materials in Product Liability Actions: A Defendant’s Primer, in CURRENT ISSUES IN LAW AND MEDICINE 36, 40 (1987).

72 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

There is, the above suggests, a glaring mismatch. The plain language in Rule 26(c), most appellate courts, and the Federal Judicial Center say that trial courts need to do one thing. But it appears that trial courts, instead, are doing quite another. In a companion paper we more fully explore this disconcerting disconnect between the law on the books and law in action.253 But for now, this basic mismatch—trial courts’ basic disobedience—contributes to at least two broader debates about American procedure-making. First, our findings raise important questions about the relative roles played by rulemaking, appellate oversight, and judicial discretion in American civil litigation. In recent decades, one of the signal developments in American civil procedure has been the steady deepening of the reservoirs of discretion within which trial judges operate.254 From the rise of “plausibility” pleading,255 to the centering of summary judgment,256 to increased attention to case management,257 to new controls and limits on trial,258 to the “Wild West” and often ad hoc administration of mammoth multidistrict

253

See supra note 191. As a spoiler, we find that, even in circuits that appear to have strict (written) standards for the review and approval of Rule 26(c) motions, trial courts quite freely grant parties’ jointly filed Rule 26(c) motions. See Peters et al., supra note 47. 254

See David Freeman Engstrom, Digital Civil Procedure, 169 U. PA. L. REV. 2243, 2246–52 (2022). 255

In Bell Atlantic Corp. v. Twombly, 550 U.S. 544, 565 (2007), the Court conferred substantial discretion on trial judges in instructing them to assess the plausibility of allegations “in light of common economic experience.” That the discretion conferred was not limited to Twombly’s antitrust setting became clear two years later in Ashcroft v. Iqbal, 556 U.S. 662, 663–64 (2009). 256

For the radical growth of summary judgment, see Nora Freeman Engstrom, The Lessons of Lone Pine, 129 YALE L.J. 2, 67 n.293 (2019). 257

Considered in tandem, Rules 16 and 26 grant enormous discretion to trial judges to control the breadth and pace of discovery. See Stephen B. Burbank & Linda J. Silberman, Civil Procedure Reform in Comparative Context: The United States of America, 45 AM. J. COMP. L. 675, 679 (1997). For the classic discussion of the growth of, and tradeoffs surrounding, trial-judge discretion, see generally Judith Resnik, Managerial Judges, 96 HARV. L. REV. 374 (1982). 258

Nora Freeman Engstrom, The Trouble with Trial Time Limits, 106 GEO. L.J. 933, 941–46 (2018).

73 Electronic copy available at: https://ssrn.com/abstract=4811151

litigations,259 American trial judges increasingly call their own tune and are, simultaneously, increasingly able to shield their day-to-day conduct from meaningful appellate scrutiny.260 Indeed, large swaths of contemporary civil procedure can be seen as a deliberate conferral of discretion on trial judges to tailor procedural rulings to individual cases or even to make procedure out of whole cloth.261 Against this broader backdrop, our striking findings concerning district court disobedience when granting stipulated POs is doubly provocative. Enriching—and complicating—current discussion of judicial discretion, our findings reveal that, even on those seemingly rare occasions when we all think trial judges’ discretion is cabined, in fact, trial judges are, instead, free agents, untethered to formal commands. Second, our findings raise vital questions about other pathways through which American procedure is made, particularly the role of rulemaking under the Rules Enabling Act. The Rules Enabling Act, supposedly, is the cornerstone of the system. 262 But we show that, whether or not rulemakers formally choose to revisit Rule 26(c), trial judges have already effected substantial, sub rosa changes to its core operation, particularly the “good cause” standard. That startling insight is important in its own right—and it also connects to wider concerns about the process of American procedure-making. A rich empirical literature 259

See Engstrom, supra note 256, at 6–7 (discussing the ascendance of MDLs and MDL judges’ persistent and potentially lawless “ad hocery”). 260

Engstrom, supra note 254, at 2246–52.

261

Engstrom, supra note 256, at 8–10 (discussing judicial ad hocery and the growth of “DIY procedures”). For a classic discussion of discretion and the Federal Rules of Civil Procedure, see Stephen N. Subrin, How Equity Conquered Common Law: The Federal Rules of Civil Procedure in Historical Perspective, 135 U. PA. L. REV. 909 (1987). 262

See 28 U.S.C. § 2072; see also Stephen B. Burbank, The Rules Enabling Act of 1934, 130 U. PA. L. REV. 1015, 1025 (1982) (offering the fullest account of the REA’s origins and purposes); Engstrom, supra note 254, at 2246–58 (cataloging increasing pressure on the REA’s rulemaking process as a result of the felt need for judicial discretion over rule implementation, separation-of-powers tensions, and the unmasking of the distributive consequences of rules via increasingly sophisticated empiricism).

74 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

examines efforts to shape American procedure. One key finding is that some of the most impactful procedural reforms have come not by way of legislation or rulemaking, but rather through court decisions, precisely because these decisions tend to fall below the public radar.263 Technical and turgid court decisions, the theory goes, flunk the “Dan Rather” test, named after the news anchor of old. A tweak to Rule 8, 12, or 56, buried deep in paragraph nineteen of a decision issued sometime in June can be hugely impactful, but it does not mobilize constituencies and, indeed, does not register with the public at all.264 Our findings enrich that idea. Recall, in the mid 1990s, the Judicial Conference considered a package of amendments to Rule 26(c). Among them was a proposal that would have allowed courts to issue a PO “for good cause shown or on stipulation of the parties.”265 The proposal, then, would have formally allowed a PO to be issued simply on the parties’ sayso.266 Yet, the proposed amendment sent shockwaves through the legal world—and ultimately, in a highly unusual 263

STEPHEN B. BURBANK & SEAN FARHANG, RIGHTS AND RETRENCHMENT: THE COUNTERREVOLUTION AGAINST FEDERAL LITIGATION 22–23 (2017). 264

Id. As just one example, the Court in Twombly achieved important changes to the pleading standard where Congress and the Advisory Committee had failed—and it did so just five years after the Supreme Court unanimously insisted, for the second time in less than a decade, that only the rulemaking process could change the very same pleading standard. Swierkiewicz v. Sorema N. A., 534 U.S. 506, 515, (2002) (“A requirement of greater specificity for particular claims is a result that must be obtained by the process of amending the Federal Rules, and not by judicial interpretation.” (quotation marks omitted)). 265

Letter of June 2, 1995, supra note 78, at 1–2.

266

Some felt that the move would merely “confirm the common practice of entering protective orders on stipulation by the parties.” Advisory Comm. on Civil Rules, October Meeting, Meeting Minutes 8 (Oct. 20, 1994) https://www.uscourts.gov/sites/default/files/fr_import/CV101994.pdf; Others disagreed. E.g., Richard B. Schmitt, Judges’ Group Kills Protective Order Rule, WALL ST. J., Mar. 15, 1995, at B4 (quoting the chairman of the executive committee, Judge Gilbert Merritt, as stating that the stipulation rule departs from long-standing practice); Letter of June 2, 1995, supra note 78, at 1–2 (“express[ing] concern that the proposed rule would change existing practice by allowing entry of protective orders without a showing of good cause”).

75 Electronic copy available at: https://ssrn.com/abstract=4811151

move, the Judicial Conference rejected the amendment in the face of blistering criticism.267 In particular, opponents lodged two core objections. Their first was procedural—that the proposal was added in the eleventh hour.268 Second, opponents argued that the rule, which would represent a “dramatic shift” from the status quo,269 would facilitate unbridled secrecy, which would harm public health.270 This was, at least in part, because “individual plaintiffs suing large corporations . . . have little choice but to agree” to POs “as the price for avoiding prolonged, expensive pretrial discovery.”271 Following the vote, Linda Greenhouse wrote a New York Times story summarizing the debate and its stakes. The story quoted Judge Gilbert S. Merritt who described his opposition to the amendment: “Federal courts shouldn’t do anything without just cause.”272 Secrecy “should not just be left to the option of the parties.”273 Yet, we show that, although a formal effort to rewrite Rule 26(c) failed, somehow, through a drip, drip, drip of district court discretion, Rule 26(c) was nevertheless rewritten. Secrecy has been left to the option of the parties. This startling insight connects to a broader set of concerns about procedure267

Letter of June 2, 1995, supra note 78, at 2; see also Arthur R. Miller, Effective Rulemaking Damaged by Politics, NAT’L L.J. May 1, 1995, at A19 (“Media and special interest group lawyers successfully mounted a last-minute, high-visibility protest against the proposal. . . .”). For representative criticism, see, e.g., Arthur Bryant, A Sneak Attack on Open Justice, NAT’L L.J., July 10, 1995, at A19; Jane Kirtley, Keeping the Public in the Dark, AM. JOURNALISM REV., July 1995, at 50. Indeed, the stipulation proposal proved so contentious that the Judicial Conference rejected the entire package of Rule 26(c) amendments— including provisions that would have clarified the standard for modification and dissolution of POs—without discussing the other reforms. See Letter of June 2, 1995, supra note 78, at 2. 268

While the other amendments in the 26(c) package had been circulated months in advance and had been the subject of extensive public comment, the stipulation provision was added at the eleventh hour. See Schmitt, supra note 266, at B4; Bryant, supra note 267, at A19. 269

Bryant, supra note 267, at A19. As noted, not everyone agreed with this characterization. See supra note 266. 270

See e.g., Bryant, supra note 267, at A19.

271

See Greenhouse, supra note 80, at A9.

272

Id. at B9.

273

Id.

76 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

making’s pathways and adds fuel to a growing critique about a derogation of the rulemaking process.274

D. Open Courts, Adversarialism, and Access to Justice Finally, evidence concerning the actual operation of Rule 26(c) contributes to a swirling debate about whether two cornerstones of the American civil justice system—the “open courts” principle and the American commitment to adversarialism—should be adapted to new legal realities.275 Indeed, our Rule 26(c) empirical findings both reflect and shed light on debate about whether and how to revise both mainstays of the American system. Debate about the meaning and wisdom of the “open courts” principle is not exactly new. It has come, first and foremost, in disputes around how best to interpret the array of constitutional, statutory, and procedural provisions from which the principle emanates.276 Debate has also come in 274

See Engstrom, Digital, supra note 254, at 2246–58 (tracing heated debate about the place of rulemaking in the American civil justice system). For further discussion, see Stephen B. Burbank, Procedure, Politics and Power: The Role of Congress, 79 NOTRE DAME L. REV. 1677 (2004) (considering the constitutional dimension); BURBANK & FARHANG, supra note 263 (offering a wide array of empirics on the operation and public perception of the rulemaking process); ROBERT BONE, JUSTIFYING LITIGATION REFORM (forthcoming 2024) (offering a full theory of procedure-making, including the REA rulemaking process); David Marcus, The Collapse of the Federal Rules System, 169 U. PA. L. REV. 2485 (2021) (examining rulemaking through the lens of core procedural values of trans-substantivity and neutrality). 275

“Open courts” is a longstanding principle within the American constitutional tradition and legal system, with doctrinal roots in privileges and immunities, free speech, due process, and equal protection and finding further expression in a large body of statutes, common-law rules, and procedural rules. See infra note 276. The traditional American commitment to adversarialism holds that a partydriven clash of proof in a structured, forensic setting before a passive, neutral decision-maker is the best way to resolve disputes. See STEPHAN LANDSMAN, THE ADVERSARY SYSTEM 2 (1984). For a further discussion of the U.S.’s commitment to the adversarial resolution of disputes, see generally ROBERT A. KAGAN, ADVERSARIAL LEGALISM: THE AMERICAN WAY OF LAW (2d ed. 2019). For rich historical perspective, see generally AMALIA D. KESSLER, INVENTING AMERICAN EXCEPTIONALISM: THE ORIGINS OF AMERICAN ADVERSARIAL LEGAL CULTURE, 1800–1877 (2017). 276

A large body of constitutional provisions, statutes, common-law rules, and procedural rules extolls the virtues of “open courts” and grants the public broad rights of access to civil court proceedings. See,

77 Electronic copy available at: https://ssrn.com/abstract=4811151

more academic forms, about the core purposes of adjudication. According to some, courts exist to resolve private disputes277—and any consideration of the public interest blurs and blunts that core purpose. Courts, it follows, should “readily accede to the parties’ mutual desire for confidentiality,”278 lest they turn courts into “information ombudsmen,” distracted from their core function of

e.g., Press-Enterprise Co. v. Superior Ct. (Press-Enterprise I), 464 U.S. 501, 508–09 (1984) (lauding the “community therapeutic value” of open proceedings); Joy v. North, 692 F.2d 880, 893 (2d Cir. 1982) (“An adjudication is a formal act of government, the basis of which should, absent exceptional circumstances, be subject to public scrutiny.”); In re Union Leader Corp., 147 809 A.2d 752, 753 (N.H. 2002) (affirming that open courtrooms and access to court records are “critical to ensure that court proceedings are conducted fairly and impartially”); see also Landmark Commc’ns v. Virginia, 435 U.S. 829, 839 (1978) (“The operations of the courts and the judicial conduct of judges are matters of utmost public concern.”); FED. R. CIV. P. 43(a) (“At trial, the witnesses’ testimony must be taken in open court unless a federal statute, the Federal Rules of Evidence, these rules, or other rules adopted by the Supreme Court provide otherwise.”); 28 U.S.C. § 452 (“All courts of the United States shall be deemed always open for the purpose of filing proper papers, issuing and returning process, and making motions and orders.”). See generally Jonathan M. Hoffman, By the Course of the Law: The Origins of the Open Courts Clause of State Constitutions, 74 OR. L. REV. 1279, 1279 (1995) (canvassing state constitutional provisions). 277

See Miller, supra note 35, at 431 (noting that a focus on public access would divert courts “from their primary mission” of resolving disputes among litigants); id. at 441 (“[T]he function of the judicial system is to resolve private disputes, not to generate information for the public.”); Marcus, supra note 101, at 468 (“[C]ourts exist to resolve disputes that are brought to them by litigants . . . .”); id. at 470 (“The primary purpose for which courts were created, distinguishing them from other organs of government, is to decide cases according to the substantive law. The collateral effects of litigation should not be allowed to supplant this primary purpose.”); see also supra notes 137–139 and accompanying text (rehearsing these arguments). 278

Doré, supra note 113, at 289; see David S. Sanson, The Pervasive Problem of Court-Sanctioned Secrecy and the Exigency of National Reform, 43 DUKE L.J. 807, 810 (2003) (“Jurisprudentially, the prosecrecy argument insists that the judicial system is a forum for private parties to resolve private disputes and is not an instrument for social justice.”); id. (“[T]he prosecrecy argument typically bases its logic on the premise that courts exist to assist private litigants in resolving their disputes in the most efficient way possible.”); cf. Erichson, supra note 54, at 360 (“A light standard of good cause for discovery confidentiality reduces the burden on the court . . . .”).

78 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

adjudicating private disputes.279 Others take the opposite view. Courts, Professor Owen Fiss has told us, “exist to give meaning to our public values, not to resolve disputes.”280 They should be no less open and transparent than other democratic institutions—perhaps even more so.281 What is different in the current moment is that the “open courts” principle is facing significant pressure and even something of a reckoning. For starters, and as noted previously, numerous high-profile scandals and public health crises have revitalized the debate about the core purposes of adjudication by shining unflattering light on the ways society’s “haves” are able to hide wrongdoing—and perpetuate it. Horrified by these revelations, at last count, more than a dozen states have enacted new laws limiting secret settlements or NDAs for certain claims.282 And, the federal government has also gotten in the act, by, for example, amending the tax laws to establish that defendants cannot deduct payments incurred to settle sexual harassment cases where an NDA shields the settlement from scrutiny.283 But that is only the beginning. A fast-digitizing legal system is also raising important questions about the terms on which courts make available their mountains of case data. Some argue that open data is essential if a growing legal tech industry is to serve impecunious clients (as well as their wealthy corporate counterparts).284 Others worry that, if made 279

See Miller, supra note 35, at 488.

280

Owen M. Fiss, The Supreme Court, 1978 Term — Foreword: The Forms of Justice, 93 HARV. L. REV. 1, 29 (1979). Others have said much the same. See Luban, supra note 120, at 2657 (describing information-generation as “a fundamental public interest” of litigation, not a side effect); Abram Chayes, The Role of the Judge in Public Law Litigation, 89 HARV. L. REV. 1281, 1313–16 (1976) (“In my view, judicial action only achieves such legitimacy by responding to, indeed by stirring, the deep and durable demand for justice in our society.”). 281

For further articulations, see supra notes 118–120 and accompanying text (rehearsing these arguments). 282

Engstrom et al., supra note 6.

283

Id. (collecting this and other recent federal efforts).

284

Gelbach, supra note 166, at 133–54 (advocating the liberation of PACER); David Freeman Engstrom & Nora Freeman Engstrom, Legal Tech and the Litigation Playing Field, in LEGAL TECH AND THE FUTURE OF CIVIL JUSTICE 133–54 (David Freeman Engstrom ed., 2023) (explaining that, currently, the “have’s” have disproportionate access to proprietary court data and that, over time, this disparity is apt to skew

79 Electronic copy available at: https://ssrn.com/abstract=4811151

easily available, court data could be downloaded in bulk and used for dubious (and perhaps discriminatory) ends.285 Add in legitimate privacy concerns, leavened with the velocity, virality, and permanence of information in the digital age, and it is clear that Rule 26(c) is part of a wider reckoning about how information flows into and, perhaps more importantly, out of America’s courts.286 The traditional American commitment to adversarial process is likewise under significant scrutiny and possible rethinking. Adversarialism—the notion that key procedural values such as accuracy, efficiency, and fairness are best optimized where opposing litigants compete by placing arguments before a neutral and passive judge—has long been a core tenet of the American the civil justice system. As Roscoe Pound told the ABA in his now-famous 1906 address: “[I]n America we take it as a matter of course that a judge should be a mere umpire . . . and that the parties should fight out their own game in their own way without judicial interference.”287

the development of substantive law); David Freeman Engstrom & R.J. Vogt, The New Judicial Governance: Courts, Data, and the Future of Civil Justice, 72 DEPAUL L. REV. 171, 198–224 (2023) (offering a wide-angle assessment of courts’ role as “data dispensers”); Aziz Z. Huz & Zachary D. Clopton, The Necessary and Proper Stewardship of Judicial Data, 76 STAN. L. REV. (forthcoming 2024) (offering a “normative evaluation of how Congress should regulate the production and dissemination of judicial data, in light of the capabilities and incentives of relevant actors”). 285

Among the more worrying and demoralizing uses of court data is the practice of landlord associations in most major cities of scraping housing court dockets in order to construct blacklists of renters who should not be rented to because they dared to try to vindicate rights in housing court. See Ronda Kaysen, How to Escape the Dreaded ‘Tenant Blacklist’, N.Y. TIMES (Apr. 13, 2019) (“There are hundreds of tenant screening bureaus, collecting names from courthouses around the country and selling the information to landlords.”). 286

See generally David S. Ardia & Anne Klinefelter, Privacy and Court Records: An Empirical Study, 30 BERKELEY TECH. L.J. 1807 (2015) (cataloging the types of sensitive information contained in court records, including locational, identity, health, and financial information as well as past involvement in criminal or civil proceedings). 287

Roscoe Pound, The Causes of Popular Dissatisfaction with the Administration of Justice, 29 ANN. REP. AM. BAR ASS’N 395, 405 (1906).

80 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

But the American commitment has long shown cracks in its foundation—and here again, the debate over Rule 26(c) both reflects and sheds light on that broader assessment. Many observers have long expressed doubt about whether an adversarial process, as opposed to an inquisitorial approach in which judges actively steer the course of litigation, is the better bet.288 Critics have long argued that adversarialism may not optimize the system’s truth-seeking function. Judge Henry Friendly famously observed that the role of counsel “is not to make sure the truth is ascertained,” and then suggested that we should wonder about a system in which “causing delay and sowing confusion not only are [counsel’s] right but may be his duty.”289 Revelations that a stunning three-quarters of civil cases filed in American courts feature at least one side without a lawyer deepen the critique and make continued adherence to unmediated adversarialism hard to defend under any reasonable theory of due process or procedural fairness.290 Our findings add another example of how adversarialism may systematically fail to achieve key systemic ends. After all, there is no constituency for disclosure in a system in which parties interested in confidentiality place arguments before a passive judge who does not consider the public interest. In economic terms, there is a negative externality—a harm borne by persons whose interests the stipulating parties can freely ignore.291 And that externality is accumulating. Put differently: The adversarial structure of the American civil justice system might leave it ill-equipped to promote anything resembling an optimal level of openness.292 288

E.g., John H. Langbein, The German Advantage in Civil Procedure, 52 U. CHI. L. REV. 823, 824 (1985). 289

Henry J. Friendly, Some Kind of Hearing, 123 U. PA. L. REV. 1267, 1288 (1975). 290

E.g., Jessica K. Steinberg, Adversary Breakdown and Judicial Role Confusion in “Small Case” Civil Justice, 2016 B.Y.U. L. REV. 899 (2016) (coining the phrase “adversary breakdown” to describe the millions of cases pitting institutional plaintiffs with lawyers against individuals without lawyers in debt, eviction, home-foreclosure, and family law, particularly child support, matters); Pamela K. Bookman & Colleen F. Shanahan, A Tale of Two Civil Procedures, 122 COLUM. L. REV. 1183 (2022) (same); Marcus, Collapse, supra note 274, at 2504– 07 (cataloging the ways debt collection litigation makes a “mockery . . . of the adversarial assumption” that underpins American procedure). 291

Ribeiro, supra note 30, at 174.

292

See Burch & Lahav, supra note 8, at 350 & 394 (explaining that, “[i]nformation revealed in lawsuits adjudicated in taxpayer-funded

81 Electronic copy available at: https://ssrn.com/abstract=4811151

Indeed, when a stipulated PO is at play, the plaintiff is not pitted against the defendant. By definition, the plaintiff and defendant, jointly, are pitted against the rest of us. CONCLUSION On an unseasonably hot afternoon in Washington, D.C., on May 17, 1990, the late Senator Herb Kohl, a perennial sponsor of the Sunshine in Litigation Act and perhaps the most steadfast champion of transparency in civil litigation, called to order one of the first hearings ever to discuss the matter. In so doing, he observed: “We must ask whether we have struck the proper balance between disclosure and secrecy.”293 Senator Kohl proceeded to highlight a few questions that needed to be answered in order to assess that broader inquiry. These included: “Is there a growing use of court secrecy in civil litigation, and if so, does it prevent disclosure of information that raises safety concerns?” And: “Do the demands of the court system encourage indiscriminate use of secrecy, and how can these competing interests best be resolved?”294 For more than a quarter-century, Senator Kohl’s simple questions have been much discussed, but they have remained stubbornly unanswered. Offering overdue clarity, this Article shows that, in fact, in contemporary federal civil litigation, the balance has tipped more sharply in favor of secrecy than has been previously recognized. Indeed, in opposing Sunshine in Litigation Acts and similar reforms, critics have, for decades, insisted that courts are “very circumspect about entering protective orders.”295 No reform courts often benefits society as a whole, but there is no incentive for private parties to reveal it” and arguing that, given this mis-match, courts ought to work harder to promote the public interest). 293 Examining the Use of Secrecy and Confidentiality of Documents by Courts in Civil Litigation, Hearing Before the S. Comm. on the Judiciary, Subcomm. on Courts and Administrative Practice, 101st Cong. 1 (1990) (statement of Sen. Herb Kohl) [hereinafter Kohl Statement]. For the weather that May afternoon, see Washington D.C. Weather in 1990, EXTREME WEATHER WATCH, https://www.extremeweatherwatch.com/cities/washington-dc/year1990. 294

Kohl Statement, supra note 293, at 1.

295

The Federal Judiciary in the 21st Century: Ensuring the Public’s Right of Access to the Courts, Hearing Before the H. Subcomm. on Courts, Intellectual Property, and the Internet of the Comm. on the Judiciary, 116th Cong. 32 (2019) (The Hon. Richard Story).

82 Electronic copy available at: https://ssrn.com/abstract=4811151

Secrecy by Stipulation

is needed, some have maintained, because judges “take their Rule 26(c) obligations very seriously.”296 Yet, as we have shown, the data suggest otherwise. Judges do not appear to cast a discerning eye on POs. They reject, on average, just 4.1% of stipulated PO requests. In fact, some judges never seem to be circumspect in entering POs. Of those federal court judges that were asked to sign more than twenty-five POs, the majority of judges’ grant rate is a bracing 100%. Beyond concerns of fidelity to law and binding precedent, these findings suggest that infamous cases—such as Beth and Ken Melton’s wrongful death lawsuit against GM, or others, such as that involving Zyprexa, Zomax, or the Dalkon Shield—are not a bug of the legal system, but rather a feature. They show how the legal system typically operates. What’s perhaps unusual about those infamous cases is that they were the rare instances when the truth was ever, belatedly, revealed.

296

83 Electronic copy available at: https://ssrn.com/abstract=4811151

Early, Incomplete Draft: Do Not Cite or Quote Civil Conspiracy, Negligence, and Recklessness in Claims Against Manufacturers of Dangerous Products John C. P. Goldberg & Benjamin C. Zipursky† Introduction Plaintiffs alleging harms caused by defective products are sometimes unable to prevail because they cannot identify the tortfeasor that injured then, cannot prove causation, or can make out a claim only against a judgment-proof defendant. In the face of these obstacles, plaintiffs’ lawyers have understandably urged courts to adopt doctrines that enable them to expand the set of potentially liable defendants. In this vein, the doctrine of market-share liability first developed by the California Supreme Court has probably received the most academic attention. However, the conditions that have been specified by the courts for its application are sufficiently stringent that it has only rarely been applied. Less well attended to, but potentially of greater practical importance, are two other doctrines: concert of action and civil conspiracy. In Prosser’s description, the function of a conspiracy claim is to extend liability in tort beyond the active wrongdoer to those who have helped plan or otherwise assisted the wrongdoer’s acts.1 In a similar vein, the Dobbs treatise treats conspiracy and the related idea of concert of action interchangeably, describing the basis for liability they identify as follows: “[t]hose who cooperate, tacitly or expressly, in particular conduct to pursue a common illegal design (or a legal design by illegal acts) are said to be acting in concert. Each of those acting in concert is liable jointly and severally for all of the intended or foreseeable resulting harm.”2 Likewise, the Second Torts Restatement treated liability for civil conspiracy as a special case of “persons acting in concert.”3

† John C.P. Goldberg is Interim Dean and Carter Professor of General Jurisprudence, Harvard Law School;

Benjamin C. Zipursky is Professor and James H. Quinn ‘49 Chair in Legal Ethics, Fordham University School of Law. 1 W. Page Keeton, et al., Prosser & Keeton on Torts § 46, at 322 (5th ed. 1984). David Owen’s products liability treatise has separate paragraphs on concert-of-action and conspiracy liability, but the description of each that is provided does not distinguish them sharply from one another. The former is described as attaching “when, under a common plan or scheme, a party assists, participates with, or actively encourages one or more wrongdoers who tortiously harm the plaintiff. David G. Owen, Products Liability Law § 11.3, at 753 (2005). In Owen’s view, liability of this sort is a species of aiding-and-abetting liability in which the abettor voluntarily participates in and furthers a group activity that the actor knows to be tortious as to the plaintiff. Id. The treatise then describes civil conspiracy liability as consisting of “a combination of two or more persons for the purposes of accomplishing by concert of action an unlawful purpose or a lawful purpose by unlawful means,” acknowledging that, under this definition, civil conspiracy is “closely related” to concert of action. Id. at 754. It adds that civil conspiracy liability requires, among other things, proof that the defendant intentionally participated in the unlawful conspiracy and did so for the purpose of doing something unlawful. Id. at 755. 2 Dan B. Dobbs, Paul T. Hayden & Ellen M. Bublick, The Law of Torts § 435 (2d ed. April 2024 update) (citations omitted). 3 Restatement (Second) of Torts § 876(a) (1979). Although the Second Restatement treats liability for aiding and abetting as a special case of concerted action, id. at § 876(b), today it is probably more common to treat aiding and abetting as a distinct grounds of liability. As explained below, this is the position taken in the draft Third Torts Restatement. We will assume for purposes of analysis that civil conspiracy and concert-of action as identical or virtually identical theories that stand apart from the notion of aiding and abetting.

In the mass tort context, plaintiffs suing for injuries caused by asbestos, DES, tobacco products, and the fuel additive MTBE, have regularly invoked concert-of-action and conspiracy claims against defendants constituting the bulk of the relevant industry. Most commonly, these suits allege that each defendant is jointly liable for having expressly or tacitly agreed with other industry members to suppress information about a hazard associated with their activities, which suppression functioned as a cause of injury to the plaintiff.4 To date, these efforts have met with mixed results. Courts routinely acknowledge the availability of these theories of liability. But they also tend to set a relatively high bar for proving claims of this sort. To establish a civil conspiracy, for example, plaintiff must prove – in some jurisdictions by clear and convincing evidence – that each defendant being sued knowingly entered into an agreement to achieve some unlawful end, or to employ unlawful means to achieve a lawful end. The mere fact that a group of businesses acted in a parallel fashion (i.e., that each failed to provide certain safety information to consumers) does not suffice, they say. Nor is it enough to show that each defendant participated in a trade association whose function is to promote the allegedly harm-causing product or activity.5 In addition, unlike in criminal law, an agreement is not sufficient to establish conspiracy liability – there must also be conduct by at least one actor pursuant to the agreement that proximately cause injury to a victim.6 Illinois law is illustrative of how the courts approach efforts by plaintiffs to reach multiple members of an industry for harms allegedly caused by one member. In a 1994 decision, the state’s high court noted the availability, in principle, of civil conspiracy liability.7 Five years later, in McLure v. Owens Corning Fiberglas Corp., it rejected civil conspiracy claims against asbestos manufacturers based on the failure of the plaintiffs to prove the requisite agreement.8 However, some subsequent Illinois intermediate appellate court decisions have rejected efforts by defendants to have civil conspiracy

4 See, e.g., Richard Ausness, Conspiracy Theories: Is There a Place for Civil Conspiracy in Products Liability Litigation?,

74 Tenn. L. Rev. 383 (2007). 5 Owen, supra note 1, at 755; see also In re: Welding Fume Prod. Liab. Litig., 526 F. Supp. 2d 775, 804-05 (N.D. Ohio) (mere attendance at trade association meetings not sufficient to establish the agreement necessary for conspiracy liability); Mowrer v. Armour Pharmaceutical Co., 1993 WL 54241 at *3 (E.D. Pa.) (“Plaintiff need not provide all the details of the alleged conspiracy but must at least ‘describe the general composition of the conspiracy, some or all of its broad objectives, and defendant’s general role in that conspiracy.’”) (cited authorities omitted); cf. Bichler v. Eli Lilly & Co., 436 N.E.2d 182, 188 (N.Y. 1982) (jury permitted to infer from evidence of “consciously parallel behavior that an implied agreement existed [among defendants] to market DES for problems of human pregnancy without first conducting [appropriate tests].”) (emphasis added). 6 See Tuttle v. Lorillard Tobacco Co., 377 F.3d 917, 926 (8th Cir. 2004) (affirming dismissal of claim alleging defendants conspired to commit fraud on the ground that the plaintiff could not prove the underlying tort of fraud given absence of any proof of reliance); Havers v. Gatke Corp., 132 Cal. Rptr. 2d 198, 201 (Ct. App. 2003) (no conspiracy liability absent underlying tortious conduct); Restatement (Third) of Torts: Miscellaneous Provisions § [Agreements Provision] cmt. f (T.D. No. 3, April 2024). 7 Adcock v. Brakegate, Ltd., 645 N.E.2d 888, 893-94 (Ill. 1994). 8 720 N.E.2d 242 (Ill. 1999) (reversing civil conspiracy verdicts against asbestos manufacturer on the ground that the evidence showed “parallel conduct” among them, not an “agreement” to suppress information about harmful effects of exposure, and further stating the sharing of information about a product among manufacturers does not suffice to establish an agreement by clear and convincing evidence, especially given evidence that manufacturers did not all pursue the same course of conduct with respect to warnings). McClure includes a review of case law from Illinois and elsewhere dealing with claims against manufacturers of DES, asbestos, and other products that declined to impose conspiracy or concert-of-action liability based only on evidence of parallel conduct.

claims dismissed at an early stage.9 In sum, civil conspiracy liability exists in principle, but it is difficult to establish in the product liability context, and there appear to be no prominent appellate court decisions affirming a jury’s imposition of such liability (though of course this may indicate that cases in which early-stage dispositive defense motions are rejected proceed to settle). Illinois appears not to be an outlier. In other jurisdictions, courts entertaining suits by persons alleging injuries from exposure to dangerous products have sometimes allowed civil conspiracy claims to reach juries.10 Yet the 2005 edition of the Owen products liability treatise concludes that “civil conspiracy claims [for product related injuries] occasionally survive motions to dismiss and for summary judgment, … [but] are routinely found wanting ….”11 Uncertainties as to both positive law and normative judgments about the applicability of civil conspiracy to claims alleging product defects, particularly failures to warn, thus call for careful and authoritative examination. To their credit, the Reporters for the Miscellaneous Provisions of the Restatement (Third) of Torts have offered a thoughtful take on these issues, helpfully distinguishing between the conditions under which liability based on negligent or reckless conduct can be imposed on a person other than the negligent or reckless actor based on “Agreements to Engage in Conduct that is Negligent or Reckless,” on the one hand, and “Aiding and Abetting Negligence Torts,” on the other.12 Notably, civil conspiracy has already received attention in several other portions of the Restatement, including those on intentional torts, liability for physical and emotional harm, and 9 See, e.g., Dukes v. Pneumo Abex Corp., 900 N.E.2d 1128 (Ill. App. 2008) (although membership of

asbestos manufacturers in a trade organization is not sufficient to establish a conspiratorial agreement, evidence of interlocking directorship and other facts would permit a jury to find the requisite agreement). 10 Waterhouse v. R.J. Reynolds Tobacco Co., 270 F. Supp.2d 678 (D. Md. 2003) (denying motion to dismiss conspiracy claim by smoker against tobacco manufacturers); Cipollone v. Liggett Group, Inc., 683 F. Supp. 1487, 1499-1500 (D.N.J. 1988) (allowing civil conspiracy claim against tobacco companies to go to jury); In re Methyl Butyl Ether (“MTBE”) Products Liability Litigation, 175 F.Supp.2d 593, 634 (S.D.N.Y. 2001) (denying motion to dismiss civil conspiracy claims given evidence that defendants “formed joint-task forces and committees … for the specific purposes of suppressing or minimizing information regarding MTBE hazards”); Rogers v. Armstrong World Indus., Inc., 744 F. Supp. 901, 905 (E.D. Ark. 1990) (denying asbestos defendants’ motion of summary judgment on civil conspiracy claims under Arkansas law, which requires for liability “a combination of two or more persons to accomplish a purpose that is unlawful or oppressive or to accomplish some purpose, not in itself unlawful, oppressive, or immoral, by unlawful, oppressive or immoral means, to the injury of another.”); Wright v. Brooke Group Ltd., 652 N.W.2d 159, 173 (Iowa 2002) (answering certified question in tobacco litigation indicating the availability, in principle, under Iowa law of civil conspiracy liability for agreements to suppress information about product risks); Farmer v. City of Newport, 748 S.W.2d 162, 164 (Ky. Ct. App. 1988) (reversing dismissal of civil conspiracy claim against mattress manufacturers agreed to produce and market unduly flammable mattresses); Abel v. Eli Lilly & Co., 343 N.W.2d 164, 176 (Mich. 1984) (affirming denial of DES manufacturers’ summary judgment motion on concert of action claims against them). Initial efforts to invoke conspiracy as grounds for imposing liability on energy companies for harm related to climate change have not yet resulted in plaintiff victories. See Native Vill. of Kivalina v. ExxonMobil Corp., 696 F.3d 849 (9th Cir. 2012) (dismissing a public nuisance lawsuit containing a civil conspiracy claim without ruling on the viability of that claim). But see Joseph Manning, Climate Torts: It’s a Conspiracy!, 62 Boston C. L. Rev. 941 (2021) (arguing for the propriety of imposing civil conspiracy liability on fossil-fuel companies for climatechange related harms); Henry Steinberg, Note: Civil Conspiracy Up in Smoke: How Similar are Cigarettes and Smokestacks?, 18 Hastings-N.W. J. Envt’l. L. & Pol’y 157 (2012) (comparing civil conspiracy claims brought against tobacco companies with those that might be brought against energy companies). 11 Owen, supra note 1, at 755 (citations omitted). For decisions issuing matter-of-law rulings in favor of defendants facing civil conspiracy claims, see In Re Asbestos School Litig., 46 F.3d 1284 (3rd Cir. 1994 (Alito, J.); [other cites]. 12 Restatement (Third) of Torts: Miscellaneous Provisions (T.D. No. 3, April 2024).

liability for economic harm. But the latter Restatements’ treatments of civil conspiracy have left an important gap: namely, the possibility of civil conspiracy liability for negligent or reckless conduct causing physical harm? As indicated earlier, such legal theories lie at the heart of much modern mass tort litigation, yet unclarity abounds. The draft Restatement’s “Agreements” section is clearly meant to cover within its terms instances of what have traditionally been called “concert of action” and “civil conspiracy” liability. Thus, the section requires for liability that: (i) an actor agree (expressly or tacitly) with at least one other to engage in conduct that is negligent or reckless; (ii) the actors engage in the agreed-upon conduct; and (iii) the conduct of at least one of them proximately causes harm to a third party. Commentary makes clear that this formulation is meant to encompass what is perhaps the paradigm “concert-ofaction” case, namely the drag race in which one driver crashes into a victim, yet both are held jointly liable.13 It also indicates that the section is meant to cover liability for civil conspiracy, yet uses the term “agreement” rather than “conspiracy” on the ground use of the latter may generate confusion because it is “associated with intentional, criminal, and especially heinous wrongdoing,” when in fact liability under the terms of this section “addresses neither intent to cause harm nor criminal conduct.”14 Unlike aiding and abetting liability, liability under the Agreements section does not require encouragement or assistance.15 Reporters’ Notes to the Agreements section candidly describe the unsettled state of the law on civil conspiracy in cases involving injuries caused in the first instance by negligence and recklessness. Nonetheless, as indicated, the text of the black letter announces a rule that entails liability when the conditions summarized above are satisfied. In terms of positioning, this lands the Draft Restatement in line with opinions from state and federal courts around the country that have denied defense motions to dismiss or summary judgment motions, but runs against a number of federal court opinions and a few prominent state high court opinions. In this essay, we are not primarily interested in case-counting or in exploring the shadow-effect of civil conspiracy doctrine and the pressures on settlement it triggers (a valuable topic in its own right).16 Instead, we wish to address a small cluster of basic doctrinal and normative questions in this area. Among them are these: Should a product manufacturer’s agreement with other manufacturers to sell a product with inadequate warnings generate liability for that manufacturer to a victim injured by a co-conspirator’s similarly defective product? Should it do so if the manufacturer’s conduct was negligent or reckless, but not fraudulent? If so, when and why; if not, why not? Our answer is predictably equivocal. Sometimes an agreement of this sort should generate liability and sometimes it should not. More helpfully, we identify two categories of agreementscenarios in which liability should attach: (a) genuine analogues to the classic concert-of-action scenarios of the drag race; (b) instances in which defendants agree to undertake the sort of 13 Id. § [agreements provision] cmt. c, Illus. 1. 14 Id. cmt. b.

15 The draft Miscellaneous Provisions’ “aiding and abetting” section allows for liability under the following

conditions: (i) an actor (A) has actual knowledge that another (O) might engage in negligent or reckless conduct toward a third party (TP); (ii) A substantially assists or encourages O to engage in such conduct, and (iii) O undertakes such conduct, thereby injuring TP. Liability of this sort does not require an agreement to undertake a course of conduct between A and O. It is enough that A acts in the knowledge that O is prepared to engage in unduly risky conduct and encourages or assists O in undertaking such conduct. Id. cmts. b – e. 16 An earlier title for our contribution to this symposium – Toxic Torts and the Shadowlands of Responsibility – contemplated an exploration of the impact of civil conspiracy allegations on settlement. We are currently inclined to regard the doctrinal questions as the ‘horse’ and the settlement topic the ‘cart,’ and have chosen not to put the cart before the horse.

misrepresentation that qualifies as deceit or fraud, including instances in which the conduct is reckless as to the risk of injuring consumers.17 To establish these propositions, we will proceed in a dialectical fashion. Part I sets forth and then rebuts a conceptual or theoretical ground that some courts have invoked as a basis for refusing to allow for the imposition of civil conspiracy or concert-of-action liability for injuries caused in the first instance by the reckless or negligent conduct of an actor. Part II then reverses course, explaining why the rebuttal offered in Part I is not of huge help to plaintiffs because it is vulnerable to a retort from defendants that would seem to limit the use of these theories to rare cases in which manufacturers coordinate their conduct in the manner of drag racers. In Part III, however, the argument takes a more plaintiff-friendly turn, harnessing settled law allowing for conspiracy-based liability for injuries caused in the first instance by conduct that amounts to deceit or fraud. Part IV points out that, while deceit is typically thought of as an intentional tort (and hence might seem to be of limited assistance to plaintiffs, given the difficulties of proving intent), the courts routinely allow for the imposition of deceit liability based on proof of misrepresentations made with reckless disregard for their falsity. On this basis, we conclude that civil conspiracy liability is and ought to be available for instances in which multiple actors agree to a market a product where the safety information provided to consumers is so inadequate or misleading that the marketing can properly be characterized as reckless with respect to consumer safety. I. The Incoherence Objection We begin by confronting what purports to be a knock-down argument against the use of civil conspiracy as a basis for holding multiple defendants liable for an injury caused by a single defendant’s defective product. According to it, the idea of conspiracy-based liability for accidentally caused harm is simply incoherent. To this objection, we respond, initially, with a put-down worthy of Torts nerds: it is negligent to suppose that there is something incoherent about the idea of conspiring to be negligent (or reckless). The logic of the incoherence argument proceeds as follows. There is no such thing as an accidental conspiracy. An actor who conspires to rob a bank necessarily does so with the intention of robbing a bank. Yet negligence and recklessness consist of unintentional injurings of others through risky conduct. To speak of a conspiracy to commit negligence is thus to posit the existence of an “intentional accident,” and there is no such thing. So goes the incoherence objection to the recognition of civil conspiracy liability in cases in which multiple defendants are alleged to have conspired to market an unduly unsafe product. Our put-down of the objection is not only nerdy, but obnoxious. After all, the objection has been embraced by prominent courts, including an illustrious panel of D.C. Circuit judges in the leading case of Halberstam v. Welch.18 Two decades after Halberstam, the Texas Supreme Court decided two cases for defendants, articulating a similar view. In Triplex Communications, Inc. v. Riley, it opined as follows:

17 With the Third Restatement Reporters, we also believe that there are grounds other than civil conspiracy or concert of action for holding a manufacturer liable for an injury caused by the negligent or reckless conduct of another manufacturer of the same product. We would describe these as instances in which a manufacturer or seller breaches an aﬀirmative duty to protect others from dangers posed by competitors’ products. 18 705 F.2d 472, 478 (D.C. Cir. 1983) (“it is diﬀicult to conceive of how a conspiracy could establish vicarious liability where the primary wrong is negligence”). We do not take a position here on whether liability for conspiracy is rightly described as vicarious or direct.

Civil conspiracy requires specific intent. For a civil conspiracy to arise, the parties must be aware of the harm or wrongful conduct at the inception of the combination or agreement. …. The “gist of a civil conspiracy” is the injury that is intended to be caused. One ‘cannot agree, either expressly or tacitly, to the commission of a wrong which he knows not of.19 A year later, in Juhl v. Airington,20 the Court was even more explicit on this point: “[b]ecause negligence by definition is not an intentional wrong, one cannot agree or conspire to be negligent.”21 A string of federal district court decisions also reject liability for conspiracy to commit negligence on the basis of the incoherence objection.22 Following the lead of other courts, and the Reporters for the Miscellaneous Provisions of the Third Torts Restatement, we deny the existence of the incoherence problem. We do so on two grounds: one logical and one doctrinal. As will become clear in Part II, however, we do not suppose that the failure of the incoherence objection entirely vindicates the Reporters or defeats all the concerns raised by courts and commentators about civil conspiracy liability for injuries caused by dangerous products. A. The Analytic Response: Distinguishing between Inadvertent and Advertent Negligence As defined in tort law, “negligence” comes in two familiar varieties: advertent and inadvertent. A driver who unwittingly crosses lanes or falls asleep while driving engages in inadvertent negligent conduct, as does the generally competent surgeon who, on an off day, incompetently nicks a nerve while performing a routine appendectomy. By contrast, a driver who is fully aware of the risks associated with driving too fast in a residential neighborhood and then fails to stop in time to avoid hitting a pedestrian, or a surgeon who delegates a procedure to a resident she knows to be insufficiently prepared to perform surgery, such that the procedure goes badly, engages in advertent negligence. In these cases, the actor adverts to the risk as they take it. Importantly, advertent negligence is still negligence; merely adverting to a risk associated with certain conduct does not of itself render the conduct reckless. A driver who knowingly drives at 45 m.p.h. in a residential area with a 30 m.p.h. speed limit, but does so at 3 a.m., when there is reason to believe that no one else is on the roads, engages in conduct that is careless, not reckless. Much of modern tort theory has focused on advertent negligence – that is, situations in which an actor consciously chooses to forego precautions that a reasonable person would take under the circumstances. Indeed, it is advertent negligence that allows for the type of cost-benefit analysis that so many academics take to be central to determinations of fault. And yet it is quite clear that a great deal of negligence liability arises from slip-ups, lapses, and other forms of inadvertent carelessness. Indeed, it is an argument against the plausibility of Hand-Formula renditions of negligence that they simply fail to capture what is negligent about garden-variety instances of inadvertent negligence. The argument that it is incoherent to speak of an agreement to engage in negligent conduct manipulates the ambiguity of the term “negligent” to reach its purportedly logical conclusion. In fact, the opposite is true. It IS intellectually challenging – perhaps even incoherent – to imagine an actor intending to engage in inadvertent negligent conduct. A driver could intentionally choose not to check their blind-spot before changing lanes, but the choice to do so would convert what is ordinarily 19 900 S.W.2d 716, 719 (Tex. 1995). 20 936 S.W.2d 640 (Tex. 1996) 21 Id. at 744.

22 See Restatement (Third) of Torts: Miscellaneous Provisions [agreement provision], Reporters’ Note to

cmt a. (T.D. No. 3, April 2024).

inadvertent negligence into advertent negligence. By contrast, it is hard to fathom how one can intend a momentary lapse in attention. But of course this concession has no application to the domain of advertent negligence (or most instances of recklessness). And, crucially, in the set of cases of present concern to us – those in which plaintiffs sue for allegedly inadequate warnings – the complaint is typically about advertent negligence. Simply put, there is nothing incoherent in an allegation that multiple manufacturers have agreed to provide to consumers or regulators a characterization of the risk-level of a product that they all sell, and that the marketing of the product with that risk characterization is negligent in the advertent sense. And this type of allegation has been central in litigation against firms that produce and sell products containing asbestos, tobacco, and the like. B. The Doctrinal Response: Concert-of-Action Liability for Drag-Racing There is another reason to reject the incoherence objection, namely, the existence of a doctrinally well-established fact pattern that gives rise to civil conspiracy liability and the tortious conduct in question is negligent (or reckless), not intentional. These are the drag racing cases that often go under the label of “concert of action.”23 In a standard drag-racing case, the driver of one car (call it the blue car) revs his engine next to the teenaged driver of another car (call it the red car), who revs his engine in response. The two proceed to race through a residential neighborhood, down a two-lane road, driving inordinately fast. The driver of the red car then loses control and crashes into a pedestrian, injuring them. In this case, the victim will have the option to sue the driver of the blue car on the ground that the two drivers were acting in concert. In other words, these are legal actors who have, by tacit agreement, intentionally engaged in a negligent or reckless course of conduct, and nobody seems to dispute that this agreement provides a sound basis for recovery against the actor who did not immediately or directly cause the plaintiff’s injury. Drag-racing cases demonstrate clearly that is not only possible but a matter of settled law that courts can impose liability on an actor for being party to an agreement to undertake a course of conduct that is careless or reckless. II. A Retort A good defense lawyer will have a response to the foregoing arguments, particularly the argument based on the drag-race example. Here is how it might go: The case of a drag race is special because it is in the nature of a race that it is competitive, and competition requires at least two parties. Without the agreement of the red car driver to race and the engagement in the race, the blue car driver could not participate in a drag race. In this respect, the agreement of the red car driver is actually a but-for cause of the blue car driver’s dangerous driving. A somewhat different analytical route also generates a foundation for liability: the unreasonably risky conduct that injures the plaintiff is the drag race itself, of which each racer’s driving is a constitutive part. The two drivers are drag racing together and it is the drag racing – taken as a unit – that injured the plaintiff. By contrast, competitors in an industry are not engaged in this sort of joint undertaking; they are competitors in a different sense. Each of them presumably would market its defective product regardless of whether the others did, and it is the dangerously defective product that has brought about the plaintiff’s injury. The agreement is 23 See. e.g., Bierczynski v. Rogers, 239 A.2d 218 (Del. 1968) (collecting cases).

not an agreement to sell these products; it is an agreement to characterize (or mischaracterize or to suppress information about) the product’s risks. Even if Firm A’s agreement with Firm B that each will supply the same inadequate warnings on their respective products is properly considered part of what renders B’s products defective, a consumer injured because of an inadequate warning on one of B’s product will not typically be able to establish that the deficient warnings on A’s products were a cause of their injury. Moreover, individual manufacturers do not sell the products that injure persons pursuant to an enterprise constituted by the group, the way a racer drives negligently pursuant to the enterprise of a drag-race. Therefore, even if it is coherent to posit an agreement or conspiracy to commit negligence, the case for imposing liability on Firm A for an injury caused by a warning defect in Firm B’s product does not actually hold up. Firm B’s selling of the defective product caused the injury, but Firm B’s action is not plausibly understood as an act of the conspiratorial group and therefore cannot be attributed to Firm A, notwithstanding the agreement between them. Conversely, while Firm B’s participation in the agreement to adhere to the party line on information about its product’s dangers can be attributed to Firm A, that participation is not plausibly depicted as a but-for cause of the plaintiff’s injury. To establish such a connection, the plaintiff would have to establish that, absent the agreement, Firm B would not have sold the product or would have provided the warnings the due care requires. The proof of a conspiracy to adhere to a party line on information does not entail that conclusion. The retort we have imagined contains two lines of argument that correspond to the two analytical routes taken in Part I to explain why it is cogent to impose liability on both drivers in dragracing cases. The simplest challenge to blue-car-driver liability is that it was the red-car-driver who struck the plaintiff and hence should be deemed to have negligently caused the plaintiff’s injury. This causation-based challenge is surmounted by pointing out that, but-for the agreement of the blue-driver to engage in the race, the red driver would not have been driving carelessly (and that such risks were entirely foreseeable). A different challenge is that the imposition of liability on the red-car driver has the flavor of vicarious liability, yet it would be unfounded to say that the red-car driver was acting as the employee or partner of the blue-car driver. The response to this concern is that the agreement between the drivers is what allows them to be treated as a unit, such that both are liable (vicariously or directly) for harms resulting from the negligent conduct of either. In other words, because the bluecar driver intentional participating in the race is constituent of it, he is liable for injuries resulting from the race, even if it was not his car that did the damage. Whether these two routes are, in the end, just different ways of establishing the same point we have not yet ascertained. It is worth noting, however, that the existence of each renders it plausible that the seemingly demanding requirements of the other need not be met. To be precise, if one thought that the vicarious liability route could stand on its own, one could – in the product-based injury scenario mentioned in the above-imagined retort – deny the necessity of a showing of but-for causation by Firm A. Conversely, if one was confident that the causation story could hold water, there would be no need to repair to a notion of vicarious liability. Understandably, the Reporters for the Tentative Draft state that the agreement need not be shown to be a cause-in-fact of the injury and that “vicarious liability” is not apt as a description of the underlying principles of civil conspiracy-based liability. While we did not contest these conclusions as such, it should be noted that there is a respectable argument that one needs at least one of these routes to hold up, and neither does in the products context.

III. Conspiracy to Commit Fraud and Products Liability Part II stated its retort in a deliberately provocative manner, leaning heavily on the defense side. The point of this presentation was to generate a more serious challenge than the incoherence objection to civil conspiracy liability for negligent or recklessness conduct, in the context of products liability cases. In this part, we respond to the challenge by identifying a doctrinal and normative basis for imposing such liability grounded in the law of deceit. It is important to recognize that in the context of the mass tort litigation highlighted by this symposium – litigation over illnesses caused by tobacco products being a prominent example – plaintiffs have sometimes credibly alleged not simply negligence or recklessness, but deceitful conduct (fraud). More generally, complaints in mass tort cases often include counts for negligence or products liability, on the one hand, and fraud, on the other. While our ultimate concern is to identify plausible versions of conspiracy liability for unintentional harms resulting to consumer from the use of products, our route to that destination will (for reasons explained below) take us through the law of civil conspiracy to defraud others. The Third Restatement of Torts – including, prominently, the its economic harm and intentional torts provisions – adopt a conventional (and arguably somewhat narrow) understanding of fraud. Section 9 of its economic harm provisions states as follows: One who fraudulently makes a material misrepresentation of fact, opinion, intention, or law, for the purpose of inducing another to act or refrain from acting, is subject to liability for economic loss caused by the other's justifiable reliance on the misrepresentation.24 In turn, this definition of fraud has been incorporated into the Restatement’s intentional torts provisions: One who fraudulently makes a material misrepresentation of fact, opinion, intention, or law for the purpose of inducing another to act or refrain from acting is subject to liability for physical harm to a person or tangible property caused by the other' justifiable reliance on the misrepresentation. The criteria for determining liability under this Section are set out in Restatement Third of Torts: Liability for Economic Harm §§ 9-15.25 Turning back to the Restatement’s economic harm provisions, one finds in Section 27 recognition of liability grounded in civil conspiracy: A defendant is subject to liability for conspiracy to commit a tort upon proof of the following elements: (a) the defendant made an agreement with another to commit a wrong; (b) a tortious or unlawful act was committed against the plaintiff in furtherance of the agreement; and (c) the plaintiff suffered economic loss as a result.26 24 Restatement (Third) of Torts: Liability for Economic Harm § 9 (Am. L. Inst. 2020). 25 Restatement (Third) of Torts: Intentional Torts to Persons § 51 (Fraud Causing Physical Harm to Person

or Property) (Am. L. Inst. 20__). 26 Restatement (Third) of Torts: Liability for Economic Harm § 27.

Taken together, these three sections would seem to indicate that the Third Restatement of Torts, taken as a whole, endorses conspiracy liability for fraud that causes physical injury and property damage.27 Assuming this interpretation is correct, it follows that, if a plaintiff can prove that: (i) two or more manufacturers of the same product (for example, tobacco cigarettes) conspired to make affirmative misrepresentations about the safety of the product to purchasers of the product, and (ii) a purchaser suffered physical harm on the basis of having justifiably relied upon the misrepresentations made by one of the manufacturers, the consumer can recover in tort from a conspirator who did not manufacture the particular product that was the source of the plaintiff’s injury. After all, an agreement to lie to the public about public safety risks would seem to count as an agreement to commit a wrong. And a tortious act was committed against the plaintiff in furtherance of the agreement so as to cause (physical) harm to the plaintiff. The application to this sort of case of conspiracy liability promises a significant expansion in liability relative to a legal regime that would not allow it. This is because a ‘direct’ (non-conspiracybased) fraud claim against an entity other than the manufacturer of the product that injured the plaintiff is unlikely to be viable. Only those who have justifiably relied upon a defendant’s misrepresentations are entitled to prevail on ‘direct’ fraud claims, and there is a good argument that the plaintiff in the imagined case did not actually rely upon a misrepresentation made by any entity other than the manufacturer of the product that injured the plaintiff. Civil conspiracy solves this problem for the plaintiff by extending liability to entities who agreed to participate in the crafting of the knowing misrepresentation made to the plaintiff by the manufacturer whose product injured the consumer.28 In short, assuming that: (i) collaboration among firms to issue a knowing misrepresentation about a product’s safety to the public, the government, and consumers counts as an agreement to commit a “wrong”; (ii) the issuance of that misrepresentation by a defendant counts as an action in furtherance of the conspiracy: and (iii) justified reliance on that misrepresentation causes the plaintiff’s injury, courts that follow the Third Restatement should be prepared to impose conspiracy liability on all collaborating firms. The foregoing analysis concerns conspiracy liability for participating in the perpetration of a fraud, but one might suppose that it can be extended to support the adoption of conspiracy liability in instances in which the relevant misrepresentation is negligent or reckless rather than made with 27 A careful parsing of the various parts of the Restatements does not necessarily yield this conclusion, but for the purposes of this article we shall assume it is valid. The problem is that the Intentional Torts Restatement contains only what appears to be a provision on joint tortfeasors or aiding and abetting liability, not civil conspiracy as such: An actor who knowingly and substantially instigates, encourages, or assists another person’s commission of an intentional tort of battery, purposeful infliction of bodily harm, assault, intentional infliction of emotional harm, or false imprisonment is subject to liability for that tort, even if the actor’s conduct does not independently satisfy all elements of the underlying tort. Restatement (Third) of Torts: Intentional Torts to Persons § 10 (“Participation in an Intentional Tort”) (T.D. No. 3) (approved 2018) (formerly §§ 109A and 109B). 28 It is worth noting, additionally, that the criteria for aiding and abetting liability will typically be satisfied a fortiori where a conspiracy to commit fraud is proved in the mass tort context. This is because the manufacturing defendant will have been assisted and encouraged substantially by co-conspirator manufacturers in formulating the fraudulent misrepresentations and ensuring that they remain viable in the informational environment. The phrase “aiding-and-abetting” connotes a secondary role, however, and it is understandable that plaintiffs’ attorneys, courts, and public health advocates more generally are disinclined to invoke a category that misleadingly suggests a merely secondary role when the defendant is one of the manufacturers that is central to the alleged conspiracy.

scienter. And yet, somewhat surprisingly, a comment to the civil conspiracy section of the Economic Harm provisions says otherwise, and seems to do so on the basis for the incoherence objection that we criticized above (and that has been rejected as unsound by the Reporters for the Miscellaneous Provisions of the Third Restatement): f. Negligence. Joining a conspiracy is an intentional act. It cannot be done negligently. Likewise, to conspire is to plan a deliberate wrong; liability does not arise for an agreement to do an act that is then found to have been negligent. But if a conspiracy is formed to commit an intentional wrong, and such a wrong is then committed, the conspirators are subject to liability for any tortious act in furtherance of the conspiracy, including an act that is wrongful because it was negligent.29 In light of the argument of Part I, this comment seems unpersuasive. Perhaps it should be interpreted as a statement about the meaning of “conspire.” Understood as such, it might be taken to say that to assert that certain actors “conspired to do X” is to imply that X was a deliberate wrong. However, even this reply is not satisfactory, for two reasons. One is that – as the Reporters for Miscellaneous Provisions explain – it is not obvious why the particular words “conspire” and “conspiracy” should control or set the substantive content of our legal norms. A second is that, as we noted above, it is entirely cogent to speak of advertent negligence having been undertaken deliberately, with knowledge that the conduct falls below the standard of reasonable care. An alternative argument in support of Comment f’s rejection of conspiracy liability for negligent misrepresentations is more aggressive. Indeed, it suggests that it is mistaken to allow conspiracy liability even for fraudulent misrepresentations – mistaken for the reasons articulated in the hypothetical retort canvassed in Part II concerning negligent misrepresentation. Even if several firms agree to adopt a warning for a particular product that contains negligent misrepresentations as to its safety, it may not be true that, each sale of a product with those warnings by one of the firms counts as an action in furtherance of the agreement among the firms. Arguably, none of the other parties to the agreement have any interest or stake in whether the seller engages in sales at all. (In a sense their only interest is that no seller adopts more fulsome warnings.) Yet, if this is true of negligent misrepresentations, why shouldn’t it also be true with respect to knowing misrepresentations? Is there anything about sales that include lies about a product’s safety that indicates such sales are – unlike sales involving negligent misrepresentations – done in furtherance of the conspiracy to adopt those lies? In fact, sales involving deliberate misrepresentations are different, and for two reasons. First, fraudulent representations, as distinct from negligent misrepresentations, typically trigger or could trigger a range of legal sanctions. Thus, the risks of making fraudulent representations are in principle much higher than with negligent misrepresentation and it is not obvious that these liability risks (regulatory and criminal) are worth taking absent some assurance that the deception will not be unearthed, perhaps via others in the industry. In this way, it may well be the case that the fraudulent sales are more tightly linked to the overall conspiracy than those done in connection with negligent misrepresentations. Note also that the Restatement’s “in furtherance” language is not legislative text; it suggests a common-law standard meant to capture an idea of connectedness that makes sense of the notion of civil conspiracy. And generally speaking, the law of torts allows for more expansive conceptions of connectedness in the domain of intentional torts as compared to unintentional torts. A second reaction cuts deeper: it suggests that the individual manufacturer-consumer transaction is not the only “unlawful act in furtherance of the conspiracy” that resulted in the plaintiff’s 29 Restatement (Third) of Torts: Liability for Economic Harm § 27, cmt. f.

injury. Many of the products at issue in the kind of litigation we are considering are subject to regulatory approval or oversight. In so far as manufacturers of such products make deliberate misrepresentations, they do so in part to deceive regulators, and the deception of regulators is itself an unlawful act. By contrast, negligent misrepresentations to regulators may not be. In other words, an agreement to lie about the dangers of a product, if used actually to deceive regulators, is a ‘completed’ wrong done in furtherance of a conspiracy to sell the product to consumers with inadequate information. Presenting to regulators a product warning in the honest but mistaken belief that it suffices to make the product reasonably safe is not a completed wrong – it only becomes one, under tort law, when the product’s warning defect causes injury to a consumer.30 IV. Recklessness as a Basis of Fraud-Like Conspiracy Liability Part II aimed to establish that the difficulties of civil conspiracy in the products liability context for “intention to engage in negligent or reckless conduct” run somewhat deeper than one might have thought. Part III, however, argued that fraud claims have a sound basis in the products liability context and that conspiracies to commit fraud are plausibly deemed actionable. At this point, a reader could be forgiven for thinking that our paper is, in the end, defense-friendly because it seems to reject civil conspiracy for negligent and reckless failures to warn of product dangers, while accepting it for fraud as a kind of consolation prize. But that is not our position. First, it may be that certain cases involving agreements to market products with inadequate warnings actually fit the drag-race/concert-of-action profile. A relatively recent case from the Eighth Circuit provides a nice example of such a case.31 Second, the background of agreement may indeed combine well with another section of the Draft Miscellaneous Provisions Restatement – Section 18A – which, in part,32 builds upon ideas like those expressed in the California Supreme Court’s Novartis opinion.33 That case holds a brand-name drug manufacturer liable for negligence in the context of misrepresentation and failure to warn by a generic drug manufacturer, against the backdrop of federal preemption doctrine that realistically supports the recognition of a duty of care from one manufacturer to the user of another company’s product. The larger picture is that the reality of relationships among manufacturers may indeed generate grounds for liability other than those that we have address here. Third and more directly to the point: products liability litigation incorporates various tort doctrines concerning misrepresentation that are not always or even typically categorized as fraud (or misrepresentation) – sometimes they may fall under the rubric of negligence or recklessness (or indeed failure to warn). In subsets of these cases, plaintiffs offering proof of negligence or recklessness may indeed satisfy the premises that provide the foundation of fraud-based civil conspiracy claims. For example, in the context of other affirmative representations, knowing concealment of safety information that changes the meaning or truth of those representations can count as fraud. In 30 It is concededly a subtle question whether the line of argument we are entertaining only applies to

formally regulated industries. We are inclined to think defrauding state or federal government actors so as to cause them not to regulate does not put the products’ manufacturers in a separate, untouchable basket. A subtler question still is whether fraud on the public, on physicians, public health experts, journalists, and consumer advocates will suﬀice. We leave that for another time. 31 Hahn v. Bader Farms, Inc., 39 F.4th 954 (8th Cir. 2022) (manufacturers of herbicide deemed liable on civil conspiracy/products liability theory based in part on joint venture to generate market for herbicide resistant seeds by promoting oﬀ-label use of herbicide). 32 See Restatement (Third) of Torts: Miscellaneous Provisions § 18A, cmts. d & q (T.D. No. 3, April 2024). 33 T.H. v. Novartis Pharm. Corp., 407 P.3d 18 (Cal. 2017).

this way, certain failures to warn are analytically akin to fraud. Relatedly, in the products liability context the marketing of a product is often understood to include a representation of safety. In some such contexts, a knowing failure to provide information that is material to safety may thus satisfy the conditions of Part III, when done as part of a conspiracy. Most importantly, as we have recently written about elsewhere, the law of deceit, like the law of defamation, has been interpreted to cover situations in which a misrepresentation is made with recklessness as to its falsity, rather than with actual knowledge of its falsity.34 In the securities fraud context, for example, lower federal courts have overwhelming held that certain forms of recklessness suﬀice to establish scienter. Recklessness in this context consists of “‘an extreme departure from ordinary care’ which presents a danger of misleading buyers or sellers that is either known to the defendant or is so obvious that the actor must have been aware of it.”35 As we have explained, in fraud law, as in the constitutional law of defamation, recklessness ﬁlls out the edges of certain otherwise ‘intentional’ torts. The willfully blind or self-servingly obtuse utterer of a defamatory statement is deemed by our law to be on par with the deliberate peddler of defamatory lies. The same goes for the reckless purveyor of misrepresentations made in the expectation that others will rely on them. Recklessness, in sum, is already elsewhere in tort law recognized as an appropriate way to round out liability for torts that, in their core instances, are intentional, including deceit (fraud). Imposition of civil conspiracy liability on manufacturers who agree to introduce into the information environment reckless misrepresentations as to a product’s safety is entirely consistent with this general approach to fraud. Conclusion [To be provided.]

34 John C. P. Goldberg & Benjamin C. Zipursky, Recklessness in Tort: Interstitial Law as Doctrinal Fine-Tuning, in INTERSTITIAL PRIVATE LAW (S. Bray, J. Goldberg, P. Miller & H. Smith eds., forthcoming). 35 Sundstrand Corp. v. Sun Chem. Corp., [1977] 553 F2d 1033, 1045 (7th Cir.) (emphasis added). Circuit courts have overwhelmingly permitted some form of recklessness to suﬀice for scienter in cases not involving forward-looking statements. 3 Thomas Lee Hazen, Treatise on the Law of Securities Regulation § 12:52 (8th ed. 2020).

20 May 2024 DRAFT

Effects of Sponsorship on Research Results Michael J. Saks ABSTRACT Does research produce different findings depending on who is sponsoring (funding) the research? If so, that is a serious problem for society as well as for science. This article reviews some of the empirical evidence on that question, illuminates some of the mechanisms through which sponsor influence might operate, and explores current and suggested strategies for mitigating the problem. INTRODUCTION In a rational and straightforward universe, science would help us figure out what in the material world is true and what is not, what works and what doesn’t, how things work, and would contribute to the more rapid solution of many of our problems.1 Science would reveal which medical interventions work, which are useless, and which cause more harm than good.2 Science would become a serious part of forensic science, helping to determine whodunnit and who didn’t.3 And, in general, science would relentlessly move our knowledge about the way the world works in one direction: forward. Honest mistakes, inadvertent incorrect turns, and dead ends would occur, of course. But they would soon be detected and corrected by researchers cheerfully committed to nothing so much as finding answers that are true. But life often doesn’t work in so constructive and straightforward a fashion. “Evidence-based medicine” is a surprisingly recent development,4 many providers are slow to follow the science, and some actively resist giving up familiar and lucrative, though ineffective, practices regardless 1 By “science” I mean nothing more elaborate than: systematic, rigorous, empirical testing. In other words, I am not talking about classic, academic, theory-driven science. 2

For a classic work on the topic, see, J.P. Bunker, B.A. Barnes, and Frederick Mosteller, COSTS, RISKS AND BENEFITS OF SURGERY (1977). 3

Michael J. Saks, Merlin and Solomon: Lessons from the Law's Formative Encounters with Forensic Identification Science, 49 HASTINGS LAW JOURNAL 1069 (1998); Michael J. Saks & Jonathan J. Koehler, The Coming Paradigm Shift in Forensic Identification Science, 309 SCIENCE 892 (2005). 4

David Sackett et al., Evidence Based Medicine: What It Is and What It Isn't, 312 BMJ 71 (1996). See, also, A.L. Cochrane, EFFECTIVENESS AND EFFICIENCY (1972).

of the evidence.5 Many areas of forensic science became established in courts long before anyone took the trouble to rigorously test them, and they have been evolving beyond their prescientific form sluggishly if at all.6 Forensic science errors (and fraud) are among the leading causes of convictions of the innocent.7 And science in general has had more difficulties than most of the public, or scientists themselves, used to imagine. Some of those failings are whole-cloth scientific fraud perpetrated in an effort to advance individual careers.8 Many others are small-scale thumbs-on-the-scale tweaks to help “find” what the researchers hoped or expected to find.9 Others are the result of unintended biases that inadvertently got woven into the system of science, such as publication bias10 and certain quirks of statistical analysis.11 Together, the varied sources of error have led to a “replication crisis” (or “reproducibility” crisis) that has tripped up and set back numerous lines of research in various scientific areas.12 The harm is not merely that many scientists and the educated13 public are growing more skeptical about the dependability of the findings that emerge from research. The greater harm is that too many of the things we thought were true turn out not to be. A survey of 1,576 researchers found that more than 70% of them have tried and failed to reproduce another scientist's experiments, and more than 50% have failed to reproduce their own experiments.14 5

See, Shannon Brownlee, Newtered, WASHINGTON MONTHLY (October 1, 2007). For a century, radical mastectomies were the preferred treatment for breast cancer, until an experiment in 1974 discovered that more conservative procedures often were as or more effective. 6

As a major report from and National Academy of Sciences notes, though they keep coming to court decade after decade, “Forensic science professionals have yet to establish either the validity of their approach or the accuracy of their conclusions, and the courts have been utterly ineffective in addressing this problem.” National Research Council, STRENGTHENING FORENSIC SCIENCE IN THE UNITED STATES: A PATH FORWARD (2009), at 53. See, also, President’s Council of Advisers on Science and Technology, FORENSIC SCIENCE IN CRIMINAL COURTS: ENSURING SCIENTIFIC VALIDITY OF FEATURE-COMPARISON METHODS (2016) and David L. Faigman et al. (eds.), MODERN SCIENTIFIC EVIDENCE: THE LAW AND SCIENCE OF EXPERT TESTIMONY (2023). 7

Saks & Koehler, supra note 3.

Stuart Ritchie, SCIENCE FICTIONS: HOW FRAUD, BIAS, NEGLIGENCE, AND HYPE UNDERMINE THE SEARCH FOR TRUTH (2020). 9

Id.

Id.; Reginald Davey, What is the Replication Crisis?, NEWS MED. LIFE SCI. 1 (2022).

Among these are work in biomedicine, economics, psychology, and others. Some or many fields have undertaken corrective measures, which provides reason for optimism. This is self-correction writ large. But it also shows how much harder it is to have consistently good science than was once thought to be the case. 13

I say “educated” public because I don’t include those members of society whose skepticism is borne of uninformed suspicion or desperation rather than serious evidence. Anti-vaxxers are a prominent example. 14

Monya Baker, 1,500 Scientists Lift the Lid on Reproducibility, 533 NATURE NEWS 452 (2016).

A world where too many medicines, surgical procedures, and other treatments don’t work as expected; where harmful products are thought to be safe; where the dangers of disease-causing chemicals and environmental pollutants are misunderstood; and so on – is a world of squandered investments, elusive safety, and less flourishing. The present article could be viewed as an examination of one particular source of potential distortion and error: Does the sponsorship of a research project influence the study’s outcome? Industry is certainly in a position to influence research, given how extensive is the involvement of U.S. industry in funding and doing research on their own products. One study examined the global landscape of industry-sponsored clinical trials by analyzing those registered on the World Health Organization’s Clinical Trials Registry.15 Their data consisted of 119,679 clinical trials in 177 countries.16 High-income countries, such as the United States, had 100 times the median number of trials per million inhabitants compared to low-income countries.17 Moreover, in highincome countries, industry sponsored three times as many trials per million inhabitants than nonindustry sponsors.18 Worldwide, industry sponsored 30.1% of the clinical trials.19 These figures might underestimate the level of industry sponsorship because they only capture the primary sponsor listed on the registry, whereas industry might be a secondary sponsor, uncounted, but still putting them in a position to influence. The intensity of sponsorship varies from one research area to another. For example, Cary Gross reported that in his field of oncology published randomized trials sponsored by industry have increased from 50% of those in high-profile medical journals to almost 90%.20 He believes that biomedicine has reached the point where industry sponsorship of clinical research has become the norm.21 Another researcher, Gary Ruskin, reported estimates that nearly 60% of medical research funding is attributable to industry.22 Pharmaceutical clinical trials are carried out by several different types of researchers and funded from several different types of sources. Pharmaceutical companies are the principal funders of 15

Ignacio Atal et al., Differential Globalization of Industry- and Non-Industry–Sponsored Clinical Trials, 10 PLOS ONE 1 (2015). 16

Id. at 4.

Id. at 1.

Id.

Id. at 4, 6, Fig. 2.

National Academies of Sciences, Engineering, and Medicine, SPONSOR INFLUENCES ON THE QUALITY AND INDEPENDENCE OF HEALTH RESEARCH: A WORKSHOP (2023), at 35. https://doi.org/10.17226/27056. 21

Id.

Id. at 51.

experiments conducted to meet requirements for regulatory approval of their products.23 Overall, about three-quarters of the researchers who conduct Phase 1 clinical trials of new drugs are employees of the company developing the drug. More than half of the researchers who conduct Phase 2 and Phase 3 trials are employees of those companies.24 Contract research organizations supply the next largest pool of researchers to conduct clinical trials, with university-based researchers not far behind. Government grants are the second largest source of funding for such research, but those grants tilt toward health conditions and treatments that are expected to be less commercially profitable. Other funding sources include biotechnology firms, venture capital and private equity, and public-private partnerships. Clearly, large investments are at risk, and those who stand to gain or lose from those investments are in a position to exercise some control over those carrying out the research that will in large part determine the fate of those new medications. What incentive does industry have to exercise influence that would advance any interests but its own? (For an overview of the many and varied influences and stakeholders in the health research ecosystem, see the diagram in Appendix 1 of this Article.) Whatever the precise extent of industry sponsorship, its prevalence could lead to a crisis of confidence in the soundness of health research in the U.S. and the world.25 To the extent that sponsorship bias weakens the evidence base used by healthcare institutions, healthcare itself is weakened, and the public’s skepticism will only grow.26 This article will discuss the ways in which sponsorship can influence research findings (Part I), review evidence that sponsorship does in fact tend to have distorting effects on research findings (Part II), and examine some suggestions about how to mitigate those undesirable influences (Part III). The focus will tend to be on industries affecting health and healthcare, though the kinds of influences found almost certainly generalize to research on any topic. I. CONCEPTUALIZING SPONSORSHIP INFLUENCE A spectrum of sponsor influence can be envisaged, from blatant demands or understandings that certain outcomes are to be produced or reported, through moderate influences, to subtle hints or no hints at all, that certain outcomes are sought.

See, FDA Development and Approval Process, https://www.fda.gov/drugs/development-approval-process-drugs

The average cost of phase 1, 2, and 3 clinical trials across therapeutic areas is around $4, 13, and 20 million respectively. 25

National Academies, supra note 20, at 51.

Brian Kennedy et al., Americans’ Trust in Scientists, Other Groups Declines, PEW RESEARCH CENTER (2022) https://www.pewresearch.org/science/2022/02/15/americans-trust-in-scientists-other-groups-declines/. (finding that only 29% of adults have great confidence in science to act in the public’s best interest, down from 40% in 2020).

A. Blatant Influences At one extreme, a sponsor of research could hire scientists for the explicit purpose of producing studies that support the sponsor’s preferred propositions, regardless of what the phenomena might truly be. Or to conduct research that produces results that are sufficiently contrary to valid, mainstream research that they raise doubts about the correct understanding of the effects of a chemical or a process or whatever the subject might be. Indeed, private research firms exist to provide exactly those kinds of services, that is, to produce whatever a corporate sponsor needs. This is what epidemiologist David Michaels refers to as the “doubt industry.”27 In his book, The Triumph of Doubt, Michaels describes the birth of the doubt industry out of the advertising industry’s efforts to protect Big Tobacco in the 1950s, and its growth and development in the service of protecting various other products and activities from regulation. Among those have also been forever chemicals, football’s encounters with chronic traumatic encephalopathy, harmful health effects of alcoholic beverages, diesel emissions, opioids, silica dust, other cancer-causing chemicals, sugar, and, of course, carbon-generated climate change.28 An energy producer, to take one example, has every right to scrutinize (and, if warranted, challenge) findings about global warming and carbon’s role in it. They can use in-house scientists or hire consulting scientists to review unwelcome research data and highlight any weaknesses found so as to raise doubts about the soundness of the conclusions. That’s good for testing claims and counterclaims, and ultimately for advancing real knowledge. Think of it as a kind of adversarial process within science – the goal of which would be to help build clearer and more accurate knowledge so society can deal constructively with our problems. But when the producer of energy is determined to produce “findings” that appear to counter the body of conventional research regardless of real evidence, the purpose of which is to muddy the scientific waters in order to fend off potential regulation, society is ill-served.29 As Michaels explains: The heart of the product defense charade is its claim to produce and publish scientific research. Research is the benchmark for true scientific expertise, and publication in a peer-reviewed journal is what makes research different from mere argument. The industry understands this very well. To help its clients who 27

David Michaels, THE TRIUMPH OF DOUBT: DARK MONEY AND THE TRIUMPH OF DECEPTION (2020). See, also, his earlier book, DOUBT IS THEIR PRODUCT: HOW INDUSTRY'S ASSAULT ON SCIENCE THREATENS YOUR HEALTH (2008) (in which Michaels quotes a tobacco executive as saying, “Doubt is our product since it is the best means of competing with the ‘body of fact’.”) See, also, Naomi Oreskes and Erik Conway, MERCHANTS OF DOUBT (2010). 28

Michaels, THE TRIUMPH OF DOUBT, supra note 27, chapters 2 and 14 (describing the work of and naming major firms in the doubt industry). 29

For an example, see Robinson Meyer, It Wasn’t Just Oil Companies Spreading Climate Denial, The Atlantic (September 7, 2022) (initially open to studying and understanding the problem, the electric energy industry later took whatever steps they could to hide the problem).

manufacture hazardous products or engage in dangerous activities, the industry has taken this model and subjected it to contortions that yield what appears to be real science. But the "studies" done by these specialists have little or nothing to do with advancing science.30 Rather than going to the trouble and expense of conducting sketchy research, they can, of course, simply lie.31 One of the best known examples is probably that of Elizabeth Holmes and her firm, Theranos, which lied about the capabilities of its magic blood testing machine, Edison.32 Tesla was sued for rigging the onboard software of its Models S and Model X cars to generate misleadingly optimistic range projections.33 A medical researcher, anesthesiologist Scott Reuben, was found to have fabricated the results of 21 studies.34 Andrew Wakefield started an epidemic of fear about vaccines causing autism with his hoax study linking the two in his 1998 publication in the medical journal The Lancet. Based on falsified medical records about a supposedly random sample of 12 children, Wakefield suggested that the measles, mumps and rubella (MMR) vaccine led to gastrointestinal symptoms, generated harmful proteins which circulated in the bloodstream, and caused autism. Numerous studies followed, involving much larger numbers of children tracked over much longer time periods, all finding no link. Later, it was learned that Wakefield had filed a patent application for his own measles vaccine and had received secret payments from a lawyer who was planning to sue MMR manufacturers and who had supplied some of the patients. Sometimes it is sufficient to simply hide harmful evidence, as the asbestos companies did for decades about the harmfulness of their product.35 B. More Moderate Influences Sponsors have more subtle ways to influence research with the goal of producing findings that are helpful to their interests. A sponsor can steer the research process in ways that tend to move it toward the sponsor’s preferred outcomes.36 30

Michaels, THE TRIUMPH OF DOUBT, supra note 27 at 23. It is possible to do both, and this might actually be the most misleading: conduct sound research and then substitute fake results for whatever the real results actually were. Years ago, while living in Boston, an acquaintance told me about his first client when he launched his political polling firm. After agreeing to the work to be done and the fees to be paid to do the work, just before they parted the client handed the young pollster a piece of paper and said: by the way, when you are finished with your research, these are the results we want you to announce. 31

John Carreyrou, BAD BLOOD: SECRETS AND LIES IN A SILICON VALLEY STARTUP (2018).

Steve Stecklow and Norihiko Shirouzu, Tesla Created Secret Team to Suppress Thousands of Driving Range Complaints, REUTERS (July 27, 2023), https://www.reuters.com/investigates/special-report/tesla-batteries-range/ 34

Brendan Borrell, A Medical Madoff: Anesthesiologist Faked Data in 21 Studies, SCIENTIFIC AMERICAN (2009).

Russell Mokhiber, Asbestos: An Industry's Lethal Cover-Up, 8 THE MULTINATIONAL MONITOR (April, 1987).

Donna H. Odierna et al., The Cycle of Bias in Health Research: A Framework and Toolbox for Critical Appraisal Training, 20 ACCOUNTABILITY IN RSCH.: POL’YS AND QUALITY ASSURANCE, 127, 130 (2013). See, also National Academies, supra note 20.

On the one hand, numerous ingenious methodological and statistical techniques have been developed to prevent bias of various kinds from entering the research process and influencing results, among them blind and double-blind procedures, representative sampling, random assignment to treatment conditions, placebos, counterbalancing, and so on. Yet their benefits can be defeated at various stages if they are omitted or corrupted. Even small distortions in the research can lead to findings that overestimate efficacy or underestimate harm.37 The cumulative impact of U.S. policies and practices on the corpus of, for example, biomedical knowledge is substantial, considering that the U.S. has been home to about 150,000 pharmaceutical clinical trials over a span of 15 years – nearly as many as the rest of the world’s top ten countries added together.38 Rasmussen et al. interviewed the lead academic investigators on 200 industry-funded drug trials, all of which contained statements assuring that the sponsor played no role in the design or conduct of the studies. Nevertheless, in interviews, those authors indicated that sponsors were involved in study design (92%), data analysis (73%), and reporting the findings (87%). Further, only a third of those interviewed stated that the author had final control over what appeared in the published report of the study and it results.39 Sponsor influence can be brought to bear at a variety of points in the research process: setting the research agenda, choosing the principal investigator to lead the research, designing the specific study, choosing the population for the study and the sampling design, conducting the study, analyzing data, drafting the report of the research, deciding whether and where to publish, and disclosing underlying data (or not) for other researchers to scrutinize and re-analyze. Each is discussed briefly below. 1. Setting the Research Agenda Sponsors can influence the research agenda in a field or an industry or in relation to concerns about a type of product or activity. They can work to limit or expand the scope of inquiry and suggest or set research objectives. For example, a sponsor whose product or service is suspected of being harmful can try to steer research in directions that will reflect less badly on their own product or draw attention to alternative causes of harm. This can be done by individual

See, Catalogue of Bias Collaboration, https://catalogofbias.org/ (citing ways that research can be manipulated to promote commercial interests: “posing a research question such that the answer is true but misleading; choosing unrepresentative study populations; administering a competitor’s drug at a non-optimal dose (in comparator trials); questionable choices made while analysing the data; non-publication of statistically nonsignificant results; selective reporting of outcomes; and multiple publication of positive results.”). 38

Leia Hawkins, Top Five Countries Running the Most Clinical Trials, PHARMA IQ, https://www.pharmaiq.com/pre-clinical-discovery-and-development/articles/top-five-countries-running-the-most-clinical-trials. 39

Kristine Rasmussen et al., Collaboration Between Academics and Industry in Clinical Trials: Cross Sectional Study of Publications and Survey of Lead Academic Authors, 363 BMJ k3654 (2018).

corporations planning their own research program40 or on a larger scale, affecting research broadly in the domain of interest. For example, industry was involved in crafting The Brussels Declaration on Ethics and Principles for Science & Society Policy-Making, shaping it to be more favorable to their interests.41 Research on gambling funded by the gambling industry looks to psychological or genetic characteristics as the causes of gambling addiction, rather than strategies employed by the industry to promote gambling behavior.42 Research funded by industries such as tobacco or highly processed foods often focuses on areas that distract from the harmful effects of their products, such as by emphasizing factors like genetic predispositions over product impact. Such efforts can produce a research picture that benefits the sponsor's commercial interests at the expense of understanding risks to public health. An analysis of studies funded by Coca-Cola found that it was much more likely to fund research on the effects of exercise than of sugar.43 Meta-research44 on nutrition studies found that foodindustry-funded studies were more likely to focus on micronutrients or minor components than on diets as a whole or dietary patterns. An oncology researcher, Cary Gross, has commented on industry sponsorship of biomedical research that, “our whole research ecosystem is so dominated by industry sponsorship that it also affects… the selection of research questions [and] the scope and objectives of clinical

See, e.g., Alice Fabbri et al., Food Industry Sponsorship of Academic Research: Investigating Commercial Bias in the Research Agenda, 21 PUBLIC HEALTH NUTRITION 3422 (2018) (finding for several food industry companies, “[a]mong the 204 publications resulting from thirty-seven disclosed research projects, the most common topic was physical activity (40·7 %), while highly processed foods were analysed in 10·8 % of the publications.”). 41

Of the 165 names on the declaration, 26 were affiliated with the tobacco or alcohol industries. The Declaration suggests, for example, that industries should have greater involvement in policymaking processes, while providing few safeguards against conflicts of interest. Jim McCambridge et al., Brussels Declaration: A Vehicle For the Advancement of Tobacco and Alcohol Industry Interests at the Science/Policy Interface?, 28 TOBACCO CONTROL 7 (2019). 42

National Academies, supra note 20, at 6.

Id.

In the present article, the terms meta-research and meta-analysis will be used interchangeably. More formally, a meta-analysis is a statistical technique which combines the results of multiple empirical studies that address the same research question. The term "meta-analysis" was coined by Gene V. Glass, Primary, Secondary, and Metaanalysis of Research, 5 EDUCATIONAL RESEARCHER 3 (1976 (writing that “meta-analysis refers to the analysis of analyses"). The first formal published meta-analysis was Mary L. Smith & Gene V. Glass, Meta-analysis of Psychotherapy Outcome Studies, 32 AMERICAN PSYCHOLOGIST 752 (1977). Meta-study is a broader concept, including but not limited to meta-analyses, embracing underlying studies and analyses of various kinds and combining them using analytic techniques of several kinds, qualitative as well as quantitative, formal studies as well as other documents and artifacts. Thus, a meta-analysis is a meta-study, but not all meta-studies are meta-analyses. Most of the research on sponsorship effects cited in the present Article, especially in Part II, are in fact quantitative meta-analyses. For present purposes, however, maintaining a careful distinction between the two terms is not worth the effort of the author or the reader.

research.”45 In other words, industry sponsorship influences not only what research gets done, but also what research needs to be done that is not being done. 2. Choosing the Principal Investigator (PI) When an industry sponsor chooses the principal investigators for the studies it funds, no one can be surprised if they prefer researchers they trust to be helpful. As noted above, over half of the research on Phase 2 and Phase 3 research in the pharmaceutical domain is conducted by employees of the company applying to the FDA to have its new drug approved. For researchers outside of the company, trust could mean people whose work the sponsor has been pleased with in the past and expects to be pleased with in the future. Or, going further, it could mean researchers who have confluences of interest, such as being paid by the industry to do other work for them. Or the company could be developing products the PI helped to create (and might even hold patents on). In many respects, cooperative relationships can be beneficial to almost everyone concerned – manufacturers, researchers, universities.46 Moreover, working together to advance the development of a useful product that is safe and effective benefits the public. Sometimes, a research question is such that a sponsor needs and wants a correct answer, whatever that answer might be. The problems arise when the confluence of interest turns into a conflict of interest: when society’s interest in rigor and candor is overshadowed by the private interest in bringing a profitable product to market.47 3. Designing the Study The design of a research project entails numerous decisions about numerous features. Normally, the whole point of study design is to try to achieve the most accurate, most unambiguous results possible concerning the research question under examination. Choosing a less ideal design, whether negligently or purposefully, can produce misleading outcomes. Among the many choices researchers have are whether the study is to be correlational (observational, epidemiological), experimental (randomized controlled trials), quasi-experimental, or something 45

National Academies, supra note 20.

By lending scientists to industry, universities gain much needed economic benefits (especially state universities, which have experienced deep cutbacks in state support over the past generation); faculty, staff, and students are able to do research they might not otherwise be able to do, some of it advancing the basic science agenda; have the opportunity to see real-world applications of their research; and have experiences that might lead faculty to develop innovations and students to find future careers. 47

The Institute of Medicine (IOM) Committee on Conflicts of Interest in Medical Research, Education, and Practice defined conflict of interest as a relationship that may place primary interests (e.g., public well-being or research integrity) at risk of being improperly influenced by the secondary, personal interests of the relationship (e.g., financial, professional, or intellectual gains). Institute of Medicine, Conflict of Interest in Medical Research, Education, and Practice (2009). See, also, Institute of Medicine, CLINICAL PRACTICE GUIDELINES WE CAN TRUST (2011) and Institute of Medicine, CONFLICT OF INTEREST AND MEDICAL INNOVATION: ENSURING INTEGRITY WHILE FACILITATING INNOVATION IN MEDICAL RESEARCH: WORKSHOP SUMMARY (2014).

else; what variables are to be included in the study; the operational measures employed to capture the chosen variables; in an experimental study, whether it is to have more or fewer conditions, embracing more or fewer comparison groups; for correlational studies, whether to employ a cohort study or a case-control study,48 which control variables are to be employed (to try to reduce the impact of confounding variables), and whether to employ more or fewer control variables; choice of population, sample size and sampling plan (affecting the ability to detect true differences and whether sub-population differences and interactions are detectable). For example, based on observational (correlational) studies of hormone replacement therapy, it appeared that estrogen treatments were quite beneficial to post-menopausal women. Later experimental designs (randomized trials) found precisely the reverse to be true: decades of misguided estrogen replacement had caused tens of thousands of unnecessary breast cancers, heart attacks, and strokes in women.49 In testing the toxicity of chemicals, researchers face a familiar dilemma which confounds population selection with study design. Because researchers usually cannot ethically expose humans to suspected dangerous substances in controlled trials, researchers face the dilemma of employing animal experiments or human correlational (epidemiological) studies, drawing data from the ordinary world of varying exposures. The animal experiments will permit strong causeeffect inferences, but uncertain generalizability to humans. Epidemiological studies have the benefit of humans, but leave unresolvable uncertainties as to cause and effect. Illustrating the impact of the operational measures, imagine two studies of family violence in the United States. One study finds that family violence is pervasive. The other study that family violence occurrs rarely. What explains the difference is the operationalization of the measure of “family violence.” The former study counted almost any kind of conflict as violence, including raising voices. In the latter study, conflict did not count as violence unless it caused injury requiring medical care.50 In studies of a new drug, effectiveness could be compared to that of the current leading medication used to treat the relevant medical condition, or the new drug could be compared to a placebo. Studies using the latter comparison group would cause the new drug to appear more effective than if it were being compared to an existing effective drug. A simple illustration of the impact of sample size is provided by a set of experiments conducted by the manufacturer of a non-prescription medication which sought to show that its product 48

A cohort study is concerned with frequency of disease in exposed and non-exposed individuals; a case-control study is concerned with the frequency and amount of exposure in subjects with a specific disease (cases) versus people without the disease (controls). 49

Jerry Avorn, POWERFUL MEDICINES: THE BENEFITS, RISKS, AND COSTS OF PRESCRIPTION DRUGS 23–38 (2004). More recent research has changed the picture again. JoAnn Manson et al., The Women’s Health Initiative Randomized Trials and Clinical Practice: A Review, JAMA (online May 1, 2024), doi:10.1001/jama.2024.6542 50

It is inevitable that setting a high threshold counts fewer of whatever is being measured, while setting a low threshold counts many more.

worked as well as the leading brand. The experiments used small sample sizes, almost guaranteed to produce a finding of no statistically significant difference. And so, when the results were in, it could be reported that in multiple randomized trials no difference was found between the two products.51 A more complicated sampling problem is illustrated by clinical trials of opioids that employed enriched enrollment randomized withdrawal (EERW).52 EERW is controversial, because participants are made physiologically dependent on the opioid in a 4- to 6-week open-label phase. Those who are not tolerating the drug well are removed from the study and the remaining participants enter a double-blind study.53 This methodology has at least two problems. First, by removing people who did not tolerate the opioid before the double-blind stage, the trial no longer represents the general population.54 Second, because participants become dependent on the drug prior to the double-blind stage, those receiving the placebo are more sensitive to pain due to withdrawal, thereby skewing apparent efficacy in favor of the drug.55 4. Conducting the Study Even if a study’s design, protocol, and methods have been specified and approved, researchers might negligently fail to conduct the study as designed or might make intentional modifications. Brian Martinson surveyed researchers and found that a third admitted to changing a study design, procedures, or analysis in response to pressure from a funding source.56 By contrast, consider how government agencies have been treating the research they contemplate funding. In an effort to ensure more rigorous research that would be transparently adhered to, the National Institutes of Health adopted a policy in 2016 known as “enhancing reproducibility through rigor and transparency.”57 Under this policy, grant applicants are required to describe their research, including planned data analysis, in great detail, and explain how their design and methods would lead to sound and unbiased results. Moreover, grant reviewers are permitted to evaluate applicants and applications based in part on the rigor of their prior research and plans to

The present article’s author was a consultant in this matter to attorneys for the established brand. The attorneys wanted to know if their statisticians could properly combine the smaller studies into a meta-analysis to increase the statistical power to reject the null hypothesis. 52

Andrew Kolodny, How FDA Failures Contributed to the Opioid Crisis, 22 AMAJ ETHICS E743 (2020), at 746.

Id. at 745.

Id. at 746.

Id.

Described in Richard Harris, RIGOR MORTIS: HOW SLOPPY SCIENCE CREATES WORTHLESS CURES, CRUSHES HOPE, AND WASTES BILLIONS (2017), at 186. Daniele Fanelli, of Stanford University, obtained similar findings. Id. 57

NIH, Enhancing Reproducibility through Rigor and Transparency, at https://grants.nih.gov/policy/reproducibility/index.htm

address weaknesses in the prior work. NIH’s principal goal is to obtain the best (most accurate) results.58 5. Analyzing Data and Interpreting Results Researchers can subtly manipulate data analyses to produce results that favor their hypotheses or the interests of their sponsors through a variety of questionable research practices (QRPs). Below are some common examples.59 P-hacking involves trying multiple statistical analyses, testing various combinations of variables, until desired statistically significant results emerge. HARKing (hypothesizing after results are known) refers to formulating hypotheses after the results are known. This makes findings appear stronger than they actually are. Selective Reporting and Cherry Picking are closely related QRPs, both involving the disclosure of data or outcomes that support a desired conclusion while omitting other relevant findings that cut the “wrong” way. Data Dredging/Mining entails searching without prior hypotheses through large datasets to find significant correlations or between-group differences. Some of these significant results will emerge by chance. (When using p<.05 as the threshold of statistical significance, one significant finding in 20 is expected by chance alone.60) While data mining can be a valid approach for generating hypotheses, presenting such findings as if they were confirmatory is misleading. Manipulating Outliers involves adjusting or removing outliers from data to make results look more uniform and statistically significant than they are. Dealing with outliers is sometimes necessary, but there should be clear justification and transparent disclosure explaining what was done, why, and with what effect. Overfitting Models refers to building a statistical model that closely fits the sample data by including too many variables. This produces apparently powerful results that are likely to fail to generalize to other data sets. Failing to Control for Multiple Comparisons refers to the fact that multiple statistical tests capitalize on chance and are likely to lead to significant results which don’t really exist in the population. Corrections for this problem exist and should be used. Another problem is how to “read” a measure or result, since measures and data do not speak for themselves. Sponsors might, for example, set the standards by which a drug is evaluated to be 58

See, sec. III.A., infra, for further description of the efforts of NIH to assure accurate, unbiased research.

See, e.g., Alex Reinhart, STATISTICS DONE WRONG: THE WOEFULLY COMPLETE GUIDE (2015).

See the cartoon in Appendix 2 which nicely illustrates this problem.

effective versus harmful, or may require the use of standards different from those used by other researchers in the same or similar fields.61 At a minimum, this complicates comparisons across studies. At worst, it can create a false impression of effectiveness of the sponsor’s drug. Publication bias complements the efforts of researchers to find favorable results by preferring to publish positive findings (effects found, hypotheses confirmed), preferably accompanied by seemingly clear and strong empirical evidence. Null effects usually are not submitted for publication and, if submitted, there is a bias against accepting them for publication. What all of the above has managed to achieve is the accumulation of a body of published studies that overwhelmingly report positive effects and hypotheses confirmed, increasing the number of those that are illusory.62 6. Research Reports and Publication Written reports memorialize the results of studies. Or not. Corporate sponsors typically require non-disclosure agreements and retain ownership or control of the products of the funded research.63 Reports might not get written at all if the results are sufficiently disappointing to the investigators or their sponsors, such as when they indicate that a drug is ineffective or has a worrisome risk of harm.64 If a report is written, sponsors might discourage or veto publication, so the data do not become part of the public scientific corpus of knowledge.65 When reports are written, they do not write themselves (not yet, anyway). Authors obviously have considerable control over which of their study’s findings are reported and which are omitted, and how those findings are expressed. They can focus attention on findings they like and bury or omit findings they don’t like.66

National Academies, supra note 20, at 6-7.

Daniele Fanelli, Negative Results Are Disappearing from Most Disciplines and Countries, 90 SCIENTOMETRICS 891 (2011) (reviewing 2,500 studies sampled from all scientific disciplines, examining how many reported positive result for the main hypothesis, finding confirmation rates varying by field from 70.2% to 91.5%). 63

Molly McCluskey, Public Universities Get an Education in Private Industry, ATLANTIC (April 3, 2017). See, also Dirk Czarnitzki et al., Delay and Secrecy: Does Industry Sponsorship Jeopardize Disclosure of Academic Research?, 24 INDUSTRIAL AND CORPORATE CHANGE 251 (2014) (finding that academic researchers’ disclosure behavior changes when the research is corporate sponsored). 64

National Academies, supra note 20, at 7-8.

See, e.g., S. Swaroop Vedula et al., Outcome Reporting in Industry-Sponsored Trials of Gabapentin for Off-Label Use, 361 NEW ENGLAND J. MED. 1963 (2009) (finding pharmaceutical companies publishing results only when favorable to conclusions of drug efficacy, or changing pre-specified primary outcomes to different outcomes when the former were not significant but the latter were); Justin E Bekelman et al., Scope and Impact of Financial Conflicts of Interest in Biomedical Research: A Systematic Review, 289 JAMA 454 (2003) (“Industry sponsorship was… associated with restrictions on publication and data sharing.”); Christopher Jones et al., Non-Publication of Large Randomized Clinical Trials: Cross Sectional Analysis, 347 BMJ f6104 (2013). 66

Odierna et al., supra note 36.

What does and does not get published, Bero has pointed out, influences the formation of additional research questions and the pictures that emerge from meta-analyses and systematic reviews.67 At a more clinical level, Mitra-Majumdar and Kesselheim note, “Reporting bias can skew the perceived risk-benefit ratio of treatments, mislead medical professionals and policy makers, and ultimately result in suboptimal medical decisions.”68 7. Disclosing Underlying Data for Reanalysis Regardless of what does or does not make it into a report of research, underlying data can be made available for scrutiny or reanalysis by others, outside of the research team that produced it, or it can be withheld.69 Numerous federal agencies require that research they have funded must be archived in databases where it will be available for re-analysis by others.70 This has the benefit of expanding what can be learned from the data, as well as helping to detect errors (or fraud) that occurred during initial analysis. Since much of our discussion has been about drug development and testing, it is noteworthy that the FDA requires data from clinical trials supporting New Drug Applications (NDAs) to be archived. The archiving is part of the FDA's efforts to make the drug review process transparent and help ensure that data supporting drug approval are reliable and subject to verification. Nevertheless, access by other researchers can be subject to restrictions, and navigating the system for gaining access can be challenging. Any data that are withheld (hidden) will not, of course, be in the archived database to begin with. In sum, the tools of research are just that: tools. In the hands of competent and honest researchers, committed to the pursuit of truth, the tools can assist in better understanding the way the world works. In the hands of those pursuing other goals, such as persuading regulators or consumers that a product or activity is as the sponsor of the research would like them to be, the tools can be manipulated to produce those impressions, or at least to generate a fog over unfavorable realities.

Bero notes that “the most stunning information comes from internal company documents…. One set, for example, showed that a sponsoring pharmaceutical company would publish the results only if they showed a statistically significant finding for drug efficacy.” When comparing “internal documents and the medical literature, they found that a cluster of studies that produced statistically insignificant (sic) results were not published, in contrast with those with statistically significant results for the same drug. In addition, they found instances where the pre-specified primary outcome was subsequently replaced with a different outcome which was statistically significant.” “Of the 19 industry-sponsored trials of gabapentin for neuropathic pain, selective outcomes were published in four papers, selective analysis occurred in 11, and seven were not published at all….” National Academies, note 20, at 7-8. 68

Mayookha Mitra-Majumdar and Aaron Kesselheim, Reporting Bias in Clinical Trials: Progress Toward Transparency and Next Steps, 19 PLOS MEDICINE e1003894 (2022). 69

Such data-sharing has long been considered an ethical obligation by scientific organizations for the benefit of scientific advancement; generally not by engineering organizations; and sometimes resisted by researchers who feel they invested a lot of themselves in the data-gathering effort and have some sort of moral claim on the data. 70

Examples of such agencies are NSF, NIH, NASA, and DOE. Archiving requirements vary somewhat in their details from agency to agency. For national security reasons, some agencies do not require, or prohibit, data sharing.

C. Invisible Influences If distorting influences all consisted of the blatant or moderate kind, and the motivations of those on both sides of the transaction (funding the research and conducting the research) were simple greed, protecting research and its results to help ensure their integrity and soundness would be simpler. That which is easy to recognize is easier to defend against. Unfortunately, influences far more subtle rival the more glaring ones in their distorting potential. Imagine a private sponsor, whose interests in the outcomes are evident, but who has agreed to remain, and has in fact remained, completely hands-off. The researchers are free to do everything they felt needed to be done to test the safety and effectiveness of, say, a new drug. And the researchers have no interest other than conducting quality research and reaching valid conclusions. Even this seemingly ideal arrangement has a substantial risk of producing outcomes that tilt in favor of the sponsor’s interests. The psychological phenomena that lead to these outcomes are subtle – so subtle that researchers are unlikely to even be aware of the biases they have taken on. Do we even ask ourselves why we root for the home team? It is almost impossible not to be influenced by one’s group associations, and by one’s consequent hopes or expectations. This subsection reviews the major types of these phenomena that threaten to distort research when it is being conducted by humans71 – which means all research. Careful thinkers have for centuries recognized that human perception and decision-making are unconsciously tainted by the hopes and expectations of the observer – a phenomenon which is therefore referred to as “observer effects” – and are not entirely the product of that which is observed. Soon after Renaissance natural philosophers began creating the scientific method, they began paying specific attention to the problem of observer effects. The writings of Sir Francis Bacon in 1620, for example, recognized the problem. Bacon suggested that "[t]he human understanding, when any proposition has once been laid down . . . forces everything else to add fresh support and confirmation; and although . . . instances may exist to the contrary, yet [the understanding] either does not observe or despises them . . . ." Bacon also posited that "it is the peculiar and perpetual error of the human understanding to be more moved and excited by affirmatives than negatives, whereas it ought duly to be impartial; nay, in establishing any true axiom, the negative instance is the most powerful." In the first passage, Bacon anticipated what modern research has shown to be the cognitive phenomenon of selective attention: the tendency of observers to seek out some information and avoid other information. In both passages, Bacon anticipated what modern cognitive scientists refer to as confirmation bias, the tendency to test a hypothesis by looking for instances that confirm it rather than by searching for potentially falsifying instances, even though most scientists and philosophers of science today agree with Bacon that the best scientific method is 71

See, generally, Harris, supra note 56; Ritchie, supra note 8. Robert Rosenthal and Ralph Rosnow, ARTIFACT IN BEHAVIORAL RESEARCH (1969), and others.

to proceed by doing the latter. Bacon adds that "[t]he human understanding resembles not a dry light, but admits a tincture of the will and passions, which generate their own system accordingly, for man always believes more readily that which he prefers." *** Bacon took a step beyond cognition and raised the issue of motivational or attitudinal effects on what a person thinks he or she has observed.72 Modern science reflects awareness of observer effects in some of its methodological conventions, most familiarly the use of placebos and double-blind experiments.73 Science abounds with examples of the problem: Sir Isaac Newton failed to report absorption lines in the prismatic solar spectrum, though they would have been clearly visible with the apparatus he was using. The most likely explanation for his failure to see them is that he held theoretically based expectations that such phenomena should not exist. Because he believed they did not exist, he failed to see them, or at least to note their presence. While Newton failed to see something that did exist, scientists of the early twentieth century saw something that did not exist. First reported by Rene Blondlot in 1903, "N-rays" appeared to make reflected light more intense. So long as they were believed to exist, the effects of N-rays were “observed” by many scientists. Of course, once it was determined that N-rays did not exist, their effects ceased to be observed. Observer effects also have been found in the reading of scales. That is, people do not always read dials and other readouts correctly, and their errors are nonrandom. Certain numbers or patterns are more likely to be “read” than others, resulting in systematic errors in the data read from the measuring instruments. For many years, laboratory technicians who counted blood cells visually were taught that correct counting would keep blood cell counts within a certain range of variation. In 1940, using a more accurate photographic method to count blood cells, researchers discovered that for years technicians had been reporting blood cell counts that were within an impossibly narrow band of variability. The technicians made observations consistent with the expectations they held, but inconsistent with reality.74

D. Michael Risinger, Michael J. Saks, Robert Rosenthal & William C. Thompson, The Daubert/Kumho Implications of Observer Effects in Forensic Science: Hidden Problems of Expectation and Suggestion, 90 U. CAL. L. REV. 1 (2002), at 6-7 [internal footnotes omitted]. 73

In a double-blind procedure, neither the participants (patients, subjects) nor the researchers administering the treatment know which condition the participant is in. 74

Risinger et al., supra note 72, at 8 [internal footnotes omitted].

By the middle of the Twentieth Century, researchers were systematically testing such phenomena and deepening our understanding of the “science of science.”75 For example, in one experiment, identical planaria (flatworms) were randomly divided into two groups. Research assistants were trained to count head turns and body contractions. Those led to believe they were observing an active group of planaria counted five times as many turns and twenty times as many contractions as did other research assistants, who had been led to believe that they were observing a relatively inactive group.76 In a more applied study, three of five fingerprint examiners who were led to believe that the latent and file prints they were about to examine probably did not come from the same person reached that conclusion, even though each was being shown a pair of prints he or she had at an earlier time examined in the course of routine casework and at that earlier time deemed to be an identification.77 In a more commonplace setting, 256 schoolteachers-in-training were asked to score exam booklets with 30 test items, each on a 4-point scale. Each test booklet had “background information” on the student, including an IQ score. Scorers gave different grades to identical performances: those whose booklets showed lower IQs gave, on average, lower scores than scorers who were led to believe the student test-taker had a higher IQ.78 Research has elaborated on the nature of these distorting observer effects, and discovered such phenomena as threshold effects, anchoring, role effects, conformity effects, and experimenter effects. Most of these phenomena can be invisible; they affect the decision-maker’s perceptions or thinking without any overt action or apparent external influence. Other times they can operate only slightly more visibly, as when a decision-maker sees evidence inconsistent with expectation, and has that evidence re-checked or re-tested (to see if it really is what it seems to be), but does not call for such re-testing of evidence when it is consistent with the decision-maker’s expectations or desires. This reflects the tendency to accept as true and correct that which comports with our expectations and desires. Though investigators think they are being highly conscientious to “double-check” the seemingly aberrant data-point or result, such behavior builds a confirmatory bias into the investigation. Many of the phenomena mentioned above fall under the heading of “confirmation bias” – the unconscious tendency to search for, interpret, be more satisfied with, and recall information in ways that tend to confirm one's preexisting beliefs or hopes while implicitly discounting 75

Robert Rosenthal, EXPERIMENTER EFFECTS IN BEHAVIORAL RESEARCH (1966).

Lucien Cordaro & James R. Ison, The Psychology of the Scientist: X. Observer Bias in Classical Conditioning of the Planarian, 13 PSYCHOLOGICAL REPORTS 787 (1963). 77

Itiel E. Dror et al., Contextual Information Renders Experts Vulnerable to Making Erroneous Identifications, 156 FORENSIC SCI. INT'L 74 (2006). 78

L.S. Cahen, An Experimental Manipulation of the "Halo Effect": A Study of Teacher Bias (1965) (unpublished manuscript, Stanford University), described in Rosenthal, supra note 75.

evidence or conclusions that go counter to expectation or desire. In the context of the research enterprise, this bias can influence everything along the path from the conception of a research question to the write-up of the report of research findings. Where desire for a preferred conclusion leads one to work harder to reason one’s way to the conclusion, that is confirmation bias’s more muscular cousin, “motivated reasoning.”79 An important source of motivation is group membership or identification. A classic experimental examination of these phenomena was a study by Hastorf and Cantril.80 After a highly controversial and rough football game between Princeton and Dartmouth, researchers at each university showed a film of the game to a sample of each school’s students. As they watched the film, they were to mark down any infractions they observed and indicate if the infraction was “mild” or “flagrant.” Though they were watching identical films, they perceived different behavior. In the eyes of Princeton students, Dartmouth players committed more than twice as many total infractions as were perceived and judged by Dartmouth students. These results have long been understood to suggest that people’s group identification affects what they perceive of the events unfolding in front of their eyes. In Sherif et al.’s famous “Robber’s Cave” experiment, youngsters at a summer camp were randomly formed into groups which then engaged in a range of activities, many of them competitive.81 Though they had been formed on a purely random basis, the groups solidified quickly, showing loyalty and solidarity, and soon showed in-group biases against the out-group. When asked to systematically evaluate the performance of individuals in tasks and games, ingroup members were seen as being better (on average) and the out-group’s members as poorer – even though they all had witnessed the same physical events. A Dutch researcher team wanted to find out how little it took to turn individuals into groups displaying classic in-group versus out-group biases.82 They, too, worked with schoolchildren, who were at random divided into two different groups, each given its own name. That was not sufficient to produce in-group/out-group effects. The next step was to tell the children that gifts were to be given to them, but not enough were available for everyone. So a flip of a coin would determine which of the two groups’ members would receive the gifts. Now in-group/out-group phenomena took hold.

Motivated reasoning has been described thus: “On occasions when we have come to be motivated to reach a particular conclusion, we draw on information selectively and interpret it in ways that bring us to the conclusions we desire to reach. This process occurs automatically and without awareness; we feel we are just thinking about the matter and arriving at the correct conclusion.” Michael J. Saks and Barbara A. Spellman, THE PSYCHOLOGICAL FOUNDATIONS OF EVIDENCE LAW (2016), at 159. 80

Albert Hastorf and Hadley Cantril, They Saw a Game: A Case Study, 49 J. ABNORMAL & SOCIAL PSYCH. 129 (1954). 81

Muzafer Sherif et al., THE ROBBERS CAVE EXPERIMENT: INTERGROUP CONFLICT AND COOPERATION (1961, 1988).

Jacob Rabbie and Murray Horwitz, Arousal of Ingroup-Outgroup Bias by a Chance Win or Loss, 13 JOURNAL OF PERSONALITY AND SOCIAL PSYCHOLOGY 269 (1969).

The instantaneity with which such affiliations form has been found among adult experts in a now-famous experiment by Murrie and colleagues.83 About 100 forensic psychologists and psychiatrists were hired and paid to review the same offender case files. Half were led to believe that they were consulting for the defense and the other half for the prosecution. The psychologists and psychiatrists reviewed each of four offender files using two well-established risk-assessment instruments. Though they were reviewing identical case files, those who thought they were working for the prosecution assigned substantially higher risk (of criminal harm) scores to offenders than did those who thought they were working for the defense. Murrie et al. dubbed this the “allegiance effect.” Thus, it takes very little for group affiliation or identification to distort the perceptions of anyone from children to well-trained professionals. Add competition (another company or research team trying to develop a similar product) or a few frustrations (regulators) and group cohesion solidifies. Litigators can see this happen with the experts witnesses they hire. Though experts formally have the judge and jury as the audience to whom the expert owes its candor and loyalty, everything about the expert-attorney relationship promotes allegiance: who hires, who pays, who interacts with the expert, whose goals are the focus, the sense that an out-group exists (the “other team”). Judges know all this (perhaps from their own experiences as lawyers), and instinctively distrust those experts. But when judges hire the same experts as the court’s witnesses under Rule 706, their trust in and regard for expert typically rockets to the skies.84 Robert Ciadini, probably the most prominent scholar of social influence, has identified a number of mechanisms through which these subtle influences work, whether in group contexts or merely dyadic relationships.85 Cialdini refers to them as “weapons of influence.” The ones most relevant to the problem we are discussing are the following.86 The most potent weapon of influence is reciprocity. People inherently feel a strong pull to return favors, gifts, information, contributions, concessions, etc. to others from whom they have received something. As far as we know, this is a universal, doubtless evolved, feature of being human. Remarkably little is required to activate reciprocity and thereby change important behaviors. For example, astonishingly small gifts from pharmaceutical company representatives is all it takes to alter doctors' prescribing patterns.87 The principle of authority is that people are 83

Daniel C. Murrie et al., Are Forensic Experts Biased by the Side That Retained Them?, 24 PSYCHOLOGICAL SCIENCE 1889 (2013). 84

Joe S. Cecil and Thomas E. Willging, COURT-APPOINTED EXPERTS: DEFINING THE ROLE OF EXPERTS APPOINTED UNDER FEDERAL RULE OF EVIDENCE 706 (Federal Judicial Center, 1993) and Michael J. Saks, The Phantom of the Courthouse, 35 JURIMETRICS J. 233 (1995) (book review of Cecil & Willging). 85

Robert Cialdini, INFLUENCE: THE PSYCHOLOGY OF PERSUASION (2006).

The weapons of influence omitted from the present discussions are commitment and consistency and scarcity.

Troyen A. Brennan et al., Health Industry Practices that Create Conflicts of Interest: A Policy Proposal for Academic Medical Centers, 295 JAMA 429 (2006). See, also, Ashley Wazana, Physicians and the Pharmaceutical Industry: Is a Gift Ever Just a Gift? 283 JAMA 373 (2000).

more willing to comply with the directions or recommendations of a person they perceive as having relevant authority or expertise. This would be relevant not only for employees of a pharmaceutical company conducting clinical trials, but also for outside researchers with whom the company has contracted. Social proof is the concept that people tend to believe what other people tell them or suggest to them. They are even more willing to take a requested or suggested course of action if they view the others as similar to themselves on relevant dimensions. Liking is the principle that people are inclined to comply with the preferences of those they know and like. All that has to happen for this effect to exert its influence is that the corporation’s representatives or supervisors with whom researchers work, or report to, be warm and friendly. Unity brings us back around to the concept of being a unit, a group, the more cohesive the better. We humans have evolved to prefer being team players. In all of the discussion above, researchers can be unaware of the influences and effects on their motivation and cognition. Sometimes, though, we are tempted to consciously cheat. Dan Ariely, one of the leading students of the psychology (and behavioral economics) of dishonesty has identified the following factors as facilitators of increased dishonesty in people who might otherwise be scrupulously honest. Ariely begins with the master influence, reciprocity, echoing Cialdini. Next is the presence of a conflict of interest. These are conflicts between what will benefit an individual and the duty that individual owes to a patient, a client, a customer, an employer, or the public. Ariely highlights the human mind’s extraordinary ability to engage in rationalization to enable the conflict to be resolved in the individual’s interest. That means justifying to oneself that one is still a good and honest person while gaining a benefit at the expense of others. If the individual is living or working within a culture where dishonesty in certain matters is regarded as acceptable, or where the individual can observe others behave dishonestly, it becomes that much easier to be that much more dishonest.88 In short, we can be complicit in producing less than honest, accurate research findings and still feel righteous about ourselves – or not even be aware we have been transformed, at least somewhat, by the situation we are in. II. EVIDENCE OF SPONSORSHIP INFLUENCE We’ve discussed blatant, moderate, and invisible influences on research and its outcomes. What is the evidence that they actually do distort research findings and conclusions? Does sponsorship, through any or all of the mechanisms described above bring about different research outcomes? There is a line of research variously known as "sponsor influence," “sponsorship bias,” or “sponsorship effects” devoted to examining how the funding of research might affect the outcomes, quality, and direction of scientific studies. The National Academies of Sciences, Engineering, and Medicine recently published the proceedings of a workshop which brought together quite a few of the people working in that area, who discussed the problem, research on it, future directions, and related issues.89 This was a valuable resource for the present article. 88

Dan Ariely, THE (HONEST) TRUTH ABOUT DISHONESTY: HOW WE LIE TO EVERYONE – ESPECIALLY OURSELVES (2012). 89

National Academies, supra note 20.

The major type of research in this area is the meta-study. One cannot do “a study” to assess whether sponsorship affects research outcomes. Looking at a single study sponsored by one source and determining whether its results would have been different if the study had been sponsored by a different source is not possible. Instead, one has to compare collections of studies which differ as little as possible except on the factor of concern: who is sponsoring the research – a charitable foundation, a government agency, a for-profit corporation. This is the world of metaresearch: studies of studies.90 Other types of research exist in this domain: looking at amounts of sponsorship, study protocols, published reports, company records (often in the context of litigation), interviews with researchers and funders – looking for indications of influence and bias. But nothing so clearly focuses on research outcomes as a function of sponsorship as meta-analyses of clinical trials of products (and procedures). The chief concern of the present review paper is that corporations, which typically have intense interest in the outcomes of research, might in one way or another influence the results of that research where more neutral sponsors would not. As we saw in the immediately preceding subsection, the sponsor need not do anything improper to engender distorting effects. Meta-research on the problem of sponsorship effects dates at least as far back as the mid-1980s.91 Those studies generally find that research sponsored by industry produces results that favor the industry’s products – finding more benefits, larger benefit, and less harm. This section will summarize some of that meta-research, focusing on more recent meta-studies.92 This cannot be a comprehensive review of the many meta-studies on the wide range of subject matters affected by the problem of sponsorship effects. It is, rather, illustrative of the research and the findings. We begin with a meta-study of meta-analyses of various drugs and medical devices. A. Drugs and Devices Lundh et al. (2017) gathered and analyzed meta-analyses of groups of studies of drugs and devices.93 Thus, theirs is a study of studies (meta-analyses) of studies (individual experimental or observational studies). A total of 75 papers (meta-analyses) were found and included. The 90

As noted earlier, see supra note 44, meta-analysis and meta-study have slightly different meanings. In the present article, however, they are used interchangeably, along with meta-research. 91

See, Ross A. Davidson, Source of Funding and Outcome of Clinical Trials, 1 J. GEN. INTERNAL MEDICINE 155 (1986). 92

Later meta-analyses typically include the studies that had been included in earlier meta-analyses, incorporating additional studies conducted after publication of the earlier analyses. 93

Andreas Lundh et al., Industry Sponsorship and Research Outcome, COCHRANE DATABASE OF SYSTEMATIC REVIEWS (2) (2017), https://doi.org/10.1002/14651858.MR000033.pub3.

median number of studies included per paper was 105 (ranging from nine to 930) – in short, thousands of underlying empirical studies. “Of the 75 papers [meta-analyses], 27 reported data on both favorable outcomes and risk of bias, 44 on favorable outcomes only, and four on risk of bias only. Twenty-five papers were non-industry funded, two were industry funded, one was funded by both industry and non-industry sources, 17 reported that they did not receive funding and 30 did not report on funding.”94 Because studies included or omitted various measures of interest (benefit, harm, bias, sponsorship, etc.), only subsets of studies could be included in any given analysis. Appendix 3 to the present Article reproduces Table 1 from Lundh et al., listing the 24 metaanalyses that measured effect sizes in industry-sponsored versus non-industry-sponsored studies. This Appendix provides readers with a more concrete sense of the range of subject matters of the underlying studies (anti-psychotics, nicotine replacements, statins, dental implants, etc.), and the differences in effects sizes found by industry-sponsored versus non-industry-sponsored studies (ranging from no significant differences to highly significant differences). In general, compared to non-industry-sponsored studies, industry-sponsored studies produced more evidence of beneficial effects (the equivalent of 638 studies per thousand favorable to 502 per thousand favorable), less evidence of harmful outcomes (the equivalent of 649 per thousand to 474 per thousand), and more favorable conclusions (the equivalent of 863 per thousand to 644 per thousand). Moreover, industry-sponsored studies reflected less agreement between the reported results and stated conclusions than did non‐industry sponsored studies. That is, the occurrence of a discrepancy between data-based results and verbally expressed conclusions was more frequent in industry-sponsored studies. Interestingly, no overall differences were found in studies’ apparent vulnerability to bias, leading Lundh et al. to conclude, “Our analyses suggest the existence of an industry bias that cannot be explained by standard ‘Risk of bias’ assessments.” Various other studies of studies focused more on particular subject matter areas. The discussion below briefly summarizes some of those other meta-studies. B. Sugar-sweetened Beverages In the area of nutrition, one major question has been the health effects of high levels of sugar. In one meta-study, Schillinger et al. (2016) reviewed research conducted over a 15-year period, finding 28 randomized controlled trials plus 32 systematic reviews or meta-analyses 95 Primary outcomes of interest were markers of diabetes and obesity. Of 26 studies that found no relationship between sugar-sweetened beverages (SSB) and disease outcomes, 25 had funding from industry. Only one of the 34 studies that found harmful health effects were industry

Id., at 9-10 (internal citations omitted).

Dean Schillinger et al., Do Sugar-Sweetened Beverages Cause Obesity and Diabetes? Industry and the Manufacture of Scientific Controversy, 165 ANNALS OF INTERNAL MEDICINE 895 (2016).

sponsored. That was a 32.7 fold greater likelihood that industry-supported research would find SSBs to produce no adverse health effects. In another study of SSB studies, Litman et al. (2018) reviewed 133 studies published between 2001 and 2013, and conducted a meta-analysis controlling for a number of potentially confounding variables.96 Their summary of findings: “A ‘strong’ or ‘qualified’ scientific conclusion was reached in 82% of independent [versus] 7% of industry-related SSB studies” (a strongly statistically significant difference). “Industry-related studies were overwhelmingly more likely to reach ‘weak/null’ conclusions compared with independent studies regarding the adverse effects of SSB consumption on health” – 57 times more likely for studies supported by industry. C. Dairy Chartres et al. (2020) examined studies of the effects of dairy intake on cardiovascular disease and mortality.97 On the one hand, “There was no clear evidence of an association between studies with industry ties… versus no industry ties… and the reporting of favourable results… [nor in] favourable conclusions….” On the other hand, the studies did differ significantly in the size of the risk of cardiovascular disease (CVD) that was reported: “Studies with industry sponsorship… showed a decreased magnitude of risk of CVD outcomes compared with studies with no industry sponsorship….” D. Various Nutritional Supplements Chartres et al. (2016) conducted a broader study of nutritional studies, including all metaanalyses and reviews of nutrition studies that could be found by searching research indexes from their inception through 2015.98 Chartes et al. looked at the relationship of sponsorship to statistically significant findings in the underlying reports, effect sizes, and the reviews’ conclusions. Two reports (meta-analyses) examined statistical significance obtained in the underlying empirical studies as a function of sponsorship. The hundreds of underlying studies addressed various nutritional issues, such as efficacy and harm of synbiotics, probiotics, prebiotics, and calcium supplementation, on eight different clinical outcomes. “Overall, industry-sponsored studies reported 20.6% (73 of 354) of all clinical outcomes as favorable compared with non– 96

Ethan A. Litman et al., Source of Bias in Sugar-Sweetened Beverage Research: A Systematic Review, 21 PUBLIC HEALTH NUTRITION 2345 (2018). Litman et al. were testing the concern that bias goes in both directions: not only that industry funding biases research in favor of finding minimal harm, but that independent funding might support researchers biased in the opposite direction (sometimes referred to as the “white hat bias”). They sought to test this dual problem “with research involving sugar-sweetened beverages (SSB) as a test case, focusing on a period during which scientific consensus about the adverse health effects of SSB emerged from uncertainty.” 97

Nicholas Chartres et al., Association of Food Industry Ties with Findings of Studies Examining the Effect of Dairy Food Intake on Cardiovascular Disease and Mortality: Systematic Review and Meta-Analysis, 10 BMJ e039036 (2020). 98

Nicholas Chartres et al., Association of Industry Sponsorship with Outcomes of Nutrition Studies: A Systematic Review and Meta-Analysis, 176 JAMA INTERNAL MEDICINE 1769 (2016).

industry-sponsored studies, which reported 16.7% (9 of 54) as favorable.” The difference is not statistically significant. This illustrates that industry-sponsored studies do not inevitably produce industry-favorable results. Notably, the great majority of the industry-sponsored studies failed to find benefits for the nutritional supplements being tested (281 of 354). Regarding effect sizes, “Only 1 report including 88 observational studies and RCTs [randomized controlled trials] examining sugar-sweetened beverages and various health outcomes assessed the relationship between industry sponsorship and effect size. For the harmful outcome of energy intake, overall effect size was smaller in industry-sponsored compared with non-industrysponsored studies” (a statistically significant difference) “and for the outcome of body weight, effect size was also smaller in industry-sponsored vs non–industry-sponsored studies” (also statistically significant). “However, no significant difference in effect size was observed among RCTs.”99 E. Anti-viral Medication Dunn et al. (2014) examined 37 assessments of studies testing the efficacy of anti-viral medication (e.g. Tamiflu) for treating influenza.100 Of those reviews conducted by authors with financial conflicts of interest with pharmaceutical manufacturers of the drugs, 7 of 8 reached favorable conclusions about efficacy 87.5%). Of reviews conducted by authors without financial conflicts of interest, 5 of 29 reached favorable conclusions (17.2%). Thus, it was five times as likely that reviews with financial ties to manufacturers would reach conclusions supportive of efficacy of the drugs. That difference is statistically significant.101 Dunn et al. concluded, “Money distorts science.” F. Passive Smoking Barnes & Bero (1998) analyzed articles reviewing research concerning the effects of secondhand smoke (on lung cancer, heart disease, respiratory disorders, etc.), comparing the results of industry-associated research to that of comparable reviews by non-tobacco-industry-affiliated authors.102 Barnes & Bero found that review studies by tobacco-industry-affiliated authors were significantly more likely to report no harmful effects of second-hand smoke. Passive smoking was concluded not to be harmful by 37% of reviews overall, of which 74% were tobaccoaffiliated reviews. Put slightly differently, 94% of tobacco-affiliated reviews concluded that second-hand smoke was not harmful, while only 13% of non-tobacco-affiliated reviews reached 99

Id. (internal citations and statistical analysis data omitted). This finding suggests that enforcing rigorous methodology would help make research more resistant to undesirable influences. 100

Adam G. Dunn et al., Financial Conflicts of Interest and Conclusions about Neuraminidase Inhibitors for Influenza: An Analysis of Systematic Reviews, 161 ANNALS OF INTERNAL MEDICINE 513 (2014). 101

The author of the present Article calculated a significance test for the data in this paragraph. Chisquare = 11.10, p<.001. 102

Deborah E. Barnes & Lisa A. Bero, Why Review Articles on the Health Effects of Passive Smoking Reach Different Conclusions, 279 JAMA 1566 (1998). Industry sponsorship was also associated with restrictions on publication and data sharing.

that conclusion (and 87% concluded passive smoking was harmful). After controlling for a number of factors, author affiliation was the only variable significantly correlated with the conclusions reached by the reviews. The industry-sponsored/affiliated research was an astonishing 88 times more likely to find no harm being caused by second-hand smoke. This linw of research is almost certainly the product of the blatant influence discussed, supra, in section I.A. The various meta-studies described above comparing research that had been industry-sponsored versus non-industry-sponsored generally supports the concern that results are affected by who pays for the research. However, the impact of sponsorship ranges widely from reflecting the extreme bias of blatant influence to the other extreme of no effect at all. Optimists will find good news there, suggesting that sponsorship effects are robust but not pervasive. Under some conditions research findings can be the same whether the study is sponsored by industry or not. III. POSSIBLE SOLUTIONS To varying degrees, bias and conflicts of interest are omnipresent. They are impossible to abolish and quite difficult to neutralize. This part of the article explores ways that various entities have undertaken to try to manage and minimize those distorting influences. The purpose of the following discussion is not to endorse any particular solutions, but to give readers a sense of the nature and range of various adopted and proposed solutions. A. NIH: Minimizing Bias in Evaluating Proposals, Awarding Funds, Requiring Registration, Reports, and Data-sharing The challenges presented by bias and conflicts of interest are nicely illustrated by the lengths to which the National Institutes of Health (NIH) go in trying to prevent such influences from infecting the numerous decisions occurring at every stage of the process of funding, conducting, analyzing, and reporting the extramural research they fund. NIH sponsors (funds) research projects proposed by researchers (principal investigators (PIs)) outside of their organization (extramural). NIH’s aim is to ensure that every decision turns on nothing but the merits of the issues and that every finding reflects nothing but the phenomena under study.103 In reviewing research proposals, NIH demands of itself “expert assessment, transparency, impartiality, fairness, confidentiality, security, integrity, and efficiency.”104 Proposals are assigned to subject-matter-appropriate sections. They are reviewed by peer experts from outside as well as inside the agency. Reviewers are required to disclose and certify their own conflicts of interest both before and after their study section meets.105 Persons with serious conflicts are prohibited from reviewing. NIH separates functions such that staff members who oversee the doing of the research are not the same people who manage the reviewing. A formal appeals process is available to PIs who believe their proposal did not receive unbiased review. Before a 103

Various other federal research funding agencies have similar procedures, but NIH’s are as rigorous as any.

104

Patricia Valdez, in National Academies, supra note 20, at 42.

105

NIH’s financial conflict of interest policy is based on HHS regulation 42 CFR part 50 subpart F.

final funding decision is made, proposals and the reviews of proposals are reviewed by the relevant institute’s or center’s advisory committee. Proposed research is required to be as rigorous methodologically as feasible given the nature of the research problem – defined as “the strict application of the scientific method to ensure unbiased and well-controlled experimental design, methodology, analysis, interpretation, and reporting of results” – and that rigorous research design will be transparently adhered to. Grant applicants are required to describe their research, including planned data analysis, in detail and explain how their design and methods will lead to sound and unbiased results. In addition, reviewers may evaluate applicants and applications based in part on the rigor of their prior research and current plans to address weaknesses in the prior work. Funded research projects are required to be promptly registered at the ClinicalTrials.gov website and a report summarizing results posted within one year of the project’s primary completion date.106 Further, pursuant to a detailed data-sharing policy, data emerging from NIH-sponsored research must be made available to other researchers through an approved data repository to permit re-analysis and extended analysis. Important to note about NIH’s funding is that the sponsor itself has no obvious bias or conflicts of interest. It is not trying to build support for a product it hopes to sell and profit from. Its purpose is to help generate valid new knowledge. Finding that something is not true or does not work is as useful to the advancement of knowledge as finding something that does work. There is, of course, the risk that staff or reviewers have their own pet theories or favor certain approaches or investigators over others. Those, too, need to be guarded against. But the advantage it has as a sponsor is that the principal message its institutional culture sends is one of seeking high quality research that discovers truths, whatever they may be. The very extent of the care taken by NIH in its funding of extramural research highlights the challenge of removing bias and its potential distorting effects from research. Private sponsors are not known to go to the same lengths. Indeed, to greater or lesser degrees, some private sponsors allow or invite bias favorable to the company’s products to enter the research process. What NIH does with extramural research is harder to achieve in other contexts, though not impossible. B. Conventional and Contemporary Strategies Many of the conventional strategies for managing bias in research are reflected in the description above of NIH’s extramural research funding. These strategies can be, and often are, used in other contexts as well.

106

Thus, whether or not the work results in publication in scientific journals, reports of the work become available to the public and other researchers, such as for use in meta-analysis.

1. Conventional strategies Some, like demands for scientific rigor and disclosure of conflicts of interest, have been around for quite some time and can be regarded as conventional strategies to try to remove or manage bias. Rigorous scientific methods. As noted earlier, researchers have many decisions to make about the design of their research. Quite apart from manipulating study design to shepherd studies in a desired direction, it has long been recognized that weaker study designs are more likely to lead to misleadingly optimistic results, only to be corrected years or decades later by more rigorous studies.107 Some sponsors, notably federal funding agencies, already insist on use of the best research designs and methods to which a research question is susceptible. Reviewers, commentators, and consumers of research also can and should give greatest weight to the best studies and least to the poorest. Disclosure of conflicts of interest. Disclosure of conflicts of interest has long been relied on in many contexts. A government funding agency or private foundation can use candid disclosures to assess whether to employ reviewers or fund proposals. Unsurprisingly, some researchers, not to mention sponsors who have a self-serving stake in findings being taken to be valid, are less than eager to make damaging disclosures. The most concerning conflicts of interest are probably the ones most likely to remain undisclosed.108 Consequently, the withholding of important information ought to be treated like other kinds of scientific misconduct. Disclosure requirements can be more or less demanding. The field of cardiology, for example, has been “proactive around issues of professionalism and ethics by identifying four essential components of a COI compliance program: disclosure of interests, assessment of interests, management of interests, and oversight and enforcement of a conflict management plan.”109 All four components were said to be essential, yet still not sufficient. The trouble with mere disclosure is that consumers of knowledge (much like everyday consumers of financial services and healthcare) don’t know what adjustments to make in their decisions and actions in response to the disclosure.110 Many consumers react to disclosures of conflicts of interest by having increased trust in the professional: the fact of disclosure is taken as a sign of trustworthiness. Disclosures of conflicts could, however, be incorporated into a metaanalysis that adjusts the weight accorded to various studies commensurate with their methodological quality. Conflicts of interest could be included to reduce a study’s weight. 107

See, e.g., Mosteller, supra note 2 and Avorn, supra note 49.

108

McCluskey, supra note 63 (providing an example involving an academic consultant to DuPont Pioneer hiding that connection in articles advancing the corporation’s interests). 109

National Academies, supra note 20; Ivor J. Benjamin, 2020 American Heart Association and American College of Cardiology Consensus Conference on Professionalism and Ethics: A Consensus Conference Report, 77 J. American College of Cardiology 3079 (2021). 110

See comments of Dunn, National Academies, supra note 20, at 27; Ariely, supra note 88, at 88-92.

2. Contemporary strategies A number of strategies have been spawned in response to the “replication crisis” or “reproducibility crisis.”111 Though they were not developed as a response to the problem of sponsorship influence, some of these strategies could be useful in confronting the sponsorship problem as well. • Preregistration of studies. Preregistering studies involves documenting research methods and hypotheses before conducting the research. This helps prevent data dredging and p-hacking. It facilitates comparison of the study design to what actually was conducted. It also makes replication by other researchers easier. • Improved statistical standards. Some fields are ratcheting up their standards for significance levels (e.g., from p<0.05 to p<0.005), promoting alternative statistics such as confidence intervals (also known as uncertainty intervals) and Bayesian methods, and encouraging more robust statistical practices.112 • Transparency. Increased transparency embraces everything a research team can do, on the assumption that the more that other researchers and the public know about a study, the better able they are to evaluate it. These transparency-promoting efforts include open sharing of research protocols, methods, data, and analyses (in an accessible repository).113 Journals are increasingly requiring authors to submit raw data along with their manuscripts for publication, and of course disclosure of conflicts of interest. • Replication. Funding bodies and journals are increasingly recognizing the importance of replication, and therefore increasingly open to publishing replication studies or negative

111

Davey, supra note 11. Sixty-seven percent of researchers in the medical sciences reported not being able to reproduce the results of at least one of their peers’ studies. Id. at 2. Bayer HealthCare reported in 2012 that it was unable to validate three-quarters of the preclinical studies it attempted to replicate as a prelude to doing further work on those topics. Amgen could reproduce the results of only 11% of the landmark cancer research studies its researchers tried to replicate, says a 2012 report. A project aimed at replicating 100 important psychology experiments obtained only a 36% success rate. Economists were unable to replicate nearly 40% of the studies that had been published in two of their leading journals. 112

See, e.g., Geoff Cumming, UNDERSTANDING THE NEW STATISTICS: EFFECT SIZES, CONFIDENCE INTERVALS, AND META-ANALYSIS (2012). 113

NIH expanded its data-sharing policies beginning in January 2023. National Institute of Health, NOT-OD-21013, Final NIH Policy for Data Management and Sharing (2020), https://grants.nih.gov/grants/guide/noticefiles/NOT-OD-21-013.html. The principle behind the new policy is that results of research funded by taxpayers should be freely available, whatever the findings. The purpose section of the data-sharing policy states that datasharing “enables researchers to rigorously test the validity of research findings, strengthen analyses through combined datasets, reuse hard-to-generate data, and explore new frontiers of discovery.” Id. The scope of the policy is broad and applies to all research, funded or conducted in whole or in part by NIH, that results in the generation of scientific data.

results.114 Historically, they have been difficult to publish, thereby reducing incentives to do replications. • Collaborative science. This involves the encouragement of collaborations across labs, disciplines, and borders to increase the diversity of thinking and scrutiny, canceling out of biases, and ultimately increasing robustness of findings. The most interesting collaborations are between researchers whose labs have been producing contradictory results. • Audit and evaluation. Some fields have introduced audits of research labs and projects. This includes checking the raw data, software code, and experimental setups to ensure they meet predefined standards. • Technological tools and platforms. New tools are developing to facilitate data sharing, version control, and collaborative analysis. • Promotion of open science. This movement encourages making scientific research, data, and dissemination accessible to all levels of an inquiring society, amateur or professional. Open access repositories and journals facilitate this by allowing free access to research outputs.115 C. More Innovative Strategies If the interests of a sponsor – no matter how hands-off the sponsor remains and how free researchers are to carry out a study of the sponsor’s research question – nevertheless affect the decisions researchers make and distort the ensuing results, then something more far-reaching is needed than what has been described to this point. Note that for the general approaches described in the following subsection, agency intrusiveness deceases from the first to the last while total research costs increase. 1. In General One approach would be to separate the researchers from the funding source. For example, a regulatory body (e.g., FDA) could require applicants to deposit funds sufficient to cover the cost of the necessary research in an account from which the agency could draw. The agency would select an independent research team (using many of the procedures used by NIH, described above) to carry out the research project. The research team would report to the agency and would

114

For one important effort to promote replication, see, After 10 Years, ‘Many Labs’ Comes to an End – But Its Success Is Replicable, https://news.virginia.edu/content/after-10-years-many-labs-comes-end-its-success-replicable 115

UNESCO describes open science as an “inclusive construct that combines various movements and practices aiming to make multilingual scientific knowledge openly available, accessible and reusable for everyone, to increase scientific collaborations and sharing of information for the benefits of science and society, and to open the processes of scientific knowledge creation, evaluation and communication to societal actors beyond the traditional scientific community.” UNESCO Recommendation on Open Science, UNESCO 2, 7 (2021).

remain blind to whose product or process they are testing. This would be a radical departure from current practice, in which the FDA is an ex post facto sorter of more-or-less biased evidence.116 Another approach might be to place monitors inside the corporate-sponsored research team. Monitors would have open access to the work and members of the research team, and thereby be privy to the design, methods, measures, analyses, draft reports, etc. The monitors would be paid by the relevant government agency (with funds provided by the corporate applicant. A third approach might be random replication. A random subset of research initially carried out under the sponsorship of a private sponsor would be replicated by an independent research team as part of a program developed for just this purpose. The principal advantage of this approach is that every industry-sponsored research team would be aware that a certain probability exists that their project will be subjected to replication. The principal disadvantage is that the great expense of reproducing a clinical trial.117 Some well-established working models exist which reflect some of the ideas described above. Two are described next. 2. The Italian Medicines Agency (AIFA) In 2004, Italy, which is ranked eighth in the world for the number of clinical trials conducted,118 created regulations and a funding mechanism to support non-profit pharmaceutical research.119 All pharmaceutical companies that operate in Italy are required to contribute an amount equal to 5% of their annual expenditures on all promotional activities. Each year, AIFA announces a call for proposals for much-needed research, beginning in the early years with a focus on orphan drugs, pharmacovigilance, and comparative studies on drugs and therapeutic strategies. Non-profit organizations apply for research funding, provided that the researchers have complete control over the study design and conduct, have no involvement in patents, promise to publish the research, and have no involvement in the drug approval process.

116

Through its Office of New Drugs within the Center for Drug Evaluation and Research (CDER), the FDA reviews new drug applications. See, Office of New Drugs, Food and Drug Admin., https://www.fda.gov/about-fda/centerdrug-evaluation-and-research-cder/office-new-drugs; FDA's Drug Review Process: Continued, Food and Drug Admin., https://www.fda.gov/drugs/information-consumers-and-patients-drugs/fdas-drug-review-process-continued; How Drugs are Developed and Approved, Food and Drug Admin., https://www.fda.gov/drugs/developmentapproval-process-drugs/how-drugs-are-developed-and-approved. 117

Also referred to as “competitive self regulation.” Roger Koppl, How to Improve Forensic Science, 20

EUROPEAN J LAW ECON 255 (2005). 118

Hawkins, supra note 38.

119

AIFA Research & Development Working Group, Feasibility and challenges of independent research on drugs: The Italian Medicines Agency (AIFA) experience, 40 EUR. J. OF CLINICAL INVESTIGATION 69 (2010).

All clinical trials in Italy must be designated as either for-profit or non-profit, which affects the uses to which research may be put.120 Non-profit research must be conducted by a public or nonprofit entity that has no economic interest in the drug and no connection to the marketing authorization holder.121 Pharmaceutical companies may provide support through an in-kind donation of the necessary products.122 The trials had to be conducted to improve clinical knowledge and practice, but – prior to 2022 – could not be used for marketing or registration purposes.123 Starting in 2022, non-profits are permitted to transfer data from their clinical trials to pharmaceutical companies for drug registration and marketing. Detailed rules control how the transfer is to be accomplished and who reimburses whom for what, always maintaining a separation of the researchers from the pharmaceutical companies, and with transparency and regulatory oversight.124 3. The Health Effects Institute (HEI) HEI was founded in 1980 at the behest of Congress125 as an independent public-private partnership. This purpose of this research organization is to conduct studies of the health effects of vehicle emissions.126 It receives balanced funding split between the Environmental Protection Agency (EPA) and the automobile industry.127 The board of HEI is appointed by the EPA administrator and core industry sponsors.128 The board then appoints the research and peer review committees without the approval of any of its

120

Roberto Cursano et al., Preclinical and Clinical Trial Requirements / Italy, Pharma Boardroom (2022), https://pharmaboardroom.com/legal-articles/preclinical-and-clinical-trial-requirements-italy/. 121

Id.

122

Id.

123

Id.

124

Laura Liguori & Elisa Stefanini, The New Decree on Non-Profit Studies: Data and Results may now be Sold, Portolano Cavallo (2022), https://portolano.it/en/blog/life-sciences/the-new-decree-on-non-profitstudies-data-and-results-may-now-be-sold. 125

42 U.S.C.A. § 7409(d).

126

Id.

127

Health Effects Institute, What is the Health Effects Institute?, https://www.healtheffects.org/about.

128

Id.

sponsors.129 Both committees are composed entirely of academics from relevant fields.130 Every five years, HEI adopts a research agenda. Priority topics can be suggested by the sponsors as well as by the scientific and environmental communities. Neither the board nor the sponsors has any control over selecting research teams, implementing the research, or reviewing the results before publication. The balanced-funding approach of HEI and its governance, much like Italy’s non-profit clinical trial laws, is thought to remove any desire researchers might have to produce favorable results for the sponsor. Having two major, fundamentally different, sponsors who are distanced from researchers by layers of governance and oversight prevents researchers from feeling they have any sponsor to please. This appears to have worked well for decades, with each center of interest – government regulators and industry – trusting the process and their respective roles in it so that they also accept the findings. Creating something like the HEI for other regulator-industry pairs, such as the FDA and the pharmaceutical industry, might be possible. CONCLUSION [To be written later]

129

Id.

130

See, Health Effects Institute, Research Committee, https://www.healtheffects.org/about/researchcommittee; See, also, Health Effects Institute, Review Committee, https://www.healtheffects.org/about/review-committee.

Appendix 1. Influences and stakeholders in the health research ecosystem.

Source: Adrian Hernandez in National Academies of Sciences, Engineering, and Medicine, Sponsor Influences on the Quality and Independence of Health Research: A Workshop (2023).

Appendix 2. An illustration of how data dredging can produce spurious statistical significance.

Source: Randall Munroe, at https://xkcd.com/882/

Appendix 3. Brief descriptions of some meta-analyses from Lundh et al. (2017)

SESSION IV: ETHICAL DILEMMAS REGARDING LEGAL REPRESENTATION Deal making and the Ethically Slippery Slope in the Quest for Global Peace Elizabeth Chamblee Burch, University of Georgia Corporate Lawyers, Disloyalty and the Opioid Crisis Elise Maizel, Michigan State University What About Purdue’s Lawyers? A Brief History of Lawyer Gatekeeping From Social Trustee Professionalism Through the 2023 ABA Model Rules Amendment Bradley Wendel, Cornell University Discussant: Myriam Gilles, Yeshiva University

(Back to main agenda)

THE 30 TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

Corporate Lawyers, Disloyalty, and the Opioid Crisis Elise Bernlohr Maizel*

Introduction ................................................................................................ 2 Opioids and Disloyal Corporate Lawyers .................................... 7

I. a.

The Manufacturers ................................................................................... 8

The Distributors ..................................................................................... 11

The Pharmacies ...................................................................................... 14

II.

Corporate Lawyers and Legal Obedience .............................. 16

a. Dominant Narratives About Lawyers for Lawbreaking Corporations .................................................................................................... 16 b.

The Ultra Vires Doctrine and Corporate Legal Obedience ........ 19

The Ultra Vires Doctrine as a Corporate Lawyer’s Constraint 21

III. A Disloyalty Reframe and its Implications for Corporate Lawyers ....................................................................................................... 23 a.

Pedagogical Interventions ................................................................... 23

Doctrinal Implications.......................................................................... 24

Expressive Function .............................................................................. 26

Conclusion ................................................................................................. 28

* Assistant Professor of Law, Michigan State University College of Law. Many thanks to Richard Brooks, Brittany Farr, Myriam Gilles, Maria Glover, J. Travis Laster, Noah Rosenblum, Cathy Sharkey, and the participants of New York University’s Lawyering Scholarship Colloquium for helpful conversations and feedback in various stages of the drafting process of this Essay. Thanks to Jessica Halpert and Giavanna LaGamba for excellent research assistance.

Introduction In 2010, in Roanoke, Virginia, a teenager named Spencer Mumpower sold heroin to a young man named Scott Roth.1 After the purchase—a casual transaction between friends who shared the same tragic habit—Scott died of an overdose. Spencer was prosecuted and ultimately sentenced to eight years in prison for his role in Scott’s death.2 Scott’s passing was one of many deaths in the Roanoke Valley caused by legal and illegal opioids, which took hold of Appalachia (and later the whole country) in the years following the introduction of Purdue Pharma’s OxyContin.3 Beginning in the early 2000s, Purdue flooded the region with the powerful opioid pills, which its lawyers (and ultimately the FDA) claimed were less likely to cause addiction because of their continuous release formula. This was not true. In fact, the drug was powerfully addictive, systemically overprescribed, prone to abuse, and frequently lethal. When law enforcement and regulators began to discover what Purdue had long known—that OxyContin was a source of deadly addiction causing a major public health crisis—it was far too late. A broad swath of Americans, including a shocking number of young people in Southwestern Virginia, were already addicted. When OxyContin supply “dried up” and many who had become addicted to prescription opioids were forced to switch to illegal substances like heroin, death and destruction of lives predictably followed. What happened to Scott and Spencer is directly traceable to corporate decisions. The stakes of corporate compliance are life and death for multitudes whose lives corporate activity will ultimately touch. It has become cliché to ask, in cases of major corporate

1 Scott Roth was my high school classmate and my family knows the Mumpower

family. I dedicate this Essay to Scott’s memory and to Spencer’s recovery. 2 Lawrence Hammack, Man to Serve 8 Years in Heroin Death, THE ROANOKE TIMES, (Mar. 1, 2012) https://roanoke.com/archive/man-to-serve-8-years-in-heroindeath/article_b3cd7337-e0b5-5ee5-ab8a-70c851748ae6.html. 3 Former Roanoke Times reporter Beth Macy’s book, Dopesick, draws a direct line between Purdue’s introduction of OxyContin in the Roanoke Valley and overdose deaths from heroin like Scott’s. BETH MACY, DOPESICK: DEALERS, DOCTORS, AND THE DRUG COMPANY THAT ADDICTED AMERICA (2018).

misconduct, “where were the lawyers?”4 As in so many past cases of societal harms brought about by corporations, in the opioid crisis, lawyers were everywhere. They were involved in the creation and perpetuation of the opioid crisis.5 And they helped corporations avoid accountability for the millions of deaths that resulted. Accordingly, this Essay asks a slightly different version of that age old question—is there something about corporate lawyers and corporate lawyering in particular that helped to create and perpetuate the opioid crisis? I propose the answer to this question is yes. There is a prevailing cultural narrative that lionizes zealous advocacy by a lawyer on behalf of their client. Aspirational young lawyers are raised on stories of brave and selfless lawyers pursing justice on behalf of wronged clients. We see these narratives play out in fiction—Atticus Finch!6 And in the true stories of the great lawyers—Bryan Stevenson!7 Many law students come to law school seeking to emulate these zealous advocates, deploying “zealous advocacy” in service of a wide variety of clients and causes. But, beneath the surface, the generally applicable norms around the kind of lawyering that is acceptable, commendable, and prized in the context of representing individuals do not map neatly onto corporate representation. A posture that might be deemed “zealous advocacy” in the individual context—for example cooperating as little as possible with regulators—fails to take on the same moral valence when a corporation is a client.8 This is not just because the 4 See, e.g., Donald C. Langevoort, Where Were the Lawyers? A Behavioral Inquiry

Into Lawyers' Responsibility for Clients' Fraud, 46 VAND. L. REV. 75 (1993); Bernard S. Carrey, Enron—Where Were the Lawyers, 27 VT. L. REV. 871 (2002); Hendricks, Ed, and Mary Berkheiser, Where Were the Lawyers?, LITIGATION 18, no. 4 (1992): 30–65. http://www.jstor.org/stable/29759558. 5 See Section I, infra. 6 See, generally, Jonathan A. Rapping, It's A Sin to Kill A Mockingbird: The Need for Idealism in the Legal Profession, 114 MICH. L. REV. 847 (2016). 7 When I taught 1Ls at NYU Law, quite a few cited Bryan Stevenson’s Just Mercy in explaining why they came to law school. See BRYAN STEVENSON, JUST MERCY: A STORY OF JUSTICE AND REDEMPTION (2014). 8 See Robin West, The Zealous Advocacy of Justice in A Less Than Ideal Legal World, 51 STAN. L. REV. 973, 973–74 (1999) (“Lawyers embark on their careers expecting to be engaged in the pursuit of justice, and by the end of a life spent representing corporate clients, come to realize and lament that they have done much the opposite.”).

individual client may be more sympathetic than the organizational one. Rather, there are important structural and theoretical differences between the two. Corporate lawyers represent a legal fiction—a non-human organization whose existence is premised on a promise that it will engage in only lawful conduct.9 State law incorporation statutes allow corporations to organize under the laws of a state with one limitation: that they will engage only in “lawful business.”10 Consistent with that promise, management is charged with running the corporation within the boundaries of the law. Under the ultra vires doctrine, engaging in illegal conduct is outside the power of the corporation and, therefore, impermissible. Under the prevailing orthodoxy of corporate law, the ultimate (and perhaps only permissible) objective of the corporation is to generate the greatest possible value for shareholders.11 This is often at odds with legal compliance. Corporate scholar Eugene Soltes colorfully observed that “it may be profitable for a firm to provide financial services to a drug cartel, but . . . [i]n return for the privilege of operating within a jurisdiction, firms agree to abide by certain rules and regulations or else face punishment.”12 In other words, the corporation must work to generate value for its shareholders but it may not do so outside the bounds of the law.

9 Elizabeth Pollman, Corporate Oversight and Disobedience, 72 VAND. L. REV. 2013,

2018 (2019) (“An endless variety of businesses may organize through the corporate form, but as a matter of fundamental principles, they are all required to engage in only lawful conduct. Corporate statutes enshrine this prime directive of obedience.”). 10 See, e.g., Del. Gen. Corp. L. 101(b). 11 For detailed discussion of this corporate purpose debate, compare LYNN STOUT, THE SHAREHOLDER VALUE MYTH (2012) and STEPHEN BAINBRIDGE, THE PROFIT MOTIVE (2023); see also Dorothy Lund & Elizabeth Pollman, Corporate Purpose, OXFORD HANDBOOK OF CORPORATE LAW AND GOVERNANCE (Jeffrey N. Gordon and Wolf-Georg Ringe eds., Oxford University Press). 12 Eugene Soltes, Evaluating the Effectiveness of Corporate Compliance Programs: Establishing A Model for Prosecutors, Courts, and Firms, 14 N.Y.U. J.L. & BUS. 965, 976–77 (2018).

When corporate lawyers13 encourage, enable, or willfully blind themselves to a corporation’s acts that violate the law, this behavior is often described as zealous advocacy.14 Scholars of corporate law and professional responsibility often describe the problem as overzealousness or an excess of loyalty.15 And the market for corporate clients prizes sharp elbows and slick practices.16 But a reading of this conduct that is consistent with corporate law theory should conceive of it as something different—in fact zealous advocacy in service of corporate misconduct is not an excess of loyalty but instead disloyalty. Legal compliance is a mandatory objective of the firm. Though corporate agents generate most of a corporation’s objectives, wants, and needs, positive law supplies the mandate of legal compliance. That requirement is a fundamental component of corporate

The term “corporate lawyers” is used in this Essay to mean lawyers who represent corporate clients in all contexts, not just transactional ones. 14 See, e.g., William T. Allen, Corporate Governance and a Business Lawyer’s Duty of Independence, 38 SUFFOLK U. L. Rev. 1, 3 (2004) (describing the “increasing domination of the zealous advocacy ideal” among corporate lawyers). 15 See, e.g., Robert W. Gordon, A New Role for Lawyers?: The Corporate Counselor After Enron, 35 CONN. L. REV. 1185, 1194 (2003) (“The classic defense of the corporate lawyer's role, both most often advanced and held in reserve if other defenses fail, is of course that we are advocates, whose duty is zealous representation of clients.”); Charles M. Yablon, The Lawyer As Accomplice: Cannabis, Uber, Airbnb, and the Ethics of Advising "Disruptive" Businesses, 104 MINN. L. REV. 309, 334 (2019) (“In recent years, there has been increasing concern that many corporate lawyers have stepped over the line from technicians to accomplices.”); Paula Schaefer, Behavioral Legal Ethics Lessons for Corporate Counsel, 69 CASE W. RES. L. REV. 975, 982 (2019) (“There is a mismatch between zealous-advocacy-within-the-bounds-of-the-law and the duty of a corporate lawyer advising about possibly fraudulent or criminal conduct.”); David B. Wilkins, Team of Rivals? Toward A New Model of the Corporate Attorney-Client Relationship, 78 FORDHAM L. REV. 2067, 2068 (2010) (lamenting the zealous advocacy norm’s application in the context of corporate representation); see also W. Bradley Wendel, Civil Obedience, 104 COLUM. L. REV. 363, 364 (2004) (describing the “zealous advocacy within the bounds of the law” as the “dominant view” and collecting sources espousing that position). 16 See J. Travis Laster & Elise B. Maizel, Discovery as a Compliance Problem, 50 J. CORP. L. __ (forthcoming 2024) (describing intense competition amongst law firms for business); Melissa Mortazavi, Code of Silence, 40 CARDOZO L. REV. 2171, 2209 (2019) (describing the market power of clients in encouraging discovery abuses and other instances of bad behavior by attorneys). 13

existence and not subject to debate.17 For agents, the only way to be loyal to the firm is to counsel the client to obey the law. Anything else is ultra vires and disloyal to the corporate client. Reconceptualizing a lawyer’s complicity in a firm’s legal violations as disloyalty flips the standard. It has significant implications and calls for novel interventions. The first is pedagogical. Legal education about professional responsibility must include enough corporate law theory to make clear that counseling acts in violation of law constitutes disloyal to the corporate client. The second is doctrinal. Reframing the counseling of law-breaking as disloyalty changes the nature of the agent’s accountability to the corporation. Currently, doctrines like in pari delicto shield lawyers who are overzealous accomplices in corporate misconduct. Reframing that conduct as betrayal undermines the doctrinal premise for the in pari delicto shield and enables successor managers to seek accountability on behalf of the corporate client. Finally, there is an expressive dimension. Reclassifying this behavior as attorney disloyalty recognizes that the same public policies that constrain corporations also constrain their attorneys. It also reinforces the truism that corporate lawyers owe their duties to the corporate entity itself, not the managerial agents who might cause the corporation to engage in misconduct.18

17 See Elizabeth Pollman, Corporate Disobedience, 68 DUKE L.J. 709, 719 (2019)

(“It is a fundamental principle of corporate law that corporations are to be engaged only in lawful activity. Put another way, corporate law is broadly enabling within the constraint that corporations act inside the bounds of the law.”); Kent Greenfield, Ultra Vires Lives! A Stakeholder Analysis of Corporate Illegality (with Notes on How Corporate Law Could Reinforce International Law Norms), 87 VA. L. REV. 1279, 1314 (2001) (describing the ultra vires doctrine’s remaining component in corporate law—the “doctrine sets off illegal activities as ultra vires for corporations.”). 18 Greenfield, supra note 17 at 1324-25; see also Pollman, supra note 17 at 2016 (explaining that the requirement of corporate legal obedience “embeds a safety valve for public policy in the obligations of fiduciaries that cannot be eliminated”); Dorothy S. Lund, Public Primacy in Corporate Law, 47 SEATTLE U. L. REV. 365, 368 (2024) (“There are, however, reasons to believe that promoting an obligation to the public . . . could be viewed as a core focus for corporate law and governance. . . . consistent with early corporate law, which required corporations to demonstrate a public benefit as a condition of receiving a charter”).

Ultimately, this framework envisions that corporate lawyers in the opioid business might have owed more to Scott Roth and to Spencer Mumpower. Although the attorneys they never knew did not represent them, those same lawyers counseled corporations to behave in ways that fundamentally changed—and in one case ended—their lives. It follows that the legal architects of the crisis that led to the Scott’s death and Spencer’s incarceration might be accountable for the harms they made possible. This Essay proceeds in three parts. Part I provides a descriptive account of some of the ways corporate lawyers created and perpetuated the opioid crisis, and how corporate law practice norms continue to shield corporations who profited from a public health epidemic. Part II compares the cultural model of a “zealous advocate” with the doctrinal constraints of corporate law, ultimately arguing that the ultra vires doctrine should be seen as actively restricting the corporate lawyer. As a result, I argue, we should conceive of aggressive lawyering in service of corporate misconduct not as “zealous advocacy” but instead as betrayal of the corporate client. Part III concludes by mapping the pedagogical, doctrinal, and expressive implications of a “disloyalty reframe.” I.

Opioids and Disloyal Corporate Lawyers

This section tells three stories of lawyers’ roles in the opioid epidemic. The first is of Howard Udell, the general counsel for Purdue Pharma. Udell’s aggressive tactics paved the way for Purdue to flood the market with OxyContin. The second is of the outside law firms who, from time to time, reviewed and approved a compliance program at AmerisourceBergen that minimized suspicious order reporting to the DEA. The third is of Walmart’s current litigation counsel who, in shareholder litigation, have made such aggressive assertions of privilege that we may never know quite how Walmart failed to prevent its own massive compliance failures. These are hardly the only stories. There are other manufacturers, distributors and pharmacies in the opioid business. But together these stories highlight a few worrying features of modern corporate lawyering. The first is the way that lawyers understand their duties to their clients. The stories of Purdue and

of Walmart highlight a zero sum, “win at all costs” approach that runs directly counter to the public interest. The second is the way corporations use outside law firms—hired under limited representations and with less accountability—to legitimize and justify corporate decision-making. The story of AmerisourceBergen demonstrates how the reputations of outside law firms can be used to create an accountability vacuum—management can point to outside approval to justify their decisions and outside advisors can, in turn, point to their outside status and limited scope to avoid responsibility. Together these features helped to create the opioid crisis, perpetuate it, and continue to help its perpetrators avoid accountability. a. The Manufacturers “It’s not the job of a lawyer to tell management that the company can’t do what it needs to do.”19 Howard Udell was a corporate lawyer whose client was almost always Purdue Pharma. 20 As general counsel for Purdue, Udell engaged in what some might call “zealous advocacy.” His aggressive tactics on behalf of the firm made it possible for a dangerous drug to hit the market without the typical regulatory scrutiny associated with the introduction a new and powerful narcotic.21 In the 1980s, Purdue began manufacturing an opioid called MS Contin—a slow release morphine pill that was the precursor to OxyContin.22 Udell directed the company to release the drug before

PATRICK RADDEN KEEFE, EMPIRE OF PAIN: THE SECRET HISTORY OF THE SACKLER DYNASTY 77 (2021) (quoting Howard Udell, General Counsel and Chief Legal Officer, Purdue Pharma). 20 U.S. v. The Purdue Frederick Company, Inc. D/B/A The Purdue Frederick Company, Michael Friedman, Howard R. Udell & Paul D. Goldenheim., 2007 WL 1423896, 1 (W.D.Va.). 21 RADDEN KEEFE, supra note 19 at 161-62. 22 “MS Contin”, FDA, https://www.accessdata.fda.gov/drugsatfda_docs/label/2010/019516s034lbl.p df. 19

filing the FDA-required New Drug application.23 When the FDA sought to compel Purdue to file the application, Udell went over the regulator’s head, appealing directly to political actors in the Reagan administration.24 By the mid-1990s, Udell and other senior executives at Purdue knew that MS Contin was being abused on the street.25 At the same time, despite problems with clinical trials, Purdue began the process of getting FDA approval for OxyContin.26 Purdue executives cultivated a relationship with an FDA official in order to finesse the approval process, ultimately allowing the company slip a highly unusual bit of text into the drug’s package insert: “Delayed absorption, as provided by OxyContin tablets, is believed to reduce the abuse liability of the drug.”27 Purdue then began an aggressive marketing campaign for the drug, instructing its sales representatives to promote OxyContin as less addictive.28 Eventually, the fact that OxyContin and its predecessor, MS Contin, were outrageously addictive came to light.29 In the personal injury lawsuits that followed, Udell bragged that Purdue had no intention of paying a penny to personal injury plaintiffs.30 He assembled a team of outside lawyers whose past experience in senior roles in government lent all their clients a patina of seriousness and respectability.31 He blamed the addicted, claiming that their problems predated OxyContin and that the company owed them 23

RADDEN KEEFE, supra note 19, at 79.

24 Id.

Barry Meier, Origins of an Epidemic: Purdue Pharma Knew Its Opioids Were Widely Abused, N.Y. TIMES (May 29, 2018), https://www.nytimes.com/2018/05/29/health/purdue-opioids-oxycontin.html. 26 Shraddha Chakradhar & Casey Ross, The History of OxyContin, Told Through Unsealed Purdue Documents, STAT NEWS (Dec. 3, 2019), https://www.statnews.com/2019/12/03/oxycontin-history-told-throughpurdue-pharma-documents/; Art Van Zee, The Promotion and Marketing of OxyContin: Commercial Triumph, Public Health Tragedy, 99(2) AM. J. PUB. HEALTH 221 (Feb. 2009). 27 RADDEN KEEFE, supra note 19, at 195-96. 28 Art Van Zee, supra note 26 (“The distribution to health care professionals of branded promotional items such as OxyContin fishing hats, stuffed plush toys, and music compact discs (“Get in the Swing With OxyContin”) was unprecedented for a schedule II opioid”). 29 Barry Meier, Origins of an Epidemic, supra note 25. 30 RADDEN KEEFE, supra note 19, at 227. 31 The murderers’ row of former government heavyweights retained by Purdue included Eric Holder, Mary Jo White, and Rudy Giuliani. Id. at 247-48. 25

nothing.32 And all the while, Udell worked within the company to limit the ways damaging information was documented, instructing sales reps not to put negative feedback about the drug in writing and destroying old documents and emails.33 Eventually the house of cards fell. The DOJ, which had been investigating Purdue since 2001, was contemplating charging three senior Purdue officers—including Udell—with a number of felonies including mail fraud, wire fraud, and money laundering.34 In 2007, Udell, Michael Friedman (Purdue’s former CEO) and Paul Goldenheim (Purdue’s chief medical officer) agreed to plead guilty to a single charge of misdemeanor misbranding.35 The misbranding charge was brought against the three executives under the “responsible corporate officer” doctrine, meaning that all three conceded guilt but denied any specific responsibility for the harm caused.36 Because of the nature of the charges, Udell faced no prison time37 but was instead sentenced to pay a fine and perform community service.38

32 RADDEN KEEFE, supra note 19 at 245-46.

33 Soo Youn, One Year After OxyContin Launched, Purdue Pharma Execs Applied

to Patent 'Self-Destructing' Emails, ABC NEWS (May 28, 2019), https://abcnews.go.com/Business/year-oxycontin-launched-purdue-pharmaexecs-applied-patent/story?id=63320323. 34 Edward Helmore, Purdue Pharma Escaped Serious Charges Over Opioid in 2006, Memo Shows, THE GUARDIAN (Aug. 19, 2020), https://www.theguardian.com/usnews/2020/aug/19/purdue-pharma-oxycontin-justice-department-memoopioid. 35 News Release: The Purdue Frederick Company, Inc. and Top Executives Plead Guilty to Misbranding Oxycontin; Will Pay Over $600 Million, United States Attorney’s Office, Western District of Virginia (May 10, 2007) https://media.defense.gov/2007/May/10/2001711223/-1/1/1/purdue_frederick_1.pdf. 36 Miriam H. Baer, The Information Shortfalls of Prosecuting Irresponsible Executives, 70 DEPAUL L. REV. 191, 198 (2021). 37 I do not mention Udell’s lack of prison time to suggest that justice would have been served by more incarceration. Rather, I note Udell’s lack of prison time because it is striking compared to the eight years Spencer Mumpower served. One wonders how many deaths Udell’s actions might have caused. 38 Id. at 193; Barry Meier, 3 Executives Spared Prison in OxyContin Case, N.Y TIMES (July 21, 2007), https://www.nytimes.com/2007/07/21/business/21pharma.html.

Udell was unusual in that he faced criminal accountability for his role in Purdue’s destructive acts, but by the time he pleaded guilty in 2007, the crisis had already been set in motion. It would take another decade for prescription opioid deaths to hit their peak, killing a record 17,029 Americans.39 b. The Distributors “There’s a crisis raging in America outside the board's doors. For them to completely ignore it and just rely on some whitewashed presentation . . . just because some law firm or advisor comes in and gives you a 40-page presentation that's very fancy and has lots of colors, it doesn't mean that it actually addresses the issue.”40 In the United States, nearly a third of all wholesale drug distribution is controlled by a one company—AmerisourceBergen.41 In the early 2000s, as governmental actors became increasingly aware of the dangers of prescription opioids, the Drug Enforcement Agency (DEA) increased its scrutiny of pharmaceutical distributors.42 These distributors were charged with “antidiversion” responsibilities and required to report suspicious orders to the government—in essence, professionals within the company had a duty to monitor to make sure the prescriptions they distributed were not being diverted into the wrong hands. The strategies AmerisourceBergen pursued to maximize profits made its anti-diversion tasks difficult. The company sought out partnerships with independent pharmacies rather than large chains because independent pharmacies had less market power.43 Drug Overdose Deaths, NATIONAL INSTITUTE OF HEALTH: NATIONAL INSTITUTE ON DRUG ABUSE (June 30, 2023) https://nida.nih.gov/research-topics/trendsstatistics/overdose-death-rates. 40 Plaintiffs’ counsel, Oral Argument on Defendants’ Motion to Dismiss, Lebanon Cnty. Employees' Ret. Fund v. Collis, No. 2021-1118-JTL, 48:2-11 (Sept. 23, 2022) (hereinafter, “Collis Oral Argument Transcript”). 41 Lebanon Cnty. Employees' Ret. Fund v. Collis, 311 A.3d 773, 778 (Del. 2023) (hereinafter, “Collis II”). 42 Lebanon Cnty. Employees' Ret. Fund v. Collis, No. 2021-1118-JTL, 2022 WL 17841215, at *5 (Del. Ch. Dec. 22, 2022) (hereinafter, “Collis I”). 43 Collis I at *5-6. 39

However, independent pharmacies also had comparatively less sophisticated and less resourced programs for suspicious order monitoring, leaving them far more vulnerable to becoming “pill mills.”44 At AmerisourceBergen, the professional in charge of antidiversion at the height of the opioid crisis was Chris Zimmerman, who served in a variety of roles for AmerisourceBergen including the role of Chief Compliance Officer. From his office outside Philadelphia, Mr. Zimmerman frequently mocked people in Appalachia and the south who had become addicted to prescription pills. He called them “pillbillies” and joked about supplying OxyContin to their children.45 An email from his employees mocked Appalachians as illiterate. Mr. Zimmerman was not a lawyer, but worked alongside in-house attorneys such as the general counsel and chief compliance counsel.46 Over the course of Mr. Zimmerman’s tenure, his division made a series of changes to the anti-diversion program’s order monitoring protocol that dramatically reduced the number of orders for opioids that would be flagged and reported to the DEA.47 Plaintiffs in a later shareholder derivative suit would claim these changes were designed to hide the true scale of opioid diversion from the DEA, “prioritiz[ing] profits over compliance.”48 The changes were all reviewed and approved by the Board of Directors. At various points between 2010 and 2020, Mr. Zimmerman’s team engaged a series of brand name law firms to review and bless AmerisourceBergen’s anti-diversion programs. In 2010, shortly before the implementation of the independent pharmacy strategy, the company engaged Davis Polk to review its compliance program and found that it was “functioning effectively.”49 In 2015, in 44 Id. at *5

45 Drug Executive Emails Mocked Appalachians as “Pillbillies”, CBS NEWS, (May 14,

2021) https://www.cbsnews.com/news/drug-executive-emails-mockedappalachians-as-pillbillies/. 46 Collis Oral Argument Transcript at 12:19-13:14. 47 Collis I at *7-8 48 Id. at *7. 49 Appellees’ Answering Brief at 9, Lebanon Cnty. Employees' Ret. Fund v. Collis, No 22, 2023 (Del. Sup. 2023).

response to a shareholder litigation demand concerning opioid diversion, AmerisourceBergen engaged the law firm Fried Frank to conduct an investigation which found there was no need for the board to pursue the shareholder’s claims.50 And in 2017, Reed Smith’s healthcare regulatory practice group concluded that AmerisourceBergen had an effective compliance program, “while also recommending improvements to make the program even more robust.”51 Despite the outside law firms’ stamp of approval, opioids continued to flood communities where addiction flourished.52 Many lawsuits followed.53 In 2022, AmerisourceBergen committed to pay more than $6 billion to as part of a nationwide settlement to resolve the multidistrict litigation in which it was a defendant.54 The company paid over $1 billion more in litigation costs.55 When the dust cleared, AmerisourceBergen’s board faced scrutiny from its shareholders for its oversight over its anti-diversion program. As justification for approving Mr. Zimmerman’s evidently inadequate compliance efforts, counsel for the company and its board members pointed to the involvement of outside counsel. When arguing to dismiss the lawsuit, AmerisourceBergen’s counsel’s stressed that the board’s decisions about the company’s anti-diversion programs relied upon “the independent findings of a highly respected law firm.”56 As litigation counsel saw it, having 50 Id. at 37.

51 Collis I at *9.

According to one report, the “big three” distributors—AmerisourceBergen, McKesson, and Cardinal Health—collectively sent more than 81 million pills to a community of 91,000 people. See Brian Mann, Was It 'Reasonable' To Ship 81 Million Opioid Pills To This Small West Virginia City?, NPR (Jul. 30, 2021) https://www.npr.org/2021/07/30/1021676306/was-it-reasonable-to-ship-81million-opioid-pills-to-this-small-west-virginia-ci. 53 See, generally, National Prescription Opiate Litigation, MDL 2804, United States District Court, Northern District of Ohio, https://www.ohnd.uscourts.gov/mdl2804. 54 Press Release: Distributors Approve Opioid Settlement Agreement, AmerisourceBergen, (Feb. 25, 2022) https://investor.amerisourcebergen.com/news/newsdetails/2022/Distributors-Approve-Opioid-Settlement-Agreement/default.aspx. 55 Collis I at *1. 56 Collis Oral Argument Transcript at 25:11-13. 52

respected outside law firms involved meant the board could approve the anti-diversion program at each stage.57 The shareholders did not sue the outside law firms, and both they and the company would invoke the attorney-client privilege to protect their work. In an apparent rebranding effort, AmerisourceBergen is now called “Cencora.” Chris Zimmerman remains employed in a highlevel compliance role.58 c. The Pharmacies “If Walmart's assertions of privilege are to be believed, then as the opioid epidemic raged, Walmart's senior compliance employees did nothing except receive and consider legal advice. They knew about the problem and took no action whatsoever. Although that seems highly unlikely, that is the record that Walmart created through its highly redacted Section 220 production.”59 Walmart operates thousands of pharmacies across the United States.60 Between 2006 and 2012, those pharmacies distributed over five billion opioid pills.61 This Essay’s story about Walmart’s pharmacies is the shortest—this is by corporate lawyers’ design. Walmart has faced significant liability for its role in the opioid epidemic. After a trial in the nationwide opioid multidistrict litigation, a jury found Walmart had “engaged in intentional and/or illegal conduct which was a substantial factor in” creating an “oversupply of legal prescription opioids, and diversion of those opioids into the illicit market.”62 In 2020, a group of shareholders

57 See id. at 24:4-8.

See Chris Zimmerman, Senior Vice President, Global Quality, Risk, and Resilience at Cencora (formerly AmerisourceBergen), LINKEDIN (last visited May 13, 2024) https://www.linkedin.com/in/chriszimmerman/. 59 Ontario Provincial Council of Carpenters' Pension Tr. Fund v. Walton, No. 20210827-JTL, 2023 WL 3093500, at *17 (Del. Ch. Apr. 26, 2023). 60 Id. at *1. 61 Id. 62 In re Nat'l Prescription Opiate Litig., 622 F. Supp. 3d 584, 593 (N.D. Ohio 2022); see also Myriam Gilles & Gary Friedman, Mdl Drano: Rule 23-Based Solutions to 58

brought an action under a Delaware law that allows shareholders to seek access to a corporation’s books and records in order to bring claims against Walmart’s officers and directors for opioid oversight failures.63 In responding to shareholders’ books and records demands, Walmart’s lawyers made it nearly impossible to know quite what Walmart’s board had known and when. In what would ordinarily be called “zealous advocacy,” Walmart’s lawyers engaged in aggressive redaction practices.64 They made broad assertions of attorney client privilege and attorney work product.65 For example, Walmart’s counsel withheld as privileged all meeting minutes from twenty meetings of the Employee Compliance Committee between 2016 and 2020.66 Indeed, Walmart’s sweeping privilege withholding decisions were so extensive that, on a motion to dismiss the shareholder fiduciary duty suit, Vice Chancellor Laster of the Delaware Court of Chancery drew a negative inference from the lack of non-privileged documents.67 If Walmart’s lawyers are successful, they will end up shielding the much of the company’s compliance decision-making from public view and scrutiny, making it difficult to draw lessons from the company’s failures.68 That litigation is ongoing. Mass Tort Buildup, LAW & CONTEMP. PROBS., 2021, at 121, 132 (describing the mechanics of the nationwide opiate MDL). 63 See Verified Complaint Pursuant to 8 Del. C. 220 to Compel Inspection of Books and Records, Police & Fire Ret. Sys. of City of Detroit v. Walmart Inc., C.A. No. 2020-0478-JTL, Dkt. 1 (Del. Ch. June 17, 2020). 64 Ontario Provincial Council of Carpenters' Pension Tr. Fund v. Walton, No. 20210827-JTL, 2023 WL 3093500, at *3 (Del. Ch. Apr. 26, 2023) (“Walmart laid the foundation for the plaintiffs to seek damaging inferences by redacting documents extensively. In many cases, only a few words survive.”). 65 This kind of blanket assertion of barely supportable privilege claims is commonplace in corporate litigation contexts. See Elise Bernlohr Maizel, The Case for Downsizing the Corporate Attorney-Client Privilege, 75 UC L. J. 373, 38992 (2024); Laster & Maizel, Discovery as a Compliance Problem, supra note [x]. 66 Id. at *19. 67 Walton, 2023 WL 3093500, at *3; see also Ann Lipton, VC Laster Goes to War, BUSINESS LAW PROF BLOG, (Apr. 28, 2023) https://lawprofessors.typepad.com/business_law/2023/04/vc-laster-goes-towar.html. 68 For the accountability and public policy reasons why these kinds of disclosures are important, see Elizabeth Chamblee Burch & Alexandra D. Lahav, Information

II.

Corporate Lawyers and Legal Obedience a. Dominant Narratives About Lawyers for Lawbreaking Corporations

In the late 1990s, the Cigarette Papers69 revealed that corporate lawyers for tobacco companies had worked as “cheerful abettor[s] of [their] clients' misdeeds.”70 The response from scholars of the legal profession was one of shock and shame. They called the behavior of tobacco industry lawyers a departure from the “wise counselor” role that corporate lawyers had previously occupied.71 This mythologized corporate lawyer of a bygone era is described in the scholarly literature a “statesman,”72 a “friend”73 to the client, or a “citizen lawyer.”74 This lawyer was independent in his75 judgement for the Common Good in Mass Torts, 70 DEPAUL L. REV. 345 (2022); see also Abbe R. Gluck, Elizabeth Chamblee Burch, Adam S. Zimmerman, Against Bankruptcy?: Public Litigation Values Versus the Endless Quest for Global Peace in Mass Litigation, 131 YALE L. J. F. 27 (forthcoming 2024) (describing the litigation value of information production). 69 STANTON A. GLANTZ ET AL., THE CIGARETTE PAPERS (1996). 70 Robert W. Gordon, A Collective Failure of Nerve: The Bar's Response to Kaye Scholer, 23 LAW & SOC. INQUIRY 315, 316 (1998). 71 See Bruce A. Green, Thoughts About Corporate Lawyers After Reading The Cigarette Papers: Has the "Wise Counselor" Given Way to the "Hired Gun"?, 51 DEPAUL L. REV. 407, 433 (2001) (“However one understands the normative shift from “wise counselor” to “hired gun,” one's discomfort with the tobacco lawyers' conduct suggests the need to consider whether the change is desirable.”); Gordon, Failure of Nerve, supra note 70; see also David B. Wilkins, Team of Rivals? Toward A New Model of the Corporate Attorney-Client Relationship, 78 FORDHAM L. REV. 2067, 2068 (2010) (describing modern corporate practice as a departure from the “wise counselor” norm of past law firm lawyers). 72 See Wilkins, supra note 71 at 2075; Yablon, Lawyer as Accomplice, supra note 14 at 320. 73 Charles Fried, The Lawyer As Friend: The Moral Foundations of the LawyerClient Relation, 85 YALE L.J. 1060 (1976). 74 Robert W. Gordon, The Citizen Lawyer-A Brief Informal History of A Myth with Some Basis in Reality, 50 WM. & MARY L. REV. 1169, 1188 (2009) (“The view of lawyers as economically self-interested actors and agents of economically selfinterested clients cannot adequately explain the rich traditions of American lawyers' civic activism-their involvements in state and legal institution building.”). 75 I use “his” as a matter of historical accuracy. The romanticized golden era of the corporate lawyer was less golden for women and people of color who were

and used his considerable influence to guide corporate decisionmaking toward outcomes that protected the long term interests of the corporation and the public.76 But, according to these tellings, the past ethic of the corporate lawyer got lost somewhere along the way. Scholarly accounts place blame for the decline of the lawyerstatesman at the foot of “zealous advocacy”—a lawyer must be singularly devoted to the interests of the client, regardless of what that means for anyone else.77 The greatest defender of unabashed zeal was Monroe Freedman, whose scholarship on the subject helped to develop the modern concept of “client-centered” lawyering.78 That model applied most persuasively to the representation of individuals, especially those facing criminal prosecution.79 Far less attention was paid to organizational representations.80 And despite the many differences between individual and organizational representation, a Freedman-like concept of professional ethics came to dominate lawyers’ popular mostly excluded from the profession. These lawyers to whom so much deference was afforded were largely white men. 76 See Wilkins, supra note 71 at 2075 (“[T]hese early lawyers aspired to be wise counselors, or “lawyer-statesmen” . . . who played a key role in shaping their clients' goals and in mediating between these private ends and the public purposes of the legal framework.”); Green, Wise Counselors, supra note 71 at 410 (“At moments in the last century when the legal profession was perceived to be on a moral decline, bar leaders looked to corporate counsel as professional exemplars. They were the profession's gray eminences.”). 77 See supra note 15. 78 See, generally, e.g., MONROE FREEDMAN AND ABBE SMITH, ZEALOUS REPRESENTATION: THE PERVASIVE ETHIC (2010); see also Abbe Smith, The Difference in Criminal Defense and the Difference It Makes, 11 WASH. U. J.L. & POL'Y 83, 87 (2003) (suggesting that Freedman may have coined the term “client-centered” and certainly established its ethical underpinning). 79 See, e.g., Monroe H. Freedman, Professional Responsibility of the Criminal Defense Lawyer: The Three Hardest Questions, 64 MICH. L. REV. 1469 (1966); Monroe H. Freedman, Ethics, Truth, and Justice in Criminal Litigation, 68 FORDHAM L. REV. 1372 (2000); Monroe H. Freedman, Getting Honest About Client Perjury, 21 GEO. J. LEGAL ETHICS 133 (2008); see also Fred C. Zacharias, Reconciling Professionalism and Client Interests, 36 WM. & MARY L. REV. 1303, 1319 (1995) (describing the work of Monroe Freedman as “upholding the dignity of individuals.”). 80 But see Monroe H. Freedman, SEC Repression of Effective Advocacy, 1 LITIGATION 3 (1975); Monroe H. Freedman, Kaye Scholer - Overzealous or Overblown?, 35 S. TEX. L. REV 577 (1994).

understanding of their role, regardless of whom (or what) they represented. According to William Simon, the “dominant view” held by most practicing attorneys is that “the only ethical duty distinctive to the lawyer’s role is loyalty to the client. Legal ethics impose no responsibility to third parties or the public different from that of the minimal obligation of legal compliance that is required of everyone.”81 The ethical backstop to a lawyer’s loyalty to client is typically assumed to be the lawyer’s obligations as an “officer of the court.”82 But the lion’s share of corporate lawyers never appear before a court. Even for those who do, the risk of attorney discipline for a lack of candor to the tribunal is small enough as to not be salient for most practicing attorneys.83 A few short years after the publication of the Cigarette Papers, inside and outside counsel representing Enron facilitated a historic fraud that led to the implosion of $70 billion worth of market capitalization.84 “Zealous advocacy” was the once again the explanation offered for the behavior of Enron’s lawyers. It was also the reason why those lawyers faced limited accountability as compared to the other professionals like auditors and accountants, many of whom faced criminal charges.85 The contrast was striking

WILLIAM H. SIMON, THE PRACTICE OF JUSTICE 7 (1998); see also Wendel, Civil Obedience, supra note 15 at 368-69 (unpacking the “dominant view”). 82 Robert W. Gordon, The Citizen Lawyer-A Brief Informal History of A Myth with Some Basis in Reality, 50 WM. & MARY L. REV. 1169, 1173 (2009) (“When involved in litigation, the citizen lawyer regards herself as an “officer of the court,” that is, a trustee for the integrity and fair operation of the basic procedures of the adversary system, the rules of the game, and their underlying purposes.”); see also MODEL RULES OF PROF’L CONDUCT R. 3.3; STEPHEN LANDSMAN, THE ADVERSARY SYSTEM: A DESCRIPTION AND DEFENSE (1984) (tracing the built-in tensions between the roles of “zealous advocate” and “officer of the court” back to the 1700s). 83 Veronica Root Martinez & Caitlin-Jean Juricic, Toward More Robust SelfRegulation Within the Legal Profession, 69 Wash. U. J.L. & Pol'y 241, 269 (2022) (noting how infrequently attorneys who violate the Rules of Professional Conduct actually face discipline for misconduct). 84 Deborah L. Rhode & Paul D. Paton, Lawyers, Ethics, and Enron, 8 STAN. J.L., BUS. & FIN. 10 (2002). 85 Julie Hilden, Scummery Judgment, SLATE (June 21, 2002, 10:45 AM), https://slate.com/news-and-politics/2002/06/how-enron-s-lawyers-wriggledoff-the-hook.html; John R. Wilke, Jonathan Weil & Alexei Barrionuevo, Andersen Ex-Party Pleads Guilty, In a Significant Blow to the Firm, WALL STREET JOURNAL (Apr. 10, 2002) (https://www.wsj.com/articles/SB1018365232603680560). 81

and voices from within and without the legal profession called for reform. In 2003, the American Bar Association responded by amending the Model Rules of Professional Conduct to create its first ethical rule specific to the corporate client—Rule 1.13. That Rule requires attorneys to report legal violations that risk causing “substantial injury to the organization” to higher authorities in the organization.86 The “up the ladder” reporting requirement assumes that if the most senior management or the board of directors knew about illegality within the corporation, then they would do something about it. But, as Carliss Chatman points out, “the best way to adhere to traditional confidentiality and privilege norms while complying with [Rule 1.13] is to avoid information—a practice that is both detrimental to the attorney-client relationship and the prevention of corporate fraud.”87 In other words, Rule 1.13 created an incentive for corporate lawyers to avoid learning anything that would require reporting. This was the state of the practice of law when Purdue began pumping out OxyContin and AmerisourceBergen began distributing more pills than there were people to parts of Appalachia.88 This is why, when we learn about acts of corporate misconduct aided by corporate lawyers, we call it an excess of loyalty or too much zeal. But what zealous advocacy means to a corporate client is undertheorized, leaving a gap in the literature as to precisely what loyalty to a corporate client requires. b. The Ultra Vires Doctrine and Corporate Legal Obedience

86 MODEL RULES OF PROF’L CONDUCT R. 1.13(b); see also Carliss N. Chatman, Myth

of the Attorney Whistleblower, 72 SMU L. REV. 669, 673 (2019) (describing the passage of the Rule 1.13). 87 Chatman, supra note 86 at 723. 88 See Brian Mann, Was It 'Reasonable' To Ship 81 Million Opioid Pills To This Small West Virginia City?, NPR (Jul. 30, 2021) https://www.npr.org/2021/07/30/1021676306/was-it-reasonable-to-ship-81million-opioid-pills-to-this-small-west-virginia-ci.

The corporation is an “artificial creature of the law”89 that comes into being through its charter with a state.90 State incorporation statutes require, at a minimum, legal obedience. For example, in Delaware, where a majority of Fortune 500 companies are incorporated, the state incorporation statute permits a corporation to organize under the laws of the state “to conduct or promote any lawful business or purposes.”91 Lawfulness is a foundational requirement. In the nineteenth and early twentieth centuries, corporate charters commonly enumerated the specific powers a corporation could exercise.92 Under the traditional ultra vires doctrine, a corporation could not exercise any authority beyond those specific powers.93 That doctrine was understood to protect the state against the excessive aggregation of corporate power and to protect investors against the misuse of their capital.94 Today, corporate charters are no longer required to specify a particular purpose, and the majority of corporations use broad boilerplate language permitting them to engage in business for “any lawful purpose.”95 However, as scholars such as Elizabeth Pollman and Kent Greenfield have argued, the ultra vires doctrine continues to mandate that firms abide by the law.96

89 Upjohn Co. v. United States, 449 U.S. 383, 389-90 (1981).

Trustees of Dartmouth College v. Woodward, 17 U.S. 518, 636 (1819) (“A corporation is an artificial being, invisible, intangible, and existing only in contemplation of law. Being the mere creature of law, it possesses only those properties which the charter of its creation confers upon it either expressly or as incidental to its very existence.”). 91 Del. Gen. Corp. Law. § 101(b) (emphasis added). 92 Greenfield, Ultra Vires Lives!, supra note 17 at 1302-04. 93 Id.; see also Elizabeth Pollman, The History and Revival of the Corporate Purpose Clause, 99 TEX. L. REV. 1423 (2021). 94 Greenfield, Ultra Vires Lives!, supra note 17 at 1304-06. 95 Pollman, supra note 93 at 1448. 96 See id.; see also Greenfield, Ultra Vires Lives!, supra note 17 at 1314; Chris Brummer & Leo E. Strine, Jr., Duty and Diversity, 75 VAND. L. REV. 1, 71 (2022) (“This affirmative obligation to honor society's laws is the foundation that permits the principled use of the enabling form of current American general corporation statutes . . . the most important example of an enabling statute, the law is not just enabling, but, at the same time, prescriptive.”) 90

This requirement is reflected in the way that corporate law looks at good faith, law abidingness, and loyalty. According to Leo Strine, Jr., former Chief Justice of the Delaware Supreme Court and Chancellor of the Delaware Court of Chancery has explained, a corporate fiduciary’s “duty of ‘legal fidelity’ . . . is already a subsidiary element of the fundamental duty of loyalty.”97 The business judgment rule—which wide latitude and protection to corporate managers to direct the affairs of the company as they see fit—does not apply when those managers engage in illegal conduct. In other words, Delaware corporate law holds that violations of positive law to violate the fiduciary duty of loyalty.98 So too, officers and directors of corporations have oversight duties under Delaware law that require them to monitor the firm for compliance with legal obligations. A breach of these so-called Caremark duties constitutes a breach of the duty of loyalty.99 Simply put, to be loyal to a corporate principal, for whom illegal conduct is ultra vires, an agent may only allow the principal to engage in lawful conduct. c. The Ultra Vires Doctrine as a Corporate Lawyer’s Constraint Why then, when it is disloyal to cause a corporation to behave illegally, do we refer to a corporate lawyer’s handiwork in service of

Guttman v. Huang, 823 A.2d 492, 506 n.34 (Del. Ch. 2003); see also Chris Brummer & Leo E. Strine, Jr., Duty and Diversity, 75 VAND. L. REV. 1, 71 (2022) (“This affirmative obligation to honor society's laws is the foundation that permits the principled use of the enabling form of current American general corporation statutes . . . the most important example of an enabling statute, the law is not just enabling, but, at the same time, prescriptive.”). 98 See Andrew Gold, The New Concept of Loyalty in Corporate Law, 43 U.C. DAVIS L. REV. 457, 475 (2009) (describing the relationship between the concepts of “good faith” and “loyalty” under Delaware corporate law). 99 See In re McDonald's Corp. S'holder Derivative Litig., 289 A.3d 343, 370-75 (Del. Ch. 2023) (explaining the relationship between the oversight duties, the duty of loyalty, and good faith under Delaware law); see also In re Massey Energy Co., 2011 WL 2176479, *20 (Del. Ch. May 31, 2011) (“[A] fiduciary of a Delaware corporation cannot be loyal to a Delaware corporation by knowingly causing it to seek profit by violating the law.”). 97

corporate misconduct as “zealous advocacy?”100 In fact, what these lawyers have done—from tobacco lawyers to Enron lawyers to opioid lawyers—is betrayal. Corporations owe their existence to an agreement with the state—a corporate charter—that promises legal obedience. This last vestige of the ultra vires doctrine is the remaining duty to the public to not violate the law. Lawyers for a corporation owe their loyalty to the corporate client. The corporate client cannot violate the law. Thus, when a corporation breaches its duty to the public by violating the law, its corporate lawyers have been disloyal. Of course, explicit violations of the law are the easy cases, but what about closer calls? Many legal commands are unclear.101 (Hence the need for many lawyers.) Simply calling a lawyer who aids in corporate illegality “disloyal” does not provide an easy answer or a bright line when legal rules are unclear. But, this reframing does offer some insight as to how a corporate lawyer might think about her obligations. Instead of seeing legal obligations as external and running counter to the corporation’s interest—an inherently adversarial proposition—a corporate lawyer should think of the client’s promise of legal obedience as operating as an internal counterweight to any proposed corporate action. Because the ultra vires doctrine “reflects a public-regarding limit on corporate activity,” 102 when corporate lawyers consider aiding their client in doing harm to the public through illegal or quasi-legal means, they should worry that this action betrays the corporate client’s fundamental interest in meeting its obligation to the state. This suggestion differs conceptually from past calls for corporate lawyers to take on a greater degree of professional See, e.g., ZEALOUS ADVOCACY, ABUSE OF PROCESS OFTEN ONE IN THE SAME, SAYS EX-CHANCELLOR, 6 No. 13 Andrews Sec. Litig. & Reg. Rep. 7 (2001) (recounting remarks by William T. Allen, ex-Chancellor of the Delaware Court of Chancery, characterizing “zealous advocacy” as a euphemism for abuse of process). 101 There is, of course, a thorough and voluminous literature on the subject of legal grey areas and what they require of lawyers. See, e.g., W. BRADLEY WENDEL, LAWYERS AND FIDELITY TO LAW (2011); Stephen L. Pepper, Counseling at the Limits of the Law: An Exercise in the Jurisprudence and Ethics of Lawyering, 104 YALE L.J. 1545 (1995). 102 Pollman, supra note 93 at 1452. 100

independence, or to work as a “team of rivals” with the firm.103 Those calls see legal compliance weighing on the corporate lawyer as a function of her role as “officer of the court” that must be balanced against a lawyer’s loyalty to the client.104 I propose something different. Because of the fundamental corporate interest in legal compliance, a duty of loyalty to a corporate client counsels in favor of corporate compliance with law. Zealous advocacy for a corporate client means advocating for its interests in legal compliance. When corporate agents try to lead the corporation astray, a zealous advocate for a corporate client has a duty to object. Loyalty requires no less.105 III.

A Disloyalty Reframe Corporate Lawyers

and

its

Implications

for

Reconceiving of corporate lawyering in service of corporate misconduct as disloyalty to the client (what I call the “disloyalty reframe”) carries a number of implications and points to potential interventions. This conceptual reframing suggests a pedagogical intervention, a doctrinal revision, and an expressive reorientation. a. Pedagogical Interventions In law firm practice, it is commonplace to refer to the agent of a corporate client—the general counsel, the CEO, the head of a particular division—as “the client.” This is useful shorthand. But 103 See Robert W. Gordon, A New Role for Lawyers?: The Corporate Counselor After

Enron, 35 CONN. L. REV. 1185 (2003) (suggesting, in the wake of Enron, the creation of a “separate professional role for a distinct type of lawyer, the Independent Counselor.”); William T. Allen, Corporate Governance and a Business Lawyer’s Duty of Independence, 38 SUFFOLK U. L. Rev. 1, 11-15 (2004) (urging “professional independence” to counteract the forces of zealous advocacy on behalf of corporate clients); JOHN C. COFFEE, JR., GATEKEEPERS: THE PROFESSIONS AND CORPORATE GOVERNANCE (2006); Wilkins, supra note 71 at 2072-73. 104 Wilkins, supra note 71 at 2072-73. 105 Bradley Wendel has described this duty as one to “protect [the corporation] from the liability and reputational damage” inherent in ethics and compliance failures. W. Bradley Wendel, The Problem of the Faithless Principal: Fiduciary Theory and the Capacities of Clients, 124 PENN ST. L. REV. 107, 132 (2019) (“The lawyers' greater responsibility was primarily to the client, but derivatively to those with whom the client stood in a relationship of accountability.”).

it reveals a trap. Because corporate clients cannot speak for themselves, and can only act through agents, the distinction between the corporate client and the client’s management can collapse. Corporate lawyers form relationships with individual corporate agents, while the client itself is a legal fiction. And individual agents decide which lawyers to hire and fire in a hotly competitive market for high dollar value legal representations. Under these conditions, it is tempting to conflate the client’s representatives and the client itself.106 For this reason, professional responsibility coursework must include sufficient doses of agency law and corporate law so that students understand who their actual client is. And, professional responsibility professors should be explicit about the inherent tensions of organizational representation when teaching Model Rule of Professional Conduct 1.13. Students should be primed to understand that a career as a corporate lawyer requires navigating these tensions. They should be taught that zealous advocacy for a corporate client is, sometimes, to advocate for things not demanded by corporate managers but instead a fight for a public-regarding interest in legality emanating from the corporate charter itself. b. Doctrinal Implications There are also doctrinal implications to a disloyalty reframe. The equitable doctrine of in pari delicto (meaning “in equal fault”) bars a plaintiff from recovering in tort when the plaintiff’s wrongdoing was equal to or greater than the defendant’s.107 In pari delicto is one of the chief defenses against malpractice claims by a corporate client against outside lawyers who help that client to violate the law.108 The theory goes, “no court should be required to serve as paymaster of the wages of crime, or referee between 106 This kind of conflation of the interests of a corporate agent and the corporation

itself often shows itself in the context of privilege disputes. See, e.g., U.S. v. Graf, 610 F.3d 1148 (2010). 107 See Kevin H. Michels, The Corporate Attorney as “Internal” Gatekeeper and the In Pari Delicto Defense: A Proposed New Standard, 4 ST. MARY’S J. LEGAL MALPRACTICE & ETHICS 318, 321 (2014). 108 See, generally, Asaf Eckstein & Roy Shapira, Compliance Gatekeepers, 41 YALE J. REG. (forthcoming 2024).

thieves.”109 Under traditional agency principles, the wrongdoing of a corporate agent is imputed to the corporation itself.110 Thus, an accomplice of a corporate wrongdoer—a corporate lawyer who acted as co-conspirator—may not be sued by the wronged client itself. In pari delicto’s very premise, that judicial resources should not be wasted settling grievances among equally guilty parties, is in tension with a theory of attorney betrayal of the firm, and in tension with traditional corporate law accountability for insiders who act disloyally. Corporate law allows stockholder plaintiffs, bankruptcy trustees, receivers, and other corporate representatives to bring derivative claims against insiders, including in-house counsel, in the wake of corporate misconduct.111 In pari delicto, however, erects a barrier to liability when shareholders or corporate successors112 seek to sue outside counsel for roles in facilitating corporate misconduct.113 When a corporate lawyer’s work in service of corporate misconduct is viewed through the lens of zealous advocacy, the in pari delicto doctrine makes sense: The corporation and lawyer are working together in service of unlawful aims. Reframing the counseling of law-breaking as the betrayal of the firm breaks the beam that supports the in pari delicto doctrine. The corporation and the disloyal agent are not co-conspirators. The corporation is the innocent victim. Under this framing, a group of disloyal agents have

109 Stone v. Freeman, 298 N.Y. 268, 271 (1948).

Courts quibble, somewhat, about whether what is imputed from corporate agent to corporation is knowledge or conduct. See Paula Schaefer, In Pari Delicto Deconstructed: Dismantling the Doctrine That Protects the Business Entity's Lawyer from Malpractice Liability, 90 ST. JOHN'S L. REV. 1003, 1014 (2016). 111 See, e.g., In re Boeing Co. Derivative Litig., No. CV 2019-0907-MTZ, 2021 WL 4059934, at *3 (Del. Ch. Sept. 7, 2021) (shareholder derivative claims against J. Michael Luttig, a former federal judge who served as Boeing’s general counsel and oversaw the company’s compliance function from 2006 until several hundred people had been killed on Boeing airplanes in late 2018 and early 2019). 112 Such as a trustee in bankruptcy. 113 See, e.g., Brickley for CryptoMetrics, Inc. Creditors' Tr. v. ScanTech Identification Beams Sys., LLC, 566 B.R. 815, 839 (W.D. Tex. 2017) (denying trustee claims against outside lawyers who aided in fraud under in pari delicto); see also Kirschner v. KPMG LLP, 626 F.3d 673, 674 (2d Cir. 2010) (affirming dismissal of bankruptcy trustee claims against the law firm Mayer Brown under in pari delicto). 110

joined together to betray the firm. This doctrinal misfit counsels in favor of an exception. In pari delicto “is subject to exception when another public policy is perceived to trump the policy basis for the doctrine itself.”114 Scholars have called for this exception to more widely apply to outside law firms who help corporations violate the law.115 A disloyalty reframe provides theoretical support for a public policy exception. This same line of thinking could support a number of other doctrinal expansions predicated on good faith, judicial discretion, and the public interest.116 For example, both the Garner doctrine and crime-fraud exceptions to the attorney-client privilege allow for waiver of a corporation’s privilege where an attorney has participated in corporate illegal acts.117 However, these exceptions are incredibly narrow.118 A disloyalty reframe suggests there are significant public policy reasons to expand them in the corporate context. c. Expressive Function Finally, and perhaps most importantly, there is expressive virtue to not valorizing attorney conduct that causes public harms. 114 In re Am. Int'l Grp., Inc., Consol. Derivative Litig., 976 A.2d 872, 888 (Del. Ch.

2009). 115 See Eckstein & Shapira, supra note 108 at 41-42; Schaefer, supra note 110 at 1061-62; Brian A. Blum, Equity's Leaded Feet in A Contest of Scoundrels: The Assertion of the in Pari Delicto Defense Against A Lawbreaking Plaintiff and Innocent Successors, 44 HOFSTRA L. REV. 781, 834-37 (2016); Kevin H. Michels, The Corporate Attorney As "Internal" Gatekeeper and the in Pari Delicto Defense: A Proposed New Standard, 4 ST. MARY'S J. LEGAL MAL. & ETHICS 318, 371 (2014). 116 The form and space limitations of this Essay do not allow for a thorough discussion of all the ways a disloyalty reframe could counsel in favor of doctrinal shifts. One possibility, far too complex to unpack here, would be an expansion of Caremark liability under Delaware law for aiding and abetting breaches of fiduciary duty, given the significant legal oversight function played by outside law firms, as illustrated by the AmerisourceBergen story. 117 Maizel, supra note 65 at 385. 118 See, e.g., Wal-Mart Stores, Inc. v. Indiana Elec. Workers Pension Tr. Fund IBEW, 95 A.3d 1264, 1278 (Del. 2014) (describing the Garner doctrine as “narrow, exacting, and intended to be very difficult to satisfy”).

“Zealous advocate” is a badge of honor. “Disloyal lawyer” is a mark of shame. The stories we tell ourselves have power and demarcate heroes from villains. Delaware corporate law succeeds largely through this expressive function. Nearly twenty years ago, Ed Rock wrote that “Delaware opinions can be understood as providing a set of parables—instructive tales—of good managers and bad managers, of good lawyers and bad lawyers, that, in combination, fill out the normative job description of these critical players.”119 So too, the cultural narratives about what makes a good lawyer instruct each new attorney as to what behavior to emulate. Calling lawyers who engage in misconduct on behalf of corporate clients “disloyal” is both descriptive and normative. As a matter of doctrine, the ultra vires doctrine restricts the scope of conduct in which a loyal lawyer for a corporation can engage. To operate outside that range of acceptable behavior is to betray that obligation. But, of course, the term “disloyal” is laden with normative meaning. It tells a far less flattering story about who these lawyers are and shifts the cultural narrative.120 Lawyers want to be seen as good people.121 Under the dominant view of professional norms, the ideal of “zealous advocacy”122 allows corporate lawyers to rationalize, even justify conduct that is harmful to the public— even conduct that laps at the shores of illegality. And when this conduct is wrapped in a narrative that valorizes corporate lawyers as “serious people” worthy Edward B. Rock, Saints and Sinners: How Does Delaware Corporate Law Work?, 44 UCLA L. REV. 1009, 1016 (1997). 120 As Naomi Mezey explains, “law shapes individual and group identity, social practices as well as the meaning of cultural symbols, but all of these things . . . also shape law by changing what is socially desirable, politically feasible, legally legitimate.” Naomi Mezey, Law As Culture, 13 YALE J.L. & HUMAN. 35, 46 (2001). 121 Indeed, legal scholars have mined the field of behavioral ethics to describe the way that people engage in misconduct and retain a positive self-image. See, e.g., Yuval Feldman, Adi Libson, Gideon Parchomovsky, Corporate Law for Good People, 115 NW. U. L. REV. 1125, 1141 (2021); Jennifer Arlen & Lewis A. Kornhauser, Battle For Our Souls: A Psychological Justification for Corporate and Individual Liability for Organizational Misconduct, 2023 U. ILL. L. REV. 673, 696 (2023). 122 See W. Bradley Wendel, Lawyer Shaming, 2022 U. ILL. L. REV. 175, 191 (2022) (describing the way “zealous advocacy” is used as a mantra to refute criticism for representing unpopular clients). 119

of deference, corporate bad actors can trade on the good name of prominent law firms to protect themselves.123 Howard Udell was certainly a villain in the story of the opioid crisis, but he was also a victim of a cultural narrative that said to be unapologetically loyal to his client, he would need to blind himself to the harm Purdue caused or to hide those harms from the world as long as he could. We don’t need to pretend it is good or necessary for corporate lawyers to look away from the harms their clients cause. A disloyalty reframe reminds the corporate lawyer that they are constrained by the same public policy forces that constrain the corporation. It suggests that corporate lawyers whose work will make an impact on the public owes something to that public. And it suggests the lawyers whose work created and perpetuated the opioid crisis might have owed more to Spencer and to Scott, whose story opened this Essay. Conclusion The story we currently tell about lawyers who help corporate clients harm others is one of zealous advocacy gone too far. But this framing often forgets a fundamental component of the corporate client’s existence—its promise of obedience to the state. We need to tell a new story, one that describes lawyers who aid and abet corporate misconduct as disloyal to that promise. While this reframing certainly does not resolve the tensions inherent in representing a corporate entity, it can change the way we think, feel, and speak about corporate lawyers duties’ and the consequences of their failures. In doing so, we can resist using heroic rhetoric to describe lawyerly acts of public harm and remind corporate lawyers of the promises their clients must make.

123 Indeed, each of the opioid defendants described in Part I was represented by

brand name law firms, whose partnership include former US attorneys, senior DOJ officials and retired federal judges.

Where Were Purdue’s Lawyers? Lawyer Gatekeeping and Large-Scale Harms W. Bradley Wendel1 2024 Cliﬀord Symposium on Tort Law and Theory: The Legacy of Industrywide Deadly Misconduct I.

Introduction: “Where Were the Lawyers?”

One of the most familiar themes in the history of the American legal profession was pithily stated by a district judge overseeing a civil action arising out of the collapse of Lincoln Savings and Loan Association in the 1980’s. “Where were these professionals?” asked Judge Stanley Sporkin, while Charles Keating was orchestrating a series of fraudulent transactions.2 He meant that the lawyers (and accountants) for Lincoln Savings should have taken more effective steps to put a stop to the funny business or, failing that, should have withdrawn from the representation. The slightly paraphrased question, “where were the lawyers?” has become iconic. It is now a commonly used shorthand for the observation that powerful actors frequently can get away with wrongdoing only if lawyers within an organization or serving as outside counsel fail to take steps to prevent the harms. Following the collapse of Enron Corporation in the early 2000’s, Susan Koniak gave an ironic answer to Judge Sporkin’s question as if it had been directed to Enron’s lawyers. In testimony before the Senate Judiciary Commiêe, Koniak said the lawyers were there all along, wearing magic caps that they believed could transport them to an alternative reality – “a law free zone, in which they are free to do anything and everything” for their clients.3 Leaning into the metaphor, Koniak said: If Enron was the Emerald City, no maêr what individual or groups of individuals end up to have been standing behind the curtain playing the Wizard of Oz (be it Enron's CEO, its CFO, some or all of its board of directors or whomever), I guarantee you one thing: A lawyer was standing beside them Edwin H. Woodruff Professor of Law, Cornell Law School. I gratefully acknowledge the research funding provided by the Judge Albert Conway Memorial Fund for Legal Research, established by the William C. and Joyce C. O’Neil Charitable Trust. Thanks to Wendy Muchman and Sari Montgomery for conversations about the process leading to the ABA’s recent adoption of amendments to Model Rule 1.16. 2 Lincoln Sav. & Loan Ass’n v. Wall, 743 F. Supp. 901, 920 (D.D.C. 1990). 3 Susan P. Koniak, Testimony, “Where Were the Lawyers? Behind the Curtain Wearing Their Magic Caps,” in Accountability Issues: Lessons Learned from Enron’s Fall — Senate Judiciary Committee Hearing (Feb. 6, 2002). 1

making sure the curtain stayed drawn and all the bells and whistles were hooked up and operating to fool Dorothy, the brainless scarecrow and the cowardly lion.4 Koniak may have over-worked the comparison a li^le. Relative to their outside counsel, companies like Enron are not the ineffectual Wizard of Oz. Owing to the increased power of inside counsel and the economic leverage possessed by large corporate clients,5 law firms are frequently in a subordinate role with respect to the client. It is certainly the case, however, that a principal wrongdoer may be aided and abeêd by the active or passive assistance of lawyers. A very recent case illustrates the role of active participation or passive acquiescence of lawyers in massive financial harms. After the cryptocurrency exchange and trading company FTX Group filed for bankruptcy, the U.S. Trustee sought and obtained the appointment of an outside examiner to investigate and determine whether, among other things, law firms representing FTX entities had assisted in any “fraud, dishonesty, incompetence, misconduct, mismanagement, or irregularity in the management of the affairs of the debtor of or by current or former management of the debtor.”6 The examiner found that one firm (identified only as Law Firm-1 in the report, but identified by media reports as Fenwick & West7), which had been employed as FTX’s primary outside general counsel, was involved in a number of instances of fraudulent conduct by FTX.8 Significantly, that law firm had been “entrusted with a bird’s-eye view”9 and could appreciate the significance of the transactions it worked on, unlike other firms who only saw bits and pieces of FTX’s overall business and may not have understood how their representation contributed to the client’s malfeasance. The premise underlying this article is simple: If we can understand that lawyers have duties to take action to prevent massive financial frauds like those that precipitated the collapse of FTX, then why shouldn’t lawyers similarly have duties to prevent large-scale injury or death caused by defective products? Having one’s life savings wiped out is a terrible thing, but developing an addiction to opioid painkillers Id. See, e.g., BEN W. HEINEMAN, JR., THE INSIDE COUNSEL REVOLUTION: RESOLVING THE PARTNER-GUARDIAN TENSION (2016). 6 11 U.S.C. § 1104(c). 7 See Justin Wise, Fenwick Complicit in FTX Fraud, Bankruptcy Examiner Says, BLOOMBERG LAW (May 24, 2024). 8 See In re FTX Trading, Ltd., Case 22-11068-JTD (Bkrtcy. D. Del.), Report of Robert J. Cleary, Examiner, Doc. 15545 (May 23, 2024), pp. 120-21. 9 Id. at 119. 4 5

or contracting ovarian cancer from asbestos-containing baby powder is also a tremendous loss. The magnitude and kind of suffering experienced by the victims of client wrongdoing cannot be what differentiates lawyer assistance in financial frauds from participation in tortious conduct resulting in personal injuries. Lawyers will resist any proposal that exposes them to potential liability for aiding and abeîng their clients’ torts. However, the landscape of lawyers’ duties to exercise some responsibility as gatekeepers to prevent client misconduct is slowly shifting. In most instances a lawyer for a product manufacturer or other client potentially subject to liability for large-scale harms will have traditionally understood duties of confidentiality and zealous advocacy. There may be some cases, however, in which gatekeeping responsibilities may extend to cases involving bodily injury, not only financial harms. The role of lawyers in the Enron transactions was the big high-profile scandal that finally broke the longstanding resistance of the American Bar Association (ABA) to permiîng lawyers to disclose confidential information of clients where disclosure would be reasonably necessary to prevent, rectify, or mitigate a client fraud for which the client had used the lawyer’s services. The ABA had successfully fought off proposals to amend the confidentiality rule in the wake of other big, high-profile scandals. Following Enron, however, the ABA approved amendments to the Model Rules of Professional Conduct, permiîng lawyers to disclose confidential client information to the extent reasonably necessary to prevent, rectify, or mitigate clients frauds, in maêrs in which their services had been used by the client.10 Lawyers now had an additional role: While continuing to provide loyal client service and competent representation, and maintaining the confidentiality of sensitive information, lawyers were also required, to a limited extent, to serve as gatekeepers. Lawyers could no longer stand by and watch while the harmful consequences of client actions, for which the services of lawyers had been used, unfolded. If lawyers became aware of a client crime or fraud that threatened substantial financial harm to others, and if their services had been used in connection with the wrongdoing, they were permiêd to disclose confidential information. This represented at least a partial resurrection of a traditional role for lawyers, as “law guardians” as well as the more modern role of “client guardians.”11 Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.6(b)(2), (b)(3). For this framing see Charles W. Wolfram, Toward a History of the Legalization of American Legal Ethics - II: The Modern Era, 15 GEO. J. LEGAL ETHICS 205, 208 (2002). I used Wolfram’s analysis in terms of the law guardian and client guardian roles to structure a historical account of the development of the field of the law governing lawyers, including the drafting of the Restatement of the Law Governing Lawyers, for a chapter in a book for the 100th anniversary of the American Law Institute. See W. Bradley Wendel, Constructing a Legal Field: The Restatement of the Law Governing Lawyers, in THE AMERICAN LAW INSTITUTE – A CENTENNIAL HISTORY 237 (Andrew S. Gold & Robert W. Gordon eds. 2023). 10 11

This article considers a very recent change to Rule 1.16(a) of the Model Rules of Professional Conduct, adopted in August 2023 by the ABA House of Delegates, requiring lawyers to inquire into the facts and circumstances of a prospective or ongoing client representation to determine whether the client “seeks to use or persists in using the lawyer’s services to commit or further a crime or fraud.”12 If the lawyer concludes, after conducting an investigation appropriate in the circumstances, that the client’s purpose is to employ the lawyer’s services for assistance in commiîng a crime or fraud, the lawyer is required to decline the representation of a prospective client or terminate an ongoing representation. There is no requirement or even permission in the new rule to disclose information to prevent, rectify, or mitigate the harm, although existing rules provide some limited avenues for disclosure.13 In other words, amended Rule 1.16(a) is a gatekeeping provision that does not implicate the related but distinct issue of whistleblowing. It therefore differs from the post-Enron rule amendments permiîng disclosure of client information as a means of enforcing the lawyer’s law guardian responsibilities. Adoption of the amendment to Rule 1.16(a) was motivated by a sense that if the ABA did not take meaningful steps to impose an obligation on lawyers to avoid assisting financial crimes such as money laundering or terrorist financing, Congress or the Treasury Department would step in and the result would be much worse. The new rule is intended to thread the needle, between the pressure from domestic and international financial regulators, on the one hand, and on the other hand the vehement opposition by lawyers to anything that looks like an obligation to “rat out” a client, serve as an adjunct to law enforcement, or betray the duty of zealous client advocacy. (The rule actually has nothing to do with advocacy; it is aimed at legal advising and transactional representation, where the client could plausibly use the lawyer’s services to commit or further a crime or fraud.14) Once again, the ABA has enhanced the law AM. B. ASS’N, MODEL RULES OF PROF’L CONDUCT, Rule 1.16(a). See, e.g., AM. B. ASS’N, MODEL RULES OF PROF’L CONDUCT, Rules 1.6(b)(1) (disclosure to prevent reasonable certain death or substantial bodily harm), 1.6(b)(2), (b)(3) (disclosure to prevent, rectify, or mitigate substantial harm to the property of others resulting from client crime or fraud for which the lawyer’s services are used), 1.13(c) (disclosure to prevent substantial injury to an organizational client caused by misconduct of a constituent of the organization). The exceptions in Rule 1.6(b) are all qualified by the language “to the extent the lawyer reasonably believes necessary.” Comment [16] to Rule 1.6 recognizes that, within a lawyer-client fiduciary relationship, characterized by duties of loyalty and reasonable communication, the lawyer should first try to talk the client into doing the right thing before disclosing confidential information. 14 As such, the scope of this article differs from others that consider aggressive defense strategies employed by lawyers in other mass tort proceedings, such as the decades-long “tobacco wars,” where defense lawyers engaged in conduct like the assertion of factually and legally unsupported privilege and 12 13

guardian role of lawyers, notwithstanding the recurring opposition of the legal profession to any role other than client guardian. The position taken by the rule amendment, or so I will argue, is that lawyers must still look after the interests of their clients, complying with demanding fiduciary duties of loyalty, competence, communication, and confidentiality, but there are limits on these client-guardian duties. Those limits are not given by the lawyer’s judgment about what the public interest requires, but by the law itself, in defining the of legal concepts of crime, fraud, and the use of the lawyer’s services.15 If the ABA has put a thumb on the law guardian side of the scale, a question naturally arises in connection with large-scale harms: Does the law guardian role apply equally to the prevention of widespread serious physical harms? Considered as a whole, at a high level of abstraction, the response of the legal system today to largescale physical harms is to muddle through the ex post resolution of an aggregation of individual claims for wrongful injury through tort litigation, employing some combination of class actions, multidistrict litigation (MDL) consolidation orders, and Chapter 11 bankruptcy reorganization proceedings.16 All of these proceedings have characteristic dysfunctions and failure modes and are pre^y clearly suboptimal from a system-design perspective.17 One might therefore ask whether it would not be a beêr thing – certainly from the point of view of potential victims of tortious wrongdoing – if the legal system had an ex ante institutional response that was geared toward preventing harm, not merely compensating it. If the legal system is understood broadly enough to include not only courts and the litigation process, along with legislatures and administrative agencies and the lawmaking process, but also the legal profession that performs an intermediating function between the law and its subjects, then we see the work product claims to prevent plaintiffs from obtaining discovery of probative evidence. See, e.g., Roger C. Cramton, Lawyer Conduct in the Tobacco Wars, 51 DEPAUL L. REV. 435 (2002); Bruce A. Green, Thoughts about Corporate Lawyers after Reading the Cigarette Papers: Has the Wise Counselor Given Way to the Hired Gun, 51 DEPAUL L. REV. 407 (2002). Green’s alludes to the “wise counselor” conception of the lawyer’s role, which is another traditional description, closely related to the law guardian role. 15 This is a position I have been defending for most of my career as a scholar. See, e.g., W. Bradley Wendel, The Limits of Positivist Legal Ethics: A Brief History, a Critique, and a Return to Foundations, 30 CAN. J.L. & JURIS. 443 (2017). For a version of my general argument that proceeds via a deep dive into fiduciary law and theory, see W. Bradley Wendel, Understanding the Complex Loyalty of Lawyers: Dual-Commission, Governance Mandate, and Intrinsic-Limit Analyses, in OXFORD STUDIES IN PRIVATE LAW THEORY: VOLUME II `59 (Paul B. Miller & John Oberdiek, eds., 2023) [hereinafter, “Wendel, Complex Loyalty”]. For an overview from the standpoint of political theory, see W. Bradley Wendel, Pluralism, Polarization, and the Common Good: The Possibility of Modus Vivendi Legal Ethics, 131 YALE L.J.F. (Oct. 24, 2021). 16 See Alexandra D. Lahav, The Continuum of Aggregation, 53 GA. L. REV. 1393 (2019 (describing and comparing these three mechanisms of large-scale dispute resolution). 17 [Cite representative sample of the zillions of articles on dysfunctional mass tort procedures.]

force of a reformulated version of Judge Sporkin’s question: Where should the lawyers be when their clients are engaging in the large-scale creation of a substantial risk of physical harm? As someone who has followed the debates on lawyer gatekeeping for over two decades, I believe the new ABA rule amendment represents a sea change in the balance between client-centered duties and obligations to protect third parties from harm. The rule does not reach all the way back to traditional conceptions of lawyers as social trustees, with duties to act in furtherance of the interests of society as a whole.18 But it does require lawyers to inquire, in appropriate circumstances, into the client’s purpose for retaining the lawyer and the risk that the client is using the lawyer’s services to commit a crime or fraud. Although the debate preceding enactment of the amendment focused on financial harms like money laundering and terrorist financing, there is nothing in the text of the rule that is limited to these categories of wrongdoing. The rule is limited to client conduct that amounts to a crime or fraud, but “fraud” is interpreted expansively in some areas of the law governing lawyers, such as the crime-fraud exception to the aôrney-client privilege, to include other types of conduct manifesting bad faith or abuse of the legal process.19 Moreover, as discussed below, nondisclosure by a public company of the risks associated with a product may constitute securities fraud, thereby implicating lawyers who advise on the public statements and SEC disclosures of the company and its officials. To make this analysis more concrete, one could ask what a law guardian would do when confronted with evidence that the client was doing something that foreseeably would result in widespread serious injury, without legal justification. Imagine, for example, that a lawyer for Purdue Pharma learned that the company had been exaggerating, or outright misrepresenting, the non-addictive quality of its flagship product, OxyContin, in promotional materials directed to physicians.20 Or suppose that a lawyer for a pharmaceutical distributor discovered that the client had been shipping opioid painkillers to a state or county in quantities greatly disproportionate to the population. Changing products, think about the situation of lawyers representing See Michael Ariens, The Rise and Fall of Social Trustee Professionalism, 2016 J. PROF. LAW. 49 (2016). See RESTATEMENT (THIRD) OF THE LAW GOVERNING LAWYERS § 82, cmt. d (2000) (noting that authorities including the Wigmore treatise on evidence and the 1942 Model Code of Evidence included advice concerning future torts within the scope of communications that would not be privileged under the crime-fraud exception). 20 See BARRY MEIER, PAIN KILLER: AN EMPIRE OF DECEIT AND THE ORIGIN OF AMERICA’S OPIOID EPIDEMIC (2018). Meier’s book was based on earlier long-form reporting he did for the New York Times. See Barry Meier, Origins of an Epidemic: Purdue Pharma Knew Its Opioids Were Widely Abused, N.Y. TIMES (May 29, 2018). 18 19

Johnson & Johnson upon learning that the company’s iconic Johnson’s Baby Powder contained asbestos fibers, mined along with the talc that was the principal ingredient in the product.21 Going back to an earlier mass tort, consider lawyers for tobacco companies who were aware that “the probabilities are that some combination of constituents of smoke will be found conducive to the onset of cancer.”22 Consistent with the lawyer’s fiduciary duties to the client, assume the lawyer in question has raised the issue with senior management within the organization who are empowered to put a stop to the harm-causing conduct but the lawyer’s advice was ignored or rebuffed. Now what? The intuition behind Judge Sporkin’s question is that in cases such as these, lawyers should do something. What they may or must do is a balance of duties in tension, as both client guardians and law guardians. Lawyers sometimes complain that these competing duties create a conflict of interest, but it is well established that lawyers may have duties that run to third parties, or limitations internal to the duties they owe to clients, and there is no conflict, only a balance or tension that must be managed.23 From a more explicitly normative point of view, lawyers also complain that they should not be subject to these duties in tension because they should be permiêd to represent their clients zealously up to the limits clearly established by law. Sometimes this position is stated even more strongly as a permission to represent their clients zealously unless they knowingly provide assistance in conduct that is a crime or fraud.24 To roughly generalize, it seems fair to say that lawyers intensely dislike law guardian duties. Nevertheless, in response to public and regulatory pressure, the law governing lawyers has incorporated duties that are intended to protect third parties or the public as a whole from harm. The new ABA amendment, if adopted by state courts,25 will See Lisa Girion, Johnson & Jonson Knew for Decades that Asbestos Lurked in Its Baby Powder, REUTERS (Dec. 14, 2018). 22 From a 1963 memorandum from the general counsel of Brown & Williamson, quoted in Richard A. Daynard, Hired Guns, Adversaries, or White-Collar Killers: Comment on Professors Green’s and Redish’s Views of Tobacco Lawyers, 51 DEPAUL L. REV. 449, 454 (2002). 23 See, e.g., Greycas, Inc. v. Proud, 826 F.2d 1560 (7th Cir. 1987). 24 See W. Bradley Wendel, Lawyers’ Constrained Fiduciary Duties: A Comment on Paul R. Tremblay, “At Your Service: Lawyer Discretion to Assist Clients in Unlawful Conduct,” 70 FLA. L. REV. F. 7 (2018) (arguing against the view that Model Rule 1.2(d), prohibiting knowing assistance in client crimes or frauds, exhausts the limitations on the assistance that lawyers may provide to clients). A better view begins with the observation that, legally speaking, lawyers are agents for their client and can exercise no greater lawful power than their client, the principal, possesses under the law applicable to the principal’s actions. See generally Deborah A. DeMott, The Lawyer as Agent, 67 FORDHAM L. REV. 301 (1998). In any case, it is absolutely critical to my understanding of legal ethics that the limits on permissible lawyer conduct are given by positive law, not the public interest or extra-legal moral considerations. 25 By itself the ABA has no regulatory authority over lawyers. It can only propose rules – hence the term “Model Rules” – for consideration by the judiciary of each state. See CHARLES W. WOLFRAM, MODERN 21

expand law guardian duties to financial crimes and frauds. The interesting question to consider here is whether the amendment provides a template for extending law guardian duties to tortious wrongdoing as well. The overall theme of this article is that it may be possible to extend existing gatekeeping or law guardian duties in a way that would obligate in-house or outside lawyers for a company like Purdue to take affirmative steps to prevent foreseeable physical harms to consumers. Given sufficiently weighty evidence of a sufficiently clear connection between an unjustified risk posed by a client’s product and serious bodily harm to those foreseeably exposed to the risk, lawyers should not be permiêd to continue providing legal services to the client in connection with that maêr. At least, lawyers should reasonably be concerned that the rule amendment will open the door to this sort of obligation. Whether that is a good thing or a bad thing depends on one’s normative priors regarding the balance of law guardian and client guardian duties. I have always advocated for serious law guardian duties as part of the role of lawyers, provided that these duties are oriented toward positive law and not considerations of the public interest or the interests of others that are not recognized as establishing legally protected interests. I suspect many lawyers will disagree and will seek to read the new version of Rule 1.16(a) as narrowly as possible. But there is a decent argument to be made that the train has left the station and lawyers need to be ready to be subject to expanded gatekeeping duties. The new amendment is constrained in three ways that will influence the impact it has on representing clients engaged in tortious wrongdoing. First, it is a gatekeeping obligation backed up by a duty to withdraw from representation. It is not a proposal for reporting out, whistleblowing, dropping a dime on the client, or disclosure of confidential information in any way. The ABA amendment is to the rule on withdrawal from representation. I do not believe it was intended to broaden the existing permission to disclose confidential information “to prevent reasonable certain death or substantial LEGAL ETHICS § 2.2 (1986). Notwithstanding frequent references to regulation by “the bar,” American lawyers are actually regulated by courts, with the organized bar playing only an advisory role. It is true that some states have so-called integrated, mandatory, or unified bars, meaning that admission to practice in the state also entails membership in the State X Bar Association. The agency of the judicial branch that investigates filed grievances and other evidence of attorney misconduct, and administers licensurerelated sanctions such as disbarment, suspension, or reprimands, is sometimes called something like the “State Bar of California,” although careful investigation will show that, no matter the name, it is an agency of the judiciary. For example, the State Bar of California’s website states that “[t]he State Bar of California is the regulatory arm of the California Supreme Court responsible for licensing and disciplining attorneys.” See “Who We Are,” https://www.calbar.ca.gov/About-Us/Who-We-Are. It is important to remember that, in formal legal terms, the ABA is nothing more than a trade association. See Richard L. Abel, Why Does the ABA Promulgate Ethical Rules?, 59 TEX. L. REV. 639, 653-67 (1981).

bodily harm.”26 Second, the new amendment requires withdrawal only where the client has used, or intends to use, the lawyer’s services in connection with the wrongful act. The amendment does not say anything about the situation of the lawyer-as-spectator, who in the course of representing a client comes to learn of wrongdoing in which the lawyer’s services are not mixed up. Third, the text of the rule is limited to misconduct defined as a crime or fraud. It remains to be seen how literally courts or disciplinary authorities read these terms. “Crime” is probably limited to actual defined statutory offenses, but “fraud” is often given a broader reading in the law of lawyering, to encompass varieties of bad-faith conduct. At the very least, I do not believe the new rule amendment forecloses the imposition of duties on the part of lawyers to act as law guardians with respect to tortious wrongdoing. Here is a brief overview of the article: Section II begins with a very brief historical account of how we got here, tracing the history of lawyer gatekeeping in the context in which it has been the most frequently discussed, namely the prevention of financial frauds. Recurring scandals have focused public aêntion on the role of lawyers in facilitating client wrongdoing, but for decades the organized bar, most prominently the ABA, had resisted any aêmpt to beef up law guardian duties. The recent Rule 1.16(a) amendments may be a watershed moment in this history. Section III will recap two episodes of corporate wrongdoing in which lawyers must have played some role. The first is alluded to in the title of the paper – the conduct of Purdue Pharma in connection with the marketing of its OxyContin opioid painkiller product. The second is the sale of Johnson & Johnson’s baby powder product, with knowledge that its principal ingredient, talc, may be mixed up with asbestos. After seîng out these scenarios, Section IV will go a li^le deeper into the doctrinal weeds, considering three questions concerning the application of Rule 1.16(a) to large-scale tortious wrongdoing: (1) whether the rule is limited to the prevention of financial crimes and frauds; (2) the surprisingly slippery distinction between lawyer participation in the client’s wrongdoing and knowledge alone, without providing substantial assistance; and (3) the mental state required to trigger gatekeeping duties such as withdrawal or disclosure. II.

The Emerging Landscape of Lawyer Gatekeeping.

Charles Wolfram, the Chief Reporter for the American Law Institute’s Restatement of the Law Governing Lawyers, traced the ascendancy of the client guardian AM. B. ASS’N, MODEL RULES OF PROF’L CONDUCT, Rule 1.6(b)(1). It may happen, of course, that a lawyer inquiring into the facts and circumstances of the representation, as required by Rule 1.16(a), might learn facts supporting a conclusion that disclosure of confidential information is reasonably necessary to prevent reasonably certain death or substantial bodily harm. This would be an indirect impact of the due diligence inquiry, however, and not something directly mandated by the amended rule. 26

ideal for lawyers to the late Nineteenth Century.27 The early history of the American Bar Association was largely characterized by efforts at professional self-regulation aimed at fending off regulation of the legal profession by Congress or state legislatures. The ABA was reacting against Progressive-era political currents such as President Theodore Roosevelt’s commencement speech at Harvard University at which he severely criticized lawyers for assisting clients in “evad[ing] the laws which are made to regulate in the interest of the public the use of great wealth.”28 Throughout most of the Twentieth Century, the ABA’s position, in the 1908 Canons of Ethics, the 1969 Model Code of Professional Responsibility, and the 1983 Model Rules of Professional Conduct, embodied a coherent normative position. As described by Geoffrey Hazard, the Reporter to the ABA Commission that prepared the 1983 Model Rules, the rules were characterized by duties of loyalty to clients partisanship in advocacy; lack of any duty to take account of the interests of third parties or the common good; and the libertarian assumption that promoting the autonomy of clients is a good thing to do.29 To the extent lawyers did serve the public interest, it was indirectly, through loyal, partisan representation of clients30 – the “zealous advocacy” that lawyers love to valorize.31 To borrow an image from Benjamin Cardozo (via William Prosser) in tort law,32 the citadel of client guardian duties came under sustained assault in the laêr decades of the Twentieth Century.33 A series of scandals focused the aêntion of journalists and considerable public ire on lawyers who did nothing while investors or bank depositors Charles W. Wolfram, Toward a History of the Legalization of American Legal Ethics - II: The Modern Era, 15 GEO. J. LEGAL ETHICS 205, 208 (2002). See also Rebecca Roiphe, The Decline of Professionalism, 29 GEO. J. LEGAL ETHICS 649 (2016) (accepting this rough chronology of the decline of the ideal that lawyers ought to serve the public interest). 28 Theodore Roosevelt, The Harvard Spirit (June 28, 1905), in IV PRESIDENTIAL ADDRESSES AND STATE PAPERS 407, 419-20 (1910). 29 Geoffrey C. Hazard Jr., The Future of Legal Ethics, 100 YALE L.J. 1239, 1242-46 (1991). 30 Geoffrey C. Hazard Jr., Rules of Legal Ethics: The Drafting Task, 36 REC. ASS'N B. CITY N.Y. 77, 93-94 (1981). 31 The motto “zealous advocacy within the bounds of law” comes from Canon 7 of the ABA Model Code. The phrase was deleted from the Model Rules, now existing only in a vestigial form in a comment to the rule prescribing diligence on behalf of clients. See AM. B. ASS’N, MODEL RULES OF PROF’L CONDUCT, Rule 1.3, cmt. [1] (“A lawyer must also act with commitment and dedication to the interests of the client and with zeal in advocacy upon the client’s behalf.”). Although the motto refers to advocacy, American lawyers generally understand it to state duties for lawyers representing clients in transactional or compliance matters as well. 32 William L. Prosser, The Assault Upon the Citadel (Strict Liability to the Consumer), 69 YALE L.J. 1099 (1960). 33 See Lawrence J. Fox, OTS vs. Kaye, Scholer: An Assault on the Citadel, 48 BUS. LAW. 1521, 1522 (1992) (using the same metaphor in the context of an enforcement action brought by a federal bank regulator against the Kaye Scholer firm, attempting to enforce some law guardian obligations, as against the proposition that “we as a nation have decided that a lawyer's primary obligation, subject to certain welldefined exceptions, is to the client”). 27

were swindled by crooked managers. The appeal of neutral partisanship and zealous advocacy turned out to be limited when lawyers had provided substantial assistance in their clients’ frauds. By the beginning of the Twenty-First Century, the ABA had responded to not only public pressure but the threat of Congressional intervention and acquiesced in changes to the Model Rules that it had long resisted. A.

A very brief recap of the history of ﬁnancial-frauds gatekeeping.

The story begins with a fraud commiêd by a company called O.P.M., which somewhat alarmingly stood for “Other People’s Money.”34 O.P.M. started out as a legitimate computer leasing business, but eventually became a Ponzi scheme, with the company’s principals using fake lease documents to secure additional loans from banks. Lawyers representing the company belatedly realized that they had drafted documents and closed several lease transactions based on fraudulent representations they passed along from their now-crooked clients. The nervous lawyers contacted two outside ethics experts who advised them, based on the rule of confidentiality then in effect, notwithstanding other provisions of the rules prohibiting providing assistance to client fraud, that they were not permiêd to disclose to investors that the transactions had closed based on fraudulent information. Significantly, the experts relied only on the state rules of professional conduct, and did not consider whether the lawyers would be liable under the common law of fraud if they did not explicitly disassociate themselves from the earlier representations made in the course of closing the leases. In the ensuing bankruptcy of O.P.M., the trustee concluded that the law firm’s failure to take affirmative steps to prevent the frauds caused an additional $85 million in damages. The law firm se^led for $10 million, the limits of its liability insurance coverage, and subsequently dissolved. As it happens, an ABA body known as the Kutak Commission was meeting at the same time to draft the Model Rules of Professional Conduct, a replacement for the Model Code of Professional Responsibility, which had come to be regarded as deeply flawed.35 The Kutak Commission recommended that lawyers be permiêd, but not This summary of the O.P.M. case is based on Susan P. Koniak, When the Hurlyburly’s Done: The Bar’s Struggle with the SEC, 103 COLUM. L. REV. 1236, 1260-64 (2003); and GEOFFREY C. HAZARD, JR., ET AL., THE LAW AND ETHICS OF LAWYERING 304-10 (3d ed. 1999), both of which cite journalistic accounts and the bankruptcy trustee’s report. 35 See Michael S. Ariens, “Playing Chicken”: An Instant History of the Battle over Exceptions to Client Confidentiality, 33 J. LEGAL PROF. 239, 257-64 (2009) [hereinafter “Ariens, Chicken”] (recounting history of confidentiality exceptions throughout history of Kutak Commission process, including impact of O.P.M. case); Ted Schneyer, Professionalism as Bar Politics, 14 LAW & SOC. INQUIRY 677 (1989) (providing a critical overview of the Kutak Commission process); Geoffrey C. Hazard, Jr., Rules of Legal Ethics: The Drafting 34

required, to disclose confidential information to prevent client fraud in situations like the O.P.M. transactions. However, the full House of Delegates put the kibosh on this proposal, adopting a strict rule of confidentiality (Rule 1.6) without an exception for disclosure to prevent client fraud.36 As a compromise it inserted a comment to the rule: If the lawyer's services will be used by the client in materially furthering a course of criminal or fraudulent conduct, the lawyer must withdraw, as stated in Rule 1.16(a)(1). After withdrawal the lawyer is required to refrain from making disclosure of the client's confidence, except as otherwise provided in Rule 1.6. Neither this rule nor Rule 1.8(b) [prohibiting disclosure of confidential information of a former client] nor Rule 1.16(d) [protecting the interests of a former client] prevents the lawyer from giving notice of the fact of withdrawal, and the lawyer may also withdraw or disaffirm any opinion, document, affirmation, or the like.37 The ABA Ethics Commiêe recognized that “[s]uch a ‘noisy’ withdrawal is, of course, likely to have the collateral consequence of disclosing, inferentially, information relating to the representation that is otherwise protected as a client confidence under Rule 1.6.”38 As I have characterized it, the comment to Rule 1.6 functions like Magriê’s caption on a picture of a pipe, “Ceci n’est pas une pipe.”39 It was a rule permiîng disclosure of confidential information to prevent, rectify, or mitigate client fraud, for which the lawyer’s services had been used by the client, only with a genteel agreement not to admit that it really was a permission to disclose. The noisy withdrawal solution was not solution at all – just a kludge employed by the ABA to avoid the problem created by the collision of the prohibition on disclosure with the real world of cases like O.P.M.40 The idea of noisy withdrawal is likely destined to be forgoên because of the post-Enron 2003 amendments to Model Rule 1.6(b), which expressly permit disclosure of confidential information to prevent, rectify, or mitigate client fraud. However, it still has some currency with lawyers overseas – my international students frequently Task, 36 REC. BAR CITY N.Y. 77 (March 1981) (recounting the process of drafting the Model Rules from the point of view of the Reporter to the Kutak Commission). 36 See Geoffrey C. Hazard Jr., Lawyers and Client Fraud: They Still Don't Get It, 6 GEO. J. LEGAL ETHICS 701, 720-26 (1993). 37 See Am. B. Ass’n, Standing Comm. on Prof’l Resp., Formal Opinion 92-366 (1992), at 7. The facts of the ABA opinion are based on the O.P.M. case. 38 See id. at 7 n.9 (“It must be recognized, therefore, that a ‘noisy’ withdrawal may result in a disclosure of ‘information relating to representation’ that is generally prohibited by Rule 1.6.”). 39 Wendel, supra note __, at 246 n.61). 40 See Hazard, Client Fraud, supra note __, at 728-36.

mention it – and is an important part of the history of lawyer gatekeeping. In addition to being amusing, the history of the noisy withdrawal approach shows the organized bar struggling to reconcile the competing demands of the client guardian and law guardian roles. Enron, however, represented a much more definitive turning point in the siege of the citadel of client guardian duties, at least where financial frauds are concerned. The Enron story is familiar in broad outlines,41 but the important takeaway for our purposes is the effect of possible external legislative and regulatory intervention on the ABA’s defense of the client guardian approach.42 In a remarkable irony, an ABA Commission, the time referred to as the Ethics 2000 Commission, happened to be debating a new set of rules of professional conduct right when a major corporate scandal broke. Like the O.P.M. case that was splashed all over the headlines during the Kutak Commission’s drafting of the Model Rules, the Enron collapse drew considerable public aêntion to the Ethics 2000 Commission’s consideration of amendments to the confidentiality rule. The Commission’s proposal would have permiêd, but not required, limited disclosure of confidential information “to prevent a client from commiîng a crime or fraud that is reasonably certain to result in substantial financial injury to another, or to rectify the consequences of a financial crime or fraud, in each case, if the lawyer’s services were utilized by the client in connection therewith.”43 However, the proposed rule went down in flames in the House of Delegates. The ABA’s rulemaking body decisively rejected the proposal on the ground that even permiîng lawyers to disclose confidential information – let alone making disclosure mandatory – would fatally undermine the relationship of trust and confidence between clients and their counsel.44 But this time, in contrast with the denouement of the O.P.M. story, Congress and the Securities and Exchange Commission were also engaged. During the process of drafting what was passed as the Sarbanes-Oxley Act, Senator Enzi of Wyoming asked a version of Judge Sporkin’s “where were the lawyers?” question. He wondered why, if 41

See generally BETHANY MCLEAN & PETER ELKIND, THE SMARTEST GUYS IN THE ROOM: THE AMAZING RISE

AND SCANDALOUS FALL OF ENRON (2003); MIMI SWARTZ & SHERRON WATKINS, POWER FAILURE: THE INSIDE

STORY OF THE COLLAPSE OF ENRON (2003); LOREN FOX, ENRON: THE RISE AND FALL (2003). 42 For accounts of the history summarized in this and the following paragraph, see Ariens, Chicken, supra note __, at 270-81; Thomas G. Bost, Corporate Lawyers after the Big Quake: The Conceptual Fault Line in the Professional Duty of Confidentiality, 19 GEO. J. LEGAL ETHICS 1089 (2006); Thomas D. Morgan, SarbanesOxley: A Complication, Not a Contribution, in the Effort to Improve Corporate Lawyers' Professional Conduct, 17 GEO. J. LEGAL ETHICS 1 (2003); Larry P. Scriggins, Legal Ethics, Confidentiality, and the Organizational Client, 58 BUS. LAW. 123 (2002). 43 Ethics 2000 Final Report, quoted in Scriggins, supra note __, at 126 (alterations committed). 44 Bost, supra note __, at 1113.

lawyers almost certainly had to be involved in drafting documents to give effect to the fraudulent transactions engaged in by Enron employees, there was nothing in the pending legislation speaking to the duties of counsel for a corporation.45 In response, Senator Edwards of North Carolina introduced a floor amendment requiring the SEC to issue regulations “seîng forth minimum standards of professional conduct for aôrneys appearing and practicing before the Commission.”46 This the SEC eventually did, but in the meantime, the ABA panicked. The House of Delegates having just rejected the Ethics 2000 Commission’s proposal for a rule permiîng limited disclosure of confidential information, the ABA President appointed a “Task Force on Corporate Responsibility,” charged with examining “systemic issues” relating to situations like the Enron collapse that “have shaken confidence in the effectiveness of the governance and disclosure systems applicable to public companies in the United States.”47 The Task Force specifically focused on the role of lawyers, along with other key participants, as part of a system of checks and balances intended to promote corporate integrity and responsibility.48 The Task Force recommended that the House of Delegates adopt the amendment it had just rejected, permiîng disclosure to prevent, rectify, or mitigate client crimes or frauds for which the client had used the lawyer’s services. It was widely assumed at the time that the Task Force hoped to forestall federal legislation that would have intruded into the regulation of lawyers, traditionally a maêr for state courts.49 In this it was unsuccessful, and Section 307 of the Sarbanes-Oxley Act was included in the final version of the legislation.50 The House of Delegates, not needing a weatherman to know

See Scriggins, supra note __, at 123. Id. at 123-24; Bost, supra note __, at 1102-03. 47 ABA Task Force on Corp. Responsibility, Report of the American Bar Association Task Force on Corporate Responsibility, 59 BUS. LAW. 145, 145-46 (2003). 48 Id. at 146. 49 See Morgan, supra note __, at 15-16. Professor Morgan is not only a highly respected scholar of legal ethics but also has considerable experience with the ABA; he would be in a good position to know whether the Task Force intended to head off regulation of lawyers by Sarbanes-Oxley. 50 A subsidiary controversy related to Sarbanes-Oxley was the consideration by the SEC of rules requiring disclosure, or “reporting out,” of client confidences, in addition to the procedure it eventually mandated, for “reporting up” information, up the corporate chain of command potentially as far as the board of directors. See, e.g., Ariens, Chicken, supra note __, at 278-81; Bost, supra note __, at 1108-10; Morgan, supra note __, at 19. At the time this was referred to as a “noisy withdrawal” procedure, and it shared a certain coyness about the violation of the duty of confidentiality; under the SEC’s proposal, a lawyer who believed a material violation of securities laws was ongoing or about to occur, and likely to result in substantial financial injury to investors, then the lawyer must withdraw from the representation, indicate that the withdrawal was based on “professional considerations” and so inform the SEC within one day. See Bost, supra note __, at 1109. 45 46

which way the wind was blowing, narrowly adopted the proposal.51 Model Rule 1.6(b)(2) and (b)(3) now contains an exception permiîng, but not requiring, disclosure of confidential information to the extent the lawyer reasonably believes necessary (2) to prevent the client from commiîng a crime or fraud that is reasonably certain to result in substantial injury to the financial interests or property of another and in furtherance of which the client has used or is using the lawyer's services; (3) to prevent, mitigate or rectify substantial injury to the financial interests or property of another that is reasonably certain to result or has resulted from the client's commission of a crime or fraud in furtherance of which the client has used the lawyer's services . . .. It is difficult to overstate the force of the opposition to even this incremental shift in the balance of client guardian and law guardian duties. The combined forces of public criticism in the wake of the Enron scandal and the threat of federal regulation were required to secure the narrow passage of these rule amendments. Defense of the traditional client guardian position was ferocious. Briefly summarizing (and toning down) a feisty article by my friend and co-author Larry Fox,52 client guardians deny that lawyers have an obligation to ensure that their clients do the right thing, or at least should be deemed equally responsible for the wrongdoing of clients.53 The obligation of lawyers is, instead, to manifest loyalty and maintain unbreakable secrecy with respect to client information, to enable the development of trust between the lawyer and client.54 Within that relationship of trust the lawyer may “remonstrate with the client to Bost, supra note __, at 1113; Morgan, supra note __, at 23. Even with the perceived importance of heading off federal regulation, the House of Delegates passed the Task Force proposal by the margin of 218 to 201. See Douglas R. Richmond, Lawyers' Duty of Confidentiality and Clients' Crimes and Frauds, 38 GA. ST. U. L. REV. 493, 498 n.18 (2022). 52 Lawrence J. Fox, The Fallout from Enron: Media Frenzy and Misguided Notions of Public Relations Are No Reason to Abandon Our Commitment to Our Clients, 2003 U. ILL. L. REV. 1243 (2003) [hereinafter, “Fox, Fallout from Enron”]. 53 Id. at 1246. 54 Compare the Supreme Court’s well-known explanation of the policy underlying the attorney-client evidentiary privilege: The purpose of the privilege “is to encourage full and frank communication between attorneys and their clients and thereby promote broader public interests in the observance of law and administration of justice.” Upjohn Co. v. United States, 449 U.S. 383, 389 (1981). In a separate article, Fox argues that unless exceptions to the duty of confidentiality are drawn as narrowly as possible “our clients will not share with us their innermost secrets and will view their lawyers with suspicion and distrust.” Lawrence J. Fox, It Takes More Than Cheek to Lose Our Way, 77 ST. JOHN’S L. REV. 277, 283 (2003) [hereinafter, “Fox, More Than Cheek”]. Of course, much of what lawyers learn about their clients’ activities is not through heart-to-heart conversations about the innermost secrets of clients, but in the ordinary 51

conform the client’s conduct to the law,”55 but that is the extent of law guardian duties that are compatible with the traditional conception of the lawyer’s role. Any concession to law guardian duties runs the risk that lawyers will react to the possibility of discipline or liability, by “skeptically auditing client conduct and erring on the side of blowing the whistle on the client in even the most marginal cases.”56 Not only do lawyers become “turncoats,” betraying their duty of loyalty to clients, but the result is not even a reduction in client fraud, because the erosion of trust in the lawyer-client relationship will mean that clients hold their lawyers at arms’ length and conceal bad facts from them.57 One of Fox’s objections to the rule amendments was the lack of any empirical evidence that lawyers’ services had, in fact, been used by Enron or other corporate clients to further their frauds.58 On the flip side, one may ask whether there is evidence that any serious corporate wrongdoing has been prevented since the amendments to Model Rule 1.6(b) were adopted. It appears that there are still instances, such as the fraud at the root of the bankruptcy of FTX, in which lawyers’ services were used to perpetrate large-scale harms. The net balance of costs and benefits of the rule amendments is difficult to assess, however, since there may be situations in which a client was deterred from a course of conduct by knowledge of lawyers’ permission to report out confidential information or even an explicit threat by a lawyer to disclose information. A lawyer’s successful interdiction of client wrongdoing may leave no paper trail.59 Thus, it is hard to know whether the regulatory response to the Enron frauds was justified. In any event, something like the Enron story has recently repeated itself in the ABA’s reaction to the threat of Congressional regulation of lawyers’ involvement in their clients’ financial crimes and frauds. B.

The ABA’s recent anti-money laundering amendment.

More recently, the ABA has been feeling pressure from domestic and international regulators, including the inter-governmental Financial Action Task Force course of working on a matter. Model Rule 1.6 covers all “information relating to representation” of a client, not only those matters divulged as if under seal of confession. 55 Fox, Fallout from Enron, supra note __, at 1246. 56 Id. at 1247. 57 Fox, More Than Cheek, supra note __, at 285-87. 58 Fox, More Than Cheek, supra note __, at 278-79; Fox, Fallout from Enron, supra note __, at 1244, 1248-49, 1253-54. 59 One thing we do know is that there have been no reported instances of discipline by the SEC of a lawyer for failing to follow the reporting-up procedures set out in the regulations promulgated pursuant to Section 307 of the Sarbanes-Oxley Act. See 17 C.F.R. Part 205.

(“FATF”) and the United States Department of the Treasury, to respond to concerns about money laundering and terrorist financing.60 Regulators in the U.S. had succeeded in promulgating regulations affecting banks, financial intermediaries such as investment advisors, decentralized finance platforms, and online gaming companies. The problem of lawyers facilitating money laundering transactions had briefly a^racted public aêntion in the spring of 2016, when an ad hoc international working group of investigative journalists, led by the German newspaper Süddeutsche Zeitung, published an analysis of a massive document leak from the Panamanian law firm Mossack Fonseca, obtained from a whistleblower inside the firm.61 The documents detailed how wealthy businesspeople and government officials in countries including Iceland, China, Russia, Brazil, and Saudi Arabia used complex structures involving offshore shell corporations to hide their true identity as owners of assets. The reporting revealed the law firm’s involvement in the creation of 214,488 offshore entities in over 200 countries but, unsurprisingly, concentrated in well-known “secrecy jurisdictions” such as the Cayman Islands, the British Virgin Islands, the Seychelles, and Panama.62 The firm had offices in many countries and in Nevada and Wyoming in the United States,63 and reportedly worked with prominent banks such as Deutsche Bank, Credit Suisse, and ABN Amro.64

See DEPARTMENT OF THE TREASURY, NATIONAL MONEY LAUNDERING RISK ASSESSMENT (Feb. 2024); FINANCIAL ACTION TASK FORCE, ANTI-MONEY LAUNDERING AND COUNTER-TERRORIST FINANCING MEASURES: UNITED STATES – 3D ENHANCED FOLLOW-UP REPORT & TECHNICAL COMPLIANCE RE-RATING (Mar. 2020). 61 See International Consortium of Investigative Journalists, Giant Leak of Offshore Financial Records Exposes Global Array of Crime and Corruption, ICIJ: THE PANAMA PAPERS (Apr. 3, 2016), https://panamapapers.icij.org/20160403-panama-papers-global-overview.html. See also Eric Lipton & Julie Cresswell, Panama Papers Show How Rich United States Clients Hid Millions Abroad, N.Y. TIMES (June. 5, 2016). 62 See Emilia Diaz-Struck, Cécile Schilis-Gallego, Beyond Panama: Unlocking the World’s Secrecy Jurisdictions, ICIJ: THE PANAMA PAPERS (May 9, 2016), https://panamapapers.icij.org/20160509-beyondpanama-secrecy-jurisdictions.html. 63 See Kevin G. Hall & Marissa Taylor, US Scolds Others About Offshores, But Looks Other Way at Home, MCCLATCHY NEWSPAPERS (Apr. 5, 2016), http://www.mcclatchydc.com/news/nationworld/national/article70008302.html. McClatchy is the U.S. member of the international consortium. In response to the McClatchy report, the Wyoming Secretary of State investigated and penalized the state office of Mossack Fonseca for failure to comply with statutory requirements concerning registered agents of corporations. Kevin G. Hall, Wyoming Investigates Panama Papers Law Firm, MCCLATCHY NEWSPAPERS (Apr. 6, 2016), http://www.mcclatchydc.com/news/nation-world/national/article70408322.html. 64 See Martha M. Hamilton and Hamish Boland-Rudder, Banks Ordered to Provide Info on Panama Dealings to NY Regulator, ICIJ: THE PANAMA PAPERS (Apr. 20, 2016), https://panamapapers.icij.org/20160420-nybanks-regulator.html. 60

` The Panama Papers story blew over fairly quickly as a maêr of public controversy, but around the same time the American legal profession was embarrassed by a sting operation run by an organization called Global Witness.65 The group hired an actor to pose as the Minister of Mines from an African country who had amassed millions of dollars in what were pre^y transparently bribes paid by companies wanting to do business in the country. The “Minister” sought assistance in buying real estate in the U.S.66 Many of the lawyers, whose meetings with the imposter were videotaped and later broadcast on 60 Minutes, seemed quite happy to provide advice on owning property through a shell corporation or straw purchaser. While the Panama Papers and Global Witness incidents were making headlines, behind the scenes domestic and international regulators were considering legal responses to the problem of lawyers providing assistance in financial crimes. Money laundering had long been a concern of regulators, including the Treasury Department’s Financial Crimes Enforcement Network (“FinCEN”). Regulators had been asking Judge Sporkin’s question: Where are the lawyers when money laundering and terrorist financing occurs? The answer is, they are frequently involved in providing assistance to clients engaging in financial wrongdoing. As in the post-Enron regulatory landscape, the ABA was concerned that if the organized bar did not take specific, tangible action to beef up regulatory constraints on lawyer assistance in financial crimes, Congress, the Treasury Department, or other regulators would step in and the result for the legal profession would be much worse. For example, Congress had debated provisions of the Corporate Transparency Act that would have required lawyers to disclose information about the beneficial owners of limited-liability entities used by their clients.67 Members of Congress had also considered regulating lawyers as “financial institutions” under the Bank Secrecy Act.68 The ABA lobbied successfully against these provisions69 but there was a sense that, without concrete steps toward enhanced anti-money laundering regulation by the state judiciary – the traditional regulator of the legal profession – the dam might burst and a flood of federal regulation might swamp lawyers. Indeed, the Treasury Department’s 2024 National Risk Assessment for Money Laundering cites the See, e.g., Debra Cassens Weiss, Group Goes Undercover at 13 Law Firms to Show How US Law Firms Facilitate Anonymous Investment, ABA J. (Feb. 1. 2016); Louise Story, Report Describes Lawyers’ Advice on Moving Suspect Funds Into U.S., N.Y. TIMES (Jan. 31, 2016). 66 This episode became the basis for Hypo #2 in the ABA ethics opinion on avoiding assisting client crimes or frauds. See Am. B. Ass’n, Standing Comm. on Ethics & Prof’l Resp., Formal Op. 491 (2020), at 12. 67 Am. B. Ass’n, Standing Comm. on Ethics and Prof’l Responsibility and Standing Comm. on Prof’l Regulation, Report to the House of Delegates and Resolution (Aug. 2023), at 4. 68 Id. Regulations under the Bank Secrecy Act for identifying beneficial owners of property are at 31 CFR § 1010.230. 69 Id. at 4-5. 65

conviction of a former law firm partner for money laundering, arising out of his establishment of a fake private equity fund to conceal what were in fact the proceeds from a cryptocurrency fraud scheme and a “prominent global law firm” involved in a civil asset forfeiture action after it allowed its trust account to be used to launder hundreds of millions of dollars from a Malaysian fund.70 The Treasury Report, rather pointedly, observes that lawyers in the United States are regulated on a state-by-state basis and that no state subjects lawyers to regulation aimed at anti-money laundering or counter-terrorist financing purposes.71 Reacting to this pressure, in August 2023, the American Bar Association’s House of Delegates adopted the following amendment to Rule 1.16(a), governing mandatory withdrawal from representation of a client: (a) A lawyer shall inquire into and assess the facts and circumstances of each representation to determine whether the lawyer may accept or continue the representation. Except as stated in paragraph (c), a lawyer shall not represent a client or, where representation has commenced, shall withdraw from the representation of a client if: (1) the representation will result in violation of the Rules of Professional Conduct or other law; (2) the lawyer's physical or mental condition materially impairs the lawyer's ability to represent the client; or (3) the lawyer is discharged; or (4) the client or prospective client seeks to use or persists in using the lawyer’s services to commit or further a crime or fraud, despite the lawyer’s discussion pursuant to Rules 1.2(d) and 1.4(a)(5) regarding the limitations on the lawyer assisting with the proposed conduct.72 Comment [1] to the rule now clarifies that the obligation to inquire into and assess the facts and circumstances of the representation applies to prospective clients – i.e. before accepting the representation – and continues throughout the representation. As I DEPARTMENT OF THE TREASURY, NATIONAL MONEY LAUNDERING RISK ASSESSMENT (NMLRA) (Feb. 2024), at 88, 92. 71 Id. at 92-93. 72 Am. B. Ass’n, Standing Comm. on Ethics and Prof’l Responsibility and Standing Comm. on Prof’l Regulation, Report to the House of Delegates and Resolution (Aug. 2023). 70

interpret the temporal scope of the duty, a lawyer may have re-assess the propriety of an existing relationship if new facts come to light suggesting the need for a renewed investigation. Amendments to Comment [2] to Rule 1.16 explain the contextual approach to assessing the risk of a client representation: A lawyer ordinarily must decline or withdraw from representation if the client demands that the lawyer engage in conduct that is illegal or violates the Rules of Professional Conduct or other law. The lawyer is not obliged to decline or withdraw simply because the client suggests such a course of conduct; a client may make such a suggestion in the hope that a lawyer will not be constrained by a professional obligation. Under paragraph (a)(4), the lawyer’s inquiry into and assessment of the facts and circumstances will be informed by the risk that the client or prospective client seeks to use or persists in using the lawyer’s services to commit or further a crime or fraud. This analysis means that the required level of a lawyer’s inquiry and assessment will vary for each client or prospective client, depending on the nature of the risk posed by each situation. Factors to be considered in determining the level of risk may include: (i) the identity of the client, such as whether the client is a natural person or an entity and, if an entity, the beneficial owners of that entity, (ii) the lawyer’s experience and familiarity with the client, (iii) the nature of the requested legal services, (iv) the relevant jurisdictions involved in the representation (for example, whether a jurisdiction is considered at high risk for money laundering or terrorist financing), and (v) the identities of those depositing into or receiving funds from the lawyer’s client trust account, or any other accounts in which client funds are held. . .. The accompanying Report provides further context. It makes frequent reference to the idea of “red flags” indicating that clients may be seeking to use a lawyer’s services for illicit purposes.73 Amended Model Rule 1.16(a) does not include an explicit mental state requirement qualifying the language, “the client or prospective client seeks to use or persists in using the lawyer’s services to commit or further a crime or fraud.” This is in contrast with Model Rule 1.2(d), which states that “[a] lawyer shall not counsel a client to engage, or assist a client, in conduct that the lawyer knows is criminal or fraudulent.”74 Even if the new rule had included an express knowledge requirement, a recent ABA ethics opinion had cautioned lawyers that, not only may knowledge be inferred from circumstances, but the knowledge standard may be 73 74

Id. at 9-10, 12, 15. Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.2(d).

satisfied by conduct amounting to willful blindness or conscious avoidance of facts.75 When added to the explicit due diligence requirement in Rule 1.16(a) – a lawyer “shall inquire into and assess the facts and circumstances of each representation” – the ethics opinion, lack of an express mens rea term in the rule, and repeated references in the Report to red flags combined to suggest something approaching a duty of inquiry to determine whether the lawyer’s services are being used to perpetuate a wrongful act. As I have said repeatedly, to my law school Professional Responsibility class and to anyone else who will listen, lawyers are not nearly as freaked out about the amendments to Model Rule 1.16(a) as they should be. That may be because most lawyers do not carefully follow developments in the ABA (and fair enough). Even if they do, a lawyer would have to be particularly vigilant to notice that an amendment to the rule on mandatory withdrawal from representation is actually a gatekeeping rule requiring significant due diligence before accepting a prospective client representation. (Wouldn’t it have made more sense as part of the rule on prohibited assistance?) Without knowing that the amendment was adopted under pressure to do something about lawyer involvement in financial crimes and frauds, the black-leêr text, ”inquire into and assess the facts and circumstances of each representation to determine whether the lawyer may accept or continue the representation,” looks relatively innocuous.76 Even a lawyer who read the text might not have read the comment, and almost certainly not the accompanying Report. Law firm general counsels probably are aware of these issues, but it will take a while before most lawyers start hearing the news through CLEs and loss-prevention materials sent around by liability insurers.

Am. B. Ass’n, Standing Comm. on Ethics & Prof’l Resp., Formal Op. 491 (2020). Lawyers would be even more nervous if they noticed that the Treasury Department’s 2024 National Money Laundering Risk Assessment hinted that federal regulation might be necessary if state courts didn’t hurry up and adopt the amendments to Model Rule 1.16(a): 75 76

The revision to these codes by states leaves attorneys substantial discretion to determine whether to accept or continue representation under the facts and circumstances of a particular case. These state codes are permissive rather than mandatory because they generally leave final decisionmaking and authority over the conduct of attorneys to a state court or a specially designated grievance or discipline committee within the state. These entities generally lack the resources or authority to conduct systematic audits, examinations, or other regulatory measures for AML/CFT purposes. At the same time, unscrupulous attorneys continue to be involved in complex money laundering, sanctions evasion, and other illicit finance schemes. DEPARTMENT OF THE TREASURY, NATIONAL MONEY LAUNDERING RISK ASSESSMENT (NMLRA) (Feb. 2024), at 93.

What maêrs in the longer term is whether the recent moves by the ABA, starting with the post-Enron amendments to the confidentiality rule (Model Rule 1.2(d)), and now including the anti-money laundering amendment to the withdrawal rule (Model Rule 1.16(a)), will begin to nudge the thinking of lawyers, courts, and regulators about the responsibilities of lawyers representing clients engaging in largescale bodily harm-creating conduct. Opponents of law guardian duties, such as Larry Fox, might contend that the legal profession is like the cliché of the frog that fails to notice the increasing temperature of a pot of water and ends up boiling to death. Those who are more favorably inclined toward a robust law guardian component to the ethical obligations of lawyers (myself included) could see this instead as the type of incremental change that may gradually win acceptance. The sky has not fallen in response to the permission to disclose information to prevent, rectify, or mitigate financial frauds, and so perhaps the profession is primed to consider a further expansion in its gatekeeping role. The remainder of the paper will consider how these recent changes may play out in the context of the types of tortious wrongdoing at issue in proceedings like the opioids MDL and the aêmpts by Johnson & Johnson to spin off its consumer-products division to divest itself of talc-related liability. III.

Red Flags at Purdue and J&J.

At the initial pretrial conference following the assignment of the opioids MDL to his court in the Northern District of Ohio, District Judge Dan Polster declared that responsibility for the crisis of opioid addiction was wide ranging: What's happening in our country with the opioid crisis is present and ongoing. I did a li^le math. Since we’re losing more than 50,000 of our citizens every year, about 150 Americans are going to die today, just today, while we’re meeting. And in my humble opinion, everyone shares some of the responsibility, and no one has done enough to abate it. That includes the manufacturers, the distributors, the pharmacies, the doctors, the federal government and state government, local governments, hospitals, third-party payors, and individuals. Just about everyone we’ve got on both sides of the equation in this case.77

See Transcript of Proceedings at 6, In re Nat'l Prescription Opiate Litig., No. 1:17-MD-02804 (N.D. Ohio Jan. 9, 2018), quoted in Howard M. Erichson, MDL and the Allure of Sidestepping Litigation, 53 GA. L. REV. 1287, 1290 (2019). Judge Polster’s clearly stated preference for a global settlement of the opioids proceedings drew a petition for mandamus, seeking his disqualification, which was denied by the Sixth Circuit. See In re National Prescription Opiate Litigation, 2019 WL 7482137 (6th Cir., Oct. 10, 2019). 77

Interestingly, however, Judge Polster did not ask Judge Sporkin’s question: Where were the lawyers for the manufacturers, distributors, pharmacies, and so on? [This section to be wriên later. It is the part of the paper where I have the least to contribute from the legal ethics perspective. At least among Clifford Symposium participants, the facts of Purdue’s knowledge of the addictive properties of OxyContin and J&J’s long-running efforts to eliminate asbestos fibers from its talc supplies are well known. A recent report in the New York Times referred to knowledge by DuPont and other manufacturers of the health risks of PFAS as far back as 1961. Hiroko Tabuchi, Lawyers to Plastics Makers: Prepare for “Astronomical” PFAS Lawsuits, N.Y. Times (May 28, 2024). Those facts will serve as hypotheticals for lawyer duties under the amended Model Rule 1.16(a). I will contend there were sufficient red flags, as the ABA Report understands that idea. The application questions will be considered in Section IV.] IV.

Lawyer Gatekeeping Beyond Financial Crimes.

The concept of lawyer gatekeeping is deeply rooted in the law and ethics of the legal profession, even if lawyers sometimes resist it. Its basis is the lawyer’s fiduciary obligation to the client to “proceed in a manner reasonably calculated to advance a client’s lawful objectives, as defined by the client after consultation.”78 As I have argued, there is a difference between that statement of the fundamental duties of the lawyer and the way many American lawyers tend to understand it: •

The usual approach to lawyer ﬁduciary duties: A lawyer may do anything at all on behalf of a client, as long as it does not constitute a crime or fraud – citing Model Rule 1.2(d).

•

A beNer understanding of the ﬁduciary role: A lawyer may do on behalf of a client only what the law actually permits the client to do.79 Now, citing Model Rule 1.16(a), a lawyer may have an obligation to withdraw from a representation that involves assisting the client in doing something that is not lawful.

Gatekeeping, on the second (be^er) understanding of the ﬁduciary role entails refusing to provide assistance in client objectives that are a legal nullity. The law establishes juridical relationships: Duties that are correlated with various sorts of claim rights, liabilities, and liberties.80 A lawyer may assist a client in establishing a juridical RESTATEMENT, supra note __, § 16(1) (emphasis added). Wendel, Complex Loyalty, supra note __, at 178. 80 Wesley N. Hohfeld, Fundamental Legal Conceptions as Applied in Judicial Reasoning, 26 YALE L.J. 710 (1917). 78 79

relationship to the extent the law empowers the client to enter into that relationship. To recycle my silly example, a lawyer who prepares a document granting a 99-year lease to a one-acre parcel on the moon is acting improperly.81 Not because there is a disciplinary rule one can point to that says a lawyer may not do that, but because as an agent of the client, the lawyer has no greater power than the principal has, and it is extremely unlikely that the principal has title to land on the moon. As applied to providing assistance in financial frauds, gatekeeping responsibility would be based on the underlying prohibition on the client’s conduct. The leases in the O.P.M. case purported to grant rights to a lender in a stream of income from an asset that did not exist. In fiduciary-theory terms, the wrongfulness of the assistance in that conduct is grounded in the lack of legal authorization on the client’s part to convey something that did not exist. Although the fraudulent transactions at Enron were much more complicated, the fiduciary critique of the lawyers’ conduct is the same: They purported to authorize a legal characterization of the transactions that was not supported by the economic reality.82 The law did not authorize the client, Enron, to do what it wanted to do, so the lawyers as agents for the principal had no lawful authorization to do what they purported to do. The fiduciary-theory analysis can be extended to explain gatekeeping responsibility with respect to tortious wrongdoing for causing bodily injury: The law does not authorize the distribution of defective products. At some point a faithful agent, exhibiting fiduciary loyalty properly understood, must refuse to offer assistance in designing and marketing products that satisfy the legal definition of defective in design or warnings. In the real world, of course, the situation is more complicated because the rights and duties of clients are frequently reasonably contestable. In most cases, lawyers will not have any kind of gatekeeping responsibility for this reason.83 Whether a transaction is lawful or whether a product is defective in design is frequently a maêr about which judgment is required. The factual record may be uncertain and a lawyer may have access to only some of the evidence they would need to conclude that the client’s activities are not legally authorized. Gatekeeping duties must be sensitive to the need, in many cases, for the exercise of reasonably informed professional judgment. Gatekeeping responsibilities also must be consistent with the representation of Wendel, Complex Loyalty, supra note __, at 182. See W. Bradley Wendel, Professionalism as Interpretation, 99 NW. U. L. REV. 1169 (2005). 83 For a longer defense of the adversarial system of litigation from the point of view of democratic political theory, with which I am largely in agreement, see Martin H. Redish, The Adversary System, Democratic Theory, and the Constitutional Role of Self-Interest: The Tobacco Wars, 1953-1971, 51 DEPAUL L. REV. 359 (2002). As discussed below, the requirement that the lawyers services have been used in furtherance of the client’s fraud insulates litigation advocates from gatekeeping responsibilities. 81 82

defendants in tort litigation and the good faith assertion of legal and factual defenses to liability. It is also important, as opponents of lawyer gatekeeping have consistently argued, that any duties to withdraw from representation or report up evidence of wrongdoing, and any permission to disclose conﬁdential information, must be narrowly tailored to avoid undue interference with the trust and conﬁdence that characterizes the lawyer-client relationship. Lawyers are not as troubled by the amendments to Rule 1.16(a) as they probably should be, but it is also not realistic to think that the amendment will remake the entire landscape of the law and theoretical foundation of the lawyer-client relationship. Let’s consider in detail some of the feature of the recent amendment to Model Rule 1.16(a). A.

Expansion of Rule 1.16(a) to Bodily Harms?

The ABA’s Report accompanying the adoption of Rule 1.16(a) states that the rule should not be limited to the core instances of money laundering and terrorist financing: While the impetus for these proposed amendments was lawyers’ unwiîng involvement in or failure to pay appropriate aêntion to signs or warnings of danger (“red flags”) relating to a client’s use of a lawyer’s services to facilitate possible money laundering and terrorist financing activities, it is clear that lawyers’ client due diligence obligations apply broadly to all lawyers.84 The Report did say the amended rule would help lawyers avoid liability for their clients’ financial crimes and frauds, but then gave a list of categories of wrongdoing that is not so limited: . . . including tax fraud, mortgage fraud, concealment from disclosure of assets in dissolution or bankruptcy proceedings, human trafficking and other human rights violations, violations of U.S. foreign policy sanctions and export controls, and U.S. national security violations.85 The reference to human trafficking and human rights in particular suggests that the rule may be expanded to include wrongful conduct including bodily harm.

84 85

ABA Report and Resolution (Aug. 2023), at 2. Id.

The history recounted in Section II pertains to financial-frauds gatekeeping, because that was the context of the debate surrounding the drafting of the 1983 Model Rules (the O.P.M. case) and the Ethics 2000 amendments (the Enron collapse). There have been similar controversies over duties of lawyers to take action to prevent physical harm. Most famously, a 1962 Minnesota case called Spaulding v. Zimmerman86 and, more recently, the 26-year-long wrongful imprisonment of Alton Logan87 have focused discussion on the narrow scope of the permission to disclose confidential information to prevent bodily harm. An exception to the confidentiality rule provides: A lawyer may reveal information relating to the representation of a client to the extent the lawyer reasonably believes necessary to prevent reasonably certain death or substantial bodily harm.88 Lawyers are not permiêd to disclose confidential information to prevent possible (i.e. not “reasonably certain”) bodily harm or the harm related to wrongful incarceration, even if it involves the harm of confinement of one’s body. But should they be? In effect this is a question pertaining to gatekeeping with respect to bodily harms. The same policy arguments that recur whenever there is a debate over preventing financial harms can be rehearsed in this context as well89: The normative foundation of the lawyer-client 116 N.W.2d 704 (Minn. 1962). See also Roger C. Cramton & Lori P. Knowles, Professional Secrecy and Its Exceptions: Spaulding v. Zimmerman Revisited, 83 MINN. L. REV. 63 (1998). Very briefly, Spaulding involved a lawyer representing the defendant in a personal-injury action arising out of an automobile accident. Through an independent medical examination, the defendant’s lawyer learned that the plaintiff’s injuries were more serious than the plaintiff or his lawyer were aware, and included an aortic aneurysm that was potentially life-threatening if not surgically repaired immediately. The case was generally discussed with reference not to the confidentiality provision existing at the time in Minnesota, based on the ABA’s 1908 Canons of Ethics, but as a way of criticizing the disclosure provision in the 1983 Model Rules. The bodily-harm exception in the 1983 Model Rules would have permitted disclosure only where necessary “to prevent the client from committing a criminal act that the lawyer believes is likely to result in imminent death or substantial bodily harm.” The Ethics 2000 Commission modified the rule so that it is now the version quoted in the text. 87 See ALTON LOGAN & BERL FALBAUM, JUSTICE FAILED: HOW “LEGAL ETHICS” KEPT ME IN PRISON FOR 26 YEARS (2018). Alton Logan was serving a sentence in Illinois state prison for a murder in Chicago when another man, convicted on a separate crime, told his two lawyers that he had committed the murder for which Logan was convicted. The other man, Wilson, wrote out and signed a confession but directed his lawyers to keep it confidential until his death. When Wilson died, the lawyers immediately moved to vacate Logan’s conviction on the basis of their deceased client’s confession. The two lawyers have stated that they concluded, with great reluctance, that the Illinois rule on confidentiality did not permit them to disclose their client’s confession without their client’s informed consent. Their position has been generally accepted as correct by lawyers and legal scholars, even though it resulted in an extended period of wrongful incarceration for Logan. 88 Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.6(b)(1). 89 As in Cramton & Knowles, supra note __, at 106-13. 86

relationship is the maintenance of a relationship of trust and confidence, not for its own sake, but toward the end of bringing clients into compliance with the law.90 The client’s reasonable expectation of confidentiality is not absolute, but dependent upon the client’s scope of lawful action. The aôrney’s withdrawal from the relationship – noisy or silent – may interfere the lawyer-client relationship but that interference is justified by the harm thereby avoided, whether physical injury or wrongful incarceration. Some limitations on and exceptions to the duty of confidentiality are justified by preventing third-party harms. If certain financial harms are sufficiently serious to justify the limitation on the duties of loyalty and confidentiality, then conceivably some physical harms may be as well. Larry Fox has a truly excellent line describing the limitation of the ABA’s role in regulating lawyers: “When the ABA House adopts rules, all that really happens is the publication of a booklet.”91 What this means is that a Model Rule has to be adopted by the judiciary of any given state in order to have any impact on the duties of lawyers admiêd to practice in that state, or providing legal services in that state. Law school Professional Responsibility courses teach, and the MPRE tests on, the Model Rules because, for the most part, the content of the rules is relatively consistent from one state to another. There are exceptions, however, and the confidentiality rule is one of the rules that tends to have the most state-by-state variation. The expansion of gatekeeping obligations with respect to bodily harms may occur expressly, through greater permission to disclose confidential information under Rule 1.6(b). I also expect there will be considerable variation in state adoption of Rule 1.16(a). It is possible that some states may elect to impose additional requirements of declining or withdrawing from representation under Rule 1.16(a), either by deleting the limitation to client conduct that amounts to a crime or fraud or by specifying substantial bodily harm as wrongful activity that would require a lawyer’s withdrawal from representation. B.

Limitations on Gatekeeping Duties. 1.

The crime or fraud limitation.

The ﬁrst limitation on expansion of amended Model Rule 1.16(a) to establish gatekeeping responsibilities with respect to large-scale tortious wrongs is the language of commi^ing or furthering a crime or fraud. A similar limitation is contained in Model Rule 1.2(d): Upjohn Co. v. United States, 449 U.S. 383, 389 (1981). Lawrence Fox, The Gang of Thirty-Three: Taking the Wrecking Ball to Client Loyalty, 121 Yale L.J. F. 567, 569 (2012). 90 91

A lawyer shall not counsel a client to engage, or assist a client, in conduct that the lawyer knows is criminal or fraudulent, but a lawyer may discuss the legal consequences of any proposed course of action with a client and may counsel or assist a client to make a good faith effort to determine the validity, scope, meaning or application of the law.92 “Crime” or “criminal” is not separately defined in the Model Rules – what else could it mean, other than conduct defined as a crime under the law of the relevant jurisdiction? – but “fraud” and “fraudulent” is defined as “conduct that is fraudulent under the substantive or procedural law of the applicable jurisdiction and has a purpose to deceive.93 Despite the intuitive reaction that most torts are not crimes or frauds, it actually may be a relatively common occurrence that instances of large-scale wrongdoing fall within the definition of the lawyer’s services being used to further a fraud.94 Consider the crashes of two Boeing 737-Max airliners, operated by Lion Air and Ethiopian Airlines, the causes of which were traced to an undisclosed feature of the flight-control system of the airplane.95 Pre^y clearly the introduction into the stream of commerce of a product with a design and warnings defect would be a tort. In addition, however, Boeing also failed to inform the FAA, during the process of certifying the modified 737 design, that it had included a software feature intended to smooth out control responses during certain flight-test maneuvers, but which was subject to erroneous activation that could potentially lead to a loss of control. This omission in its submission to the FAA exposed Boeing to prosecution for conspiracy to defraud the United States.96 The Justice Department entered into a deferred prosecution agreement with Boeing, which included the payment of a $2.5 billion fine and an undertaking by the company to “design, implement, and enforce a compliance and ethics program to prevent and

Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.2(d). Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.0(d). 94 This discussion is inspired by a running sort-of joke in Matt Levine’s Money Stuff newsletter, that everything is securities fraud. 95 I have written two articles on these accidents, the latter of which includes an update on some of the details in the former. See W. Bradley Wendel, Malice or SNAFU? Punitive Damages and Organizational Culture Defects, 72 DEPAUL L. REV. 457 (2022); W. Bradley Wendel, Technological Solutions to Human Error and How They Can Kill You: Understanding the Boeing 737 Max Products Liability Litigation, 84 J. AIR L. & COM. 379 (2019). 96 See Criminal Information, United States v. The Boeing Co., Case 4:21-cr-00005-O (N.D. Tex., Jan. 7, 2021). The information charged one count of violation of 18 U.S.C. § 371. 92 93

detect violations of the U.S. fraud laws throughout its operations.”97 Boeing is also the defendant in a civil securities class action, alleging securities fraud in the form of misstatements and concealment of safety issues related to the 737-Max.98 Not surprisingly, Johnson and Johnson is also the defendant in a securities class action lawsuit by shareholders alleging that the company’s stock price fell when it was hit with jury verdicts in 2017 and 2018 finding that the talc in its Baby Powder product was contaminated with asbestos.99 As discussed further below, it is important to focus on what, specifically, are the acts alleged to have been fraudulent when considering the ramifications of tortious conduct by a company. Securities fraud actions can be based on representations made by corporations in periodic filings with the Securities and Exchange Commission, such as on Form 10-K, or they may be based on public statements made by company representatives.100 In the context of gatekeeping with respect to financial frauds, both the ABA’s post-Enron amendment to the confidentiality rule and the SEC’s regulations promulgated under the Sarbanes-Oxley Act limit disciplinary exposure to lawyers whose services were used in connection with the fraud.101 The point to emphasize here, however, is that, at least at a medium level of generality, there is not a clean line between tortious conduct resulting in bodily harm and conduct amounting to fraud.

See Department of Justice, Press Release, https://www.justice.gov/opa/pr/boeing-charged-737-maxfraud-conspiracy-and-agrees-pay-over-25-billion; Jack Coffee severely criticized the deferred prosecution agreement in a paper he presented at a previous Clifford Symposium. See John C. Coffee, Jr., Nosedive: Boeing and the Corruption of the Deferred Prosecution Agreement (June 6, 2022). Available at SSRN: https://ssrn.com/abstract=4105514. In May 2024 the Justice Department notified Boeing that it had breached the agreement by failing to put into place an effective program to rectify shortcomings in its operations. The agreement was scheduled to expire two days after a panel blew out of an Alaska Airlines 737-Max climbing out of Portland, Oregon, causing no injuries but scaring the willies out of the passengers on the flight. See David Shepardson & Mike Spector, US Says Boeing Breached 2021 737 MAX Criminal Prosecution Deal, REUTERS (May 14, 2024). 98 See In re The Boeing Company Aircraft Securities Litigation, Case no. 1:19-cv-02394 (N.D. Ill.). 99 See Hall v Johnson & Johnson, Case no. 18-01833 (D.N.J.). 100 See, e.g., the Appendix to the district court’s order on Boeing’s motion to dismiss the shareholder action in which the court went through 49 different statements to determine whether they are potentially actionable or instead whether they were not false (e.g. “There are new systems on the airplane that are designed to take advantage of the capabilities of the airplane and provide control capability in high angle of attack conditions”); not made with scienter (“[T]hat’s part of the training manual. It’s an existing procedure. So the bulletin we put out again last weekend . . . pointed to that existing flight procedure”); were not material (“Boeing continues to have full confidence in the safety of the 737 MAX”); or constitute opinion or puffery (e.g. “the airplane is safe”). 101 See Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.6(b)(2), (b)(3); 17 C.F.R. § 205.2(a) (defining attorney “appearing and practicing before the Commission” as, among other things, providing advice in respect of federal securities laws regarding a document the attorney has notice will be filed with the SEC). 97

Mens rea requirement.

Arguing against the amendment to Model Rule 1.6(b) permiîng disclosure of confidential information to prevent, rectify, or mitigate financial crimes or frauds where the lawyer’s services had been used by the client, Larry Fox objects to the premise that fraud is apparent on its face: “Facts are ambiguous, hindsight is 20-20, and the ability of a lawyer to identify a good fraud is at a very low order of magnitude. That is why it’s called fraud.”102 The ABA Task Force that proposed the amendments that were eventually adopted as Model Rule 1.6(b)(2) and (b)(3) had proposed a duty to act when the lawyer “reasonably should know” the client’s conduct was improper; again, this was likely intended to fend off proposals then pending in Congress for federal regulation of lawyers in connection with client frauds.103 The current rule permits disclosure only to the extent the lawyer reasonably believes necessary to effectuate the purpose of preventing, rectifying, or mitigating fraud. However, the Model Rules are not particularly helpful in explaining how much information is necessary to support a lawyer’s reasonable belief.104 The lack of precision with mens rea standards is an underappreciated problem with the ABA’s Model Rules.105 There are some areas in which courts have stepped in to clarify the meaning of a mental-state term in the rules. For example, where a lawyer is considering whether to take remedial measures to correct false testimony offered by a criminal defendant, some courts have held that, in order to safeguard the constitutional rights of the defendant, the “knowledge” required by Rule 3.3(a)(3) is practically equivalent to “absolute certainty”: “Absent the most extraordinary circumstances, such knowledge must be based on the client’s expressed admission of intent to testify untruthfully.”106 With respect to the “reasonably believes necessary” standard in the exception to the confidentiality rule for financial frauds, a lawyer should probably regard this as a general reasonableness standard, taking into account what a lawyer of ordinary prudence would consider before drawing a conclusion.107 But there is no problem of ambiguity with respect to mens rea standards in connection with the amendments to Rule 1.16(a). Comment [2] to the amended rule sets out a multi-factorial test to determine whether the risk that the representation will Fox, More Than Cheek, supra note __, at 284. See Morgan, supra note __, at 17-18. 104 See Richmond, supra note __, at 507-08. 105 See Nancy J. Moore, Mens Rea Standards in Lawyer Disciplinary Codes, 23 GEO. J. LEGAL ETHICS 1 (2010). 106 State v. McDowell, 681 N.W.2d 500, 513 (Wis. 2004). 107 See RESTATEMENT, supra note __ § 52(1) (“a lawyer who owes a duty of care must exercise the competence and diligence normally exercised by lawyers in similar circumstances”). 102 103

result in a crime or fraud is sufficiently serious to warrant further investigation. In addition to this comment, the ABA ethics opinion on counseling or assisting client crimes or frauds in non-litigation seîngs includes numerous hypothetical scenarios illustrating the risks presented by certain recurring types of representations.108 For example, Hypo #4 (which I embellish a bit when teaching it) involves a general practitioner in North Dakota meeting with a prospective client who says he is interested in buying a ranch. From his dress and demeanor, the lawyer can tell the client has never been near a ranch in his life. However, the client says he just won a fortune at a casino in Las Vegas and thought it would be fun to own a few hundred head of ca^le. The opinion rightly concludes that additional investigation is required to ensure that the transaction is not being used as part of a scheme to launder the proceeds of a crime or fraud.109 One of the staunchest critics of lawyer gatekeeping responsibilities, Larry Fox, is right that fraud is seldom apparent on its face. However, the amendment to Rule 1.16(a) contemplates more than a facial assessment of the likelihood of fraud. Rather, the rule, comment, and earlier ethics opinion set out a two-step process. First, are there red flags requiring further investigation? Second, upon conclusion of the further investigation, would a reasonable lawyer conclude that the client is seeking or using the lawyer’s services to commit a crime or fraud? Fox gives an example to illustrate that the facts suggesting fraud may be ambiguous: [I]magine a lawyer asked to defend a pharmaceutical company in liability claims arising from a new miracle drug. In the course of the lawyer's investigation, he learns of a company lab employee who believes the drug causes liver damage. When the lawyer asks the head of research about these views, lawyer is told the lab guy is a nut and his charges baseless. . . . [T]he lawyer not only would be required to determine whether the nut characterization is true, a tough task for a lawyer who hardly has any expertise on this topic, but also disclose this nut's opinion if the lawyer thinks there is any chance the liability cases might be lost and some shareholder will come along contending that the liver problem should have been disclosed sooner, the lawyer should have known about it, and the lawyer failed to disclose it.110 His concern is that the lawyer may be deemed, in hindsight, to have had a reasonable belief that the client’s conduct was fraudulent. Without more, he insists, it would be Am. B. Ass’n, Standing Comm. on Ethics & Prof’l Resp., Formal Op. 491 (2020), at 11-13. Id. at 12 n.53. 110 Fox, More Than Cheek, supra note __, at 287. 108 109

unfair to hold the lawyer potentially subject to discipline or civil liability for failing to take steps to prevent the fraud, including potentially disclosure of confidential information: If it turns out that by some action the client engaged in a fraud or made a misrepresentation and the usual suspects are lined up to be held responsible, if the lawyer was on the scene, it will be asserted that the lawyer should have known, even if the lawyer did not know what was about to occur. Why wasn't the lawyer suspicious? Why didn't the lawyer ask more questions? Seek documentation? Check further? Contact third parties? The lawyer’s situation may be fraught in this way with respect to the disclosure rule, but the amendment to Rule 1.16(a) addresses this ambiguity. In response to Fox’s rhetorical questions, the answer is yes: the lawyer may be required to ask more questions, seek documentation, or check further. It may not be sufficient to accept the statement of the head of research that the lab guy is a nut. In fact, the discrepancy between the lab guy’s statement about the risk of liver cancer and the dismissive aîtude of the head of research may itself be a red flag requiring further investigation. (Note, however, that this would be the case only if the lawyer were doing something other than defending the company in litigation, which would not be services in furtherance of the company’s fraud.) An additional feature of the amendment to keep in mind: Rule 1.16 pertains to withdrawal from representation of a client. The Report talks about due diligence and red flags, so it would be natural to see the duties in the rule as applying only at the outset, when deciding whether to represent a client, Comment [1] to the rule clarifies, however, that the obligation to assess whether the lawyer’s services are being used to further wrongdoing applies throughout the course of the representation. A lawyer may have to re-assess the propriety of the representation and determine whether to withdraw, even from a professional relationship of long standing. At the outset of the representation and throughout its course, a lawyer must be alert to red flags and prepared to re-assess the propriety of providing legal services to the client. 3.

The requirement of the lawyer’s services having been used.

Lawyers who object to gatekeeping responsibilities sometimes worry that they are being asked to serve as deputized governmental regulators, constantly monitoring their clients’ behavior for compliance with the law. Or, more pejoratively, they claim to be required to snitch on their clients. It is important to be clear, however, that even with its red-ﬂags-plus-due-diligence requirements, amended Rule 1.16(a) applies only to 32

lawyers providing legal services in furtherance of the client’s crime or fraud. The permissive-disclosure provisions of Rule 1.6(b)(2) and (b)(3) similarly require that the lawyer’s services have been used in furtherance of the fraud as a condition of permiêd disclosure. Susan Koniak’s metaphor of the Wizard of Oz, with which this paper began, is more accurate than the comparison to rats or government agents. A lawyer who happens to run across evidence that a client is commiîng a crime or fraud is not a potential gatekeeper. As mentioned in the previous section, in connection with Larry Fox’s hypo of the possibly nu^y guy in the lab who believes the company’s drug causes liver cancer, lawyer-gatekeeping rules specifically require that the lawyer’s services be used in furtherance of the client’s crime or fraud. Representing a defendant in products liability litigation, contending that a product is not defective in its design or warnings, is not the provision of financial assistance in connection with a fraud, even if the fraud consists of the company’s concealment of its knowledge of risks on a Form 10-K filed with the SEC. This conclusion is supported by the caselaw on the crime-fraud exception to the aôrney-client privilege. No privilege aâches to a confidential communication between a lawyer and the lawyer’s client if the communication is for the client’s purpose of obtaining assistance in commiîng a crime or fraud.111 It is clear, however, that “[t]he exception does not apply to communications about client criminal or fraudulent acts that occurred in the past.”112 It follows that a lawyer furthering the defense of a past action by the client – including designing and marketing a defective product – is not furthering a fraudulent act by the client. Gatekeeping duties must be assessed lawyer-by-lawyer and maêr-by-maêr. It is not furthering fraud for an outside law firm to represent Purdue in litigation and asserting that the company should not be liable because (i) OxyContin was duly approved by the FDA and this approval should preempt state design-defect and failureto-warn claims; (ii) the cause of the plaintiff’s death could have been ingestion of opioids distributed by a different manufacturer; or (iii) the plaintiff’s own actions constituted contributory negligence or assumption of risk and should serve as a defense to liability. On the paêrn of the shareholder actions brought against Boeing and J&J, however, it may be fraud for an in-house lawyer at Purdue to approve a Form 10-K filing that makes a false statement or material omission about the company’s knowledge of the risks to consumers posed by OxyContin. Moreover, given the language in Rule 1.16(a) and comments about red flags, this in-house lawyer cannot ignore evidence such as greatly excessive shipments of the product to rural areas or reports by law 111 112

See RESTATEMENT, supra note __, § 82(a). Id., cmt. e.

enforcement of users crushing the tablets to defeat the time-release feature. Further due diligence may be required depending on the credibility and specificity of these reports. Finally, it is worth emphasizing that gatekeeping duties do not entail immediate withdrawal from representation or disclosure of confidential information. Model Rule 1.13 reinforces duties that exist as a maêr of the common law of agency, reminding lawyers that if retained by an organization, their client is the organization itself, not any individual constituent.113 The head of research in Larry Fox’s hypothetical who reassures the lawyer that the lab guy is a nut is not the lawyer’s client. If the lawyer knows that a constituent of the organization is engaged in action that is a violation of a legal duty to the organization, or a violation of law that reasonably might be imputed to the organization (such as distributing a defective product), then the lawyer must proceed in the best interests of the organizational client, which generally will involve “up the ladder” reporting, potentially to the board of directors.114 The lawyer must also “consult with the client about any relevant limitations on the lawyer’s conduct when the lawyer knows that the client expects assistance not permiêd by the Rules of Professional Conduct or other law.”115 A great deal of counseling, remonstrating, and if necessary strongly worded warnings should go on behind the scenes before the lawyer considers withdrawing from the representation, even though withdrawal might ultimately be required by Rule 1.16(a). Fox is right that the lawyer-client relationship is a highly fiduciary one, characterized by the importance of trust and confidentiality. Lawyers should certainly respect these ideals while seeking to counsel their clients to conform their conduct to the law. Lawyers should not jump the gun and disclose or withdraw in situations where the client would have complied with the law given sufficient time. If, however, in the end the client refuses to heed the lawyer’s advice, the new Rule 1.16(a) makes clear that the lawyer must cease to represent the client. Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.13(a). Id., Rule 1.13(b). Internally, corporations can be a chaotic mess of decentralized decisionmakers trying to overcome information silos and diffused responsibility. The GM ignition switch fiasco was largely a story of unclear lines of authority, committees with overlapping responsibility, and no centralized accountability mechanism. See ANTON R. VALUKAS, FOR JENNER & BLOCK, REPORT TO BOARD OF DIRECTORS OF GENERAL MOTORS COMPANY REGARDING IGNITION SWITCH RECALLS (2014). The internal dysfunction was memorably captured in congressional testimony given by CEO Mary Barra, describing the “G.M. nod,” when everyone in the room agrees with a proposed plan of action, but no one does anything to make it happen, and the “G.M. salute,” which consists of crossed arms with fingers pointing toward others, to whom responsibility is being punted See Peter J. Henning, How G.M.’s Lawyers Failed in Their Duties, N.Y. TIMES (June 9, 2014). Thus, up-the-ladder reporting and any subsequent follow-up by the lawyer to make sure that corrective action has been taken is anything but a straightforward matter. 115 Am. B. Ass’n, Model Rules of Prof’l Conduct, Rule 1.13(a), Rule 1.4(b)(5). 113 114

Impact of rules of professional conduct on other duties.

In addition to the need for the amendments to Model Rule 1.16(a) to be picked up by the states, there is a further question of what impact the rule will have beyond the disciplinary process of a state. Lawyers may dismiss the amended rule as unlikely to be enforced by the disciplinary apparatus of their state, which may be under-staffed, under-funded, and over-worked. Importantly, however, state rules of professional conduct may inform the application of other standards that exist as a maêr of the common law of tort, agency, or fiduciary law. This is not to say that the violation of a rule is negligence per se, or the rules create an implied private right of action. But violation of a rule of professional conduct “may be considered by a trier of fact as an aid in understanding and applying” the generally applicable standard of ordinary competence and diligence.116 Common law standards establishing and governing the lawyer-client relationship would apply even if state courts had not promulgated specific rules of conduct for lawyers. Once those more specific rules are in effect, however, a court may draw from the disciplinary rules to inform the application of the standard of care. V.

Conclusion.

As reports of the involvement of Fenwick & West in the FTX fraud shows, the post-Enron amendments to the Model Rules did not put an end to financial shenanigans commiêd by clients with the assistance of lawyers. Even if every state adopted the new version of Model Rule 1.16(a) tomorrow, and every lawyer read and understood it, I am not so naïve to think that there will be a significant impact on the frequency and severity of tortious wrongdoing by product manufacturers and other corporations. At the margin, however, there may be an effect. Some lawyer, either inside the organization or in an outside law firm, may get nervous about their involvement in a fraud related to concealment of the company’s awareness of the risks posed by a product and take their concerns up the corporate chain of command. Of course, lawyers could have done this last year as well, before the rule was amended, but the recent action by the ABA may indicate to lawyers that regulators are losing patience with the longstanding aîtude of the organized bar that lawyers have only client guardian responsibilities. The law guardian duties recognized by amended Rule 1.16(a) are not so onerous as to fundamentally remake the lawyer-client relationship. At most the represent a thumb on the scale in the direction of law guardian duties. But sometimes that is all that is needed to contribute to behavioral changes. 116

See RESTATEMENT, supra note __, § 52(2)(c).

Whether such behavioral changes are a good thing is, of course, a normative question on which lawyers and scholars can differ. I have tried to sidestep it here as much as possible, hoping to focus aêntion on the ABA’s history with gatekeeping rules and the new Model Rule 1.16(a) amendment, which I believe do have significance for lawyers advising product manufacturers and other corporations in connection with tortious conduct. I have consistently advocated for a normative position that can be summarized as “fidelity to law”: Lawyers have an obligation to represent clients with reference to the rights and duties that are allocated to clients by the positive of the applicable jurisdiction.117 Lawyers should not see themselves as having obligations, or even permission, to seek to promote the public interest or take into account the interests of third parties, except to the extent those interests are given recognition by the law. As I see it, the gatekeeping provisions in the ABA Model Rules, now including Rule 1.16(a), are broadly consistent with this approach. But even lawyers who disagree with that normative account should be aêntive to what I believe is a subtle but unmistakable trend in the ABA’s aîtude toward the law guardian dimension of the lawyer’s role.

117

See W. BRADLEY WENDEL, LAWYERS AND FIDELITY TO LAW (2010).

SESSION V: RESPONSES On the Relation Between Causation and Misconduct Alexandra Lahav, Cornell University sing Public Nuisance Litigation to Address Industry Misconduct: U Common Law Statutes, Delegation Doctrine and Systems Theory Timothy Lytton, Georgia State University Hillel Levin, University of Georgia Microlocal Litigation for a Mass Tort World Sarah Swan, Rutgers University Discussant: Robert Rabin, Stanford University

(Back to main agenda)

THE 30 TH ANNUAL

CLIFFORD SYMPOSIUM

ON TORT LAW AND SOCIAL POLICY

Early Draft

The Evolving Relationship Between Causation and Misconduct Alexandra D. Lahav1 Clifford Symposium, June 2024

INTRODUCTION There are three important differences between causation in law and causation in science.2 First, causation in law is attributive. In a lawsuit, causation is determined in service of assigning liability or fault to one party. By contrast, causation in science is explanatory. It does not make a normative judgment about who is to blame. Both involve judgment, but they are different kinds of judgments. Second, a legal case is ordinarily decided at one moment in time. A determination of causation and consequent liability may seem correct at time one, but science may evolve at time two to disprove that causal link. For example, initially scientists believed that ulcers were caused by acidic foods; they later learned that they are caused by bacteria.3 Past treatments can become obsolete, past causal explanations outdated, but a legal decision still binds. Third, a legal determination of liability is deterministic and binary: either the defendant is liable or they are not. Causation in science, particularly in the epidemiologic sciences, is probabilistic. Epidemiologists speak in the language of probabilities, and discuss the likelihood that a drug or device causes an increase in adverse outcomes. Because it is a science of populations, it cannot be used to determine whether any particular individual’s harm was caused by the drug or toxin. Juries are asked to use this probabilistic evidence to determine whether (by a preponderance of the evidence) a given fact is true. Notably, the burden of proof determines the strength of the evidence, not the truth of its content.4 The ultimate determination is deterministic under current law. If these three points are correct, then it is a category error to think of cause in law as being the same as cause in science.5 Yet in the 1980s judges began to think of causation in 1 Anthony W. and Lulu C. Wang Professor of Law, Cornell Law School. Many thanks to Bob Rabin for helpful comments and Malcolm Drenttel, whose research on this project moved it forward tremendously. 2 See Nancy Cartwright, Causation: One Word, Many Things, 71 Phil. Sci. 805, 814 (2004)(setting out the idea that the definition of causation can differ by context); Jennifer K. Robbennholt & Valerie P. Hans, The Psychology of Tort Law 65-82 (2016) (describing different understandings of causation among lay people as they relate to tort). 3 Richard G. “Bugs” Stevens, Does Monsanto’s Roundup Cause Cancer? The Conversation, August 3, 2018 (available at https://theconversation.com/does-monsantos-roundup-cause-cancer-trial-highlights-the-difficulty-of-proving-a-link-100875) (comparing the method of finding scientific consensus to the jury deliberation process). 4 See Alexandra D. Lahav, Chancy Causation in Tort Law, 15 J. Tort L. 109, 119 (2022); Mark Parascandola, Evidence and Association: Evidence of Confusion in Toxic Tort Law, 63 Phil. of Sci. 168, 174 (1996) (“One cannot justify reasoning from epistemic probabilities to physical chances and actual frequencies.”). 5 The observation that factual causation in law rests on policy questions has been made many times over the decades. See, e.g., Henry W. Edgerton, Legal Cause, 72 U. Penn. L. Rev. 211 (1924) (critiquing legal approaches to causation that claim to provide certainty and arguing that at bottom the causation inquiry rests on policy); Wex Malone, Ruminations on Cause-In-

science as a separable inquiry within a given lawsuit. Up until that time, causation and liability were intermingled and were presented to the jury in a unitary trial. This essay is an exploration of that change. The separation of the scientific causal inquiry from the liability inquiry is historically contingent. This development was made possible by two factors: the rise of strict liability and the tort reform movement. First, the rise of strict liability, which set to the side questions of defendant’s careless conduct in favor of other rationales for tort liability, laid the groundwork for isolating the causation inquiry. Such a separation of causation from fault would have been much harder absent strict liability. Second, the development of epidemiology in the middle of the twentieth century, particularly the epidemiology of chronic or non-infectious disease, at heart a form of probabilistic knowledge, created a tension between the determinism of the law and the probabilistic reasoning of the relevant science. This tension was resolved by separation at a political moment where industry arguments that tort law was out of control became increasingly powerful. The process of separation entailed a denial of the socially contingent nature of the scientific inquiry because evidence of the context in which the science was developed overlaps with evidence of misconduct. In service of the thesis that strict liability laid the legal groundwork and epidemiology the catalyst for the separation of causation from fault, this essay recounts two stories. The first is about a trial that proceeded wholistically, as a narrative, where causation was clear but fault was not. That was a trial against the pharmaceutical company Cutter for distributing a vaccine with a live, virulent polio virus that injured thousands of children, some severely. In a rare move, the jury wrote the judge a note that they found the company not at fault, but nevertheless felt obligated to hold it liable under a breach of warranty theory: the company had promised a killed virus and instead had delivered a live one. Although not proceeding on a strict products liability theory, this case came to stand for the idea of strict products liability in pharmaceuticals through the doctrines of express and implied warranty. Years later, it was discovered that the company had been careless in manufacturing the vaccine, but the jury had on way of knowing this at the time. The jury’s finding that the company was not at fault, mostly because it had followed government regulations in manufacturing the vaccine, was socially contingent. It was a product of the political and social understandings of its time, characterized by trust in both government and the private sector. The jury’s letter to the judge illustrates that to them biological cause was not enough, although the law permitted liability under a warranty theory. Instead, these jurors believed that the company should have been at fault in order to be liable. The second story is about a series of trials involving the anti-nausea drug Bendectin, one of the early multi-district products liability litigations. Bendectin was prescribed to Fact, 9 Stan. L. Rev. 60 (1956) (criticizing claims that cause in fact can be definitively determined); Guido Calabresi, Concerning Cause and the Law of Torts: An Article for Harry Kalven, Jr., 43 U. Chi. L. Rev. 69 (1975); Steven Shavell, An Analysis of Causation and the Scope of Liability in the Law of Torts, 9 J. Legal Stud. 463 (1980) (arguing that where causation is uncertain, the determination of causation should be made based on economic analysis and providing such an analysis).

pregnant women and was alleged to cause birth defects. In the course of the litigation, after some liability determinations, more rigorous studies showed that Bendectin was not a teratogen. Because the company was initially found to be liable on the basis of preliminary epidemiologic testimony that showed a weak causal link, the Bendectin story has been often used as a stand in for the concept of “junk science.” But, in fact, liability in the Bendectin cases was inextricably linked with evidence presented about the defendant’s conduct, including evidence that although it was the same company that had attempted to sell Thalidomide in the United States, it had not tested Bendectin to determine whether it too was a teratogen. That testing came only after the fact. The Bendectin litigation was also a case of political and social understandings influencing liability determinations, although this time some juries did find fault. Rather than being understood as a case of “junk” science, the Bendectin litigation should be reframed as an object lesson on the linkage between science and social mores in the development of products liability. These two cases, which came out different ways, are both illustrations of how science cannot be separated from the social and political context in which it is produced, at least in the case of medical and epidemiologic science. Attempting to separate science and conduct at trial produces narratives that miss important threads. Yet it is tempting to elevate science above the law, to consider science more objective in opposition to the subjective and unpredictable determinations of juries.6 The judicial approach of the 1980s was rooted in the idea that where the courts had failed, epidemiologists would succeed.7 As it turns out, misconduct matters to adjudicators. It matters in evaluating scientific evidence whether a study was ghost written, whether scientists reran the data until they found the result they were looking for, and whether they suppressed data because it showed things they did not want to find.8 A. An Overview of the Chronology Liability Standards and Epidemiology This part presents an overview of the evolution of liability standards and of epidemiology. The law tends to be behind science and the case of epidemiology was no different. Interestingly, the relationship between law and science was inverse: the more probabilistic epidemiology became, the more law insisted on determinism. Strict products liability followed closely after the development of epidemiology as an academic and scientific discipline in the mid-twentieth century. These two developments 6 One study of bifurcation of punitive damages found that “[w]ith respect to the compensatory liability decision, bifurcation

significantly reduced the percentage of those willing to find the defendant liable (55.2% versus 42.8%).” Stephan Landsman, Shari Diamond, Linda Dimitropoulos, Michael J. Saks, Be Careful What You Wish for: The Paradoxical Effects of Bifurcating Claims for Punitive Damages, 1998 Wis. L. Rev. 297, 334 (1998). This study also found that when defendants were found liable, jurors under the bifurcated trial condition awarded punitive damages more frequently and in higher amounts than jurors in the unitary trial condition. Id. at 335. Another study found that defendants were more likely to prevail when causation was separated from liability, but that damages were higher in bifurcated trials when defendants were found liable. Kenneth S. Bordens & Irwin A. Horowitz, Mass Tort Civil Litigation: The Impact of Procedural Changes on Jury Decisions, 73 Judicature 22, 26 (1989). 7 This is a continuation of the New Deal ideology of expertise. As Rebecca Roiphe argues: “Ultimately . . . experts would do what, in the minds of legal realists, the courts had failed so miserably to accomplish. Not only would expertise solve the intricate problems of modern industrial America, it would police the boundaries of state power while ensuring the welfare of all.” Rebecca Roiphe, The Most Dangerous Profession, 39 Conn. L. Rev. 603, 627 (2006). 8 Wendy Wagner, When A Corporation's Deliberate Ignorance Causes Harm: Charting A New Role for Tort Law, 72 DePaul L. Rev. 413, 436 (2023).

are generally understood to have no relationship other than the fact that later in time lawyers came to use epidemiologic evidence in trials. But strict liability can be understood to be a doctrine that relies on the same population level understanding of causation as epidemiology because it relies on the idea that when industry imposes costs on a population, it should be required either to internalize those costs or serve as a kind of social insurer by spreading them across the affected population. The more the doctrine of strict liability took hold the more pressure was put on epidemiologic science to provide certainty, ultimately resulting in the categorical separation of causation from misconduct. There were, of course, other developments in the 1980s that influenced these events, and in particular that spurred tort reform. These include the rise of deregulation, the increased concern that the tort system was suppressing innovation, and the concern that American companies were not competitive with foreign counterparts.9 The pre-history of strict products liability The history of the development of liability for dangerous products in the early twentieth century is substantially a story of exploding bottles. From the early 1900s through the 1950s, plaintiff victims of exploding soda bottles often sued. Sometimes courts found the manufacturer could be held liable even when a plaintiff could not show how the manufacturer was at fault.10 Other times courts held that manufacturers had a duty to test their bottles.11 Overall, courts began to recognize that sometimes a bottle, although it had been part of the same manufacturing process as all the other bottles, was just faulty and had the propensity to explode. As the North Carolina Supreme Court announced in 1918: “if these goods are so inherently dangerous from their frequent explosion and liability to cause damage, as by putting out the eye of the plaintiff, that they cannot be made safe, then placing them upon the market is indictable, as well as makes the manufacturers and all vendors liable to actions for any damage accruing.”12 The most famous of the bottle cases is, of course, Escola v. Coca-Cola.13 In that 1944 case Justice Traynor announced a theory of strict liability in a concurrence that became the basis for the Restatement (Second) of Products Liability, spread like wildfire across the country and has remained a subject of debate ever since. But by the time the famous concurrence was penned, it had already been the case that courts for years had seen exploding bottles as a problem for negligence law and had moved towards liability without fault. Escola is often considered the beginning of the strict liability revolution, although it is more aptly described as the best statement of where the law had already been going for

9 See Robert L. Rabin, The Torts History Scholarship of Gary Schwartz: A Commentary, 50 UCLA L. Rev. 461, 476 (2002). 10 See, e.g., Dail v. Taylor, 151 N.C. 284, 66 S.E. 135, 137 (1909).

In Dail a Coca-Cola bottle broke and injured a young girl. A witness testified that “that all along for the last two years witness had seen these bottles from defendant's works explode in the store.” Id. While the court rejected the idea that the burden should be shifted to the defendant under a res ipsa loquitor theory, it held that this was sufficient evidence for the case to go to the jury on a negligence theory. But see Glaser v. Seitz, 35 Misc. 341, 341, 71 N.Y.S. 942, 943 (Sup. Ct. 1901) (“There is no evidence in this case that the bottle, which was not manufactured, but filled, by the defendant, was not properly constructed, or that it was constructed differently from bottles in which seltzer water is usually sold.”). 11 Torgesen v. Schultz, 192 N.Y. 156, 161, 84 N.E. 956, 958 (1908) (jury could find defendant liable for inadequately testing bottles under pressure so that they were liable to explode). 12 Grant v. Graham Chero-Cola Bottling Co., 176 N.C. 256, 97 S.E. 27, 29 (1918). 13 Escola v. Coca Cola Bottling Co. of Fresno, 24 Cal. 2d 453, 150 P.2d 436 (1944)

many years before it was decided.14 State courts had for a long time recognized that the burden of accidents should rest of bottle manufacturers by the time the Escola decision was written. Scholars advocated strict products liability starting about ten years later.15 In some ways those bottles presented an easy causation problem: of course the exploding bottle had caused the injury in the minimal sense. But did the manufacturer do, or fail to do, anything that caused that explosion? Courts increasingly answered that plaintiff need not show what the defendant did, based on a variety of legal theories that all bottomed out on some version of responsibility for inevitable accidents.16 As that North Carolina court explained: It may be that the defendant could have used wicker covering for the bottles such as is used for champagne bottles, or wire mesh cases as is used for certain goods of explosive nature. These would not prevent explosions, but would prevent the fragments of the glass doing much damage. Or the goods might be packed in sawdust, as is done with some goods such as aerated water liable to explosion. Or there might be some harmless ingredient put in the decoction to prevent sudden expansion causing explosions--a device that is not unusual. Or thicker bottles might be used, or there may be still other devices in this age in which “men have sought out many inventions.” Ecclesiastes, c. vii, v. 28. But what is the best protection is one which the defendant must ascertain and use.17 The science of bottle explosions was like the science of epidemiology in the sense that it was probabilistic and population-based. In a population using soda bottles, some number would suffer injuries as a result of a bottle explosion. This was a certainty. But who would be injured was unknown nor could it be known why precisely this bottle injured them when so many others did not explode. The pre-history of epidemiology The 1940s, when Escola was decided, were a critical period for the development of epidemiology as well. The science of public health was about to experience a massive breakthrough with the publication of two large studies in 1950 linking smoking to lung cancer.18 These would present new problems for law once lawsuits against the tobacco companies started to be brought. In potted histories of epidemiology, the next greatest moment in population health sciences after the tobacco studies was the testing of the Salk vaccine in 1954. A few other key moments in the middle of the twentieth century stand out: the advent of the first textbook on study design in 196019 and the landmark Surgeon General’s Report on Smoking 14 Roscoe Pound, The Problem of the Exploding Bottle, 40 BU L Rev. 167, 169-172 (1960). 15 Fleming James, Jr., Products Liability, 34 Tex. L. Rev. 192, 228 (1955); ALBERT A EHRENZWEIG, NEGLIGENCE WITHOUT

FAULT 81 (1951). 16 For a discussion of the different theories see Pound, The Problem of the Exploding Bottle, supra note __ at 169-171. 17 Grant v. Graham Chero-Cola Bottling Co., 176 N.C. 256, 97 S.E. 27, 28 (1918) 18 Wynder and Graham, Tobacco smoking as a possible etiologic factor in bronchogenic carcinoma: a study of six hundred and eighty-four proved cases, 143 JAMA 329-336 (1950); R Doll and A Hill, Smoking and carcinoma of the lung: preliminary report, 2 British Med. J. 739-748 (1950). 19 Brian MacMahon, Epidemiologic Methods (1960).

and Health published in 1964,20 which established standards for demonstrating causation along with the Bradford-Hill criteria first published in 1965.21 That same year, section 405(A) of the Restatement (Second) of Torts was published, ushering in the era of strict products liability. Epidemiology after 1950 had two strands: studies of infectious diseases and studies of chronic diseases.22 The biggest struggle for epidemiologists is identifying causal factors. In the 1950s and 1960s, methods for establishing causation of chronic diseases (such as a link between smoking and lung cancer) were debated. By contrast, establishing causation of acute viral or bacterial diseases, such as polio, was relatively easy. By the 1970s, epidemiologists had begun to debate more rigorously how precisely causation would be established, with the first book on causality in population health sciences published in 1973.23 These epidemiologic arguments about causality in chronic or long-latency diseases found their way into courtrooms in the 1980s. Yet as in epidemiology, where broader ecological or environmental factors have to be addressed even if they complicate the picture, the environmental, social and political factors that contributed to assessments of fault remained salient. To deal with this tension between the idea of scientific objectivity and the role of social context in the production of science, judges innovated by separating out the causal inquiry from the rest of the evidence presented at trial through bi- or trifurcation. The narrative of “junk science” and an increased hope (at odds with scientific possibility) of deterministic causal findings justified the separation. The intersection of epidemiology and law The remainder of this essay considers two events in the history of the intersection between epidemiology and law: the Cutter incident and the Bendectin litigation. The Cutter incident epitomizes the easy case of causation as we shall see. It hails from the era of epidemiology as a study of infectious diseases and the period that ushered in strict products liability. Bendectin represents the hard case for causation. It required careful epidemiologic study to determine whether the product was a teratogen. In both cases the relationship between causation and blame was complicated by the social context in which both the injury and the litigation occurred, and legal histories have yet to fully grapple with the ways these complexities played out and affected the law going forward. As we shall see, as epidemiologic science developed more sophisticated methods for demonstrating causation, the pressure to de-link narratives of misconduct and causation rose. This strategy was available due to the structure of strict liability which does not require any showing of fault. The strategy was successful because it fit into a narrative of juror naiveté, a belief that juries would believe “junk science” in cases where verdicts where U.S. Dep’t of Health, Education and Welfare, Smoking and Health: Report of the Advisory Committee to the Surgeon General of the Health Service (1964). 21 Austin Bradford Hill, The Environment and Disease: Association or Causation?, 58 Proceedings of the Royal Society of Medicine, 295-300 (1965). 22 See Louis H. Kuller, Epidemiology: Then and Now, 183 American Journal of Epidemiology, 372–380 (2016). 23 See MERVYN SUSSER, CAUSAL THINKING IN THE HEALTH SCIENCES (1973). See also J.S. Kaufman and C. Poole, Looking Back on ‘Causal Thinking in the Health Sciences,’ 21 Annual Rev. of Public Health 101-119 (2000). Kaufman and Poole point out that while there have been refinements, the terms of the debate were set in the 1960s. 20

the verdict was explained not by the mistaken belief in junk science but instead in a moral reaction to the conduct that gave rise to the harm. B. The Salk Vaccine and A Hidden History of Misconduct The key thing to know about the Salk vaccine was that it involved killing a very virulent strain of the polio virus. The exposure to the killed virus created immunity, and it also carried a very high risk if the virus was not killed. Killing the virus required a meticulous process. But some of the private companies that manufactured the vaccine, and in particular Cutter, a small company located in Oakland, California, were not equipped to follow this process as they should have. The rollout of the polio vaccine should have been the FDA’s proudest moment. Every spring and summer, thousands of children had died or been paralyzed. Parents lived in fear. When it seemed that the hunt for a vaccine to prevent polio was over, the country breathed a sigh of relief. The dread that families experienced every spring — the empty pools, beaches, and movie theaters — would end. Their children would be safe. The vaccine was famously developed by Jonas Salk and funded by the National Foundation for Infant Paralysis. The National Foundation funded infant care, but a great deal of its money was invested in developing a vaccine. As David Oshinsky describes it in his Pulitzer Prize winning book, Polio: An American Story, there was a kind of race between scientists to produce a vaccine. One group, led by Salk and based on research by Isabel Morgan, wanted to produce a vaccine by killing the virus. This “killed vaccine” used formaldehyde to destroy a virulent strain of the virus, so that it would be inert when it entered the body but would still produce antibodies that would protect children from infection. Another group, whose chief advocate was Albert Sabin, wanted to develop a “livevirus” vaccine, using weakened strains of the polio virus. The National Foundation generously funded the labs of scientists on both sides of this divide, including Salk and Sabin, with the goal of obtaining a vaccine as soon as possible. The trouble with the killed-virus approach was that scientists needed to kill all the virus, otherwise it would paralyze the patient, or worse, and there was some question of how long immunity from that vaccine would last. The trouble with the live-virus approach was that scientists had to develop a mild enough strain that would produce an immune reaction without infection, and that strain had to remain mild even as it reproduced, so that others wouldn’t get the virulent form of the disease. It turned out that the process of creating a strain that started out and stayed sufficiently weak took a very long time, whereas it was much quicker to produce a killed virus vaccine. The National Foundation privileged speed, so it bet on the killed virus, despite the concerns of the live-virus scientists that it would not be as effective and might be more dangerous than a live-virus vaccine. Salk had won the race in the public eye. His vaccine was ready for testing in 1953 and the National Foundation oversaw a massive testing project that year. The process of testing the vaccine was rife with ethical violations. Even at the time, the other scientists were appalled that Jonas Salk tested his vaccine on mentally disabled 7

children at the Polk School for the Retarded and the Feeble-Minded, where many of the children could not consent because of their disabilities and lacked parents or guardians who could consent to participation in the experiment on their behalf.24 But these secret tests were a success, demonstrating both that the killed-virus vaccine was safe and that it stimulated a high antibody response. Salk and the National Foundation were ready to move on to national testing with an eye towards national vaccination. Other virologists, however, were still not so sure. They worried about the fact that Salk’s vaccine used the most virulent polio virus — known as the Mahoney strain — and they worried about the inactivation process.25 If all the virus wasn’t killed, the vaccine would infect children with polio instead of preventing the disease. The process of killing the virus that Salk was in the process of perfecting, even as the national testing project got underway, involved multiple steps. First, the virus had to be grown using monkey cells. Then it had to be filtered through exceedingly fine filters so nothing but the virus remained. Any cells or cellular debris left over from the production process could result in the virus being able to “hide” behind them, and evade the formaldehyde solution that was the final, and critical, step in production because it killed the virus. (You may recall from biology class that cells are hundreds of times larger than viruses, which is why a cell might shield the virus from formaldehyde.) How could Salk know that the virus was completely inactivated, even when producing millions of doses of vaccine? He developed a theory called the “straight line theory.” As Paul Offit, who wrote the definitive book on the events that happened next, explains, the straight line theory involved plotting the amount of virus against the length of time the virus was treated with formaldehyde. The theory was that if the line remained straight, the virus would be fully inactivated. Salk believed that the line would remain straight, but other scientists did not. They worried that the line might “curve backwards to baseline” in which case live virus might be present in the vaccine. The straight line theory was a matter of faith.26 Even as the National Foundation was testing the vaccine on hundreds of thousands of American children, scientists were disputing how long it would take to inactivate the virus. A highly regarded virologist from Sweden (and a member of the Nobel Prize committee) named Sven Gard warned that he was unable to replicate Salk’s straight line in his lab. Whereas Salk’s theory predicted that the virus would be killed in nine days, and he was able to achieve this result, it took the scientists in Gard’s lab twelve weeks to kill all the virus. Gard believed that the straight line theory must be wrong and that as the virus was treated with formaldehyde for longer periods, the line started to curve backwards, leaving live virus in the vaccine. He wasn’t the only one. A year earlier, a doctor at a Chicago hospital had reported similar problems.27 Still, the National Foundation’s field trial went forward using Salk’s method. Who was to make the vaccine? The consensus at the National Foundation was that it 24 Polio: An American Story at 157-58. 25 Polio: An American Story at 166-68, 182. 26 Offit, The Cutter Incident, at 41-42. 27 Offit, The Cutter Incident, at 42-43.

should be private industry, rather than the government, that would make the vaccine for the trials and ultimately for the general public if the trials were successful. The idea was that the companies would provide the vaccine for free for the field trial involving hundreds of thousands of child volunteers. Afterwards, they would be given the right to sell the vaccine at a profit. This was part of the cold war mentality of the time, a desire to avoid the step towards “socialized medicine” that government production and distribution would create. It marked the beginning of the fraught relationship between industry, regulators and courts going forward. As companies were vetted by the National Foundation in late 1953 and early 1954, Salk didn’t have a precise production method to give them. Executives and scientists inside the companies worried that despite all the talk, Salk didn’t have a vaccine that was ready for the rollout that the National Foundation was promising. Four companies had nevertheless signed on to produce the vaccine by February 1954: Cutter Laboratories, Eli Lilly, ParkeDavis, Pitman-Moore, and Wyeth Laboratories.28 That same month, Salk put together a fifty five page document detailing how to manufacture the vaccine, but as Paul Offit writes, many details were missing, such as “the type of filter to be used prior to inactivation, the maximum quantity of virus to be inactivated, and the exact intervals during inactivation when samples should be tested.”29 It seems Salk thought the companies should adapt his theories to their own processes, that he conceived of inactivation as more an art than a science. He likened the method to developing a recipe for cake. “She starts with an idea and certain ingredients, then experiments, a little more of this and a little less of that, and keeps changing things until she finally gets a good recipe,” he explained.30 What Salk seems to have missed, perhaps because he didn’t cook, is that baking needs to be precise. And novice bakers follow recipes to the letter, otherwise their cake might not rise. It turns out that is what the pharmaceutical companies wanted too, but they did not get it. As one researcher in a commercial laboratory put it “every batch is a damned research project.”31 In preparation for the field trial, every lot of vaccine was tested by the manufacturers, Salk’s laboratory, and the National Institutes of Health Laboratory of Biologics Control, which would ultimately approve the vaccine if the trials were successful. There were problems from the beginning. The companies were having trouble reproducing the vaccine. Some lots of the vaccine contained live virus. One of the companies’ vaccines (Cutter) did not contain all the types of virus that were required for maximum immunity. The head of the Laboratory of Biologics Control at the NIH, who was in charge of approving the vaccine, wrote to Salk about these problems. Ultimately, it was decided that in order for the vaccine to be accepted, companies “had to make eleven consecutive lots of vaccine that passed safety tests.”32 And that was how they proceeded. The trials were a success. At least, they were successful enough to move to the next level of immunization of children across the United States. In April 1955 the NIH approved the vaccine. The five companies were ready to make the vaccine, this time at a profit. 28 Offit, The Cutter Incident at 44-45 29 Offit, The Cutter Incident at 47. 30 Offit, The Cutter Incident at 47. 31 Offit, The Cutter Incident at 47. 32 Offit, The Cutter Incident at 48-49.

But there was a terrible omission in the approval process. A scientist at the Laboratory of Biologics Control, Bernice Eddy, tested lots of the vaccine that the five companies who were to be licensed to make it had provided several months earlier. Three of the six lots — fully half — submitted by Cutter Laboratories contained live polio virus. When Eddy tested them on monkeys, they developed polio. At that point it should have been evident that there was something very wrong with the process Cutter was using to make the vaccine, but William Workman, the head of the Laboratory, never told the committee. Eddy predicted there would be a terrible disaster.33 She was right. The problems that had plagued production continued, in part because the government’s conditions were looser than those imposed in the trials. Recall that Salk and the National Foundation had produced a fifty-five page how-to manual about his vaccine for the manufacturers in the trial, but the regulators only provided a five page document of “minimum requirements” for producing the vaccine.34 The National Foundation, at the behest of government scientists, had required eleven consecutive lots of vaccine that were safe before a company could distribute it for the trials. Upon approval, the government did not impose such a specific requirement, but only the general requirement that the method used result in a consistently effective and reliable vaccine. If any lots included live virus, these lots did not need to be reported to the regulator. The idea was the companies might produce some lots with live virus, but if they caught those dangerous vaccines there would be no problem.35 The companies had every incentive to produce a vaccine free of live virus. Cutter Laboratories may have had every incentive to produce a safe vaccine — they even gave it to the children of their own scientists and other employees — but that did not mean they had the ability to do so. As production was heating up, in April 1955, Cutter asked one of the members of Salk’s lab who had played a key role in developing the vaccine to come by while he was in San Francisco. They were having trouble producing a consistent and reliable killed virus. When that scientist, Julius Youngner, toured the facility he found all kinds of problems, from sloppy records to bad storage practices. Years later, when he was interviewed about the experience he said “They looked like they didn’t know what the hell they were doing.”36 Youngner went back to the lab and told Salk. He volunteered to write letters to the Laboratory of Biologics Control and to the National Foundation, but Salk said he would do it. So Youngner never wrote the letter, and apparently neither did Salk. This information did not come out until more than fifty years later, when Paul Offit interviewed Youngner for his book about the Cutter incident.37 It is possible that Youngner’s memory was colored by what occurred later, but his recollection is important because it shows that even at that early point, people who understood what was required to make a killed vaccine were worried about Cutter Laboratories. And they had good reason to worry. Cutter wasn’t able to reliably produce a killed vaccine, but they produced one anyway. By April 26, 1955, Workman, the head of the Laboratory of Biologics, started getting 33 Offit, The Cutter Incident at 62-63 34 Offit, The Cutter Incident at 60 35 Offit, The Cutter Incident at 61 36 Offit, The Cutter Incident at 65. 37 Offit, The Cutter incident at 65.

calls reporting that children who had been given the polio vaccine were developing polio.38 The progression of the disease was not characteristic; it started in the arms where they had received their shot. Despite the tests and the promises, they had been given a vaccine with a live, virulent strain of polio, not an inactivated one. Everything Cutter reported to the government seemed in order, but they didn’t report the lots that had failed their internal safety tests. Because nobody knew how bad Cutter’s processes were, they didn’t know whether the problem was the Cutter vaccine or Salk’s method of inactivation. It would take years to figure out precisely what had gone wrong. In the meantime, Workman wrote to Cutter and asked them to recall the vaccine. They did. But not before three hundred and eighty thousand children had been given the Cutter vaccine. Of those, approximately two hundred and twenty thousand were infected with a live vaccine, seventy thousand developed muscle weakness, one hundred and sixty four were paralyzed, and ten died.39 Cutter Laboratories wasn’t the only one to fail, but it failed most tragically and spectacularly. It turned out the Wyeth Laboratories had also had problems inactiviating the virus. Its vaccine also killed and paralyzed people, although many fewer than Cutter’s. But only a small number of people knew about Wyeth’s problems.40 It was Cutter that became infamous for the tragedy that killed children with a vaccine that was supposed to protect them. There were a number of problems with Cutter’s manufacturing process, some of which were spotted by Salk’s colleague Julius Youngner when he visited the lab. But he did not know, and indeed couldn’t have known, of the main problem. In order for the formaldehyde to inactivate the virus, all cell and cell debris had to be removed. Salk did this with a special kind of filter called a Seitz filter, which was very effective. Cutter and Wyeth did this with glass filters, which were not consistent in their ability to filter out fine debris. The glass filters’ efficacy depended on how they were made. It is likely that the virus hid behind the cell debris in some of these lots filtered through glass filters, never being fully inactivated, waiting for its victims. But the filter itself does not explain the difference between Cutter’s deadly vaccine and Wyeth’s. Both companies had problems, but Cutter’s were tenfold worse. Why was that? Part of it was likely luck. Cutter’s glass filters were especially bad. But Cutter also made other errors. After filtering, it waited to inactivate the virus for longer periods than other companies, potentially allowing the virus to multiply and making it harder to kill.41 And it never figured out how to operationalize Salk’s straight line theory. As Paul Offit explains “Cutter Laboratories never determined when live virus was first eliminated and, therefore, couldn’t determine how long to treat with formaldehyde.”42 They could have consulted other polio researchers, or Salk himself, but other than reaching out briefly to Youngner, they didn’t, despite the fact that it was evident they were never able to reproduce Salk’s inactivation results. Offit concludes: “No company showed greater disregard for Salk and

38 Offit, The Cutter incident at 67 39 Offit, the Cutter Incident at 89 40 Offit, the Cutter incident at 102-103. 41 Offit, The Cutter Incident at 110 42 Id.

his theories than Cutter Laboratories.”43 The thing is, Cutter complied with every requirement that the government had placed on vaccine manufacturers. After the Cutter Incident, the government required that filtered virus sit no longer than 72 hours before being inactivated. After the incident, the government required better filters and mandated reporting of lots tested that contained live virus. But all these additional regulations were added after the incident. This raised a question that the law was ultimately required to answer: what was Cutter’s responsibility to the children who were given its vaccine? What role should the fact that Cutter complied with all the government’s requirements play in that determination? These questions would be litigated at trial in a case brought by the parents of Ann Gottsdanker against Cutter Laboratories. Wyeth, which had less exposure because its vaccine caused fewer injuries, settled all its cases out of court. But Cutter decided to go to trial on the claims against it. Gottsdanker was represented by one of the most famous tort lawyers of the age, Melvin Belli. The Gottsdanker case had all the features that would characterize modern products liability suits. The scope of harm was wide, threatening to drive Cutter into bankruptcy (although the company emerged relatively unscathed). The scientific issues were complex, requiring virologists to testify about the validity of the virus inactivation method and threatening to confuse the judge and jury. The plaintiff’s lawyers were well capitalized. And the interplay between government regulation and private industry in the manufacture and dissemination of products at once useful and dangerous forced a reckoning over responsibility. The polio vaccine had been developed and tested by reputable scientists. It was regulated by the National Institutes of Health Laboratory of Biologics Control. There was a set of rules about how to make it and Cutter had followed all of them. From Cutter’s perspective, why should it be held responsible for mistakes the federal government had made, especially since it was not the only company whose vaccine contained live virus, even if it was the worst offender? One of the defense lawyers for Cutter told Paul Offit years later that the lawyers had thought of suing the United States Government or even Jonas Salk for their regulatory failures. Suing the government would “give the jury someone else to look at as the responsible party,” he explained. “There was also some thought to [suing] Dr. Salk, but Dr. Salk was such a cult figure, he was such a hero at that time, that I thought it would be counterproductive.” Then he went on: “I wish we had done it.”44 Cutter didn’t need to sue the government or Salk because they had what lawyers call an empty chair defense. They could blame absent parties: the government for its five pages of insufficient and flawed guidance and Dr. Salk for his inaccurate straight line theory and incomplete directions on how to inactivate the virus. It might be better, even, if the true culprits were not there to defend themselves. Salk was there, however, testifying for the 43 Offit, The Cutter Incident at 112 44 Offit, Cutter Incident at 144-45

plaintiff that it was Cutter’s decisions that were flawed, not his methodology. But counterexperts questioned Salk’s straight line theory, which at that time had not yet been proven, although he would eventually be vindicated. These experts thought that killed virus vaccines were a mistake to begin with, and among them were Nobel laureates. At that time Cutter couldn’t have sued the government for its role in the debacle. For most of United States history, one couldn’t sue the federal government or federal officers for personal injury because the doctrine of sovereign immunity protected the government from suits. The only way for a victim of government negligence to sue was to get a special dispensation from Congress, through a private bill.45 In 1946, Congress passed the Federal Tort Claims Act (FTCA) which allowed people to sue the federal government or officers for negligence. This would not, however, have allowed Cutter to sue the United States. The statute expressly prohibited (and prohibits to this day) lawsuits against the government when the government is performing a “discretionary function.”46 What is a discretionary function? In 1953 the Supreme Court defined it as “determinations made by executives or administrators in establishing plans, specifications or schedules of operations.”47 Once a plan was made, and the discretion to make policy exercised, the implementation of policy might be subject to suit. So, for example, in 1955 the Supreme Court held that a person could sue the government for negligently maintaining a lighthouse. The decision to build and maintain a lighthouse in the first place was discretionary and could not be grounds for a lawsuit. But once built, a failure to maintain that lighthouse in good working order could be.48 If Cutter had sued the United States, the question for the court would have been: was the approval of the Salk vaccine more like the decision to build a lighthouse or the maintenance of one? Nearly thirty years later a family brought just such a lawsuit against the federal government. Kevan Berkovitz received the oral polio vaccine (which contained a live weakened virus) when he was two months old. It was widely known that with a live virus vaccine some minority of people contract polio. By that time, having had a long experience with polio vaccines, the Bureau of Biologics (as the Laboratory of Biologics Control was later called) had a number of specific regulations that vaccine manufacturers were required to meet. For example, manufacturers were required to submit test data about lots of vaccine for review by the regulator. In Berkovitz’ case, the Bureau had not reviewed Lederle Laboratories’ test data before approving the lot of vaccine that paralyzed Kevan, as was required by law.49 The Supreme Court held that the government could be sued for the 45 United States v. Testan, 424 U.S. 392, 399 (1976) (quoting United States v. Sherwood, 312 U.S. 584, 587-88 (1941)) (“Thus,

except as Congress has consented to a cause of action against the United States, ‘there is no jurisdiction ... in any ... court to entertain suits against the United States.”’); United States v. McLemore, 45 U.S. 286, 288 (1846) (“[T]he government is not liable to be sued, except with its own consent, given by law.”); see also Cohens v. Virginia, 19 U.S. 264, 411-12 (1821) (“The universally received opinion is, that no suit can be commenced or prosecuted against the United States; that the judiciary act does not authorize such suits.”). 46 See 28 U.S.C. § 2680(a) (stating that the FTCA’s waiver of sovereign immunity “shall not apply to . . . [a]ny claim based upon an act or omission of an employee of the Government, exercising due care, in the execution of a statute or regulation, whether or not such statute or regulation be valid, or based upon the exercise or performance or the failure to exercise or perform a discretionary function or duty on the part of a federal agency or an employee of the Government, whether or not the discretion involved be abused”). 47 Dalehite v. United States, 346 U.S. 15, 35–36, 73 S. Ct. 956, 968, 97 L. Ed. 1427 (1953) 48 Indian Towing Co. v. United States, 350 U.S. 61, 67, 76 S. Ct. 122, 126, 100 L. Ed. 48 (1955) 49 Berkovitz by Berkovitz v. U.S., 486 U.S. 531, 543, 108 S. Ct. 1954, 1962(1988)

failure to comply with its own regulations, as this was the functional equivalent of maintaining a lighthouse. But, the Court explained, “the discretionary function exception bars any claims that challenge the Bureau's formulation of policy as to the appropriate way in which to regulate the release of vaccine lots.”50 Had Cutter tried to litigate this defense (and had it existed in the 1950s), the problem would have been that it was the NIH’s decision to impose minimal regulations, not its failure to require compliance with those regulations, that laid the groundwork for the debacle. The only possible theory for Cutter to argue that the government was negligent in following through on its own regulations would have been that when Bernice Eddy discovered that three out of the six lots Cutter submitted for review contained live virus, this defect should have doomed their vaccine. But there were no regulations in place that required that every lot, or some number of lots, be inactivated — the decision to move forward was at the government’s discretion. And the government decided that so long as the lots of live virus were discarded, this risk as acceptable. That was a policy decision protected from suit. So, too, the five-page manual on how to inactivate the virus, with its inadequate description, was akin to the decision to build a lighthouse rather than its maintenance. No regulation required Cutter to submit the number of lots that it was making which failed its internal tests for a live virus. When the government didn’t request those data, the government was not in violation of any regulation. From a moral perspective, the government was to some extent at fault. In retrospect, it should have required more and different testing. When Bernice Eddy raised the alarm that half of Cutter’s lots contained live virus, the government should have denied the company a license to produce the vaccine. Further, the NIH had the model of the requirements of the National Foundation, which had mandated that eleven consecutive lots pass testing for live viruses and provided more detailed directions on how to manufacture the vaccine. But instead of using those, it provided much looser guidance, counting on industry to develop its own, perhaps better, methods of production. A key problem was that everyone — the government and Salk included — trusted industry to be able to produce the vaccine at scale using its own scientific acumen. That trust was profoundly misplaced. Ironically, despite talk of suing the government for inadequate regulations, Cutter’s best defense was that it followed those regulations to the letter. The company argued that since it had met government requirements, however minimal, it was free from fault. In the language of law, the company argued that its duty was to comply with government regulations of vaccine production and it had complied with that duty. Negligence is about failing to take the right amount of care. If the right amount of care is whatever the government says it ought to be, Cutter had done that. To the extent that some lots contained live virus, this was beyond their capability to prevent, they argued. It helped that they had evidence that Wyeth Laboratories also produced some lots containing live vaccine that injured children, albeit at a much lower rate than Cutter had. The idea was a familiar one from the bottle cases: accidents will happen. And the argument that Cutter

50 Berkovitz, 108 S. Ct. at 1964

made was that the harm should lie where it falls,51 in this case on the child and her family. The problem was that courts had been holding for years in bottle cases that although accidents will happen, the obligation to either prevent the harm or pay for the ensuing loss lay on the manufacturer not the consumer. Melvin Belli, Gottsdanker’s lawyer, didn’t have access to all the information about what happened at Cutter. The straight-line theory was still controversial. He didn’t know that Wyeth had also had difficulties inactivating the virus or the difference in magnitude of Wyeth’s mistake and Cutter’s, or the fact that a key problem (albeit not the only problem) was the inadequate glass filters, or that companies that had better practices did not have live virus in their vaccines. Perhaps as a result of this, or perhaps because they thought that compliance with government regulations should be enough, the jury ruled for Cutter on the negligence claim. But there was a second claim in the lawsuit as well. The plaintiff argued that Cutter, in selling the vaccine, had impliedly warrantied it to be safe. In contract, a warranty is a promise by the seller that the product is what it purports to be. A warranty in contract is express, that is, it is a promise that is expressly made by the manufacturer, usually in writing. Today warranties are usually included in small appliances and other products on a piece of paper or electronically. But a warranty can also be implied, which is to say imposed by the court rather than promised by the manufacturer expressly. Implied warranties for defective products are considered tort causes of action. The importance of this is that unlike in contract, where the two parties to the contract enter into some kind of agreement, in tort the promise extends to any user, even one who didn’t obtain the vaccine directly from the manufacturer. After all, it was doctors who purchased the vaccine for sale to patients. Cutter had put what might be described as an express warranty, a written promise, as a label on the vaccine it sold. “This vaccine is prepared in accordance with the requirements of the National Institutes of Health of the U. S. Public Health Service.” It read. “Local and other untoward reactions have been very minimal using this material.”52 But, as the California appellate court pointed out, Cutter’s warning labels also explained that the vaccine was inactivated with formaldehyde. They “affirmatively represent that the virus has been inactivated, and thus are completely consistent with the implied warranties here in issue.”53 Cutter had represented that the virus was inactivated, and it would be held to that promise. The court and jury agreed that Cutter breached that implied warranty when it produced a vaccine that infected patients with the disease it was meant to prevent. Usually, when a jury decides a case, it does not provide reasoning for its decision. But in the Gottsdanker case the jury gave the judge a note along with its verdict. In that note the foreman of the jury explained that all of the jurors but one (the verdict did not need to 51 This idea has a long history. It was most famously stated by the future Justice Holmes: “[t]he general principle of our law

is that loss from accident must lie where it falls.” Oliver Wendell Holmes, Jr., The Common Law 94 (Boston, Little, Brown & Co., 1923) (1881). As I have demonstrated in other work, this was Justice Holme’s idea, not a reflection of the state of the law during the period he was writing, at least with respect to liability for injurious products. See Alexandra D. Lahav, A Revisionist History of Products Liability, 122 Mich. L. Rev. 509, 559 (2023). 52 Gottsdanker v. Cutter Lab’s, 182 Cal. App. 2d 602, 610, 6 Cal. Rptr. 320, 325 (Ct. App. 1960) 53 Gottsdanker, 182 Cal. App. 2d at 611

be unanimous under California law) agreed that Cutter was not negligent. But they “felt constrained” to rule against Cutter on the warranty issue. It seems that the jury was reluctant to hold Cutter liable at all, but since the judge had told the jury that Cutter had promised an inactivated vaccine, and the facts showed it had delivered a live one, the jurors felt they had to find Cutter liable despite their misgivings. This protest from the jurors illustrates the power of Cutter’s arguments at trial that they had done all that they could have done, that this was a terrible accident but not their fault, and that for this reason, the company should not be held responsible. They couldn’t have improved their process, the argument went, and they shouldn’t be required to do the impossible. Cutter had a strong policy argument as well: “that development of medicines will be retarded if manufacturers are held to strict liability for their defects.”54 The courts rejected this argument on the grounds that it was acceptable to limit liability in cases where the medicine was ineffective, or, in this case, if the vaccine had not prevented polio, but quite another thing when it came to an “assurance that the product will not actively cause the very disease it was designed to prevent.”55 Commentators at the time agreed with Cutter. A comment in the Stanford Law Review described the situation this way: Cutter was undoubtedly aware that negligently produced vaccine might cause extensive liability. It closely followed the Salk process and the government's standardized tests designed to exclude active virus from the marketable vaccine. These standardized precautionary measures were not found inadequate until after the Cutter vaccine was withdrawn from the market. Therefore it is unlikely that Cutter could have used greater care and still have released its vaccine at the time authorized by the government.56 The strange thing about the Cutter case is that historical investigation has proven that it was, in fact, a case of negligence. The jury was wrong because it lacked all the information it would have needed. The legal claim of negligence bottoms out on the idea of carelessness, and Cutter was careless. Unlike the bottlers who had been routinely held liable for explosions of glass pop bottles, this was not a case of “accidents will happen.” It may be that vaccines are imperfect, but Cutter engaged in practices that increased the risk of harm by magnitudes. Because of the way it was litigated, in particular because the lawyer for the plaintiffs, Melvin Belli, did not know and did not prepare for the fact that one other companies had also been careless, did not know that Cutter had called in Youngner, did not know about the errors in the timing of the process and the use of inadquate glass filters, he wasn’t able to convince the jury of Cutter’s negligence. The full story of what had happened was not available, including some information that was not unearthed until Offit’s book was published in 2007. As a result, the case came to stand for the proposition that pharmaceutical companies could be held liable for the damage caused by their products even if they had taken 54 Gottsdanker, 182 Cal. App. 2d at 611. 55 Gottsdanker, 182 Cal. App. 2d at 611. 56 Comment, Strict Liability for Drug Manufacturers: Public Policy Misconceived, 13 Stan. L. Rev. 645, 647 (1961).

maximum care despite the fact that in the particular case, the company had not taken what would be fairly described as ordinary care. This fed into a larger mythology about strict liability, that innocent companies were being unfairly exposed to existential risk due to a new strict products liability regime. By the time the California court of appeals decided the Gottsdanker case in 1960 the idea of strict products liability had ascended. We have already seen its rise in the bottle cases, albeit without the use of the term strict liability. Fleming James had made the case in 1955 that the inevitability of accidents and the cost-spreading rationale should require manufacturers to pay for injuries regardless of fault.57 Courts made similar arguments earlier still, although still under the rubric of negligence, positing that it is a breach of duty not to take additional precautions when products were likely to injure. One lesson of the litigation following on the Cutter incident are that sometimes the uncertainty in law is not about causation but about misconduct. The kind of research lawyers need to do in order to unearth what happened is not always done, or cannot be done at the time because people do not want to talk to the lawyers, defendants can hide information, and lawyers may be unsophisticated with respect to technical issues. Sometimes, as in the Cutter case, even the government with its substantial expertise is not quite sure what the problem is. Thus, as litigated, the case turned away from the question of fault to the question of safety.58 To what extent should a promise of safety be enforced? As the North Carolina supreme court wrote in 1918, if goods are so dangerous that they cannot be made safe, then “placing them upon the market is indictable.”59 The California appellate court basically agreed with this sentiment in Gottsdanker. The Gottsdanker case had little influence over the trajectory of tort law, but it was part of a larger strict liability movement that began in California. It was the subject of interest in 1955, when the lawsuit was filed. A Yale Law Journal student note explained how the theory of strict liability advocated that same year by Fleming James should be applied in the case of vaccines.60 A 1967 article on the rise of strict products liability called it “[t]he most famous drug case[.]”61 It was an example of strict products liability used by academics at the time, but is only cited in 121 cases and 56 law review articles. Still, it illuminates the debate during this period between the (briefly) successful insurance rationale for products liability law and the (ultimately) dominant negligence rationale.

57 Fleming James, Jr., Products Liability, 34 Tex. L. Rev. 192, 228 (1955). 58 Legal commentators at the time thought that the negligence cause of action would not even go to the jury.

A note in the Yale Law Journal explained: “the methods Cutter used to inactivate and test its vaccine were generally believed to assure complete removal of live virus,10 and seem to satisfy the standards of care to which the manufacturing druggist has traditionally been held.11 Therefore it is likely that Cutter will be able to overcome the usual presumptions of negligence and thereby prevent the issue from going to the jury.” The Cutter Polio Vaccine Incident: A Case Study of Manufacturers' Liability Without Fault in Tort and Warranty, 65 Yale L.J. 262, 263 (1955) 59 Grant v. Graham Chero-Cola Bottling Co., 176 N.C. 256, 97 S.E. 27, 29 (1918). 60 The Cutter Polio Vaccine Incident: A Case Study of Manufacturers' Liability Without Fault in Tort and Warranty, 65 Yale L.J. 262, 263 (1955) 61 Friedrich Kessler, Products Liability, 76 Yale L.J. 887, 938 (1967). See also Comment, Strict Liability for Drug Manufacturers: Public Policy Misconceived, 13 Stan. L. Rev. 645, 645–46 (1961) (stating that the case was “the first case to sustain an implied warranty action by a party not in privity where the defective product was a drug.”).

Compare, as one student’s Comment did, the outcome of Gottsdanker with Green v. American Tobacco, a tobacco case litigated in the same period.62 Both cases were litigated on breach of warranty theories. The issue of privity was dispensed with on the theory that the ultimate user of the product could sue for implied warranty. The probable link between cigarette smoking and lung cancer satisfied the causation requirement. So what was left was the question of fault. The Fifth Circuit in Green v. American Tobacco, decided in 1962, stated that under Florida law: “The defendant could not be held liable as an absolute insurer against consequences of which no developed human skill and foresight could afford knowledge.”63 The overarching theory was that these were credence goods and the least cost avoider was the manufacturer. A manufacturer of a packaged product is in the better position to ensure product safety, but only for those dangers that it could foresee and prevent.64 But as we know from the subsequent history, the story of tobacco is, in fact, a story of misconduct.65 As Robert Rabin points out, tobacco litigation had two phases.66 The first phase, in the 1950s and 1960s, which produced Green v. American Tobacco, plaintiffs lost on both negligence and implied warranty theories. Rabin explains how in this first phase of tobacco litigation, there would have been sufficient evidence to sustain a negligence claim (the statement in Green v. American Tobacco notwithstanding, as he convincingly demonstrates,) but the tobacco companies’ litigation strategies wore out the opposition, and the plaintiffs bar to some extent mismanaged the litigation.67 Causation was important to this strategy. In the tobacco cases, although lack of causation was not a successful legal argument for defendants on the merits, the industry’s refusal to concede causality “worked to its advantage principally by imposing an enormous cost burden on its adversaries in waging a battle of the experts.”68 During this early period, the individual responsibility argument that would later dominate litigation around tobacco had yet to develop, in part because smokers before 1950 would not have been aware of the link to cancer.69 In the 1980s, after the star of strict products liability had risen, tobacco plaintiffs tried again.70 By then, the tort system “appeared to be in disarray from rising costs and award levels, with products liability law regarded as a prime villain.”71 I turn to that historical moment now, but in the context of pharmaceutical litigation rather than tobacco, and a case where ultimately the epidemiologic consensus was that the drug did not cause injury.

62 Comment, Cigarettes and Vaccine: Unforeseeable Risks in Manufacturers' Liability Under Implied Warranty, 63 Colum. L.

Rev. 515 (1963). 63 Green v. Am. Tobacco Co., 304 F.2d 70, 76 (5th Cir. 1962), certified question answered, 154 So. 2d 169 (Fla. 1963). 64 Here the Fifth Circuit relied on cases involving packaged foods, particularly canned foods. Id. See also Comment, Cigarettes and Vaccine, supra note __ at 535, quoting Karl Llewellyn: “This is a question of consumer. Of helpless consumer. Of consumer who takes what he gets, because he does not know enough, technically, to test even what is before his eyes.” Karl Llewellyn, On Warranty of Quality, And Society: II, 37 Colum. L. Rev. 341, 404 (1937). 65 For a general overview see Robert L. Rabin, A Sociolegal History of the Tobacco Tort Litigation, 44 Stan. L. Rev. 853 (1992). 66 Rabin, A Sociolegal History, 44 Stan. L. Rev. at 854. 67 Rabin, A Sociolegal History, 44 Stan. L. Rev. at 860. 68 Id. at 861. 69 Rabin, A Sociolegal History, 44 Stan. L. Rev. at 863 70 Rabin, A Sociolegal History, 44 Stan. L. Rev. at 854. 71 Id.

C. Bendectin and Categorical Separation of Causation and Misconduct The Cutter trial was conducted as all torts cases were during the period – as a continuous narrative. But the second case study presented here, the Bendectin litigation, was the first mass tort litigation to bifurcate trials in order to isolate the issue of causation from the other questions in the case. This approach presupposes a particular approach to science: that it can be separated from the social context in which it is developed. This view of scientific inquiry reached its peak in the late 1970s. By the 1980s, scientists were debating the role of society in scientific inquiry.72 And increasingly some scientists came to recognize that social assumptions and mores played a role in the development of scientific inquiry. In 1993, for example, the FDA finally required that pharmaceuticals be tested on women. The fact that this had never been required before was not a product of empirical rigor but of a blind spot among scientists, that is, it was the product of the social context in which research was being done. Because law always trails other disciplines, it has still not fully grappled with the impact of the context in which science is produced on the quality of the scientific result. Bendectin was a drug marketed to pregnant women to treat nausea in pregnancy. Betty Mekdeci, whose son David was born with birth defects, developed a theory that her son’s birth defects were caused by the Bendectin she took when she was pregnant.73 Melvin Belli, who had litigated the Gottsdanker case about twenty years earlier, agreed to take the case (although ultimately Mekdeci was represented by other lawyers).74 Lawyers were interested in Bendectin because the manufacturer, William Merrell Company, had previously produced Thalidomide in the United States and Canada and the first cholesterol-lowering drug (MER /29) which was pulled from the market after it came out that the company had suppressed evidence that it caused irreversible vision loss and other severe adverse reactions.75 Merrell’s prior malfeasance at least justified further investigation, the lawyers thought.76 But the issue in any Bendectin case would not be Merrell’s past misconduct but what happened with Bendectin. The drug posed a difficult causation issue because it presented the problem of weak effects. What if Bendectin increased the incidence of birth defects over the background rate, but this was virtually impossible to detect?77 Unlike Thalidomide, which had a signature injury, or the defective Cutter vaccine which caused the progression of the disease to begin in the arm that had received the shot, there wasn’t a clear causal connection in the Bendectin case. As Anita Bernstein explains, thalidomide “is the pinnacle

72 See, e.g., Trevor J. Pinch and Wiebe E. Bijker, The social construction of facts and artefacts: or How the sociology of science

and the sociology of technology might benefit each other, 14 Social Studies of Science 399-441 (1984). https://doi.org/10.1177/030631284014003004 73 MICHAEL D. GREEN, BENDECTIN AND BIRTH DEFECTS: THE CHALLENGES OF MASS TOXIC SUBSTANCES LITIGATION, 98-99 (1996). 74 Id. at 112. Two other law firms were supposed to do the pretrial work in the case and Belli ultimately did not sign a retainer with Mekdeci. The result was that later in the case, when all the lawyers sought to withdraw, Belli was not bound to the representation. Id. at 150-152. There is a fascinating story of how relationships with clients fall apart here, but that is beyond the scope of this essay. 75 Id. at 117. 76 Id. at 117-118. 77 Id. at 108.

toxin.”78 The effect of the drug was so substantial that a causal link was intuitive. “Limb reduction birth defects (one among many harms thalidomide caused) appeared in about half the children born to mothers who took thalidomide during the fifteen-day period when organs are formed; the background rate of this birth defect is less than one per thousand.”79 This close causal relationship created an expectation about proof of causation that would be very difficult to meet in many cases.80 As Michael Green explains: [A] basic difficulty with epidemiology, especially cohort studies, is its inability to discover small increases of risk above a background rate. If the toxic effect is powerful, epidemiology may be unnecessary. . . . But if the agent has only a small or modest effect, it will go undetected in a cohort study unless a massive number of subjects are included. The power of cohort studies in this situation is quite low.81 There exists a very good way to prove small or modest effects: randomized controlled trials. In a randomized controlled trial, a treatment group is given the intervention and a control group is not. The two groups are compared to see what effect the intervention has on the human body. This is considered the most rigorous method of testing scientific hypotheses, but it presents three problems relevant here. The first is that under certain circumstances it can be unethical. For example, imagine testing an unknown substance on pregnant women to see if it has teratogenic effects. Second, with latent conditions results can take a long time. Third, it is expensive. There was no randomized controlled trial of Bendectin before it was put on the market nor were there after-market studies.82 The rationale was that the components of the drug had not had any documented effects in past use, such that the combination should not either. This lack of studies, combined with the history of Thalidomide and Merrell’s involvement in that history, is the best explanation for what happened in the Bendectin litigation. Instead, the case has been framed as one of “junk science,” feeding into (perhaps aimed at creating) a perception that juries were illegitimate and the tort system out of control. Note, however, that if a drug is able to be put on the market because its components are considered sufficiently safe, from an ethical perspective it should also be able to be tested in a randomized controlled trial. 78 Anita Bernstein, Formed by Thalidomide: Mass Torts as A False Cure for Toxic Exposure, 97 Colum. L. Rev. 2153, 2165–66

(1997). 79 Id. 80 This could be described as a kind of “CSI effect” for causation. For a discussion see Tom R. Tyler, Viewing CSI and the Threshold of Guilt: Managing Truth and Justice in Reality and Fiction, 115 Yale L.J. 1050, 1052 (2006) (describing the CSI effect as the idea, presented mostly in popular media, that watching the television show “CSI: Crime Scene Investigators” caused potential jurors to develop “unrealistic expectations about the type of evidence typically available during trials, which, in turn, increases the likelihood that they will have a ‘reasonable doubt’ about a defendant's guilt.”) For a study of whether there is an actual CSI effect that favors criminal defendants see Simon A. Cole & Rachel Dioso-Villa, Investigating the "CSI Effect' Effect: Media and Litigation Crisis in Criminal Law, 61 Stan. L. Rev. 1335, 1337-1343 (2009) (demonstrating there is scant evidence that a CSI effect exists and positing that the popular discussion of that effect “embodies anxiety about science's threat to the law’s role in society as a truth-generating institution.”). 81 Michael D. Green, Expert Witnesses and Sufficiency of Evidence in Toxic Substances Litigation: The Legacy of Agent Orange and Bendectin Litigation, 86 Nw. U. L. Rev. 643, 653 (1992). 82 On the immense costs to corporations of testing products with some suggestions for reform see Wendy Wagner, When A Corporation's Deliberate Ignorance Causes Harm: Charting A New Role for Tort Law, 72 DEPAUL L. REV. 413, 417 (2023) (“Tort law is structured to deter corporations from producing products that foreseeably cause harm to individuals, but corporations' privileged role over controlling the information relevant to those claims requires some doctrinal adjustments to ensure that tort goals are effectuated.”)

The story of Bendectin is interlaced in multiple ways with that of Thalidomide. Both drugs were developed around the same time, marketed in the U.S. by the same company, accused of causing similar effects in utero, untested before marketing, and involving the same personalities. The story of Thalidomide in the United States is a famous one because it has a heroine: Frances Kelsey, a young mother and doctor working at the Food and Drug Administration. She delayed approval of the drug pending animal studies showing it did not have teratogenic effects. During the period that the drug’s dissemination in the United States was delayed, reports began to surface in Europe that it was causing birth defects. It was pulled from the market everywhere. Thalidomide was discovered to be a teratogen in 1960, the same year that Gottsdanker was decided. But the aftermath of Kelsey’s story is less often told. Although she was awarded the Medal of Freedom from President Kennedy and widely hailed as a savior of babies, she was sidelined in the Food and Drug Administration, which saw her as too careful and a burden on innovation and development of new drugs.83 Bendectin was developed and marketed before Thalidomide, but was litigated in the post-Thalidomide world. In 1956 Merrell filed a new drug application for Bendectin which was swiftly approved. It contained no warnings, “although its three ingredients had been the subject of only causal toxicity testing and no reproductive tests.”84 The labeling was not changed for twenty years. As Green explains: “Because the three ingredients had each been in use and because the prevailing pharmacologic at the time was that the fetus was protected by the placenta, Bendectin was not tested for teratogenicity before it was marketed as a prescription drug in 1956.”85 Merrell ghostwrote an article for Dr. Ray Nulsen, a supposedly independent obstetrician in 1957, lauding Bendectin. Nulsen had also been the author of a ghostwritten article favorably reviewing Thalidomide.86 Documents produced in later litigation showed that Merrell was aware that Bendectin studies were inadequate. The company did internal studies of its effects in animals and some of the studies found teratogenic effects, but ultimately Merrell scientists concluded they were not attributable to the drug.87 In 1973 it hired a French teratologist, Dr. Tuchmann-Duplessis, who reviewed Bendectin teratology studies and informed the company that “in his view the studies available were inadequate to justify a conclusion about Bendectin’s lack of teratogenicity.”88 The studies he recommended “were not concluded for almost ten years, until 1982.”89 At that point the litigation was well underway. Suppose that Bendectin was a weak teratogen, only increasing the risk of birth defects by five percent, as was alleged in some of the cases. If this were true, should it be enough to demonstrate that a plaintiff’s birth defects were caused by the drug? The courts and 83 This will all have citations in the final version! 84 GREEN, BENDECTIN, supra note __ at 90. 85 Id. 86 GREEN, BENDECTIN, supra note __ at 176. 87 Joseph Sanders, From Science to Evidence: The Testimony on Causation in the Bendectin Cases, 46 Stan. L. Rev. 1, 21

(1993) (describing animal studies). 88 GREEN, BENDECTIN, supra note __ at 176. 89 Id.

many commentators thought that this should not be sufficient as a matter of law.90 That is a legal judgment about a fact in the world, not a statement of the truth of the matter. The legal question of whether very modest effects and great uncertainty together should ever lead to liability is a different type of question than the question of what is knowable and provable without a randomized controlled trial. Eight hundred and eighteen Bendectin cases were ultimately consolidated in an MDL. The MDL is important because the Judge in the MDL trials issued a pioneering trifurcation order. He trifurcated the trials into three phases: general causation, liability, specific causation and damages.91 As Joseph Sanders, who studied the trial transcripts of the MDL cases in depth, explained “The MDL trial provide[d] an opportunity to observe a ‘pure’ general causation case, unobscured by other considerations.”92 This assertion raises the crucial question, is it possible to evaluate ‘pure’ general causation in a case involving the kinds of uncertainty present in Bendectin? What does purity mean in a case involving genuine conflicts about how to read the evidence available? Very shortly after the Bendectin litigation was emerging as the quintessential example of junk science in the courts, the late 1980s to the mid-1990s, the “story model” of jury decision-making also rose to prominence. Under this model, the jury uses information from both the trial and their own experience to create a narrative of what occurred to which they apply the legal standard.93 As Sanders explains in his analysis of Bendectin: “plaintiffs often attempted to commingle elements, thereby bolstering weak evidence on causation with stronger proof of breach of duty and damages. If Pennington and Hastie's story model of decisionmaking is correct, jurors are likely to trade proof of one element for proof of another, making this technique successful.”94 Indeed, Sanders demonstrates that in the MDL cases where presentation of general causation was completely cordoned off from all other evidence, the jury found for the defendant on that issue.95 The general view of legal scholars and judges is the one Sanders adopts: comingling elements is a mistake. Each element of the cause of action must be met, and therefore each must be considered separately. The ideal is to separate each into a stand-alone, “pure” inquiry. But what if, to use a metaphor from a different context, there cannot be acoustic separation between the policy question of breach of duty and the scientific question of

90 See, e.g., Richardson v. Richardson-Merrell, Inc., 649 F. Supp. 799 (D.D.C. 1986) (granting judgment as a matter of law

after jury trial), aff’d, 857 F.2d 823 (D.C. Cir. 1988), cert. denied 493 U.S. 882 (1989); Ealy v. Richardson-Merrell, Inc., 897 F.2d 1159 (D.C. Cir. 1990) (granting judgment as a matter of law after jury verdict), cert. denied, 498 U.S. 950 (1990); Raynor v. Richardson-Merrell, Inc., 643 F. Supp. 238, 244 (D.D.C. 1986) (decided on a motion for summary judgment without the plaintiff contesting key facts). See also Joseph Sanders, From Science to Evidence: The Testimony on Causation in the Bendectin Cases, 46 STAN. L. REV. 1, 27 (1993) (“A definitive case exonerating Bendectin or conclusively linking it to birth defects cannot be made. Still, the scientific findings disputing Bendectin's teratogenicity are substantial. If one employs traditional standards for proving the causal question, the science does not support a plaintiff's verdict.”). 91 GREEN, BENDECTIN, supra note _ at 197-98. 92 Sanders, From Science to Evidence, supra note __ at 29–30. 93 See Nancy Pennington & Reid Hastie, A Cognitive Theory of Juror Decision Making: The Story Model, 13 CARDOZO L. REV. 519, 520-29 (1991) (describing the process of cognitive construction of a story as part of juror decision-making). See also Richard L. Jolly, Valerie Hans, and Robert Peck, Democratic Renewal and the Civil Jury, 57 GA. L. REV. 79, 162 (2022) (connecting the story model to the juror’s broader social role). 94 Sanders, From Science to Evidence, supra note __ at 53. 95 Sanders, From Science to Evidence, supra note _ at 54.

causation?96 What if the causal question, particularly in cases of probabilistic cause, is inextricably linked with an evaluation of the conduct at issue and the stakes of the case? Consider the trial court’s decision to exclude “a portion of a letter prepared by one of the testers in Merrell's 1970's study, the Smithells study.”97 The letter “written by Smithells to Merrell Dow that mentioned his hope that publication of his study would save the defendant large sums of money defending California litigation.”98 The district judge ruled that this letter “went to classical bias of the person conducting the study, and not scientific bias that would go to whether the study was actually accurate.”99 He explained “As long as the methodology was correct, the fact that somebody might have intended to use the results in a particular way would be of no consequence.”100 The problem with the judge’s approach is that interpreting epidemiologic evidence requires judgment.101 The concept of “p-hacking” or “data diving” illustrates the point.102 While collecting and analyzing data, researchers have many decisions to make, including whether to collect more data, which outliers to exclude, which measure(s) to analyze, which covariates to use, and so on. If these decisions are not made in advance but rather are made as the data are being analyzed, then researchers may make them in ways that self-servingly increase their odds of publishing (Kunda, 1990). Thus, rather than placing entire studies in the file-drawer, researchers may file merely the subsets of analyses that produce nonsignificant results. We refer to such behavior as phacking.103 The legal system is less concerned with researchers looking to obtain interesting results to get a publication, and more concerned with researchers trying to find the “right” result of their client or their patron. This is why a document like the Smithells letter could be important. When professional judgment is at issue in a causation determination, so is the integrity of the person making that judgment. This is why separating the cause from context can distort attributive decision-making. For another example of a study that reflects the interrelationship between science and society, consider the only extant published epidemiologic study at the time Betty Mekdeci, the first Bendectin plaintiff, started to research the question of Bendectin’s effects. It was a 96 The reference is to Meir Dan-Cohen, Decision Rules and Conduct Rules: On Acoustic Separation in Criminal Law, 97 HARV.

L. REV. 625 (1986). 97 In re Bendectin Litig., 857 F.2d 290, 318 (6th Cir. 1988). 98 Id. 99 Id. 100 Id. 101 See Edward K. Cheng, The Consensus Rule: A New Approach to Scientific Evidence, 75 Vand. L. Rev. 407, 419 (2022) (“These questions ultimately boil down to scientific judgment--perhaps judgments aided by principles or factors, but judgments just the same. And such scientific judgment requires expertise. Indeed, the law's desire to “transmute [ ] scientific subtleties into formulaic legal shibboleths” both is hypocritical and shows a profound lack of professional respect for scientists.”). See also Richard G. “Bugs” Stevens, Does Monsanto’s Roundup Cause Cancer? The Conversation, August 3, 2018 (available at https://theconversation.com/does-monsantos-roundup-cause-cancer-trial-highlights-the-difficulty-of-proving-alink-100875) (comparing the method of finding scientific consensus to the jury deliberation process). 102 Simonsohn, U., Nelson, L. D., & Simmons, J. P. (2014). P-curve: A key to the file-drawer. Journal of Experimental Psychology: General, 143(2), 534–547. https://doi.org/10.1037/a0033242. 103 Id.

cohort study conducted by Merrell employees (Drs. Bunde and Bowles) and published in a third-tier medical journal. Green notes a number of serious flaws: The exposed and unexposed cohorts came from records of private obstetricians recommended by the Merrell marketing department who were known to be major Bendectin users. . . . The control group was not matched with the exposed cohort in terms of other risk factors, such as age, other drugs taken, and diabetes. Nor did the study account for those factors and determine whether they played a role in the outcome. The study examined only live births, thereby neglecting the possibility of an excess rate of miscarriages or stillbirths. The definition of exposure was too broad, including the entire first trimester, thereby including many women in the exposed cohort who should not have been. Moreover, the overall defect rate among the exposed cohort and all women was .5 and .72 percent respectively, rates that were so low compared to the general population figures that they strongly suggested substantial ascertainment error. . . .” One wonders how this could be the only published study of a drug given to pregnant women during pregnancy. Many better studies followed this one in the late 1970s and 1980s, but that delay too has a social meaning. One would imagine that after the Thalidomide experience there would be a much greater emphasis on testing, but something about a combination of scientific hubris and the relevant population meant that the simply wasn’t the kind of testing one might expect. In sum, Bendectin teaches two opposite lessons. First, there is a social aspect to scientific studies – who is studied, what is accepted as sufficient proof, how much skepticism of findings is warranted are all questions that can only be answered by judgment in context of its production. Second, judges thought in the 1980s and still think today, as evidenced by the trifurcation of the Roundup cases in the Northern District of California, that causation can be determined finally, objectively and definitively when in fact the science of epidemiology is an ongoing project of adjusting knowledge to fit new facts, is based on consensus judgments, and is probabilistic rather than determined. There is, in other words, a lack of fit between the community of scientists and the community of lawyers; we lack a shared understanding of the project at hand. Separating out the science from the law does not solve this tension, and may even exacerbate it.

C. The Relationship Between Expertise, Causation, and Fault Edward Cheng argues that lawyers and judges have a tendency to idealize science, either because of hubris or because their distance from it enchants them and they come to believe that science offers an unequivocal answer in a given situation.104 But in complex issues of toxins, for example, there are rarely unequivocal answers. Instead, knowledge is accrued over time until scientists come to a consensus about the likelihood that a toxin 104 Edward K. Cheng, The Consensus Rule: A New Approach to Scientific Evidence, 75 Vand. L. Rev. 407, 418, 429 (2022)

(explaining lawyer’s idealization of science as well as the idea that “distance leads to enchantment” such that science may appear more certain than it is.)

causes cancer, for example, or that a drug causes birth defects. The hardest cases are those where the science is uncertain, and harder still when the scientific results were obtained in ways and by individuals whose objectivity a factfinder cannot trust. In those cases, the kind of purity of science that lawyers wish existed is impossible to obtain. The Bendectin litigation has come to stand for the proposition that the risk that “junk science” is used to create claims of causation where none should lie – although that position has also been criticized.105 But there is much more than that to the case. It raises crucial questions such as who should bear the cost of injury under conditions of uncertainty when a product has not been tested? Who should bear the costs of testing products? What is the role of social influences on scientific inquiry, both in the decision not to test and in the decision to reject a remedy? Of those social influences, what role does the subordinate status of women as research subjects and objects of intervention play, in addition to the societal expectation of perfect motherhood, which includes producing a perfectly healthy child? With respect to the first set of questions, scholars have considered ways to address the problem of both uncertainty and deliberate ignorance or failure to test. For example, Edward Cheng suggests a rule that would require courts to defer to scientific consensus in cases where there was such consensus – such as for example the general agreement among epidemiologists today that Bendectin does not cause birth defects – but would allow the jury to determine the question in cases where there is no consensus – such as for example the disagreement in the late 1970s on the question of Bendectin’s risks of terateginicity.106 Wendy Wagner argues that the remedial structure of torts can and should address the issue of uncertain causation more directly, rather than through the law of evidence.107 Wagner explains that incentives of both the tort and capitalist systems push corporations to either hide evidence of causation or not create it at all.108 She proposes a new tort, called “deliberate ignorance,” which would include the following elements: (1) the “creation of a preventable (latent), material risk of harm to plaintiff”; (2) “by a wrongful act committed by a company with resources/expertise to conduct the research”; (3) “that causes plaintiff ongoing harm as a result of significant exposure.”109 A company could present an affirmative defense that it had followed the state of the art in its testing.110 The idea is that the costs and incentives to test would be aligned with public policy goals of making sure products are not injurious.

105 Daubert (9th Cir) at 1317 (citing PETER W. HUBER, GALILEO'S REVENGE: JUNK SCIENCE IN THE COURTROOM 206–09 (1991)

and explaining that “the prevalent practice of expert-shopping leads to bad science.”). See also Heidi Li Feldman, Science and Uncertainty in Mass Exposure Litigation, 74 TEX. L. REV. 1, 1 (1995) (describing commentary taking the “junk science” perspective); Edward K. Cheng, The Consensus Rule: A New Approach to Scientific Evidence, 75 VAND. L. REV. 407, 412 (2022) (describing the role of the idea of “junk science” in Bendectin litigation). 106 Cheng, The Consensus Rule, supra note __ at 440. 107 Wendy Wagner, When A Corporation's Deliberate Ignorance Causes Harm: Charting A New Role for Tort Law, 72 DePaul L. Rev. 413, 436 (2023). See also Alexandra D. Lahav, The Knowledge Remedy, 98 Tex. L. Rev. 1361 (2020). 108 Id. at 425-432 (she explains that corporations do this by maintaining ignorance, producing unreliable information, and attacking third party research). 109 Id. at 441. 110 “This state-of-the-art defense would place the burden on the defendant to establish that, at the very least, its own internal assessment and post-market monitoring efforts met or exceeded the best practices of an expert manufacturer in that same business.” Id. at 446.

As to the second, sociological set of questions, both the Cutter incident and the Bendectin story involved children, and illustrate the legal system’s reflection of the expectation of safety and intolerance of poor outcomes when it comes to children’s health in the modern era.111 With respect to the polio vaccine, it is hard to know how much the very real danger of polio affected the jury’s view that Cutter’s process was careful enough (despite Jonas Salk testifying in favor of the plaintiff in that trial). The rush to production of that vaccine was, after all, a reaction to the coming polio season and an attempt to save thousands of children’s lives. It, too, was a process that ended up being bound up with the politics of the day, particularly the need to avoid any hint of socialized medicine and therefore to allow private companies without requisite expertise to make this very complicated product. With respect to Bendectin, it seems likely that a combination of the legacy of Thalidomide, the evidence of lack of care on the part of Merrell, and the need to blame something other than terrible luck for birth defects were key drivers of that litigation. Had Merrell not engaged in manipulation of the science, had they not been intimately involved in Thalidomide, had they tested the product carefully because it was being administered to such a vulnerable population, it is likely that the outcome would have been different. One has to ignore quite a lot of evidence to isolate problems with epidemiologic method or to blame a jury’s inability to understand the relative strengths and weaknesses of different methodologies for determining toxicity as the primary causes of the outcomes of the early Bendectin trials. Ironically, the critics of the system on junk science grounds are making a mistake very similar to the one that they accuse jurors of making, which is to say they too manipulate the facts to fit their narrative. The story of junk science may be convenient, but it is only a small piece of a broader story, and the verdicts are better explained by social approbation for the company’s conduct which cannot be separated from the scientific research. CONCLUSION This essay has argued that judicial understandings of legal cause, and particularly the relationship between causation and fault, is historically contingent. The separation of the two in the 1980s was the result of several factors. First, developments in the substantive law, particularly the rise of strict liability, made separation possible. Second, a tension developed between the law’s need for determinism and the growth of epidemiology as a probabilistic science. The judiciary resolved this tension with an overreliance on scientific expertise and a belief that scientific objectivity could be divorced from the social context in which research was conducted. This resolution was driven by political forces, particularly the tort reform movement of the 1980s, but it is not exclusively explained by those forces. One of the defining features of legal analysis is the use of categories. Events are broken up, examined one by one, placed into categories, then (sometimes) put together again. This characteristic of legal thought was famously summarized by Thomas Reed Powell: “If you think that you can think about a thing inextricably attached to something else without 111 See generally VIVIANA ZELIZER, PRICING THE PRICELESS CHILD (1985) (presenting an account of the changing understanding

of children from useless to “priceless” between the nineteenth and twentieth centuries).

thinking of the thing which it is attached to, then you have a legal mind.”112 It is a quintessential legalistic approach to try to separate the “pure” causation question from the question of how that knowledge of causation came about, what the expectations of care are among the actors in the legal story of misconduct, and the background knowledge of social expectations. Understanding that causation differs in law and science, that causation in law is attributive, singular, deterministic and binary, helps to understand why context in the legal evaluation of scientific evidence is so important, and therefore why the attempt to separate science from context in the legal setting is a kind of category error. Perhaps the attempt to separate causation from the rest of the story is something only a lawyer can do, but it does not mean it is something a lawyer should do.

112 The quote is from Thomas Reed Powell, reproduced in Thomas Arnold, The Symbols of Government 101 (1935) (internal quotation marks omitted). Also quoted in Fisher v. Univ. of Texas at Austin, 570 U.S. 297, 336 (2013) (Ginsburg, J. dissenting).

DRAFT: DO NOT CIRCULATE OR CITE WITHOUT PERMISSION

USING PUBLIC NUISANCE LITIGATION TO ADDRESS INDUSTRYWIDE MISCONDUCT: COMMON-LAW STATUTES, NONDELEGATION DOCTRINE, AND REGULATION BY LITIGATION Hillel Y. Levin* & Timothy D. Lytton** ABSTRACT In this Essay, we assert that broad, open-ended public nuisance statutes offer a legitimate legal tool to curb public health harms caused by the unreasonable design, marketing, and distribution of practices product manufacturers. We argue that these statutes are “common-law statutes” by which legislatures authorize courts to apply the concept of public nuisance to new situations using common-law-style reasoning. We demonstrate that this understanding of public nuisance statues is consistent with well-established interpretive traditions regarding the codification of common law. Statutory delegation of broad interpretive discretion to courts concerning the application of public nuisance law constitutes a legislative choice to regulate through litigation. We contend that the delegation to courts to elaborate on public nuisance law is no less justified than delegation to administrative agencies to exercise policy discretion through adjudication and rulemaking. TABLE OF CONTENTS INTRODUCTION ................................................................................................ 2 I. STATUTORY PUBLIC NUISANCE CLAIMS ...................................................... 4 A. Statutes ................................................................................................... 4 B. Claims..................................................................................................... 6 1. Firearms .............................................................................................. 6 2. Lead Paint ........................................................................................... 8 3. Opioids ................................................................................................ 8 4. E-Cigarettes......................................................................................... 9 5. Social Media Platforms ..................................................................... 10 II. NONDELEGATION DOCTRINE .................................................................... 11 III. BROAD PUBLIC NUISANCE STATUTES AS COMMON-LAW STATUTES ....... 19 A. Common-Law Statutes.......................................................................... 19 B. Codification and “Common-Law Elasticity” ....................................... 22 C. Delegating Regulation by Litigation .................................................... 25 CONCLUSION ................................................................................................. 32 * Alex W. Smith Professor of Law, University of Georgia. ** Regents’ Professor of Law, Georgia State University.

5/29/2024]

PUBLIC NUISANCE LITIGATION

INTRODUCTION Public nuisance lawsuits have become a familiar tool in efforts to hold product manufacturers liable for public health problems such as tobaccorelated illness, 1 firearm violence, 2 lead poisoning, 3 opioid addiction, 4 ecigarette use by minors, 5 and social-media-induced adolescent depression and anxiety. 6 Scholarly debate about these claims has focused primarily on whether they distort common-law public nuisance doctrine. 7 However, some of the claims are based on statutes, and these have received less attention from legal scholars. 8 In this Essay, we argue that statutory public nuisance claims offer a legitimate legal tool to address design, marketing, and distribution practices that harm public health. We are particularly interested in statutes that define public nuisance as conduct or conditions that are injurious to public health, safety, or welfare without providing specific definitions of those terms. 9 Such Donald G. Gifford, Public Nuisance as a Mass Products Liability Tort, 71, CIN. L. REV. 741, 747 (2003). 2 Hillel Y. Levin & Timothy D. Lytton, The Contours of Gun Industry Immunity: Separation of Powers, Federalism, and the Second Amendment, 75 FLORIDA L. REV. 833, 852 (2023). 3 Victor E. Schwartz, Phil Goldberg & Corey Schaecher, Game Over? Why Recent State Supreme Court Decisions Should End the Attempted Expansion of Public Nuisance Law, 62 OKLA. L. REV. 629, 643 (2010). 4 Leslie Kendrick, The Perils and Promise of Public Nuisance, 132 YALE L. J. 702, 731 (2023). 5 The People of the State of California v. JUUL Labs, Inc., First Amended Complaint, No. RG19043543, March 15, 2022. 6 In re Social Media Adolescent Addiction/Personal Injury Products Liability Litigation, Plaintiffs’ Master Complaint (Local Government and School District), No. 4:22-MD-3047, Dec. 18, 2023, at 297. 7 E.g., Gifford, supra note _; Schwartz, Goldberg & Schaecher, supra note _; Thomas W. Merrill, Is Public Nuisance a Tort? 4 J. OF TORT L. 1, 52 (2011); Richard A. Epstein, The Private Law Connections to Public Nuisance Law: Some Realism about Today’s Intellectual Nominalism, 17 J. of L. ECON. & POL’Y 282 (2022); Kendrick, supra note _; Thomas W. Merrill, The New Public Nuisance: Illegitimate and Dysfunctional, YALE L. J. FORUM, Feb. 20, 2023; John C. P. Goldberg, On Being a Nuisance, 99 N.Y.U. L. Rev. (2024). 8 For examples of statutory public nuisance claims, see infra Part I.B. 9 E.g., Indianna Code 32-30-6-6 (“Whatever is (1) injurious to health; (2) indecent; (3) offensive to the senses, or (4) an obstruction to the free use of property; so as essentially to interfere with the comfortable enjoyment of life or property, is a nuisance, and the subject of an action.”); New Mexico Stat §30-8-1 (“A public nuisance consists of knowingly creating, performing or maintaining anything affecting any number of citizens without lawful authority which is either: A. injurious to public health, safety, morals or welfare; or B. interferes with the exercise and enjoyment of public rights, including the right to use public property.”). For additional examples, see infra Part I. Compare Merrill, [2011], supra note _ at 16 (insisting that public nuisance denotes “an unlawful condition, not … any particular type of conduct”). 1

5/29/2024]

PUBLIC NUISANCE LITIGATION

statutory language raises a fundamental interpretive question: do these statutes merely ratify the common law doctrine of public nuisance as it existed at the time of legislative enactment, or does their broad, open-ended language authorize courts to apply them expansively in new contexts? 10 Thomas Merrill, a leading critic of public nuisance lawsuits against product manufacturers, argues that courts should interpret public nuisance statutes narrowly. 11 Merrill asserts that a literal reading of broad, open-ended provisions would violate state nondelegation principles because it would outsource too much policymaking authority to courts. 12 Merrill proposes that courts avoid this problem by adopting a “canon of non-dynamic interpretation of public nuisance law.” 13 According to this canon, courts should construe open-ended public nuisance statutes as limited in scope by the common law of public nuisance as it existed at the time of enactment. 14 We maintain that Merrill’s nondelegation concerns are unfounded and that broad, open-ended statutory provisions confer upon courts wide interpretive discretion in applying public nuisance law. Based on textual and historical evidence, we demonstrate that broad, open-ended public nuisance provisions belong to a category of statutes known as “common-law statutes.” These are statutes that delegate discretion to courts to elaborate on the meaning of open-ended statutory terms in common-law fashion. Notable examples of such statutes include the Sherman Antitrust Act, which empowers judges to define “restraint of trade,” 15 and the Federal Trade Commission Act, which relies on judges to define “unfair or deceptive acts or practices in or affecting commerce.” 16 Similarly, some public nuisance statutes require judges to define conduct or conditions that are “injurious to health” 17 or cause “hurt, inconvenience, or damage to another.” 18 The Essay proceeds as follows. In Part I, we provide a taxonomy of statutory public nuisance provisions to more clearly define the class of broad, open-ended provisions that are the object of our analysis. We also provide examples of litigation that feature such provisions. In Part II, we challenge Merrill’s approach to interpreting broad, openended public nuisance statutes. We argue that Merrill’s nondelegation We use the term “broad” to denote a statutory provision’s scope of application and the term “open-ended” to refer to the provision’s lack of specificity. 11 Merrill, [2011] supra, note _ at 52. 12 Id. 13 Id. 14 Merrill, [2011] supra note _ at 52. 15 15 U.S.C. §1. 16 15 U.S.C §3059. 17 Cal. Civ. Code §3479. 18 Alabama Code 1975 §6-5-120. 10

5/29/2024]

PUBLIC NUISANCE LITIGATION

analysis begs important questions to advance a highly contested conception of the separation of powers and does not accurately reflect nondelegation doctrine. Additionally, we demonstrate that Merrill’s approach disregards established principles of statutory interpretation and betrays an anti-plaintiff bias. In Part III, we argue that broad, open-ended statutory public nuisance provisions are common-law statutes. We begin in Part III.A. by describing the defining features of common-law statutes and demonstrating that broad, open-ended public nuisance statutes possess these features. In Part III.B. we demonstrate that reading public nuisance statutes as common-law statutes finds support in well-established interpretive traditions regarding the codification of common law. In Part III.C. we explain that common-law statutes, which authorize courts to elaborate on broad, open-ended statutory terms, reflect a legislative choice to regulate through litigation. We contend that this choice is no less defensible than broad delegation of policymaking discretion to executive branch agencies. Our analysis reveals that the virtues of regulation by litigation are especially well suited to addressing industrywide misconduct that harms public health and safety. We conclude with some thoughts about the implications of our analysis for the use of other common-law statutes, such as unfair trade practices laws, as a basis for regulation by litigation. I. STATUTORY PUBLIC NUISANCE CLAIMS Public nuisance statutes take various forms. In this Part, we describe a variety of statutory provisions and illustrate how broad, open-ended statutory public nuisance provisions feature in lawsuits against product manufacturers. A. Statutes The earliest records of English common-law public nuisance actions are from the late twelfth century. 19 Parliamentary statutes deeming specific conduct or conditions a public nuisance date back a least to the seventeenth century. 20 In the United States, statutory public nuisance provisions first

Gifford, supra note _ at 791. Id. at 798. Nuisance is variously defined as wrongful conduct or a harmful condition. Compare Arizona Revised Statutes §13-2917(B) (deeming conduct to be a public nuisance) and Arizona Revised Statutes §36-601 (listing a series of conditions as public nuisances). Compare also Goldberg, supra note _ (characterizing nuisance as a form of wrongful conduct) and Dan B. Dobbs, Paul T. Hayden & Ellen M. Bublick, THE LAW OF TORTS §403 (2nd ed. 2024) (characterizing nuisance as an unreasonable condition). 19 20

5/29/2024]

PUBLIC NUISANCE LITIGATION

appeared in Florida in 1832. 21 Today, the laws of every state include some form of public nuisance legislation. 22 Public nuisance statutes often contain multiple provisions which may vary in their generality. The generality of statutory public nuisance provisions varies along two dimensions: the scope of the prohibition and the liability standard. Some provisions prohibit specific conduct or conditions, for example, those that deem prostitution or noxious odors a public nuisance.23 Others are more general, proscribing unreasonable interference with the use and enjoyment of land or doing anything injurious to public health. 24 Public nuisance provisions also rely on different liability standards. Some impose strict liability, which entails a clear prohibition, 25 while others require negligence, which requires an unspecified level of care. 26 Figure 1 illustrates these distinctions. FIGURE 1. VARIATION IN STATUTORY PUBLIC NUISANCE PROVISIONS

Although we recognize that generality is not binary but rather a matter of degree, we offer this schematic categorization to distinguish between specific statutory prohibitions that provide relatively clear direction to judges in applying the law and more open-ended provisions that require greater judicial elaboration when applying them. Our primary interest lies in public nuisance

Act Feb. 10, 1832, § 47, Acts of the Legislative Council of the Territory of Florida, 1832 at 71 (1988), http://edocs.dlis.state.fl.us/fldocs/leg/actterritory/1832.pdf. 22 See, e.g., [STRING CITATION TO STATE LAWS]. 23 See, e.g., Iowa Code §657.2 (deeming “noxious exhalations” and “unreasonably offensive smells” as nuisances). 24 See, e.g., Indianna Code 32-30-6-6, supra note _ (deeming “[w]hatever is (1) injurious to health; (2) indecent; (3) offensive to the senses, or (4) an obstruction to the free use of property” as a nuisance). 25 See, e.g., Arizona Statutes §36-601 (imposing strict liability for the creation or maintenance of certain conditions deemed to be public nuisances). 26 See, e.g., Gary ex rel. King v. Smith & Wesson, Corp., 801 N.E.2d 1222, 1230 (2003) (interpreting Indianna’s public nuisance statute as incorporating a common-law reasonableness standard). 21

5/29/2024]

PUBLIC NUISANCE LITIGATION

laws that prohibit broad categories conduct and impose unspecified reasonableness standards. The shaded box in Figure 1 denotes these laws. In addition to prohibitions and liability standards, public nuisance statutes prescribe various procedures and remedies. Some public nuisance statutes rely on public prosecution and subject violators to criminal penalties. 27 These statutes may or may not create private rights of action. 28 Other statutes authorize civil actions by public officials or private parties for damages or equitable remedies, including diverse forms of summary abatement. 29 Scholars disagree over whether statutory public nuisance is best understood as a crime, a tort, or a regulatory offense. 30 We suspect that statutory public nuisance is, in many cases, a hybrid that does not entirely conform to these formal distinctions. However, for the purposes of this Essay, we take no position on the much-debated question of whether statutory public nuisance is a tort. B. Claims In the past several decades, individuals and government entities have filed public nuisance claims against product manufacturers alleging variously that their design decisions, marketing strategies, and distribution practices constitute a public nuisance. Some of these claims rely on statutory public nuisance provisions. We present examples below involving the manufacturers of tobacco, firearms, lead paint, opioids, e-cigarettes, and social media platforms. We offer these examples to illustrate, first, that these statutes have served as the basis for some lawsuits against product manufacturers and, second, that some courts have interpreted these statutes broadly and elastically. We make no claims regarding the validity of the allegations regarding industry practices or their causal connection to plaintiffs’ injuries. Some of the cases we cite are doctrinal outliers or sui generis, and many of the interpretive theories are untested by state supreme courts due to settlement of the claims or their dismissal on other grounds. 1. Firearms Shooting victims and cities have sued several firearm manufacturers for statutory public nuisance. 31 Plaintiffs’ complaints in these cases detail how See, e.g., Montana Code 45-8-111 (classifying the creation or maintenance of a public nuisance a criminal offense). 28 Compare, e.g., Alabama Code §6-5-123 (granting private right of action for public nuisance) and N.J. 2C:58-33 (restricting right to bring statutory cause of action to the state attorney general for public nuisance involving the sale of firearms). 29 See, e.g., Idaho Code 52-111. 30 Compare Merrill, [2011] supra note _ at _ and Kendrick, supra note _ at _. 31 E.g., Ileto v. Glock, 349 F.3d 1191 (2003) (lawsuit filed by shooting victim, subsequently dismissed based on preemption by the Protection of Lawful Commerce in Arms 27

5/29/2024]

PUBLIC NUISANCE LITIGATION

defendants’ marketing and distribution practices allegedly increase the risk of illegal sales that facilitate criminal misuse of their weapons. 32 In Ileto v. Glock, shooting victims brought a claim based on a broad statutory public nuisance provision in the California Code, which states: Anything which is injurious to health, including, but not limited to, the illegal sale of controlled substances, or is indecent or offensive to the senses, or an obstruction to the free use of property, so as to interfere with the comfortable enjoyment of life or property, or unlawfully obstructs the free passage or use, in the customary manner, of any navigable lake, or river, bay, stream, canal, or basin, or any public park, square, street, or highway, is a nuisance. 33 In reversing the district court’s dismissal of the plaintiffs’ public nuisance claim, the United States Court of Appeals for the Ninth Circuit interpreted this California Code provision in light of the Restatement (Second) of Torts definition of public nuisance to cover significant and unreasonable interference with public rights to health, safety, peace, comfort, or convenience. 34 The court held that the plaintiffs’ allegations were sufficient to support a claim for public nuisance under the statute. 35 Although the Ninth Circuit recognized the validity of the plaintiff’s public nuisance theory, the Court subsequently dismissed the lawsuit based on the Protection of Lawful Commerce in Arms Act, passed by Congress in 2005, which grants firearm manufacturers and sellers broad immunity from civil lawsuits arising out of criminal misuse of a weapon. 36 A similar case filed by the City of Gary, Indianna, against handgun manufacturers relied on the Indianna statute: 37 Whatever is: (1) injurious to health; (2) indecent; (3) offensive to the senses; or Act); Gary ex rel. King v. Smith & Wesson, supra note _ (lawsuit filed by the City of Gary, Indianna, subsequently dismissed based on a 2024 Indianna statute depriving local government entities from filing lawsuits against firearms manufacturers). 32 Ileto v. Glock, supra note _ at 1195-1199 (detailing allegations that firearm manufacturer knowingly marketed and distributed weapons to profit from illegal downstream sales). See Robert L. Rabin, Enabling Torts, 49 DePaul L. Rev. 435 (1999) (analyzing such theories as an example of “enabling torts”). 33 Cal. Civ. Code §3479. 34 Ileto v. Glock, supra note _ at 1211. 35 Id. 36 15 U.S.C. §§7901-7903. 37 Gary ex rel. King v. Smith & Wesson, supra note _ at 1299.

5/29/2024]

PUBLIC NUISANCE LITIGATION

(4) an obstruction to the free use of property; so as essentially to interfere with the comfortable enjoyment of life or property, is a nuisance, and the subject of an action. 38 The Indianna Supreme Court held that firearm manufacturers were subject to liability for statutory public nuisance based on their distribution practices.39 Following the recent passage by the Indianna legislature of a law prohibiting local governments from filing lawsuits against gun manufacturers, the lawsuit currently appears headed for dismissal. 40 2. Lead Paint Local governments have filed lawsuits against manufacturers of lead paint for public nuisance. They argue that the companies promoted the use of lead paint In County of Santa Clara v. Atlantic Richfield, a group of local governments in California alleged that manufacturers promoted the use of lead paint for interior residential use and engaged in a concerted effort to conceal the products’ health risks, resulting in widespread lead poisoning. 41 A state appellate court reversed the trial court’s dismissal of the plaintiffs’ public nuisance claim, and the case ultimately settled. 42 The plaintiffs’ claims invoked the same broad California Code public nuisance provision relied upon in Ileto. 43 3. Opioids State public nuisance statutes have also played a role in lawsuits against drug manufacturers for design, marketing, and distribution practices that have contributed to the opioid crisis. 44 In federal multidistrict litigation proceedings, which collected more than 2,400 such claims, a Magistrate Judge’s report analyzed Montana’s statutory public nuisance provision, which states: Anything that is injurious to health, indecent or offensive to the senses, or an obstruction to the free use of property, so as to interfere with the comfortable enjoyment of life or property, or that unlawfully obstructs the free passage or use, in the customary manner, of any navigable lake, Indianna Code 32-30-6-6. Id. 40 Carole Carlson, Gun Makers Move to Dismiss Gary Lawsuit in Wake of New State Law, CHICAGO TRIB. Mar. 21, 2024. 41 County of Santa Clara v. Atlantic Richfield Co., 40 Cal. Rptr. 3d 313 (2006). 42 $305 Million Settlement Reached for California Cities & Counties in Historic Lead Pigment Litigation, Motley Rice, July 17, 2019, https://perma.cc/JV38-LABD. 43 County of Santa Clara v. Atlantic Richfield Co., supra note _, at 321. See also People v. ConAgra Grocery Prod. Co., 227 Cal. Rptr. 3d 499 (2017). 44 E.g., State of Montana v. Purdue Pharma, complaint, No. ADV-2019-949 (2017). 38 39

5/29/2024]

PUBLIC NUISANCE LITIGATION

river, bay, stream, canal, or basin or any public park, square, street, or highway is a nuisance. 45 In rejecting the defendants’ argument that this provision should be read as limited to the misuse or interference with land or real property interests, the Magistrate Judge pointed to the plain language of the statute and interpretation of the statute by the Montana Supreme Court declaring that it would follow the plain language of the statute absent express limitations set by the legislature. 46 The Magistrate Judge also cited a 1998 Montana trial court opinion recognizing that marketing practices by tobacco companies could constitute a public nuisance. 47 Following the report, the State of Montana settled its claims with the defendants. 48 A federal district court allowed a similar statutory public nuisance claim by California local governments against Purdue Pharma to proceed. 49 4. E-Cigarettes Individuals and government entities have filed statutory public nuisance claims against manufacturers of e-cigarettes. 50 In one such case, the Whatcom County School District in Washington State sued JUUL Labs, alleging that the company designed its e-cigarettes to appeal to minors and increase their addictiveness, aimed their marketing at kids, concealed the health risks associated with vaping, and established distribution channels designed to capitalize on illegal sales, ultimately causing widespread nicotine use and addiction among school-age children. 51 The plaintiffs’ claim relied on Washington’s public nuisance statute:

Mont. Code 27-30-101(1). Report and Recommendation by Magistrate Judge David A. Ruiz, MDL No. 1:17-cv02804, Case no. 1:18-op-45749 25 (2019) 25 (citing Tarlton v. Kaufman, 199 P.3d 263 at 270: “Since the Montana Legislature has not narrowed the definition of what constitutes a nuisance, we are not at liberty to do so here.”). 47 Id. at 26 (citing State v. Phillip Morris, DA 07-0299 at 12 (2009) (holding that defendant tobacco companies’ “alleged deceitful and misleading conduct in promoting tobacco products; the alleged deceptive manipulation of nicotine in tobacco products; the alleged targeting of minors; and the alleged misrepresentation to the public regarding the safety of tobacco products … adequately pled a cause of action for public nuisance”). 48 Compare Report and Recommendation by Magistrate Judge David A. Ruiz, MDL No. 1:17-cv-02804, MDL No. 1:18-op-45459 (2019) and State ex rel. Hunter v. Johnson & Johnson, 499 P3d. 719, (2021) (holding that the Oklahoma public nuisance statute did not cover alleged misconduct in the marketing, sale, and distribution of opioid products and dismissing the state’s public nuisance claims against drug manufacturers). 49 City & County of San Francisco v. Purdue Pharma, 620 F. Supp. 3d 936 (N.D. CA 2022). 50 E.g., People of the State of California v. JUUL Labs, Inc., First Amended Complaint, Case no. RG19043543 (2022). 51 In re JUUL Labs, Inc., 2021 WL5276162 (N.D. Cal. 2021). 45 46

5/29/2024]

PUBLIC NUISANCE LITIGATION

Nuisance consists in unlawfully doing an act, or omitting to perform a duty, which act or omission either annoys, injures or endangers the comfort, repose, health or safety of others, offends decency, or unlawfully interferes with, obstructs or tends to obstruct, or render dangerous for passage, any lake or navigable river, bay, stream, canal or basin, or any public park, square, street or highway; or in any way renders other persons insecure in life, or in the use of property. 52 JUUL has settled dozens of public entity lawsuits and thousands of individual claims for more than a billion dollars so far. 53 5. Social Media Platforms Local governments, school districts, and state attorneys general across the country have filed public nuisance claims against social media platforms. 54 The plaintiffs claim that social media companies design their products and market them in ways intended to foster addiction and that features of the companies’ apps cause anxiety, eating disorders, depression and, in some cases, suicide, among young users. 55 The complaint in that case references multiple broad, open-ended state statutory public nuisance provisions, including, for example, New Mexico’s: A public nuisance consists of knowingly creating, performing or maintaining anything affecting any number of citizens without lawful authority which is either: A. injurious to public health, safety, morals or welfare; or B. interferes with the exercise and enjoyment of public rights, including the right to use public property. 56 This and other cases are currently pending in state and federal courts around the country. These various examples illustrate how broad, open-ended public nuisance statutory provisions feature in claims against product manufacturers. In some cases, courts have interpreted these provisions using common-law-style reasoning to cover novel forms of conduct or conditions. In the next Part, we Rev. Code of Washington 7.48.120. Christy Bieber, Juul Lawsuit Update, Forbes Advisor, May 22, 2023, https://perma.cc/6YYX-5NN3. 54 Cecilia Kang & Natasha Singer, Meta Accused by States of Using Features to Lure Children to Instagram and Facebook, N.Y, TIMES, Oct. 24, 2023. E.g., In re Social Media Adolescent Addiction supra note _ at 297 (citing multiple state statutory public nuisance provisions in lawsuits by local governments against social media platforms for public nuisance). 55 Id. at 18-296. 56 N.M. Stat. §30-8-1, in In re Social Media Adolescent Addiction supra note _ at 297. 52 53

5/29/2024]

PUBLIC NUISANCE LITIGATION

critically examine the claim that this interpretive approach runs afoul of nondelegation doctrine. II. NONDELEGATION DOCTRINE Thomas Merrill argues that a literal reading of broad, open-ended public nuisance statutes violates the principle of nondelegation and therefore must be avoided. 57 To him, using public nuisance statutes against manufacturers of firearms, lead paint, opioids, e-cigarettes, and social media platforms is illegitimate, and courts that have allowed such claims have erred. According to Merrill, under the nondelegation doctrine, legislatures “may not delegate too much discretion to another branch” of government; however, read literally, broad public nuisance statutes do just that. 58 To address this “delegation deficit,” Merrill proposes that courts avoid potential delegation problems by adopting a “canon of non-dynamic interpretation of public nuisance law.” 59 This canon would require courts to treat broad, open-ended Merrill, Public Nuisance as Risk Regulation, 17 J. L. ECON. & POL’Y 347, 367 (2022) (“some of the state statutes authorizing public nuisance actions may flunk the nondelegation doctrine, which is often applied more rigorously under state constitutional law than it has been as a matter of federal constitutional law”). Merrill also asserts that a public nuisance statute that adopted the Restatement Second of Torts definition of public nuisance would be unconstitutional. Merrill, [2011] supra note _ at 35 fn.134 (“A legislative directive to courts to determine what constitutes ‘unreasonable interference with a right common to the general public’ would press the limits of even the most elastic conception of the nondelegation doctrine.”). See also Merrill [2023], supra note _ at 997 (asserting that if Congress enacted a statute conferring the power on an agency to determine what constitutes “an unreasonable interference with a right common to the general public’ … the current Court would likely hold it unconstitutional” on nondelegation grounds). Whether Merrill is correct that a public nuisance statute that adopted the Restatement definition encompassing any “right common to the general public” without qualification would violate the nondelegation doctrine is of little practical relevance since even the most open-ended state public nuisance provisions refer to more circumscribed categories of interests in health, safety, comfort, and public morality. It is worth noting that the Restatement definition provides a list of factors that arguably circumscribe the category of rights common to the general public. See RESTATEMENT (SECOND) TORTS, §821B Public Nuisance. 58 Merrill, [2023] supra note _ at 997; Merrill, [2022], supra note _ at 367; Merrill, [2011] supra note _ at 35 fn. 134. 59 Merrill, [2011] supra note _ at 52. Merrill’s interpretive approach to broad public nuisance statutes is a specific application of the Constitutional-Doubt Canon. See Congressional Research Service, The Constitutional Avoidance Doctrine: The Constitutional-Doubt Canon (Part 3 of 3), Legal Sidebar, LSB10722, March 29, 2022. See also Margaret H. Lemos, The Other Delegate: Judicially Administered Statutes and the Nondelegation Doctrine, 81 S. CAL. L. REV. 405, (2008) (discussing how the U.S. Supreme Court “experimented with the possibility of enforcing the intelligible principle requirement [of the nondelegation doctrine] through statutory construction”); Cass Sunstein, Nondelegation Canons, 67 Chi. L. Rev. 315 (2000) (analyzing the use of statutory construction canons as a means of limiting Congressional delegation). 57

5/29/2024]

PUBLIC NUISANCE LITIGATION

public nuisance statutes as “an implicit ratification by the legislature of the common law of public nuisance as of the time the statute was adopted.” 60 The canon would impose a clear statement rule on statutory public nuisance claims. As Merrill explains: Established forms of liability would be covered, and close analogues. Absent some clear directive that the legislature intends to delegate authority to the courts to read the statute in a dynamic fashion, however, the list of proscribed circumstances should be regarded as closed. New applications would require more specific legislative authorization. 61 When public nuisance statutes prohibit specific conduct or conditions, Merrill insists that they should be strictly construed: “courts should enforce these statutes by their terms, but should not read them as implicitly authorizing public nuisance actions for other types of conditions not covered by the statute.” 62 Thus, when interpreting public nuisance statutes, “absent some clear indication that the legislature intends courts to read the statute dynamically, the list of circumstances covered would be regarded as fixed by the common law in effect when the statute was adopted.” 63 Merrill asserts that his approach would eliminate private rights of action for public nuisance “absent evidence that the legislature intended to ratify the special injury rule.” 64 We find Merrill’s approach unconvincing. To begin with, he justifies his proposed canon of nondynamic interpretation of public nuisance law by assuming without any evidence that broad, open-ended statutory public nuisance provisions violate state nondelegation doctrines. 65 Merrill’s canon relies on an impressionistic constitutional claim devoid of rigorous constitutional analysis, which invalidates statutory applications that might well survive constitutional review. 66 To the extent that state nondelegation doctrines track the federal standard, it does not seem obvious or intuitive that prohibitions on unreasonable interference with health, safety, comfort, and public morality fail to provide an “intelligible principle” comparable to other well-established provisions in common-law statutes, such as the Sherman

Merrill, [2011] supra note _ at 52. Id. 62 Id. 63 Id. at 53. 64 Id. 65 Merrill, [2022] supra note _ at 367. 66 See John F. Manning & Mathew C. Stephenson, LEGISLATION AND REGULATION: CASES AND MATERIALS 402 (4th ed. 2021) (articulating the view that “it is troubling that the avoidance canon apparently allows courts to make quasi-constitutional decisions on the cheap”). 60 61

5/29/2024]

PUBLIC NUISANCE LITIGATION

Act’s prohibition on unreasonable restraint of trade or the Federal Trade Commission Act’s prohibition on unfair trade practices. 67 Moreover, although state courts reportedly invalidate about 10 statutes per year on nondelegation grounds 68—typically under stricter standards than that of the federal nondelegation doctrine 69—Merrill cites and we are aware of no cases striking down open-ended public nuisance provisions on nondelegation grounds. Nor is there reason to believe that state courts would be amenable to reliance on statutory construction to avoid nondelegation problems. According to one recent analysis, “state supreme courts tend to use constitutional avoidance canons in nondelegation cases quite sparingly.” 70 Perhaps more to the point, common-law statutes delegate common-law powers to courts, and Merrill’s canon of nondynamic interpretation is inconsistent with the very idea of this kind of statutory delegation. 71 However, such delegations are well-entrenched and widely recognized in American jurisprudence. 72 Moreover, as we will demonstrate below in Part III, wellFor additional examples, see Lemos, supra note _ at 429-430 fn. 22. We express no opinion at this point about whether these and other common-law statutory provisions would pass muster under the stricter nondelegation test set forth in Justice Gorsuch’s dissent in Gundy v. United States, 139 S. Ct. 2116, 2131 (2019) (Justice Gorsuch dissenting). 68 Benjamin Silver, Nondelegation in the States, 75 VANDERBILT L. REV. 1211, 1214 (2022), citing Jason Iuliano & Keith E. Whittington, The Nondelegation Doctrine: Alive and Well, 93 NOTRE DAME L. REV. 619,636 (2017) (estimating that ten statutes per year are invalidated on nondelegation grounds). 69 Merrill, [2023] supra note _ at 997 (state courts “have always been stricter about delegations” than federal courts). Note that Merrill here considers the broad language of the Restatement (Second) of Torts, not the more circumscribed language of state public nuisance statues. See note _ [“Whether Merrill…”] supra. 70 Silver, supra note _ at 1224. 71 See Lemos, supra note _ at 428-432 (analyzing common-law statutes as broad legislative delegations to courts). 72 See Margaret H. Lemos, Interpretive Methodology and Delegations to Courts: Are “Common-Law Statutes” Different? in INTELLECTUAL PROPERTY AND THE COMMON LAW, Shyam Balganesh, ed. 9 fn. 61 (2012), citing Pierre N. Leval, Trademark: Champion of Free Speech, 27 COLUM. J. L. & ARTS 197 (2004) (“Regardless . . . of whether the text is detailed or consists of only a vague generalized reference to the body of doctrine, if the intention of the statute was not to make law but to give recognition in statutory form to a previously developed body of court-made law, proper interpretation of the statute demands that it be so understood.”); Thomas W. Merrill, The Common Law Powers of Federal Courts, 52 U. CHI. L. REV. 1, 44 (1985) (“[A]n inference of delegation arises when . . . Congress adopts a legal standard that is borrowed from (or ‘codifies’) the common law.”); Richard A. Posner, Statutory Interpretation—in the Classroom and in the Courtroom, 50 U. CHI. L. REV. 800, 818 (1983) (“If the legislature enacts into statute law a common law concept, . . . that is a clue that the courts are to interpret the statute with the freedom with which they would construe and apply a common law principle—in which event the values of the framers may not be controlling after all.”); Daniel A. Farber & Brett H. McDonnell, ’Is There a Text in this Class?” The Conflict Between Textualism and Antitrust, 14 J. Contemp. Legal Issues 619, 658 (2004) (“One of the main arguments for the delegation theory is that the Sherman 67

5/29/2024]

PUBLIC NUISANCE LITIGATION

established interpretive traditions understand legislative codification of common law using broad, open-ended language as granting courts broad authority to elaborate on statutory meaning using common-law-style reasoning. 73 Merrill’s antipathy to open-ended statutory public nuisance provisions 74 is resonant with efforts to restrict broad Congressional legislative delegations more generally by revising longstanding nondelegation doctrine and developing new rules of statutory construction such as the Major Questions Doctrine. 75 He claims that “public nuisance law, in its recent incarnation as a vehicle for social reform litigation, is out of step with widely shared precepts about the proper assignment of roles among different legal institutions in our society.” 76 To the contrary, Merrill’s attack on broad statutory delegations of common-law power to courts adopts a conception of separation of powers that has long been contested in American legal tradition. 77 He casts himself as a Act uses common law terms, and must therefore contemplate a common-law process of law development.”); Peter S. Menell, Forty Years of Wondering in the Wilderness and No Closer to the Promised Land: Bilski’s Superficial Textualism and the Missed Opportunity to Ground Patent Law Interpretation and Return Patent Law to its Technology Mooring, 63 Stan. L. Rev. 1289, 1311 (2011) (arguing that the common-law history of the Patent Act demands that courts “perpetuate the common law-type jurisprudential tradition”). (Citations and parentheticals in this footnote are from Lemos.) 73 See, e.g., Joseph Story, Codification of the Common Law 706 (1837) in The Miscellaneous Writings of Joseph Story, ed. William W. Story 1852 https://www.google.com/books/edition/The_Miscellaneous_Writings_of_Joseph_Sto/g6AQ AQAAMAAJ?hl=en&gbpv=1&printsec=frontcover, discussed in Charles M. Cook, THE AMERICAN CODIFICATION MOVEMENT: A STORY OF ANTEBELLUM LEGAL REFORM 177 (1981). For further elaboration, see infra Part III.B. 74 See Merrill, [2011] supra note _ at 50 (“Ideally, all public nuisance liability and all statutory references to public nuisance should be repealed. In their place legislatures should enumerate what kinds of activities are prohibited, what penalties apply to violators, and who may institute enforcement actions.” 75 See, e.g., West Virginia v. Environmental Protection Agency, 597 U.S. 697, 780 (2022) (Justice Kagan dissenting, discussing the Court’s use of the Major Questions Doctrine to limit Congressional delegation); Riley T. Svikahart, Note, “Major Questions” as Major Opportunities, 92 NOTRE DAME L. REV. 1873, 1874 (calling on the Court to embrace the Major Questions Doctrine as a means of compensating for “a toothless nondelegation doctrine”). See also, Lemos, [The Other Delegate] supra note _ at 455; Sunstein, supra note _ (both discussing the use of canons of statutory interpretation to limit Congressional delegation) 76 Merrill, [2011] supra note _ at 5. 77 See Cook, supra note _ (detailing ongoing debate throughout the nineteenth century between anti-codifiers, moderates, and radicals concerning the desirability of common law adjudication). See also Morton Horowitz, THE TRANSFORMATION OF AMERICAN LAW, 17801860 22 (1977) (documenting how an instrumentalist view of common law led to a conception of judges as agents of legal change and delegation of a legislative function to courts.); William J. Novak, THE PEOPLES’ WELFARE: LAW AND REGULATION IN NINETEENTH CENTURY AMERICA 38-40 (1996) (describing the widely held view that state statutes should

5/29/2024]

PUBLIC NUISANCE LITIGATION

guardian of widely shared precepts. 78 However, when it comes to public nuisance statutes, he appears to be an advocate for a highly controversial conception of nondelegation advanced by conservative legal scholars and judges seeking to restrict administrative regulation generally. 79 Although dynamic interpretation of broad, open-ended public nuisance statutes might violate the version of the nondelegation doctrine that Merrill would prefer, he provides no evidence that they violate nondelegation doctrine as it is currently applied by courts. Merrill’s seemingly modest canon of nondynamic interpretation of public nuisance law purports to prevent the excessive exercise of policymaking power by courts and restore that power to legislatures. He claims that it merely reflects “an administrative law professor's reaction to the idea that courts, rather than legislatures or administrative agencies, should be regarded as having inherent authority to regulate complex social problems like tobacco smoking, gun control, lead paint removal, MTBE contamination, transboundary pollution, and climate change.” However, his approach might also be fairly characterized as enlarging the power of courts to interpret statutes in ways that deprive legislatures of the power to delegate broad policymaking powers. 80 That is, while purporting to protect the policymaking prerogatives of the legislature, Merrill is, in fact, calling on courts to deprive legislatures of the power of broad delegation to courts to regulate industry via

be interpreted in reference to the common law and that the common law was dynamic and constantly changing to meet new demands). 78 See, e.g., Merrill, [2011] supra note _ at 6 (grounding his arguments against public nuisance claims in “conventional understandings”). 79 See, e.g., Philip Hamberger, Nondelegation Blues, 92 Geo. Wash. L. Rev. 1083 (2023) (arguing that the Constitution bars Congressional transfer of legislative power to other branches of government). See also, Gundy v. United States, supra note _ at 2131 (Justice Gorsuch dissenting). 80 David A. Dana, Public Nuisance Law when Politics Fail, 83 OHIO ST. L. J. 61, 75 (2022) (“because state legislatures have broadly defined public nuisances, it would be contrary to the legislative delegation—it arguably would be undemocratic—for courts to rewrite the statutes and proclaim that they only have authority to declare public nuisances in a very limited set of cases that fit the original quasi-crime conception of nuisance”). See West Virginia v. Environmental Protection Agency, supra note _, at 783-784 (Justice Kagan dissenting, arguing that, by applying the Major Questions Doctrine, the Court deprives Congress of the power to “do its job” by “an ability to delegate power under broad general directives”). See also Lemos, [The Other Delegate] supra note _ at 460 (pointing out that by narrowing the scope of delegation to courts, a court cannot escape the exercise of the type of policy discretion that it seeks to deny itself); Sunstein, supra note _, at 330 (arguing that when courts interpret ambiguous statutory language, “there is no escaping delegation”).

5/29/2024]

PUBLIC NUISANCE LITIGATION

private lawsuits by refusing to recognize the common-law elasticity of broad statutory public nuisance provisions. 81 Merrill’s views also betray a general hostility to common law adjudication. In a footnote to his most recent essay on public nuisance, he makes the remarkable claim that “[t]he assumption that courts have broad discretionary authority to modify the common law is a widespread but largely unexamined premise of the American legal system,” citing Li v. Yellow Cab Co. 82 as an example. 83 It is unclear what Merrill means by “unexamined” given the court’s citation to the widely respected nineteenth-century legal scholar John Norton Pomeroy’s extensive 70-page justification for interpreting the California Code as “primarily and mainly a declaration and enactment of common law rules” that “retain the elasticity, the power of expansion and of adaptation to new facts and circumstances which belong to them in their original form as portions of the ‘unwritten’ common law,—or law promulgated by judicial decisions.” 84 Pomeroy is, of course, merely one voice in a long American tradition of careful examination and defense of the dynamic nature of the common law, spanning from the Founding Era to today and including such prominent jurists as James Wilson, 85 Joseph Story, 86

Lemos 460: conceptually impossible for courts to enforce nondelegation doctrine when the broad delegation is to themselves (Lemos 460), see Dana Public Nuisance Law 14, 25 (?) 82 532 P. 2d. 1226 (1975). 83 Merrill, [2023], supra note _ at 993 fn 39. See also Id. at 994 (criticizing the democratic deficit of common law policymaking); Id. at 992 (citing Frederick Schauer, Is the Common Law Law?, 77 CALIF. L. REV. 455 (1989) (arguing that common law violates fundamental principles of the rule of law). 84 Li v. Yellow Cab Co., supra note _ at 1234. See John Norton Pomeroy, THE “CIVIL CODE” OF CALIFORNIA 55 (1885), citing. 85 James Wilson, THE WORKS OF JAMES WILSON, Robert Green McClosky, ed. 1: 180184 (1967) (celebrating the evolutionary nature of common law). See Novak, supra note _ at 37 (characterizing the conception of common law among early American jurists including Wilson as “dynamic, social, and visionary” in which “rights and duties were inextricably intertwined in a conception of a well-ordered, constantly changing, society”). See also Horowitz, supra note _ at 22 (describing early American views of the common law as an instrument to advance public policy and a conception of judges as agents of legal change). 86 Story, supra note _ at 701-702 (describing common law as “a system of elementary principles and general juridical truths, which are continually expanding with the progress of society, and adapting themselves to the gradual changes of trade, and commerce, and the mechanic arts, and the exigencies and usages of the country”). 81

5/29/2024]

PUBLIC NUISANCE LITIGATION

Benjamin Cardozo, 87 Roger Traynor, 88 Jack Weinstein, 89 and Guido Calabresi. 90 As this distinguished list suggests, the “widely shared precepts about the proper assignment of roles among different legal institutions in our society” in which Merrill roots his attack on broad, open-ended statutory public nuisance provisions are actually contested views that have been the subject of ongoing debate among American jurists and scholars since the Founding Era. Merrill’s views echo the views of nineteenth century radical reformers who envisioned codification as a means of reducing the judicial discretion afforded by common law and replacing it with a comprehensive set of legislative commands—views that, ultimately, did not prevail in the long-running debate over codification. 91 Similarly, Merrill’s assertion that “[c]ourts are duty bound ‘to say what the law is,’ and not to make it up,” 92 reflects a formalism that has arguably been on the defensive against an instrumental conception that sees common-law judges as agents of legal change and common-law adjudication as a means of advancing social policy that also dates to the Founding Era. 93 Benjamin Cardozo, THE NATURE OF THE JUDICIAL PROCESS _ (1921) (judges act as legislators when they fill in gaps in the common law). 88 Roger Traynor, Prepared Observations on “The Courts and Lawmaking” in LEGAL INSTITUTIONS TODAY AND TOMORROW 60 (1959) (endorsing a conception of the common law as “dynamic”). 89 Jack B. Weinstein, INDIVIDUAL JUSTICE IN MASS TORT LITIGATION (1995) (describing creative solutions in adjudicating mass tort cases); Robert L. Rabin, Judge Jack Weinstein and the World of Tort: Institutional and Historical Perspectives, 64 DEPAUL L. REV.641 (2015).; see also divergent views on the democratic deficit of common law (White, rising perception of democratic legitimacy of legislation vs. Fruend greater faith in the legitimacy of court rulings (Fruend 286); Cook at 40 (common law stable & fixed, statutory law unreliable) 90 Guido Calabresi, A COMMON LAW FOR THE AGE OF STATUTES 92 (1982) (defending court-made law). For divergent views on the democratic legitimacy of common law, compare Cook, supra note _ at 40 (describing the view of conservative lawyers that the common law is stable and fixed, whereas statutory interpretation is subject to the lawless discretion of judges) and Ernst Freund, STANDARDS OF LEGISLATION 286-287 (1917) (“in comparing legislative with judicial justice in American it is necessary to bear in mind that the courts represent our best in government while our legislatures do not”) with G. Edward White, LAW IN AMERICAN HISTORY, VOLUME 3, 1930-2000 212-213 (2019) (documenting the rising perception that legislation enjoyed greater democratic legitimacy than common law). 91 Cook, supra note _ at 86-87. In the nineteenth century it was conservatives who championed the common law against radicals who wished to replace it with statutory commands; whereas, today it is conservatives who seek to displace the delegation of common law powers to courts by insisting on replacing them with narrow statutory commands. 92 Merrill, [2023] supra note_ at 995. 93 Horowitz, supra note_ at 1-23; Cook supra note _ at 52; Novak, supra note _ at 38. For contemporary articulations of this instrumentalist conception of tort law, see, e.g., Dobbs et al. supra note _ at §§13-14 (describing the risk deterrence and loss spreading functions of 87

5/29/2024]

PUBLIC NUISANCE LITIGATION

Merrill pretends that his conception of the separation of powers reflects a longstanding consensus, which he describes as “widely shared precepts about the proper assignment of roles among different legal institutions in society.”94 He seeks to leverage this consensus to restore “conventional understandings about public nuisance” that support his cannon of nondynamic interpretation. 95 However, the proper assignment of roles among different legal institutions in the American constitutional structure has always been contested. 96 We reject the assumption shared by many so-called originalists and conservatives “that cultures are so fully integrated that it is possible to speak of a collective mind or a shared set of meaning and values, which would be necessary in order to achieve a return to those values.” 97 Instead, we believe that intellectual history is more accurately characterized as the study of “communities of discourse that are united by a shared set of questions rather than answers.” 98 In his attempt to leverage legal history to advance his claims, Merrill flattens it. As a final note, Merrill’s suggestion that “absent evidence that the legislature intended to ratify the special injury rule, [public nuisance] statutes should not be construed as allowing private actions for either injunctive relief of damages” 99 betrays a rank anti-plaintiff bias. Merrill’s canon of nondynamic interpretation of public nuisance law would require courts, in the absence of a clear statement to the contrary, to “treat the statute as an implicit ratification by the legislature of the common law of public nuisance as of the time the statute was adopted.” 100 The special injury rule has been part of the common law of public nuisance since the sixteenth century 101 and was adopted almost universally in the United States. 102 Merrill’s extensive tort law); Kenneth S. Abraham, THE FORMS AND FUNCTIONS OF TORT LAW, 6th ed. 16-24 (2022) (characterizing the nature and functions of tort law as a mixed system of justice and policy principles); Catherine M. Sharkey, Modern Tort Law: Preventing Harms, Not Recognizing Wrongs, 134 HARV. L. REV. 1423, 1425 (celebrating the instrumental goals of tort law). 94 Merrill, [2011], supra note _ at 5. 95 Id. at 6. 96 See Eric. A. Posner, Balance-of-Powers Arguments and the Structural Constitution, 164 U. PENN. L. REV., 1677 (2016); Nikolas Bowie & Daphna Renan, The Separation-ofPowers Counterrevolution, 131 YALE L. J. 2020 (2022). 97 David Sehat, On Legal Fundamentalism, in AMERICAN LABYRINTH: INTELLECTUAL HISTORY FOR COMPLICATED TIMES, Raymond Haberski, Jr. & Andrew Hartman, eds. 32 (2018). 98 Id. 99 Merrill, [2023] supra note _ at 52. 100 Id. 101 Denise E. Antolini, Modernizing Public Nuisance: Solving the Paradox of the Special Injury Rule, 28 ECOLOGY L. Q. 755, 790 (2001); J.R. Spencer, Public Nuisance—A Critical Examination, 48 CAMBRIDGE L. J. 55, 73-76 (1989); Gifford, supra note _ at 799-800. 102 Antolini, supra note _ at 761.

5/29/2024]

PUBLIC NUISANCE LITIGATION

analysis of why the special injury rule is inconsistent with important distinctions between public and private law suggests that, when it comes to plaintiff standing, Merrill ironically favors a dynamic approach to statutory interpretation of public nuisance statutes that departs from common law at the time the statute was adopted. 103 In sum, we reject Merrill’s canon of nondynamic interpretation of public nuisance law because we believe that his justification for it begs essential questions about the separation of powers, disregards established principles of statutory interpretation, and betrays an anti-plaintiff bias. We turn next to make the affirmative case for interpreting public nuisance statutes dynamically. III. BROAD PUBLIC NUISANCE STATUTES AS COMMON-LAW STATUTES Broad, open-ended public nuisance provisions belong to a category of statutes called common-law statutes. These statutes employ sweeping, general terms that authorize courts to elaborate on their provisions using common-law-style reasoning. 104 In this Part, we briefly describe the features of common-law statutes and demonstrate that broad, open-ended public nuisance statutes share these features. We then show that interpreting public nuisance provisions dynamically is consistent with a longstanding historical tradition that attributes “common-law elasticity” to statutory codification of common law. Next, we argue that common-law statutes, which authorize courts to elaborate on broad, open-ended statutory terms reflect a justifiable legislative choice to regulate through litigation. A. Common-Law Statutes There is no universally agreed-upon or readily identifiable group of common law statutes. However, there is broad consensus that such a class of statutes exists, that statutes within this class share certain distinguishing characteristics, and that the normal rules of statutory interpretation do not apply to statutes within the class. 105 Unlike other statutes, common-law statutes should not be interpreted solely as a product of the time of their enactment or according to strict rules of statutory interpretation. Instead, they should be understood as delegations to courts to Merrill, [2011] supra note _ at 12-16. Lemos, [Interpretive Methodology] supra note _ at 6 (“Perhaps the most conventional way of describing common-law statutes is by reference to how they are written. Commonlaw statutes, we are told, are ‘phrased in sweeping, general terms.’ Congress did not define the precise content of common-law statutes, but instead ‘expect[ed] the federal courts to interpret them by developing legal rules on a case-by-case basis in the common law tradition.’” (citing Guardians Assoc. v. Civil Serv. Comm’n 463 U.S. 582, 641 & n. 12 (1982) (Stevens, J., dissenting)). 105 Lemos, [Interpretive Methodology], supra note _ at 1. 103 104

5/29/2024]

PUBLIC NUISANCE LITIGATION

develop the law and fill in statutory gaps using the type of policy-informed analysis that is a feature of common-law adjudication. 106 The paradigmatic example of a common-law statute is the Sherman Antitrust Act. The primary substantive provision of the Sherman Act declares that “[e]very contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal.” 107 None of the operative terms in this provision is defined or circumscribed in the Sherman Act’s original text. Consequently, this language requires interpretation and gap-filling. Indeed, read literally, this sweeping provision could outlaw every private contract. Given the implausibility of such a reading, courts have “long held that the operative question is whether the conduct at issue restrained trade unreasonably.” 108 This standard “plainly invites [yet further] judicial policymaking,” since it demands “elaboration to clarify its contents.” 109 Judge Frank Easterbrook explains that broad, open-ended statutes like the Sherman Act “effectively authorize courts to create new lines of commonlaw.” 110 William Baxter, Assistant Attorney General in Charge of the Department of Justice Antitrust Division during the Regan Administration, enumerates three characteristics that make the Sherman Act a common-law statute. First, the act’s extensive “jurisdictional reach … allows courts to scrutinize the full range of business conduct.” 111 Second, “the substantive terms within the statute are either common-law origin or otherwise readily susceptible to judicial interpretation.” 112 Third, “Congress provided little if any extrastatutory guidance to direct interpretation of the basic antitrust

Hillel Y. Levin & Michael L. Wells, Qualified Immunity and Statutory Interpretation: A Response to William Baude, 9 CAL. L. REV. ONLINE 40, 45–46 (2018) (citing Lemos, supra note __, at 91); William N. Eskridge, Public Values in Statutory Interpretation, 137 U. PENN. L. REV. 1007, 1052 (1989) (“The Court's invocation of common law principles to fill in gaps within fairly detailed statutes such as FELA is a regular occurrence, but the meta-rule is even more critical for several older, generally worded federal statutes that Congress has substantially left to the courts to develop. These ‘common law statutes' include Section 1983 and several of the other civil rights measures enacted after the Civil War, the Sherman Act of 1890, the anti-fraud provisions of the securities laws, and Section 301 of the Taft-Hartley Act of 1947.”). On the role of policy analysis in commonlaw reasoning, see Melvin Eisenberg, LEGAL REASONING 41-59 (2022). 107 15 U.S.C. §1. 108 Lemos, [Interpretive Methodology] supra note _ at 3. 109 Id. 110 Frank Easterbrook, Statutes’ Domains, 50 U. CHI. L. REV. 533, 544 (1983). 111 William F. Baxter, Separation of Powers, Prosecutorial Discretion, and the “Common Law” Nature of Antitrust Law, 60 TEX. L. REV. 661, 663-664 (1982). 112 Id. at 664. 106

5/29/2024]

PUBLIC NUISANCE LITIGATION

provisions,” relying instead on judges to work out how best to advance the statute’s broad goals of competition and free enterprise. 113 The antitrust laws were written with awareness of the diversity of business conduct and with the knowledge that the detailed statutes designed to prohibit socially undesirable conduct would lack the flexibility needed to permit (and at times even encourage) desirable conduct. As one leading antitrust scholar explains, as a common-law statute, “the Sherman Act can be regarded as ‘enabling’ legislation—an invitation to the federal courts to learn how businesses and markets work and formulate a set of rules that will make them work in socially efficient ways.” 114 As Professor Margaret Lemos puts it, “common-law statutes represent delegations to courts” that “authorize courts to make law.” 115 To do so, they are “written in broad terms and build on a tradition of commonlawmaking” 116 Even Justice Scalia, not one to look favorably on loose judicial interpretation of statutory texts, maintained that common-law statutes like the Sherman Act must be interpreted and applied by judges in a dynamic manner. Writing for the Court in Business Electronics Corp. v. Sharp Electronics Corp, he explained that the Sherman Act “adopted the term ‘restraint of trade’ along with its dynamic potential. It invokes the common law itself, and not merely the static content that the common law had assigned to the term in 1890.” 117 Like the Sherman Act and other common-law statutes, broad, open-ended statutory public nuisance provisions authorize courts to use common-lawstyle reasoning to define the scope of prohibited conduct within generally defined categories. Merrill anticipates this suggestion, writing, “[c]onceivably, these statutes could be interpreted as delegating authority to courts to develop the scope of public nuisance law in the common-law fashion” and comparing them to the Sherman Act. 118 However, he rejects this possibility because “delegation of such common-law authority to courts is rare.” 119 Instead, he declares it “far more plausible [to suppose] that the legislatures enacted public-nuisance statutes because they intended to preserve public authority to bring actions to abate the kinds of conditions that had long been understood to be public nuisances (e.g., blocking highways) 113

Id. Herbert Hovenkamp, ECONOMICS AND FEDERAL ANTITRUST LAW 52 (student ed. 1985). See also Baxter, supra n. __ at 661. 115 Lemos, [Interpretive Methodology] supra note _ at 2. For examples of other federal common-law statutes, see Lemos, [The Other Delegate] supra note _ at 429-430 fn. 122. 116 Lemos, [Interpretive Methodology] supra note _ at 2. 117 485 U.S. 717, 732 (1988). 118 Merrill, [2023] supra note _ at 996. 119 Id. 114

5/29/2024]

PUBLIC NUISANCE LITIGATION

and to indict activities added to the conventional list by the legislature (e.g., storing gunpowder in cities).” 120 But these public nuisance statutes certainly look like textbook commonlaw statutes. Indeed, they possess Baxter’s three defining characteristics of common-law statutes. First, they have extensive jurisdictional reach that allows courts to scrutinize the full range of conduct that has the potential to compromise public health, safety, and welfare, depending on the language of the statutory provision. 121 Second, substantive terms such as “injurious,” “offensive,” and “duty” derive from common-law concepts and are readily susceptible to judicial interpretation. 122 Third, state legislatures typically provide little if any extrastatutory guidance to direct interpretation of public nuisance provisions. Moreover, the broad, open-ended language of statutory public nuisance provisions arguably suggests that they were written with awareness of the diversity of antisocial conduct and with the knowledge that detailed statutes would lack the flexibility needed to discourage new forms of antisocial conduct. 123 Finally, it is unclear by what measure Merrill considers common-law statutes “rare” or why that would be reason to believe that broad, open-ended public nuisance statutes should not be counted among them. In the next section, we find additional support for treating public nuisance statutes as common-law statutes in longstanding interpretive traditions regarding codification of the common law. B. Codification and “Common-Law Elasticity” In the United States, the first broad, open-ended public nuisance statutes were enacted in the nineteenth century. 124 Legal historian William Novak has extensively documented how the common law of public nuisance served as the primary instrument for regulating public health, safety, commerce, order, 120

Id. See, e.g., Indiana Code 32-30-6-6, supra note _; New Mexico Stat § 30-80-1, supra note _; supra notes _ - _ [Part I.A.] and accompanying discussion (providing examples of statutory public nuisance provisions). 122 E.g., Cal. Civ. Code §3479 (“Anything which is injurious to health, including, but not limited to, the illegal sale of controlled substances, or is indecent or offensive to the senses … is a nuisance.”); Rev. Code of Washington 7.48.120 (“Nuisance consists in unlawfully doing an act, or omitting to perform a duty ….”). 123 See, Novak, supra note _ at 44 (“nuisance law was one of the most important regulatory tools of the nineteenth-century American state”); Id. at 37 (characterizing the conception of common law among early American jurists including Wilson as “dynamic, social, and visionary” in which “rights and duties were inextricably intertwined in a conception of a well-ordered, constantly changing, society”); Id. at 38-40 (1996) (describing the widely held view that state statutes should be interpreted in reference to the common law and that the common law was dynamic and constantly changing to meet new demands). 124 Supra notes _ - _ and accompanying discussion. 121

5/29/2024]

PUBLIC NUISANCE LITIGATION

and morality in the United States throughout the nineteenth century.125 Although nineteenth-century attitudes towards the common law were by no means uniform, Novak and others demonstrate that legal professionals and laypersons alike defended the common law as an instrument of social regulation because of its consonance with established community norms, its adherence to enduring legal principles, its adaptability to changing circumstances, and its capacity to advance public values. 126 This high regard for the common law informed the insistence by many early American jurists that statutes should be interpreted in reference to common law. 127 Legal historian Charles Cook chronicles how many advocates of codification in Antebellum America conceived of it “as a means of better expressing, fixing, and actually strengthening the common law by putting it into a more permanent and exacting form” while “the content would largely remain the same.” 128 According to this widely held view, championed by prominent legal figures such as Joseph Story, [A] code would be treated as a clear and systematic, but perforce incomplete, statement of the common law that judges could authoritatively interpret, expound, and correct as they had before the law was codified. In other words, the decisions of the courts would continue to be creative and have the same obligatory force. Of course, this meant that a corpus of decisional law would develop subsequent to codification ….” 129 This vision of codification was by no means universal. Cook describes sharp disagreement between anti-codifiers, who feared that statutory law would corrupt the integrity of common law, radicals, who envisioned statutory law as entirely replacing the common law, and moderates, like Story, who believed that codification would provide a new, better organized form of state statutory and common law to make them more accessible to judges, Novak, supra note _. See also Freund, supra note _ at 66 (“the law of nuisance is the common law of the police power, striking at all gross violations of health, safety, order, and morals,”). 126 See generally, Novak, supra note _ at 37-41. See also, Horowitz Transformation 1922; Cook, supra note _ at 52. 127 Novak, supra note _ at 38 fn. 95, quoting Founding Era jurists James Kent (“Statutes are likewise to be construed in reference to the principles of the common law, for it is not presumed the legislature intended to make any innovation upon the common law, further than the case absolutely required), James Sullivan (“If we examine the laws now existing, as acts of Legislature in the State, we shall find the use of them principally designed for the directing and rendering efficacious, the system of the Common Law”) and Francis Hilliard (“With regard to the statutes of the State legislatures, it may be remarked that the object and effect of them Is not in general to alter or abrogate the principles of common law”). 128 Cook, supra note _ at 81. 129 Id. at 82. 125

5/29/2024]

PUBLIC NUISANCE LITIGATION

lawyers, and the public, while preserving the common law in “both content and technique” as the rule and method of decision. 130 In the end, most jurisdictions engaged in partial codification along the lines set forth by moderates. 131 Writing in 1884, leading legal scholar John Norton Pomeroy insisted that “the true method and principles of interpretation” of the California Civil Code, first adopted in 1872, require judges to treat it “primarily and mainly as a declaration and enactment of common law rules.” 132 Pomeroy believed that common law doctrines, “although clothed with a statutory form … would still retain the elasticity, the power of expansion and of adaptation to new facts and circumstances, and the comprehensiveness which belonged to them in their original form as portions of the ‘unwritten’ common law,—or law promulgated by judicial decisions.” 133 California courts adopted Pomeroy’s approach to the code in 1888, and the legislature enshrined them in the code itself in 1901, stating in Section 5: “The provisions of this Code, so far as they are substantially the same as existing statutes or the common law, must be construed as continuations thereof, and not as new enactments.” 134 In the 1975 case, of Li v. Yellow Cab Co., the California Supreme Court reaffirmed this understanding of the code, explaining that it was not the intention of the Legislature in enacting section 1714 of the Civil Code, as well as other sections of that code declarative of the common law, to insulate the matters therein expressed from further judicial development; rather it was the intention of the Legislature to announce and formulate existing common law principles and definitions for purposes of orderly and concise presentation and with a distinct view toward continuing judicial evolution. 135

Id. at 177. Id. at 209-210, 177. 132 Pomeroy, supra note _ at 6, 50. 133 Id. at 55. 134 Lewis Grossman, Codification and the California Mentality, 45 HASTINGS L. J. 617, 620 (1994) (citing Sharon v. Sharon, 16 P. 345, 354-355 (Cal. 1888) and quoting Cal. Civ. Code §5 (1901)). See also OKLA. STAT. title 12 § 2 (“The common law, as modified by constitutional and statutory law, judicial decisions and the condition and wants of the people, shall remain in force in aid of the general statutes of Oklahoma.”). 135 532 P.2d 1226, 1233 (1975). See also Id. at 1238 (“the peculiar nature of the 1872 Civil Code as an avowed Continuation of the common law has rendered it particularly flexible and adaptable in its response to changing circumstances and conditions”) (“the code's incompleteness, both in scope and detail, have provided ample room for judicial development of important new systems of rules, frequently built upon Code foundations.” quoting Arvo Van Alstyne, WEST’S ANNOTATED CIVIL CODE 36-37 (1954) (internal quotations omitted)). 130 131

5/29/2024]

PUBLIC NUISANCE LITIGATION

This view of codification as a means of organizing and clarifying the common law but not freezing it implies that code provisions have what one leading commentator called “common law elasticity.” 136 Common-law elasticity allows for judicial evolution that may expand or constrain the scope of statutory public nuisance provisions. Courts may expansively apply public nuisance law to unspecified forms of conduct or conditions. Conversely, the common-law elasticity of public nuisance statutes also may instead lead courts to interpret them more narrowly than a literal reading would suggest. For example, although Indianna’s nuisance statute 137 makes no reference to reasonableness, Indiana courts have interpreted it, since its adoption in 1881, to include an unstated and narrowing common-law reasonableness standard. 138 Similarly, courts today routinely interpret broad statutory nuisance provisions to conform with constraints articulated in the Restatement (Second) of Torts analysis of common-law public nuisance doctrine. 139 Our claim in this section is that the contemporary application of broad, open-ended statutory public nuisance provisions to conduct or conditions unknown at the time of enactment—for example, modern mass-marketing strategies or distribution practices by manufacturers that are injurious to public health, safety, or welfare—are consistent with a well-established tradition regarding the proper interpretation of statutory codification of common law. When it comes to statutory causes of action, “the new public nuisance” has an historical pedigree that Merrill and other critics have failed to acknowledge. 140 The common-law elasticity of broad, open-ended public nuisance statutes seems to us a more well-grounded “conventional understanding of public nuisance” than the highly controversial conception of separation-of-powers invoked by Merrill to justify his canon of nondynamic interpretation of public nuisance law. 141 C. Delegating Regulation by Litigation We have argued that broad, open-ended statutory public nuisance provisions are common-law statutes. As such, they should be understood as Van Alstyne, supra note _ at 36-37, quoted in Li v. Yellow Cab, supra note _ at 1234. Indianna Code 32-30-6-6. 138 Gary ex rel. King v. Smith & Wesson, supra note _ at 1230. 139 See, e.g., Ileto v. Glock, supra note _ at 1210. 140 See Merrill, [2022] supra note _ at 986 (“the new public nuisance”); Gifford, supra note _ at 746 (the “new” public nuisance tort”); Schwartz, Goldberg & Schaecher, supra note _ at 645 (“an entirely new and unfounded cause of action”); Epstein, supra note _ at 306 (“the new theory” of public nuisance). 141 See supra notes _ - _ and accompanying discussion. 136 137

5/29/2024]

PUBLIC NUISANCE LITIGATION

legislative “delegations to courts” that “authorize courts to make law.” 142 That is, these statutory provisions delegate to courts the power to implement regulation by litigation. In considering the legitimacy of common-law statutes as a regulatory instrument, Lemos observes that delegations to courts should be subject to the same constitutional restrictions that apply to delegations to executive agencies. 143 She explains that “[p]ro-delegation commentators offer functional arguments in defense of agency lawmaking, focusing on institutional characteristics of agencies that make them valuable partners in the lawmaking enterprise-and in some respects better lawmakers than Congress.” These characteristics include “expertise, accountability, accessibility, and flexibility.” 144 For pro-delegation commentators, “the Court’s refusal to enforce the nondelegation doctrine directly in the administrative context is not a necessary evil but a positive good.” 145 On this account, “the costs of nonenforcement of the nondelegation doctrine [are] outweighed by the benefits of the administrative state.” 146 However, according to Lemos, “delegations to courts cannot be defended on the same functional grounds as delegations to agencies” because courts lack the institutional characteristics that justify delegation to agencies. 147 “If broad delegations to courts are to be tolerated,” she concludes, “it must be on grounds other than those typically used to defend the work of agencies.”148 Her analysis places the burden on those seeking to justify broad delegation to courts: “It may be that delegations to courts can be defended on different functional grounds, focusing on the unique institutional capacities of courts. But our current nondelegation law lacks such a theory, and the theories that have been offered to justify delegations to agencies do not work for courts.”149 We agree with Lemos that delegation to courts requires justification no less than does delegation to agencies, and in this Section, we take on the challenge of providing one. We suggest that civil litigation has institutional virtues different from, but comparable in force to, those of agency adjudication and rulemaking that support functional arguments capable of justifying delegation to courts. To highlight these virtues, we contrast them to

Lemos, [Interpretive Methodology] supra note _ at 2. Lemos, [Other Delegate] supra note _ at 443. 144 Id. at 445-455. 145 Id. at 444-445. 146 Id. at 454. 147 Id. at 445. 148 Id. at 445. 149 Id. at 455. 142 143

5/29/2024]

PUBLIC NUISANCE LITIGATION

the unfavorable characterization of public nuisance litigation advanced by Merrill and reflective of sustained criticism by other scholars. 150 Merrill briefly considers and dismisses the functionalist idea that public nuisance litigation can be justified as a catalyst to stimulate new legislation and administrative regulation. 151 He asserts that it is illegitimate for courts to entertain a public nuisance claim “in the belief that it will serve as a ‘catalyst’ for what the judge thinks is the desirable course of future political change.”152 Moreover, he contends that public nuisance litigation has failed to stimulate intended policy change. 153 For example, Merrill points out that lawsuits against gun makers elicited a legislative backlash that resulted in sweeping immunity for the industry against civil lawsuits. 154 When it comes to tobacco and opioid litigation, he concedes that “the discovery of damaging facts during the litigation” helped create “a climate more favorable to administrative regulation” but cautions that “one cannot always expect civil discovery to yield up facts that will sway public opinion” and contends that “the causal connection between the litigation and enhanced administrative regulation is not entirely clear. 155 “Arguably,” he speculates, “rising public dismay about the depth and breadth of the opioid crisis has been the driving force behind both the litigation and the invigorated administrative regulation.” 156 Merrill also denounces the political economy of public nuisance litigation. “The most startling feature of the new-public-nuisance litigation is that its objective is to obtain damages.” 157 He asserts that the prospect of large settlements and damage awards has fostered “joint ventures between public prosecutors and personal-injury firms” that promise big payouts for private attorneys and political prestige for public officials. 158 Although “[s]ome of these actors may also hope that their efforts will serve as a catalyst for regulatory change,” argues Merrill, “the dominant motivation is a type of rentseeking,” which means that “the only defendants that get sued are corporations with fat balance sheets” while “other bad actors—like pill-mill doctors and drug smugglers—do not get sued because they do not have enough assets or because it would be too hard to collect a judgment if one

See, e.g., Gifford, supra note _; Schwartz, Goldberg & Schaecher, supra note _; Epstein, supra note _. 151 Merrill, [2022] supra note _ at 365, 370. 152 Id. at 366. 153 Id. at 366. 154 Id. at 366. 155 Id. at 366. 156 Id. at 366. 157 Merrill [2023] supra note _ at 1002, Merrill, [2022] supra note _ at 367. 158 Merrill, [2023] supra note _ at 1003-1004, 1007. 150

5/29/2024]

PUBLIC NUISANCE LITIGATION

were obtained.” 159 In Merrill’s estimation, “[t]his novel form of joint venture has seen several failures: the lead-paint and gun litigation largely went nowhere,” while the tobacco and opioid settlements have done little more than encourage additional litigation. 160 Merrill asserts that reliance on public nuisance litigation to regulate risk is anachronistic. He grants that authorizing public nuisance litigation might have been an attractive regulatory option to state legislatures during the nineteenth and early twentieth centuries when states lacked administrative capacity but that this “litigation model” was gradually replaced by administrative regulation, which was better suited to address increasingly complex risks that emerged in the late twentieth and early twenty-first centuries. 161 Regulation by litigation, he concludes, is “woefully inadequate” to provide “meaningful regulation” of the complex risks that have emerged in the late twentieth and early twenty-first centuries. 162 Merrill declares boldly: “Today, no one would think of responding to a new type of risk by urging the legislature to label the risk a public nuisance and entrust prosecutors and judges to use litigation to eliminate the risk.” 163 As a matter of fact, they would. In 2003, state legislatures New York, California, New Jersey, Maryland, Delaware, Washington, and Hawaii passed laws deeming it a public nuisance for a gun manufacturer to sell weapons without requiring downstream distributors and retailers of its products to institute “reasonable controls” to reduce the risk of illegal misuse of its products and authorizing public officials and private parties to institute actions for damages and injunctive relief. 164 Although these public nuisance statutes target firearm sales, they delegate to courts the power to use common-law-style reasoning to specify the meaning of “reasonable controls.” More importantly, they reflect the ongoing vitality of using open-ended public nuisance statutes as a regulatory instrument. The following responses to Merrill’s dim view of public nuisance litigation highlight some of the reasons why. Merrill misconstrues the nature of litigation as a catalyst to policy change. Plaintiffs in public nuisance claims do not ask judges to assess the capacity of litigation to prompt legislative or administrative action; they ask judges to Merrill, [2022] supra note _ at 367; Merrill, [2023] supra note _ at 1005, 1008. Merrill, [2023] supra note _ at 1004. Merrill poses additional objections based on overdeterrence (Merrill, [2023] supra note _ at 1001) and vagueness (Id. at 987) which we do not address in this Essay. 161 Merrill, [2002] supra note _ at 361-364 (characterizing the litigation model as “woefully inadequate”). 162 Id. at 361, 365. 163 Id. at 370. 164 E.g., N.Y. GEN. BUS. LAW Art. 39-DDDD, Sec. 898-A-E. 159 160

5/29/2024]

PUBLIC NUISANCE LITIGATION

recognize that the defendant’s conduct constitutes a wrong that caused harm. 165 Public nuisance claims are catalysts to policy change insofar as they make a convincing case that wrongful conduct on the part of the defendant caused harm to the plaintiff. Public nuisance litigation frames policy problems in ways that focus attention on the role of manufacturers’ wrongful conduct in causing gun violence, residential lead poisoning, opioid addiction, ecigarette use among minors, and the psychological damage that social media inflicts on children. 166 The civil discovery process offers a means of uncovering policy-relevant information closely held by manufacturers concerning what they know about the relationship between their design choices, marketing strategies, and distribution practices on the one hand and these public health problems on the other. 167 The media attention attendant to public nuisance litigation also plays a role in holding the attention of the public and increasing or maintaining the salience of public health issues on legislative agendas. 168 Of course, judges will not always agree that a plaintiff’s allegations are sufficient to establish a claim of wrongdoing, and juries may not believe that the plaintiff’s evidence is convincing proof of wrongdoing or causation. 169 The point is that in asking judges and juries to evaluate such claims of wrongdoing, the litigation process itself—through issue framing, information generation, and agenda setting—can and has, in some instances, influenced policymaking outside of the courtroom. 170 See John C. P. Goldberg & Benjamin C. Zipursky, Torts as Wrongs, 88 TEX. L. REV. 917 (2010); Scott Hershovitz, Treating Wrongs as Wrongs: An Expressive Argument for Tort Law, 10 J. TORT L. 1 (2017). 166 See Lytton, supra note _ at 1861 (discussing the framing function of tort litigation). 167 See Wendy Wagner, Stubborn Information Problems & the Regulatory Benefits of Gun Litigation, in SUING THE GUN INDUSTRY: A BATTLE AT THE CROSSROADS OF GUN CONTROL AND MASS TORTS, Timothy D. Lytton, ed., 271 (2005) (analyzing the potential of tort litigation to reveal policy-relevant information through the civil discovery process). 168 See Lytton, supra note _ at 1861 (documenting the agenda-setting function of tort litigation). 169 E.g., Timothy D. Lytton, Halberstam v. Daniel and the Uncertain Future of Negligent Marketing Claims against Firearms Manufacturers, 64 BROOK. L. REV. 681, 682 (1998) (describing jury finding of insufficient evidence of proximate cause in claim by shooting victim against negligent firearm manufacturer). 170 For a detailed analysis of the mechanisms by which tort litigation frames issues, uncovers policy-relevant information, influences policy agendas, and mobilizes stakeholders, see Timothy D. Lytton, A Systems Theory of Tort Law (in progress), See also Nora Freeman Engstrom & Robert L. Rabin, Pursuing the Public Health Through Litigation: Lessons from Tobacco and Opioids, 73 STAN. L. REV. 285 (documenting role of tobacco and opioid litigation as a catalyst to policy change); Denise Antolini, Attacking Bananas and Defending Environmental Law, 58 CASE WESTERN RESERVE L. REV. 663, 664 (2008) (describing the “use of common law litigation as a catalyst for policy change.”) Merrill rightly points out that this influence can sometimes backfire on plaintiffs—as it did in the case of lawsuits against gun makers. Merrill, [2023] supra note _ at 998 fn. 64. Notably, the states’ legislative response in this case was to immunize a single industry, not to repeal their open-ended public 165

5/29/2024]

PUBLIC NUISANCE LITIGATION

In suggesting that the “objective” of public nuisance litigation is to obtain damage awards, Merrill obscures an important distinction between the business model of public nuisance litigation and its function as a regulatory instrument. The availability of contingency fees and the prospect of large damage awards motivate plaintiffs’ attorneys to file lawsuits. They have, in many cases, emboldened plaintiffs’ attorneys to sue powerful institutions that use their influence to stymie popular legislation, discourage aggressive enforcement, or blunt media criticism. 171 The resulting litigation process has, in some cases, catalyzed policy change. 172 In response to Merrill’s concern that collusion between private plaintiffs’ firms and public officials has led to blaming large corporations for major social problems, we think that this might be viewed as a corrective. In many instances, public officials blame small targets more visibly connected to harm—e.g., street criminals for gun violence, landlords for lead paint poisoning, doctors for opioid addiction, schools for e-cigarette use, and parents for social media use—while avoiding confrontation with powerful industries whose conduct may be an important source of the problem that is harder to trace given the complexity of causation. 173 Moreover, partnerships nuisance statutes. We take this as some evidence that state legislatures did not want to throw the baby out with the bathwater—that is, they wished to carefully circumscribe the application of open-ended public nuisance statutes to immunize gun makers while otherwise leaving it unchanged. 171 See, e.g., HOLDING BISHOPS ACCOUNTABLE: HOW LAWSUITS HELPED THE CATHOLIC CHURCH CONFRONT CLERGY SEXUAL ABUSE (2008) (documenting how contingency fee attorneys filed lawsuits to hold church officials liable for covering up child sexual abuse after years during which church officials successfully blocked legislation, criminal investigation, and media coverage). On the use of public nuisance litigation as a response to agency capture and other forms of regulatory failure, see Kendrick, supra note _ at 709-711, 778-787; Dana, supra note _ at 3; Albert C. Lin, Dodging Public Nuisance, 11 U. C. IRVINE L. REV. 489, 531 (2020). 172 See supra note _ [For a detailed analysis…]. 173 See, e.g., Samuel Peterson & Shawn Bushway, Law Enforcement Approaches for Reducing Gun Violence, RAND, April 22, 2020, https://perma.cc/F2AL-7UN3 (cataloguing gun violence prevention policies that target criminals and retail sellers but not manufacturers). We believe that the partnership between state attorneys’ general and plaintiffs’ attorneys seems less nefarious when viewed in the larger context of public-private partnerships that characterize much of the regulatory system at both the federal and state levels. For example, government agencies rely on private firms or organizations to help them implement inspection regimes in the certification of professional services, hospitals, and universities. See Lesley K. McAllister, Harnessing Private Regulation, 3 MICH. J. ENVTL. & ADMIN. L. 291 (2014). Regulators are aware that these arrangements carry with them potential conflicts of interest, and they create safeguards to address this problem. See, e.g., Timothy D. Lytton & Lesley McAllister, Oversight in Private Food Safety Auditing: Addressing Auditor Conflict of Interest, 2014 WISC. L. REV. 289 (2014) (surveying and comparing various mechanisms to counteract conflict of interest by private third-party food safety inspectors).

5/29/2024]

PUBLIC NUISANCE LITIGATION

between public prosecutors and private plaintiffs’ firms that empower public prosecutors counterbalance corporate power, especially in cases of industrywide misconduct, where a small group of well-organized companies with a lot at stake exercise considerable influence over issue framing, information dissemination, legislative agendas and public enforcement. 174 Given the existence of broad, open-ended public nuisance statutes on the books and the evidence of interest-group influence on the legislative and regulatory process, we believe that the burden of repealing current statutes or securing industry-specific immunity should fall on industry. Examples of such legislative responses in the context of public nuisance include right-tofarm statutes 175 and gun industry immunity laws. 176 This seems a more democratic approach to determining the scope of public nuisance statutes than the use of restrictive interpretive canons by courts. 177 Finally, we believe that Merrill’s suggestion that regulation by litigation is anachronistic in an age of administrative regulation suffers from an idealization of modern administrative regulation. In practice, legislatures routinely charge administrative agencies with broad regulatory responsibilities and provide inadequate resources for implementation, enforcement, and policy evaluation, notwithstanding Merrill’s blithe suggestion that “insofar as enforcement of public rights has been increasingly given over to administrative bodies, only the legislature has the resources to create and fund the appropriate administrative offices.” 178 Merrill’s idealization of administrative regulation is an inaccurate description of a reality in which limited agency resources mean that a great deal of regulatory compliance activity is private and decentralized through managerial oversight, supply chain management, and insurance underwriting—much of which is motivated by liability exposure. 179 Merrill’s conception of regulation as a tri-partite enterprise between legislatures, executive agencies, and courts fails to capture the full spectrum of institutional actors in regulatory governance. Public nuisance litigation, especially that brought by private See James Q Wilson, THE POLITICS OF REGULATION (1980) (describing how concentrated interests form groups that block popular legislation that benefits diffuse stakeholders). 175 See, e.g., IOWA CODE Title 9, Section 352.11. 176 Gun Industry Immunity, GIFFORDS LAW CENTER, https://perma.cc/F2AL-7UN3 (last visited May 23, 2024) (surveying federal and state firearms industry immunity statutes). 177 Compare Merrill, [2011] supra note _ at 31-32 (invoking the “democratic ethos” of legislation to justify a canon of nondynamic interpretation of public nuisance law aimed at narrowing the scope of broad, open-ended statutory provisions). 178 Merrill, [2011] supra note _ at 35. But see Peter H. Shuck, WHY GOVERNMENT FAILS SO OFTEN AND HOW IT CAN DO BETTER 62 (2014) (describing how chronic underfunding of agencies charged with broad regulatory mandates is a common cause of regulatory failure). 179 See Lytton, [Systems Theory] supra note _. 174

5/29/2024]

PUBLIC NUISANCE LITIGATION

litigants, should be evaluated in light of this reality of public-private “risk regimes.” 180 We do not purport here to offer a full-throated or unqualified defense of public nuisance litigation as a regulatory instrument. A complete accounting of the benefits of regulation by litigation is beyond the scope of this Essay, and readers seeking a more detailed analysis should consult sources cited in the footnotes. Moreover, we offer generalized claims about the institutional advantages of regulation by litigation recognizing that, like all regulatory instruments, civil litigation performs better in some contexts than others and that comparative institutional analysis should be sensitive to context. 181 We offer this cursory presentation and reliance on general claims in service to a specific aim. By highlighting the virtues of broad, open-ended public nuisance statutes as a regulatory instrument, we have outlined a functional justification for common-law statutes, and we have argued that the legislative choice to regulate through litigation is no less defensible than broad delegation of policymaking discretion to administrative agencies. Moreover, we have suggested that the virtues of regulation by litigation are especially well suited to identifying, exposing, and addressing industrywide misconduct that harms public health and safety. CONCLUSION In this Essay we have defended several interrelated claims regarding statutory public nuisance lawsuits against product manufacturers. We argued that broad, open-ended statutory public nuisance provisions should be understood as common-law statutes that authorize courts to apply them to new situations using common-law-style reasoning. We demonstrated that this view of statutory public nuisance law is consistent with well-established interpretive traditions regarding the codification of common law. We explained why the use of these statutes to delegate regulatory power to courts through the adjudication of lawsuits is no less justified than delegations to administrative agencies to exercise policy discretion through adjudication and rulemaking. Taken together, these claims support the use of statutory public For foundational work on the concept of risk regimes, see CHRISTOPHER HOOD, HENRY ROTHSTEIN & ROBERT BALDWIN, THE GOVERNMENT OF RISK: UNDERSTANDING RISK REGULATION REGIMES (2001); Jody Freeman, Private Parties, Public Functions and the New Administrative Law, 52 ADMIN. L. REV. 857 (2000). See also David Levi-Faur, Regulation and Regulatory Governance, in HANDBOOK ON THE POLITICS OF REGULATION 13-14 (David Levi-Faur, ed., 2011); Colin Scott, Regulating Everything: From Mega- to Meta-Regulation, 60 ADMINISTRATION 61, 67 (2012). 181 See Timothy D. Lytton, Using Tort Litigation to Enhance Regulatory Policy Making: Evaluating Climate-Change Litigation in Light of Lessons from Gun-Industry and ClergySexual-Abuse Lawsuits, 86 TEX. L. REV. 1837 (2008) (applying empirical metrics for contextbased comparative institutional analysis). 180

5/29/2024]

PUBLIC NUISANCE LITIGATION

nuisance claims against product manufacturers to address harms to public health, safety, and welfare caused by unreasonably dangerous design decisions, marketing strategies, and distribution practices. There are several important and relevant issues that we have not discussed in this Essay and that merit consideration. We have not addressed concerns that the open-ended nature of statutory nuisance provisions violates constitutional norms of vagueness. 182 Nor have we answered charges that public nuisance claims circumvent well-established product liability doctrines designed to limit recovery against manufacturers for product-related harms. 183 Additionally, our account remains uncommitted on the question of whether the scope of statutory public nuisance provision is coextensive with common-law public nuisance. 184 We hope to address these important issues in future work. In addition to developing a more complete analysis of public nuisance lawsuits against product manufacturers for injury to public health, safety, and welfare, we plan to extend our analysis to other common-law statutes—for example, to unfair trade practices acts. Public nuisance claims against product manufacturers are typically accompanied by statutory claims alleging unfair trade practices. The proper resolution of these lawsuits requires more than merely ordinary statutory construction or the application of common-law doctrines. It requires a deeper understanding of common-law statutes and the virtues of regulation by litigation.

See e.g., Merrill, [2011] supra note _ at 26-28; Gifford, supra note _ at 786-787. See, e.g., Schwartz, Goldberg at Schaecher, supra note _ at 630. 184 See Goldberg, supra note _ at 46 (arguing that the power of a legislature to declare conduct or conditions a public nuisance is constrained by the scope of the concept of public nuisance in common law). 182 183

Microlocal Litigation for a Mass Tort World Sarah L. Swan† Draft – Please Do Not Cite

When public entities began suing companies for creating harmful products like tobacco and asbestos, it was mainly states leading the efforts. Cities soon joined the fray though, eventually emerging as their own significant force bringing numerous claims to remedy the harms caused by products like guns, lead paint, and opioids. More recently, even smaller local government units have been quite active. School districts have sued vaping and social media companies, water districts have sued industrial manufacturers for polluting drinking water supplies, and very small, rural localities have sued for big harms like climate change. Further, some very small localities have gone even further, and attempted to change the litigation playing field itself, by empowering additional entities like ecosystems to bring legal actions, and by establishing new statutory causes of action. This Article explores these microlocal litigative actions, and argues that the bold moves of these tiny governments may offer a promising source of solutions for widespread problems. Introduction………………………………………..………..........2 I. Microlocal Plaintiff Litigation………..………..…………….4 A. School District Suits….….…………………………..5 i. Litigation against Vaping Companies………....5 ii. Litigation against Social Media Defendants...10 B. Additional Microlocal Plaintiff Suits …………....15 II. Microlocal Changes to the Litigation Landscape………17 A. Stripping Corporate Rights……….……………….18 B. Empowering New Parties and New Basis for Claims……………………………………………...…19 III. Legitimacy and Democracy in Microlocal Litigation…26 A. Alignments Between the Harm and the Microlocal Entity……………………………………………………………..27 † Professor of Law and Dean’s Civil Governance Scholar, Rutgers Law School (Newark).

2 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

B. Tiny Governments and Democracy…….……….29 Conclusion…………………………………………..…………...33

INTRODUCTION When the massive harms caused by products like tobacco, asbestos, and opioids proved all but impervious to suits brought by private plaintiffs, public plaintiffs like states and cities stepped in to seek remedy for these forms of corporate wrongdoing.1 Bringing claims against “Big Tobacco” and “Big Pharma” corporations, these public plaintiffs were able to achieve what private ones could not, receiving billions of dollars in settlement agreements.2 The legitimacy of these “public tort” lawsuits, though, continues to be contested. Tort reformists and corporate advocates assail these public lawsuits, derisively referring to them as “big government lawsuits” and arguing that this “anti-democratic” litigations gives governments “the chance to achieve a regulatory objective that the majority of the electorate, as represented by their legislators, does not support.”3 Such “regulation by litigation,” they argue, “occurs at the cost of making our frail democracy even weaker,” as these lawsuits are tantamount to “faux legislation, which

1 See Nora Freeman Engstrom & Robert L. Rabin, Pursuing Public Health

Through Litigation, 73 STAN. L. REV. 285, 291 (2021), explaining that “starting in 1954 and continuing for four decades, hundreds of personal injury suits were filed against the cigarette industry, but the industry’s dominance was so complete that no individual plaintiff prevailed.” 2 Id., noting that “in 1998 plaintiffs forced the tobacco industry to the negotiating table and entered into a $206 billion ‘master settlement agreement’ (MSA) – the largest tort payment, by far, in American history.” 3 See, e.g., Victor E. Schwartz, Mark A. Behrens, and Leah Lorber, Tort Reform Past, Present and Future: Solving Old Problems and Dealing with ‘New Style’ Litigation, 27 WILLIAM MITCHELL L. REV. 237 (2020). The authors devote a large Part to “Tobacco Litigation: The Birth of ‘Big Government’ Lawsuits,” and note that “the state attorneys general Medicaid recoupment litigation against the tobacco industry was the genesis of the “Big Government” lawsuit trend. Id.

3 20xx]

Draft – Please Do Not Cite

Draft

sacrifices democracy to the discretion of administration officials operating in secrecy.”4 For state-driven public tort lawsuits, an easy retort to these arguments is that the parens patriae doctrine, which requires states to act in the best interests of their citizens, serves as robust grounding for these state-led claims.5 At the city level, numerous paths to standing provide city claims with legal legitimacy,6 structures of local control give the claims political legitimacy,7 and the corporate misconduct revealed through these litigative efforts lend the claims sociological legitimacy as well.8 Further, as this Article shows, a significant wave of public plaintiff litigation seeking remedy for harms of massive corporate wrongdoing cannot fairly be called “big government” at all. Rather, such litigation is increasingly brought by tiny governments.9 Microlocal government entities like school districts, or very small towns and villages, have been bringing forward claims against massive corporations for widespread harms that impact their communities. This microlocal litigation offers at least two benefits for entities considering combating injurious corporate harms. First, the plaintiffs are such discrete sorts of political communities that arguments that that these litigative moves are anti-democratic have next to no persuasive purchase. 4 Id.

See Margaret H. Lemos, Aggregate Litigation Goes Public: Representative Suits by State Attorneys General, 126 HARV. L. REV. 486, 492 (2012) ( noting that “The authority for such [state] litigation stems from several sources. Most straightforward are state statutes that empower the attorney general to seek restitution for citizens injured by violations of state consumer protection law. Attorneys general also have authority to sue as parens patriae to recover damages for citizens injured by certain violations of state and federal law.” 6 Sarah L. Swan, Plaintiff Cities, 71 VAND. L. REV. 1227 (2019). 7 Id. 8 Id. 9 In the opioid litigation, for example, many small localities followed the lead of the states and larger cities and filed suits as well. See Nino C. Monea, Cities v. Big Pharma: Municipal Affirmative Litigation and the Opioid Crisis, 50 URB. LAWYER 87 (2019). 5

4 Draft – Please Do Not Cite

Draft

[Vol. XX:nnn

Second, this small scale allows some of these microlocal entities to engage in radical responses that would be more difficult to initially achieve at larger scales. For example, hundreds of tiny towns across the nation have enacted ordinances which purport to grant standing to towns and their residents to bring new statutory claims, and to strip corporations of their corporate personhood if those corporations engage in certain forms of wrongdoing.10 These ordinances are almost certainly not technically legally enforceable under current legal frameworks, but they offer an imaginative window into a world of possible responses, and open up new imaginative horizons through which to consider community responses to corporate wrongdoing, in tort and beyond. To make these arguments about this microlocal litigative activty, this Article proceeds as follows: Part I maps the landscape of recent microlocal litigative activity, exploring claims like those brought by school boards against vaping and social media companies, and small towns against industrial polluters. Part II explores a wave of small-town activity targeting corporate environmental wrongdoing through ordinances altering the structural relationship and the litigative interactions between small government and big corporations. Part III connects the microlocal activity described in the preceding Parts to critiques and defenses of public tort action more generally, which often sound in the registers of legitimacy and democracy. Ultimately, the struggles of these small microlocal communities to protect themselves through litigation reveal the deep connections between litigation and self-governance, and offer bold new possibilities for a vision of non-corporate-dominated political communities. I.

MICROLOCAL PLAINTIFF LITIGATION

This Part describes how tiny governments are increasingly important players on the public plaintiff litigation stage. As two prominent examples, school districts have been first 10 See infra, Part I.

5 20xx]

Draft – Please Do Not Cite

Draft

movers in litigation against vaping and social media companies, and small localities and other tiny local governments like water districts have been suing industrial manufacturers for polluting water supplies, and “big oil” companies for climate-change related harms.11 A. SCHOOL DISTRICT SUITS Affirmative litigation, meaning litigation brought by public entities to address widespread wrongdoing, was once a domain run mostly by states. In the tobacco litigation of the 1990s, states, through their attorney generals, were the major plaintiffs, with only a couple of major cities joining them.12 Over time, though, cities have increasingly been engaging in affirmative litigation, bringing suits against lead paint, opioid, and gun manufacturers for example.13 Now, as smaller local government units are becoming more significant actors in local government generally, and “localism in contemporary American law is more local than ever before,”14 these smaller public units are becoming increasingly active on the public litigation stage. School districts in particular are emerging as major players, bringing litigation against vaping and social media companies. i. Litigation against Vaping Companies

11 Engstrom and Rabin, supra note 1 12 Swan, supra note 6. 13 Two school districts in and around Houston, Texas also brought suit

against lead paint manufacturers. In Spring Branch Independent School District v. Lead Industries Association, Inc., et al--Appeal from 11th District Court of Harris County, the court dismissed the claim of the district on summary judgment, on the basis that the plaintiff could not specifically connect the defendant to the paint used on school buildings between 1950 and 1977. See Amber E. Dean, Lead Paint Public Entity Lawsuits: Has the Broad Stroke of Tobacco and Firearms Litigation Painted a Troubling Picture for Lead Paint Manufacturers? 28 PEPP. L. REV. 915 (2001). 14 Nadav Shoked, The New Local, 100 VA. L. REV. 1323 (2014).

6 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

In the storied tobacco litigation of the 1990s, a key vulnerability for the big tobacco defendants turned out to be that they targeted children in their advertising.15 Yet for the tobacco companies, the allure of creating this market of future long-term consumers, despite the risks involved, has continually proven too much to resist. When the original tobacco litigation resulted in numerous limitations on how tobacco companies can advertise to children, tobacco manufacturers pivoted and brought a new product to market: vapes and e-cigarettes.16 In these products, a battery heats up a nicotine-containing liquid, transforming it into an aerosol that the user inhales.17 An estimated 9 million people in the United States vape, and children and youth do so with alarming frequency.18 In fact, the 2022 National Youth Tobacco Survey reported that “more than 1 in 4 middle and high school students use e-cigarettes daily, with nearly 85% of them using fruit-flavored e-cigarettes.”19 The vast majority of these children and youth vape with products produced by JUUL, a company held mostly by other traditional tobacco manufacturers.20 JUUL put vaping devices on the market in 2015, claiming that they were meant to “improve the lives of the one billion adult smokers” on the planet.21 While JUUL claimed its primary mission was to “help those already smoking traditional cigarettes and tobacco products to quit smoking,” it soon became clear that these new products created their own market of users, a market which included a large number of children and 15 Engstrom and Rabin, supra note 1, at 300. 16 E-cigarette company Juul makes the most commonly used e-cigarettes.

In 2018, tobacco company Altria, which manufacturers Marlboro cigarettes, bought a 35% stake in Juul for $13 billion. Jamie Ducharme, I Wrote the Book On Juul in 2021. Here’s What’s Happened to the Vaping Industry Since, TIME (Oct.12, 2023). 17 Id. 18 Id. 19 Syracuse City School Districts Settles Lawsuit Against Juul for More than $130k, CNYCENTRAL (Oct. 24, 2023). These numbers are fairly consistent with earlier ones, which suggested that “27.5% of high school students and 11.5% of middle school students vape.” Id. 20 Id. 21 Id.

7 20xx]

Draft – Please Do Not Cite

Draft

youth.22 Vaping poses particular risk to children, as it leads to nicotine addiction, deficits in brain development (particularly in the areas of “learning, memory and attention), and an increased risk for “future addiction to other drugs” and to regular cigarettes.23 Vaping can also cause nicotine poisoning, which involves symptoms like “sweating, dizziness, vomiting and increased heart rate.”24 Further, many of the ingredients in the liquid vape solution are known carcinogens, and although research is developing, it seems probable that cancer may result.25 School districts—smaller local government entities created by state authority and uniquely tasked with educating children and youth and ensuring appropriate child development—”spent significant and unexpected levels of time and resources to address the pervasiveness of youth ecigarette use and had to divert resources and deploy new ones to combat the problem.”26 The impact on their limited budgets was enormous, ranging from having to “educate about the dangers of youth e-cigarette use, including staff training and new counseling services;” to dealing with the cost of discard of the vaping items “either from youth improperly disposing of [the products] by littering or throwing them in the trash or toilets, or because teachers and school staff must confiscate and store them.”27 as they incurred and costs related to the discard of the items, 22 Id. 23 Id. 24 Id. 25 Id. 26 In re JUUL Labs, Inc., Mktg., Sales Pracs., & Prod. Liab. Litig., 497 F.

Supp. 3d 552 (N.D. Cal. 2020). 27 Additional costs include “costs associated with discipline and suspensions related to incidents of e-cigarette use in schools, including additional legal expenses in connection with disciplinary hearings and costs for home-tutoring services to suspended students at risk of falling behind academically,” along with costs of “monitoring bathrooms for pervasive e-cigarette use as well as installing e-cigarette detectors, additional cameras and anti-vaping signs around the property,” and costs of lost “personnel hours of school staff who had to direct time and activity towards discipline, education, and health care attention associated with e-cigarette use.” Id.

8 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

To recoup these expenses and obtain remedy for the harms done, school districts in states like Kansas, Missouri, and New York began to file suits against JUUL in October 2019.28 Hundreds of districts followed, as did some municipalities, and some state attorney generals also eventually joined.29 The suits were corralled into multi-district litigation (MDL) in federal district court in California, with an additional consolidated litigation procedure simultaneously taking place in California state court.30 A Judicial Council Coordination Proceeding managed the multiple trials in both federal and state court. The plaintiffs argued that JUUL and its associated corporate defendants should be liable negligence, public nuisance, and for violations of federal RICO.31 They argued the defendants “intentionally created a device to hook a new generation on nicotine,” and “targeted youth with its aggressive advertising campaigns, fraudulently marketing its devices and products and failing to warn of their highly addictive nature.”32 School district plaintiffs were particularly well-positioned to bring forward these claims, and may have had some distinct doctrinal advantages to other types of plaintiffs. In some of the districts’ claim for negligence, for example, at least initially the MDL judge noted that “public policy factors” determined whether a defendant owes a duty of care in each of the jurisdictions at play.33 When considering whether the duty test was met under the tests applicable in states like Florida, New York, and California, the court noted Mark Walsh, School Districts’ Legal Battle with Juul Isn’t Over, EDWEEK (Sept. 8, 2022). In the litigation, school districts in over twenty states, represented by California-based law firm, the Frantz Law Group, filed suit against Juul. 29 Id. 30 “The JUUL multidistrict litigation (MDL) was formed on Oct. 2, 2019, and assigned to Judge William Orrick in the Northern District of California.” The MDL contains hundreds of cases. Id. 31 Joe Malinconico, Paterson Schools Become Third NJ School District to File Vaping Lawsuit, PATERSONPRESS (May 24, 2021). 32 Id. 33 In re JUUL Labs, Inc., Mktg., Sales Pracs., & Prod. Liab. Litig., 497 F. Supp. 3d 552 (N.D. Cal. 2020). 28

9 20xx]

Draft – Please Do Not Cite

Draft

that “strong ‘moral blame’ attached to [JUUL’s] conduct in targeting school children with a dangerous product” and that “a strong policy in favor of preventing future harm and curbing teen vaping” pushed in favor of a duty.34 In response to the JUUL defendants’ argument that the core of the claim was simply that the defendants “manufactured a dangerous product that caused them downstream harm,” the court stated that the heart of the claim was instead that the defendants’ “lack of reasonable care in the marketing and sales of JUUL created an illicit youth market and the heavy promotion of JUUL to teen users caused the school districts reasonably predictable harm.”35 In the court’s words, “[a]ltogether, the school districts do not seek to impose a duty on [the JUUL defendants] as members of the general public. Their specific allegations answer JLI's concerns about limitless or unbounded liability.”36 The court found that this is not a “situation where the school districts are ‘strangers’ claiming economic harm based on [the JUUL defendants’] negligence towards other people. Instead, the school districts plausibly plead that [the defendants] directly owed them a common-law duty given the foreseeability and public policy considerations.” The plaintiffs’ status as school districts also assisted in their public nuisance claims. To establish public nuisance (“generally defined as ‘an unreasonable interference with a right common to the general public”),37 a plaintiff who is not a state government may be required to show a “special injury” entitling them to bring forward the claim, meaning that their damages must be “different in kind from that suffered by the general public.”38 The court noted that the school district plaintiffs successfully “alleged specific details of the harm 34 Id. (emphasis mine). 35 Id. (emphasis mine). 36 Id. (emphasis mine). 37 Id. Conduct that may constitute an “unreasonable interference with a

right common to the general public” includes conduct “that involves a significant interference with the public health.” Restatement (Second) of Torts § 821B. 38 Id. (representative v. non-representative local government public nuisance suits)

10 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

that the alleged intentional marketing of the [JUUL] products to youth caused to schools, and how that harm was different than any harm suffered by the public at large.”39 The harm alleged by the school districts was “unique to schools and different in kind from that suffered by the general public in their community.”40 Plaintiff school districts were largely successful in their efforts to hold JUUL responsible for the vaping harms. As the lawsuits mounted, “JUUL dropped all U.S. advertising and discontinued most of its flavors in 2019.”41 Further, it has agreed to pay nearly $3 billion in legal settlements.”42 A $438 million settlement that was made between the state attorney generals included “changes to the company’s marketing practices that allegedly target youths,”43 and school districts across the nation have been announcing additional settlements.44 ii. Litigation Against Social Media Companies School districts have also brought significant litigation against major social media companies. Facebook (Meta), Instagram (Meta), Snapchat (Snap), Youtube (Alphabet), and TikTok (ByteDance) are frequently named as defendants in the majority of these suits, which allege that social media companies have created mass mental health harms for children and youth.45 School districts allege that these “platforms have fueled a youth mental health crisis with addictive technology and harmful targeted content,”46 and 39 Id. 40 Id. 41 Michelle Chapman, Altria invests $2.75B in rival startup NJOY after

Juul Exit, ASSOCIATED PRESS (March 26, 2023). 42 Ducharme, supra note 36. 43 Mark Walsh, School Districts’ Legal Battle with Juul Isn’t Over, EDWEEK (Sept. 8, 2022). 44 However, many suits are still active: “As of May 2024, a total of 5,102 lawsuits remain active in the California MDL.” https://www.consumernotice.org/legal/juul-lawsuits/ . 45 Arianna Prothero, Seattle District Sues Social Media Companies Over Students’ Mental Health, EDWEEK.ORG (Jan. 9, 2023). 46 Pennsylvania County Lawsuit

11 20xx]

Draft – Please Do Not Cite

Draft

that they “manipulate users by hooking them in as they browse algorithmically personalized content and advertisements.”47 Children and youth “are especially vulnerable” to such manipulations “because their brains are still developing.”48 These lawsuits allege that the defendant companies like Meta are “exploiting young people’s vulnerabilities by developing algorithms intended to keep users on the platform as long as possible, even compulsively; creating visual filters it knows can contribute to body dysmorphia; and presenting content in an “infinite scroll” format that makes it hard for children to disengage.”49 In what school district plaintiffs say is a result of these social media company practices, the mental health of children and youth has plummeted.50 For example, health screenings in one community found that “34% of school-aged youth were at risk for moderate-to-severe depression, more than 25% of youth had a history of suicide ideation, and 40% were at risk for significant anxiety.”51 Further, “[a]cross the US, youth suicide rates were stable from 2000 to 2007, but increased by 57% in 2018, according to the Centers for Disease Control and Prevention.” Medical experts like the US Surgeon General have cautioned that “social media poses serious potential risk to children’s mental health,” and “there is increasing evidence that social media use during adolescence—a critical stage of brain development—is associated with harm to mental health and well-being.”52 Indeed—with significant echoes of the tobacco litigation, where whistleblower evidence was the catalyst for the litigation—whistleblower evidence in these social media suits suggests that Meta was aware of the negative impact of social media on children’s mental health, but chose to do nothing

47 Id. 48 Alyson Klein & Arianna Prothero, 41 States Sue Meta Over the Social

Media Giant’s Impact on Kids, EDWEEK (October 24, 2023). 49 Id. 50 Id. 51 Id. 52 Id. The whistleblower, Frances Haugen, testified to these matters in front of Congress.

12 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

with this information.53 In 2021, a whistleblower offered “[t]housands of pages of documents” showing “Meta conducted extensive research on the negative impact of its platforms on children’s mental health and the spread of false information, but failed to act on any of those findings.”54 Like with tobacco, children are often a core part of corporate strategy for social media companies: “Facebook understands that if they want to continue to grow, they have to find new users, they have to make sure that that next generation is just as engaged on Instagram as the current one […] And the way they’ll do that is by making sure that children establish habits before they have good selfregulation … It’s just like cigarettes. [Teenagers] say they feel bad when they use Instagram but can’t stop.”55 Children and youth spend a significant amount of their free time on social media, with the pandemic increasing these amounts exponentially.56 One study suggests that “kids ages 8-12 spend “a total of about five and a half hours a day” on social media, and for teens, that number can be as high as eight and a half hours.57 Additionally, “[m]ore than 90% of children ages 13 to 17 report using social media, so teens are an important market for tech companies, according to the lawsuit.”58 School districts have incurred significant expense and cost as a result. Their “teachers and administrators waste valuable time responding to cyberbullying and other disciplinary problems.” Schools continually add “new training and school policies around social-media use,” and human and other resources for “counseling youths whose addiction to online apps is leading to anxiety, depression or suicidal thoughts.”59 Educational costs in general go up, as “worsening mental health and behavioral disorders including 53 Id. 54 Id. 55 Id. 56 Id. 57 Walsh, supra note 30. 58 Pennsylvania County Lawsuit 59 Elizabeth Nolan Brown, Nearly 200 School Districts Sue Tech Giants

Over Harm Caused by Social Media, REASON.COM (July 24, 2023).

13 20xx]

Draft – Please Do Not Cite

Draft

anxiety, depression, disordered eating and cyberbullying” increase the complexity of the already challenging task of educating students.60 Accordingly, in 2023, the Seattle School District sued a number of social media companies, becoming the first school district in the nation to do so.61 Numerous families had already brought forward claims seeking remedy for the harms of social media on their children, but the school district suit brought the litigation into the world of public plaintiff litigation, and relied on public nuisance62 as a cause of action.63 Since it did so, hundreds of additional school districts have brought suits as well.64 Now joined by additional local government plaintiffs, like cities, counties, and at least two tribal nation plaintiffs,65 an MDL is proceeding the U.S. District Court in Oakland, California, and over forty states have now sued as well, mostly in their respective state courts.66 60 Gene Johnson, School Lawsuits Over Social Media Harm Face Tough

Legal Road, ASSOCIATED PRESS (Jan. 11, 2023). 61 Id. 62 Cyrus Farivar, School Districts Took on Juul with a Novel Legal Strategy. Now they’re Going After Social Media Giants, FORBES (April 18, 2023). 63 Id. 64 Johnson, supra note 60, describing how “[n]early 200 school districts are taking on Facebook, TikTok, Snapchat and YouTube, joining a lawsuit claiming the social media outlets harm students and distract teachers with cyberbullying and mental health issues.” 65 These cities include Providence, Rhode Island; New York City, New York; Two tribal nations have sued. See Graham Lee Brewer, Haleluya Hadero, and Shawn Chen, Two Tribal Nations Sue Social Media Companies over Native Youth Suicides, ASSOCIATED PRESS (April 9, 2024). Four Canadian school districts, the Toronto District School Board, Toronto Catholic School Board, Peel District School Board and the Ottawa-Carleton District School, are also suing. They represent a total of 600,000 student and are seeking damages in the amount of $4.5 billion (CAN). Reuters, Canadian School Boards Sue Social Media Giants for Over C$4 bln in Damages, REUTERS (Marhc 28, 2024). 66 “Arizona, California, Colorado, Connecticut, Delaware, Georgia, Hawaii, Idaho, Illinois, Indiana, Kansas, Kentucky, Louisiana, Maine, Maryland, Michigan, Minnesota, Missouri, Nebraska, New Jersey, New York, North Carolina, North Dakota, Ohio, Oregon, Pennsylvania, Rhode

14 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

The Seattle School District’s suit was quite clearly a direct descendant of the vaping litigation. However, this social media litigation has two complications that the vaping litigation did not. First, section 230 of the Communications Decency Act and the First Amendment loom large over this litigation.67 Second, the causation argument is more difficult.68 In regards to causation, in the vaping litigation, “drawing a line between e-cigarettes and the physiological harm to children” was relatively easy to do, particularly in a posttobacco litigation landscape.69 Establishing “causality between the social media’s rise and the worsening of teenagers’ mental health,” on the other hand, is more tenuous.70 The use of social media “has been linked to anxiety and depression, among other psychological ailments” in children, but “there are a lot of other factors contributing to the sorry state of kids’ mental health at the moment,” including the pandemic and the accompanying remote learning.71 Additionally, there has been “decreased access to mental health providers in the school and in the community,” and “for the last several years there’s been racial unrest and there’s been economic unrest.”72 Ultimately, “[i[t’s hard to tease out one reason” from others regarding the decline in children’s mental health.73 Further, tort law has traditionally disfavored claims for psychological, as opposed to physical, injury. Nevertheless, just as the public plaintiffs were able to bring causation arguments in a different register than the private Island, South Carolina, South Dakota, Virginia, Washington, West Virginia, and Wisconsin have joined the broader lawsuit filed in California’s federal court,” and “Florida, Massachusetts, Mississippi, New Hampshire, Oklahoma, Tennessee, Utah, Vermont, and the District of Columbia, filed similar action in their own state courts. Johnson, supra note 60. 67 Farivar, supra note 62. 68 Id. 69 Id. 70 Id. 71 Id. 72 Johnson, supra note 60. 73 Id.

15 20xx]

Draft – Please Do Not Cite

Draft

ones in the tobacco litigation, school districts are also a persuasive public plaintiff. As one commentator noted, although a school district may be “an unexpected plaintiff,” they are likely positioned such that they will have “higher odds of success than individual families,” given that districts can draw on “broader trend data” to strengthen their arguments.74 As this litigation unfolds, the school district plaintiff will likely share in some of the doctrinal advantages they did in the vaping litigation, though whether that will be sufficient for a finding of liability (or a settlement) in these factual circumstances remains unclear. B. ADDITIONAL MICROLOCAL PLAINTIFF SUITS Like school districts, other microlocal governmental units, including other kinds of districts and tiny towns, have also been using litigation to advance the interests of themselves and their communities. As one example, a sewer district in York, Maine, recently brought suit against nearly a dozen industrial manufacturing companies like 3M and DuPont after polyfluoroalkyl substances (PFAS) were discovered in the town’s wastewater.75 Three additional water and sewer districts in Maine have also brought similar suits.76 Over in North Carolina, the small town of Pittsboro (population under 3000), also sued a number of industrial manufacturers for contaminating their water supply with PFAS, as did Carneys Point Township (population approximately 8000). Lee, Massachusetts, a town of less than 6000 people according to the 2020 census, sued Monsanto and General Electric for the “catastrophic disaster” the small community has experienced because of the toxic consequences of polychlorinated biphenyls (PCBs) that polluted the 74 Julian Shen-Barro, As Seattle Schools Sue Social Media Companies,

Legal Experts Split on Potential Impact, CHALKBEAT (Jan. 17, 2023). 75 Id. 76 Anna Coon, York Sewer District Sues Manufacturers, Chemical Companies Over PFAS Contamination, FOX 23 (Oct. 24, 2023). “York Sewer District is one of four water systems in Maine represented by SL Environmental Law Group.” Id.

16 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

Housatonic River there. Additionally, Roxana, Illinois, with a population of less than 1500 people, “sued Shell, ConocoPhillips and local refinery, accusing them of releasing millions of pounds of cancer-causing pollutants and failing to repair broken pipelines and other problems at the refinery.”77 These lawsuits are reminiscent of the famous David-andGoliath story dramatized in the 1990s movie Erin Brokovich. In real life, Brokovich was a resident of Hinkley, California, a small town whose water supply was contaminated by chromium 6 as a result of the activities of Pacific Gas & Electric. “In 1996, the case settled for $333 million” –the then largest settlement on record for this kind of lawsuit.78 In Alaska, the tiny Native Village of Kivalina, with a population of approximately 400 people, sued “dozens of oil, utility and mining companies, alleging that their actions cause global warming.”79 The complaint alleged that the companies have caused global warming, which has prevented the normal formation of protective ice around the island, leaving it vulnerable to erosion and extreme weather events.80 The town may be forced to essentially close and “relocate to the Alaskan mainland,” a feat will cost an estimated $95 to $400 million.”81 The Village sued in public nuisance and civil conspiracy, alleging that the companies were aware of the impact of their activities on global warming but nevertheless “conspired to mislead the public with regard to the science of global warming.”82 John Badman, Illinois Village Sues Oil Firms Over Pollution, LONGVIEW NEWS J. (March 21, 2012). 78 Erin Brokovich: The Real Story of the Town Three Decades Later 79 The defendants are "(ExxonMobil, Chevron, BP, ConocoPhillips, and Royal Dutch Shell), coal-fired utilities (AES, AEP, DTE, Duke Energy, Dynegy, Edison International, MidAmerican, Mirant, NRG, Pinnacle West, Reliant, Southern Company, and Xcel Energy), and from Peabody Energy, the nation’s largest coal company. 80 Id. 81 Id. 82 Native Vill. of Kivalina v. ExxonMobil Corp., 663 F. Supp. 2d 863, 87980 (N.D. Cal. 2009), aff'd, 696 F.3d 849 (9th Cir. 2012). “The district court dismissed the nuisance claims for lack of standing and for presenting nonjusticiable questions. The Ninth Circuit affirmed but did not rely on the 77

17 20xx]

Draft – Please Do Not Cite

Draft

Through these claims, tiny governments and districts attempt to confront massive corporate harms using traditional litigation tools. While not as well-resourced as larger public entities like states and cities, non-profits organizations and contingency-fee private attorneys assist these small governments in these efforts. II.

MICROLOCAL LANDSCAPE

CHANGES

THE

LITIGATION

Other microlocal entities have chosen a more radical path to combat similar harms. Rather than continuing to use traditional litigative tools, these microlocal entities are seeking to alter the litigative playing field itself. Small, discrete communities bearing negative impacts of regulatory failures have been passing radical laws and ordinances that combat large corporate harms by altering the terms on which litigation about them takes place. As a first wave strategy, these microlocal entities began disempowering corporations, stripping corporate rights and declaring in ordinances that the constitutional perks corporations have come to enjoy are not applicable in their towns. This later evolved into ordinances through which these microlocal entities empower

doctrines of standing or justiciability. Rather, like the Supreme Court's opinion in American Electric Power, the Ninth Circuit's panel ruling focused solely on the statutory displacement of the federal common law of nuisance.” Luther J. Strange III, A Prescription for Disaster: How Local Governments’ Abuse of Public Nuisance Claims Wrongly Elevates Courts and Litigants into a Policy-Making Role and Subverts the Equitable Administration of Justice, 70 S.C.L. REV. 517 (2019). Native Vill. of Kivalina v. ExxonMobil Corp., 663 F. Supp. 2d 863, 879-80 (N.D. Cal. 2009), aff'd, 696 F.3d 849 (9th Cir. 2012). “The lawsuit is similar to three other global warming nuisance cases that have been rejected by federal courts in New York, Mississippi, and even in the Northern District of California, where the Village of Kivalina filed its lawsuit. Those courts have generally agreed that the judiciary is illequipped to address global climate change issues. All three cases are currently on appeal in separate federal courts of appeals.” Troutman Pepper, Alaskan Village Sues Energy Companies for Global Warming Nuisance, TROUTMAN PEPPER INSIGHTS (Feb. 29, 2008).

18 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

new parties to sue, and create new causes of action on which they can do so.

A. Stripping Corporate Rights In 1994, in rural Pennsylvania, eleven year old Tony Behun died after spending a day biking through a field that had, unbeknownst to him or anyone else in the town, been subject to an “EPA-approved practice of sludge spreading, whereby sewage—including medical and industrial waste that is “treated” to remove some but not all of the deadly pathogens—is applied to farm land.”83 After another boy, Daniel Pennock, died in a nearby town from a similar exposure, towns in the area began agitating for more regulations over the practice, and some began passing ordinances limiting sludge spreading.84 Unhappy with this attempted regulation, one sludge-spreading corporation “sued one of the towns that had tried to crack down on the procedure—on the basis that the local rules infringed on the corporation’s “personhood rights.”85 When another town tried to pass a “tipping fee”—assigning the cost of testing containers of toxic sludge to the companies doing the dumping—they too found themselves in court, as Synagro corporation, one of the “largest sludge haulers in the country,” sued not only the town, but also the town supervisors, seeking $1 million in damages from each of them personally.86 These small communities then mounted a “rebellion against the idea of corporate rights” itself.87 In Porter Township, this took the form of a drastic measure: “Porter Township made history by becoming the first community in

KEVIN DANAHER, SHANNON BIGGS, AND JASON MARK, BUILDING THE GREEN ECONOMY: SUCCESS STORIES FROM THE GRASSROOTS (2016). 84 Id. 85 Id. 86 Id. 87 Id. 83

19 20xx]

Draft – Please Do Not Cite

Draft

the United States to pass an ordinance stripping corporate constitutional protections within the township.”88 At a public hearing on December 9, 2002, the town unanimously agreed to pass an ordinance stating that that in their town, “Corporations shall not be considered to be persons protected by the Constitution of the United States or the Constitution of the Commonwealth of Pennsylvania.”89 A few months later, Licking, with a population of only 479 people, followed Porter’s lead, passing its own “Corporate Rights Elimination Ordinance.”90 B. Empowering New Parties and New Basis for Claims Like Porter and Licking, other rural Pennsylvania townships also became interested in pushing back against corporate power through local ordinances. Those ordinances eventually became focused not only on removing corporate powers and personhood, but on empowering new parties to sue and establishing new basis for legal claims. One example of this occurred in 2006, in Tamaqua Borough, Pennsylvania, a town with a population of about 7000 people. The town became enraged after corporate dumping of raw sewage into an open pit mine near the town’s water source threatened their water supply.91 Although this dumping was illegal under state law, the state was not pursuing the violation, and the Borough decided to independently react. The Borough enacted a new ordinance that did four important things. First, the ordinance “incorporated into local law the state's prohibition against dumping sludge, thereby laying the foundation for local enforcement.”92 Second, the ordinance declared it “unlawful for any corporation or its directors, officers, owners, or managers to interfere with the existence and flourishing of natural communities or ecosystems, or to cause damage to those

88 Id. 89 Danaher, supra note 83. 90 Id. 91 Id. 92 Id.

20 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

natural communities and ecosystems.”93 Third, the ordinance set forth the world's first “rights of nature” law, giving natural ecosystems and resources “inalienable and fundamental rights to exist, and defining “borough residents, natural communities, and ecosystems” as “persons” for the purposes of enforcing “the civil rights of those residents, natural communities, and ecosystems.” Fourth and finally, like its predecessor ordinances in Porter and Licking, the ordinance “provided that corporate violators would lose their rights secured by state law and by the Contracts and Commerce Clauses of the U.S. Constitution.”94 In other words, this ordinance created a new cause of action, ensured residents and small local governments have standing to bring that cause of action, and removed corporate constitutional and civil rights. Similar ordinances also often add additional legal penalties to corporate misconduct.95 All of these townships, Porter, Licking, and Tamaqua Borough, were assisted by the Community Environmental Legal Defense Fund (CELDF), “a left-leaning” public interest organization focused on environmental justice, which also advocates for “positions remarkably similar to some far-right libertarian groups that prioritize local control” in seeking to promote environmental and community rights.96 Since Tamaqua Borough passed its ordinance, CELDF has assisted hundreds of other towns and even larger cities in passing similar ordinances. In fact, “[f]rom small New England hamlets to major mid-Atlantic cities to sea-side California counties,” close to 200 towns and cities now have these laws.97 What began as a microlocal intervention has moved not only horizontally, across other tiny towns, but also vertically, up 93 Id. 94 Desmond Nicols, After LEBOR: Can the Rights of Nature Movement

Stand Back Up? 74 FLA. L. REV. 699 (2022). 95 Id. 96 Indeed, even prior to the Tamaqua ordinance, CELDF had assisted other small towns in passing similar ordinances. Id. 97 The Rights of Nature Movement. Currently, the list of participating municipalities includes “Santa Monica, California; Mountain Lake Park, Maryland; Halifax, Virginia; Barnstead, New Hampshire; and Shapleigh, Maine. Nicols, supra note 94.

21 20xx]

Draft – Please Do Not Cite

Draft

to larger cities. In 2010, Pittsburgh, Pennsylvania became “the world's first large city to adopt a law recognizing Nature's rights” such a law,98 followed later by Toledo, Ohio.99 In Toledo, the city enacted the Lake Erie Bill of Rights. The Lake Erie Bill of Rights granted the Lake Erie Ecosystem special protections and prohibited corporations from violating these new rights.100 Like the Tamaque ordinance, the Lake Erie Bill of Rights also granted standing to the city and its residents to sue for such damages. It also enhanced the available damages, declaring that “violators of the Lake Erie Bill of Rights would be subject to fines and strict liability for ‘all harms and rights violations,’ including damages for the ‘full and complete restoration of the Lake Erie Ecosystem and its constituent parts.’”101 And like the Tamaqua ordinance, the Lake Erie Bill of Rights “also removes personhood from any corporation that violates or seeks to violate the law,” stating that “[c]orporations that violate the Lake Erie Bill of Rights shall neither be deemed to be persons nor ‘possess any other legal rights, powers, privileges, immunities, or duties’ to the extent that corporate personhood would ‘interfere with the rights or prohibitions enumerated by this law.’102 Not surprisingly given its radical nature, the Lake Erie Bill of Rights lived a short life. The Ohio State Legislature Ironically, Pittsburgh actually generated most of the toxic sludge harming the rural areas of the state: “most of the sludge entering rural Pennsylvania isn’t local; it comes from the city of Pittsburgh. With a metropolitan population of about 1.7 million people, Pittsburgh generates tons of waste every day from homes, hospitals, and industry. Dumping it in the landfill has always been an expensive proposition for waste management companies. So, following heavy industry lobbying, legislators in Pennsylvania lowered waste disposal standards in 1997 to accommodate the interests of big business. The lowering of the statewide standard opened the floodgates for sludge-spreading in rural areas. It also automatically repealed more than 100 existing local regulations that had restricted factory farms from setting up shop.” Id. 99 Id. 100 Id. 101 Anne Choike, Local Corporate Law (draft on file with author). 102 Id. 98

22 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

soon nullified the Lake Erie Bill of Rights, and a district court found it unconstitutional for reasons of vagueness and exceeding the scope of local power.103 Despite setbacks like this, the movement has inspired other communities to also take bold action to counteract corporate power.104 For instance, again with the assistance of CELDF, at least ten towns across the nation (including Halifax, Virginia, population 1,099, and Nottingham, New Hampshire, population 4,785) have enacted ordinances prohibiting chemical or toxic trespass in their towns. The model “Corporate Chemical Trespass Ordinance” states that the “deposition of toxic chemicals or potentially toxic chemicals within the body is a form of trespass for which a resident in these towns may seek remedy.105 A corporation that engages in chemical trespass (on a strict liability basis) can be liable for “trespass damages, compensatory damages, 103 “After LEBOR passed in Toledo, investigative reporting revealed that

the Ohio Chamber of Commerce secured the cooperation of a key lawmaker to slip an amendment in the Ohio budget to nullify the ordinance. An investigative journalist obtained the email between the Ohio Chamber of Commerce and an Ohio State Representative, which showed the Chamber of Commerce asserting that ‘[l]anguage in this amendment stating that [nature and ecosystems] do not have standing is essential to what we're trying to accomplish.’ This amendment to the budget bill was reported by local media the next day as seeking to “nullify Toledo's Lake Erie Bill of Rights, which allows city residents to sue on behalf of the lake to, among other things, curb farmers' agricultural runoff that helps promote toxic algal blooms. The budget passed with the text of the amendment stating in most pertinent part: “Nature or any ecosystem does not have standing to participate in or bring an action in any court of common pleas.” Nicols, supra note 83. 104 Additionally, in Pennsylvania, proponents of these ordinances had a state constitutional amendment introduced into the state House of Representatives. The state constitutional amendment would have stated that: “The right to local self-government includes, without limitation, the power to enact local laws: (1) protecting health, safety and welfare by establishing the rights of people, their communities and Nature and by securing those rights using prohibitions and other means; and (2) establishing, defining, altering or eliminating the rights, powers and duties of corporations and other business entities operating or seeking to operate in the community.” It did not pass. Id. 105 Steven N. Geise and Hollis R. Peterson, Toxic Trespass: Lead Us Not Into Litigation; Rights of Nature Timeline.

23 20xx]

Draft – Please Do Not Cite

Draft

punitive damages, and the instatement of permanent injunctive relief” under the ordinance.106 To be clear, these kinds of tiny government ordinances empowering towns and residents to bring forward novel civil claims and disempowering corporations are “almost certainly” unenforceable, both as a matter of state and local government law, and as a constitutional one.107 Indeed, “absent an upheaval of Supreme Court precedent that restructures state and federal power, as well as the rights of corporations, these rationales will certainly fail in the courts.”108 To use the context of the Lake Erie Bill of Rights as an example, the corporations who wanted to pollute made a number of arguments that sounded in the registers of both constitutional law and local government law. Amongst other things, Plaintiff corporations argued that LEBOR was preempted, because, like in most states, Ohio does not actually allow municipalities or other local government units to create new causes of action or legislate in the area of private law at all.109 Similarly, municipalities and other local government units do not have the ability to make corporate law, or determine what forms are acceptable for incorporation purposes, and, at any rate, “the power to confer statutory standing to parties in Ohio State Courts is reserved to the Ohio Legislature,” so Toledo could not empower itself nor its residents with standing to bring novel claims.110 In fact, as a matter of state and local government law, local government power is derivative, meaning the scope and extent of local government power is determined by state law, and state law usually imposes significant restrictions on it. Even in “home rule” jurisdictions, where localities exercise the most potential power, local government entities have relatively limited abilities to pass laws, to tax, and to raise revenues in other ways. 106 Id. 107 Id. 108 Id. 109 Id. 110 See Nicols, supra note 83, noting that “the Ohio Constitution reserves

exclusive power to the Ohio legislature to enact legislation governing the formation, structure, and organization of corporations.”

24 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

In the Lake Erie litigation, Toledo argued that, for its part, the town could make such governance choices, as a fundamental part of an innate right to self-government.111 Their brief in the case “reached back into colonial history, when local governments were founded without express authority from any centralized power, to show that ‘the right of self-government was one of the basic, fundamental values upon which this country was founded.’” They also turned to state constitutional law to argue that each state contained “principles of the right of local self-government in every state constitution.”112 As a matter of technical legal rules, though, the corporations were almost certainly right: under current legal frameworks, the ordinances stand little to no chance of surviving a challenge. Moreover, even if there were a chance that these ordinances could be upheld as within the scope of local government power, the federal constitution would no doubt quickly override them. These ordinances are not close to the line issues: the problems with them are “readily apparent.” A local ordinance that “strips corporations of their legal rights will almost certainly be invalidated as a violation of the prevailing (and longstanding) interpretation of the Fourteenth Amendment. Indeed, corporations were viewed as artificial persons dating back to at least the mideighteenth century. Corporations eventually “secured nearly all the same rights as individuals through a two-centurieslong effort concentrated on the Supreme Court,” culimating in Burwell. V. Hobby Lobby Stores, Inc., in 2014.” 113 Under this legal framework, “there is little to no hope for a personhood-stripping ordinance, such as LEBOR, to survive judicial review.”114 A similar story played out in relation to an ordinance passed in Grant Township. In June 2014, the townspeople of Grant Township, hoping to keep wastewater injection wells out of their community, “voted unanimously to adopt a 111 Id 112 Id. 113 Id. 114 Id.

25 20xx]

Draft – Please Do Not Cite

Draft

Community Bill of Rights Ordinance” similar to the other CELDF organized ordinance. Pennsylvania General Energy Company (PGE) then sued Grant Township, and argued that their adoption of such an ordinance “is deliberate, arbitrary, irrational, exceeding the limits of governmental authority, amounts to an abuse of official power and shocks the conscience.”115 PGE also argued that the ordinance “was ‘unconstitutional’ because it violates the corporate ‘person’s’ civil rights under the 1st and 14th amendments, as well as the Commerce and Supremacy Clauses of the U.S. Constitution.”116 Not surprisingly, PGE was successful on its claims. PGE also “filed a motion for attorneys' fees based on its status as the prevailing party in a federal civil rights action.” The company “produced detailed attorney billing records supporting its claim for over $600,000 in fees,” but eventually agreed to instead “accept $100,000 in fees to avoid bankrupting Grant Township.”117 III.

LEGITIMACY

AND

DEMOCRACY

MICROLOCAL

LITIGATION

115 Id. 116 Id. 117 The Rights of Nature. The case of Highland Township and Seneca

Resources Corporation also resulted in corporate victory. The Seneca corporation sued the town when the tow banned disposal injection wells. When the town’s governance changed, however, support for the ordinance dissolved, and the town “entered into a consent decree providing that the ordinance was unconstitutional, unenforceable, and adopted in excess of the town's authority, and the district court entered judgment for Seneca.” Nevertheless, advocates for the ordinance “were undeterred,” and “three months later,” they succeeded in enacting “a Home Rule Charter, which included the community rights provisions of the former ordinance. The Seneca company then sued again, and the district court judge partially granted a motion for judgment, “finding that the charter provision prohibiting the injection of fracking waste was preempted by the federal Safe Drinking Water Act and by state law. The judge also found that community rights provisions violated Seneca's First Amendment right to seek redress in courts, as well as principles of Substantive Due Process, which protected Seneca against arbitrary government action.” The case eventually settled. The Rights of Nature Movement.

26 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

These arguments that corporate defendants have relied on in challenging these ordinances are remarkably similar to an argument tort reformists make against public entity litigation more generally: corporate defendants have argued that these ordinances violate their corporate constitutional rights, and scholars sometimes critique public litigation settlements on similar grounds.118 As one legal scholar wrote in critique of public tort settlements, when states in particular sue companies, those companies “may be forced to accept settlement provisions that subject them to regulations that would be considered unconstitutional if they were imposed by legislation. This is particularly true of attempts to restrict commercial speech by manufacturers. Although commercial speech is not treated like other forms of expression, it is entitled to some protection under the First Amendment."119 Specifically, under current constitutional law, “the government can regulate commercial speech only if it is deceptive, intended to serve an illegal purpose, or: (1) the regulation in question is necessary to advance a substantial governmental interest; (2) the regulation directly advances that interest; and (3) the regulation is no broader than necessary to advance that interest.”120 But a settlement like the tobacco master settlement agreement “contained a number of content-based restrictions on marketing and advertising that arguably would have violated the tobacco companies' right to commercial speech if they had been imposed by law.”121 On this telling, then, massive state power is brought to bear on vulnerable corporations through public entity litigation, and those corporations have little ability to do anything but succumb to such pressure. Microlocal litigation disrupts this narrative, and others. Underlying the above lament about corporations agreeing to

118 Id. 119 Id. 120 Id. 121 Richard C. Ausness, Public Tort Litigation: Public Benefit or Public

Nuisance, 77 TEMPLE L. REV. 825 (2004).

27 20xx]

Draft – Please Do Not Cite

Draft

speech restrictions is a basic assertion that in public tort litigation, states are simply too powerful to be fair plaintiffs.122 Thus, when states bring a case, “an unpopular industry may be forced to pay a large settlement solely because the states have decided to unite together against it” and industries will end up paying “regardless of their blameworthiness or the severity of their conduct.”123 But microlocal plaintiffs, even united together, have significantly less resources and powers than states. It is highly unlikely that any large corporation could plausibly feel cowed into settling despite a lack of wrongdoing when facing a suit brought by such plaintiffs, rendering this argument inapposite in this context. In fact, microlocal, tiny governments thwart many of the other arguments often levied against public entity litigation generally. Along with the argument above, microlocal litigative activity upends the most common arguments against public entity litigation, namely that the public entities are too removed from the harm directly created, and that such lawsuits are an affront to democracy. A. Alignments Between the Harm and the Microlocal Entity Public tort suits have been criticized on multiple grounds.124 One commonly-voiced critique is that public entities are simply inappropriate plaintiffs: they are not properly harmed and do not fit into the standard singledefendant-directly-harms-single-plaintiff paradigm.125

122 Id. 123 See Dean, supra note 13, describing how “[d]ecisions to file multistate

suits are made at the quarterly National Association of Attorneys General meetings. During those meetings, each state attorney general may present ideas about the proposed suits, and lead states are appointed to conduct investigations.' When those lead states make a decision regarding whether to sue, they pass the suit to all other interested states.' Those states may either agree jointly to pursue the action or decline to participate.” 124 Id. 125 Id.

28 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

Critics of public entity litigation thus often highlight a perceived mismatch between the harm and the public entity bringing suit. For example, they argue there is little connection between an individual’s “choice” to use illicit street drugs following a prescribed round of opioid treatment, and a purported harm to a city or state.126 These arguments are not persuasive in any context, but they have even less salience when microlocal plaintiffs are at play. School districts bringing claims for harms to school children, water districts bringing claims for harms to the drinking water, and tiny towns trying to remedy harms caused to residents by the toxic sludge literally in their backyards are able to rhetorically align harms and entities. The “different levels of government – national, state, county, and municipal” – and, as this Article suggests, the microlocal – speak on behalf of ever more specific political communities.”127 These smaller political communities are more obviously impacted by the types of harms being claimed. Cities, for example, are able to persuasively argue that they are harmed by things like social media companies targeting children with algorithms that they can’t possibly resist, but it takes more work than it does for a school district to show its relationship to the harm.128 This was clear in the court’s discussion of the distinction between the harms to a school district and the harms to the general public, as the court was easily convinced that the harm being asserted belonged specifically to the microlocal entity bringing the claim.129

126 Id. 127 Eric Miller, Municipal Exclusion, 89 FORDHAM L. REV. ONLINE 193

(2020). 128 Id. 129 See infra, Part I.

29 20xx]

Draft – Please Do Not Cite

Draft

B. Tiny Governments and Democracy Tiny governments also disrupt another long-running critique, that public entity litigation is inherently undemocratic. According to this argument, litigation brought by public entities is not “democratic,” because only state legislators may legitimately express the regulatory will of the voters, and if they have failed to regulate in a particular area, that is because that is the majority will.130 Indeed, commentators argue that such ‘regulation through litigation’ ‘makes a parody of the democratic process’ and ‘leads to dubious and one-sided public policy.’131 Although this argument is spurious in any context, as it ignores the well-documented reality of regulatory capture at the state level by interest groups,132 it is particularly inapt at the level the level of microlocal or tiny governments.133 Microlocal, tiny governments are widely understood to be one of the most accessible forms of democratic participation available for the average American.134 In the words of one commentator, “The most direct input citizens can have is at the local level. You can go to your township meeting and as an individual can have a substantial impact on the way things are considered. At the state level, other voices are coming in that may or may not care about the local community. Citizens’ rights get drowned out in that larger process. It shouldn’t happen, but it does. By the time you get to the federal level, you are way out of inﬂuence in decisionmaking.”135 Scaling up to a large size government “reduces 130 Dean, supra note 13, describing how “[r]egulation through litigation

poses many significant problems, and dissenters argue that tort law should not be manipulated to do in court what we as a democratic society lack the popular support to do through the political process.” 131 Id., citing Richard Epstein, Lawsuits Aimed at Guns Probably Won’t Hit Crime, WALL ST. J. (Dec. 9 1999). 132 Swan, supra note 6. 133 Id. 134 Nevertheless, this is an oversimplified telling of the connections between democracy and the microlocal: Tiny governments have a more complicated relationship with democratic participation that is more visible in other contexts. Shoked, supra note 3. 135 Id.

30 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

the potential for participation;” scaling down drastically increases it.136 Indeed, in many of these tiny towns passing CELDFinspired ordinances, “the ordinances have been passed into law through ballot initiatives,” a tool of direct democracy that typically requires at least 60% popular support to pass.137 Many of the ordinances are described as passing “unanimously,” with high bi-partisan support.138 In fact, in addition to the high level of popular support in the tiny governments where these bills pass, it is also notable that these ordinances tend to easily overcome the usual partisan lines associated with both litigation and regulation. Townships like Porter and Licking are, like most rural localities, politically conservative enclaves.139 The ordinances they produce nevertheless “often seek to establish local power while diminishing corporate power,” and they happen because of dissatisfaction with regulatory regimes that allow things like toxic sludge to continue to the detriment of residents.140 It is perhaps not surprising that a large city like Pittsburgh would pass such an ordinance,141 since American cities are almost universally progressive politically, but the salience of these new enforcement tools in politically conservative tiny governments suggests that richly nuanced democratic engagement is occurring there. Turning to the microlocal plaintiff suits, school districts are normally products of representational democracy, but at a level that is still so close to the voting public that its actions can often lay claim to being “democratic” in a way different than, for example, a state. Entities like water and sewer districts share in this as well: they are not glamorous Richard Briffault, The Rise of Sublocal Structures in Urban Governance, 82 MINN. L. REV. 503, 507 (1997). 137 Stephen R. Miller, Community Rights and the Municipal Police Power, 55 SANTA CLARA L. REV. 675 (2015). 138 Id. 139 Id. 140 Id. 141 “The Pittsburgh ordinance “was enacted by a unanimous 9-0 vote of the city council,” after “strong community organizing in support of the measure.” Rights of Nature Timeline 136

31 20xx]

Draft – Please Do Not Cite

Draft

entities, but their leaders are often democratically-elected and members of the voting public can usually easily access their leadership.142 Tiny governments in general are “viewed as particularly propitious for erecting governments that accurately conform to citizens’ desires since the closer government is to the citizen the better the citizen can convey her preferences to her representatives. Small-scale government can do still more to attenuate, even abolish, the government-citizen divide by turning to direct democracy, which may be impracticable for large-scale traditional government.”143 Interestingly, the very act of fighting against corporate harm can itself assist in democratic community-making. At the microlocal level, “citizens learn[] to assert control over the institutions that govern their lives” and the lives of children in their communities.144 When a common enemy wreaks havoc on a community, that experience can create cohesion across what otherwise might otherwise be dividing partisan lines. As one commentator noted, in some ways, “struggle itself” can be an “incubator that create[s] community.”145 Struggles like this can spur broader democratic engagements, for “[d]emocracy in its best sense is much more than a mechanism for choosing leaders. At its heart it is a method, a way of living based on the principle that “sovereignty” (ultimate political authority) resides in the people.”146 Much of this microlocal litigation action eschews standard American individualism in favor of a “collective responsibility” that appears particularly salient at the microlocal level, and is at the heart of this idea of democracy.147 Through bringing claims or changing the

See, e.g., https://www.chimneyhillmud.com/trash-water/water/whatexactly-is-a-water-district/ 143 Shoked, supra note 3, at 1384. 144 Danaher, supra note 83, at 6. 145 Id. at 9. 146 Id. 147 Id. at 2. 142

32 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

playing field on which claims are brought, these tiny governments are engaged in a deeply democratic activity.148 Further, in some instances, bringing forward litigation can reaffirm the value of a particular level of public entity.149 In the case of school districts, “the primacy of school districts” as the most important entity in the context of education law at least has recently receded.150 Instead, the Supreme Court has “moved past the traditional view of school districts as homogenous and unitary entities,” and issued decisions that have “implied that individual school boundaries, not school district boundaries, are the smallest meaningful boundaries in education law.”151 State law reforms, too, have broken apart the notion of school districts as unified entities and transformed “the individual school – the micro-local unit – into a legal entity.”152 But when a school district acts as a plaintiff, it reaffirms itself as a meaningful legal entity that

Another argument that is levied against public tort litigation is that these entities are not really engaged in democratic behavior at all; rather, they are the puppets of unscrupulous plaintiff attorneys who prey on public entities and are only interested in a quick pay day. A recent article on school districts suing social media companies, for example, opens with the subheadline: “Lawyers seek to cash in on tech panic by enlisting school districts to sue.” Brown, supra note 70. This argument disparages the intellectual capacities of government entities. Entities of any scale are capable of determining whether bringing forward a public tort suit is a desirable option for them. Notably, most of these tiny governments have few resources, and contingency arrangements are likely one of the few mechanisms they can access to fund such claims. Again, as the claims between these tiny governments and multi-national corporations bring into sharp relief, these are very differently-resourced entities. Local governments are highly restricted in how they can raise revenue and funding, and they pay their legal expenses, if any, through public funds. For additional discussion of some of the issues raised by contingency fee attorneys in the public litigation realm, see Margaret H. Lemos, Privatizing Public Litigation, 104 GEORGETOWN L. J. 515 (2016). 149 Swan, supra note 8. 150 Id. 151 Shoked, supra note 3, at 1338. 152 Id. at 1339. These reforms include “(1) the ‘right to attend a specific – as opposed to any – school in the district; (2) individual school management; and (3) the power to dispute a district’s decision to close a school.” 148

33 20xx]

Draft – Please Do Not Cite

Draft

corresponds with its own distinct “public.”153 Similarly, rural towns can sometimes dismissed as essentially non-places:154 these litigative moves and ordinances reassert these communities as relevant entities.155 CONCLUSION Litigative activities at the microlocal level are riddled with ironies of scale. The reason why these entities resort to litigative moves in the first place is that they have few other formal legal tools to rely on: because they are small they cannot regulate as larger governments can (school districts cannot regulate social media companies, for example), yet they bear the harms created by those regulatory failures. Then, their small size is what essentially allows these entities to serve as tiny “laboratories of democracy” and engage in different litigative activities to see what might be successful in solving a particular problem. Ironically, again, though, their small size can sometimes defeat the efficacy of their efforts: in the context of the CELDF ordinances, their small size enables entities to experiment with ordinances that could face significantly more resistance in larger government entities, but is also because of their size that these ordinances are more expressive than legally meaningful. At any rate, however, the litigation brought by these tiny governments or microlocal units constitutes an important move in public tort development, and the CELDF-inspired ordinances may in fact constitute one of “one of the most rebellious, and radical” moves happening “in American local government today.”156 The tiny government CELDF ordinances, along with the more traditional microlocal litigation occurring between school districts and vaping and social media companies, and other districts and corporate

City v State How State Law Prevents Towns From Keeping Out Businesses, ALU (Aug. 24, 2018). 154 Id. 155 Id. 156 Miller, supra note 130. 153

34 Draft

Draft – Please Do Not Cite

[Vol. XX:nnn

misfeasors, represent a new layer of democratic action in the public litigation realm. They also invite richer thinking and suggest new visions for power balances between communities and corporations. While microlocal units like school districts may be able to achieve litigative success in the courtroom to help offset the harms currently caused by corporate activities and products, the radical solutions arising from the tiny governments passing radical new ordinances fundamentally overhauling the corporate-community relationship invite all of us to think bigger when it comes to lasting solutions for the seemingly interminable issue of massive corporate wrongs.