ILR Press 2018 Sampler

ILR P R ES S • • CORNELLL UNIVERSITY P R ESS

CORNEL L PRE SS.CORN ELL.EDU

SAMPLER 201 8

This sampler contains short excerpts from the following forthcoming books published by ILR Press, an imprint of Cornell University Press. Full details about each book are available on our website, cornellpress.cornell.edu.

Michael Evan Gold, A Primer on Legal Reasoning, available November

Julee T. Flood and Terry L. Leap, Managing Risk in High-Stakes Faculty Employment Decisions, available November

Ines Wagner, Workers without Borders: Posted Work and Precarity in the EU, available November

Philip Rathgeb, Strong Governments, Precarious Workers: Labor Market Policy in the Era of Liberalization, available December

ILR PRESS SA M P LE R

2 01 8

[FREE]

CORNELL UNIVERSITY PRESS

A Primer on Legal Reasoning

Michael Evan Gold

ILR Press an imprint of Cornell University Press Ithaca and London

3050-2726-Fullbook.indb 3

6/4/2018 7:40:00 PM

Copyright Š 2018 by Cornell University All rights reserved. Except for brief quotations in a review, this book, or parts thereof, must not be reproduced in any form without permission in writing from the publisher. For information, address Cornell University Press, Sage House, 512 East State Street, Ithaca, New York 14850. Visit our website at cornellpress.cornell.edu.

First published 2018 by Cornell University Press Printed in the United States of America Library of Congress Cataloging-in-Publication Data <CIP to come>

3050-2726-Fullbook.indb 4

6/4/2018 7:40:00 PM

Brief Contents

Introduction 1 § 1. Issues

3

§ 2. Identifying the Governing Rule of Law

22

§ 3. Levels of Abstraction

35

§ 4. Deduction

65

§ 5. Induction

88

§ 6. Arguments in General

105

§ 7. Arguments Classified by Function

114

§ 8. Arguments Based on Evidence

122

§ 9. Policy Arguments

124

3050-2726-Fullbook.indb 5

6/4/2018 7:40:00 PM

v i Bri e f Conte nts § 10. Doctrinal Arguments

139

§ 11. Analogies and Precedents

150

§ 12. Distinctions

180

§ 13. Holding and Dictum

209

§ 14. Reductios ad Absurdum

213

§ 15. Subjective and Objective Standards

223

§ 16. Interpreting Statutes

232

§ 17. Prima Facie Case, Affirmative Defense, Burden of Proof

245

§ 18. Application of Law to Fact

259

§ 19. A Model of Legal Argument

285

Answers 305 Index 341

3050-2726-Fullbook.indb 6

6/4/2018 7:40:00 PM

Introduction

Lawyers, particularly law professors speaking to prospective or beginning law students, are fond of stating that thinking like a lawyer is something special, something that is powerful, esoteric, and unique. Such statements are partly true. Thinking like a lawyer—analytical thinking, as lawyers call it (everyone else calls it “critical thinking”)—is a powerful tool because it allows one to understand and evaluate others’ arguments and to make sound arguments of one’s own. Further, analytical thinking is esoteric because most persons need advanced education to learn how to do it, and the subject matter to which it is applied is broad and complex. But the statements are partly false because analytical thinking is by no means unique. It is practiced in many disciplines, from engineering to economics, from statistics to philosophy. (For this reason, we wrote that learning analytical thinking requires advanced education, not law school.) The first premise of this book is that legal reasoning is simply rigorous thinking applied to the domain of law.

3050-2726-Fullbook.indb 1

6/4/2018 7:40:00 PM

2 Introduction Legal reasoning is typically taught via the Socratic Method. In the early dialogues of Plato, the philosopher Socrates asked questions but gave no answers. Perhaps the questions he asked lacked good answers. Perhaps he wanted his interlocutors to figure out the answers for themselves. And perhaps (given that he was called a gadfly) he wanted to show them that they did not know as much as they thought they did. Most classes in law school resemble a dialogue with Socrates. The professor asks many questions and explains very little, either about the content of the law or about legal reasoning. A synonym for the “Socratic Method” as practiced in many law schools is “hide the ball.”

The second premise of this book is that the ball need not be hidden, that analytical thinking can be taught in as transparent and orderly a manner as is mathematics or any other academic subject. Courses on critical thinking have precisely the same premise. Thus, this book is an introduction to critical thinking in the context of law. This book is not a shortcut to becoming a lawyer; it is not a substitute for law school. Rather, the book is a discussion of some of the most important techniques of reasoning used by lawyers—and by other rigorous thinkers.

The author believes that this book can be useful to two classes of reader.

The first class comprises college students who are preparing to go to law school and law students who are in the first months of their legal education. Being familiar with the forms of legal argument will help students understand cases and respond to professors’ hypothetical questions. The second class comprises anyone else who wants to understand and evaluate others’ arguments and to make sound arguments of one’s own. All citizens in a democracy need to be able to think critically about public issues, to cut through politicians’ rhetoric and judge the validity of their assertions.

Thinking analytically is a necessary step toward that end.

3050-2726-Fullbook.indb 2

6/4/2018 7:40:00 PM

§1

Issues

The purpose of this section is to introduce the student to the concept of issues. The student will learn how issues are created and how they are classified. In addition, the student will learn how to identify an issue.

I. Definitions Issues and resolutions go by various names, depending on the field of inquiry. In daily life, an issue is often called a “question,” and the resolution of a question is an “answer.” question Should Lennis and Phela paint the den pale blue, light green, or off white?

3050-2726-Fullbook.indb 3

6/4/2018 7:40:00 PM

4 § 1 answer Pale blue.

In science, the resolution of an issue is a theory. issue What is the mechanism of evolution? theory Changes in species are explained by natural selection.1 In law, the resolution of an issue by a tribunal such as a court or an administrative agency is a finding or a holding. Some holdings pertain to what the law should be. issue Under the National Labor Relations Act, is an employer responsible for an unfair labor practice committed by a supervisor? holding Yes.2

1. Before it is proved, a scientific theory is called a “hypothesis.” When a scientific theory is well supported by evidence, the theory is often called a “law” or a “law of nature.”

2. This book focuses on the interpretation of existing law, not the creation of new statutes. Nonetheless, it may be noted that the resolution of issues by a legislature is similar to resolution by a tribunal, though the terminology differs. When a legislature enacts a bill, the result is a statute or an ordinance. For example— issue Should the legislature enact a statute that outlaws sending spam? statute Sending spam shall be a felony punishable by thirty years of torture followed by death by kimchi.

As they consider whether to enact statutes, legislatures often make determinations about facts of the past, present, or future. Such determinations are called “findings.” For example— issue Is spam harmful? finding Spam costs America 666 gazillion dollars per day in lost productivity.

3050-2726-Fullbook.indb 4

6/4/2018 7:40:00 PM

I s s u es 5 Other holdings pertain to facts. issue Did supervisor S of employer ER discharge worker WR because WR was soliciting coworkers to join a union? holding Yes. Still other holdings pertain to the legal significance of facts. issue Is employer ER guilty of an unfair labor practice because supervisor S discharged worker WR for soliciting for a union? holding Yes. A law case typically boils down to one or a small number of issues.

When a tribunal makes holdings that resolve the issues, one party wins the case and the other party loses it.

As the foregoing illustrations demonstrate, issues and resolutions come in pairs. A resolution applies to the issue that generates the resolution, but not to other issues. Lennis and Phela’s answer applies to the den in their house at this time; their answer does not apply to another time or another room. The theory of natural selection may explain evolution, but does not explain the orbit of the earth or the size of the gross domestic product. The holding that an employer is responsible for an unfair labor practice committed by a supervisor applies to the Labor Act, but not necessarily to other statutes.3 The finding that S discharged WR for soliciting for a union applies to these persons at a particular moment, but not to other moments or to other persons.

An issue-and-resolution pair is often related to other issue-and-­ resolution pairs. For example, Lennis and Phela’s decision about the color to paint their den relates to the issues of whether they should do anything to the den or leave it alone, whether they should paint the den or wallpaper it, whether they should do the work themselves or pay someone else to 3. For example, this holding does not tell us whether an employer is responsible under Title VII of the Civil Rights Act of 1964 for a supervisor’s sexual harassment of a subordinate.

3050-2726-Fullbook.indb 5

6/4/2018 7:40:00 PM

6 § 1 do it, whether they should remodel the den this year or next, and whether they should borrow the necessary money or take it from their savings. Of course, many other decisions that Lennis and Phela need to make are not related to these issues, for example, whether to eat Thanksgiving dinner with Lennis’s family or Phela’s, or whether to tell their daughter the truth about Santa Claus.

II. Creation of Issues If an issue is something that persons disagree about, issues are created by disagreements. But which aspect of a disagreement creates the issue? The answer is the facts. Issues grow out of facts, and facts define issues.

Lennis and Phela want to paint the den. Lennis prefers off white; Phela prefers pale blue. These facts generate the issue, which color should the den be painted? These facts do not generate the issue, should the den be painted or wall papered, nor the issue, should Lennis, Phela, or a professional painter paint the den? Of course, if the facts changed, these issues might arise. For example, suppose that after Lennis and Phela decide to paint the den pale blue, Lennis wants to paint the room himself and Phela prefers to hire a painter. From these facts arises the issue, who should paint the den?

Here is a legal example. Supervisor S and worker WR are employed by ER. S discharges WR, who has been soliciting coworkers to join a union.

When WR asks why she is being discharged, S replies that her job performance has become unsatisfactory, but WR believes the true reason is her union activity. These facts generate the issue, did S discharge WR because of WR’s union activity? These facts do not generate the issue, is ER in the class of employers to whom the Labor Act applies, nor the issue, is ER responsible for S’s act? The reason is that nothing in the facts indicates that the parties disagree about whether the Labor Act covers ER or whether ER is responsible for S’s act. But the facts do show that the parties disagree about why S fired WR: ER asserts that the reason was WR’s poor job performance, and WR asserts that the reason was her union activity.4 4. If the facts were different, other issues would arise. For example, suppose that S privately supports unionization of the firm, and S and WR secretly conspire to manufacture an issue that would make ER look bad in the eyes of the workers. These facts would generate the issue, is ER responsible for S’s discharge of WR?

3050-2726-Fullbook.indb 6

6/4/2018 7:40:00 PM

I s s u es 7

III. Resolution of Issues When parties cannot themselves settle a dispute or disagreement, and wish to settle it peaceably, they ask someone else to resolve it. The one who resolves the dispute may be a person such as a relative, a friend, or an arbitrator, or an institution such as a court or an administrative agency such as the National Labor Relations Board. In this book, the term tribunal will refer to any person or institution that parties ask to resolve a dispute.

A dispute is composed of one or more issues.5 To resolve a dispute in a rational way is to resolve those issues. A tribunal must perform four functions in order to resolve an issue. The functions are— • Frame the issue in the dispute • Find the facts that pertain to the issue • Identify and interpret, or if necessary create, the authority that governs the issue • Apply law to fact, i.e., use the authority to determine the legal consequences of the facts of the dispute Issues may be classified according to the function they perform.

A. Frame the Issue in the Dispute The first function a tribunal performs in resolving a dispute is to identify the issue. Parties usually agree on the issue in a case. However, sometimes it is advantageous to one party to frame the issue in a particular way and advantageous to the other party to frame the issue in another way. For example, suppose the forensic society holds a debate over outlawing automatic firearms. The opponent of such a law might frame the issue as, may a state infringe on citizens’ right to bear arms under the Second Amendment? The advocate of such a law might frame the issue as, may a state

5. This book deals only with disputes that can be settled in a rational way. Of course, not all disputes can be settled rationally. Many disputes are emotional. The issue in an emotional dispute may never be identified. Indeed, there may be no issue in the sense in which we are using the term. Even if one exists and is identified, the issue in an emotional dispute may be settled without resolving (and often without addressing) the issue. The parties decide to make up with one another and go on with life.

3050-2726-Fullbook.indb 7

6/4/2018 7:40:00 PM

8 § 1 protect its citizens from mass murder? This disagreement generates an issue as to the issue in the debate.

Here is a legal example. In the late nineteenth century and first third of the twentieth century, the federal courts played an active role in reviewing federal and state statutes pertaining to employment, and the courts struck down as unconstitutional many (albeit not all) such statutes.6 The basis of these decisions was the Fifth and Fourteenth Amendments to the Constitution, which prohibit state and federal government from “depriv[ing] any person of life, liberty, or property, without due process of law.” The Supreme Court held that the word “liberty” in these amendments includes the right to enter contracts on any terms the parties may choose. The government may not limit liberty of contract, that is, may not prohibit certain terms in a contract, except when the limitation is justified by a strong public policy. The law regards employment as a contract between an employer and a worker. In Holden v. Hardy Utah enacted a statute limiting the workday of underground miners to eight hours. Although the statute infringed on the parties' liberty of contract by prohibiting them from agreeing to a workday longer than eight hours, the Court upheld the statute. The Court reasoned a state has a strong interest in protecting the safety of miners. Accidents increase with workers’ fatigue; accordingly, Utah could reduce fatigue and, therefore, injuries by limiting hours at work. But in Lochner v. New York a state statute limited the workday of bakers to ten hours, and the Supreme Court invalidated this statute. The Court held that the statute interfered with employers’ and workers’ liberty to agree on the length of the workday, and New York had failed to demonstrate that a longer workday jeopardized the safety of bakers.7

6. Today the courts are much less active in this regard; they accept legislative judgments about the need for regulation of employment. Nonetheless, the example that follows in the text is still valid because legal reasoning has not changed.

7. The theory that the employment relationship is contractual is an accurate description when an employer and a worker actually negotiate (exchange offers) over the terms of employment (wages, hours, etc.). In the typical case, however, the employer dictates the terms of employment, and the worker accepts them or looks for another job. Seeming to ignore this fact, the law presumes that the employer makes an offer of a job to a worker. The offer contains terms of employment. If the worker takes the job, the worker accepts the employer’s offer, and a contract results.

3050-2726-Fullbook.indb 8

6/4/2018 7:40:00 PM

I s s u es 9 Now suppose that shortly after these cases were decided, a state promulgated a regulation requiring that a hair stylist must wear gloves when applying dye to a customer’s hair and that the owner of the salon must supply gloves at no cost to stylists. The association of salon owners challenges the regulation, and the state defends it. The owners frame the issue as, does the regulation interfere with liberty of contract? They frame the issue this way because Lochner v. New York prohibits a state from interfering with liberty of contract, and the regulation restricts owners and stylists from agreeing that stylists need not wear gloves or that stylists will pay for their gloves. By contrast, the state frames the issue as, does the regulation protect the safety of workers? The state frames the issue this way because Holden v. Hardy held that a state may protect the safety of workers, and gloves protect stylists from chemicals in hair dye.

Because the parties disagree on the issue, the tribunal will frame it.

B. Find the Facts That Pertain to the Issue A tribunal’s second function in resolving an issue is to find the facts of the events involving the parties. Sometimes the parties agree on one or more facts; in this event, they stipulate to those facts, and no issue of fact regarding those facts arises. Other times the parties disagree about the facts.

“I didn’t change the channel.”

“Yes he did.” *** “You made a snarky remark behind my back.”

“No, I didn’t. It must have been some other girl.” *** “You did it on purpose.”

“No, I didn’t. It was an accident.”8 ***

8. Note that the fact in dispute in this example is a party’s state of mind. State of mind— that is, what a party was thinking—can be crucial in a case. The student may be thinking

3050-2726-Fullbook.indb 9

6/4/2018 7:40:00 PM

1 0 § 1 “You never told me what you were going to do.”

“I told you, but you weren’t listening.”

Such disagreements give rise to issues of fact. When parties disagree about facts, the tribunal must determine what happened—or, as a lawyer would say, the tribunal must find the facts.

C. Identify and Interpret, or If Necessary Create, the Authority That Governs the Issue A tribunal’s third function in resolving a dispute is to identify and interpret the authority (i.e., the law) that governs the issue. Authorities abound in daily life, but in law three types of authority predominate: • An authoritative text such as a constitution or statute.

The Internal Revenue Code provides that federal income taxes must be paid by April 15.

• A precedent, that is, a tribunal’s previous resolution of an analogous dispute.9 “When Johnny threw his broccoli at his sister, I told him that was naughty of him; and now you’re doing the same thing.”

• A principle of public policy or the common good.10 “Protective tariffs are good for the economy of Lilliput.”

that we can never know what goes on in another person’s mind; but humans have made such judgments from time immemorial, and the law regards state of mind as a fact. For example, although the following statement is not recorded in the Iliad, it may well have occurred. Suppose Helen sued Hecuba for unkindness. (Unkindness is not a legal claim today, but let us suppose it was a legal claim in ancient Troy.) The basis of Helen’s suit was that Hecuba had said to Helen, “You look awful today.” Now, a good critic is one’s best friend, but a malicious critic is an abomination. If Hecuba wanted Helen to look good in order to hearten Paris, the criticism was good; if Hecuba wanted to make Helen feel bad for causing the Trojan War, the criticism was bad. It all depended on Hecuba’s state of mind. The tribunal deciding Helen’s case would consider Hecuba’s state of mind to be a fact that could be found as any other fact is found.

9. The word “precedent” is sometimes used to mean any prior decision. In better usage, “precedent” has a more limited meaning, namely, a prior decision that controls the present case. The difference between these meanings is apparent in the sentence, “Up v. Down was decided long ago and is not a precedent for the case at bar.”

10. Synonyms for “public policy” include “social policy,” “public welfare,” and “social welfare.”

3050-2726-Fullbook.indb 10

6/4/2018 7:40:01 PM

I s s u es 1 1 Parties may disagree about which authority governs an issue. A says that authority α applies to the issue, and B says that authority β applies.

“This case is governed by Pennsylvania’s law of contract.”

“No, this case is governed by the National Labor Relations Act.” Parties may also disagree on how to interpret an authority. A and B agree that authority γ applies; but A says γ means this and that, whereas B says γ means the other thing.

“Thou shalt not kill.”

“True, but what does ‘kill’ mean? Does it include self-defense?”

Both types of disagreement are issues of law, which a tribunal must resolve.

Sometimes no relevant authority exists. In this event, the tribunal’s decision may become a new authority.

D. A pply Law to Fact (i.e., use the authority to determine the legal consequences of the facts of the dispute) The fourth function in resolving a dispute is applying the governing authority to the facts of the dispute. After the tribunal has found the facts, and has identified and interpreted the governing authority, the question arises, what did the law require these parties to do in these circumstances?

The parties may disagree about the answer to this question. A says that, under the authorities, he did the right thing. B says that, under the same authorities, A did the wrong thing.

“You were negligent to install a swimming pool without putting a fence around it.”

“I was not negligent. There was a fence around my yard, and the pool was in the yard.”

This sort of disagreement is an issue of application of law to fact. The tribunal determines the legal consequences of the parties’ behavior.

IV. Identifying Issues Perhaps the most daunting task facing a beginning student of law is identifying or “spotting” issues. Curiously, we all are able to identify issues in daily

3050-2726-Fullbook.indb 11

6/4/2018 7:40:01 PM

1 2 § 1 life, but we do not know how we do it (or even realize that we are doing it).

In fact, identifying issues is simple and, once the student understands how it is done and practices it, identifying legal issues will not be difficult.

A student needs to identify an issue in two different contexts. The first context is reading a document such as a law case, an article (whether in a magazine or a professional journal), or a statute. The student must know what issue(s) the author is addressing. The second context is listening to or reading a set of facts. The student must know what issue(s) the facts generate. Let us begin with the first context.

A. Identifying Issues in Documents Issues occur in various types of documents. For example, suppose one encounters the following editorial in a newspaper.

The College Daily An American is legally an adult at eighteen years of age. One may vote, run for political office, be conscripted into the armed forces, and be held fully accountable in a court of law for one’s criminal behavior. One may leave home and be free of parental supervision. One may marry, divorce, obtain a credit card, and in general enter into binding legal relationships.

One may even view or purchase pornography. But the law also provides that one may not consume alcoholic beverages until one turns twenty-one.

Need anything more be said?

Doubtlessly, the student knows that the issue that this passage addresses is whether the drinking age should be reduced to eighteen. But how does one know this?

Identifying issues is essentially a matter of drawing analogies. Analogies and distinctions are discussed in detail in § 11 Analogies and Precedents and § 12 Distinctions, but the student’s intuitive understanding of analogies will suffice for present purposes.

The student recognized the issue in the foregoing editorial because of an analogy. The student knew that people in the past have disagreed about what the minimum drinking age should be, or about whether there should be one at all. Knowing this history, the student (probably without realizing it) drew an analogy between the past and the present. The student

3050-2726-Fullbook.indb 12

6/4/2018 7:40:01 PM

I s s u es 1 3 found that the facts of the past and the present were similar; and knowing that the issue arose in the past, the student concluded that the same facts generated the same issue in the present. Therefore, the student knew what issue the author of the editorial was addressing.

One cannot draw an analogy between the past and the present without knowing the past. Thus, the student’s first task is to learn what issues have arisen in the past.11 Unfortunately, no book in the library contains lists of issues. The student will have to discover the issue(s) in a case. Fortunately, the student already has some ability to recognize disagreements, to sense when points of view clash; and a disagreement is another name for an issue. Accordingly, a student reading a case should make a list of each point of disagreement (issue) between the parties or the judges. The list should include the facts underlying the disagreement.

How can the student spot a disagreement? A disagreement may be ex­­ plicit in an opinion. Blessed are the judges who write, “The issue is. . . .”

Often, however, disagreements are implicit, and the student must learn to recognize signals. A common signal of a disagreement or issue is, “One party argues . . . ,” followed by either, “The other party argues . . .” or “This argument is mistaken because . . .” But (too often!) the opinion gives no signals and merely presents the judge’s reasoning. In this event, the student must dig beneath the surface of the opinion.

Perhaps the best tool for digging into an opinion is to ask— • What claim is the author trying to make?

A claim (or point or position) is the resolution of an issue. Therefore, by changing a claim into a question, the student will identify an issue.

For example, in the Case of the College Daily, the editorial contains an implied claim that eighteen-year-olds should be allowed to consume alcohol. Accordingly, the issue was, should the law be changed to allow persons aged eighteen and older to consume alcohol?

Other useful tools for digging into an opinion are questions such as— • Why is the tribunal saying this? What obstacle is the tribunal is trying to surmount?

11. Of course, new issues arise, but they are usually easy to recognize. Commonly, they are explicitly identified as questions. Does a woman have a constitutional right to an abortion? Should “intelligent design” be taught alongside of evolution in biology courses?

3050-2726-Fullbook.indb 13

6/4/2018 7:40:01 PM

1 4 § 1 • What would the counterargument be? How might one disagree with this?

• If the tribunal wanted to reach the opposite result, what would the tribunal have said?

• Wherein might a reasonable person disagree with this opinion?

Could the outcome be different?

Answers to these questions will lead to the issues that the tribunal was confronting.

Another useful tool for digging issues out of legal opinions is to list each rule of law stated in a case. Whenever the opinion announces a rule of law, the student knows that an issue lurks behind the rule: every rule of law represents the resolution of an issue over which parties, at one time or another, have disagreed. Change a rule of law into a question, and it becomes an issue. For example, the rule, “Thou shalt not steal” resolves the issue, “Should a commandment forbid stealing?” Or consider the following e: When Is an Offer Accepted?

On July 1, S sent letters to several friends, offering to sell her dining room furniture for $4,000. The letter stated, “If you want it and accept my price, mail me your check for $4,000 and tell me when you’d like to pick up the furniture.” B found the letter in her mailbox when she came home from work in the evening of July 3. She immediately wrote a check, put it in a stamped envelope addressed to S, and deposited it in a mailbox at 6:00 p.m. Normally, the Post Office would have picked up the letter at 9:00 a.m. of the following morning, but, observing the Independence Day holiday, the Post Office collected the letter on July 5 and delivered it to S the following day. In the meantime, S held a party on July 4, and she mentioned to her guests that her dining room furniture was for sale. “How much do you want for it?” asked X. “Oh, I don’t know, it’s very nice, but I don’t know anything about these things,” replied S coyly. “Will you take $5,000 for it?” offered X. “Yes, I will,” answered S, and X immediately wrote a check. B, having always admired this furniture, was upset to learn that S had accepted X’s check, and she went to a lawyer. “That furniture belongs to me,” said B. “I mailed my check on July 3, before X even knew the furniture was for sale. The lawyer told B that she was right. “The rule

3050-2726-Fullbook.indb 14

6/4/2018 7:40:01 PM

I s s u es 1 5 in this state,” said the lawyer, “is that, when an offer contemplates that it will be accepted by mail, a contract is formed when the party accepting the offer [in legal parlance, the ‘offeree’] deposits into a mailbox a properly stamped letter of acceptance addressed to the party who made the offer [the ‘offeror’].”

Like every rule, this rule resolves an issue, and the student knows the issue as soon as one knows the rule. The rule was that an offer is accepted when the offeree mails the acceptance. Therefore, the issue was, “At what point in time is a contract formed if the offer contemplates acceptance by mail?”

A problem with identifying the issue by listing each rule of law is that most rules that are mentioned in a case, or that are not mentioned but are conceptually necessary in order to decide the case, are not in dispute.

The parties accept these rules as valid. For example, a rule of law that is necessary in every civil case is, the plaintiff must initiate the action by filing a complaint and serving it on the defendant. The beginning student’s list of the rules of law in a case will probably contain some such “background” rules. These rules are surely worth learning for their own sake, but they will not lead to the issue in the case. How does a student distinguish between the background rules and the rules that lead to the issue(s) in a case? One way is to identify the claim (or point or position) that the tribunal is trying to make (or defend). Usually a claim is the resolution of a disputed issue in the case. Another way to distinguish between background rules and the issue in a case is to follow the action. Feel the heat. Look for the disagreement, the conflict, the clash of ideas—the legal reason why the case is in court.12 Identifying issues in statutes is similar to identifying issues in cases. One knows that a passage in a statute generates an issue because one knows that a similar passage in another statute generated that issue. For example, Title VII of the Civil Rights Act of 1964 uses the term “employee.” Does this term generate any issues? The answer is yes, as we know because the term

12. The legal reason must be distinguished from the personal reason. Parties have their personal reasons with which the law is not concerned. A’s personal reason for suing B may be that B insulted A. The legal reason the case is in court must be that an issue that the law considers significant has arisen between the parties; for example, did B purposefully misrepresent material facts on which A reasonably relied and thereby lost money? The parties may be unaware of the legal reasons, but the lawyers and the judges know those reasons.

3050-2726-Fullbook.indb 15

6/4/2018 7:40:01 PM

1 6 § 1 “employee” also appears in the National Labor Relations Act and generated issues under that act. Thus, under the Labor Act the issue arose, “Is a supervisor an ‘employee’?” and the same issue later arose under Title VII.13 B. Identifying Issues in Facts So far this section has discussed identifying issues in a document. Now let us discuss identifying issues in a set of facts. The student must know which (if any) issues that a set of facts generates. Of course, a lawyer must identify the issue(s) in a story told by a client who asks for advice. Other persons also must identify the issue(s) in the situations that occur in daily life, as illustrated below.

Like identifying issues in a document, identifying issues that lurk in facts requires drawing analogies, and drawing analogies requires knowledge of the past. Consider the following summary of a tale that Jonathan Swift relates in chapter 4 of Gulliver’s Travels.

The inhabitants of Lilliput opened their soft-boiled eggs at the small end; the denizens of Blefuscu opened their eggs at the big end. Thousands of lives had been lost in battles, and another was about to commence, over which end of an egg should be opened.

This issue had not arisen in England, and therefore Gulliver had no idea that the issue existed until he was informed about it. Had he been offered an egg to eat, he would not have known that cracking open one end or the other would involve him in a ferocious controversy.

Gulliver did not know the history that preceded the situation in which he found himself and, therefore, did not know the issue that was about to affect his life.

It follows that identifying the issues in a set of facts is essentially the same as identifying the issues in a reported case. The student asks oneself,

13. Under the Labor Act, a supervisor is not an “employee” and is not protected by the act. Under Title VII, a supervisor is an “employee” and is protected by the act. The resolutions of the issues are different because the purposes of the statutes are different, but the important point for present purposes is that the same issue arose under both statutes.

3050-2726-Fullbook.indb 16

6/4/2018 7:40:01 PM

I s s u es 1 7 are the facts of this situation or case the same as, or similar to, the facts of issues with which I am familiar? The guiding principle is the same in both contexts: • Similar facts beget similar issues.

When Ham and Shem came to a fork in the road, they disagreed about which was the better route to Zuzim. Ham said the left was shorter; Shem said the right was faster. When Meshech and Shadrach came to the same fork, the same issue may well have arisen.

Now consider these not entirely hypothetical cases:

The Coconut Cases Case 1 A sophomore at an American university in New York State approaches the Coconut Bar, which is near the campus. A hefty bouncer standing near the entrance says, “Let me see your ID.” What is the issue generated by these facts? Is it whether the sophomore resides in the county in which the bar is located? Is it whether the sophomore meets minimum height and weight requirements? Of course not.

The student knows that the issue in Case 1 is whether the sophomore is at least twenty-one years old, and the student knows this is the issue because the student (probably without realizing it) has drawn an analogy. The facts of Case 1 are analogous to Cases A, B, C . . . in the past in which bouncers were employed to check the identification of young persons who sought to enter bars. The student knows that the issue in Cases A, B, C . . . was whether the potential patrons were legally old enough to drink. Applying the principle that similar facts beget similar issues, the student has concluded that the issue that arose in Cases A, B, C . . . also arises in Case 1.

Case 2 An octogenarian approaches the same bar. Without saying a word, the bouncer opens the door.

3050-2726-Fullbook.indb 17

6/4/2018 7:40:01 PM

1 8 § 1 Why does the bouncer fail to check the octogenarian’s identification? Is the bouncer negligent about his duty? Is the octogenarian a relative of the bouncer? Of course not. The student knows that the bouncer does not check the octogenarian’s identification because Case 2 is analogous to Cases R, S, T . . . in which bouncers did not check the identification of elderly patrons. Like the elderly patrons in Cases R, S, T . . . , the octogenarian in Case 2 is obviously old enough to drink. Therefore, concludes the student, Cases R, S, T . . . and Case 2 are analogous. The octogenarian is plainly old enough to drink; the issue of minimum age does not arise, and the bouncer has no need to check the octogenarian’s identification.

Case 3 A twenty-two-year-old student from Denmark enrolls as a graduate student in an American university in New York State and wants to enter the Coconut Bar. The bouncer asks her for identification. She is not carrying satisfactory proof of age and is turned away. She asks, “What’s going on here?” Recognizing a foreign accent and having experience with foreigners, the bouncer understands the question and explains.

The woman from Denmark does not carry proof of age because she has entered bars at home without having to prove her age. She draws an analogy between bars in Denmark and bars in America and concludes that she can enter an American bar without proof of age. The analogy is false, however, because Denmark has no minimum drinking age whereas America does.

Case 4 Subsequently, the Danish student is preparing to accompany friends to the Palm Bar, a tavern popular among graduate students. She puts her passport in her purse.

Without consciously thinking it, she has drawn an analogy between Cases 3 and 4. The Palm and the Coconut are both bars, and they are both in America. Knowing that similar facts beget similar issues, she knows that she will need to prove her age in order to enter the Palm.

3050-2726-Fullbook.indb 18

6/4/2018 7:40:01 PM

I s s u es 1 9 Case 5 Our Dane plans to go to dinner in the Acorn, an upscale restaurant. Being a European, she wishes to order wine with her meal. She is unsure whether she needs to take her passport with her.14 She asks herself the question (probably consciously this time), “Are the facts of this situation analogous to my experience at the bars?” The answer is not obvious because she can think of arguments on both sides of the matter: “On the one hand, alcohol is alcohol, so perhaps the silly American rule applies to restaurants. [In other words, Case 5 is analogous to Cases 3 and 4.] On the other hand, alcohol consumed with a meal has a lesser effect than alcohol consumed on an empty stomach, and the environment of a restaurant is more controlled than the environment of a bar; so perhaps the rule does not apply to restaurants. [In other words, Case 5 is not analogous to Cases 3 and 4.]”

A lawyer is frequently in the same position as the Danish student in the foregoing cases. A client tells a story, and the lawyer must identify the issues raised by the facts of the story. The practicing lawyer draws analogies. The lawyer asks, “Are the facts of my client’s story analogous to the facts of cases in the past or the facts contemplated by a statute?” As ever, the guiding principle is— • Similar facts beget similar issues.

V. Review Questions More than one answer may be correct to questions 1–5.

1. An issue is— □ (a) a flimsy piece of paper. □ (b) a question with different reasonable answers. □ (c) a point of contention between parties. □ (d) the reason a case comes before a tribunal. □ (e) generated by facts.

14. In this example, the question is trivial. Taking her passport with her costs nothing, so she may as well take it in the event she has to prove her age. But we are considering theoretical, not practical questions. Our question is, does she need to take her passport with her? To make the question important practically, let us assume that the Danish government charges ten thousand Krone to replace a lost or stolen passport, and our student carries it with her only when necessary.

3050-2726-Fullbook-FM-Ch5.indd 19

6/5/2018 10:01:40 PM

2 0 § 1 2. In law the resolution of an issue is called— □ (a) the answer. □ (b) the theory. □ (c) the judgment. □ (d) the holding.

3. The steps in resolving any issue are— □ (a) identify the issue in the dispute. □ (b) find the facts that pertain to the issue. □ (c) identify and interpret the authority that governs the dispute. □ (d) apply the authority to the facts of the dispute. □ (e) prevent future injury. □ (f) determine which party is offering the larger bribe.

4. The guiding principle in recognizing an issue is— □ (a) anything can be an issue in any case. □ (b) look for words in italics or boldface. □ (c) similar facts beget similar issues.

5. One recognizes an issue by— □ (a) remembering (or finding) an analogous case in the past; an issue that arose in that case probably arises in the case at hand. □ (b) listing the points of disagreement between the parties or judges. □ (c) looking for phrases such as, “one party argues . . .” or “party A is mistaken because . . .” □ (d) asking, what claim is the author trying to establish? □ (e) asking, why did the judge write this? □ (f) asking, what is the counterargument? □ (g) asking, could a reasonable person disagree with this? □ (h) asking, could the law be different? □ (i) focusing on the footnotes, where everything important appears. □ (j) making a list of each rule of law in the case. □ (k) calling your mother.

6.

Sally’s Case Sally transferred as a sophomore to First Choice U. She had spent her freshone year at Second Choice U., where she had rushed and was accepted

3050-2726-Fullbook.indb 20

6/4/2018 7:40:01 PM

I s s u es 2 1 into membership in βΦΦ, a sorority. Upon enrolling in First Choice, she applied and was accepted into βΦΦ’s chapter at First Choice without going through the chapter’s usual initiation ritual.

Red is the primary color of First Choice’s archrival. A bylaw of βΦΦ at First Choice states, “No member of this chapter shall wear red to an athletic event. Any sister who violates this bylaw shall wash dinner dishes for a week.”

Sally attended the rally preceding the first game of the season. It was a chilly evening, and she decided to wear her warmest socks, which were red.

The next day Sister accused Sally of violating the bylaw. Sally replied that she did not know about it. Sister read it to her. Sally replied that a rally is not an athletic event.

What issues must be addressed by the judicial committee of βΦΦ?

7.

The Case of the Research Papers ta:

Professor, I’m almost through grading the research assignment, but I think you should read two of the papers yourself. prof: What’s the problem? ta: They’re pretty similar. prof: Well, we did say the students could work together on the research, although each student had to write one’s paper by oneself. So we can expect some similarities. ta: I think one copied from the other. prof: What they actually did certainly matters. The punishment for copying is usually heavier than the punishment for writing together.

What issues must the professor resolve before deciding whether to file charges against the students with the Academic Standards Committee?

VI. References Holden v. Hardy, 169 U.S. 366 (1898).

Lochner v. New York, 198 U.S. 45 (1905).

3050-2726-Fullbook.indb 21

6/4/2018 7:40:01 PM

Managing Risk in High-Stakes Faculty Employment Decisions

Julee T. Flood and Terry L. Leap

ILR Press an imprint of Cornell University Press Ithaca and London

3050-2728-1pass-0FM-r03.indd 3

5/16/2018 8:29:29 AM

Copyright © 2018 by Cornell University All rights reserved. Except for brief quotations in a review, this book, or parts thereof, must not be reproduced in any form without permission in writing from the publisher. For information, address Cornell University Press, Sage House, 512 East State Street, Ithaca, New York 14850. Visit our website at cornellpress.cornell.edu.

First published 2018 by Cornell University Press Printed in the United States of America Library of Congress Cataloging-in-Publication Data Names: Flood, Julee T., author. | Leap, Terry L., 1948– author.

Title: Managing risk in high-stakes faculty employment decisions / Julee T. Flood and Terry L. Leap.

Description: 1st edition. | Ithaca [New York] : ILR Press, an imprint of Cornell University Press, 2018. | Includes bibliographical references and index.

Identifiers: LCCN 2018013448 (print) | LCCN 2018015230 (ebook) | ISBN 9781501728969 (pdf) | ISBN 9781501728976 (epub/mobi) | ISBN 9781501728952 | ISBN 9781501728952 (pbk. : alk. paper) Subjects: LCSH: College teachers—Legal status, laws, etc.—United States. | College teachers—Selection and appointment—United States. | Universities and colleges—Law and legislation—United States. | Tort liability of universities and colleges—United States. | College teachers—Tenure—United States.

Classification: LCC KF4240 (ebook) | LCC KF4240. F56 2018 (print) | DDC 344.7301/01—dc23 LC record available at https://lccn.loc.gov/2018013448

3050-2728-1pass-0FM-r03.indd 4

5/16/2018 8:29:29 AM

Contents

Preface ix Acknowledgments xi Introduction 1 1. E stablishing a Career in Academia: Faculty Hiring, Evaluation, and Pay

28

2. Risk, Biases, and Logical Fallacies

52

3. Faculty Contracts

80

4. From Contracts to Constitutions: Faculty Free Speech Issues

108

5. Collegiality: An Enigma

133

Conclusion 155 Notes 171 References 185 Index 193

3050-2728-1pass-0FM-r03.indd 7

5/16/2018 8:29:29 AM

Introduction

Recruiting, hiring, and retaining college and university faculty members raises a number of important issues. Where do these institutions look to find the best faculty candidates, and how do these institutions identify the best candidates from the applicant pool while avoiding faculty who may later pose problems? On the one hand, did the University of Illinois hiring committee that extended a job offer to Rosalyn Sussman Yalow in the 1940s have any way of knowing that over three decades later she would win the Nobel Prize? On the other hand, would a more careful vetting have prevented the murderous rampage of biology professor Amy Bishop, who killed three of her colleagues—and wounded six—at the University of Alabama in Huntsville after she had been denied tenure there? Why did a background check not uncover Bishop’s erratic and violent history?

Likewise, would a more informed screening have identified serial sexual harassers before they had the opportunity to cause disruptions at some of the most prestigious universities in the United States?1 Or could more prudent risk management have avoided the overeager hiring of a “next

3050-2728-1pass-0Intro-r03.indd 1

5/16/2018 6:45:34 AM

2 Introduc ti on academic superstar” faculty member whose research productivity quickly fizzled out? These difficult questions indicate a need for better risk management by colleges and universities. For instance, should colleges and universities do a better job of screening faculty members before they are hired, looking beyond what formal background checks may or may not reveal? Is it possible to dig deeper into a faculty candidate’s past than is normally the case with a routine background check? Would Bishop’s banishment from campus—perhaps while still paying her for the academic year—have prevented her murderous rampage?

The vast majority of problem faculty members demonstrate considerably less egregious behaviors than those of an Amy Bishop or a serial sexual harasser. Yet the public pronouncements and embarrassing behaviors of faculty members might still discredit the institutions where they work. What risks, for example, did a major medical school take when it unknowingly hired an academically prominent but ill-tempered dean who led a secret double life as a partier who abused drugs and participated in orgies?2 Could this institution, which had garnered ample warnings about his behavior, have done more to avoid the embarrassing publicity that this dean inflicted upon it?3 How tempting is it for an institution to look the other way when a rogue faculty member is also a gigantic rainmaker who brings in millions of dollars in research grants? Challenging human resource management issues such as these continue to plague U.S. colleges and universities. How do we avoid hiring faculty who turn out to be nonproductive, counterproductive, and even toxic? Bad faculty hires may exhibit a disappointing level of research productivity, incompetent teaching, disruptive behaviors, a litigious bent, or possibly actions that pose physical threats to themselves and others.

Hiring and retaining the most productive faculty members also requires that institutional decision makers and counsel be mindful of the academic landscape of U.S. higher education. What are the latest trends in tenure, free speech, and collegiality? What about the emerging multitiered system of tenured tenure track faculty, non–tenure track faculty with contracts, and adjunct faculty? The third group may teach the same classes and work only feet away from their more privileged colleagues, but they might as well be on another planet when it comes to how poorly they are often treated.

Risk management plays an important role in finding answers (or partial answers) to these questions. This book seeks to provide advice on

3050-2728-1pass-0Intro-r03.indd 2

5/16/2018 6:45:34 AM

I n tr o d u ctio n 3 how to prevent, minimize, or avoid issues pertaining to the hiring and management of faculty members. We believe that a better understanding of legal and human resource management practices, framed within the unique culture of academia, may help decrease institutional risk exposure. The audience for whom a discussion of the risks associated with faculty employment may be most useful includes attorneys and counsel who have substantial legal knowledge, but little knowledge of the culture of academia; faculty who have accepted administrative positions and who understand the culture of academia, but have little knowledge of or training in legal requisites; and faculty members who face or who may face hiring, retention, or promotion issues.

Faculty and administrators usually have limited training in human resource management issues (hiring, developing, evaluating, and rewarding staff) or legal issues (equal employment opportunity law, contractual matters, or basic constitutional protections). A newly appointed provost with a professional background in nuclear engineering cannot be expected to understand the intricacies, or even the basics, of employment law. Since colleges and universities provide little or no legal or human resource management training for newly appointed administrators, these administrators are often forced to learn on the job, perhaps with the help of university legal counsel and academic consultants.

This lack of training can become a quite risky and expensive problem for the institution, especially if it leads to an inadvertent violation of civil rights, contract, tort, or constitutional law, resulting in a lawsuit. For the faculty member, litigation costs can be highly variable and potentially financially ruinous. The psychological trauma associated with a civil suit can be devastating as these cases drag on, moving in and out of court over a period of several years. Immersing one’s life and financial resources into a highly complex, drawn out, and stressful process where the odds of winning are low takes its toll both personally and professionally. For large institutions that employ full-time in-house legal counsel, the expense of litigation is largely fixed and is viewed simply as another cost of doing business, but the public relations fallout is something that most schools prefer to avoid.

The landscape of U.S. higher education is in a state of flux with many uncertainties on the horizon. Before we address the issue of risk management, we believe it would be helpful to describe the categories of

3050-2728-1pass-0Intro-r03.indd 3

5/16/2018 6:45:34 AM

4 â&#x20AC;&#x192;â&#x20AC;&#x192;â&#x20AC;&#x192;Introduc ti on institutions along with some of the major issues and trends affecting U.S. higher education. Readers who work outside academia or who are new to it may find the issues summarized here to be of special interest. We will then delve into some of the salient risks that arise on U.S. campuses.

The Academic Landscape U.S. colleges and universities are regarded collectively as the best in the world, and many of these schools are steeped in rich traditions of academic excellence. These institutions of higher learning disseminate vast amounts of knowledge, prepare students to become responsible and productive citizens, sponsor major cultural and athletic events, and create knowledge through basic and applied research.

Academia can also be a wonderful place to work. Faculty members often find themselves surrounded by intellectually stimulating colleagues and an internationally diverse student body. An especially attractive feature of academic life is being able to work in an environment that encourages creativity. Professors have traditionally had the academic freedom to pursue their professional interests and to teach their classes the way they want. They also have a great deal of discretion with regard to the research agendas they pursue. Furthermore, many colleges and universities have beautiful architecture and landscapes that enhance the quality of life for those fortunate enough to work and study there.4 But, as is the case with every occupation, there are drawbacks. Not everyone is comfortable working in an unstructured environment with little day-to-day supervision and accountability. College and university faculty do not punch time clocks and, with few restrictions, they come and go as they please. Although tenured faculty members have a strong degree of job security, faculty members on the tenure track must eventually face an up-or-out tenure evaluation process that can be contentious and even brutal. Meanwhile faculty salaries, with the possible exception of those in the professional schools of medicine, engineering, business, and law, fall well below the salaries of the corporate world. Roughly measured, the level of education and effort required to secure a university faculty position vastly exceeds the pay and benefits of that position.

3050-2728-1pass-0Intro-r03.indd 4

5/16/2018 6:45:34 AM

I n tr o d u ctio n 5

The Kaleidoscope of U.S. Colleges and Universities Most large universities fall into the category of research school. Faculty members at these schools are expected to present the results of their research and creative work in books, academic journals, government scientific reports, exhibits, and concerts. Elite private universities, including the eight Ivy League schools plus the Massachusetts Institute of Technology (MIT), the University of Chicago, Stanford University, and the California Institute of Technology (Cal Tech), along with top-tier public universities such as the University of California, Berkeley, the University of California, Los Angeles (UCLA), the University of North Carolina at Chapel Hill, and the University of Michigan, expect their faculty to produce research of the highest quality. While much of this creative work escapes the public eye, a great deal of it has led to advances in the biological, physical, and social sciences as well as in the arts and humanities. Not surprisingly, the work of Nobel laureates often begins on a college campus.

Faculty members at research institutions typically become experts in a narrow segment of an academic discipline. For example, an entomologist at a major research school may be known worldwide for her expertise on certain species of bees found in the Amazon River basin, a biologist may devote his professional life to the study of fungi, a psychologist may spend her career studying the personality development of children, or a law school professor might be known for her expertise on legal issues in the music industry. Although the top-tier research schools support and pay their faculty well, the standards for promotion and tenure at these schools are often extremely high. In fact, the well-worn phrase “publish or perish” is the norm at nearly all research institutions.

Research schools also expect excellence in graduate teaching. A university that is serious about the research productivity of its faculty, however, usually offers teaching loads of no more than two classes a semester. Most research universities also provide tenured faculty with periodic paid sabbaticals.5 Sabbaticals allow for uninterrupted time for research, and many faculty members spend their sabbaticals working at other universities, usually with colleagues who share similar research interests.

Teaching schools focus primarily on teaching undergraduate students.

Professors at these schools conduct little or no research. Instead, they

3050-2728-1pass-0Intro-r03.indd 5

5/16/2018 6:45:34 AM

6 Introduc ti on teach seven or eight classes during an academic year, and their teaching load may cover a wide variety of subjects. A history professor at a teaching school, for example, might teach courses in ancient history, Chinese history, the history of religion, and U.S. Civil War history, all within the span of one or two academic years. Heavy teaching loads and multiple class preparations make it difficult for faculty at these schools to conduct research or to write journal articles and books. Professors in community colleges, small private colleges, and specialized institutions such as technical schools devote their professional lives primarily to sharing existing knowledge and preparing students for the workplace—that is, they typically disseminate rather than create knowledge. Promotion, tenure, and job retention at these schools is based on a faculty member’s teaching quality, the demand for his courses, and his participation in school service activities such as student advising.

In reality, the distinction between teaching and research schools is not always clear. Colleges and universities have different teaching and research balances—not only among different institutions, but also within these same institutions. Even within a single department, teaching and research expectations may vary from one faculty member to another.

These differences should be carefully explained to each faculty member so that misunderstandings about expectations are avoided. The Carnegie Foundation for the Advancement of Teaching has classified institutions of higher learning. Each of these categories, in turn, contains subcategories that further differentiate these institutions.6 The classifications are as follows: associate’s colleges, baccalaureate colleges, master’s colleges and universities, doctorate-granting universities (this category has three subcategories, including schools that awarded at least twenty research doctoral degrees), special focus institutions, and tribal colleges.

The point that we want to make here is that colleges and universities come in all sizes and shapes, which makes it difficult to generalize about their missions and modes of operation. It is the job of an institution’s leaders to determine how faculty members are hired, trained, and evaluated as well as the specific tasks, duties, and responsibilities to which they are assigned. A school’s board of trustees, along with its president, chancellor, provost, and collegiate deans, is responsible for a school’s broad strategic initiatives. Other administrators, such as associate and assistant deans, institute directors, department heads, and nonacademic staff, usually work in the trenches, overseeing a school’s day-to-day operations.

3050-2728-1pass-0Intro-r03.indd 6

5/16/2018 6:45:34 AM

I n tr o d u ctio n 7

Issues Facing U.S. Colleges and Universities in the Early Twenty-First Century U.S. institutions of higher learning face a complex set of issues. Some of the more prominent ones are: (1) the soaring cost of higher education and the heavy debt loads faced by many college graduates; (2) concerns over eroding standards and less rigor in college classrooms; (3) the growing number of nonacademic administrators who hold positions of increasing power and who are thought to usurp the academic autonomy of the faculty; (4) the zealous pursuit of higher and higher institutional rankings by colleges and universities—especially those published annually by the U.S. News & World Report; (5) the controversies about academic freedom, free speech, and political correctness in institutions of higher learning; (6) the long-standing debate surrounding what constitutes the “proper” curriculum and goals of a college education; (7) the growing problem of criminal behaviors on college campuses; (8) the ongoing and sometimes contentious debate about academic tenure; and (9) the plight of adjunct faculty.

The High Cost of a College Education The cost of higher education in the United States has increased substantially faster than the rate of inflation, and the tuition and fees charged by many colleges and universities place a heavy financial burden on middle-class families. Added to the steep tuitions are the exorbitant costs of textbooks and computer technology. Although most textbook prices fall in the $100 to $200 range, CBS MoneyWatch has identified at least two textbooks that have a retail price tag of over $1,000 and at least a dozen others that cost $500 or more.7 The tab can increase by another $2,000 to $3,000 when a student is required to purchase a laptop computer and course software.

Three quarters of the students graduating from U.S. colleges and universities carry an average indebtedness of $35,051. When the debts are added up for all students, the total reaches an amazing $1.2 trillion.8 Staggering debt coupled with dim job prospects and meager projected incomes for many students—especially those with degrees in the humanities—means that student loan debt will plague many graduates for years or even decades after they leave campus. Onerous debt loads can create personal financial

3050-2728-1pass-0Intro-r03.indd 7

5/16/2018 6:45:34 AM

8 Introduc ti on dilemmas that have serious implications for major life decisions such as marriage, starting a family, purchasing a home, or seeking a postgraduate degree. Even declaring bankruptcy will not erase these debts. Some politicians, however, have advocated legislation that will tailor loan repayments to a debtor’s post-graduation income, with the government picking up the tab on any remaining outstanding debt after a defined repayment period such as twenty years.9

Eroding Academic Standards An unacceptably large percentage of undergraduate students who are entering college in the early years of the twenty-first century are taking more than four years to earn their baccalaureate degrees. Students who fail to complete their studies in a timely fashion not only increase the costs of their own education, but they make it more difficult for a college or university to accommodate new students. Delayed graduations are sometimes the result of a student switching majors or losing transfer credits, but this trend also suggests that colleges and universities are admitting students who are either academically unprepared or who are in no hurry to leave the comforts of a college campus and face the workaday world. Highly selective schools, almost without exception, have higher four-year graduation rates than their less selective counterparts. According to the College Board, for example, the highly selective and academically rigorous MIT has a four-year graduation rate of 84 percent, whereas the nearby University of Massachusetts Boston graduates only 13 percent of its students in four years.10 All professors—even those with only a modicum of classroom teaching experience—can share stories about students who rarely, if ever, attend class. And when it comes to identifying and correcting irresponsible student behaviors, faculty members and the families of students find themselves hamstrung by federal law. The Family Educational Rights and Privacy Act (FERPA) restricts parents and other interested parties from obtaining academic information about a college student. Thus, parents who are paying huge sums to send their son or daughter to college might have no idea that their investment in tuition, room, and board has been squandered. A cab driver related to one of us that she worked at three jobs to save the $26,000 needed to send her daughter to a state university in

3050-2728-1pass-0Intro-r03.indd 8

5/16/2018 6:45:34 AM

I n tr o d u ctio n 9 Florida. Months later she discovered to her horror that her daughter had spent her first year of college attending numerous parties but never setting foot in a classroom. Her freshman grade point average was a perfect 0.00.

Had the woman inquired, she would have been told by school officials that FERPA would not allow them to disclose her daughter’s academic progress (or, in this case, a complete lack of progress).

But even for the diligent student, the road to a college degree may contain obstacles, most notably the need to hold a part-time job to help cover the costs of tuition, food, rent, and other necessities. Employment outside the classroom not only cuts into study time, but a late-night job may lead to chronic sleep deprivation, a further impediment to academic progress.

Although college graduates stand to earn a much larger income over their lifetimes than do people holding only a high school diploma, some critics continue to pursue the argument that college today is not worth the money, claiming that there has been a gradual but significant erosion of the academic core. These critics—members of the media who write for publications such as The Chronicle of Higher Education or Inside Higher Ed as well as faculty and administrators who follow educational trends and events—bolster their position by pointing to how colleges pamper students by offering them posh dorms, access to gourmet dining, no early-morning classes, and state-of-the art weight rooms and rock climbing walls. Other critics sound off about the distracting emphasis of intercollegiate athletics and the surface appeal of extracurricular functions that contribute little to learning. Many critics cap their argument by citing evidence that college graduates lack the ability to think critically or to communicate effectively, both of which diminish their employability.11 Paradoxically, delayed graduations do not seem to be the result of heavier academic demands. To the contrary, there are accusations that courses and curricula have been watered down and are less rigorous than in decades past. College students today spend significantly less time studying than did students of their parents’ generation. Baby boomers attending college in the late 1960s and early 1970s regarded the grade of C as an acceptable evaluation of their work. But the Generation Y students, who now live with a grade distribution that has been truncated by years of relaxed standards, view the grade of C as unacceptable.

The conservative columnist and Stanford University professor Thomas Sowell has made reference to the “touchy-feely mush” and “trendy social

3050-2728-1pass-0Intro-r03.indd 9

5/16/2018 6:45:34 AM

1 0 Introduc ti on projects” that have become a part and parcel of K–12 education.12 Sowell could have easily blamed these same practices for the erosion of academic standards at U.S. colleges and universities. Furthermore, U.S. students are avoiding the rigorous STEM majors (science, technology, engineering, and math) that are now heavily populated by students from outside the United States. In fact, international students—not students from the United States—earn more than half of the advanced STEM degrees that are awarded on U.S. campuses.13 We tend to overlook the fact that students spend only about five percent of their life span attending college. So, how should this narrow window of opportunity best be used? Do students benefit more by studying philosophy and calculus or by picking up roadside litter as part of a college-sanctioned public service project? Perhaps academic pursuits should trump social programs since college graduates will still have four or five decades after graduation to do whatever community volunteer work they find rewarding.

Given concerns with eroding standards and excessive time spent earning a degree, colleges and universities should take a hard look at extracurricular student activities and decide whether these activities enhance or diminish the core academic mission of an institution of higher learning.

Academic standards may also be lowered by the professor’s desire to stay on good terms with students so that his teaching evaluations remain positive. Because promotion and tenure committees scrutinize student assessments of teaching effectiveness, an untenured professor may be especially fearful that a handful of disgruntled students may rate him harshly if he imposes uncomfortably high academic standards or awards too many low grades. Furthermore, an anonymous student can damage a professor’s reputation through postings on Internet sites such as RateMyProfessors. com. These sites were originally designed to enable students to exchange useful information about professors and their courses. But these same sites can be used as a tool for cyberbullying.14

Administrative Bloat The public is less aware of a major problem on many U.S. campuses—the rapid growth of campus bureaucracies that divert money from educational endeavors and distract faculty from their scholarly work. The organizational structures at most colleges and universities have become

3050-2728-1pass-0Intro-r03.indd 10

5/16/2018 6:45:34 AM

I n tr o d u ctio n 1 1 overrun with administrators, and at some schools administrators outnumber faculty members. The Johns Hopkins University political science professor Benjamin Ginsberg has pointed to the multilayered administrative structures where administrators have inserted themselves into faculty hiring, curriculum development, grading, and other academic matters—tasks that have traditionally been the purview of academics.15 One author has described such administrative creep as a “cancerous disease.”16 Growing academic bureaucracies and their increasing span of control are forcing institutions of higher learning to spend large sums of money on compliance efforts. Nicholas Zeppos, the chancellor of Vanderbilt University, broke down the compliance costs at his school and claimed that these costs annually amounted to $11,000 per student. Vanderbilt, a highly respected research university that also covets excellence in teaching, had total compliance costs of $146 million in 2014, of which $117 million were for research compliance. Also included in this cost estimate was the $14 million needed to cover the costs of higher education–specific costs, including the costs of accreditation.17 A New York Times article offers this view of administrative bloat: By contrast, a major factor driving increasing costs is the constant expansion of university administration. According to the Department of Education data, administrative positions at colleges and universities grew by 60 percent between 1993 and 2009, which Bloomberg reported was 10 times the rate of growth of tenured faculty positions.

Even more strikingly, an analysis by a professor at California Polytechnic University, Pomona, found that, while the total number of full-time faculty members in the C.S.U. system grew from 11,614 to 12,019 between 1975 and 2008, the total number of administrators grew from 3,800 to 12,183—a 221 percent increase.

The rapid increase in college enrollment can be defended by intellectually respectable arguments. Even the explosion in administration personnel is, at least in theory, defensible. On the other hand, there are no valid arguments to support the recent trend toward seven-figure salaries for high-ranking university administrators, unless one considers evidence-free assertions about “the market” to be intellectually rigorous.18

Most families are willing to write a tuition check for several thousand dollars, believing that their student will receive a quality education

3050-2728-1pass-0Intro-r03.indd 11

5/16/2018 6:45:34 AM

1 2 Introduc ti on from knowledgeable professors. These same families, however, might be outraged to learn that a huge chunk of their hard-earned dollars has been diverted from their student’s learning to cover the salaries of an ever-expanding cadre of administrators, most of whom never teach a class.

The Rankings Game Brand name and reputation are clearly important to U.S. institutions of higher learning. In the late summer months, academic administrators can be found anxiously awaiting the release of the latest U.S. News & World Report rankings of colleges and universities. These rankings have become a driving force in U.S. higher education as administrators develop strategic plans and focus on metrics that they hope will enhance their school’s chances of moving up the rankings ladder. And it seems that advancing one or two places in the rankings might trigger a major celebration, even though such progress may be nothing more than random measurement error.

The U.S. News uses over a dozen criteria to rank colleges and universities, including admission acceptance rates, alumni giving, graduation rates, the number of small classes offered, expenditures made per student, and peer assessments from other schools. It is certainly not lost on college administrators that these metrics are susceptible to manipulation.

Applicant acceptance rates can be lowered—making the school appear to be more selective than it really is—by encouraging unqualified applicants to apply, knowing full well that these students have no chance of being admitted. The U.S. News rankings also favor class sizes of fewer than twenty students. So, another way of manipulating the rankings is to create one huge class section of, say, 200 freshman English students and then to schedule several small class sections of the same course. To encourage low enrollments, a university might offer the small sections of the freshman English course at inconvenient times (for instance, at 8:00 a.m. or 6:00 p.m.). As far as we can tell, however, the U.S News ranking experts seem to care little about the quality of the instructors who actually teach these class sections, though students enrolled in a class of 100 taught by an award-winning instructor stand to learn much more than students in a class of fifteen taught by an inept one.

The U.S. News is not shy about making side-by-side comparisons of schools that are vastly different. An article in The New Yorker by

3050-2728-1pass-0Intro-r03.indd 12

5/16/2018 6:45:34 AM

I n tr o d u ctio n 1 3 Malcolm Gladwell noted that placing two dissimilar schools—Penn State University and Yeshiva University—under the same ranking system makes little sense because these schools have different missions, locations, and modes of operation.19 One is large, one is relatively small. One is rural, the other urban. One is public, the other private. One places an extremely heavy emphasis on NCAA Division 1 athletics while the other does not.

This example illustrates an apples-and-oranges comparison that is at best meaningless and at worst grossly misleading.

When one looks at the math behind these rankings, the ranking process appears to be even more nonsensical. One public institution that was recently ranked among the U.S. News top twenty had no undergraduate colleges with a national top twenty ranking. Equally perplexing was that this institution’s administration was accused, when completing their peer assessments, of ranking competing schools as “below average” as a way of elevating their own ranking.

Rankings have long precipitated heated debate, regardless of what (or who) is being ranked. Interestingly, significant increases or declines in institutional rankings are not common, even when highly publicized academic, athletic, or sexual harassment scandals taint the reputation of a campus. Not long ago, one of us was asked to complete a ranking questionnaire that solicited our opinion about schools with which we had little or no familiarity. We were able to assess our own institution (with no biases, of course) and we were able to evaluate—with some semblance of accuracy—a school where one of us recently worked. Beyond that, however, we knew next to nothing about the schools we were being asked to rate.

A question that academic administrators, especially those with social science backgrounds, should ask themselves is this: if a PhD candidate developed a dissertation proposal that ranked business corporations using a methodology that was similar to the U.S. News methodology, would any self-respecting social scientist approve this project? We suspect that the answer to this question would be a resounding “no.” A corporation is evaluated on end-results measures such as the price of its stock, consumer ratings of its products and services, or its profits. But the U.S. News ranks colleges and universities primarily on process measures such as class sizes and expenditures per student rather than on the achievements of their students or the prominence of their faculty. A good measure of student

3050-2728-1pass-0Intro-r03.indd 13

5/16/2018 6:45:34 AM

1 4 Introduc ti on achievement is the percentage of an institution’s graduates that later earn a postgraduate degree, and a good measure of the quality of a school’s faculty is the number of professors holding fellowships in prestigious academic societies. It can be argued that even these successes are more of a reflection on the caliber of those who are hired or admitted to an institution rather than on what the institution contributed to their achievements.

Nevertheless, process measures that are poorly linked to end results tell us little about the true quality of an institution. But, at least in the eyes of the U.S. News analysts, processes speak louder than end results.

Academic Freedom and Academic License Academic freedom is supposed to allow professors to pursue the truth, no matter where that pursuit might lead. Yet, contrary to popular belief, academic freedom is not necessarily synonymous with free speech, and faculty members can no longer count on these ideals to protect their employment at a college or university. At one end of the spectrum, for example, a professor who asserts that Descartes’s philosophy is “full of logical flaws” may astound other academics who believe that Descartes’s thinking forms the foundation of modern philosophy. Despite the heated rhetoric that might emerge from this debate, academic freedom supports such honorable, but conflicting, views. At the other end of the spectrum is the problem of academic license, where a faculty member engages in unprofessional or irresponsible behavior and expects naively that the principle of academic freedom will serve as a free pass. A professor, for example, cannot use the idea of academic freedom to get away with canceling most of his classes and giving students high grades for doing little or no work. Nor can a professor be allowed to take unfair advantage of a student by requiring him to provide free babysitting or domestic help as part of a research assistantship. Somewhere in the middle of the academic freedom-academic license spectrum is where the protections afforded to certain types of speech become less certain. A state university professor who is tempted to level accusations of incompetence against the central administration on her campus is advised to proceed with extreme caution. Based on recent litigation (to be discussed in chapter 4), such criticisms appear to fall outside the scope of academic freedom and protected speech. Furthermore, faculty members at private colleges cannot rely at

3050-2728-1pass-0Intro-r03.indd 14

5/16/2018 6:45:35 AM

I n tr o d u ctio n 1 5 all on free speech rights to protect their jobs. A professor at a denominational college who speaks out in support of gay marriage may find himself in a precarious position if the college’s president is offended by his speech.

A professor at a state university making the same speech, however, could invoke her constitutional free speech rights to protect her job.20 Anyone who has spent time on a college campus in recent years is aware of the growing pressures for faculty and students to comport themselves in ways that will ensure conformity with political correctness. One infamous case is the Water Buffalo scandal at the University of Pennsylvania.21 A dormitory resident at Penn asked a group of African American sorority sisters who were gathered outside his window to be quiet so that he could concentrate on writing a term paper. When they ignored his pleas, he yelled, “Shut up, you water buffalo!” This utterance was construed to be racist and it set off a firestorm on the Penn campus.

More recently, leaders at the flagship campus of the University of Missouri in Columbia were forced to resign amid accusations that they were not doing enough to stifle racism. In the wake of the Missouri debacle, students at Ithaca College, Yale University, Claremont McKenna College, and other small schools demonstrated over racial issues that appear to be far from egregious, citing stray remarks, inappropriate Halloween costumes, and “micro-aggressions” as evidence of “systematic oppression.”22 While overt racism, sexist behaviors, or ethnic discrimination are to be condemned, some campus protesters are willing to make specious accusations of racism and mistreatment based on evidence that falls well short of the “beyond a reasonable doubt” or “preponderance of evidence” standards of criminal and civil cases, seemingly ignoring the tenets of due process in their haste to censure and punish those whose views they dislike.23 The University of New Hampshire’s Bias-Free Language Guide—an eight-page guide written to inform faculty, staff, and students as to how to make speech more inclusive—provided a glossary of acceptable and unacceptable terminology, as well as a Gender Pronoun Guide and examples of microaggressions and macroaggressions.24 As reflected in the guide, some members of the university community were apparently offended by the university’s use of the terms “mothering” and “fathering,” suggesting that “parenting” was the politically correct word of choice.25 The supposedly neutral term of “parenting” overlooks the fact that mothers and fathers have for centuries served different biological, social, and cultural roles.

3050-2728-1pass-0Intro-r03.indd 15

5/16/2018 6:45:35 AM

1 6 Introduc ti on Although the university has since distanced itself from the guide, as demonstrated by its removal from the university’s website, one has to wonder whether the notion of academic freedom might be in jeopardy if only the most outspoken purveyors of political correctness try to impose their opinions to the exclusion of the free speech rights of others.

The Curriculum Debate A related concern is the age-old question of whether an ideal college curriculum exists. Academic debates on curriculum matters often address a number of interconnected questions. Does a college education provide a springboard for lifelong learning? Is there an ideal mix between courses in the arts and sciences and courses that are designed to enhance a student’s employability after graduation? What is the appropriate level of rigor for an undergraduate education? Should colleges be more assertive in forcing students to take challenging courses in mathematics, the hard sciences (physics or chemistry), and foreign languages, even when these requirements fly in the face of U. S. News criteria, lengthening a student’s time to graduation and lowering an institution’s graduation rate? If so, will students avoid enrolling at colleges that impose heavy academic demands in favor of those that have more modest standards? Discussions of curriculum matters can become extremely heated, especially when those debates threaten the hidden agendas and turf wars of deans, department heads, and faculty who are trying to protect their courses, curricula, and student enrollments.

A critical issue facing institutions in the 2010s is trying to predict how online courses and distance learning will affect higher education over the following decade. Colleges and universities in the United States and beyond are diving headfirst into Internet technology to deliver a variety of courses and educational programs. The edX venture, initially a partnership between MIT and Harvard, has a growing number of prestigious universities on its list of affiliates. The venture provides high-quality courses—known as MOOCs (massive open online courses)—at little or no cost to the learner. Upon completing a course with a passing grade, an edX learner receives a certificate. Businesses may reward employees who complete a series of courses and earn a “micro masters” or equivalent online credential in subjects such as project management, business analytics, supply chain fundamentals, or computer programming.

3050-2728-1pass-0Intro-r03.indd 16

5/16/2018 6:45:35 AM

I n tr o d u ctio n 1 7 Another MOOC provider that awards certificates of completion is Coursera. Coursera has teamed with the University of Illinois to provide a quality iMBA degree at a cost that is well below that of most on-campus MBA programs. The proliferation of online courses and degree programs raises the question as to whether traditional on-campus programs will experience declining enrollments as the number of online courses continues to grow and diversify. Earning a degree from a prestigious university program—once a pipe dream for many—may now be within the reach of thousands of talented and motivated students.26

Dangers on Campus From the perspective of a criminologist, college campuses are crime facilitative organizations that provide an ideal climate for robberies, burglaries, and assaults. Institutions of higher learning are easily accessible to the public, and perpetrators from both inside and outside the campus can use classrooms, campus streets, green spaces, walking and hiking trails, dorm rooms, and parking garages as their hunting grounds. Criminals may also target victims at nearby off-campus sites such as bars, apartment complexes, and public parks. In addition, late-night parties that get out of hand, alcohol and drug abuse that lowers inhibitions, broken romances that turn suitors into stalkers, and the abundance of immature and entitled minds easily tempted to engage in acts of sexual harassment are all issues that campus administrators and law enforcement agents must face.

Title IX, which was passed in 1972, prohibits sex discrimination at educational institutions that receive public funds. Although Title IX is best known for addressing gender equity in sports programs, the law has also been applied to cases of sexual violence and harassment. The U.S. Department of Education’s Office of Civil Rights (OCR) has issued multiple letters of guidance to colleges and universities describing their responsibilities under the act. Although it is clear that colleges and universities (and their surrounding communities) have social and legal obligations to create a safe campus environment, it is not clear how these institutions should proceed in dealing with the dark side of campus life. Laws regulating college campuses do not exist in isolation. It is a complicated task to develop a campus-wide safety plan that is fair, impartial, and in compliance with a web of federal laws, advisory letters, and reporting policies.

3050-2728-1pass-0Intro-r03.indd 17

5/16/2018 6:45:35 AM

1 8 Introduc ti on The efforts to control crime and violence, however, require a group effort from administrators, student organizations, health care staff, and even campus clergy. In addition, the support and coordination of local and state law enforcement is critical.

The Tenure Minefield Perhaps the most hotly debated topic in U.S. higher education is faculty tenure. This debate has generated dozens of books, hundreds of journal articles, and countless editorials. Within these publications and, we suspect, within faculty lounges, professional meetings, and university administrative offices throughout the country, eloquent arguments have been made supporting the retention of tenure and equally eloquent arguments have been made encouraging its abolishment. An examination of both these arguments and current practices in higher education administration suggests that tenure may be on the gradual decline, but it is not going to disappear overnight.

The tenure system is supposed to protect professors from being terminated because of controversies surrounding their research findings, classroom teaching, political views, or public pronouncements. Admittedly, concern has been expressed from some corners that professors are using the classroom as a bully pulpit to indoctrinate students to their political views.27 But the long-standing rationale behind tenure is that professors should be allowed to pursue the truth through their research and writings no matter where that truth might lead, and they must be able to do so without fear of reprisal.

Suppose that a faculty member at a state university conducts research showing that organized labor has improved the wages, hours, and working conditions of the state’s manufacturing workers. These research findings, however, might be at odds with the agenda of politicians who are trying to attract industry and promote economic development. If tenure works as advertised, the faculty member should be protected from those who want her dismissed—even when those calling for her head come from such lofty places as the governor’s mansion, the state senate, or the headquarters of a powerful lobbying group.

But even among tenured faculty members themselves, the usefulness of tenure is often questioned. Those supporting tenure point to it as the

3050-2728-1pass-0Intro-r03.indd 18

5/16/2018 6:45:35 AM

I n tr o d u ctio n 1 9 vehicle for promoting academic freedom and maintaining academic rigor.

Yet the terms “academic freedom” and “academic rigor” defy precise definition. Does academic freedom mean the same thing for faculty in civil engineering as it does for faculty in political science? How does one compare the rigor of mathematics with the rigor expected of faculty and students in English or art history? And are there really that many cases where the protections afforded by tenure even need to be invoked?

Those rejecting the tenure system feel that it restricts an institution’s ability to allocate resources and control costs as it sees fit. Opponents of tenure argue that the system can slowly, over time, create a faculty group that has a diminished interest in quality teaching or meaningful research. They claim that older professors who refuse to step aside make it difficult for institutions of higher learning to replenish their ranks with younger and, presumably, more engaged teachers and more productive scholars. The vocal and radical opponents of tenure, we believe, have unfairly painted tenured faculty members with a broad brush, characterizing them as overpaid, aging deadwood who add little value to the institution’s academic mission. In reality, the vast majority of tenured professors, especially those beyond middle age, do not “retire on the job,” and their performance ranges from very solid to monumental, even extending to winning the Nobel Prize.

Every year hundreds of tenure track faculty members fail to clear the tenure bar, usually because their research fails to pass muster or because their teaching is inadequate. In a few cases, a tenure track professor may be derailed because of problems with collegiality—they simply cannot get along with their fellow faculty, staff, or students. In chapter 5, we further explore the difficult concept of collegiality, especially as it pertains to faculty retention, promotion, and tenure. Once a decision is made to deny tenure, a faculty member is usually placed on a one-year terminal contract and is forced to start looking for a job elsewhere. Since the late 1960s, plaintiffs facing tenure denials or other adverse actions have used federal civil rights laws as their primary legal option. These laws, known commonly as equal employment opportunity or EEO laws, include Title VII of the 1964 Civil Rights Act (race, gender, religion, national origin, and skin color), the Age Discrimination in Employment Act (for plaintiffs age 40 and above), and the Americans With Disabilities Act (for plaintiffs with physical or mental disabilities).

3050-2728-1pass-0Intro-r03.indd 19

5/16/2018 6:45:35 AM

2 0 Introduc ti on Cases may arise where other laws also apply. These laws might include the Family and Medical Leave Act, the Pregnancy Discrimination Act, the National Labor Relations Act (regulating labor activities at private colleges and universities), the Occupational Safety and Health Act and state workers’ compensation laws (when faculty are injured on the job), state privacy statutes, and laws regulating employee health care and retirement benefits (such as the Employee Retirement Income Security Act). And, as colleges and universities expand their presence outside of the United States and open campuses abroad, laws such as the Immigration Reform and Control Act may come into play.

We have seen from the extensive litigation of past years that laws prohibiting race and gender discrimination have not, to any significant extent, helped faculty who have been dismissed from their professorships. Judges simply do not want to second-guess the decisions of university faculty and administrators, and jurors have a hard time understanding why anyone is entitled to a lifetime of job security. Although judicial deference to academic decision makers has clearly been the norm over the past three decades, exceptions occasionally do arise where the faculty member prevails with at least a partial settlement. And, most important, we believe that contractual issues—separate from EEO law—will play an increasingly prominent role in the dismissal of tenured faculty.

The Rise of Non–Tenure Track Positions As noted, the percentage of tenured faculty in U.S. institutions of higher learning appears to be declining—albeit slowly. One projected trend is that the number of tenured faculty will decline and eventually hit a plateau where only a few high profile professors are granted tenure. Dozens of articles lament the decline of tenure, and most of these articles suggest that the only remaining tenured faculty members may be those who are star researchers who are hired into tenured or tenure track faculty slots.28 Another current trend, however, is already crystal clear: tenured faculty members, when they quit or retire, are being replaced more and more by non–tenure track faculty who work under renewable contracts. Although non–tenure track faculty members have no job security beyond the life of their contracts, those who competently teach courses with robust

3050-2728-1pass-0Intro-r03.indd 20

5/16/2018 6:45:35 AM

I n tr o d u ctio n 2 1 enrollments may have their contracts renewed year after year, giving them a degree of job security that is on par with their tenured colleagues.

For colleges and universities, the major benefit of hiring non–tenure track faculty members is their low cost and flexibility. Non–tenure track faculty work for significantly lower pay and fewer benefits, and they usually receive no summer research stipends, graduate assistants, or expensive laboratory or computer equipment. A university may be able to hire two or three non–tenure track faculty members for the price of one tenured faculty member. And, since non–tenure track faculty members are not expected to engage in research, they usually carry teaching loads that are twice as heavy as their tenured counterparts. Furthermore, if enrollments decline or shift toward other disciplines, non–tenure track faculty members who no longer add significant value can be let go as soon as their contracts expire. For these reasons, non–tenure track faculty, with their high teaching loads, low salaries, and promise of flexible staffing, are a resource that many colleges and universities find too good to resist.29

The Adjunct Professor Paradox Historically, tenured and tenure track faculty have enjoyed a middle-class lifestyle with good pay, benefits, and job security. More recently, non– tenure track faculty, including clinical professors and professors of practice, have emerged. Life, in terms of pay, benefits, and security, has also been good for this group, as long as their teaching is of sufficient quality and the demand for their courses remains strong.

The fastest growing groups of professors, however, are adjunct faculty members—graduate assistants and part-timers. According to the U.S. House of Representatives eForum report, adjunct faculty now comprise over half of U.S. college and university faculty, up from 20 percent in 1970. The U.S. report generated 845 responses from adjuncts that were highly educated and willing to take on heavy teaching loads for modest pay—pay that the report described as piece rate because adjuncts are typically paid by the course or credit hour. It is also quite likely that these same faculty members receive zero pay for tasks such as advising students and service activities outside of class, and zero promises of retention beyond the current semester. The eForum respondents had a variety of

3050-2728-1pass-0Intro-r03.indd 21

5/16/2018 6:45:35 AM

2 2 Introduc ti on stories to tell about what it was like to be overworked, underpaid, and completely taken for granted.30 The dilemma faced by adjuncts is especially puzzling when one considers the difficulty of earning a postgraduate degree and the growing demand for higher education services in the U.S. economy. Scarce skills along with a high demand for those skills should bode well for an aspiring adjunct faculty member. Instead, this combination seems more likely to lead to stress and its associated health problems as well as a high level of frustration that might even encourage dysfunctional behaviors. The Colby College sociology professor Neil Gross wrote an op-ed piece for the New York Times that documented inappropriate and threatening remarks made by some adjunct faculty as well as a feeling of adjunct alienation on the part of frustrated part-time faculty members that have been made to feel like second-class citizens. According to Gross, “Social scientists have found that when aspiring intellectuals face highly restrictive employment opportunities, they often take refuge in extreme politics.”31 The day-to-day economic plight of adjunct instructors is often nothing short of horrible. Stories abound of adjuncts living in poverty, often homeless and forced to reside in shacks or even in cars that are often old, poorly maintained, and unreliable. There is at least one instance of a female adjunct faculty member who resorted to making money by soliciting customers for sex. Her biggest fear was that one of her customers would turn out to be one of her students. Other adjuncts have had to deal with substandard housing, defaulting on student loans, going without basic healthcare, and hiding their homelessness. To make ends meet, many adjuncts teach at several schools within driving distance of their homes. It is mind boggling to imagine the chronic fatigue that these faculty members must endure while working thirteen-hour days, day after day, to meet their teaching obligations.32 The treatment of adjunct faculty might be the single most critical issue currently facing U.S. higher education.

Accountability, Assessment, and Deprofessionalism Throughout the history and development of American higher education, institutions have been held accountable to and influenced by people who work away from campus. These stakeholders include governing boards,

3050-2728-1pass-0Intro-r03.indd 22

5/16/2018 6:45:35 AM

I n tr o d u ctio n 2 3 taxpayers, politicians, benefactors, and alumni. Ironically, while public institutions have had their budgets cut dramatically by state legislatures, these same bodies are asking for greater accountability and oversight of their state colleges and universities—an emerging doctrine of increasing control with decreasing financial support. On top of the growing amount of regulation by federal and state agencies, regional and professional accrediting agencies are imposing yet another layer of policies and assessment measures on institutions of higher learning.

All of these measures have created an unintended consequence—the deprofessionalization of the professoriate. Faculty members who once had the freedom to teach as they saw fit, assign grades, counsel students, pursue research agendas, and make other academic decisions based on their professional judgments are now challenged—seemingly at every turn—by an arsenal of laws, regulations, and policies. Some faculty are now being forced to adhere to structured syllabi and lesson plans with predetermined “outcome measures.” Others, as we have mentioned, are forbidden by federal law from talking to concerned parents about their son or daughter’s academic progress without first getting the student’s approval. In the past two decades, professors have been called on increasingly to explain their grading methods to academic grievance boards. And, any professor conducting research on human subjects knows that they must have their proposal scrutinized by institutional research offices—a process that requires multiple approvals from various administrators, most of whom know little about the research being conducted.

Accrediting bodies that serve U.S. colleges and universities have become increasingly intrusive.33 Although accreditation agencies may perform a highly useful quality control function, they are demanding more and more detailed reports and analyses. All of these documents require thousands of personnel hours to produce and at least some of it may be little more than busy work. Assessment measures imposed on faculty may go beyond the usual testing that occurs in a semester-long class, requiring a professor to complete additional assessments and related documentation.

Strangely, these same accrediting bodies are content to allow professors to grade often redundant exercises that add little or no value to student learning—something akin to eliminating baseball umpires and allowing players to call their own balls and strikes.

3050-2728-1pass-0Intro-r03.indd 23

5/16/2018 6:45:35 AM

2 4 Introduc ti on

Managing the Risks The conflicts and controversies described here require administrators and senior faculty members—perhaps without realizing it—to engage in the amorphous exercise of risk management. Providing some context in this chapter will help frame our subsequent discussion of faculty employment and management practices in higher education.

Private industry executives regularly factor risk management principles into the strategic plans of their corporations. In business management parlance, risks may be assumed (litigation is accepted as the cost of doing business), reduced (corporate decisions are reviewed by legal counsel), transferred (another firm is contracted to handle certain business responsibilities), and shared (risks flow to an insurance carrier).

Tasking legal counsel when reviewing the decisions of academics, accepting that litigation is an expected business cost, or passively allowing risks to flow to a well-funded insurance policy does little to promote a sense of academic community. Traditional industry thinking about risk management—as we might see in the financial or insurance sectors—does not necessarily provide a useful set of practices for colleges and universities, whose employment practices are deeply entrenched in the unique cultural and historical contexts described in this chapter. For these institutions, effectively managing risk requires defining what risk is and developing a system that identifies and confronts risks at multiple levels. Poorly managing the risks associated with faculty hiring, retention, or dismissal can lead to unnecessary legal expenditures; divisiveness among faculty, staff, and administrators; and unfavorable media attention.

Baker and Moss delineate several principles of effective government risk management.34 Although their analysis focuses on the broader aspects of management in the public sector, three of these principles are especially relevant to U.S. colleges and universities: 1. Shift from loss control to prevention. Loss control is ex post, whereas loss prevention is ex ante. Loss control encompasses actions such as severance pay, grievance hearings, legal proceedings, and the skillful use of public relations and media exposure tactics. Loss prevention is preemptive, and it involves actions such as careful hiring practices, fair performance evaluation processes, and equitable pay structures.

3050-2728-1pass-0Intro-r03.indd 24

5/16/2018 6:45:35 AM

I n tr o d u ctio n 2 5 2. Link responsibility with control. This approach places responsibility on those who are in a position to eliminate or mitigate a risk. Department heads, deans, provosts, and others who are guilty of a perfunctory or slipshod application of human resource management policies should be called upon to explain and justify their actions. Although most litigation in higher education is directed at the institution, plaintiffs often name individual administrators and faculty as defendants in these lawsuits.

3. Manage moral hazard. When responsibility is detached from control, moral hazard may arise. The insurance industry takes great care to minimize moral hazard by life and property insurance policy holders.

Automobile owners, for example, have a great deal of control over the damage history or theft of their car. If an automobile insurer replaces wrecked clunkers with brand new automobiles, the owners of the clunkers would be strongly tempted to have their nearly worthless automobiles wrecked or stolen. Similarly, tenure committees have a great deal of control over who is granted tenure and who is denied tenure, but the members of the committee are not likely to be punished professionally or be held responsible financially for a lack of due diligence in the event that a tenure or promotion denial results in expensive litigation. For this reason, colleges and universities, along, with administrators and senior faculty, must bear some of the liability for the damage they might cause to a faculty member’s career. Institutionally, moral hazard might be controlled through insurance deductibles and limited legal expense reimbursements. Individually, decision makers should be held at least partially accountable for their improper actions. This accountability usually does not lead to a liability judgment against a decision maker, but it is likely to force them into enduring long, drawn out, and contentious legal proceedings and it may irreparably damage their professional reputations and careers.

The crux of risk management is that human behavior can be quite unpredictable and, during the campus interview, it is important to take every precaution to vet thoroughly the credentials and background of each faculty candidate. When a hiring or tenure decision is made, the decision makers are actually placing a bet that the new hire will be a productive member of their faculty with tenure being awarded in due

3050-2728-1pass-0Intro-r03.indd 25

5/16/2018 6:45:35 AM

2 6 Introduc ti on time for a tenure track hire. But they cannot be absolutely certain of this outcome. The candidate who seemed so polite, polished, and scholarly when invited to campus for a job interview may, after being hired, turn out to be an academic superstar. Conversely, this seemingly well-qualified candidate may turn out to be a highly dysfunctional wolf in sheep’s clothing. Careful hiring, mentoring, and evaluating become key risk management tools.

Drawing on our combined legal and managerial experience in higher education, our goal is to help readers (administrators, faculty, and legal counsel) identify, address, and prevent risks associated with the hiring and retention of faculty members. Although the list is not exhaustive, examples of major risks include: 1. Litigation costs associated with adverse employment decisions. Promotion and tenure denials may lead to highly unpleasant, embarrassing, and expensive litigation that forces the institution to pay hundreds of thousands of dollars in litigation expenses and damages.35 2. Human resource management costs associated with faculty turnover and substandard pay and benefits. Colleges and universities sometimes confuse labor costs and labor rates. By paying higher salaries and by providing superior benefits packages (labor rates), institutions of higher learning may actually reduce total labor costs by shrinking or eliminating recruitment, selection, and training costs. Even when expensive litigation is avoided, faculty and staff turnover that is often traceable to poor hiring, development, and evaluation practices can create hidden, but significant, costs.

3. Acrimonious relationships among faculty, staff, and administrators that waste time and detract from academic pursuits. These risks arise as the result of a lack of poor administrative decisions such as the distribution of merit pay or other resources. Whether the injustice is real or perceived, certain administrative decisions may be regarded as personal affronts that lead to morale problems. A toxic work environment is rarely a productive one.

4. The continued erosion of academic standards. An erosion of standards diminishes the value of a college degree and generates faculty morale problems and cynicism. The economist Walter Williams contends that the erosion of standards, both in K–12 and in U.S. higher education,

3050-2728-1pass-0Intro-r03.indd 26

5/16/2018 6:45:35 AM

I n tr o d u ctio n 2 7 has resulted in the awarding of “fraudulent” diplomas and degrees that, in reality, signify grossly inflated levels of individual achievement.

5. Adverse media publicity arising from faculty and administrative disputes, athletic scandals, and campus criminal activity. A notable example is the scandal involving allegations of academic fraud that centered on athletes and the African and Afro-American studies program at the University of North Carolina at Chapel Hill.36 Not only did a professor fail to hold classes, he relegated the assignment and grading of papers to his administrative assistant. The assistant, in turn, awarded generously high grades for what amounted to be mediocre quality work. The intense media publicity surrounding the UNC scandal not only damaged the reputation of a highly regarded institution of higher learning, but it also tarnished the value of the degrees of thousands of students who were not part of the scandal and who did not cut corners while earning their diplomas.

This chapter provides a view of the landscape and salient features of U.S. higher education, and it will serve as a springboard for the chapters that will follow. The news media are interested in the goings on not only in the highly visible, academically prestigious schools, but also in local community colleges and lesser-known state and private institutions.

A lesser-known state university might employ thousands of faculty and staff who often have disparate views and interests, with a budget that might approach or exceed a billion dollars.

We treat issues such as decision making, faculty or union contracts, free speech boundaries, and collegiality in separate chapters, but these topics are frequently intertwined. A professor who is fired for making a controversial speech in a public forum may encounter questions not only about her free speech rights, but also about her contractual rights and about the opinions that fellow faculty members hold regarding her collegiality. Concerns may also arise about the fairness of the decision to fire her. Did she receive due process? Was her case examined carefully with attention paid to avoiding biases that might creep in and distort decisions about her future professional life? This book examines a multitude of issues concerning higher education. We will thus look at the critical interfaces of the phenomena we discuss and, at times, we will use segues to shift back and forth through the chapters.

3050-2728-1pass-0Intro-r03.indd 27

5/16/2018 6:45:35 AM

Workers without Borders Posted Work and Precarity in the EU

Ines Wagner

ILR Press An imprint of Cornell University Press Ithaca and London

552-73361_ch00_1P.indd 3

—-1 —0 —+1

04/09/18 5:58 PM

Copyright © 2018 by Cornell University All rights reserved. Except for brief quotations in a review, this book, or parts thereof, must not be reproduced in any form without permission in writing from the publisher. For information, address Cornell University Press, Sage House, 512 East State Street, Ithaca, New York 14850.

First published 2018 by Cornell University Press Printed in the United States of Amer­i­ca Library of Congress Cataloging-­in-­Publication Data Names: Wagner, Ines, 1984–­author.

Title: Workers without borders : posted work and precarity in the EU / Ines Wagner.

Description: Ithaca : ILR Press, an imprint of Cornell University Press, 2018. | Includes bibliographical references and index.

Identifiers: LCCN 2018013445 (print) | LCCN 2018013885 (ebook) | ISBN 9781501729164 (epub/mobi) | ISBN 9781501729171 (pdf) | ISBN 9781501729157 | ISBN 9781501729157 (cloth : alk. paper) Subjects: LCSH: Foreign workers—­European Union countries. | Foreign workers—­Germany. | Employee rights—­European Union countries. | Employee rights—­Germany. | Precarious employment—­ European Union countries. | Precarious employment—­Germany.

Classification: LCC HD8378.5.A2 (ebook) | LCC HD8378.5.A2 W34 2018 (print) | DDC 331.5/44094—­dc23 LC rec­ord available at https://­lccn​.­loc​.g­ ov​/­2018013445

-1— 0— +1—

Cornell University Press strives to use environmentally responsible suppliers and materials to the fullest extent pos­si­ble in the publishing of its books. Such materials include vegetable-­based, low-­VOC inks and acid-­free papers that are recycled, totally chlorine-­free, or partly composed of nonwood fibers. For further information, visit our website at cornellpress​.­cornell​.­edu.

552-73361_ch00_1P.indd 4

04/09/18 5:58 PM

Contents

Acknowl­edgments List of Abbreviations

ix xiii

Introduction 1 1. Methods and Data Collection

23

2. Posted Work and Transnational Workspaces in Germany

36

3. Management Strategies in Transnational Workspaces

55

4. Posted Worker Voice and Transnational Action

76

5. Borders in a Eu­ro­pean L ­ abor Market

96

6. Broadening the Scope

552-73361_ch00_1P.indd 7

115

—-1 —0 —+1

04/09/18 5:58 PM

v iii Contents

Appendix I: Article 3 of the Posting of Workers Directive

131

Appendix II: Overview of Interviews

133

Notes 135 References 137 Index 000

-1— 0— +1—

552-73361_ch00_1P.indd 8

04/09/18 5:58 PM

Introduction

In 2012, as I was visiting a housing site in Northern Germany, I met a ­woman named Maria. Maria had moved ­there from Romania a few months earlier to work in the meat industry. As we got talking, Maria shared with me a concern that is at the heart of this book. “We,” she said, referring to posted workers, “are now part of the Eu­ro­pean Union, but it does not feel like it; we have no repre­sen­ta­tion, no voice, h ­ ere.” She felt excluded ­because her expectations of a pan-­European ­labor market did not match her a­ ctual experiences. She and her colleagues received less pay from the employer than promised in their home countries, worked long or unreliable hours, faced management intimidation, experienced inadequate regulatory oversight, lacked health coverage or the time to attend to medical emergencies, and lived in substandard housing conditions. Maria and her colleagues conceived the Eu­ro­pean Union (EU) and Germany’s place within it to be well regulated. Being part of the Eu­ro­pean Single Market and constituents of the workforce supporting economic growth and wealth, they equally expected to be protected by the Eu­ro­pean ­labor market rights framework. In practice,

552-73361_ch01_1P.indd 1

—-1 —0 —+1

04/09/18 5:58 PM

2 Introduction

-1— 0— +1—

the workers’ experience within the pan-­European l­abor market was rather one characterized by fragility and contingency. As Maria pointed out, she and her colleagues are indeed situated within the Eu­ro­pean l­abor market, but many of the rights established within this context are rendered inaccessible to them.

In the Eu­ro­pean Single Market, ­labor can move individually via the ­free movement of ­labor, or firms can move workers around via the ­free movement of ser­vices. Posted Work is a central feature of employment practices via the freedom of ser­vices. Workers are “posted” by their employer to carry out work in one country, usually for a limited period of time, but they remain employed in another country. Workers, so it seems, are without borders. Yet while the real­ity of a borderless Eu­rope for workers is within reach for many, for a large group of ­people, this real­ity seems further away than ever. State borders may have “dis­appeared.” In both theory and practice, the border for the movement of ser­vices within the EU is no longer consistent with the edges of the physical territory of the member states. Yet borders still exist. They just exist elsewhere: in unequal pay, in lack of access to collective channels of repre­sen­ta­tion, or in the inability to claim rights. For example, the rebordering pro­cess of states intersects with the significant transformations of ­labor markets in Organisation for Economic Co-­operation and Development (OECD) countries since the 1970s. A key change in this pro­cess has been the increased flexibility and use of aty­pi­cal employment contracts such as posted work, substantially altering the orga­nizational activity of the main contracting firm and differentiating between rights of employees working at the main firms and t­ hose working at subcontracting firms. This book looks at how workers experience their rights when nation-­states have given up large parts of sovereignty over their l­ abor markets, while employed in a precarious employment relationship that stretches across state jurisdictions in the EU. In theory, posted workers have rights according to the EU Posting of Workers Directive (PWD). W ­ hether this is true in practice, however, is another question. Is a Polish citizen hired by a subcontractor in Cyprus for a job in Germany able to claim the minimum wage? What if the worker was sent from a Bulgarian com­pany to a German meat slaughter­house? Do EU mi­grants know which rights are available to them? Can they always access them? How do socioeconomic and cultural ­factors facilitate or complicate

552-73361_ch01_1P.indd 2

04/09/18 5:58 PM

Introduction 3

access? The answer to each of t­ hese questions is unclear b­ ecause of limited research on real-­world experiences of intra-­EU mi­grants.

This book reports on interviews with and participant observation of posted workers regarding how they experience the posting relationship, the mechanisms that enable access or denial to their rights, their ability to voice concerns over exploitative practices, and their interactions with institutions that should in theory enforce their rights. An actor-­centered research strategy helps identify the ways actors make sense of ­these regulatory pro­cesses at the workplace level. This approach can help generate an understanding of the dynamics of change in transnational workspaces in relation to the usage of rules; voice and exit; the possibilities of re­sis­tance; and, more generally, how such a pan-­European l­abor market is structured.

Using a bottom-up lens, this book examines how actors interact with institutions (Scharpf 1999) at the workplace level. This book complements the dominant research on EU integration, which is largely based on the belief that individuals act as rational beings and according to the intentions of the policy in question (Kauppi 2010). It considers the actors involved in the posting relationship to do more than produce automatic responses to the regulatory framework but to utilize this framework creatively. By focusing on actor strategies in response to the creation of a pan-­European ­labor market, the book highlights how “actors engage with, interpret, appropriate or ignore the dynamics of Eu­ro­pean integration” (Woll and Jacquot 2010, 116).

My win­dow to how posted workers experience intra-­EU mobility is Germany. This is a country that has traditionally been characterized as having both high ­labor standards and well-­functioning collective repre­sen­ta­tion rights for ­labor. It has been central to the policy discussion on the PWD. Empirically, a higher number of workers are posted to Germany than to other EU countries (Pacolet and De Wispelaere 2016). Germany also has a history of facilitating the inflow of temporary foreign worker schemes, creating a low-­wage work sector. The examination of posted worker experiences takes place in the two sectors where posting is most prevalent: the construction and meat slaughtering industries. The focus on Germany builds on the social science tradition of using changes in the “German model” to theorize broader changes. Germany has always been an impor­ tant case in the development of the comparative po­liti­cal economy and industrial relations lit­er­a­tures (Unger 2015; Bamber, Lansbury, and Wailes

552-73361_ch01_1P.indd 3

—-1 —0 —+1

04/09/18 5:58 PM

4 Introduction

2011; Hall and Soskice 2001), and building on this tradition enhances the empirical and theoretical relevance of this book. Germany’s current role and embeddedness in the Eu­ro­pean context position it with the potential to again inform the revision of existing theories.

Posted Work, the Nation-­State, and Eu­ro­pean Integration

-1— 0— +1—

Just over a c­ entury ago, Max Weber addressed the relationship between the nation-­state, economic regulation, and seasonal l­abor at his inaugural lecture in Freiburg. In his po­liti­cally troubling words, the “swarms of nomads”—­that is, Polish seasonal laborers brought in by middlemen in Russia—­appeared desirable to employers not only ­because employers could save on “workers’ ” dwellings, on taxes to support the poor, and on social obligations, but also ­because their precarious position as foreigners put them in the landowners’ hands (Weber 1994, 9). Yet, he argued, this seasonal ­labor was preventing unemployed German peasants from reentering employment. The “state’s” economic policies, Weber (1994, 9) demanded in a nationalist and discriminatory timbre, o­ ught to rise to the challenge of “defending” the German race and should shut the borders to migration.

More than a ­century ­later, the debate is still the same: employers prefer temporary mi­grant workers as a cheap, exploitable source of ­labor; agents channel mi­grants across borders; and, unfortunately, l­abor mi­grants are, still ­today, blamed for rising unemployment and for degrading the nation through “benefit tourism” or “poverty migration.” However, the structure in which ­these developments take place has changed considerably. The Weberian nation-­state, able to close the borders of the territory over which it had full authority and employ economic policy as it saw fit, was reconfigured by Eu­ro­pean integration. The EU has created a single market with reduced regulatory barriers for firms and workers. Part of this l­abor mobility takes the form of posted work, in which firms based in one EU member state “post” their employees temporarily to another EU member state to fulfill a ser­vice.

On the one hand, the PWD defines contractual terms and conditions for posting and establishes that, while posted workers’ social insurance and taxes are paid in the sending country, they should receive a minimum wage if it exists in a given industrial sector. On the other hand, posted workers

552-73361_ch01_1P.indd 4

04/09/18 5:58 PM

Introduction 5

themselves—as well as trade ­unions, works councilors, ­labor inspectorates, and the state—­are poorly equipped to ensure that ­these regulatory standards are upheld. This in-­between space and the de facto and de jure rights that posted workers hold within this space render them “borderline citizens.”1 To put it differently, while posted workers might be conceptualized as integral to and well regulated within the Eu­ro­pean and national ­labor markets at one moment, this could quickly evaporate in the face of certain management practices, a lack of enforcement, exit options being used by employers, and a lack of proper voice options for workers.

Thinking about posted workers as “borderline citizens” points to the contingent, conditional, and even vague place they inhabit within the nation state, as the most contentious issue around the posting regulations was, and still is, the question of which regulatory framework applies to posted workers at their place of work. Polish, Romanian, Portuguese, and Spanish (to name but a few) companies post workers to Germany whose wages and contracts are signed u ­ nder de facto Polish or Romanian laws, creating islands of foreign law (Hanau 1997) in the territory of the receiving country. Even though the PWD regulates posted workers’ inclusion, the structure to claim their rights is still not conclusive. As Cott (1998) noted, “Formal inclusion . . . ​ is never as decisive and determinative as formal exclusion” (1473). The aim of this book is to illuminate and analyze this in-­between space and the lives of posted workers within it.

This is not to ignore the considerable research on the ­free movement of ser­vices and posted work. Vari­ous studies have investigated the encounters between EU regulation and national l­ abor markets and their highly diverse industrial relations systems, public policies, and l­egal ­orders (Kall and Lillie 2017; Cremers 2011; Lillie 2010; Dølvik and Visser 2009; Barnard 2008; Cremers, Dølvik, and Bosch 2007; Eichhorst 2000). ­These studies discuss how institutions change through the policy pro­cess and how power imbalances are created and re-­created between the nation state and the EU policymaking body. Other studies have examined how EU member states try to re-­ regulate their ­labor markets in light of EU politics ­toward ­labor mobility (Alsos and Eldring 2008; Höpner and Schäfer 2007; Dølvik and Eldring 2006; Lefebvre 2006; Menz 2005; Eichhorst 2000). Further, ­there exist analyses of the tactics of capital and l­ abor in the national re-­regulatory pro­cesses (Greer, Ciupijus, and Lillie 2013; Refslund 2012; Afonso 2012; Lillie 2010; Krings 2009; Lillie and Greer 2007; Kahmann 2006). Moreover, a vast ­legal

552-73361_ch01_1P.indd 5

—-1 —0 —+1

04/09/18 5:58 PM

6 Introduction

-1— 0— +1—

lit­er­a­ture has discussed the impact of contentious Eu­ro­pean Court of Justice (ECJ) rulings on the scope of po­liti­cal economies to regulate their ­labor markets (Joerges and Rödl 2009; Kilpatrick 2009; Barnard 2008; Davies 2008; Ahlberg, Bruun, and Malmberg 2006).

Scholars have observed that the ­free movement of ser­vices in the form of worker posting has generated a transnational Eu­ro­pean market for low-­skill ­labor (Dølvik and Visser 2009). Meardi, Martin, and Lozano Riera (2012) noted that worker posting has facilitated “the creation of a hyper-­flexible buffer of mi­grant workers who, being disposable in case of downturn, can carry most of the uncertainty burden without causing po­liti­cal prob­lems” (5).

Even though the overall assessment of the f­ree movement of ser­vices and worker posting has indeed been negative, we know very l­ittle about how posted workers themselves experience the posting relationship (for notable exceptions see Lillie 2010; Berntsen 2016; Danaj and Sippola 2015).

By contrast, this book examines how posted workers and actors involved in the posting relationship actually utilize and experience the Eu­ro­pean posting framework. Empirically, the book shifts the attention from actors ­ atter: at the policymaking level to ­those who are the subjects of the m transnational posted workers themselves. This distinguishes the book from macro-­and national-­focused approaches in comparative po­liti­cal economy and industrial relations by zooming in on the workplace dynamics in a transnational setting. Theoretically, the book draws on adjacent disciplines— industrial relations, comparative po­liti­cal economy, Eu­ro­pean integration, and migration studies, as well as on discussions on the de-­territorialization of the nation state—to magnify their abilities to understand the interplay between changes in work, mobility, and citizenship in con­temporary Eu­ro­ pean ­labor markets.

The issue of posted work lends itself well to this kind of endeavor ­because, in the posting framework, the regulatory context is to a large extent delinked from insular national territorial regulations. At the same time, the transnational flexibilization of employment relations increases. A central theme of this book is thus the coevolution of changes in territorial po­liti­cal economies and modifications in employment relations through transnational workspaces from the bottom up. The central argument is that national policies are not simply downloaded at the Eu­ro­pean level (Kauppi 2013), but neither are Eu­ro­pean and/or national policies transferred as intended by the rule

552-73361_ch01_1P.indd 6

04/09/18 5:58 PM

Introduction 7

makers to the workplace level. Regulatory dynamics may take dif­fer­ent forms throughout multiple levels in the EU. Special attention has to be paid to the embedded and embodied nature of the regulatory framework and the forms of industrial relations structures at the supranational, national, and workplace levels. ­These levels and the actors at ­these levels do not exist in isolation to each other but are mutually constitutive of the regulatory space in the EU.

The Regulatory Environment The context of inclusion and exclusion for posted workers has shifted over time. In the 1960s it was primarily state managed, though in the 1990s and 2000s it became increasingly driven by the autonomous actions of workers and firms (Arnholtz 2013). The market in which posted workers migrate transformed itself from an international to a transnational ­labor market. This raises the question of how much posted workers differ from transnational mi­grants, if at all. They certainly do share certain characteristics. Similarities of posted workers to transnational mi­grants include being si­mul­ta­neously embedded in two socie­ties; they ­settle, to a greater or lesser degree, into the socie­ties of the countries where they work but remain embedded in their home countries at the same time (Glick Schiller, Basch, and Szanton Blanc 1995; Roberts, Frank, and Lozano-­Ascencio 1999). However, even though posted workers reside for limited periods in one or multiple countries, their migration pro­cess is arranged by their employer, and they often refrain from embedding themselves in countries other than their home country (Caro et al. 2015). T ­ hese ­factors significantly impact the in-­between position of posted workers as well as the regulatory spaces they inhabit.

Mi­grants who migrate and find a job individually must arrange, among other ­things, a place to live; learn the language and set up financial ser­vices; and interact with authorities (Datta 2009; Spencer et al. 2007). The posting relationship is an employer-­arranged migration context in the sense that employers mediate posted workers’ interactions with host socie­ties. This at once facilitates access to the host society’s administrative apparatus and limits interaction with the host society, triggering social segregation. Posted workers have to actively seek out contact with their host-­society surroundings.

552-73361_ch01_1P.indd 7

—-1 —0 —+1

04/09/18 5:58 PM

8 Introduction

-1— 0— +1—

In general, this means that their lives remain quite disconnected from the host society and region where they temporarily reside. ­Here, employer-­arranged migration reinforces connections with conationals, as posted workers from the same country often share the same work and accommodation environments (Caro et al. 2015). It also means that the rights of posted workers are increasingly related to their employer instead of the sending or receiving country (Guild 2001). Posted workers, even though they work in the host country’s territory, are disconnected to a large extent from the host country’s institutional system and ­labor relations. Language barriers with the host society foster more intensive contact with conationals. For posted workers, foreign language skills are often not a necessity; work teams are oftentimes arranged on the basis of nationality, so, within the workplace, it usually suffices if one member of a workgroup can speak the common workplace language.

The temporary nature of posted work means that mi­grants are not part of the target group for integration policies in the same way as more permanent mi­grant groups (Phillips 2010). The rights that posted workers can draw on are part of the regulatory framework of the PWD and its implementation into national laws. The start of the discussion concerning the PWD can be traced to the late 1980s. At that time, the Eu­ro­pean building ­unions pleaded for a social clause to guarantee compliance with working conditions and collective agreements in the host country in procurement rules for ­ abour Organ­ public works, in line with Convention 94 of the International L ization (ILO) and the Davis-­Bacon Act in the United States (Cremers 2009).

Despite broad support in the Eu­ro­pean Parliament, the Council of Ministers reduced the proposal to a voluntary act instead of an obligatory clause. In response, the Eu­ro­pean Commission introduced a proposal for a directive on the posting of workers in 1991 (Cremers, Dølvik, and Bosch 2007). At the center of the policy strug­gle has been the question regarding how much the protection of posted workers is on par with the ­free provision of ser­vices (Cremers, Dølvik, and Bosch 2007; Höpner and Schäfer 2012). The PWD was fi­ nally ­adopted in 1996—­five years ­after its first official proposal.

The PWD lays down a bare minimum of employment conditions for workers moving within three posting situations.2 The under­lying condition is an employment relationship between the undertaking making the posting and the posted worker during the period of posting, namely:

552-73361_ch01_1P.indd 8

04/09/18 5:58 PM

Introduction 9

1. ​Posting ­under a contract concluded between the undertaking making the posting and the party for whom the ser­vices are intended 2. ​Posting to an establishment or an undertaking owned by the group 3. Posting by a temporary employment undertaking to a user undertaking operating in a member state other than that of the undertaking making the posting Posting to an establishment or an undertaking owned by the group usually involves employees generally being regarded as “expatriates” rather than “posted workers” and is a more common situation for technical and managerial staff (Pedersini 2010). This category of employees often receives a number of benefits for their work abroad. Officially, a posted worker, as defined by the PWD, is a worker who, for a limited period, carries out their work in the territory of a member state other than the state in which they normally work (Directive 96/71/EC). The provision of ser­vices is thus only for a limited period of time.

In real­ity, the “posted worker” classification is often ambiguous. In the best scenario, posted workers are part of a genuine Eu­ro­pean division of ­labor between contractors and specialized subcontractors. In the worst scenario, posted workers are bogusly employed as temporary l­ abor. Even though posting does not lead to substandard working conditions per se, it has created an opening for new forms of recruitment that are part of a Eu­ro­pean market for low-­skilled l­abor in which both the bound­aries between mobile ­labor, posted work, and self-­employment are fluid and the line between “temporary” and “permanent” is frequently blurred (Faist 2008). For the sake of clarity, this book thus refers to workers as “posted workers” when they are sent by their employer to work in another country u ­ nder a ser­vice contract. The interviews tried to discern in which country the workers w ­ ere paying social security contributions as an indication of the posting relationship. In that sense, the term “posted work” is used as a conceptual tool to demarcate the transnational employment relationship and a de facto dependent employment relationship for workers from ­those moving ­under the ­free movement of ­people. Thus, when referring to posted workers, the par­tic­u­lar de facto employment situation is referred to instead of the strict ­legal definition, which is in itself a l­egal construction, as multiple rulings by the ECJ have shown.

552-73361_ch01_1P.indd 9

—-1 —0 —+1

04/09/18 5:58 PM

10 Introduction

The Role of the ECJ

-1— 0— +1—

Intra-­EU mobility has been contentious since its conception. ­After the six member states signed the Treaty of Rome, Lannes (1956) argued that “a scheme entailing freedom of movement such as that introduced by the Scandinavian countries seems to be out of the question as long as the marked unbalance in the economic structures and social conditions of Western Eu­ rope persists” (150). Nevertheless, at the time of conception, workers moved to and from l­abor markets and welfare states at broadly similar levels of income, and EU regulation ensured that they could secure their social rights in pensions and healthcare. The twenty-­eight member states that make up t­ oday’s EU3 have very heterogeneous social security provisions, levels of income, systems of public policy and enforcement, and industrial relations practices, not to mention the cultural and language differences among them.

This heterogeneity is perceived to result in slow policy negotiations, resulting in minor policy changes over time. By contrast, integration via the ECJ takes less time and can advance more radical decisions b­ ecause of the absence of po­liti­cal negotiation (Höpner and Schäfer 2012). First and foremost, the ECJ defined posting in such a way that it falls u ­ nder the ­free movement of ser­vices instead of the ­free movement of workers. The ECJ regarded posted workers as a “manpower ser­vice” and not as ­labor mi­grants. Therefore, posted workers are regulated ­under the ­free movement of ser­vices instead of the ­free movement of ­people. The ECJ justified this decision by reasoning that “such a worker [would return] a­ fter the completion of the ser­vice and [would] not at any time gain access to the ­labor market of the host state” (Case C-113/89 Rush Portuguesa §15), manifesting the isolated nature of posted workers in the host-­country context. Liberal voices welcomed the greater competitive pressure posted work produced that brought an im­ thers, such as trade provement in the allocation of capital (Kahanec 2013). O ­unions, feared “wage dumping” and the erosion of member states’ capacities to regulate their l­abor markets and social policies.

Moreover, in the cases Viking, Laval, and Rüffert, as well as Commission vs. Luxembourg, the ECJ fundamentally reconfigured the posting policy previously agreed on during multilateral negotiations (Scharpf 2008). T ­ hese decisions have received much attention b­ ecause they w ­ ere interpreted as landmark decisions on the strug­gle between economic freedoms and social regulation in the Eu­ro­pean common market. In the context of this book,

552-73361_ch01_1P.indd 10

04/09/18 5:58 PM

Introduction 11

two aspects are impor­tant. The first aspect is the reinterpretation of the PWD. In Laval, the ECJ referred to the list in Article 3 (1) as defining the “ceiling” on the “maximum” standards that member states are allowed to impose on posted employees4 (Kilpatrick 2009, 845–49). With this judicial reinterpretation, the ECJ effectively limited the host countries’ room for maneuvering to regulate the ­labor market. This constraint accelerates races to the bottom in the field of ­labor standards, a prob­lem that increased as the heterogeneity among member states increased (Scharpf 2006).

The second aspect concerns the restriction on private bodies (such as trade ­unions) in defending the rights of posted workers. In Viking (Case C-438/05) and Laval (Case C-341/05), the ECJ ruled that industrial action aimed at representing posted workers from a foreign undertaking could violate the com­pany’s freedom to provide ser­vices across borders. By ruling as such, the ECJ effectively curtailed trade ­unions’ right to strike (Kilpatrick 2009, 845–49). The PWD and the Ser­vices Directive are the outcomes of the EU demo­cratic pro­cess. The ECJ has, however, refined the details of posting regulations through judicial integration. This is problematic due to the inherent difficulty in reversing ­these decisions through the EU demo­cratic pro­cess (Höpner and Schäfer 2007). ­These decisions have impacted member states’ abilities to respond to EU legislation according to national institutional systems. First, the rulings regarded the par­tic­u­lar regulatory nature of national industrial relations systems as a constraint on freedom of ser­vices (Barnard 2009). This undermines the ability of national industrial relations systems to set collective standards according to their respective traditions, resulting in a “clash of capitalisms” (Höpner and Schäfer 2012; Kilpatrick 2009; Joerges and Rödl 2009).

Second, the ECJ intervened in par­tic­u­lar national industrial relations institutions, such as the right to strike in Sweden. Third, the cases established that the minimum set of rights as set out in the PWD is a maximum set of rights, meaning that member states are constrained from enforcing conditions for posted workers beyond the minimum conditions set down in law or in extended collective agreements. ­These decisions are impor­tant for the context of this book b­ ecause they constrain member states’ abilities to regulate ­labor markets and protect workers’ rights. As a result, posted workers work in regulatory loopholes where companies can draw on dif­fer­ent sets of business practices and worker

552-73361_ch01_1P.indd 11

—-1 —0 —+1

04/09/18 5:58 PM

12 Introduction

expectations. Lillie and Greer (2007) note that this posted worker isolation can be exploited by firms as a business strategy to shield themselves from national ­labor regulations and to “isolate mi­grants from the economic and social norms of the host society” (552). Lillie (2011), drawing on Ong (2006) and Palan (2003), compares this regulatory situation to a “space of exception” (694)—­a condition in which the regular law is exempted. However, ­these spaces still inhabit a par­tic­u­lar structure that the following chapters try to illuminate.

Approach and Outline of the Book

-1— 0— +1—

This book addresses the complexities of transnational posted work through three key topics. First, it examines how the de-­territorialization of national models and employment relations systems opens up exit options for management, enabling them to use the regulatory framework creatively and at a disadvantage for workers. Chapter 2 shows how transnational regulation and de-­territorialization impact employment relations in a German case of two industries in which posted workers are most pres­ent: the construction and the meat slaughtering industries. Chapter 3 combines institutional theory with strategic perspectives drawn from the sociology of organ­izations to examine how industrial relations actors “enact” EU rules at the microlevel of the workplace.

Second, it discusses how re-­territorialization, or re­sis­tance, is pos­si­ble within t­ hese spaces. Chapter 4 shifts the perspective to power and mobilization theory to demonstrate how workers foster community and media support to address contentious workplace issues within the transnational space.

Third, the book analyzes the contours of the new structure for employment relations that emerges within the pan-­European ­labor market and its implications for worker voice, regulatory enforcement, and management power.

Thus, chapter 5 adopts a more explic­itly spatial perspective and looks at how borders are constructed in both regulatory and workplace terms. Chapter 6 summarizes and concludes.

To analyze (1) the implications of the de-­territorialization of “bounded” national l­abor markets, (2) the usage of the regulatory framework and its implications for ­labor market regulations, (3) the possibilities for re­sis­tance within transnational workspaces, and (4) the shifting relationship between

552-73361_ch01_1P.indd 12

04/09/18 5:58 PM

Introduction 13

the changes in the territorial nation-­state and its institutional apparatus and among the changes in employment relations in the EU, this book draws on interviews with posted workers from Poland, Romania, Bulgaria, Portugal, and Spain aged between 17 and 53 years working in the construction and meat slaughtering industries in Germany. Interviews ­were conducted with posted workers, native workers, works councilors, inspectors, and policymakers as well as firms, trade u ­ nions, and nongovernmental organ­izations (NGOs). In total, I conducted 111 interviews with key in­for­mants (see Appendix II). Most of this work was carried out over three years between 2011 and 2013, with follow-up interviews occurring in 2014 and 2015 to track more recent developments within the industries.

The book conducts analyses at the firm, industry, country, and EU levels.

Such a perspective concentrates on how local actors implement Eu­ro­pean rules and opportunities (Pasquier and Weisbein 2004) to analyze the evolving balance of power induced by the EU around policy issues (Woll and Jacquot 2010). The approach binding the chapters of this book has been informed by an understanding of the need to link the institutional setting to the a­ ctual practices of actors involved to comprehend the full meaning of an institution and its changing nature (Kauppi 2010; Woll and Jacquot 2010; Streeck and Thelen 2005). Concentrating on formal rules when considering po­liti­cal conflict in the EU “would leave some of the most striking features of this transformation in the dark” (Woll and Jacquot 2010, 120). Moreover, the sectoral approach to analyzing power relationships distinguishes this book from macro-­and national-­focused approaches in comparative po­liti­cal economy lit­er­a­ture, providing a multifaceted perspective of po­liti­cal dynamics in transnational workspaces. The following ­will concisely explicate the main points of discussion relevant for the chapters of this book.

Nation-­States and Flexible Management Practices of Transnational Workspaces Posted work is an example of how Eu­ro­pean integration restructures relationships between states and actors in ways that challenge the traditional understanding of “nationally or­ga­nized” economic systems in mainstream comparative po­liti­cal economy. Historically, the formation of borders has had a clear purpose. Territorial demarcation was a necessary prerequisite for

552-73361_ch01_1P.indd 13

—-1 —0 —+1

04/09/18 5:58 PM

14 Introduction

-1— 0— +1—

the formation of the modern nation-­state system. Borders delineated a given territory over which authority was consolidated and order enforced (Weber 1947). Next to a central bureaucracy, citizenship, and the abilities to collect taxes and enforce order, borders ­were the prerequisite for any state-­like formation (Zielonka 2000). The nation-­state had a mono­poly of control over the movement across the borders of its territory and decisions on who could participate in its l­ abor market (Torpey 1998). Within t­ hese borders, each nation-­ state created a distinct set of institutions that have in turn structured actor be­hav­ior. Institutionalist scholars have connected t­ hese to national socioeconomic outcomes.

For example, the comparative po­liti­cal economy and industrial relations lit­er­a­tures are inclined to embed actors in nationally bounded sets of relations (Bamber, Lansbury, and Wailes 2011; Hall and Soskice 2001). In ­these models, institutional complementarities have a tendency to reinforce each other, making a par­tic­u­lar set of institutions within dif­fer­ent national models resistant to change (Hall and Soskice 2001). With regard to the regulation of posted work, Menz (2005) contends that national va­ri­e­ties of capitalism filter the impact of Eu­ro­pean regulations. This mediation results in dif­fer­ent regulatory outcomes to EU-­wide policies for regulating the provision of ser­ vices. The institutional setup of nationally or­ga­nized po­liti­cal economies thus influences how EU-­level regulation is implemented. This research does not take issue with the view on the potential of national systems to explain the diversity of response to changes in EU regulation; the evidence ­here underlines the fragility of capital’s support for national va­ri­e­ties of capitalism. The aim ­here is to show that we cannot fully understand EU or national regulation ­until we connect the way policy is designed and renegotiated with the way actors use it in everyday life and, in this case, in a transnational context.

Like comparative po­liti­cal economy, industrial relations, as a discipline, focuses on national systems as sets of relationships between workplace-­based actors within enclosed territorial spaces corresponding to national borders (Lillie and Greer 2007). Research on the dualization of the l­abor market shows that the parallel cir­cuit of subcontracted workers results in the undercut of wages, reduction of skill formation, and erosion of u ­ nion repre­sen­ta­ tion (Doellgast and Greer 2007). By its very nature, this industrial relations perspective remains closely tied to the nation-­state as a unit of analy­sis. It does not consider how a worker employed via a subcontracting relationship

552-73361_ch01_1P.indd 14

04/09/18 5:58 PM

Introduction 15

might collect pieces of protections from more than one nation-­state or how nation-­states might protect and provide for mobile citizens. Native workers are embedded within the po­liti­cal economy where the work takes place. By contrast, posted workers are legally separated both from the institutional structure of the location of the work and from native workers with whom they share a job site while they are si­mul­ta­neously alienated from the sending country (Wagner and Lillie 2014).

Moreover, globalization and Eu­ro­pean integration have started to dissociate, or de-­territorialize, the bonds that tied economics, politics, and culture to fixed spatial configurations. Certain rights have been established at the Eu­ro­pean level for posted workers, but the effects of t­ hese rights are ambiguous, since they are established in the absence of “bounded” mechanisms for social solidarity (Wagner and Lillie 2014). This relates to the way internal borders have been removed in the EU integration pro­cess. Eu­ro­pe­ anization can be said to “disembed” markets and detach social relations previously regulated via national social bargains (Höpner and Schäfer 2012) or to “embed neoliberalism,” in which market-­embedding institutions remain at the national level but are increasingly targeted by supranational liberalization attempts (Van Apeldoorn 2009). Eu­ro­pean integration opens up exit options for capital but isolates posted workers from collective channels of worker repre­sen­ta­tion. In light of the de-­territorializing capacity of posting, specific national re-­regulation often remains in­effec­tive at the workplace level. The way in which firms create and exploit transnational workspaces reduces the capabilities of state-­centered institutional systems to regulate within their own territories. To understand the evolution of any regulatory institution, it is impor­tant to examine the contexts as well as the actions of actors, as they are mutually constitutive (Jackson 2010). Chapter 3 discusses how power imbalances between capital and ­labor as created at the policy level are translated to the workplace level. The aim of the chapter is twofold: first, to study how firms creatively enact the posting framework and, second, to examine how ­these mechanisms initiate a pro­cess of institutional change through power dynamics at the microlevel, a pro­cess generally relevant for theories on institutional change. Findings show that the possibility for firms to diverge from rules is accelerated in a transnational setting due not only to the unequal power dynamics between firms and workers but also to the inability to publicly or collectively enforce rules. The examination of how actors engage with this transnational institution contributes to institutional

552-73361_ch01_1P.indd 15

—-1 —0 —+1

04/09/18 5:58 PM

16 Introduction

-1— 0— +1—

change theory by bridging the gap between institutional context and its appropriation by firms, posted workers, and u ­ nions.

DeCerteau (1984) has pointed out that theory is incomplete without connecting the experience of “everyday life” with the po­liti­cal framework it is embedded in. For example, in his analyses about the politics of the city, De­­­Certeau (1984) distinguishes between how the “voyeur” and the “walker” perceive the city. While the “voyeur” has an overview of the city from a height, it is one that is distanced from urban activities—­the overcrowding and the congestion, the dangers, and the noises; it is a decontextualized and, indeed, illusionary version of the city. DeCerteau (1984) stresses the importance of analyzing phenomena through the view of the “walker” and examining the interactions between t­ hose who made the structure and t­ hose who use it to create real­ity. Over the past few de­cades, the common approach in comparative po­liti­cal economy has been to use the point of view of the voyeur. For instance, discussions that focus on how institutions change have analyzed the change of the design of the institution as opposed to its implementation in everyday life (Hall and Soskice 2001). It is equally impor­tant, however, to go beyond the macro economy and assess institutional practices in a more detailed fashion at the individual level (Deeg and Jackson 2007; Jackson 2010). Institutions both constrain and enable action and are dynamic.

For example, actors may not conform to the institutional setting if their interests come out of alignment with the expectation and reward structures of the subsystem in which they take part (Greif and Laitin 2004). This becomes problematic when alternative legitimate frameworks pres­ent themselves, which actors can access to protect themselves from enforcement.

The posting regulation pres­ents such options to actors. The delinking of ­labor market regulation from insular national territories creates a space where the regulatory system can be avoided. Firms adhere to the norms on paper but not in practice. This is relevant ­because the appearance of conformity is often sufficient to attain legitimacy (Oliver 1991). The rule system can be upheld ­because of the inability of ­unions and ­labor inspection authorities to control t­ hese gaps. T ­ hese discrepancies can lead to institutional transformation, even when no sudden shocks or breaking points are prominent (Greif and Laitin 2004). Indeed, Djelic and Quack (2002) have suggested that central foreign actors may become missionaries of institutional change. Beyond simply playing according to their own rules of the game, foreign firms may help institutionalize their rules in new contexts.

552-73361_ch01_1P.indd 16

04/09/18 5:58 PM

Introduction 17

­ ere, national mediation of the Eu­ro­pean liberalization of ser­vices can H only partially take place. State-­centered border regimes remain foundational ele­ments of the system (Sassen 2005), but their ability to regulate within ­those borders is limited. Held (1992) usefully distinguishes between “de jure” and “de facto” sovereignty. De facto sovereignty is used in the negative to refer to a loss of authority or control, while de jure sovereignty refers to having supreme power over a given territory. De-­territorialization of ­labor regulation allows for a loss in control over the host-­country territory to emerge.

This is why it is necessary to examine how actors utilize transnational institutions in ­these specific situations. The real meaning and function of an institution ultimately emerges only in the course of how it has been interpreted and practically applied by actors (Streeck and Thelen 2005).

Delinking territorial contingency with access to po­liti­cal and social protections removes posted workers from the protection they would normally enjoy as regular l­abor mi­grants. This provides firms with leeway to creatively exploit regulatory gaps in their cross-­border activity. Undermining the nation-­state system also undermines the collective goods and stability states provide. Meardi, Martin, and Lozano Riera (2012) discuss the way new ­labor mi­grants are in a place of maximum risks and minimum “voice,” which is the topic of chapter 4.

Posted Worker Voice and Transnational Action Trade ­unions play an ambiguous role in the development of l­abor market segmentation (Emmenegger et al. 2012; Palier and Thelen 2010). While trade ­unions have preferred to resist employer pressure t­ oward dualization, they have increasingly concentrated on core members b­ ecause of a weakening of their position in the social market economy (Palier and Thelen 2010).

Therefore, they helped allow for orga­nizational mea­sures and reforms that protect insiders and negatively affect outsiders (Emmenegger et al. 2012).

Other researchers have connected increasing market inequalities to the weakening of trade ­union power. This loss in power impedes effective re­sis­ tance, resulting in dualization (Benassi and Dorigatti 2015; Korpi 2006).

In ­these accounts, employees’ representatives have been much less involved in the ­labor market liberalization pro­cesses (Streeck 2009). Posting is part of the segmentation pro­cess b­ ecause the relationships with l­abor intermediaries

552-73361_ch01_1P.indd 17

—-1 —0 —+1

04/09/18 5:58 PM

18 Introduction

-1— 0— +1—

reinforce national/ethnic hierarchies (Refslund and Wagner 2018). In some cases, trade u ­ nions have tried to resist employers’ segmentation strategies by means of organ­izing campaigns and collective bargaining targeted to outsiders.

Previous studies have discussed the ability of ­unions in vari­ous host countries to or­ga­nize mi­grant workers ­after the enlargement of the EU in 2004 (Bengtsson 2013; Krings 2009; Fitzgerald and Hardy 2010). Depending on institutional strength, sectoral characteristics, and the ability to innovate (Marino, Roosblad, and Penninx 2017; Baccaro, Hamann, and Turner 2003; Frege and Kelly 2003), u ­ nions in, for example, the United Kingdom (UK), Denmark, and Sweden have managed to include mi­grant workers by employing staff with relevant language skills or by cooperating with other organ­izations.

Traditionally, German trade u ­ nions are known for their po­liti­cal strength within the coordinated market economy embedded in an encompassing institutional framework. However, in certain sectors, t­ here has been a decline in coverage of and a growing in­equality within this ideal model. While ­ nionism and its institutional posimuch of the research on German trade u tion is based on the German metal industry, it is equally in­ter­est­ing and necessary to look at labor-­intensive sectors such as construction and meat slaughtering and pro­cessing to analyze the position and strategy of trade ­unions in the context of a pan-­European ­labor market. Moreover, in the context of organ­izing hypermobile workers, the perspectives of transnational EU posted workers, their own experiences with temporary work, and their everyday practices to cope with the exploitative and uncertain nature of employment have rarely been the subject of attention. This unduly narrows the scope of analy­sis when trying to comprehend the con­temporary contours of power and government in transnational workspaces. While isolation from host-­country trade u ­ nion repre­sen­ta­tion results in the loss of collective voice, workers may use other means to challenge malpractices within transnational workspaces. The vari­ous ways they appropriate their rights as EU citizens may differ from traditional channels of voice such as joining ­unions or voting in ­union and works council elections.

Given posted workers’ predominant exclusion from institutionalized voice channels in the German industrial relations system, chapter 4 explores the conditions for posted worker re­sis­tance. As a counterpoint to the lit­er­a­ ture on institutional stability and change, the chapter traces the pro­cess of

552-73361_ch01_1P.indd 18

04/09/18 5:58 PM

Introduction 19

forming an alliance between the trade ­union, a community organ­ization, and posted workers and examines the conditions ­under which it can evolve.

The chapter also carves out more explic­itly how mi­grant workers employ strategies to rework (Berntsen 2016) the employment situation to their advantage but do not challenge the structure of the employment relation per se, thereby contributing to the continuation of ­these employment practices.

The case study shows the willingness and need of German trade ­unions to build co­ali­tions in areas vital to u ­ nion interests, even though they have traditionally abstained from ­doing so ­because of their strong institutional entrenchment. Re­sis­tance ­here is conceptualized as re-­territorialization to analyze how workers “disembedded” from the regulatory framework (de-­ territorialization) can be inserted into an inclusionary framework with collective interest repre­sen­ta­tion (re-­territorialization). For the purposes of this book, de-­territorialization relates to the disconnect and reconnect of national territory from and to workers in relation to ­labor market integration.

For example, de-­territorialization is directly related to the ­labor pro­cess in the example of the Enclosures Acts in ­England. The act disconnected peasants from grazing land and re-­territorialized them onto textiles in the burgeoning garment industry (Elden 2005). In the posting of workers discussion, the term “de-­territorialization” was mainly used by ­legal scholars to denote the decontextualization of ­labor law and industrial relations systems from par­tic­u­lar territorial ties (Mundlak 2009). The concept of territorial embeddedness is used as a tool to further the understanding of industrial relations and ­labor pro­cesses in cases of transnational mobility of both capital and l­abor. It allows questions of location-­bound and non-­ location-­bound actions and the impact of the place on employment relations to be addressed. At the same time, it points to pro­cesses of disembedding or de-­territorialization of par­tic­u­lar importance in this context.

The Po­liti­cal Economy of Borders in the Eu­ro­pean ­Labor Market Borders as lines of demarcation have become so integrated in the way we think that we rarely notice or question them. In light of the vari­ous mobility practices in the EU, we have to take a new look at how borders relate to territory, order, and security (Bigo 2013) as well as, indeed, employment relations. Chapter 5 adopts a more explic­itly spatial perspective investigating the

552-73361_ch01_1P.indd 19

—-1 —0 —+1

04/09/18 5:58 PM

20 Introduction

-1— 0— +1—

reconfiguration of borders in a pan-­European ­labor market. Posted work incorporates two trends that have been explored separately in the po­liti­cal science, po­liti­cal economy, and industrial relations lit­er­a­tures. On the one hand, the impact of Eu­ro­pean integration on the territorial order of modern nation-­states is at the heart of a key debate in international relations and comparative politics. On the other hand, the increasing employment of workers via subcontractors or temporary work agencies as a way to weaken ­labor power and segment the ­labor force through institutional change dynamics has been explored in the industrial relations and po­liti­cal economy lit­er­a­tures.

The notion of “border” has been the topic of many discussions. Within this book, borders are understood as institutions. They are constructed of demarcating lines that refer to a territorial or functional definition delimiting membership (Bigo 2013). ­Because of their constitutive nature, borders can give insights into the social ordering of a territory (Hetherington 2003, 64). While po­liti­cal geographers used to think of borders as physical lines on the ground separating one nation-­state from another (Minghi 1963), Eu­ro­ pean integration has shown that borders are flexible and constructed by actors (Paasi 1996). Borders can serve and be constructed not only by the state (Berman 2003) but also by transnational companies, diasporas, or other actors (Adamson and Demetriou 2007). Workers, employers, capital, and ­others—­each with their own concerns—­construct borders (Herod 1998; McGrath-­Champ, Herod, and Rainnie 2010). In a pan-­European ­labor market, state borders shift. Posting via subcontractors and temporary agency firms also shifts the borders of the firm. In fact, Raess and Burgoon (2013) found a positive correlation between EU immigration influences and the incidence of employment flexibility. As ­labor markets transnationalize, the authority bounded in monolithic concepts of nation-­states is disaggregated and vari­ous state and nonstate actors are then allowed to claim t­ hese competencies, forming new territorial borders.

In the field of migration and mobility, the issue of borders is discussed in relation to how new technologies facilitate controlling the movement of ­people at the national and external EU borders (Huysmans 2006). For example, Verstraete (2001) provided a vivid account of the involvement of private companies in marketing ­human detection technologies to the Belgian port of Zeebrugge. This has given rise to an “emerging market in the ­removal of illegal refugees” (Verstraete 2001, 27). O ­ thers have focused on

552-73361_ch01_1P.indd 20

04/09/18 5:58 PM

Introduction 21

airports as sites where the intersection of technologies, subjectivities, migrations, and markets occurs, giving rise to par­tic­u­lar practices of mobility and immobility (Fuller 2003). The discussion on internal EU borders is then particularly geared t­oward non-­EU or illegal ­labor mi­grants. Nevertheless, t­hese practices point to the constitutive role of border regimes in giving rise to semipermanent, vulnerable sectors of workers. Borders do not actually prevent the movement of workers but shape the terms ­under which their movements and subsequent existences take place (Favell and Hansen 2002). In that sense, the EU’s border regime can be regarded as a major ele­ment in the flexibilization of work (Samers 2003).

Posted work embodies the abolishment of nation-­state borders induced by the EU single market. Paradoxically, the nature of the posted workers’ regulatory context creates borders within transnational workspaces for workers. ­These are not as clear-­cut as territorial borders but rather disintegrate into a multiplicity of fragmented borders. Borders are not manifested physically but the result of the movement of workers and their interactions with other actors (Guild 2009; Favell 2008). In relation to posting, ­these borders are very much related to the posting firm. Posted workers do not experience controls on their way to the posting country but at their worksite. Even though the movement across sovereign state borders no longer activates a border for EU citizens, borders still exist in the daily lives of citizens and create a system of “differentiated” memberships for workers.

The reconfiguration of po­liti­cal space lies at the heart of the Eu­ro­pean proj­ect. The new shape of the EU is the subject of manifold studies, but relatively ­little attention has been paid to the way borders are deconstructed and reconstructed in a pan-­European ­labor market. The assumption seems to be that few borders remain in the single market; however, other researchers have noted that many borders endure (Steen Knudsen 2005). Even though the EU’s four freedoms have created a common market without internal borders for l­abor and ser­vices, this book argues that borders have not become obsolete in the context of EU l­abor market integration. The de-­and re-­territorialization of state borders intersect with significant changes in ­labor markets and shifts in the borders of a firm. In developing a framework for understanding the relationship between changes in sovereign borders and changes in employment relations in the EU, chapter 5 draws on Eu­ro­pean integration lit­er­a­ture as well as on comparative institutional analy­sis and ­labor geography. It explores the position and creation of borders

552-73361_ch01_1P.indd 21

—-1 —0 —+1

04/09/18 5:58 PM

22 Introduction

in a pan-­European ­labor market. It studies the reshaping of the nation-­state from the bottom up from the point of view of actors involved in the posting relationship. Findings show the significance of borders to ­labor market regulation and firm borders to posted workers. This has negative implications for ­labor rights and transnational solidarity in the EU.

-1— 0— +1—

552-73361_ch01_1P.indd 22

04/09/18 5:58 PM

Strong Governments, Precarious Workers

Labor Market Policy in the Era of Liberalization

Philip Rathgeb

ILR Press An imprint of Cornell University Press Ithaca and London

Copyright Š 2018 by Cornell University All rights reserved. Except for brief quotations in a review, this book, or parts thereof, must not be reproduced in any form without permission in writing from the publisher. For information, address Cornell University Press, Sage House, 512 East State Street, Ithaca, New York 14850.

First published 2018 by Cornell University Press Printed in the United States of America Library of Congress Cataloging-in-Publication Data <CIP to come> Cornell University Press strives to use environmentally responsible suppliers and materials to the fullest extent possible in the publishing of its books. Such materials include vegetable-based, low-VOC inks and acid-free papers that are recycled, totally chlorine-free, or partly composed of nonwood fibers. For further information, visit our website at cornellpress.cornell.edu.

Contents Preface Acknowledgments Abbreviations 1. The Protection of Outsiders in the Era of Liberalization 2. Labor Market Policy in Austria, Denmark, and Sweden 3. Relying on the Weak: Austrian Unions and Smoothed Dualization 4. Strengthened Governments and the Erosion of Danish Flexicurity 5. Goodbye to Swedish Social Democracy and Welfare Universalism 6. Strong Governments and Precarious Workers in the Era of Liberalization Appendix: List of Interviews Bibliography

Chapter 1 The Protection of Outsiders in the Era of Liberalization <A> Introduction This book is about the losers of the liberalization era. Social scientists often call them “outsiders”, because their precarious labor market situation excludes them from the employment and social rights enjoyed by “insiders” on regular jobs. The resulting lack of security is associated with several trends that are adverse in their implications for democracy and society: declining voter turnout and political resignation (Schäfer 2013), diverging life chances and growing poverty (Tomlinson/Walker 2012) as well as poor health, and even an increased relative risk of suicide (Nordt et al. 2015). The willingness of the state to protect workers from the risks of being unemployed or ‘atypically employed’ is thus of great political and social significance.

This particular expression of economic inequality can be traced to the retreat of the state from its political commitment to ensure full employment in response to the inflation crisis of the late 1970s (Harvey 2010, Streeck 2011). Faced with rising unemployment rates, during the 1980s, the welfare states of Western Europe expanded exit routes out of the labor market such as early retirement schemes and incapacity benefits, especially for people who found themselves excluded from the rapidly growing service sector due to low or obsolete skills, chronic health problems, and weak labor demand (Ebbinghaus 2006). In this context, labor supply management by means of pre-retirement options took the place of Keynesian aggregate demand management.

With the tightening constraints of fiscal austerity, however, European welfare states of the 1990s no longer had the public resources available to ease the situation of mass unemployment through ‘labor shedding’ strategies. As a result, the main option left to adjust national models of capitalism in the interest of successful employment performance was the one of liberalization; including the differentiation of wage levels, the deregulation of employment contracts, as well as cuts in social security. Part of this transformation in welfare statehood is the turn towards ‘activation’, whereby the provision of social security for the unemployed has become conditional on active job-search, the willingness to take up any job deemed ‘suitable’, or participation in training (Clasen/Clegg 2006, 2011; Bonoli 2010). Given the growing emergence of precarious employment and welfare standards, the active re-configuration of labor market institutions to social needs of outsiders has become crucial to the redistributive capacity of the welfare state. Figure 1.1 sheds some light on the growing number of outsiders in the European Union. < insert Figure 1.1>

Despite the common reform trajectory described above, political actors could still shape the liberalization of work and welfare in different ways. That is to say, even though the neoliberal transformation of global capitalism pointed to a general expansion of market mechanisms and economic inequality over time (Streeck 2009, Baccaro/Howell 2011), the political practice of liberalization at the national level has resulted in divergent distributive outcomes for outsiders (Emmenegger et al. 2012, Thelen 2014). Therefore, the common liberalization of Keynesian post-war capitalism did not rule out variation in the reform trajectories of national welfare states, nor did it imply convergence in the redistributive capacities of European public policy regimes.

One example of this lack of convergence can be seen in the case of Austria, which deviated from the Continental European adjustment path of “pronounced dualization” between workers with stable employment (‘insiders’) versus unemployed and atypically employed workers (‘outsiders’) (Obinger et al. 2012).

In fact, the Austrian policy output ran directly counter to the reinforcement of status divisions in such diverse areas as job security arrangements, social insurance coverage, and active labor market policy spending (ibid.). Outsideroriented patterns of liberalization appear puzzling in a prototype of the Conservative welfare regime with a segmented and male-dominated institutional legacy. Existing theory would expect otherwise. Interestingly, by contrast, the reform trajectory in the Social Democratic prototype of Sweden differed profoundly from that of Austria: Sweden substantially curtailed both the coverage and generosity of benefit entitlements, while deregulating the job security of temporary workers at the same time (Davidsson 2011). As I will show, labor market policy has therefore become more egalitarian in Austria than in Sweden. Denmark, by contrast, stands in between Austria and Sweden by revealing substantial variation over time. It attracted widespread attention for undermining social divides with a policy combination of inclusive social security, human capital formation, and flexible job security arrangements. During the 2000s, however, the security-related components of this ‘Flexicurity model’ came under strain, while legislative interventions additionally curtailed the long-term power base of the union movement.

Despite the many similarities between their political economies – small size, trade openness, corporatist legacy, relative macroeconomic success, periods of center-right coalitions, and mature welfare states – the reform trajectories of Austria, Denmark, and Sweden not only differed markedly but also evolved in very counterintuitive directions. Conventional theories of partisanship (Rueda 2007, Huo 2009, Häusermann 2010) and producer group coalitions (Palier/Thelen 2010, Martin/Swank 2012, Thelen 2014) generally suggest continued regime variation between Social Democratic-Nordic and

Conservative-Continental welfare states in the distributive outcomes of economic adjustment. These two dominant lines of research in comparative political economy, however, do not hold with respect to the puzzling policy outputs observed in three small states of Western Europe. Austria and Sweden, for example, crossed levels of outsider protection in opposite directions, while the recent Danish reform trajectory underlying the erosion of Flexicurity has also been less ‘solidaristic’ relative to the Austrian case. <A> Research question and argument in brief The brief presentation of three nationally distinct political responses to the distributional tensions of the neoliberal era poses a general question of interest to academia and society alike: Why do some European welfare states protect outsiders better than others? I examine this broad question through a comparative investigation of reform trajectories in the area of labor market policy. The rationale for this choice is that labor market policy represents a crucial area that may mitigate, reproduce, or reinforce the socially corrosive effects of economic liberalization on the weakest segments of society (e.g. Emmenegger et al. 2012, Rovny 2014, Rueda 2015).

I analyze reform trajectories in three dimensions of labor market policy: employment protection legislation (EPL), unemployment insurance (UI), and active labor market policy (ALMP). The main reason for investigating change in all three dimensions is that they closely interact with each other in shaping the material situation and employment prospects of outsiders. UI and ALMP are distributive policy areas in that they provide income support and re-integration programs for the unemployed. Conventional welfare state research examines these two policy dimensions to find out how policy-makers react to the emergence of post-industrial labor markets (Clasen/Clegg 2011). Yet, as Crouch/Keune (2012) rightly point out, welfare states not only react to changing risk patterns emanating from post-industrial work and family patterns, but also co-shape them by regulating the ‘hiring and firing’ conditions for different employment contracts. Employment protection legislation may also exist as a functional equivalent to unemployment benefits in the provision of economic security (Locke/Thelen 1995, Bonoli 2003). For example, the maintenance of strict employment protection for regular workers in combination with residual unemployment support contributes to a dualization between well-protected insiders and precarious outsiders in the Mediterranean and Continental European political economies (Palier 2006). The direction of institutional change in the three areas described above has thus a profound impact on the distributive character of economic liberalization in a given capitalist regime (OECD 2008, 2011, 2015; Thelen 2014).

To explain why political actors in some European welfare regimes protect outsiders better than others, I compare the reform trajectories of Austria, Denmark, and Sweden over the past three decades, complemented with shadow case studies of Italy and Spain. The reason for concentrating on these three countries in particular is that they represent similar corporatist legacies of economic governance and three different reform trajectories. Following a mostsimilar systems design, the three cases are thus picked from among the small states of North-Western Europe that are particularly challenged by the vagaries of global capitalism and have small populations with tightly knit policy-making elites (Katzenstein 1985). The shadow case studies of Italy and Spain serve to illustrate the generalizability of the bookâ&#x20AC;&#x2122;s argument beyond the small corporatist states of North-Western Europe.

Challenging conventional explanations, I argue that the variation observed between Austria, Denmark, and Sweden does not result from differences in partisanship or producer group coalitions, but from differences in the powerdistributional interaction between trade unions and governments. High levels of inclusiveness continue to provide trade unions with an acute interest in the protection of outsiders. But the causal significance of union preferences declined in the course of the past three decades, with a shift in the balance of class power from labor to capital across the advanced capitalist countries of the Western world. National governments have accordingly prioritized the flexibility demands of employers over the social protection of outsiders. When they had a united majority of seats in parliament, they were therefore strong enough to pursue a unilateral reform strategy that excludes unions to the detriment of outsiders.

When they were weakened by (i) intra-coalitional divides or (ii) a hung parliament, on the other hand, they negotiated political deals with trade unions to mobilize an extra-parliamentary channel of consensus mobilization. This kind of weakness was instrumental in forcing governments to compensate outsiders for labor market liberalization.

The core argument of this book can therefore be summarized in one sentence: the weaker the government, the stronger the capacity of inclusive trade unions to enhance the protection of outsiders through an extension of job security regulations, unemployment benefit entitlements, and active labor market policy spending. <A> Varieties of liberalization and the protection of outsiders Liberalization refers to the expansion of market mechanisms in the allocation of material resources and life chances as the political response to the crisis of post-war democratic capitalism (Streeck 2009, Harvey 2010). Faced with rising inflation in the late 1970s, Anglo-American governments put an end to the Keynesian class compromise and initiated the evolutionary stages of

liberalization in capital, product, and labor markets (Hacker/Pierson 2010). The neoliberal era took off with the rise of monetarism (late 1970s), followed by a gradual departure from the commitment of the state to provide full employment (1980s), and continued with fiscal consolidation in tandem with the deregulation of capital markets (1990s) and the subsequent global financial crisis (post-2008) (Streeck 2011). National trajectories of liberalization had in common a conversion of the function of labor market institutions from market-constraining social rights towards market-conforming economic competitiveness (Baccaro/Howell 2011). Central to this process was the removal of legally binding constraints on employers’ strategies of capital accumulation. In terms of concrete institutional changes, this political practice implied a (partial) deregulation of job security regulations and the re-commodification of labor through activation strategies targeted at people out of work.

The socially corrosive effects of liberalization have become visible in the growing emergence of workers in precarious employment and welfare standards – that is, the so-called ‘outsiders’. Following Rueda’s definition (2006, 387), I use the term outsiders to capture workers that “are either unemployed or hold jobs that are characterized by low levels of protection and employment rights, lower salaries, and precarious levels of benefits and social security regulations”. The literature on ‘new social risks’ pointed to the social groups that face the highest probability of being unemployed or atypically employed on post-industrial labor markets. First, low-skilled workers in both the manufacturing and service sectors are exposed to greater (long-term) unemployment risks and flexibility demands.

Second, female workers, largely irrespective of skill levels, are often pushed into atypical employment contracts with reduced working hours to reconcile work and family life. Finally, young workers face difficulties in entering the primary labor market due to weak labor demand and the deregulation of fixed-term employment (Taylor-Gooby 2004, Bonoli 2007, Häusermann/Schwander 2015).

Liberalization, however, took divergent forms and distributive outcomes in different national contexts due to the mediating impact of the interplay between historically evolved institutions and domestic politics. Kathleen Thelen famously animates this empirical identification of different regimes of social solidarity, arguing that the common trend of liberalization did not rule out institutional variation in the redistributive capacities of national public policy regimes (Thelen 2014). All the same, her understanding of social solidarity is markedly different from traditional notions of de-commodification or social citizenship, and this perhaps best explains why she observes pronounced cross-country variation.

Drawing on a vast body of welfare state scholarship, Thelen claims that the institutional heritage of the industrial post-war era ceased to produce ‘solidaristic’ effects in a post-industrial environment. In this perspective, ongoing high levels of social solidarity require an institutional re-configuration of European welfare

states towards more emphasis on market-conforming human capital formation and universal minimum benefit entitlements. This argument relates to the perceived mismatch between post-industrial labor market change and industrial welfare arrangements that can be said to have caused a declining scope in and generosity of risk protection (Hacker 2004). The erosion of manufacturing employment implied greater demands for wage differentiation and (re-)training, given the lower productivity growth potential and higher qualification requirements of the service sector. At the same time, the erosion of the male breadwinner model and demographic ageing called for state support in care services, while the concomitant rise of atypical employment contracts reinforced problems of in-work poverty and insufficient benefit entitlements.

In terms of empirical outcomes, Thelen finds three different varieties of liberalization in the areas of industrial relations, vocational training and labor market policy. From a distributive point of view, her typology resonates in characterizing Esping-Andersen’s three worlds of welfare capitalism: social solidarity in Scandinavia, status protection for a (shrinking) manufacturing sector in Continental Europe, and deregulation across-the-board in Anglo-American countries. Table 1.1 summarizes her typology with a focus on labor market policy. < insert Table 1.1> First, liberalization through embedded flexibilization describes a combination of policy choices that may be subsumed under the (admittedly vague) concept of ‘Flexicurity’ (Clasen/Viebrock 2009, Burroni/Keune 2011, Rathgeb 2017). It combines inclusive social security and re-integration programs (‘security’) with liberal job security arrangements (‘flexibility’). In this strategy of supply-side solidarity, risk collectivization rests on welfare state support to enable outsiders finding poverty-free jobs and receiving security in the event of joblessness (Baccaro/Locke 1998). Second, liberalization through dualization describes increased differentiation between insiders and outsiders in the three dimensions of the dependent variable. The maintenance of job and social security for insiders goes hand-in-hand with institutional deregulation for outsiders. Finally, liberalization through deregulation describes the full dismantling of job security and welfare arrangements, while providing no compensation for outsiders through an expansion of active labor market policies at the same time. It thus implies an individualization of risks at the expense of labor as a whole.

Following Thelen’s definition of ‘embedded flexibilisation’, I conceptualize an expansion in the protection of outsiders as a set of institutional changes consisting of three elements: (1) the extension of benefit entitlements to workers with employment contracts who previously lacked coverage, (2) the expansion

of spending on active labor market policy, and (3) the homogenization of job security regulations between permanent and temporary workers. Conceived in this way, ‘embedded flexibilization’ enhances the social security of outsiders in the event of unemployment, while smoothening the chance of re-employment into the primary labor market and impeding the growth of precarious employment contracts at the fringes of the labor market. <A> Conventional explanations Why should national models of capitalism diverge in their scope of risk protection, despite powerful structural pressures for cross-national convergence?

International competition, European integration, de-industrialisation, demographic ageing, ‘financialisation’ and declining growth of the real economy, inter alia, pose undeniable constraints on the domestic capacity of national sovereign democratic states to reconcile the distributional tension between social solidarity and capital accumulation (Garrett 1995, Rhodes 1996, Kaufmann 1997, Pierson 1998, Scharpf 2000, Alber 2002, Stockhammer 2004). Simplifying somewhat, the principal empirical studies of comparative political economy offer two different theoretical approaches to the question of cross-national variation in risk protection, which can be summarised under the rubric of (1) producer group theory versus (2) partisanship theory.

The first theoretical approach – producer group theory – highlights the role of employers and unions in shaping interest group coalitions to achieve their preferred policy output. Leaving aside important details for the moment, the primary focus of this research agenda is to explain how different strategies of interest group action emerge in the face of a given set of domestic institutions and power relationships. Recent contributions have placed strong emphasis on the inclusiveness of producer groups as the dominant cause of variation in the protection of outsiders (Martin/Swank 2012, Thelen 2014), whereas contemporary classics were more attuned to differences in class power (Korpi 1978, Esping-Andersen 1985).

The second theoretical explanation – partisanship theory – emphasizes the vote-seeking action of political parties. Putting voters and elections front and center, this research agenda looks at how parties respond to electoral demands against the backdrop of fiscal constraints and institutional legacies. Central to electoral theory is the assumption that differences in the electoral alliance behind party coalitions reflect the dominant cause of variation in the protection of outsiders (Rueda 2007, Pontusson 2009, Beramendi et al. 2015). In sum, conventional theories look at either interest groups or political parties. Neither approach, however, allows for an interaction between these two types of agencies. Perhaps even more importantly, however, both lines of research tend to underestimate shift in the capitalist context in which the politics of labor

market reform plays out. I will now review these existing explanations to show that my empirical puzzle empirically contradicts their theoretically derived predictions. Second, I will delineate my alternative argument from these two lines of research. <B> Producer group theory The classic interpretation among producer group theory is that redistributive state action hinges on labor’s power resources; that is, the relative strength of the alliance between Social Democratic parties and union movements (Korpi 1978, 1983). According to this power resource approach, differences in labor power explain variation in welfare state size. A central assumption is that industrial welfare states are not only a product of working class mobilization, but, once in place, also turn into a power resource per se, because social security arrangements have a de-commodifying impact on labor (Esping-Andersen 1985, 1990). It follows that the power resource approach assigns employers the role of profitseeking antagonists against the welfare state, given the strict relationship between labor power and politically achieved social rights (Korpi 2006; see also Streeck 1997).

The general claim that employers are hostile to welfare states evoked a revisionist counter-attack in the institutionalist study of national capitalisms. At its most basic, the literature along the lines of the varieties of capitalism framework (VoC) linked institutional variation to differences in national production strategies of firms, and not to the historically specific strength of labor vis-à-vis capital (Hall/Soskice 2001). Accordingly, centralized business associations help recalibrate domestic welfare arrangements with a view to fostering skill development and competitive advantages, thereby creating opportunities for states to draw on cross-class cooperation and pursue social investment policies (Swenson 2002, Mares 2003). All-encompassing neocorporatist policy-forums – composed of centralized peak level associations and the state – are thus posited to be the necessary condition for protecting the most vulnerable social segments of society (Martin/Swank 2012, Martin/Thelen 2007).

Yet, the firm-centered argument of the VoC framework does not hold with respect to the forceful demands of employers’ associations for market-clearing adjustment strategies in the neoliberal era in general (Hassel 2009) and the forceful escape of Swedish employers from corporatist bargaining arrangements in particular (Lindvall/Sebring 2005). If anything, in fact, a growing amount of contemporary literature powerfully contests the central role assigned to organized employers by VoC approaches, arguing that business consented to welfare state expansion only in the face of political constraints in order to pre-

empt further market correction (Hacker/Pierson 2002, Korpi 2006, Baccaro/Howell 2011, Paster 2012, Emmenegger 2014).

Unlike the revisionist literature, Thelen (2012, 2014) claims that the protection of outsiders – or “embedded flexibilisation” as she termed it – rests on the capacity of unions to incorporate the growing number of service-sector workers into their traditionally manufacturing-dominated membership base.

Here, encompassing unions are seen as the only type of organization that can mobilize social support for redistributive state action on behalf of outsiders and new social risk groups. By contrast, shrinking trade unions in economies with large industrial outputs in countries such as France and Germany prioritize the interests of manufacturing core workers (Palier/Thelen 2010). Accordingly, the second necessary condition of “embedded flexibilisation” is the capacity of the state to coerce peak level associations into an agreement when tripartite negotiations appear deadlocked or employers are simply unwilling to participate in corporatist bargaining. This kind of state capacity resonates with the “shadow of hierarchy” argument, providing governments with legislative and executive tools of economic governance (Scharpf 1997).

Central to Kathleen Thelen’s coalitional theory is the claim that the inclusiveness of national peak level associations is the key factor in explaining institutional variation: the more encompassing, the more egalitarian the policy output. While the presence of centralized business associations remains important in her account, Thelen places stronger emphasis on high levels of labor organization than revisionist contributions inspired by the VoC-framework.

Therefore, accordingly, we should expect more ‘solidaristic’ reform trajectories in Scandinavia than in Austria. Yet, this is not the case: today, outsider protection through labor market policy is stronger in Austria than in Sweden, while Denmark has started to move from egalitarian levels into a (modest) dualist direction in the past decade.

Thelen’s work was pioneering in highlighting the relationship between the distributive outcomes of liberalization and the mobilization of producer group coalitions. Her approach appears persuasive in explaining continued regime variation in wage inequality, given the relative membership strength of Scandinavian peak level associations vis-à-vis their Continental European counterparts. Yet, my empirical focus is on public policy, and not industrial relations. Thus, the membership strength emphasized by Thelen (2014) or the administrative power position derived from the Ghent system highlighted by Clegg (2012) does not necessarily translate into strong union influence on the design of public policy changes. Union-based explanations ignore the fundamental fact that governments may turn against negotiated reform to exclude organized labor from the policy-making process, irrespective of membership strength and corporatist legacies. Scholarship on social pacts and

corporatism demonstrates that ideological and programmatic unity facilitate the capacity of majority governments to impose unilateral reform ambitions on organized labor (Baccaro/Simoni 2008, Afonso 2013). The omitted factor in producer group theory is thus the political choice of whether to share policymaking authority with organized labor. This is a serious shortcoming for a theory of institutional change, because it assumes away the possibility of a strong government with an autonomous reform agenda.

The general claim emerging from my emphasis on unilateral government action is that change in labor market policies does not necessarily originate in industrial relations. Producer group approaches argue that the political roots of institutional dualisms are to be found in the shrinking coverage of collective bargaining (Palier/Thelen 2010). The logic behind this argument stems from the notion of â&#x20AC;&#x153;institutional complementaritiesâ&#x20AC;? (Hall/Soskice 2001), whereby the gradual erosion of industrial relations systems unravels the economic viability of labor market and welfare policy arrangements. In essence, this mechanism attributes dualistic labor market and welfare reforms to functional pressures generated by industrial relations, leaving little room for government agency. For sure, these policy domains are tightly interrelated in the functioning of economic models, but the mechanism underlying dualization may also work precisely the other way around. That is to say, unilateral government action may set in motion dualistic tendencies in state policy without any preceding impetus from industrial relations. The Swedish case is illustrative in this regard. Thanks to encompassing unionization and collective bargaining coverage, Thelen herself finds that Sweden displayed the second lowest percentage of employees in low-wage work in 2010 (Thelen 2014, 130). Yet, the picture looks very different in the area of labor market policy: dualistic job security arrangements go hand in hand with sharply declining unemployment insurance coverage and spending on active labor market policy. Related to this policy output, Sweden experienced the fastest growing relative poverty rate in the OECD, slipping from the most poverty-free country in 1995 to 14th place behind Germany and Ireland in 2010 (OECD 2013). In a similar vein, the Danish center-right government (2001-2011) passed various institutional reforms that threatened the viability of encompassing unionization, which might well impact on collective bargaining coverage in the future (Ibsen et al. 2013). Theories of producer group action fail to capture these trends. I will now turn to previous research of partisanship theory to detail existing explanations based on the policy-making action of political parties. <B> Partisanship theory Contemporary partisanship theory scholarship highlights the impact of sociostructural change on the interaction patterns between parties and voters (Oesch 2006, Kriesi et al. 2008, Häusermann 2010, Beramendi et al. 2015). De-

industrialization, immigration, demographic ageing, and rising female labor market participation reshaped the traditional class- and family- basis, on which parties had competed for votes in the post-war era. This new post-industrial context drove a wedge between the policy preferences of low-skilled and loweducated constituencies on the one hand, and the highly educated “sociocultural professionals” mostly employed by the public sector on the other. While the former classic working class prefers income transfers and high job security to cope with current social needs, the latter group prefers social investment policies that have favorable long-term effects on their capacity to reconcile work-family life and lifelong learning. Part of the explanation for growing intra-labor divides comes from the emergence of conflicts over post-material values and immigration (Kriesi et al. 2008). In political-economic terms, the explanation comes from the heterogeneity of post-industrial risk profiles, which assumes divergent policy demands from different labor constituencies (Rueda 2007). In this perspective, the protection of outsiders hinges on the capacity of parties to forge an electoral alliance between the traditional working class and sociocultural professionals (Beramendi et al. 2015, 29 f.).

The recognition of intra-labor divides in the electoral arena is an essential contribution to the study of electoral politics and economic adjustment. Yet, among partisan theorists, there is disagreement over the effects of labor heterogeneity on the policy-making action of parties, especially those with a social-democratic character. This disagreement plays out between the labor dualism theory of David Rueda (2007) and the classic partisanship theory reinvigorated by Jonas Pontusson (2009). Rueda’s powerful contention is that social-democratic parties have strong incentives to prioritize the interests of insiders over outsiders. He argues that, faced with fiscal constraints, social democracy lost the capacity to pursue maximalist solutions on behalf of insiders and outsiders, thereby facing a new electoral trade-off. Ponutsson (2009), however, points to evidence showing that (Nordic) political economies with strong social-democratic parties remain more inclusive and egalitarian than (Continental European) countries with strong Christian-democratic parties.

More fundamentally, Schwander (2012) questions the very potential for a full politicization of the insider-outsider divide. Accordingly, the heterogeneity of workers in class, age, gender, and skills interact in shaping welfare state preferences, thereby pointing to other cleavages than just employment risk.

This contested debate about the role of social democracy appears indicative of the fact that even center-left governments are prone to pursuing dualization.

In the most extensive volume on dualization to the present day, the authors conclude that “the chapters in this book do not report strong differences between left- and right-wing governments with regard to dualization” (Emmenegger et al. 2012, 311). In theory, however, the protection of outsiders

– in terms of job security, unemployment benefits, and training opportunities – would be an essential vote-seeking device that reflects the egalitarian preference structure of the Nordic electorate (Larsen 2008, Svallfors 2011). Universalist welfare state legacies, low tax burdens for business, and broad public sector employment traditionally appealed to labor as a whole, holding together an encompassing pro-welfare coalition between ‘old’ and ‘new’ electoral constituencies (Beramendi et al. 2015, 30). Hence, an electoral logic of labor market reform should translate into strong incentives for Nordic governments, especially those from the center-left, to combat growing insider-outsider divides.

However, the Danish and Swedish center-right coalitions managed to deuniversalize unemployment protection and cut active labor market policy spending (Goul Andersen 2012a), while in Austria they homogenized job security arrangements and expanded training. Austrian grand coalitions improved the social protection of outsiders, while the opposite is the case for Swedish social democratic single-party governments (Obinger et al. 2012) and, to a smaller extent, Danish center-left governments (Goul Anderssen 2012, Rathgeb 2017).

It seems that the partisanship narrative about the left-right cleavage is too simplistic to capture these developments.

An obvious problem with theories of vote-seeking agency is that they tend to ignore shifts in class power and its implications for organized interests. Laborinclusive negotiations over the design and implementation of institutional changes condition the strategic calculations of partisan reform ambitions and provide trade unions with access to public policy-making. Trade unions have important stakes in this policy domain, because labor market reforms interact with and impinge on the institutional landscape of collective bargaining and effective labor power. By contrast, electoral mobilization on grounds of unemployment support has traditionally been constrained by low political solidarity for the unemployed (Alber 1982). As Pierson (1994, 103) argues, voters pay less attention to labor market policy than to social security for pensioners and the sick or disabled, thereby underscoring the explanatory significance of class power. Therefore, the historical context of declining union strength does not naturally “give more prominence to electoral partisan competition” (Beramendi et al. 2015, 26), but may rather help explain the common liberalizing trend in labor market reform. Moreover, the neo-corporatist literature points to the diverse mechanisms through which governments continue to fall back on negotiated reform in the neoliberal era (Visser/Hemerijck 1998, Ebbinghaus/Hassel 2000, Culpepper 2002, Traxler 2010, Afonso 2013).

Second, the partisanship literature tends to conflate trade unions with insiders that resist any attempt to enhance the material situation of outsiders in order to defend their own status quo (Rueda 2007, Häusermann 2010). Yet, a number of studies show that high levels of labor organization and unity create an acute

interest in the representation of outsiders (Becher/Pontusson 2011, Thelen 2014, Gordon 2015), which can be underpinned by institutional incentives attached to the administrative setup of the unemployment protection system (Clegg 2012). An inclusive representational outlook, however, is not necessarily confined to strong unions with encompassing membership bases.

Naczyk/Seeleib-Kaiser (2015), for example, demonstrate how pre-institutional settings led labor movements in such diverse cases as Belgium, Britain, France, and Germany to push for an extension in the coverage of supplementary pension schemes to less privileged segments of labor. Moreover, as Vlandas/Benassi (2016) points out, the ideological working-class orientation of unions in the Mediterranean countries may trump an insider-oriented, occupational egoism in the preference-formation process. This claim resonates with findings from the union revitalization literature, which documents that weaker unions opened up their bargaining and recruiting strategies to the non-unionized margins of the workforce (Baccaro et al. 2003, Heery/Adler 2004, Jódar et al. 2011).

That shrinking trade unions did accept policy deals to the detriment of outsiders often reveals their long-term power considerations in the face of declining union power, and certainly not their policy preference (Emmenegger 2014). Previous studies showed that even the rapidly declining union movement of Austria was the most forceful advocate of outsider efforts in labor and employment law (Tálos 1999, Obinger et al. 2012). Unions in the Nordic countries have traditionally been the central actors with stakes of an organizational (Ghent system) and representational (encompassing membership) nature on behalf of the weakest segments of the labor market. If anything, in fact, universal access to benefits and training had been the core demands of union confederations in Denmark and Sweden long before the idea of ‘Flexicurity’ had gained prominence in scholarly debates and policy circles.

The capacity of organized labor, however, to assert favorable concessions for outsiders rests on the political choice of parties to share policy-making authority; that is, the calculus of governments to ensure successful economic performance, and thus their electoral fortunes, by means of political exchange and labor acquiescence (Baccaro/Lim 2007, Baccaro/Simoni 2008, Afonso 2013). Even if permanent austerity and globalization challenge the policy-making autonomy of governments, they may still opt for different actor configurations at the bargaining table: they can work with unions or pursue reforms unilaterally. This political choice has important implications for the power relationships at the bargaining table. Union exclusion eases the adoption of welfare cuts and employment deregulations that stimulate an internal devaluation of the economy, while union inclusion requires lengthy and costly political exchanges between the state, capital, and labor. Conventional approaches focusing on either partisanship

or producer group coalitions miss this power-distributional interaction between political parties and organized labor. <A> The Argument: Unions and Governments in the Era of Liberalization Partisanship and producer group approaches share two features. First, both argue that the viability of social solidarity rests on encompassing pro-welfare coalitions. The causal mechanism behind these claims is that only a united intralabor coalition has the capacity to push for inclusive policy choices that satisfy divergent social demands. Partisan scholars such as Beramendi et al. (2015) highlight the partisan task of reconciling the policy demands of divided electoral constituencies, while producer group scholars like Thelen (2014) emphasize the unionist task of incorporating service-sector workers into the traditionally manufacturing-dominated union movement. Although one of these theories looks at parties, and the other one at unions, they have in common the same underlying logic. Second, they focus on either parties or unions to understand the national dynamics of change in OECD capitalism. Both lines of research have been very instructive insofar, as their arguments have helped us to understand the electoral calculations of political parties and material interests of producer groups in the politics of economic adjustment. However, neither approach focuses on the interaction between parties and producer groups under the influence of shifts in class power.

Challenging conventional explanations, I argue that the variation observed between Austria, Denmark, and Sweden does not result from differences in partisanship or producer group coalitions, but from differences in the powerdistributional interaction between governments and unions. In the Keynesian post-war era, governments in all three countries routinely involved organized labor to cope with the vulnerability of small nation states in the international capitalist economy (Katzenstein 1985). During the era of liberalization, however, the distribution of policy-making influence between governing parties and unions evolved in very different directions. Austrian governments, largely irrespective of partisanship, continued to work with unions through the process of policy concertation, thereby securing organized labor the opportunity to strike important concessions for outsiders. After turbulent years of party system change, Danish center-right governments returned to consensual modes of adjustment with organized labor, enabling unions to demand strong state support for training in exchange for welfare-to-work activation. The pattern of union involvement in Denmark, however, turned out to be more fragile than in Austria, paving the way for unilateral welfare cuts at the cost of encompassing social security. By contrast, Swedish unions were gradually marginalized from the policy-making process over time. Union influence has traditionally been reliant on social-democratic government partisanship, but party-union ties came under

strain and the center-right bloc gained strength as well as unity over time. While the policy preferences of Austrian, Danish, and Swedish unions were very similar, their political capacity to influence the reform process differed markedly.

My theoretical framework attempts to account for this variation in the capacity of organized labor to protect outsiders. I argue that the capacity of trade unions to achieve concessions for outsiders is conditional on the interaction with a politically weak government. Weakness, in this context, refers to a low level of autonomous reform capacity, making governments unable to formulate or pass a unilateral reform strategy that excludes unions. As I will show, weakness often results from intra-coalitional divisions between ideologically divided governing parties or a lack of parliamentary support behind minority governments. Trade unions are influential under these conditions, because their support provides weak governments with an extra-parliamentary channel of consensus mobilization. This creates powerful incentives for governments to delegate policy-making authority to trilateral negotiations between the state, capital, and labor. Unions can credibly demand substantial concessions from a weak government, because their assent to new policies can make the difference between a reformâ&#x20AC;&#x2122;s success and failure.

The presence of a weak government became necessary for union influence, because unions themselves lost the structural power to impose tight constraints on the unilateral reform ambitions of partisan actors. Overall, unions lost members, unity, and thus electoral significance, while the global turn to monetarism and austerity under the conditions of exogenous competitive pressures disciplined their political bargaining power vis-Ă -vis business as well as governments. I therefore expect the preferred policy output of parties to prevail under the conditions of a strong government, because a united majority allows them to pursue their preferred policy output against weakened unions. In the era of liberalization, where parties of all complexions face a structural incentive to reduce the reservation wage and increase labor market flexibility, the policy output of strong governments precludes concessions to unions to the detriment of outsiders.

The argument I present below in more detail draws on a variety of earlier contributions, especially the power resource approach of Walter Korpi (1978, 1983) and Gøsta Esping-Andersen (1985) as well as the liberalization theory advanced by Wolfgang Streeck (2009). It builds on the assumption that differences in the balance of class power are central to the causal dynamics of national liberalization paths. But it also deviates from the focus of the power resource approach on social-democratic partisanship by highlighting instead the role of organized labor in capitalist development. Neoliberal economic ideas, fiscal austerity, globalized market competition, and changes in the electorate, among others, cut across the left-right divide in important ways (Blyth 2001,

Rueda 2007, Mair 2013), so that even social-democratic parties have proved willing to dualize labor markets and welfare. By contrast, inclusive union confederations continue to mobilize political support for outsiders, because their organizations incorporate workers that are hit hardest by inequality in employment and social rights (Thelen 2014, Gordon 2015).

Yet, the preferences of unions are not causally omnipotent, given their loss of structural power vis-Ă -vis capital in the neoliberal era. My argument therefore emphasizes the political opportunities unions face when governments are weak.

Under conditions of weakness, governments are more responsive to the demands of unions, because they need support from extra-parliamentary actors in the pursuit of consensus mobilization. This claim about weak governments draws on the social-pact literature and extends to governments of the partisan right as well as the left (Baccaro/Lim 2007, Baccaro/Simoni 2008, Afonso 2013).

The presentation of my argument proceeds as follows. First, I will consider how the gradual stages of liberalization shifted the balance of class power from labor to capital. An appreciation of this shift is necessary to understand why governments had a structural incentive to prioritize the preference of employers in pursuing dualistic policy choices to the detriment of outsiders. Second, I will shift the analysis to the role of organized labor and explain why inclusive union movements â&#x20AC;&#x201C; characterized by high density and organizational unity â&#x20AC;&#x201C; resist partisan attempts aimed at increased inequality through dualization. Finally, after having outlined the contrasting preferences between governments and unions, I will discuss the opportunities unions face to influence the distributive outcomes of national paths of liberalization when governments are weak. <B> Why governments followed the preference of capital for dualization Governments operating in capitalist democracies have to reconcile the distributive tension between capital accumulation and social stability (Polanyi 1944 [1957]). But they found themselves increasingly unable to perform this dual task when egalitarian redistribution seemed no longer conducive to capitalist growth. This historical change in the capitalist context implied not only a fundamental constraint on the policy-making autonomy of governments. It also meant a shift in the balance of class power by creating opportunities for employers to push governments in relatively egalitarian countries towards dualization. This is not the place for a comprehensive review of the reasons for this shift in the capitalist context. For the purpose of this book, it suffices to discuss the class power implications of the transition from the Keynesian publicpolicy regime to the gradual stages of liberalization over the past roughly four decades (Silver 2003, Glyn 2007, Streeck 2011, Emmenegger 2014, Baccaro/Howell 2017).

The era of liberalization started with the collapse of Keynesian macro-economic management in response to the two oil price shocks of the 1970s. One of the main problem posed by the economic turmoil of the time was that flexible monetary policy adjustment in combination with full employment policies no longer seemed capable of controlling inflation. The Keynesian class compromise therefore began to be seen as the cause of the capital accumulation crisis rather than as its solution. This rethinking ultimately led the American Federal Reserve Bank under Paul Volcker to switch to a restrictive monetary policy in an attempt to put the wage demands of unions under severe pressure and thereby undermine inflationary wage-price spirals, retrospectively called the “Volcker shock”.

National governments in Western Europe were quick to follow the American lead in sharply raising interest rates and containing government spending.

The global shift to monetarism called into question the functional effectiveness of Keynesian policy instruments for national governments. This had two profound consequences for the political assertiveness of national trade unions vis-à-vis capital. On the one hand, it implied that unions had to discipline their wage demands in order to avoid increasing unemployment. As monetary policy was used to combat inflation, it could no longer serve the unions in stabilizing the employment performance. This led to the return of mass unemployment in the advanced capitalist countries (Lindvall 2010). On the other hand, the shift to a restrictive monetary policy undermined the incentive for governments to trade wage restraint for policy concessions to unions (Scharpf 1987). However, this capacity to hold back wage demands in the interest of inflation control was the main asset that unions with an encompassing, centralized, and hierarchical structure could lay before governments to promote their inclusion in the policymaking process (Lehmbruch/Schmitter 1982). A non-accommodating monetary policy thus called into question the viability of the corporatist ‘political exchange’ between governments and unions (Goldthorpe 1984), because governments were no longer reliant on the capacity (and willingness) of unions to internalize wage restraint for successful economic performance.

In addition to monetarism, nationally anchored unions and governments were confronted with the accelerated expansion of market relations beyond national frontiers, also known as globalization (Garrett 1995). As employers became less dependent on their home countries for the pursuit of profits, they could threaten unions and governments alike to exit the domestic arena by shifting production sites to more favorable political jurisdictions with lower tax obligations and non-wage labor costs (Rhodes 1996, Scharpf 2000). This change in power relations brought about by global trade and capital account liberalization went hand in hand with a massive increase in the supply of lowpaid workers in the global South that came to compete with high-paid workers of the global North in the manufacturing sector (Milanovic 2016). Globalization

was thus linked to the subsequent de-industrialization of Europe, because it allowed for cheaper production opportunities outside European borders. In other words, increased capital mobility and foreign direct investment (FDI) under the conditions of freer trade pushed national economies into a situation of strong international competitiveness pressures. Capital could therefore threaten to punish unions and governments by refraining from domestic investment and thereby causing growing unemployment. The credibility of this threat was buttressed by a concerted counter-offensive on the part of employers’ associations and their affiliated think tanks to advance neoliberal economic ideas and reverse the tide of squeezed profits against ever more demanding union movements (for Europe, see e.g. Pontusson 1992, Glyn 2006; for the US, see e.g.

Lafer 2017, MacLean 2017).

While monetarism and globalization empowered capital vis-à-vis labor, the emerging fiscal crisis of the 1990s enhanced the political assertiveness of employers in their demands for cuts in the reservation wage and pension entitlements (Pierson 1998, Bonoli 2007). The creation of the European Economic and Monetary Union (EMU) contributed to this shift in class power by delegating authority over monetary policy to the European Central Bank (which is designed to be even more independent than the German Bundesbank) and requiring the achievement of economic convergence criteria as well as the removal of all barriers to capital mobility (Baccaro/Howell 2017, 187-191). The subsequent Stability Pact, which obliged governments to rein in public spending and inflation rates, therefore underpinned the fiscal consolidation agenda of this period (Scharpf 2002). In the absence of strong economic growth and Keynesian aggregate demand management, national governments had to impose social spending cuts on organized labor to consolidate the public budget without lifting taxes on business to prevent capital flight. Declining productivity growth due to deindustrialization as well as demographic ageing aggravated the problem of reducing public debt, so that a strategy of ‘labor shedding’ or ‘public deficit spending’ was no longer a viable option (Ebbingshaus 2006).

To reconcile the popular demands from voters and unions with the structural demands of capital accumulation, governments liberalized their finance markets instead. As Crouch (2009) argues, this liberalization was used to stimulate access to cheap credits when governments could no longer embed market relations with public spending, thereby essentially ‘privatizing’ Keynesian deficit spending. The socialization of bad loans in response to the collapse of the American financial system reinforced the problem of fiscal austerity, because it led to suspicions about the sustainability of record high public debt levels. Global financial investors and institutions therefore required sovereign nation states to consolidate public finances in exchange for access to sound money (Streeck 2013).

Finally, de-industrialization, growing white-collar and female employment as well as ethnic heterogeneity, in tandem with an increasingly individualist lifestyle led to a heterogenization of the working class (Oesch 2006). Even though a number of unions remained externally encompassing (e.g. Denmark and Sweden) and internally centralized (e.g. Austria), these social changes nevertheless weakened the mobilizing capacity of organized labor. To be sure, governments continued to pursue negotiated reforms with organized interests as unions possessed an institutionalized veto-position on social security boards while partisan actors pursued a political cover for unpopular spending cuts, especially in the area of pension policy (Pierson 1996). But the strength of unions to achieve substantial concessions in return for their consent to controversial policy packages was in decline, when labor became more fragmented in terms of skills, earnings, and ideological orientation. The impact of liberalization on domestic class power relationships was thus complemented by a fragmentation and decline of the unionized workforce.

To summarize, the stages of liberalization – the monetarist turn and trade/capital account liberalization (late 1970s), the political acceptance of unemployment (1980s), the fiscal consolidation agenda and creation of EMU (1990s), and the socialization of bad loans (post-2008) – gradually altered the distribution of class power and thus facilitated shifts in labor market policy towards the preferred direction of employers. Increased wage differentiation and labor market flexibility have thus become the dominant policy instruments available to stimulate job creation (Culpepper/Regan 2014, Baccaro/Howell 2017). In countries previously committed to high levels of social solidarity, the resulting pressures for liberalization started at the fringes of the labor market and thus affected outsiders in particular (Palier/Thelen 2010).

Although partisan differences have declined in light of these structural constraints, they have certainly not disappeared. Political parties still cater to diverse electoral constituencies and thus fight over policy choices. Socialdemocratic parties, for example, are increasingly composed of (high-skilled) outsiders and should thus be more moderate than the political right with respect to dualization (Gingrich/Häusermann 2015, Häusermann et al. 2015). Such an electoral assumption would be in line with the declared goal of the socialdemocratic “Third Way” to promote labor market flexibility only to the extent that it reconciles social solidarity with the competitiveness demands of globalization (Blair/Schröder 1998). The social-democratic motive behind (moderate) dualization contrasts with the more general objective of center-right parties to individualize the risks of becoming unemployed or atypically employed (Jensen 2014, Ch. 6). We should thus expect center-right parties to be more radical than their center-left counterparts in pursuing insider-outsider divides.

My argument about the impact of class power on the policy preference of

governments is thus compatible with a common direction of partisan reform ambitions, but also with remaining partisan differences in the degree of dualization.

While the shift in the balance of class power from labor to capital was a common trend in the Western world, its effect on union preferences was not equally pronounced or influential across different national models of capitalism.

What we have yet to discuss are therefore the conditions under which unions retain the strategic capacity to incorporate outsiders into their representational outlook and resist dualizing reform plans by governments. In the next section of this chapter, I will elaborate on this question. <B>Why inclusive union movements resist dualization A core finding in the dualization literature is the recognition that encompassing and centralized unions are better able to mobilize political support for outsiders than small and decentralized unions (Palier/Thelen, 2010, Thelen 2014, Gordon 2015). First, encompassing unions organize a higher share of outsiders, thereby directly incorporating their demands into the interest formation process. Second, centralized unions prevent the formation of particularistic policy priorities, because the confederal elite incorporates the policy demands from union affiliates that are exposed to the growth of outsiders.

In other words, centralized unions give voice to the sectors that are hit hardest by atypical employment and unemployment. Moreover, administrative roles in the provision of unemployment insurance give trade unions strategic stakes in the protection of the unemployed (Clegg 2012). Drawing on Gordon (2015), my three cases form part of the five most inclusive unions in the OECD (see Table 1.2). His index of inclusive unionism comprises measures of average union density rates, centralization rates, and involvement in unemployment benefit administration for the period from 1985 to 2005. < insert Tabel 1.2> Sweden and Denmark rank first and second, scoring 78.7 and 73.6, with very high density and a central union role in the unemployment benefit system.

Austria ranks fifth, at 59.2, well above the remaining countries. We should expect the relatively high levels of inclusiveness in labor movements in Austria, Denmark, and Sweden to produce a stake in the protection of outsiders. Danish and Swedish unions display the highest density rates, while Austrian unions boast the highest levels of concentration and centralization in the OECD. High union density rates in Denmark and Sweden are in large part a product of unionsâ&#x20AC;&#x2122; responsibility over the administration of voluntary and state-regulated unemployment insurance; that is, the so-called Ghent system (Rothstein 1992).

The cost-benefit attractiveness of joining unemployment insurance traditionally acts as a recruitment device for trade unions in the Ghent countries of Denmark, Sweden, Finland, and Belgium (see Clasen/Viebrock 2008 on this mechanism in detail). By contrast, Austrian unions cannot rely on a “Ghent effect”, given that membership in unemployment insurance is mandatory and the social partners in the corporatist institutional setting share responsibility for its administration. The Austrian labor movement has experienced a significant membership loss over time, with unionization declining from almost 68 percent in 1960 to a mere 28 percent in the early 2010s. Yet, unlike the Ghent countries, the Austrian union confederation (Österreichischer Gewerkschaftsbund, ÖGB) can draw on its unquestioned leadership and political mandate on behalf of its union affiliates, providing levels of concentration and centralization that are unmatched from an international comparative perspective (Traxler/Pernicka 2007).

The implication is that organizational unity through centralization and concentration can be an effective functional equivalent to high density rates in the political representation of outsiders. As in the case of Austria, it allows the weaker unions to push the confederal elite to resist dualization by offsetting an unequal distribution of sectoral intra-union power in the interest formation process (Thelen 2014; see also Western 1997, Ch. 3). Union merging further enhanced organizational unity in the interest of outsiders, because it made a growing share of member unions affected by the usage of ‘atypical’ employment contracts, often characterized by a lack of social rights. Notably, the number of member unions gradually declined from 16 in 1978 to 7 in 2009, with the result that even the traditionally male-dominated metalworkers union – the Produktionsgewerkschaft (PRO-GE, established in 2008) – came to include sectors with female part-time workers, albeit to a moderate degree relative to the service sector.

While organizational unity implicates Austrian unions in the fate of outsiders despite a significant membership loss, precisely the opposite relationship seems to underpin labor solidarity in Denmark and Sweden. High density rates incorporate a high number of workers at risk of unemployment directly into their membership base (Becher/Pontusson 2011), but union fragmentation between the three separate confederations for manual, white-collar, and professional employees poses a challenge to the effective representation of outsiders. This problem is particularly acute in Sweden, where the main voice for outsiders – the blue-collar peak union confederation (LO) – lost its hegemonic position due to a transfer in membership to white-collar unions, mainly emanating from deindustrialization. As a result, the two non-manual confederations have come to exceed the membership levels of the LO. In Denmark, the LO faced a similar membership shift, but was more successful in maintaining its dominant position as it has traditionally included lower-level white-collar grades too. That said,

generous and inclusive unemployment support has not only been central to the interest representation of the Danish and Swedish LO’s in particular, but to the membership recruitment strategy of the labor movements as a whole (Clasen/Viebrock 2008).

In addition to high rates of organizational unity and density, recent contributions find two alternative mechanisms underlying the political support of unions for outsiders. First, as Vlandas (2013) finds, French unions pushed for regulation of temporary work contracts to undermine the ‘replaceability’ of permanent workers. He attributes this political choice to the presence of general skills, low wage coordination, and similar educational attainments, which all together increased the competition posed by temporary workers vis-à-vis permanent workers. This claim, as he himself points out, does not hold in cases with a more specific skill set and high levels of wage coordination. Second, as Benassi/Vlandas (2016) argue, a strong ideological working-class orientation broadened the representational outlook of unions in Southern Europe towards temporary agency and marginal worker groups. But such an argument can also not be behind the choices of unions in the corporatist economies of NorthWestern Europe, because their post-war identity evolved on the basis of class cooperation and state involvement, not class conflict. Still, these findings are in line with my claim that unions are not necessarily dualizing forces in the era of liberalization, as suggested by Rueda (2007).

The formation of inclusive union preferences, however, is not sufficient for the successful protection of outsiders, because the gradual decline in union power has weakened the assertiveness of labor in the policy-making process. My argument is that the decline in union power enhances the relevance of national governments in conditioning the level of union influence on policy outputs. They can still grant concessions to unions, but a unilateral reform strategy that excludes unions has become less risky for successful re-election and economic performance than in the Keynesian post-war era. Yet, despite a decline in union power, governments are not always the driver behind the reform process. They are often unable to shape public-policy outputs, when they are internally divided or lack a parliamentary majority. In this situation, governments face powerful incentives to share policy-making authority with unions, because they are too weak to formulate and pass a common reform agenda independent from an extra-parliamentary channel of consensus mobilization. The presence of a weak government is therefore the main condition under which unions can gain influence despite a decline in power resources. <B> Government strength and union influence Weakness often results from intra-coalitional divisions between ideologically divided governing parties or a lack of parliamentary support behind minority

governments. Both conditions increase the reliance of a government on the support of others for consensus mobilization, which creates opportunities for unions to influence state policies. To assess the strength of a government, I look at its vote share and partisan composition. Put simply, when governing parties come from the same partisan left-right bloc and enjoy a majority in parliament, they have a high capacity to pursue a unilateral reform path that excludes union influence, and vice versa. The following section proceeds to illustrate the logic of this operationalization by discussing different formations of government according to their autonomous reform capacity, which I call government strength. I will thereby draw on the literature of government formation (Crombez 1996, Laver/Shepsle 1996), social pacts (Baccaro/Lim 2007, Hamann Kelly 2007, Baccaro/Simoni 2008), and recent contributions to the politics of consensus building (Afonso 2013, Alexiadou 2013, Lindvall/Knotz 2015).

First, a single-party majority government obviously reflects the strongest possible formation in democratic political systems. It may legislate its preferred policy output without any need to rely on the support of other parties or interest groups.

Such governments are common in Anglo-liberal countries with first-past-thepost electoral systems that usually produce stable majorities for single parties.

Second, a single-party minority government implies the presence of a hegemonic party that dominates the parliamentary majority-building process. Otherwise, it would aim to build a coalition government with other parties. Such a powerful position facilitates the governing party’s capacity to find support for its preferred policy output. The Swedish SAP government (1994-2006) is a case in point, especially from 1994 to 1998. Thanks to a vote share of 45.25, percent it was able to rely on the support of only one party among a diverse set of six opposition parties.

Judging from my argument, we should expect even the SAP to use its high level of government strength to downgrade union influence for dualistic policy choices in the interest of job creation. By contrast, the formation of a Liberal single-party minority government in Denmark (2015-2016) resulted from failed coalition talks with the other three parties from the center-right bloc. With a vote share of 26.70 percent, it had a very weak support base, increasing the likelihood of a labor-inclusive reform strategy (Rathgeb 2017).

Third, ideologically united majority coalitions are composed of parties from either the left or the right bloc of the parliament. A cohesive ideological outlook enhances their capacity to formulate a common reform agenda, which they are able to pass in parliament due to a majority of seats. The Austrian center-right coalition of the Christian-democratic/Conservative ÖVP and the right-wing populist FPÖ (2000-2006) on the one hand, and the first tenure of the Swedish four-party center-right ‘Alliance’ (2006-2010) on the other, reflect this type of government formation. Notably, the latter did not rely on support from the farright Sweden Democrats, which should enhance its internal unity relative to the

ÖVP-FPÖ coalition. This is because right-wing populist parties – like the FPÖ and the Sweden Democrats – might face an electoral incentive to mitigate the welfare retrenchment typically pursued by center-right parties, as they have attracted an increasing share of working-class voters in the past roughly two decades (Afonso 2015, Afonso et al. 2017).

Fourth, the government strength of ideologically united minority coalitions is contingent on the willingness of the opposition to cooperate with the government holding office. In a polarized party system, such a government would face difficulties to achieve a majority in parliament. The resulting hung parliament provides opportunities for unions to achieve policy influence, because the government requires an extra-parliamentary channel of consensus mobilization (Baccaro/Lim 2007). By contrast, in a consensual party system, such a government has a stronger autonomy in relation to organized labor (Alexiadou 2013). The Danish case demonstrates both propositions, as the frequency of minority coalitions led party leaders to establish norms of crossbloc agreements and inter-party cooperation, thereby augmenting government strength over time (Green-Pedersen 2002). Finally, ideologically divided majority coalitions refer to grand coalitions that are composed of heterogeneous policy preferences across the left-right divide. They should thus be less able to find an intra-coalitional compromise relative to pure left- or right-wing governments, because they have to reconcile a broad range of partisan interests, especially when they have an equally strong share of parliamentary seats. To overcome internal divisions, grand coalitions often delegate contested issues to trilateral policy forums and thereby draw on a less politicized channel of consensus mobilization (Hamann/Kelly 2007, Afonso 2013).

In sum, we should expect minority coalitions in polarized party systems and grand coalitions between similarly strong parties to be more prone to fall back on a labor-inclusive reform strategy than the other types of government, because the former has a low capacity to pass its preferred policy output in parliament, whereas the latter has a low capacity to formulate a common reform agenda. The presence of either of these two weaknesses creates opportunities for unions to enhance the protections of outsiders. <A> Case Selection To explain why some countries protect outsiders better than others, this book compares the reform trajectories of Austria, Denmark, and Sweden, complemented with shadow case studies on Italy and Spain. The rationale for comparing these three countries in particular is that they represent cases that are similar in important theoretical respects but differ on the empirical outcome of interest. Following a most-similar systems design, the principal objective behind this strategy is to evaluate my argument for divergent reform trajectories while

controlling for alternative explanations at the same time. The three cases I have selected are similar in three conditions that are proposed to cause outsideroriented policy choices: small size and corporatist legacies (Martin/Swank 2012), inclusive union movements (Thelen 2014), and strong social-democratic parties (Rueda 2007 vs. Pontusson 2009).

First, the three cases represent small states of Western Europe that are characterized by economic openness, relatively simple political-institutional environments, and corporatist legacies. In his seminal study, Katzenstein (1985) shows how the common perception of economic and political vulnerability in small West European states translated into an ideology of social partnership that paved the way for tripartite power sharing. Small size allowed for the cooperation between centralised elite networks that could coordinate policy adjustments more flexibly than larger states with complex and pluralist intermediation patterns. Yet, among the small states of Western Europe, the distributive outcome of democratic corporatism varied markedly. In consequence, Katzenstein distinguishes between “liberal” and “social” types of corporatism.

Austria, Denmark, and Sweden shared the core trait of social corporatism in that they had strong union movements. Unlike Austria and Denmark, however, Sweden represented a mixed type due to the presence of a strong union movement (social) as well as a strong business association (liberal). However, as of the 1990s, the three cases diverged in their patterns of interest mediation, moving Sweden and, to a lesser extent, Denmark to a pluralist direction while Austria revived the corporatist decision-making patterns of the Keynesian postwar era (Öberg et al. 2011, Afonso 2013). To take these different pathways into account, we should speak of common corporatist legacies, and not static structures. However, it is clear that small size and the presence of centralized peak level associations reflect methodologically important similarities, since these two factors are often portrayed as important conditions for the improved protection of outsiders (Martin/Swank 2012, Wilensky 2012). That is to say, the variation in reform trajectories that we observe in the three cases must be attributed to other factors than country size or corporatist legacies.

Second, and related to the first point, all three countries have inclusive union movements from an international comparative perspective. As we saw above, this inclusiveness has different sources: Austria has the highest level of union centralization, while Denmark and Sweden have the highest levels of union density in the OECD. Union centralization enhances bargaining power, reduces coordination costs, and gives greater voice to the less organized union affiliates, thereby boosting the strategic capacity of union confederations to support outsider policies. Union density, on the other hand, provides for financial, political, and organizational means to further labor’s interests. Despite their undeniable retreat in the past decades, the three union movements thus remain

relatively inclusive from an international comparative perspective. This is another important methodological similarity, given that Thelen (2014) points to the necessity of encompassing unionization for the protection of outsiders. If anything, in fact, the higher levels of inclusiveness would suggest a more “solidaristic” direction of reform in Sweden relative to the case of Austria. From this theoretical perspective, therefore, it would be impossible to attribute crossnational variation between the three cases to differences in the level of unionization.

Third, all three cases share an electorally strong and organizationally united political left in the form of social democratic parties. In Austria, the alliance between the Catholic Church and liberal elites during the historical struggles for nation building (against Protestant Prussia and affiliated Hapsburg peoples) pre-empted the possibility of a split between different factions of the working-class movement (Lipset/Rokkan 1990, 132; Bartolini 2000, 552). By contrast, the structural origin of working-class unity in Denmark and Sweden lay largely in Protestant state religiosity and cultural homogeneity (Castles 1978). As Bartolini (2000, 304-305) shows for the period between 1918 and 1985, the three cases represented the most organized and electorally successful socialist parties in Western Europe. Partisan state penetration, tight linkages to centralised union confederations, and the historical absence of intra-labor divides were the common core factors underlying this exceptional strength. Unsurprisingly, the Social Democratic parties of Austria, Denmark, and Sweden played an influential role in the neoliberal era as well. Between 1970 and 2015, the Austrian SPÖ (Sozialdemokratische Partei Österreichs) and the Swedish SAP (Sveriges socialdemokratiska arbetareparti) led the government for 39 and 28 years, respectively. The Danish Social Democrats (Socialdemokraterne) were less successful in terms of government participation (Esping-Andersen 1985), even though they provided the Prime Minister for 22 years during the same period.

Leaving aside important differences for the moment, the relative strength and unity of center-left parties forms another similarity that adds to the puzzle of divergent reform trajectories in Austria, Denmark, and Sweden. While the causal predictions derived from Social Democratic partisanship remain contested (Rueda 2007 vs. Pontusson 2009), it seems fair to say that additional factors need to be taken into account to resolve this puzzle of cross-national variation.

Taken together, the reform trajectories of Austria, Denmark, and Sweden evolved in markedly different directions, although all three cases are similar in important theoretical respects: small size and corporatist legacies, relatively strong union movements and Social Democratic parties. Against the backdrop of these similarities, the outcome of interest would have been possible in all three countries (Mahoney/Goertz 2004). Yet, only Austria expanded the protection of outsiders, whereas Sweden and, to a smaller extent, Denmark did the opposite.

This variation is definitely counter-intuitive within contemporary political economy and welfare state research. Conventional producer group and partisan explanations generally suggest continued regime variation between the Social Democratic-Nordic regime (Denmark and Sweden) versus the ConservativeContinental regime (Austria).

While Austria represents the positive case of outsider protection, Denmark and Sweden highlight how temporal variations in government strength explain differences even within the same welfare regime. In both countries, trade unions are very inclusive and build on a social-democratic institutional legacy. Yet, in crucial moments of time, they faced diverse opportunities to compensate outsiders for liberalizing pressures. In Denmark, they benefitted from the presence of weak center-right governments, which were reliant on union support for consensus mobilization. In Sweden, by contrast, they faced a hegemonic social-democratic government becoming more receptive to neo-classical economists than union demands. As a result, during the 1990s, national reform trajectories came to diverge: Unions secured â&#x20AC;&#x153;Flexicurityâ&#x20AC;? in Denmark, while being sidelined in Sweden. However, the Danish case also highlights the importance of time and within-case variation to my argument. An increase in government strength, during the 2000s, led to a decline in union influence and thus social solidarity in labor market policy â&#x20AC;&#x201C; like a decade earlier in Sweden.

The obvious problem of selecting these three cases is the one of generalizability. It could be argued that they are too similar to reveal findings that could be applicable to other cases. I therefore complement my selection with shadow case studies of two countries that are neither small in size and corporatist nor characterized by the presence of strong social-democratic parties and inclusive union movements: Italy and Spain. Although these two countries are very different from the small corporatist states of North-Western Europe, they reveal similar observations about the impact of variations in government strength on union influence. Only when governments were weak did these unions have the capacity to influence labor market reforms (Italy), whereas under conditions of strong government the outcome precluded concessions to unions and, as a result, any protections for the workforce (Spain). Yet, union influence tended to prioritize the defense of achievements for insiders, while consenting to deregulation for outsiders. This suggests that union movements in such cases are less effective in fighting for outsiders, even though they also use their influence, as long as it exists, to prevent growing divisions within the workforce. But they not only face institutional legacies and liberalizing pressures that are hostile to the protection of outsiders; without high levels of union membership and centralization they are also less able to represent them. The shadow cases of Italy and Spain therefore buttress the relevance of high levels of union inclusiveness

and suggest that variations in government strength have explanatory merit for a larger universe of cases among the rich democracies. <A> Methodology The methodological approach used here is primarily one of qualitative case study research over time and space. This research strategy – often called “process-tracing” – focuses on historical trajectories within a particular case rather than correlations of data across many cases at one point in time (George/Bennett 2005). In this way, the mechanism of policy change can be more specifically observed, since the values of the explanatory variables as well as the historical contexts in which they operate fluctuate over time. Previous theories focusing on producer group coalitions (Martin/Swank 2012, Thelen 2014) or partisanship (Rueda 2007, Pontusson 2011) outline expectations about the political process we should observe if their mechanisms are the most adequate explanations of policy change. I will compare their expectations with those derived from my ‘power-distributional interaction framework’ and confront them with the empirical observations in each case. More specifically, in Hall’s terms, “the point is to see if the multiple actions and statements of the actors at each stage of the causal process are consistent with the image of the world implied by each theory” (Hall 2008, 312). Much information would get lost if I aggregated such complex political dynamics into country-year observations on a limited number of explanatory and dependent variables.

Counterfactual analysis in the concluding sections of each empirical chapter attempts to back up the causal inference drawn from my methodological approach. In counterfactual analysis, the objective is to lend additional plausibility to the causal significance of an argument by making an explicit statement about what would have happened to the dependent variable Y if the factor X had not been present. For example, my claim is that Austrian unions had the political capacity to protect outsiders because intra-coalitional divisions made governments unable to pursue a unilateral reform strategy that excludes them. So what would have happened without the presence of weak governments?

The 2003 pension reform provides a sound factual scenario in which the government was not weak; that is, it could draw on an autonomous reform capacity generated by intra-coalitional cohesiveness. The 2003 pension reform thus reflects a useful and instructive piece of evidence on which I can build a counterfactual claim from a different point in time. To back up my claim, then, we need to find evidence showing that under the conditions of a strong government the policy output that emerges does not cater to outsiders.

To assess the explanatory merits of different theories I mainly use primary and secondary sources as well as semi-structured interviews with policy-making elites (party spokespeople, interest group representatives, bureaucrats, and

academic country experts). Following the operationalization described above, I measure government strength in terms of the size and unity of the governmentâ&#x20AC;&#x2122;s parliamentary support base. In the case of minority governments, I also look at the number of options available to build up a majority for their preferred policy in parliament. More specifically, a small number of parliamentary seats composed of parties from different party families suggest a weak government, and vice versa.

The collection of data from semi-structured interviews was necessary to understand the political processes leading to institutional change in labor market policy. I conducted 21 interviews in December 2013 and August 2014 in Copenhagen, 15 interviews in December 2014 and spring 2015 (various months) in Vienna and nine interviews in May and September 2015 in Stockholm (one interview via Skype). For each key event in the causal chain, I gathered interview evidence from at least one representative of the strongest governing party, the national peak-level employersâ&#x20AC;&#x2122; association, the national peak union confederation, and the Ministry of Labor Market Affairs. The spokespeople of the respective actors were contacted for an interview once I had sufficient information documenting that they were directly involved in the reform process.

Most of my interviewees were therefore spokespeople for labor market and social policy. When the responses of different actors were consistent with each other as well as with primary and secondary sources, I ceased to collect additional interview evidence.

My interview questions attempted to grasp the policy-making influence of different actors and the extent to which the final policy output corresponded to the initial policy demands of these actors. For instance, I asked party representatives and bureaucrats about the calculations behind the inclusion (or exclusion) of organized labor in the reform process. Moreover, I asked party and interest group representatives about the initial policy preferences prior to the reform negotiations and the strategic preferences developed in the face of political constraints; that is, second-best choices. Importantly, I adapted my questionnaire to the period in which my interviewees were official representatives of the organization of interest. Following this strategy, I attempted to enhance the reliability of the interview evidence for the historical reconstruction of the reform trajectories. I recorded almost all interviews and evaluated them through reports immediately written down afterwards. On average, the interviews lasted between 45 minutes and one hour.

The usage of interview evidence required a careful interpretive consideration of various empirical sources, given that political actors may tend to give stylized justifications of their action. Therefore, the quotes used from my interviews are the result of extensive triangulation with written documents and other interviews.

I used direct interview quotes when they seemed to illustrate the core mechanism

at work. I sent all direct quotes to the interviewees for approval in order to check the accuracy of their statements and rule out any misunderstanding. A few quotes had to be modified slightly, but none of these changes altered the meaning of the statements substantially. The list of interviewees can be found in the Appendix.

Moreover, I sought advice from experts of the countries I studied to avoid any misunderstanding and draw on existing findings (Emmerich Tálos for Austria, Jørgen Goul Andersen and Henning Jørgensen for Denmark, Johannes Lindvall and Johan Bo Davidsson for Sweden). <A> Outline of the book The plan for the book is as follows. Chapter 2 provides a quantitative overview of institutional change in labor market policy for Austria, Denmark, and Sweden from a comparative West European perspective. To assess the direction of reform in labor market policy, it relies on aggregate data on change (and stability) in employment protection legislation, unemployment protection, and active labor market policy spending. As noted previously, and in contrast to the conventional wisdom, it shows that labor market policy has become more “solidaristic” in Austria than in Sweden, with the “Flexicurity” model of Denmark coming under strain in the 2000s and early 2010s.

The following three chapters constitute the empirical core of the book and address the puzzle of variation described above. Chapter 3 is an inquiry into the reasons why Austrian political actors enhanced the social protection of outsiders despite tightened fiscal constraints (1990s), neoliberal assaults from the political right (2000-2006), and the onset of the Great Recession (post-2008). Challenging the partisan explanatory perspective, it shows that governing parties were not important players in this process. Across partisan differences, they faced difficulties in coming to issue-specific agreements and thus found it more expedient to fall back on a labor-inclusive reform strategy, which enabled unions to strike concessions for outsiders. A business-oriented explanatory perspective is also not borne out by the evidence. Employers reacted strongly against concessions for outsiders, but were unable to overcome the resistance posed by organized labor in the face of weak governments. The chapter instead highlights the strategic capacity of concentrated and centralized unions in supporting the social demands of outsiders. Despite a substantial decline in membership, the Austrian union confederation (ÖGB) not only fought hard for an extension of prevailing labor market protections to ‘atypical’ workers who previously lacked coverage; it also managed to remain influential due to the intra-coalitional divisions of Austrian governments that ruled out a unilateral reform strategy.

Chapter 4 explores why Danish political actors enhanced the protection of outsiders in the 1990s, but did the opposite in the 2000s and early 2010s. Against its ideological preferences, the center-right government of the late 1980s and

early 1990s resorted to labor-inclusive negotiations in order to overcome a hung parliament between two opposing party blocs. This way the unions secured a “Flexicurity”-deal with the political right, which was immediately implemented by the subsequent center-left government. Unlike in Austria, the weakness of Danish governments was not intra-coalitional divisions. What caused a reform deadlock was a lack of parliamentary support in a divided party system. However, subsequent minority governments gained less by involving unions, because enhanced flexibility in the parliamentary majority-building process allowed them to seek their preferred policy output independent from union consent. The onset of the Great Recession therefore allowed governments of the right as well as the left to dismiss the one single actor that mobilized political support for outsiders.

Pace partisanship theory, partisan differences were of minor importance, because center-right as well as center-left governments were in power during the rise of Flexicurity as well as its erosion. Pace producer group theory, policy deals between unions and employers lost causal influence with the dominance of parliamentary decision-making processes. Organized interests were actively involved only so long as minority governments were unable to find a parliamentary majority (late 1980s, early 1990s). When that changed, union influence declined, followed by the security-related components of Flexicurity. Unlike unions, employers did not resist unilateral reform changes; governments moved state policy in their preferred direction anyway.

Chapter 5 addresses the puzzle of why Sweden departed from universalism towards a German-like path of pronounced dualization, even though it displays the strongest social-democratic party and union movement in the world. Sweden shared with Austria and Denmark a labor movement that advocated outsideroriented policy demands. As in Denmark, the core institutional recruitment mechanism of unions in Sweden is the Ghent system, which contributed to the member-based strength of Swedish unions and thus shaped their policy preference for universal labor market protections. However, Swedish governments were neither ideologically divided (as in Austria) nor in a position of a hung parliament (as in Denmark during the late 1980s and early 1990s). They were thus the clear driver of reform at a time when the SAP came to consider moderate dualisms necessary for job creation (1990-2006), while the center-right bloc gained unity and strength to reinforce them (2006-2014). The apparent loser of this process was the labor movement, especially the LO, with its demands for outsider protection. This was in the interest of employers, which not only supported the government’s direction of reform, but also forcefully withdrew from corporatist policy-forums as a way of undermining union influence.

The final chapter of the book revisits the theoretical argument, discusses its limitations and extensions through shadow case studies of Italy and Spain, and considers the general lessons drawn from the book’s findings. It concludes with

a reflection on opportunities to re-center class power in the study of politics and the relationship between unions and governments in contemporary capitalism.