JIOS Spring Issue 2015 by Henrike Landré

Journal of

International Organizations Studies

VOLUME 6, ISSUE 1, SPRING 2015

Journal of

International Organizations Studies

VOLUME 6, ISSUE 1, SPRING 2015

Journal of International Organizations Studies (JIOS) The Journal of International Organizations Studies is the peer-reviewed journal of the United Nations Studies Association (UNSA), published in cooperation with the David M. Kennedy Center for International Studies at Brigham Young University. JIOS provides a forum for scholars who work on international organizations in a variety of disciplines. The journal aims to provide a window into the state of the art of research on international governmental organizations, supporting innovative approaches and interdisciplinary dialogue. The journal’s mission is to explore new grounds and transcend the traditional perspective of international organizations as merely the sum of its members and their policies.

Details on Submission and Review

JIOS is published twice annually, in March and September, online and print-on-demand. Submission deadline for the fall issue is 1 May each year and for the spring issue is 1 November of the previous year. JIOS publishes three types of articles: • Research papers (8,000–10,000 words, including footnotes and references) • Insider’s View (3,000–7,000 words, including footnotes and references): contributions from practitioners illuminating the inner workings of international organizations. • Reviews of literature, disciplinary approaches or panels/workshops/conferences (single book reviews, panel or workshop reviews: 800–1,200 words, multiple book or subject reviews: 2,000–3,000 words, including footnotes and references) Please send submissions to editors@journal-iostudies.org. For submissions and formatting guidelines, please see http://www.journal-iostudies.org/how-submit-your-paper. All papers will be reviewed by two or three external reviewers and then either accepted, rejected, or returned to the author(s) with the invitation to make minor corrections or revise and resubmit (medium to major changes). The final decision on acceptance of submissions rests solely with the editors.

Contact Information

UN Studies Association, c/o coconets, Hannoversche Str. 2, 10115 Berlin, Germany David M. Kennedy Center for International Studies Brigham Young University, Provo, UT 84602, USA E-mail: editors@journal-iostudies.org

Web: http://www.journal-iostudies.org

Kendall W. Stiles, Brigham Young University; John Mathiason, Cornell University REVIEW EDITOR

Nicola Contessi, McGill University INSIDER’S VIEW EDITOR

Tama Shockley, UNICEF MANAGING EDITOR

Cory Leonard, Brigham Young University EDITORIAL BOARD Christopher Ankersen, Royal Military College of Canada Jonathan Koppell, Yale University Malte Brosig, University of the Witwatersrand

Henrike Landré, DIAS/UN Studies Association

Roger Coate, Georgia College and State University

Craig Murphy, Wellesley College

Donald C.F. Daniel, Georgetown University

Slawomir Redo, UN Office on Drugs and Crime

Martin S. Edwards, Seton Hall University

Bob Reinalda, Radboud University Nijmegen

Dirk Growe, Global Policy Forum Europe

Manuela Scheuermann, University of Würzburg

Julia Harfensteller, University of Bremen

Tamara Shockley, Independent Researcher

Darren Hawkins, Brigham Young University

Courtney Smith, Seton Hall University School of Diplomacy and International Relations

Ian Hurd, Northwestern University Christer Jonsson, Lund University Adam Kamradt-Scott, University of Sydney Kent Kille, The College of Wooster Ina Klein, University of Osnabrück

Kendall Stiles, Brigham Young University Jonathan R. Strand, University of Nevada Markus Thiel, Florida International University Andrew Williams, University of St. Andrews

Table of Contents Contributors....................................................................................................................... 3 FEATURE ARTICLES On Networks, International Organizations, and Institutional Hegemony Anna Ohanyan..................................................................................................................... 5 Judicial Voting Behavior at the Appeals Chambers of the International Tribunals James Meernik, , Kimi King, Melissa McKay, and Rosa Aloisi......................................... 29 Plurilateralism: The Key to Saving the World Trade Organization and Boosting Global Trade Liberalization? Michel Dupuy..................................................................................................................... 47 REVIEWS Rick Fawn, International Organizations and International Conditionality: Making Norms Matter. New York: Palgrave Macmillan, 2013 Heidi Hardt........................................................................................................................ 63 SĂŠverine Autesserre, Peaceland: Conflict Resolution and the Everyday Politics of International Intervention. New York: Cambridge University Press, 2014 Costantino Pischedda........................................................................................................ 67 Gugliemo Verdirame, The UN and Human Rights: Who Guards the Guardians? Cambridge: Cambridge University Press, 2011 Francesco Rindi................................................................................................................. 69 David P. Auserwald and Stephen Saideman, NATO in Afghanistan: Fighting Together, Fighting Alone. Princeton: Princeton University Press, 2014 Romain Malejacq............................................................................................................... 71 Carsten Stahn and Henning Melber, eds. Peace Diplomacy, Global Justice and International Agency: Rethinking Human Security and Ethics in the Spirit of Dag HammarskjĂśld. Cambridge: Cambridge University Press, 2014 Kent J. Kille....................................................................................................................... 73

CONTRIBUTORS Anna Ohanyan is an associate professor and chair of the Department of Political Science and International Studies at Stonehill College. Ohanyan’s second and most recent book is Networked Regionalism as Conflict Management (Stanford University Press 2015). She has published articles widely on international organizations, conflict resolution, and peacebuilding in such settings as Kosovo, Bosnia and Herzegovina, and Afghanistan. She has served as a Fulbright scholar (2012–13) and a fellow at the Kennedy School of Government at Harvard University (2002–04). Her research has been supported by organizations such as IREX, the Woodrow Wilson International Center for Scholars, and the German Marshall Fund, among others. James Meernik is a professor of political science and director of the Castleberry Peace Institute at the University of North Texas. Meernik specializes in research on international justice and international courts, post-conflict peace building, and U.S. foreign policy. His research focuses on the judgments of the international criminal tribunals and the impact these decisions have on local populations. Currently, Meernik is working on projects related to victim and witness testimony at international tribunals, public opinion about international tribunals, and international human rights. Kimi King is an associate professor of political science at the University of North Texas and has been visiting The Hague, leading study abroad programs, and conducting research on the International War Crimes Tribunal for the former Yugoslavia (ICTY) for almost two decades. King is conducting a comprehensive survey to examine the short- and long-term impacts of testifying on ICTY war crimes witnesses. Her research has been published in Social Science Quarterly and International Studies Perspectives. Melissa McKay is a PhD graduate student at the University of North Texas and has focused her research on international criminal law including individual and state accountability, sentencing determinants and practices, and witness feedback. McKay has presented at the International Studies Association Annual Conference and has been a guest lecturer for internationally based organizations on the laws of war and international criminal tribunals. Rosa Aloisi is an assistant professor in the Department of Political Science at Trinity University. Aloisi’s research focuses on the work of international tribunals, human-rights institutions, the evolution of international criminal law, and the perpetration of gender crimes during conflicts. Currently, Aloisi is working on projects related to the implementation of human-rights agreements. Michel Dupuy is a professor of international economics at the University of Bordeaux (France). Dupuy’s research interests include international trade, international finance, macroeconomic aspects of international trade, and international macroeconomics. He has published many books and articles in these fields. His recent publications focus on the exchange rate effects of quantitative easing and the modes of international trade organization.

JIOS, VOL. 6, ISSUE 1, 2015

On Networks, International Organizations, and Institutional Hegemony by Anna Ohanyan, Stonehill College The autonomy of international organizations (IOs) has long been the subject of heated debates in international relations theory. IO autonomy is generally examined relative to states that create them and often in contexts of dyadic relationships. The broader yet immediate institutional environment in which IOs are embedded, composed of the IO links and ties with state and nonstate actors alike, has been insufficiently studied to date. This research seeks to fill that void. In particular, it delineates between state-centric and network approaches to understanding IO autonomy. It then examines how the immediate institutional environment of an IO conditions its ability to resist the political and often parochial pressures from donor and beneficiary states in conflict zones. Do networked IOs more effectively leverage the network power to enhance their autonomy, or do the networks constrain IO behavior? To explore this question, this study draws from the European Agency for Reconstruction in the Balkans and the microfinance programs of the World Bank in Bosnia and Herzegovina and Afghanistan.

Introduction The autonomy of international organizations (IOs) has long been considered an ultimate proxy to demonstrate the significance and agency of IOs in world politics. Autonomy of IOs is often understood as the range of potential independent action available to an IO after its principals (often understood as states) have established mechanisms of control (Hawkins et al. 2006). The wider the scope of independent action by an IO relative to states, the more autonomous and more significant in global politics the IOs are considered to be. This framing of IO autonomy has been characterized by two main features. First, the discussion has been rather state-centric. Michael Barnett and Martha Finnemore define IO autonomy as independence from states. Similarly, Yoram Z. Haftel and Alexander Thompson define autonomy as IO ability to “operate in a manner that is insulated from the influence of other political actors—especially states” (Haftel and Thompson 2006: 256). It has been viewed primarily relative to states that are politically powerful and financially significant contributors to a given IO. The influence of smaller and “weaker” states has been largely omitted (McCormick 1982). There is little understanding as to how IOs interface with governments in the developing world and with states politically marginal in the world that have insignificant financial contributions to IOs. This is a particularly problematic gap in the face of anecdotal evidence, indicating a “revolving door” between the field offices of international organizations and former governmental officials. Second, the evaluations of IO autonomy have mostly been carried out internally, often focusing on such attributes as the regulative and executive capacities of the organization, its organizational expertise, and legitimacy (Pentland 1976; Barnett and Finnemore 2004) but focusing little on how these attributes and markers of organizational autonomy are shaped by institutional environments external to the organization. Why is it that IOs can exert such regulative and executive capacities in one country and not in another? The JIOS, VOL. 6, ISSUE 1, 2015

OHANYAN

variance in cases of autonomy of a single organization across different countries and issue areas has received less attention to date than comparisons of various IOs to one another and to nation states. From the policy perspective, such state-focused understanding of IO autonomy is problematic, because IOs are increasingly operating as parts of larger systems of organizations, including nongovernmental organizations (NGOs), private consulting firms, and state agencies. Their accurate evaluation requires treating them as part of such interorganizational systems and networks, and the consensus on this point is solidifying among academics and practitioners alike (Moore, Eng and Daniel 2003; OECD 1999; Ohanyan 2008, 2007; Orsini, Morin, and Young 2013). Understanding IO embeddedness in the broader network systems in which they operate is essential for a more comprehensive analysis of international development and global governance (Atouba and Shumate 2010; Jordan, Archer, Granger, and Ordes 2001). Network-based operation of IOs (Ohanyan 2008, 2009) is a new and indispensable feature of organizational behavior necessary for IO survival and goal attainment (Atouba and Shumate 2010). The state-centric and organizational bias of existing scholarship on IOs holds us back from probing the increasingly complex role IOs play in international development, particularly as sites and frontlines of global public administration. IOs are enmeshed with various types of actors, state and non-state, and often with poorly delineated institutional boundaries between them. Importantly, the dominant state-centric frameworks of IO studies privilege earlier stages of policy development, such as problem definition and agenda setting (Rochefort and Cobb 1994; Stone 1997; Schneider and Ingram 1997). The role of IOs in global policy implementation, and the intense power struggles for autonomy at this stage, is often omitted (Ohanyan 2004). The state-centric understandings of IO autonomy, regardless of oneâ&#x20AC;&#x2122;s position on how much IOs are independent of states, obscures the organizational complexity surrounding the IO. On the contrary, the starting point is acknowledging that IOs are institutionally embedded into the networks and coalitions of nongovernmental organizations (NGOs), the national public institutions of donor and beneficiary states, private consulting firms, and other IOs. The state is one of many actors that, under certain conditions, may influence the IO autonomy.1 In addition, this approach also explores the IO autonomy in the trenches of policy implementation where the ultimate beneficiaries of IO policies reside, as opposed to the distant Western capitals where major IOs are headquartered and where only initial stages of policy development processes occur. Similar to states, IOs are organizationally complex, with some of their parts being more responsive to state needs while others are more driven by professional expertise. This study moves the organizational autonomy discussion to a level of project implementation where political pressures from donor and target states acquire particular significance, and the tensions between IO and nation-states play out in subtler ways than during project formulation and broad-agenda setting that occur in IOâ&#x20AC;&#x2122;s headquarters. The network approach to IO studies developed here maintains that IO autonomy is variable, contingent on the varied institutional possibilities of IO embeddedness in such networks. Each network manifests different patterns of authority structures and produces various kinds of environmental pressures on IO autonomy. I argue that IO autonomy is contingent on its immediate network-based institutional environment. Against this backdrop is the question of whether networked politics for IOs are assets or liabilities in terms of their organizational autonomy. Do frequently networked IOs more effectively leverage the network power to enhance their autonomy, or do the networks constrain IO behavior? 1. For instance, the IO embeddedness in networks is manifest in terms of IO centrality in the network relative to other members: its reliance on states for funding and political support and their dependence on NGOs for organizational resources in program implementation. Higher reliance on such actors and networks can be assessed in terms of the size of externally generated funding and the number of NGO contracts in a given project implementation.

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

I investigate these questions in conflict zones Afghanistan and Bosnia and Herzegovina, where IOs are major actors of public policy. Conflict and post-conflict areas are settings with heightened political sensitivities, where each penny and every policy position can become a source of instability and division. In these internationalized policy domains, multiple donors have their own national interests and specific policy priorities, which often conflict with the goals of IOs and post-conflict states. In such settings, IO autonomy is defined by its ability to resist political pressures and parochial preferences from donor and beneficiary state and non-state actors. It is also defined by the extent to which an IO exerts control over the policy process relative to the other players in a given network of which an IO is a part. The empirical material introduced in this work illustrates that some IOs are more networked than others. As a result, they are empowered, to different degrees, to manage the sometimes overwhelming political influence exerted by states. The empirical section delineates the specific tactics and strategies IOs use in withstanding political influences from states. It draws on the World Bank’s engagement of the microfinance sectors of Afghanistan and Bosnia and Herzegovina and the engagement of the European Commission’s European Agency for Reconstruction in the Balkan countries. The interviews conducted with representatives of these agencies reveal new dimensions of power as expressed through the opportunities for the occasionally hegemonic network management by some IOs. Depending on certain network conditions, some IOs enjoy greater bargaining power and opportunities to advance their goals and missions relative to other players within a given network. Such network conditions allow some IOs to evolve into institutional hegemons that tend to centralize power within a network, largely because of the centrality of their nodal position within that network. Networks are the intermediary institutions between IOs and states. In that role, they can buffer political influences from the state, in some cases insulating the IO from state pressures. IOs are embedded differently in various arrangements of network-based governance between IOs and donor states, with some IOs occupying central positions and others marginal positions in the network. Comparing the three models in the empirical section illustrates the various coping strategies IOs use to exert their autonomy in complex policy settings. The comparison shows how different structures and institutional designs external to IOs shape and filter the ability of individual players, states, or IOs to advance their narrow parochial interests relative to the rich mosaic of interests pursued by the many other players. The article will start by discussing the rationale of the network approach for IO studies. The second section will focus on the network literature as it relates to the issue of the organizational autonomy of IOs. The third section will demonstrate the network model of IO autonomy. The fourth section will present the empirical cases of the European Agency for Reconstruction in the Balkans and the World Bank’s microfinance programs in post– Dayton Accord Bosnia and Herzegovina and in Afghanistan post- 9/11 U.S. intervention. The conclusion will delineate the theoretical implications for understanding IO autonomy in world politics. Network Application of IO Autonomy: Theoretical Contours The issue of organizational autonomy has stimulated vibrant and often interdisciplinary research in the fields of organizational studies, public policy and administration, and international relations, to name a few. While the discussion in each disciplinary context differs in terms of the stakeholders involved, understanding the issue of power and control over policy outcomes is a shared concern across such disciplinary divides. Johnson (2013) highlights how the issue of autonomy in American politics centers on relationships between federal bureaucracies relative to legislators and presidents (Kiewiet and McCubbins 1991; Lewis 2003). In comparative politics, the autonomy discussions often center on the European Union

OHANYAN

(Moravscik 1999; Posner 2009; Sandholz and Zysman 1989) or the extent of governmental control in the developing world (Migdal 1988; Evans 1995) in the face of nepotistic and oligarchic pressures from the society. In the field of international relations, it is the relationship between IOs relative to states that has generated the most scholarship on the issue of autonomy. Furthermore, the extent of IO autonomy is often considered a proxy of its significance as an actor in world politics. The dominant framework of IO autonomy in international relations is state-centric, regardless of whether or not the state is considered influential in shaping IO behavior. Realist and rationalist approaches to the issue highlight the primacy of states in designing IOs and setting agendas (Abbott and Snidal 1998; Hawkins et al. 2006; Mearscheimer 1994/1995 cited in Johnson 2013), which then determines state control over policy outcomes as the argument goes. The alternative explanations from social constructivist perspectives have highlighted the role of IOs as international bureaucracies and further demonstrated the significant scope of independent action which IOs have claimed since their inception (Barnett and Finnemore 2004; Loescher 2001). While the former views IOs as instruments of national policy, with a strong bias for great powers in its analysis, the latter highlights IOs as systemic modifiers of state behavior at the least and as autonomous actors at the most (McCormick 1982). Defining IO autonomy relative to states is a shared attribute among all of these definitions, which I argue obscures the institutional complexity of the links and relationships that transcend the state, in which IOs are embedded. In this context, autonomy is defined as the extent of IO agency relative to state and non-state actors alike. Autonomy is an IOâ&#x20AC;&#x2122;s ability to advance its goals while drawing resources from these actors, shaping preferences through coercion and control and cultivating compliance in cases where the interests of the IO and other actors diverge. In this respect, this definition reflects another conceptualization, according to which IO autonomy â&#x20AC;&#x153;is not an end in itself but rather a means for reducing insecurity and enacting expert policyâ&#x20AC;? (Barnett and Coleman 2005: 598). Here autonomy is understood to be a tool for organizational survival rather than being ontologically predetermined. Principal-agent theory also has furthered our understanding concerning the conditions under which IOs may claim an autonomous space for independent action from their masters. It has pushed for greater delineation of institutional conditions under which either the states or IOs can be more assertive in advancing their goals in world politics. Haftel and Thompson (2006) refer to principal-agent scholarship that has illustrated cases when IOs as agents possess a significant degree of autonomy and discretion, and when they can engage in actions not anticipated by the states (Hawkins et al. 2006; Pollack 2003; Nielson and Tierney 2003; Majone 2001). Principal-agent theory was critical of moving the debates on IO autonomy into institutional analysis. Here, the literature acknowledged the variation in institutional design within the IO. Haftel and Thompson (2006) identified several lines of scholarships on institutional design features. In particular, the degree of formality (Lipson 1991; Abbott and Snidal 1998), hierarchy (Lake 1999), legalization (Goldstein et al. 2001) and institutionalization (Stone Sweet, Sandholtz, and Fligstein 2001; Haftel 2004b) are some of them. More concrete indicators of institutional design in IOs include the existing mechanisms of state control, such as state financing, oversight meetings, and veto players (Johnson 2013) among many others. The modeled IO autonomy builds on the existing scholarship on institutional design of IOs, but it moves the discourse into a relatively underexplored area of external, as opposed to internal, institutional design features of IOs, such as the patterns of organization and engagement IOs develops with other organizations in various cycles of policy development and project implementation. Network embeddedness of IOs as an independent variable is an insti-

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

tutional design feature, but one that tells of the external environment of the IO, developed by design by the IO, or one that has emerged by default, rather organically, and with different outcomes for IO autonomy. As indicated earlier, this approach also departs from existing scholarship in two respects. First, the dominant scholarship on IOs treats them as self-contained systems of rules (Hawkins et al. 2006) and actors with clear organizational boundaries, whereas the network models of IO autonomy are treated as open systems with malleable boundaries (Ohanyan 2012). It highlights the institutional embeddedness of IOs in their immediate interorganizational links and relationships with state and non-state actors alike. The proposed framework moves away from seeing organizations as formal and self-contained units (Jönsson 2001). Instead, this approach treats them as constituent parts of internally differentiated interorganizational systems and structures (Ness and Brechin 1988) built around interorganizational bargaining, conflict, and cooperation (Ohanyan 2008). Within the organization theory, the network approach signifies a shift away from treating organizations as closed systems to seeing them open and interacting with their environment. The second distinction of the proposed approach of IO studies rests with post-sovereign focus of analysis. It abandons state-focused applications of IO autonomy research in favor of an organizationally multilayered approach that factors in state and non-state actors as forces influencing IO autonomy. The state-centric applications of IO autonomy often view IOs in zero-sum relationships with states, and any gains in autonomy for IOs relative to states are often understood as losses for the states. Formal strategies of state control, such as managing resources, institutional oversight, or decision-making practices, are usually developed relative to states only, often discounting availability of financial and institutional resources that IOs can access from other sources. State-centric applications of IO autonomy also disregard the powerful control over final policy outcomes in the implementation stage that IOs tend to exert. Their relationship with partner NGOs implementing policies is usually the missing link in state-centric applications of IO autonomy (Ohanyan 2008, 2009). Network Model of IO Autonomy The embeddedness of the IO in its immediate interorganizational network is a feature of institutional design. It further contends that such network embeddedness is consequential in expanding and/or limiting IO autonomy in world politics. As an institutional design feature, network embeddedness of IOs is about its institutional location within the network. The IO autonomy will depend on the network conditions and the specific position that it occupies in the network relative to many network members, as opposed to a single actor (often the state). It is contingent on the structural position of the IO relative to the other members of the network. That is, the ability of IOs to exercise their authority claims is contingent on their nodal position within a given network in a certain issue area. It also follows from the network approach to IO studies that the autonomy of an IO is not fixed; it is subject to specific structures of delegation developed around a given IO. Discounting the nodal position of an IO in its immediate institutional environment leads to frameworks that struggle to factor in the multilayered environmental pressures on IO autonomy. Building on my prior research on transnational networks and global governance (2004, 2008, 2009, 2013), I define networks as interorganizational coalitions among IOs, NGOs, and bilateral aid agencies. Networks are multi-sectoral, bound together by financial resource flows and formalized relationships (Ohanyan 2003, 2009). They are structures of international public administration, because they are institutionalized over time through repetitive and enduring transactions (DiMaggio and Powell 1991). Joel M. Podolny and Karen L. Page (1998) define a network as “any collection of actors (N≥2) that pursue repeated, enduring exchange relations with one another and, at the same time, lack a legitimate organizational authority to arbitrate and resolve disputes that may arise during the exchange.” Networks may be distinct players

OHANYAN

in world politics in their own right (Kahler 2009), because they influence the behavior of their members, whether they be NGOs (Ohanyan 2008, 2009) or intergovernmental organizations (Hafner-Burton 2006). Network embeddedness of IO as an institutional design feature is conceptualized in terms of two variables: centrality of IO position in its immediate interorganizational network and centralization of its relationship with other network members. While centrality describes the institutional location of an IO within the network, centralization describes the extent to which an IO centralizes its relationships with network members. Centrality as a variable is pre-determined for the IO, which is largely a result of often unpredictable forces that produce a particular network. In contrast, centralization as a variable is an administrative feature, informing the dominant patterns and practices of organizational performance by the IO relative to its network partners. In particular, centrality is a network-level index that captures the extent to which the network has one central and many peripheral nodes (Wasserman and Faust 1994). Borrowed from the organizational network literature, it measures the number and strength of ties that an organization has with other organizations in the network (Moore, Eng, and Daniel 2003). Resource dependency perspective has maintained a strong positive association between an organization’s level of centrality in a network and its influence in “community affairs” of the network (Galaskiewicz 1985). Others hypothesized that higher centrality represents more enhanced “communicative activity,” which positions them better with regard to the flow of information and coordination of their activities (Moore, Eng, Daniel, 2003). Higher centrality2 essentially signifies a more hierarchical network dominated by one or two major players. IOs with higher levels of centrality within a network have the most links to most members in the network. Such links and relationships between the given IO and other network members are created, cultivated, and cemented by financial and institutional resources flowing between them. As a variable describing the structural position of IO in the network, variance in centrality produces different scenarios, ranging from a given IO being the key provider of resources in the network to cases in which several IOs/donors are providing needed resources in the network. In the latter case, the network is less hierarchal, because it houses several major players, of which a given IO is only one. This, therefore, reduces the centrality of the IO, undermining its prospects of autonomous behavior in a given network. In this context, centrality of an IO in the network is determined by several indicators: • The extent to which a particular IO provides the bulk of the financial resources in a given network. • The extent to which it maintains an institutional presence in the area of program operation. • The extent to which it serves as a hub of expertise and knowledge development within the network. • Whether or not the sudden withdrawal of the given IO would produce a collapse or disintegration of the network as a whole. As I argue, this structural feature of IO embeddedness in its network, while important, on its own has limited explanatory power for IO autonomy—the dependent variable. Foreign-aid industry is littered with cases of large IOs with significant financial resources and numerous ties with implementing partners being pulled different directions by these very same partners (which are often NGOs or state agencies in target countries). Centrality, while describing the structural position (central or marginal) of the IO in the network, provides only a partial explanation for IO autonomy. Only when combined with centrality, the second feature of network embeddedness of IO, does a fuller picture transpires. Centralization is an attribute of organizational design in which the organization gathers and reasserts its authority and decision making over resource distribution, strategic and long2. “Centrality” and “centralization” are qualitatively different. Centrality refers to the network position of an organization, while centralization refers to the capacity and willingness of a single network member to centralize the network operation in a given issue area.

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

term policy development, and implementation, leaving only tactical and day-to-day managerial issues to its partners in the network. It is a commonly used variable, often found in wide-ranging fields of study, such as business administration, organizational management, and public administration. Centralization in a network is measured by several indicators, including but not limited to the following: • Tight administrative instruments of controlling the implementing partners • Frequent monitoring • In-house evaluation tools • Centralized decision making and decentralized implementation • Centralization of expertise Centralization is often confused with centrality and used interchangeably (Orsini, Morin, and Young 2013). It is often perceived that those organizations that enjoy a high level of centrality in a network will also be able to centralize their operations and that only autonomous organizations that enjoy a position of centrality may centralize their operations in a network. True, IOs that enjoy greater centrality within the network are more powerful, are more capable of resisting political influences, and have more chances to shield the professionalization of their policies. And indeed, autonomous organizations are better positioned to centralize their operations. What comes first, centralization/centrality or autonomy? The practice of IOs creating networks to achieve their goals in international policy systems, which is an important measure of organizational autonomy, is frequently observed. These are cases of network development by design. Often, large and powerful organizations strategically deploy large networks with NGOs and other types of agencies, and they may dominate these networks and centralize their operations due to their sheer power and resource endowments. In network development by design, there is extensive and systematic coordination among the members of a given network, IOs, NGOs, and state agencies alike. They are characterized by close coordination among donors as they determine the type of implementing actors to work with. In such cases, an IO has independent effects on network by formation. In such instances, it is the organizational power and autonomy that drive the patterns of network formation and ultimately shape the institutional features of the network. Focus turns to a second, and much more common, phenomenon of network formation by default. These are cases when network formation occurs rather sporadically, often driven by forces of organizational inertia and opportunistic forces of convenience. This can occur because of partners in a country where incompatibility of organizational cultures between network members can complicate deep coordination. In cases of network formation by default, no serious coordination exists, and the institutional features of a given network either constrain or empower the IO operation. Importantly, the institutional attributes of the network can craft the contours of IO autonomy. The structural position of an IO in the network (i.e., the levels of its centrality) by itself fails to predict the extent of IO autonomy. Some IOs will take advantage of that position, but there are many who do not. Instead, centrality as a dimension of IO embeddedness in its network needs to be supplemented with the feature of centralization. When an IO that already enjoys a favorable structural position of centrality in the network moves to also centralize its operations and engagements with its partners in the network, its autonomy and ability to resist outside pressures is significantly enhanced. It is the combination of both features that can serve to predict the IO autonomy in a given network. The network model of IO autonomy presented here maintains that IOs are more powerful in terms of advancing their goals relative to other players in world politics when they enjoy a favorable structural position of centrality and opt to operate through centralized networks. The variation of centrality and centralization of IO embeddedness in its network produces four possible institutional outcomes, presented in Table 1. Table 1 presents four institutional scenarios of IO autonomy. The “Marginal Power” scenario is an outcome of unfavorable structural position of low centrality of the IO in the net-

OHANYAN

Table 1: Institutional Scenarios of IO Autonomy

Low

Marginal Power World Bank-BiH

Hollow Shell

Institutional Hegemony EAR-Balkans; World Bank-Afghanistan

Unclaimed Power

Low

High

Centrality Structural position due to IO resource endowments

Centralization Centralization of its administrative practices by the IO

work, but in which the IO choses to centralize its operations relative to its network members. In this case, the IO is proactive but with little structural opportunities to exercise full autonomy over its programs and their implementation. The next case of “Unclaimed Power” shows when an IO fails to take advantage of the structural opportunities in the network it enjoys. This scenario captures cases when large IOs with significant financial and institutional resources in a given country fails to exercise control over its operations. Often in such networks, the IO’s power is diluted by other players in the network and often driven into unplanned and unforeseen directions by their partner NGOs and other donors. The “Hollow Shell” indicates cases in which an IO finds itself part of a larger network, in which it neither provides significant resources nor takes ownership of its projects. Such outcomes tend to develop in networks built around highly technical development issues, such as microfinance in post-war settings, where IO donors join the network by providing financial resources but cannot exert any control over the larger program, because they lack the internal expertise. The fourth scenario of “Institutional Hegemony” happens if the IO enjoys a favorable structural condition of high centrality in the network and selects to centralize its administrative and management of programs in a particular country. In each case, the opportunities and prospects for IO autonomy are different, discussed later in this section. IO autonomy, the dependent variable, is defined as the ability of an IO to advance its organizational goals relative to the politically conceived national priorities of the donors that fund them and the organizational goals of NGOs that implement IO policies. IO autonomy may vary in terms of the “politics-expertise” nexus. IOs usually possess expert authority (Barnett and Finnemore 2004), and they seek to advance their detailed, specialized knowledge about specific tasks (Barnett and Coleman 2005). The more autonomous IOs can withstand narrow and parochial political pressures coming from different stakeholders as they assert their technical expertise and further policies around it. While state-aid agencies are also endowed with “knowledge resources” and possess significant “expert authority,” they are more driven by the national interests and foreign-policy considerations of the states they represent. The IO auton-

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

omy in conflict regions is often manifested by its ability to advance professionalizing policies and shield these policies from political influences from donor and recipient states. Autonomy is assessed in terms of an IO’s ability to develop and implement its own goals, preferences, and mandates (Helfer 2006), regardless of the goal of their donors and implementing partners (Ohanyan 2009, 2012). The autonomy of an IO in each institutional scenario is quite variable. The empirical section below presents three cases for two out of the four possible institutional scenarios of IO autonomy. In the “Hollow Shell” scenario, the IO lacks favorable structural conditions to exercise autonomy as well as the political will to do so. In the “Unclaimed Power” scenario, the conditions for IO autonomy are favorable, but it is the political will to claim that power that results in limited IO autonomy. The “Marginal Power” includes cases when the structural conditions are unfavorable for IO autonomy, but the IO possesses the needed political will to claim as much power as possible, thereby exerting some control over its policies and preferences and exercising a certain level of autonomy in doing do. The condition of “Institutional Hegemony” includes cases when favorable structural conditions intersect with strong IO will to exercise power over its preferences. It is in such cases, that IO exerts the most autonomy from other stakeholders and members in the network. Empirical Findings The fieldwork for this study was conducted in the Balkans (2001–03), in Brussels, Belgium (2006, 2008), Thessaloniki, Greece (2008), and Washington, D.C. (2008). It draws on more than fifty structured interviews conducted with officials from U.S. bilateral aid agencies and policymakers from the World Bank, European Commission, European Agency for Reconstruction (EAR), and numerous transnational NGOs and their branch offices in the Balkan region and Afghanistan. This comparative qualitative study is focused on two organizations and their immediate interorganizational networks in post-conflict regions—EAR in the Balkans, which was created by the European Commission, and the World Bank in Bosnia and Herzegovina and Afghanistan. The cases imperfectly reflect the possible diversity of institutional scenarios with regard to IO autonomy, as reported in Table 1. These two organizations allow us to learn from three cases: 1) EAR in the Balkans, 2) the World Bank in Bosnia and Herzegovina and in Afghanistan. In brief, EAR’s presence in the Balkans and the World Bank’s microfinance projects in Afghanistan are representative of institutional hegemony, while the World Bank’s presence in Bosnia and Herzegovina is best captured by the “Marginal Power” scenario. Due to certain logistical difficulties of the field work, the case selection and the empirical section are only meant to provide the general contours of the network approach to IO autonomy. The development of new cases is necessary in order to further understanding of network influences on IO autonomy. This is one step in that direction. Both EAR-Balkans and World Bank-Afghanistan operated in networks in which they enjoyed high levels of centrality and both proceeded to centralize their operations. This created the condition of institutional hegemony discussed later in this section. World Bank-BiH operated in a network of increasingly declining centrality, albeit with persistent centralization by the World Bank of its operation—an attribute of marginal power. Despite ranges of differences in their internal organizational cultures, scopes of mandates, and membership rules, among other characteristics, EU/EAR and the World Bank are highly comparable. Specifically, both the World Bank and EAR have an organizational locus that supports a particular program-based intervention. And, regardless of their differences, both organizations condense their bureaucratic authority into an agency, and it is this locus of authority and administration studied in this article. Network-based operation of IOs is programbased. The same organization, such as the World Bank, can be involved in numerous separate and distinct networks around many issue areas and programs it maintains around the world.

OHANYAN

Therefore, the network formation in processes of global policy formulation occurs around a specific organizational site within an IO. In this respect, EAR is the organizational site within the EU that has been in charge of program implementation in the Balkan area, and as such, it is comparable to the World Bank, which is more distinct in its organizational boundaries than the EU. EAR in the Balkans Since its creation by the European Commission, EAR has been a central node within its broader interorganizational network that was composed of the European Commission, the European Union (EU) delegations in the Balkan countries, and numerous NGOs and private consulting firms with whom it worked to implement its projects in the region (Figure 1). Over the span of nearly a decade, EAR expanded its authority within the network and emerged as an institutional hegemon. In the aftermath of the war in the Balkans, the European Commission, in its push and pull with its member states, carved out significant institutional space for itself in this network, connecting the other actors in the network. The creation of the Rapid Response Mechanism in 2001â&#x20AC;&#x201C;02 was the first step to that end, a step driven by the realization that the organizational complexity of the EU was a major stumbling block in its responses to the new crises in the post-communist space (Byrne 2008; Cartopassi 2008; Mustard 2008). Figure 1. EAR Network Structure in the Balkans

European Commision

European Parliament

Donor Governments

EAR Thessaloniki HQ Implementing Agencies (NGOs and private consulting firms)

EAR field offices

National Governments in the Balkans

By means of the new mechanism, the European Commission could bypass the lengthy negotiation and consensus-building processes normally required within the EU and expedite its capacities in conflict response. But the Rapid Response Mechanism had a mandate of only six months, and so in 2006 it was replaced by the Instrument for Stability, which had a more expanded mandate of eighteen months. Adopting this instrument was essentially the second step in consolidating EARâ&#x20AC;&#x2122;s centrality in its network. The creation of EAR has been the most explicit step taken by the European Council to enhance the conflict-management capacities of the European Commission. Placed outside the Euro-

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

pean Commission’s governance structures, it was more hierarchical and centralized in an organizational sense than the usual European Commission operations deployed elsewhere (Andreoni Dicorrado 2008). Certain political pressures in the Balkans contributed to the high centrality enjoyed by EAR in its network. The creation of EAR was associated with the end of the conflict in Kosovo and the mounting socioeconomic and political needs in the region. Immediately after the end of the war in Kosovo, the European Council announced the EU’s intention to assume a leadership role in the reconstruction of Kosovo. Subsequently, the European Commission took responsibility for setting up an agency to implement the European community’s reconstruction programs (Temcheff and Caputo 2004). EAR, a new European institution, was largely beyond the existing and conventional channels of accountability (Cartopassi 2008) within the European Commission. EAR was not an institutional approach by the European Commission, meaning that it was an ad hoc response to a single conflict region. This relatively light and flexible structure was intended to “speed things up” and rapidly translate the aid commitments into projects (Cartopassi 2008). In terms of the specific indicators of centrality discussed earlier, the case of EAR in the Balkans is that of high centrality. It has emerged as the organization that provided the bulk of the financial resources in its network, largely channeling the European Commission funding allocated for post-war reconstruction in the region. The agency maintained a field presence in the region, which allowed its staff to be in constant communication with their partners throughout project implementation. The Rapid Response Mechanism helped EAR to maintain a field presence but also to bypass EU bureaucracy and become a key institutional hub for expertise and knowledge development in the network through creating an in-house evaluation unit and relying on professionals with deep expertise in their respective fields. Importantly, EAR’s centrality in the network is also supported by the resource dependency theory application: if EAR chose to withdraw from its network, the network would be unable to survive and sustain itself. EAR supplemented its highly favorable structural position of centrality in the network with increased levels of centralization of its operation, which was mainly achieved due to the urgency of the situation and the need to implement the humanitarian programs swiftly. And EAR represented the agency model of aid delivery (Temcheff and Caputo 2004).3 However, the centralization of EAR’s administrative practices, which was achieved through several pathways, should not be taken for granted. Among the first and most obvious ways that the EAR centralized its operations related to the centralization of responsibilities of the director, which entailed tighter administrative control over program implementation by senior level managers. In addition, the program managers and task managers inside the organization exercised frequent monitoring of programs, while also having greater say over various components of the policy implementation as carried out by their partners. The strategic choice of hiring professionals with at least thirty years of expertise was a way to centralize expert authority and maintain leverage over contents of policies and projects in the implemented stage. This also allowed the agency to carry out policy evaluations, maintaining accountability over their implementing partners. Also important in terms of governance structures EAR created was its ability to centralize a range of functions and many cycles of project management, which in international development are generally outsourced to NGOs and other non-state actors. According to Temcheff and Caputo: The “Agency Model” is characterized by the “indirect centralized management” model in which the Agency is fully and solely responsible for the full cycle of the project management (drawing up Terms of Reference, preparing and evaluating invitations to tender, 3. The EAR was created within the framework of Council Regulation 2666/2000, Community Assistance for Reconstruction, Development and Stabilisation (CARDS).

OHANYAN

signing contracts, concluding financial agreements, awarding contracts, evaluating programmes, checking implementation and effecting payments) for delivering aid. (Temcheff and Caputo 2004: 47) EAR powers over policy formulation and agenda setting, enabled by greater centralization of operations, were also formidable. EARâ&#x20AC;&#x2122;s work in budget allocations, assigning priority between sectors and regions inside the entity, was guided by the Country Strategic Papers and Multi-Annual Indicative Programs, which served as a basis for EARâ&#x20AC;&#x2122;s own Action Programs. In this work, EAR consulted with the European Commission, the local member-state representatives, international financial institutions, bilateral donor representatives, and representatives from the beneficiary governments. Respondents from the agency argued that EAR coordinated with and listened to the beneficiary governments (Almqvist 2008), and independent evaluators charged that the centralization of the policy process within EAR was not sufficiently inclusive of the local actors, which generated little local ownership and potentially alienating the local players. Frequently monitoring implementing partners, a measure of centralization, was possible due to the presence of operational centers of EAR in each country of operation. Interestingly, representatives of the member states of the EU constituted the governing board of EAR, and as such, the national interests of the European states were represented within EAR, and EAR also coordinated with the bilateral aid agencies of the member states. EAR further centralized its authority and expanded the scope of its autonomous action by concentrating its implementation in operational centers located in each area where it was active. This was an arrangement of centralized decision making and decentralized implementation. EAR was headquartered in Thessaloniki, Greece, but most of its human and organizational resources were located in the field (i.e., in countries of operation). Such a local presence allowed the agency to monitor the programs closely and promptly translate feedback into policies, facilitating organizational learning (Almqvist 2008) and, most importantly, removing and shielding its policy implementations for the political reach of the donors that created it.4 In policy terms, by means of the agency model, EAR transcended the regular bureaucratic machinery of the European Commission and brought that machinery closer to the region. The organization was more responsive to the local conflict dynamics and the day-to-day operations of state building in various areas of public administration and policy issue domains, ranging from agriculture and education to training judges. Establishing an in-house evaluation unit was one of the major instruments allowing further centralization of EAR authority within the network relative to the donor and recipient states. Interestingly, EAR created its own internal evaluation unit, although it was simultaneously subject to evaluations and audits by the European Court of Auditors and frequent visits from the members of the European Parliament. The highly decentralized (relative to the European Commissionâ&#x20AC;&#x2122;s evaluation systems) in-house evaluation unit enabled EAR to integrate its work in project implementation. The director of the Evaluation Unit at EAR noted: We covered the whole area of our work, not only in the centers. We learned the environment in which we are doing our programs. When we are evaluating the results of the projects this is not possible to do from Brussels, even if you send someone to do it. This means we could adjust the development of the agency, and we could adjust the change in situations in our evaluation. So being close in implementation was one of our advantages. Also, we had a short timeline in evaluations (shorter feedback loop), so we could improve the programs quickly based on our evaluation. (Almqvist 2008) Having an in-house evaluation unit based in the region added tremendous value to EAR as a knowledge-driven, expert-led institution. The in-house evaluation unit enabled EAR to effec4. When EAR was established, its mandate was expanded to include Serbia and Montenegro and, in 2002, the former Yugoslav Republic of Macedonia, each of which was represented with a separate operational center.

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

tively deflect political pressures from donor and beneficiary states, an effort that was helped by the relative insulation of the agency from political considerations.5 Such high-level centralization and centrality of EAR in its network produced a significant scope of autonomy for the organization. Autonomy, defined as an ability to resist political influences from donors and target governments while advancing expert-driven policies, was significantly boosted by the particular features of the institutional design of the network in which EAR was a key component. Within the European Commission-EU delegations-EAR triangle, there has been a clear division of labor: EU delegations were in charge of the political dialogue with the beneficiary-country governments, and EAR was focused on program management. In other territories in which the EU was delivering foreign assistance, EU delegations would execute the assistance while also dealing with the political dialogue with the beneficiary country. In such cases, the EU delegation would maintain political relations while playing the donor role (Simmross 2008). Such coupling of public administration and politics often negatively affects the former and could compromise the organizational autonomy of EAR if it failed to centralize its operations and maintain a high measure of centrality in its network. EAR, operating in parallel with the EU delegations, was not set up to be a political interlocutor between the EU and the local government (Mustard 2008). EAR’s engagement with the governmental institutions was limited purely to project implementation and focused primarily on fiscal relations. The ability of EAR to consolidate its superior network positions allowed the organization to shield itself from political pressures, to enjoy organizational autonomy, and to become responsive to epistemic communities and expertise on the ground. All of the respondents interviewed inside the agency noted this as an advantage, pointing out that separating the carrying out the political dialogue and the delivering aid enhanced the agency’s effectiveness. Upon closing EAR in December 2008, all of its activities were handed over to EU delegations, which had been in charge of the political dialogue and now had to combine both tasks.6 Some within EAR ascribed the agency’s success to its relative insulation from politics, and they worried that combining both tasks for EU delegations would undercut the latter (Simmross 2008). With the new arrangement, the head of the delegation would see its governance capacities strained by political considerations. The potential for sacrificing policy administration for political purposes was the core concern of many respondents. The former arrangement allowed EAR to expand its autonomy over the policy process and strengthen its advancement of expert authority by shielding this policy process from political pressures coming from various sides. The autonomy of EAR was further enhanced by concentrating expert authority inside the organization and strengthening the control over the implementing partners. The agency hired as project managers professionals with more than thirty years of experience, which helped the agency to maintain significant control over program implementation even when outside consultants and contractors were hired for a particular project. One of the respondents from EAR noted: By hiring experienced project managers, we did not put out our house to the consultants. When it came to [the] operational side, we selected managers with long experience (water engineers, for example) to be staff members, people with SME (small and medium enterprise background, civil society background. We made them program managers. We tried to assemble operational teams that reflected the subject at hand. (Mustard 2008) 5. The transition marked by EAR’s closure at the end of 2008, when it handed its operations over to the EU delegations in each country, had many ramifications for the issue of political influences on policymaking. Once program administration is handed over to the EU delegations, policymaking becomes vulnerable to political influences, because the institutional space for experts and professional program managers becomes limited. For one thing, EU delegations do not intend to continue EAR policies of hiring professionals with at least thirty years of experience; instead, they will opt for staff with at least two years of experience. In 2008, an anonymous respondent from EAR expressed his worry that “when the Delegation takes over, we will not be able to deliver as fast and effective as before. This is a political issue, and also [a] bureaucratic issue: the administration within the Commission is of course favoring its own system. They even see us [EAR] as unfair competitors, and sometimes they are not happy when we can show better results. It seems that we expose their deficiencies.” 6. EAR mandate has been temporary from the start, and its closure, which has been postponed several times, should not be understood as a sign of organizational weakness.

OHANYAN

By hiring experts as project managers, the agency avoided principal-agent problems in which the delegation of responsibilities to consultants or contractors generally translated into loss of control over the project. Instead, using experts and specialized professionals as in-house project managers enhanced the agency’s ability to monitor program implementers, in spite of the agency using other organizations to implement programs implementation. In short, the types of networks the agency used were much more centralized. From Institutional Hegemony to Marginal Power: The World Bank in BIH In the first few years of introducing microfinance to BiH, the World Bank enjoyed higher levels of centrality in networks than it mobilized with NGOs, but as new donors for microfinance NGOs emerged in the country, the level of centrality of BiH declined over the years. The World Bank provided more than twenty NGOs with seed capital for their microfinance programs after the 1995 Dayton Accords. Within two years, the World Bank consolidated the sector by terminating funding lines to financially unsustainable NGOs, which eventually merged with the financially successful NGOs that continued to enjoy the World Bank’s support (World Bank 2000). Overall, at the sectoral/country level, the World Bank centrality in microfinance networks in BiH were evidenced by the fact that in the mid-1990s, the World Bank provided the bulk of the financial resources to microfinance NGOs and was the main supporter of microfinance networks. Its centrality was cemented in these networks by its field presence, which allowed for frequent communication with its implementing NGO partners in the networks. Similarly consequential was the fact that it possessed internal expertise, which was shared with NGO partners in networks, elevating the World Bank’s central position in the network. Importantly, its sudden withdrawal from the microfinance networks with NGOs would challenge the longterm survival of the microfinance networks. While its centrality in networks has declined over the years, the World Bank has always maintained a high level of centralization of its operations in microfinance NGO networks. In all of these, the World Bank has used tight administrative instruments in managing relations with partners. The respondents from NGOs with which the World Bank worked would often complain of frequent and stringent reporting requirements imposed by the bank. Similar to EAR in the Balkans, the World Bank has maintained an in-house evaluation system, which was used in its decision-making and policy-implementation processes. In-house evaluation systems also allowed the bank to centralize its expertise and become less dependent on outside evaluation systems common in international development. Although the World Bank proceeded to centralize its operations with the microfinance NGOs in BiH, its centrality was still more limited relative to EAR network conditions in the region. The World Bank’s position in the Bosnian microfinance sector was less central, primarily because the NGO partners in the networks were experts in significant microfinance and over time developed new partnerships with others donors. Both of these factors affected the World Bank’s centrality, and limited its scope of autonomy at the sectoral level, as well as within particular NGO networks. The World Bank could then only moderately advance expert authority and professionalization of policies while withstanding political influences (which are measures of its autonomy). Some NGOs through which the World Bank disbursed microfinance funding would give in to parochial pressures from other donors and try to meet the conditions of the latter donor group, often deviating from the explicit emphasis on commercialization as advocated by the World Bank. Unlike EAR, the World Bank did not capitalize on its early institutional centrality in microfinance networks in BiH. As the sector matured and the structural position of the World Bank became compromised with more donors emerging and diluting World Bank centrality, the organization became vulnerable to the vagaries of horizontal-decentralized network politics, and its autonomy suffered. Contrasting this sector-level behavior of the World Bank with four specific NGO networks in which the World Bank operated reveals a more detailed picture of the overall nar-

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

rative on IO autonomy (Figure 2). The microfinance policies in the four NGO–IO donor networks under consideration were being implemented by Partner/Mercy Corps, Mikrofin/ CARE, the Mi-BOSPO/Norwegian Refugee Council, and EKI/World Vision. In each case, the single implementing NGO was financially supported not only by the World Bank but also by many other bilateral donors and other NGOs such as the Office of the United Nations High Commissioner for Refugees (UNHCR), the United Nations Development Programme, the U.S. Agency for International Development (USAID), the Swedish International Development Agency, and the United Methodist Committee on Relief (UMCOR). Each of these players had different preferences for and visions about the use of microfinance in postwar BiH. UNHCR favored refugees and internally displaced persons, UMCOR had particular geographical preferences, and the World Bank focused on local residents to achieve financial sustainability of microfinance institutions at a faster pace and with less risk in the process. The World Bank was most autonomous in advancing its policies and preferences in the Mikrofin and Mi-BOSPO networks. This is evidenced by stronger financial sustainability of these NGOs than their social performance compared to that of the other NGOs operating in the field (Ohanyan 2009), and this was more consistent with the policy preferences and objectives of the World Bank (Tinjič 2001; Forster 2002; Campara 2001). It is important to note that in both of these networks the World Bank enjoyed superior institutional hegemony, because neither Mikcrofin nor Mi-BOSPO had many donors, and the World Bank provided the disproportionately larger share of financial resources and institutional support throughout the implementation of the microfinance program—important attributes of the World Bank centrality in both networks. In both cases, the World Bank’s power was not offset by other, more formidable donors, reflecting a higher level of Bank centrality in the network that allowed the World Bank to advance its preferences with relative ease. This was not true, however, for the Partner/Mercy Corps network, in which the NGO partner managed to advance both financial and social performance, because it enjoyed a broader donor base and used a variety of strategies to appease all donors while asserting its own organizational goals. Figure 2. World Bank in Microfinance Sector of Bosnia and Herzegovina

Bospo/MiBospo

Mercy Corps/ Partner

World Vision/EKI

World Bank

CARE/ Mikrofin

The World Bank was the only supporter of Mi-BOSPO during its first two years (1996– 98). In the third year, UNHCR also began to support Mi-BOSPO. Church World Services and Women World Banking came on board during the fourth year of operation, and UMCOR

OHANYAN

during the seventh year. Despite the infusion of new funding, the World Bank was still providing more than three-quarters of the total funding for this network during that period, reflecting its centrality in this network. Moreover, in these favorable network conditions the World Bank proceeded to centralize its administration by maintaining frequent reporting requirements from the NGO and highly formalized relationships. This early institutional hegemony within this particular network was subsequently offset over the years by the new and equally institutionalized accountability channels and frequent monitoring from the other donors that joined this network. Nevertheless, despite a modest decline in its provision of resources in the network, the World Bank still maintained its hegemonic position relative to the network’s other donors. It then managed to deflect political pressures and targeting requirements from other national and international donors, particularly from UNHCR. Similarly, in the Mikrofin/CARE network over the period 1997–2001, the World Bank was the only donor in the first year, but in each year that followed it was joined by UNHCR, the U.S. Department of Agriculture, and Ireland Aid. This network was smaller than the Partner/Mercy Corps network, which enjoyed the support of more than seven donors. Although the share of financial contributions of the World Bank declined over the years, eventually reaching only one-third, its institutional support was still significant; the bank’s monitoring and evaluation mechanisms were more institutionalized, centralized, and frequent than those of the other donors. Because of these supports, the bank managed to maintain its institutional hegemony within the network. Similar to its role in the Mi-BOSPO network, the World Bank successfully eclipsed the targeting priorities of other, more socially oriented donors, such as UNHCR and Ireland Aid. An independent evaluation of UNHCR’s microfinance activities in postwar BiH also indicates that Mikrofin and Mi-BOSPO were least effective in advancing UNHCR’s goals of targeting specific social groups (Klinkhamer 2001), which reflects the World Bank’s ability to advance its goals of commercialization in microfinance in these networks. Institutional Hegemony: The World Bank in Afghanistan Figure 3. World Bank in the Microfinance Sector of Afghanistan

Implementing NGO Canadian International Development Agency Implementing NGO Swedish International Development Cooperation Agency Implementing NGO

World Bank/MISFA British Department for International Development

Implementing NGO Oxfam Novib, Netherlands Implementing NGO

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

In Afghanistan, in close cooperation with other major American and European bilateral donors and multilateral institutions, the World Bank helped to create a multi-donor trust fund, Microfinance Investment Support Facility for Afghanistan (MISFA) (Figure 3). All of the implementing NGOs receive their funding from MISFA rather than from their own independent donors (Arbab 2008). This centralization of donor networks into a single organization, often described as an “apex institution,” has been a powerful boost for the World Bank in deflecting political influences from individual donors or the government agencies and politicians within Afghanistan. In this network environment of institutional hegemony, the World Bank has enjoyed significantly more autonomy than it did in the Bosnian microfinance sector. This particular arrangement of institutional hegemony in Afghanistan created an organizational barrier between the NGOs and the community of donors, with MISFA acting as the intermediary agency. This arrangement reflected the World Bank’s centrality in the network as well as the World Bank’s successful efforts in centralization of microfinance programs in the country. From the beginning, MISFA committed its partner NGOs to ensuring strict financial discipline in its operations and to achieving operational sustainability by its fifth year of operations (Arbab 2008)—key goals of the World Bank in microfinance sectors around the world. In this arrangement, the World Bank and its affiliate Consultative Group to Assist the Poor (CGAP) channeled donor funding earmarked for microfinance from agencies such as the Canadian International Development Agency, the British Department for International Development, the Swedish International Development Cooperation Agency, USAID, and Oxfam Novib of the Netherlands (Figure 3). Most importantly, by creating this apex institution and centralizing the networks with NGOs and donors, the World Bank exhibited enhanced autonomy as it advanced its principles and priorities in the financial sustainability of microfinance institutions. In doing so, it has successfully deflected the more parochial pressures for targeting specific social groups that individual national donors would normally target. In this case, such pressures were coming from members of Parliament to direct microfinance funding into their respective regions, even if the geographical conditions made it nearly impossible to start effective microfinance programs, as noted by one of the consultants to this projects. This institutional approach to microfinance (Barcus 2007; Hashemi 2007) clearly established the institutional hegemony of the World Bank in the microfinance sector of Afghanistan. Moreover, this arrangement was more open to epistemic communities, in this case microfinance economists and independent consultants, some of whom were advising the World Bank in BiH and were then sought out to help create the microfinance sector in Afghanistan. Theoretical Implications: The Politics of Institutional Hegemony of IOS Amandine Orsini, Jean-Frédéric Morin, and Oran Young (2013) emphasize that the rapid growth in the number of international institutions often creates tensions, conflicts, and adjustments between the regime complexes these institutions represent. Researching variance of IO embeddedness in its network, an emerging feature of institutional design, is one step closer toward understanding how these IOs with conflicting goals on the one hand, and interorganizational dependencies among them on the other, cohabitate in the infrastructure of global governance. In particular, research shows the push and pull involved with IOs seeking to exert their autonomy from the very same network members, which are often provide various resources for the IO. The nature of IO embededdness in its network can shape the scope of IO autonomy from these network members. The four institutional scenarios of IO autonomy (Table 1) reveal a variance of network conditions in which IOs operate, providing a more nuanced understanding of global policy implementation processes, of which IO are central components. Particular theoretical implications include the following: 1) network effects on IO autonomy, 2) delineation of institutional hegemons from other IOs, and 3) the mechanisms by which IOs translate their structural power (i.e., superior nodal position in the network) into political might. The three empirical cases—EAR, World Bank-Afghanistan, and World Bank-BiH—

OHANYAN

described here indicate that the autonomy of IOs, as expressed through their ability to withstand political pressures and advance their organizational goals and purposes, is contingent on their immediate institutional environment—that is, the network-based structures of delegation in which they are embedded. In respect to the effects of network environments on IO performance, the cases examined here are telling. EAR and World Bank-Afghanistan cases are examples of IOs that reflect “institutional hegemony” as a type of their network embeddedness; they occupied a position of centrality in their respective networks and proceeded to centralize their relationships with other actors in the network. In contrast, the World Bank-BiH case is an example of an organization that enjoyed lesser levels of centrality, finding itself more vulnerable to the fluctuations and uncertainties of more decentralized network politics. This was a case of “marginal power” as a type of its network embeddedness. This study lacks cases of “hollow shell” and “unclaimed power” as types of IO embeddedness in their networks, which presents a fruitful area for future research. The development of institutional ways IOs can embed themselves in their networks illuminates what distinguishes institutional hegemons from other IOs. It is evident from these cases that the centrality of the nodal position of a given IO within its network is a major precondition for any IO to establish institutional hegemony relative to other state-centric and nonstate actors in the network. The centrality of the nodal position of an IO indicates a superior bargaining power relative to other members, which allowed these organizations to centralize their operations with the other actors in their respective networks. Among the cases studied here, EAR and the World Bank-Afghanistan (i.e., MISFA) were the most effective in controlling their respective networks of policy implementers, which is largely a measure of the centrality of their nodal positions and their ability to centralize their operations in them. In both cases, the knowledge, norms, and key principles of program implementation were centralized, albeit through different strategies. EAR used professional program managers as a way to build its in-house capacities to monitor consultants and establish an in-house evaluation unit. Unlike EAR, MISFA continued to use NGOs as key implementing agencies. However, it also engaged in similar centralization of expert authority by establishing clear standards and principles of program implementation in microfinance. The network politics of the World Bank in BiH had lower levels of centrality and more decentralized operations. Consequently, the bank was more subject to the vagaries and uncertainties of network politics. Other donors in the decentralized networks in BiH, as described earlier, had access to NGOs as implementing actors in a way that was comparable to the World Bank. The centrality of the World Bank in these networks was much lower and its bargaining position relative to other players more compromised. Here, the World Bank was more vulnerable to political influences from Western donors and local actors, and it was more vulnerable to network politics than it was in Afghanistan. The empirical section exposed the specific strategies IOs utilize when seeking to transform their superior nodal position and structural power into political power. To that end, IOs establish a culture of legitimacy, learning, and liability within a network. First, the IO legitimizes its ideas and policy prescriptions within the network, regardless of whether or not the network members share broad policy positions (which tends to occur during the policy implementation stage). The institutional hegemons within a network generally succeed in legitimating their ideas. In the case of the microfinance programs in BiH, the World Bank was less effective. Here the bank’s emphasis on the commercialization of microfinance (Drake and Rhyne 2002) was greatly challenged by those (Rankin 2006) advocating alternative approaches to using microfinance in a post-conflict entity. Moreover, the specific goals of microfinance programs were not shared by all the IOs within each network. Instead, the various ideas and principles on the use of microfinance policies clashed, and the structural composition of each network played a big role in determining which one of those ideas would eventually dominate (Ohan-

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

yan 2008). Such a contest of policy principles did not emerge in the cases of the World BankAfghanistan and EAR in the Balkans. In essence, IOs can legitimate “the network’s ideas, consolidating and strengthening the networks, and providing the logistics by which governments actually convert the management concepts into concrete policy measures” (Haas 2000). When legitimating the network’s ideas, IOs contribute to the professionalization of policies and emerge as key actors to socially construct the discourse on the “right policy” in a given region (Haas 2000). Second, as institutional hegemons, IOs may assert their autonomy and withstand political pressures by establishing clear criteria and standards for program management. They do so by empowering the epistemic communities in a given issue area and emboldening the international civil servants relative to the politicians from donor or beneficiary states; they try to institutionalize public administrative capacities of networks by weakening political control from donor states over the network. For example, in the case of the EAR, which is described shortly, the respondent noted the following: Having systems of setting strategy, priorities led by local governments, and having a basic template to which to return is important. Yes, there is always conflict. It may not be the donor but someone from the government; this is [a] common thing in most assistance programs, even in post-conflict countries. We respond to such pressures by saying that “this is a multilevel program, we have agreed with you, we reflect the priorities as already agreed with you, so we can’t change the program and its details.” . . . You have to have a clear strategy and process . . . a semblance of order. You have to empower your partners in the government who are working for integration. . . . Giving them more capacity in discussing these priorities instead of having individuals have a go with any donor is important. (Mustard 2008) Much uniformity characterized the operations under way in Afghanistan and in various areas of EAR’s program in the Balkans. In Afghanistan, the World Bank took the lead in establishing best practice standards in the microfinance sector as a way of preventing the haphazard entry of various types of donors, most of which had little experience in microfinance (Hashemi 2007). This approach was not followed in the World Bank’s microfinance programs in BiH. Third, establishing a culture of learning within a network is a crucial mechanism for IOs to advance their institutional hegemony within a network. This is often accomplished through using evaluations performed by IOs, which can then become a tool for deflecting political pressures from various sources. EAR is particularly noteworthy. Its in-house evaluation unit, as described earlier, ensured that EAR would respond to the local settings and generated a quick feedback loop between the organization and its policy outcomes (Almqvist 2008; Mustard 2008). The microfinance programs of the World Bank in BiH and Afghanistan were also subjected to frequent evaluations, mostly carried out by independent auditing and rating agencies and professional associations. Fourth, establishing a culture of liability within a network, an IO can counterintuitively strengthen its political position relative to other state-centric and non-state actors in the network. By establishing a transparent policy process within the network, IOs tend to strengthen their bureaucratic prowess and add credibility to their organizational purpose. The frequent charge of democratic deficit leveled at IOs has been met over the years with more open policy processes. Interestingly, this response has helped IOs to strengthen their own positions and independence relative to their donors. The culture of liability can also be cultivated by establishing multiple accountability lines by institutional hegemons. In the cases studied here, this approach has served IOs well, deflecting political influences from donor and beneficiary states. In a way, the IOs have purposely constrained and shielded themselves through developing multiple institutional ties, which helped them bolster their positions as expert-led and knowledge-based institutions.

OHANYAN

EAR’s experience has been particularly noteworthy. This agency was accountable to the European Commission’s Council of Ministers, Parliament, independent auditors, and member states. This arrangement complemented EAR’s organizational independence, because the agency was in a position to play off one stakeholder against the other. Conclusion In terms of its theoretical contributions, the network approach to IOs focuses on the institutional conditions under which the ideas or material interests of donor states prevail during the policymaking process. Indeed, casting the debate on IO autonomy in terms of zero-sum wins for states or IOs is not always productive. Some have already introduced a variance within the internal attributes of organization, highlighting their variable impact on the ground (Boehmer, Gartzke, and Nordstrom 2004). But the state centrism that still underpins the debate over IO autonomy dramatically limits the alternative explanations and discounts the nuances of the relationships between the state and the IO. To this end, the immediate institutional environment of an IO—that is, the external links and organizational relationships of IOs with other actors—is perhaps one of the least-treated variables for IO behavior but one that is crucial in gaining a more holistic understanding of IO behavior in world politics. The network approach to IOs developed adds to this debate. It demonstrates that the institutional environment immediately external to an IO can structure its engagement with states and shape its opportunities for independent action. Institutional forms and designs interact with ideas and norms, counterbalancing political pressures from more powerful players—either states or other IOs—in a given setting. In this respect, the network structure emerges as an independent variable of its own that can empower a particular IO at the expense of other organizations or states or vice versa. Consequently, the institutional space for ideas, knowledge, and norms is also variable, and network conditions as institutional arrangements can determine the normative discourse that IOs advance (Koremenos, Lipson, and Snidal 2001). REFERENCES Abbott, Kenneth W., and Duncan Snidal. (1998) ‘‘Why States Act through Formal International Organizations.’’ Journal of Conflict Resolution 42 (1): 3–32. Almqvist, Lennart. (2008) Interview by the author. Thessaloniki, Greece, July 3. Andreoni Dicorrado, Daniella. (2008) Interview by the author. Brussels, Belgium, July 8. Arbab, Amjad. (2008) Telephone interview by the author. Microfinance Investment Support Facility for Afghanistan (MISFA), Kabul, Afghanistan, August 30. Atouba, Yannick and Michelle Shumate. (2010) Interorganizational Networking Patterns among Development Organizations. Journal of Communication 60: 293–317. Baltodan, Andres Perez. (1997) The Study of Public Administration in Times of Global Interpenetration: A Historical Rationale for a Theoretical Model. Journal of Public Administration Research and Theory 4: 615–38. Barcus, Leslie. (2007) Interview by the author. Microfinance Management Institute, Washington, DC, June 20. Barnett, Michael and Liv Coleman. (2005) Designing Police: Interpol and the Study of Change in International Organizations. International Studies Quarterly 49: 593–619. Barnett, Michael and Raymond Duvall. (2005) Power in International Politics. International Organization 59 (Winter): 39–75.

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

Barnett, Michael and Martha Finnemore. (2004) Rules for the World: International Organizations in Global Politics. Ithaca, NY: Cornell University Press. Boehmer, Charles, Erik Gartzke, and Timothy Nordstrom. (2004) Do Intergovernmental Organizations Promote Peace? World Politics 57: 1–38. Börzel, A. Tanja. (1998) Organizing Babylon—On Different Conceptions of Policy Networks. Public Administration 76: 253–73. Byrne, Andrew. (2008) Interview by the author. Brussels, Belgium, July 9. Campara, Samin. (2001) Interview by the author. Catholic Relief Services, Sarajevo, Bosnia and Herzegovina, July 17. Carlarne, Cinnamon and John Carlarne. (2006) In-Credible Government: Legitimacy, Democracy, and Non-Governmental Organizations. Public Organization Review 6: 347–71. Cartopassi, Giovanni. (2008) Interview by the author. Thessaloniki, Greece, July 3. Castells, Manuel. (2005) Global Governance and Global Politics. PS: Political Science and Politics (January): 9–16. DiMaggio, Paul G. and Walter W. Powell. (1991) The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields. In The New Institutionalism in Organizational Analysis, edited by P. G. DiMaggio and W. W. Powell. Chicago: University of Chicago Press. Drake, Deborah, and Elizabeth Rhyne, Eds. (2002) The Commercialization of Microfinance: Balancing Business and Development. Bloomfield, CT: Kumarian Press. Evans, Peter. “Embedded Autonomy: States and Industrial Transformation.” Princeton University Press, 1995. Forster, Sarah. (2002) Telephone interview by the author. London, August, 30. Galaskiewicz, J. (1985). Interorganizational Relations. Annual Review of Sociology 11: 281–304. Gil Loescher. (2001) “The UNHCR and World Politics: State Interests vs. Institutional Autonomy” International Migration Review 35 (1): 33-56. Goldstein, Judith, Miles Kahler, Robert Keohane, and Anne-Marie Slaughter, eds. (2001) Legalization and world politics. Cambridge, MA: MIT Press. Granovetter, M. S. (1973) The Strength of Weak Ties. American Journal of Sociology 78 (6): 1360–80. Haas, M. Peter. (2000) International Institutions and Social Learning in the Management of Global Environmental Risks. Policy Studies Journal 28 (3): 558–75. Hafner-Burton, Emilie M. (2006) Power Positions: International Organizations, Social Networks, and Conflict. Journal of Conflict Resolution 50 (1): 3–27. Haftel, Yoram Z. (2004) “From the outside looking in: the effect of trading blocs on trade disputes in the GATT/WTO.” International Studies Quarterly 48: 121–42. Haftel, Yoram Z., and Alexander Thompson. (2006) The Independence of International Organizations: Concepts and Applications. Journal of Conflict Resolution 50 (2):253–75. Hashemi, Syed. (2007) Interview by the author. Consultative Group to Assist the Poor, Washington, DC, June, 6. Hawkins, G. Darren, A. David Lake, L. Daniel Nielson, and J. Michael Tierney. (2006) Delegation under Anarchy: States, International Organizations, and Principal-Agent Theory. In Delegation and Agency in International Organizations, edited by G. D. Hawkins, A. D. Lake, L. D. Nielson, and J. M. Tierney. Cambridge, UK: Cambridge University Press. Helfer, Laurence R. (2006) Understanding Change in International Organizations: Globalization and Innovation in the ILO. Vanderbilt Law Review 59 (3): 649–726.

OHANYAN

Hwang, Hokyu. (2006) Planning Development: Globalization and the Shifting Locus of Planning. In Globalization and Organization: World Society and Organizational Change, edited by S. G. Drori, M. J. Meyer and H. Hwang. New York, NY: Oxford University Press. Johnson, Tana. (2013) “Institutional Design and Bureaucrats’ Impact on Political Control.” The Journal of Politics 75 (1): 183–197. Jönsson, Christer. (1993) International Organization and Co-operation: An Interorganizational Perspective. International Social Science Journal 45 (4): 463–678. Jordan, R. S., Archer, C., Granger, G. P. and Ordes, K. (2001) International Organizations: A Comparative Approach to the Management of Cooperation (4th ed). Westport, CT: Praeger. Kahler, Miles. (2009) “Networked Politics: Agency, Power, and Governance. In Networked Politics: Agency, Power, and Governance, edited by M. Kahler. Ithaca, NY: Cornell University Press. Keck, Margaret E., and Kathryn Sikkink. (1998) Activists beyond Borders: Advocacy Networks. In International Politics. Ithaca, NY: Cornell University Press. Kiewiet, D. Roderick, and Mathew McCubbins. (1991) The Logic of Delegation: Congressional Parties and the Appropriations Process. Chicago: University of Chicago Press. Klinkhamer, Madeleine. (2001) Evaluation of Microfinance Institutions: UNHCR Bosnia Microcredit Programme 1998–1999. United Nations High Commissioner for Refugees. Koremenos, Barbara, Charles Lipson, and Duncan Snidal. (2001) The Rational Design of International Institutions. International Organization 55 (4):761–99, 768. Lake, David. (1999) Entangling relations: American foreign policy in its century. Princeton, NJ: Princeton University Press. Lewis, David. (2003) Presidents and the Politics of Agency Design: Political Insulation in the United States Government Bureaucracy, 1946–1997. Palo Alto, CA: Stanford University Press. Lipson, Charles. (1991) “Why are some international agreements informal?” International Organization 45: 495–538. Luke, Jeff S. (1992) Managing Interconnectedness: The New Challenge for Public Administration. In Public Management in an Interconnected World: Essays in the Minnowbrook Tradition, edited by M. T. Bailey and R. T. Mayer. Westport, CT: Greenwood Press. Majone, Giandomenico. (2001) “Two logics of delegation: Agency and fiduciary relations in EU governance.” European Union Politics 2: 103–21. McCormick, James M. (1982) Approaches to Evaluating International Organizations: Some Research Directions. Polity 14 (3): 531–47. Mearsheimer, John J. “The False Promise of International Institutions,” International Security, Vol. 19, No. 3 (Winter 1994/1995), pp. 5–49. Meyer, W. John, S. Gili Driori, and Hokyu Hwang. (2006) World Society and the Proliferation of Formal Organization. In Globalization and Organization: World Society and Organizational Change, edited by S. G. Drori, M. J. Meyer and H. Hwang. New York: Oxford University Press. Migdal, Joel S. “Strong Societies and Weak States: State-Society Relations and State Capabilities in the Third World.” Princeton University Press, 1988. Moore, Spencer, Eugenia Eng and Mark Daniel. (2003) International NGOs and the Role of Network Centrality in Humanitarian Aid Operations: A Case Study of Coordination During the 2000 Mozambique Floods. Disaster 27 (4): 305–18. Moravcsik, Andrew. (1999) ‘‘A New Statecraft: Supranational Entrepreneurs and International Cooperation.’’ International Organization 53 (2): 267–306. Mustard, Guy. (2008) Interview by the author. Thessaloniki, Greece, July 3.

ON NETWORKS, INTERNATIONAL ORGANIZATIONS, AND INSTITUTIONAL HEGEMONY

Ness, Gayl D. and Stephen R. Brechin. (1988) Bridging the Gap: International Organizations as Organizations. International Organization 42 (2): 245–73. Nielson, Daniel, and Michael Tierney. (2003) “Delegation to international organizations: agency theory and World Bank environmental reform.” International Organization 57: 241–76. Ohanyan, Anna. (2003) Winning Global Policies: The Network-Based Operation of Microfinance NGOs in Bosnia and Herzegovina, 1996-2002. Syracuse, NY: Political Science Department, Syracuse University. ———. (2008) NGOs, IGOs, and the Network Mechanisms of Post-Conflict Global Governance in Microfinance. New York: Palgrave Macmillan. ———. (2009) Network Model of NGO Behavior. International Studies Review 11 (3): 475–501. ———. (2012) Network Institutionalism and NGO Studies. International Studies Perspectives 13: 366–89. Organization for Economic Development and Cooperation (OECD). (1999). Guidance for Evaluating Humanitarian Assistance in Complex Emergencies. OECD, Paris. Orsini, Amandine, Jean-Frédéric Morin, and Oran Young. (2013) Regime Complexes: A Buzz, a Boom, or a Boost for Global Governance? Global Governance 19: 27–39. Pentland, Charles. (1976) International Organizations. In World Politics: An Introduction, edited by James N. Rosenau, Kenneth W. Thompson, and Gavin Boyd. New York, NY: Free Press, pp. 62–59. Pfeffer, Jeffery, and Gerald R. Salancik. (1978) The External Control of Organizations: A Resource Dependence Perspective. New York: Harper and Row. Podolny, Joel M., and Karen L. Page. (1998) Network Forms of Organizations. Annual Review of Sociology 24: 57–76. Pollack, Mark A. (2002) ‘‘Learning from the Americanists (Again): Theory and Method in the Study of Delegation.’’ West European Politics 25 (1): 200–19. Posner, Elliot. (2009) ‘‘Making Rules for Global Finance: Transatlantic Regulatory Cooperation at the Turn of the Millennium.’’ International Organization 63 (4): 665–99. Putnam, Robert. (1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Rankin, N. Katherine. (2006) Social Capital, Microfinance, and the Politics of Development. In Microfinance: Perils and Prospects, edited by L. J. Fernando. London: Routledge. Rhodes, R. A. W. (1986) European Policy-Making Implementation and Sub-Central Governments. Maastricht, Netherlands: European Institute of Public Administration. Ring, P. S., and A. H. Van de Ven. (1992) Structuring Cooperative Relationships between Organizations. Strategic Management Journal 13: 483–98. Rochefort, David A. and Roger W. Cobb. (1994) Problem Definition: An Emerging Perspective, in The Politics of Problem Definition: Shaping the Policy Agenda, ed. By David A. Rochefort and Roger W. Cobb. University Press of Kansas: Lawrence, Kansas. Sandholz, Wayne, and John Zysman. (1989) ‘‘1992: Recasting the European Bargain.’’ World Politics 42 (1): 95–128. Shields, John, and Mitchell B. Evans. (1998) Shrinking the State: Globalization and Public Administration “Reform.” Halifax, Nova Scotia: Frenwood. Schneider, Anne Larason and Helen Ingram. (1997). Policy Design for Democracy. University Press of Kansas: Lawrence, Kansas. Simmross, Sabine. (2008) Interview by the author. Thessaloniki, Greece, July 3. Slaughter, Anne-Marie. (2004) A New World Order. Princeton, NJ: Princeton University Press.

OHANYAN

Stone, Deborah. (1997). Policy Paradox: The Art of Political Decision Making. W.W. Norton and Company. New York, NY. Stone Sweet, Alex, Wayne Sandholtz, and Neil Fligstein, eds. (2001) The institutionalization of Europe. New York: Oxford University Press. Temcheff, Constantin, and Enzo Caputo. (2004). Evaluation of the Implementation of Council Regulation 2667/2000 on the European Agency for Reconstruction: Synthesis Report. European Commission and Development Researcher’s Consortium. World Bank. (2000) The 1999 Social Assessment: Bosnia and Herzegovina. Washington, DC: The World Bank Group. Tinjič, Goran. (2001) Interview by the author. The World Bank,Sarajevo, Bosnia and Herzegovina, July 12. Wasserman, S., and K. Faust. (1994) Social Network Analysis: Methods and Applications. Cambridge, UK: Cambridge University Press. Wendt, Alexander. (1995) Constructing International Politics. International Security 20 (1): 71–81. Zaheer, A., B. McEvily, and V. Perrone. (1998) Does Trust Matter? Exploring the Role of Interorganizational and Interpersonal Trust on Performance. Organization Science 9 (2): 141–59. Zaheer, Akbar, Remzi Gözübüyük, and Hana Milanov. (2010) It’s the Connections: The Network Perspective in Interorganizational Research. Academy of Management Perspectives 24 (1):62–77.

Judicial Voting Behavior at the Appeals Chambers of the International Tribunals by James Meernik, Kimi King, Melissa McKay, University of North Texas, and Rosa Aloisi, Trinity University Judges at the international criminal tribunals are revolutionizing the development of international law. Given the complexity of the relevant conflicts and crimes being prosecuted, the recent creation of the ad hoc tribunals and the embryonic state of international law, it is not surprising that many appeals are lodged against the initial decisions of these courts. The appeals chambers have confirmed and codified many of the trial chamber rulings, but they have also reversed decisions and progressively developed international law to expand its reach and to confront the complexity of these human rights atrocities. We seek to understand the judicial decision making of the appellate judges by analyzing their votes to uphold or reverse the decisions of the trial chambers. We propose a new depiction of judicial behavior that accounts for the propensity of judges to uphold the decisions of their trial chambers, or reverse these judgments by upholding the appeals of the prosecution or the defense.

Introduction Judges at the international criminal tribunals are revolutionizing the development of international law. The ad hoc tribunals—the International Criminal Tribunal for the Former Yugoslavia (ICTY), the International Criminal Tribunal for Rwanda (ICTR), and the Special Court for Sierra Leone (SCSL)—have been at the forefront of what Kathryn Sikkink (2011) terms, the “justice cascade,” the international movement to protect the rights of individuals and to hold leaders criminally accountable for their actions. When the promise of the Nuremburg and Tokyo tribunals to apply law to exact justice rather than vengeance in the aftermath of war foundered on the shoals of the Cold War, international humanitarian and human-rights law remained mostly dormant for the next forty years. But as the ethnic cleansing of the Balkan wars, the genocide of Rwanda and the horrible atrocities of Sierra Leone’s civil war demonstrated horrific violence was neither a relic of World War II nor the Cold War, and international law was given new life. Indeed, since the establishment of the ICTY in 1993, these three tribunals have handed down verdicts against 188 individuals. The judgments and the sentences meted out to the guilty have been widely visible and garnered a great deal of scholarly and political attention. Now, as the tribunals prepare to conclude the last of their trials and shutter their doors, it is timely and important to develop a theoretical account of their decision making. Each of the three ad hoc tribunals is organized into several trial chambers, which hear cases brought against those charged with violating international laws, and an appeals chamber, which hears appeals brought by both the defense and the prosecution. The trial chambers, whose Rules of Procedure and Evidence and courtroom management are an amalgam of common and civil law, issue verdicts and sentences and apply international law to novel situations, often for the very first time. Indeed, the Genocide Convention had never been applied internationally until the ICTR handed down its first judgment in 1998. Crimes against humanity have JIOS, VOL. 6, ISSUE 1, 2015

MEERNIK, ALOISI, KING, AND MCKAY

existed in practice since the dawn of humanity, but there remains no international codification of their number, meaning, and elements. The tribunal judges have been at the forefront of the clarification, deepening, and expansion of international law. Given the complexity of the relevant conflicts and crimes being prosecuted, the recent creation of the ad hoc tribunals, and the embryonic state of international law, it is not surprising that many appeals are lodged against the decisions of the trial chambers. The appeals chambers have confirmed and codified many of the trial chamber rulings, but they have also reversed decisions and progressively developed international law to expand its reach and to confront the complexity of these human rights atrocities. And as the courts of last resort for these international laws, their decisions are final. We seek to understand the judicial decision making of the appellate judges by analyzing their votes to uphold or reverse the decisions of the trial chambers. Research on international courts has examined the propensity of judges to issue decisions that reflect the interests of their national governments but has mostly concluded that while there is some evidence such strategic voting occurs on some issues, it is mostly not a widespread phenomenon (Voeten 2008). Analogous research on domestic courts, particularly in the U.S., has long debated the extent to which judicial ideology—measured on a conservative-liberal spectrum—characterizes judicial behavior, or whether the law itself in the form of policy or precedent guides judicial interpretation. We believe that both international and domestic courts research can inform our understanding of the appellate decisions of the ad hoc tribunals. Our goal, however, is to propose a new depiction of judicial behavior that accounts for votes according to judges’ propensity to uphold the decisions of their trial chambers or reverse these judgments by upholding the appeals of the prosecution or the defense. By analyzing the individual votes of judges in the appeals chambers of ad hoc tribunals for all of the claims advanced by both the defense and the prosecution, we seek to determine the degree to which the voting behavior of individual judges falls into recognizable patterns. Specifically, we propose that judges may be located along three dimensions based on their votes. We must emphasize that these proposed dimensions are hypothesized perspectives through which we may, hopefully, understand judicial decision making. Other decisionmaking dimensions undoubtedly exist as well. We also do not suggest that judges necessarily seek to vote in a particular fashion—we are not arguing that these categories necessarily reflect intentional behavior. Rather, our aim is to determine if these dimensions have any utility in accounting for judicial votes on appellate issues that may then help us better explain the outcomes of these votes. The first of our proposed dimensions is the institutionalist perspective. To the extent that such a dimension conveys a judicial ideology, we argue that judges so classified will tend to favor protecting the interests of their tribunal and will generally be reluctant to grant the appeals advanced by either the defense or the prosecution. Appeals judges have, on several occasions, underlined the importance of creating a coherent and authoritative body of law by showing deference to trial-chamber decisions and intervening only when errors of law have caused a miscarriage of justice.1 Their main interest remains that of preserving the legitimacy of the institutions by developing a consistent body of law and ensuring continuity in the vertical hierarchy of judicial decision making. The second dimension we propose is deference to the prosecution. We suggest that judges who may fall into this category will be more likely to grant the appeal claims of the Office of the Prosecutor in order to advance the mandate of the tribunals to provide justice to the victims and the international community and to promote the general deterrence of international crimes. The ad hoc tribunals represent the first unequivocal attempt since Nuremberg and Tokyo at holding accountable those that have com1. See Eliezer Niyitegeka v. Prosecutor, Case No. ICTR-96-14-A at paragraph 8 “the Appeals Chamber must give deference to the Trial Chamber that received the evidence at trial, and it will only interfere in those findings where no reasonable trier of fact could have reached the same finding or where the finding is wholly erroneous. Furthermore, the erroneous finding will be revoked or revised only if the error occasioned a miscarriage of justice.”

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

mitted the most heinous human-rights violations. Hence, there is substantial concern for ending impunity for those who violate international law and punishing violations of human rights. The third dimension we propose to describe in judicial voting behavior is concern for protecting the rights of the defendants. We suggest that the voting behavior of some judges may be categorized as receptive to the appeals of the defense. Such judges may place a high priority on the importance of protecting defendantsâ&#x20AC;&#x2122; procedural rights, and protecting defendants against nullum crimen sine lege and nullum crimen sine poena. It is our general contention that such judicial decision-making dimensions can serve as useful categories for classifying voting behavior and that scholars can utilize this classification system to better understand and predict judicial decision making, the development of international law, and long-term viability of international justice. Hereafter, our paper is organized as follows: We provide background into the appellate process and review the relevant literature, especially on the international courts to determine the degree to which such research can inform our understanding of international criminal courts. Following that, we develop in full our classification system of judicial philosophies. This is followed by our data. We collected information on every ground of appeal raised by both the prosecution and defense and recorded the votes of each of the judges, as well as their professional and home country characteristics. Then we explore the extent to which this system aptly characterizes judicial voting behavior. We also briefly evaluate the degree to which judicial background and home country characteristics influence the propensity of judges to fall into one of our categories. We caution that this is a preliminary and descriptive analysis. We are chiefly interested in introducing our data and examining general trends rather than hypothesis testing. We have saved that for a series of subsequent papers. We conclude with a discussion of this research agenda. Background Overview of Appellate Process At the conclusion of the trials at the ad hoc tribunals the three-judge panels that hear these cases issue their judgments. These judgments review the evidence and assess its probative value. They analyze the various interpretations of the relevant international law in order to resolve legal issues regarding the liability of the defendant(s) and conclude with a disposition that formally announces the verdicts and the sentences for those found guilty. These verdicts are issued on each of the counts on which an individual has been charged in the final indictment. The judgments also indicate whether any of the three judges disagreed with the majority (the affirmative votes of two of three judges are required for a guilty verdict). Often, there are dissenting opinions as well as annexes containing more detailed descriptions of the evidence or other pertinent matters. Those adjudged guilty are given a sentence, which generally reflects the sum total of their guilty conduct rather than a series of penalties individualized for each particular guilty count. All those found guilty appeal their verdicts, with the exception of those who have pled guilty (and even some of these individuals have appealed their sentences). None of those found guilty agree with the reasoning provided by the judges. As is typical of the ad hoc tribunals, according to the ICTY Statute, art. 25: The Appeals Chamber shall hear appeals from persons convicted by the Trial Chambers or from the Prosecutor on the following grounds: (a) an error on a question of law invalidating the decision; or (b) an error of fact which has occasioned a miscarriage of justice. In the ICTY Rules of Procedure and Evidence, the judges elaborate further on the appellate process: A party seeking to appeal a judgment shall, not more than thirty days from the date on which the judgment was pronounced, file a notice of appeal, setting forth the grounds. The Appellant should also identify the order, decision or ruling challenged with specific

MEERNIK, ALOISI, KING, AND MCKAY

reference to the date of its filing, and/or the transcript page, and indicate the substance of the alleged errors and the relief sought. The Appeals Chamber may, on good cause being shown by motion, authorise a variation of the grounds of appeal. (Rule 108) Defendants appeal as many aspects of the trial chambers’ factual and legal reasoning as possible. The defendants’ appeals are packed with allegations of everything from judicial misconduct to faulty legal reasoning and from misapplication of command and control doctrine to failure to consider their mitigating circumstances properly when sentencing. The Office of the Prosecutor can appeal verdicts and sentences as well and has shown little reluctance to do so, especially when it perceives a defendant’s sentence is too lenient or the defendant has been acquitted on a particular count. The OTP may also file appeals in cases where a defendant is entirely acquitted. The Appeals Chambers of the ICTY and ICTR are comprised of the same judges. There are a total of seven judges who make up the Appeals Chamber with five coming from the ICTY and two from the ICTR. When appeals arise from the ICTY, five of these judges make up what is officially known as the Appeals Chamber of the ICTY. The same process occurs with regard to the ICTR. The Appeals Chambers issue their lengthy decisions regarding all of the issues raised on appeal by both parties. The judges typically review each and every appeal raised by the parties down to each discrete issue that may make up a broader category of appeals issues. For example, a defendant may issue an appeal against sentence, which then consists of a series of more specific appeals on issues like improper consideration of mitigating circumstances, allegations that elements of the crimes that must be proven to establish guilt were also used to establish aggravating circumstances, failure to consider the sentencing practices of the former Yugoslavia, etc. These appeals issues may then be further broken down into even more specific issues, such as failure to properly consider evidence of a defendant’s cooperation with the Prosecution in mitigation of sentence or failure to properly consider the impact of incarceration on a defendant’s health. Thus, there are many specific appeals that make up broad categories of appeals. All are issues on which the judges then reach decisions to either grant or reject the specific argument. The purpose of our analysis is to assess judicial-voting behavior at the level of the most specific issues on which the judges render a decision. Prior Research Research on international judicial decision making has advanced slowly despite the emerging role played by international courts in conflict management and resolution. In particular, studies of judges at the international tribunals have certainly not kept pace with research on the legal and political developments emanating from domestic criminal courts (e.g., Darcy and Powderly 2010; Drumbl 2007; Sluiter, Swart, and Zahar 2011, and see Terris, Romano, and Swigart 2007). Some studies have analyzed the propensity of judges to vote for more severe sentences (Meernik, Aloisi, and Ding 2012; Meernik, King, and Dancy 2005), while others have begun to develop more generalized models of judicial decision making (Jodoin 2010) that are informed by studies of the U.S. Supreme Court. While these studies do help establish a rationale for studying international judicial decision making and a foundation for theoretical development, we have many more questions about why judges might vote along any ideological or political grounds and even whether there are any identifiable voting patterns. Sebastien Jodoin’s work (2010) is among the first to posit a theoretical framework for understanding the decisions of the ad hoc tribunals. He suggests two alternative conceptions of judicial decision-making. The attitudinal model, borrowed from research on domestic courts and the U.S. Supreme Court in particular (see especially Segal and Spaeth 1993, 2002), suggests that judges have policy preferences they seek to advance through interpretation of the law. Because judges at the international tribunals, and in particular the appeals chambers we study, share key features in common with the U.S. Supreme Court (e.g., they are a court of last resort), Jodoin suggests the ICTY and ICTR judges may seek to advance policy objectives,

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

although these will differ from the types of ideological preferences one finds in the domestic courts. Indeed, Jodoin suggests that international judges have even greater latitude than their domestic counterparts in interpreting the law and advancing ideological or political interests. He writes: The traditional left-right categories along which judicial preferences are aligned at the domestic level may need to be modified when moving from the domestic to the international context. In the context of international criminal law for example, in a complete reversal of the situation present at the domestic level, the left has been the strongest proponent of international prosecution while the right has been more recalcitrant. Likewise, the categories of activism and conservatism through which judicial roles at the domestic level are often understood may also not translate to the international level. (Jodoin 2010: 7) Instead, Jodoin contends that the judicial philosophy we are most likely to find in evidence in these types of international courts concerns judicial orientation on the relationship between state sovereignty and international law. Judges who apply international humanitarian, human rights, and international criminal law must determine the extent to which such law can be interpreted to increase the authority of these tribunals to weigh into matters that pertain directly to state sovereignty and security matters. Those who favor a more activist approach are more willing to expansively interpret customary international law or interpret treaties “in a teleological manner.” Such judges would carve out an increasingly consequential role for tribunals and the law in the conduct of conflict. Statist, or more conservative, judges accord greater value and deference to state sovereignty and a more limited expansion of international law (Jodoin 2012: 8). There would be less legal justification for regulation of conflict behavior through international law and courts. Thus, this ideological dimension mirrors in many ways the judicial activism and restraint philosophies one finds dominating studies of U.S. courts. Jodoin also suggests that judges may vote according to strategic rationales. Strategic decision making is quite flexible and assumes that judges have more far-reaching personal and institutional interests they seek to advance through their decisions. Such objectives might be generalizable to the international tribunals. While Jodoin looks at the propensity of judges to reach decisions that achieve either institutional or personal interests, our research is more concerned with judicial efforts to advance organizational interests. As Jodoin (2010: 13) argues, these international courts are: Most concerned with extending their authority, standing, independence and influence in the particular regime or region in which they operate. The notion of organizational self-interest advanced here is not focused on the mere survival of an organization, but instead on its non-material interests in terms of ideational commitments, reputation, and effectiveness. We believe such interests will be especially relevant at the ad hoc tribunals, which, because of the new and revolutionary role in the interpretation of international law, have been especially concerned with their power and legacy (King and Meernik 2011). We will return to this theme in more detail when we discuss our proposed typology of judicial philosophy. Several scholars have suggested that one strategic objective judges may seek to advance are the interests of their own nation. Yet the evidence that national interests affect judicial voting is not consistent. While Eric Voeten (2008) finds some evidence to suggest that some judges are more likely to support their home governments on political sensitive cases, in general he notes the evidence of nationalist voting on the European Court of Human Rights is modest. Eric Posner and Miguel de Figueirido (2005), however, find strong evidence to suggest that judges vote along state interest lines on cases before the International Court of Justice. They write, “Judges vote for their home states about 90 percent of the time. When their home states are not involved, judges vote for states that are similar to their home states—along the dimensions of wealth, culture, and political regime” (Posner & de Figueiredo 2005). But in their study of judicial voting behavior at the ICTY, James Meernik, Kimi King, and Geoffrey Dancy

MEERNIK, ALOISI, KING, AND MCKAY

do not find evidence of any kind of home country bias among judges from the NATO states that took part in military campaigns against parties in the wars of the former Yugoslavia. They looked to determine if such judges would punish Serbs, as the ethnic group targeted by their home governments, more severely, but they did not find evidence of any strong impact. More recently, Meernik (2011) argued that judicial decision making regarding sentencing behavior at the international tribunals is driven by concerns for retribution, deterrence, and peacebuilding. However, this line of research is focused more on a fairly specific outcome of the Trial Chamber judgment and is not the type of judicial philosophy in evidence in higher-level, appellate courts. In sum, the nascent research on judicial decision making at the international tribunals has only begun to develop theoretical models borrowed from domestic courts and the International Court of Justice. Our purpose here is to begin building on this initial work to suggest three dimensions or philosophies of judicial decision making we are likely to observe in the outcomes of appellate judgments at the international criminal tribunals. After having explained the proposed philosophies, we analyze data on judges’ votes on all of the issues raised in appeals. Our goal is to both determine if individual judges exhibit characteristics of these philosophies and to determine if these philosophies are more prevalent among certain types of judges based on home country characteristics. Three Judicial Decision Making Dimensions—A Modest Proposal Before proceeding, we must emphasize that the judicial decision-making dimensions or philosophies we propose are ideal types that exist along a dimension or continuum. The end points of these dimensions represent the most pure ideological variant of the dimensions. Hence, it is unlikely that in the real world there are individuals who occupy these positions. As such, these hypothesized philosophies are intended to convey the exemplars and most noteworthy features of these dimensions. It is extremely unlikely that any one judge’s decision making strictly conforms to such depictions. Nor do we wish to suggest that these are necessarily conscious ideologies or preferences judges intentionally pursue. Rather, we seek to describe below the broad contours of three possible judicial decision-making dimensions grounded in previous research and subsequently examine the empirical record to ascertain their validity. Institutionalism We conceive of the institutionalist philosophy, to the extent that such a hypothesized voting dimension may be found to exist among judges of the appellate chambers, as a preference for protecting the interests and decisions of the tribunals and the facts and precedents established in the judgments of the trial chambers. Accordingly, such a philosophy would conform to the notion that the decisions of the trial chambers should rarely be disturbed, and only then when there are compelling factual and/or legal errors at issue. Voeten (2008: 422) finds such a philosophy in evidence at the European Court of Human Rights. While we expect that judges would generally accord the opinions of the trial chambers substantial deference and would not engage in de novo review of their decisions, we suggest that some judges might hold even stronger preferences for not disturbing these prior decisions. Furthermore, in the idealized view of this decision-making type, we should find that intuitionalist judges would reject, more or less equally, the appeals of both the prosecution and defense. They would demonstrate a degree of deference toward the trial chambers’ decisions and reasoning, and would tend to set a higher bar for determining when factual and legal errors are substantial enough to warrant overturning a decision of the trial chamber. Thus, in our idealized conception of judicial philosophies, judges may be placed somewhere along a dimension represented by two extremes: one by a substantial reluctance to overturn the decisions of the trial chamber because they believe strongly in the importance of institutional interests and the other by a more “activist” proclivity in which their threshold requirement for reversing lower court decisions is significantly lower.

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

Institutionalist judges may base such preferences on several interrelated rationales. First, judges who practice judicial restraint may be generally reluctant to overturn trial chamber rulings and judgments because of a respect for precedent. Somewhat similar in form to a conservative or judicial restraint philosophy, such judges may view their role as ensuring continuity in the interpretation of the law lest the law, especially in an already rapidly evolving domain, be perceived as lacking a firm anchor in reasoning and subject to the whims of shifting judicial preferences and various combinations of judges in the appeals chamber. Such change might also undermine the legitimacy of international law and the tribunals if the international community viewed the international justice project as unpredictable and subject to the preferences of an elite and removed corpus of judges. And again, given the recentness of their creation, we might expect some judges to be especially solicitous of the need to establish the legitimacy and authority of the tribunals in the adjudication of conflict. As Cassese writes (390), “Although none of the statutes prescribe a system of precedence, where Appeals decisions are formally binding on Trial Chambers, this hierarchy has been adhered to in practice.” Second, we argue that institutionalist judges are less likely to overturn the rulings of trial chambers, because they recognize these lower court judges have a substantially greater understanding of the complex political and military issues and facts that enter into the trial record. We would expect that judges who believe in such restraint would caution against substituting the judgment of the appeals chamber and grant greater deference to the trial chambers’ familiarity with the evidentiary record. Third, we believe that judges who subscribe to a philosophy of judicial restraint may do so because they seek to bequeath to the international legal community a solid foundation upon which to gradually solidify the influence of international courts and the legitimacy of international law in contemporary international relations. Such judges may perceive that a clear and consistent jurisprudence that provides a foundational legacy for future cases may be the most important contribution the tribunals can make to international law. Deference to the Prosecution Our second, hypothesized type of judicial philosophy is deference to the prosecution. We suggest that some judges’ voting behavior in appeals chamber is characterized by a preference for the substantive objectives of their respective tribunals as outlined in their founding documents (e.g., the United Nations Security Council resolutions authorizing the ICTY and ICTR and their founding statutes). In particular, these tribunals were intended to contribute to the establishment of peace, the deterrence of future violations of international law, and most importantly of all the provision of justice for the victims of international crimes. We hypothesize that some judges will be especially mindful of this concern for human rights and justice and so will typically be reluctant to accede to the appeals of defendants. Such a philosophy we believe might stem from both an interest in the legacy of the tribunals in contributing positively to conflict amelioration and because of a concern for human rights. If the tribunals are perceived as having meted out sentences that do not adequately convey the severity of the crimes, or if their judges prove too willing to let arcane legal issues subvert the broader purposes of justice by acquitting individuals on “technical” grounds, the purpose of the judicial enterprise of international crime may be compromised. Hence, we believe some number of judges will be more likely to be especially solicitous of the appeals advanced by the Office of the Prosecutor and exhibit greater reluctance to question their evidence and arguments, lest such choices undermine the very purposes of the tribunals. These judges, we contend, will be interested in ensuring that the tribunals mete out just deserts to defendants to demonstrate that such conduct will not be tolerated and to signal to local communities that guilty leaders have been punished and that peace and reconciliation are possible. Therefore, we believe the voting behavior of judges may also be located along this continuum in which at one end judges will give significant deference to the Office of the Prosecutor and grant most of their appeals while others will not exhibit such deference and voting behavior.

MEERNIK, ALOISI, KING, AND MCKAY

Judges may also be more inclined to grant the appeals raised by the OTP because of a special interest in protecting and promoting human rights. The judicial record is heavily seasoned with arguments about the importance of protecting and advancing human rights. For example, one finds many appeals to and application of the Martens Clause2 by the ad hoc tribunals to cases in which deliberate and violent attacks on civilians were perpetrated. The Martens clause represents one of the most important principles in which considerations about human rights and humanitarian law find a common ground of application. Since their establishment, the ICTY and ICTR have been instrumental in the implementation of basic rules of respect for humanity as mentioned in the Martens clause. In the Martic case at the ICTY, the Trial Chamber found that war should not be unlimited in its violence and that the basic respect of humanity “constitutes the foundation[s] of the entire body of international humanitarian law applicable to all armed conflict.”3 Thus, human rights and humanitarian concerns have been at the center of international judicial reasoning. This concern, we contend, is a principal impetus for the need to expand the law and employ its language to speak to the law’s higher moral purposes. Concern for Defendants’ Rights Our third proposed dimension of judicial decision making concerns the extent to which judges attach importance to the procedural and legal rights of defendants. Such judges, we might surmise, would be most concerned with ensuring that proper procedures and an equality of arms between the prosecution and defense exist so that the tribunals do not exact a victor’s or victim’s justice at the expense of the rights of the defendant (Ellis 1997; Johnson 1998; Meernik 2003). Indeed, there has been a traditional criticism dating back at least to the Nuremburg and Tokyo tribunals that international justice is victor’s justice and that such tribunals have been “organized to convict.” In truth, the wealth of resources and expertise available to the prosecution, with their long history of investigation, evidentiary base, and legal argumentation dwarfs the resources and infrastructure available to the defense. Therefore, we suggest that some judges may be especially mindful of the many challenges confronting the defense and so will take seriously their claims of improper procedures (e.g., withholding of evidence by the OTP), the inadequacy of their ability to mount a case (e.g., the massive volumes of information they are given), and the use of novel and shifting legal standards and definitions (e.g., the meaning of “protected groups” in the context of genocide, the use of the joint criminal enterprise approach to charge individuals, cumulative convictions). As well, such judges may also be concerned with the legacy of the tribunals in international legal circles as well as in the affected communities (e.g., Bosnia, Rwanda) if the tribunals are perceived as having rushed to judgment or having not been impartial in their treatment of defendants. Thus, we would expect such judges to more often grant the appeals brought forward by the defense. Data Analysis Our goal is to assess the individual voting behavior of judges at the ad hoc criminal tribunals. We examine the ICTY and the ICTR as they share a common appeals chamber. As we indicated above, both defendants and the Office of the Prosecutor may issue appeals on many and various grounds, and our goal is to identify each instance whereby judges rule on the grounds of appeal. Using such a framework means that our analysis of judicial decision making goes into considerably more depth than previous studies of the appeals process that have looked at 2. The Martens Clause was first introduced in the preamble to the 1899 Hague Convention as a minimum yardstick for the protection of civilians, even in the absence of any international treaty. It establishes that “Until a more complete code of the laws of war is issued, the High Contracting Parties think it right to declare that in cases not included in the Regulations adopted by them, populations and belligerents remain under the protection and empire of the principles of international law, as they result from the usages established between civilized nations, from the laws of humanity and the requirements of the public conscience.” 3. Prosecutor v. Martic. Case No. IT-95-11-R61, Decision under Rule 61 of the ICTY Rules of Procedure and Evidence, (March 8, 1996) paragraph 13. See also Meron (2000) and Cassese (2000) discussing the evolution and relevance of the Martens clause.

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

ultimate outcomes, such as whether the defendant was acquitted or whether the defendant’s sentence was reduced on appeal (Meernik 2007). In fact, using such a classification system in which we measure each specific ground of appeal results in a data set of 3,100 cases across seventy-five trials at the ICTY and ICTR, decided through February 2014. Since we wish to measure the votes of each of the five judges who sit on an appeals case, this number of 3,100 issues must be further multiplied by five judges per case to create a data set of 15,500 total observations. In addition to the judges’ votes, we also measured the issues raised by the appellants; which party was appealing a particular outcome; the overall vote of the Appeals Chamber on each issue; and the home country characteristics of the judges. Data on the characteristics of the judges’ home countries are measured for the year of each of the appeals chamber decisions. Thus, a judge’s vote on an appeals judgment handed down in 2011 is combined with the country information (democracy, judicial independence, human rights protections, etc., described later) for that particular year. We describe each of these below as we assess the judges voting behavior on the issues. Table 1. Appeals Chamber Judges Votes on All Appeals Issues Raised, 1995–2014 ALL APPELLATE VOTES

OTP APPEALS VOTES

DEFENSE APPEALS VOTES

Judge

Percent

Number of Votes

Percent

Number of Votes

Percent

Number of Votes

Agius

0.239

471

0.574

0.196

417

Güney

0.166

2474

0.182

312

0.163

2100

Jorda

0.135

184

0.012

161

Liu

0.217

1013

0.308

107

0.206

905

Meron

0.165

1822

0.184

265

0.161

1496

Mumba

0.142

646

0.062

0.156

550

Nieto-Navia

0.183

174

0.705

0.127

157

Pocar

0.151

2331

0.145

282

0.151

1987

Ramaroson

0.325

160

0.25

0.329

155

Robinson

0.168

409

0.32

0.147

359

Schomburg

0.151

1214

0.186

241

0.142

973

Shahabuddeen

0.121

1747

0.325

209

0.09

1477

Vaz

0.135

1101

0.18

200

0.121

840

Vohrah

0.145

144

0.857

0.109

137

Weinberg de Roca

0.204

752

0.234

658

AVERAGE

0.168

15,138

0.228

2048

0.158

12,960

Table 2. Appeals Chamber Judges Votes on Appeals Issues by Tribunal, 1995–2014 ICTY

ICTR

Percent

Number

Percent

Number

All Appeals

0.157**

8880

0.183**

6438

Prosecution

0.261**

1605

0.108**

443

Defense

0.132**

6970

0.188

5990

T-Test *= p < .05 ** = p < .01

MEERNIK, ALOISI, KING, AND MCKAY

Overall Trends We begin by examining the voting behavior of the individual judges at the ICTY and ICTR Appeals Chambers. We look only at those judges who have voted on at least one hundred issues raised on appeal. We look first at all issues raised on appeal by both prosecution and defense to describe judges’ individual likelihood of reversing a trial-chamber decision before analyzing their proclivity to reverse the trial chamber on issues raised by the Office of the Prosecutor or the defense. The average rate at which judges reverse trial chambers’ decisions is 16.8% (15,138 votes, given missing information on some cases). Of the fourteen judges who voted, at least one hundred times, on such issues, Judge Arlette Ramaroson is most likely of all to vote to reverse (32.5% of 160 votes), followed by Judge Carmel Agius, who voted to reverse 23.9% of the time (471 votes), and Judge Liu Daqun, who voted to reverse 21.7% of the time (1013 votes). At the other end of the spectrum, Judge Mohamed Shahabuddeen voted to reverse only 12% of the time while Judges Claude Jorda and Andrésia Vaz each voted to reverse 13.5% of the time. More interesting patterns emerge in the judges’ voting behavior when we confine our analysis to the issues raised on appeal by the OTP. While the judges’ overall probability of voting to support OTP appeals is 22.8% (2048 votes), the range is remarkable. We see Judge Jorda voting to reverse in favor of the prosecution on every single issue it raises (23 votes) on which he voted, while Judge Inés Weinberg de Roca never votes to reverse on any Office of the Prosecutor appeal (94 votes). Indeed, there are several other judges who grant the OTP appeals more than half the time including Judges Lal C. Vohrah (85% of 7 votes); Rafael Nieto-Navia (70% of 17 votes); and Carmel Agius (57% of 54 votes). At the other end of the spectrum, Judge Florence Mumba grants the OTP appeals in only 6% of the issues on which she has voted (96 votes). We do not witness such divergence when we examine the judges’ propensity to vote to grant the appeals claims raised by the defense. Overall, the likelihood any given judge will grant an appeal raised by defense is 15.8% (12,960 votes). The judge scoring highest in this regard, Judge Ramaroson, voted to support the defense on 32.9% of the issues (155 votes). The next closest are Judge Weinberg de Roca at 23.9% (658 votes) and Judge Liu at 20.6% (905 votes). Judge Jorda voted to grant defense appeals only 1.2% of the time (161 votes), while Judge Shahabuddeen granted defense requests only 9% of the time (1,477 votes). Table 2 contains the breakdown of support for appeals claims at the two tribunals. We note that there are statistically significant differences across virtually all of the categories of support. The most visible difference concerns the propensity for judges to grant the appeals of the prosecution. On cases coming up from ICTR, judges typically grant OTP appeals approximately 10.8% of the time, while on ICTY cases OTP appeals are granted 26.1% of the time. Perhaps some of this may be due to the fact that ICTR defendants are typically given a life sentence, which may not precipitate further OTP appeals to obtain a more severe sentence. At ICTY, on the other hand, not only have more defendants been acquitted, but the sentences are generally lighter, which may cause the OTP to file more appeals when it believes a more severe sanction is warranted. This observation, however, concerns only the propensity of the OTP to file appeals. It does not speak directly to the reasons why the OTP fares significantly better with the ICTY Appeals Chamber. We see also in Table 2 that ICTR defendants are successful on appeals 18.8% of the time (5990 cases), while ICTY defendants are successful 13.2% of the time (8880 cases). Discussion Our proposed set of judicial decision-making dimensions, while not tested explicitly as a group of hypotheses, shows promise as a methodology by which to characterize judicial behavior. Most of the judges hew fairly closely with the average levels of support given to the decisions of the trial chambers, which is 16.8%. These findings are represented graphically

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

Figure 1. Percentage Support Given to All Appeals with Standard Deviation Ranges

Figure 2. Percentage Support Given to OTP Appeals with Standard Deviation Ranges

Figure 3. Percentage Support Given to Defense Appeals with Standard Deviation Ranges

MEERNIK, ALOISI, KING, AND MCKAY

in Figure 1. Judges Theodore Meron and Mehmet Güney hew most closely to the average, and only one judge, Ramaroson, falls outside the standard deviation range. There are some judges at either end of the spectrum who appear more inclined to grant appeals than their colleagues, including Ramaroson, Agius, and Liu, and some who are more apt to reject appeals, such as Vaz, Jorda, and Shahabuddeen. When we examine these trends according to whether the OTP or defense lodged the appeal, we see more visually who the outliers may be. There are noticeably more judges who are outliers once we make these distinctions, especially in Figure 2, which shows the proclivity of judges to support OTP appeals. Judges Jorda, Vohrah, and Nieto-Navia all fall outside the standard deviation range by according OTP appeals substantially more deference than their colleagues. At the other end of the spectrum, Judge Weinberg De Roca falls outside the standard deviation range in turning down OTP appeals, although we must caution that the number of votes on which these are based is rather small in some cases. We see in Figure 3 that on defense appeals it is once again Judge Jorda who is at the far end of the spectrum in granting requests. Jorda has granted only 1 percent of defense appeals. Judge Shahabuddeen is very close to Jorda in his voting preferences, although he does not fall outside of the standard deviation range. At the other end of the continuum we find Judge Ramaroson falling outside of the standard deviation range by supporting defense appeals much more than her colleagues. It would appear that there are several judges whose voting behavior is significantly different from their colleagues. While their numbers are not substantial, these findings do lend support to the notion that judges may be categorized only in the three dimensions outlined in this paper. The overall tendency is for judges to fall close to the averages, especially on defense appeals, which comprise the vast majority of all such requests. Most judges would appear to be institutionalists who grant few appeals and uphold the decisions of their colleagues in the trial chambers with a much smaller number who might be tentatively grouped together as those who defer to the OTP or are more likely to protect the rights of the defendants. We next explore what factors might be associated with these types of voting behavior. Judges’ Home Country Characteristics and Voting Behavior We examine judicial voting behavior according to the type of judicial system and the level of judicial independence in judges’ home countries. We categorized judges as hailing from either a common-law system, a civil-law system, or a mixed system that was not wholly one or the other. Judges also come from other types of systems, but because we do not have a sufficient number of judges in many of these system types, we cannot distinguish the influence of the legal system from the individual. The judicial system in which the judge was trained, practiced, and socialized should exert an important influence on the judge’s understanding of his role and responsibility in the courtroom and in the decision-making process (Balas et al. 2008; la Porta et al.1998, 2004). Civil law systems place a premium on the discovery of the truth in which the role of the judge is to acquire as complete a picture as possible of the criminal allegations to determine the veracity of the charges against the individual. Common-law systems entrust the proof of charges to the lawyers arguing the case and assume that through an adversarial competition in the courtroom the relevant facts will be made known to the judge and jury. But while common-law judges, who act more as neutral umpires in this process, are not nearly as active in the courtroom as their civil law peers, they do have a long tradition of developing legal doctrine from the bench to fill in gaps and ambiguities in the law while also evaluating the constitutionality of statutory law. Perhaps this tradition of activism will make them more inclined generally to reverse trial chamber decisions. In fact, we do not see evidence of significant variation among the legal systems represented here that would permit us to draw firm conclusions regarding whether those arriving from one system rather than another are more likely to reverse an appeal. Common-law judges

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

Table 3. Appeals Chamber Judges Votes on Appeals Issues by Legal System, 1995–2014 Common Law

Civil Law

Mixed

Percent

Number

Percent

Number

Percent

Number

All Appeals

0.173

2330

0.162

7219

0.165

5397

Prosecution

0.233

330

0.186**

986

0.279**

697

Defense

0.163

1939

0.158

6109

0.147

4576

T-Test *= p < .05 ** = p < .01

Table 4. Appeals Chamber Judges Votes on Appeals Issues by Judicial Independence, 1995–2014 No Judicial Independence

Some Judicial Independence

Judicial Independence

Percent

Number

Percent

Number

Percent

Number

All

0.17

2252

0.152

6121

0.16

5134

Prosecution

0.291

285

0.198

830

0.223

827

Defense

0.15

1845

0.144

5230

0.147

4185

T-Test *= p < .05 ** = p < .01

are only slightly more likely to reverse on appeal—17.3% (2330 votes)—versus civil-law judges who reverse 16.2% of the time (7219 votes) and judges from “mixed legal systems, who reverse approximately 16.5% (5397 votes). When the prosecution raises an issue on appeal, the judges from the mixed systems are most likely to grant their claims (27.9% of 697 votes), while judges from civil-law systems are least likely (18.6% on 986 votes). Judges from common-law states side with the prosecution 23.3% of the time (330 votes). There is little variation apparent in the willingness of judges across all three systems to grant the appeals of the defendants. Perhaps a more crucial indicator of a judge’s willingness to overrule the judgment of a trial chamber is the level of judicial independence enjoyed in his or her home state. Judges who are accustomed to a high degree of autonomy and freedom from political interference may be more comfortable in asserting their independent judgment in decisions. In order to evaluate the propensity of the appellate judges to overturn the decisions of the trial chambers, we use the measure of judicial independence developed by the CIRI Human Rights project, which also categorizes nations according to their level of human rights protections.4 Judicial independence is measured by identifying whether governments exhibit no judicial independence, partial judicial independence, or judicial independence.5 The general differences among the judges in voting behavior on appellant claims based on the level of judicial independence in their home states is not substantively meaningful. Those from the nations with the greatest level of judicial independence support appellant 4. As found at www.humanrightsdata.com 5. According to the CIRI web site at, independent judiciaries possess the following characteristics: 1. It has the right to rule on the constitutionality of legislative acts and executive decrees. 2. Judges at the highest level of courts have a minimum of a seven-year tenure. 3. The President or Minister of Justice cannot directly appoint or remove judges. 4. The removal of judges is restricted (e.g., allowed for criminal misconduct). 5. Actions of the executive and legislative branch can be challenged in the courts. 6. All court hearings are public. 7. Judgeships are held by professionals. Those nations with less than complete judicial independence fall somewhere short of these characteristics in whole or in part.

MEERNIK, ALOISI, KING, AND MCKAY

claims in general 16% of the time (5134 votes), while those in the partially independent states vote thusly 15.2% of the time (6121 votes), and those in the least independent category grant appeals approximately 17% of the time (2252 votes). Judges from judicially independent states support the prosecution and defense 22.3% and 14.7% of the time, respectively. Those hailing from partially judicially independent states support the prosecution and defense 19.8% and 14.4% of the time, respectively, while those from the least independent states support these actors 29.1% (OTP) and 15% (defense) of the time. But though the differences across all appeals are not large, it is interesting to note that judges from the least independent states are most likely to grant the prosecution’s appeals. Whether this is evidence of judicial deference to the prosecution (analogously, the executive branch in the home state) or the result of some other attribute judges from these nations share (perhaps they come from states with fewer human-rights protections), we cannot yet say. Table 5. Appeals Chamber Judges’ Votes on Appeals Issues by Political System 1995–2014 Non Democracy

Democracy Percent

Number

Percent

Number

ll Appeals

0.166**

13,443

0.184

1875

Prosecution

0.224

1826

0.256

222

Defense

0.156**

11,307

0.174

1653

T-Test *= p < .05 ** = p < .01

Table 6. Appeals Chamber Judges’ Votes on Appeals Issues by Human Rights Protections 1995–2014 Bottom 3rd Human Rights Score

Middle 3rd Human Rights Score

Top 3rd Human Rights Score

Percent

Number

Percent

Number

Percent

Number

All Appeals

0.172

1845

0.152

7214

0.162

4448

Prosecution

0.385

161

0.2

1140

0.221

641

Defense

0.15

1623

0.141

5830

0.152

3807

T-Test *= p < .05 ** = p < .01

Next examined is whether the type of political system judges work within exercises any effect on their voting behavior. Do judges from democratic states grant more appeals as they are familiar with the open political systems in which executives and legislatures are freely chosen and can be challenged? Are those judges coming from undemocratic states more accustomed to working in a politically and legally circumscribed environment and thus less willing to grant defense appeals? We do not find any substantial differences between judges from democratic versus judges from non-democratic states. The democratic group supports appeal requests 16.6% of the time in all cases (13,443 votes), 22.4% of the time when the OTP raises an issue (1826 votes), and 15.6% of the time when the defense appeals (11,307). Judges from non-democratic states grant appeals 18.4% of the time (1875 votes); they support the OTP 25.6% of the time (222 votes) and grant the defense appeals 17.4% of the time (1653 votes). The subsequent focus examines judicial voting behavior according to the level of humanrights protection found in a society. We find some of the same trends in evidence we saw earlier regarding the proclivity of judges from states with less judicial independence to grant

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

appeals. First, we use the CIRI data on human rights involving the right to personal integrity. The CIRI scale ranges from 0 to 8, where “0” indicates no government respect for human rights and “8” indicates full government respect for such rights.6 There do not appear to be any meaningful differences when we examine all appeal claims. Nor do we see much variation in voting on issues raised by the defense. However, we do see noticeable differences in voting behavior on appeals brought by the prosecution. Those judges from states that fall in the top one-third of states with the highest level of human-rights protections and those judges who are in the middle tier of states support the OTP 22% and 20% of the time, respectively. However, judges from the states that fall into the bottom tier with the most questionable human-rights records grant the OTP appeals 38.5% of the time. Judges from more authoritarian governments, as measured by their level of judicial independence and human rights records, accord greater deference to the prosecution claims. We suggest that such trends may indicate that these judges are more accustomed to a strong executive branch to which they grant significant political and judicial deference. Next, we examine the issues involved in the appeals. In the judgments of the appellate chambers the matters raised in the defense and OTP appeals can be identified according to their place in the outlines the tribunals use to organize their judgments. The appeals chambers typically use a standard outline format to classify the issues raised on appeals. At the level of the Roman numeral in the outlines, one finds the issues organized at their most general level, which is typically not useful for determining the issues involved. For example, the first Roman numeral may be listed as “I. Defendant’s Appeals” or “II. The Crimes.” At the next outline level (capital letters) we almost always find the issues at stake expressly stated, such as “I.A. Alleged Errors of Fact” or “II. B. Command and Control,” albeit at still a fairly general level. Subsequent levels of the outline (e.g., Arabic numerals and lowercase letters) break issues down even further to describe, for example, how the defendant or the OTP is appealing alleged errors of fact involving a particular military attack, or the defendant’s role in aiding and abetting. The outlines sometimes break the appeals down into fairly specific claims either party may be appealing. We consider each of these specific grounds as separate issues on appeal, as long as there is an identifiable record of judges voting on these appeals. For purposes of coding the issues and analyzing the data, however, we rely on the issue codes contained in the second and third levels of the outline. As explained above, at the first, Roman numeral level of the outline, the descriptions are typically quite general and do not allow us to identify what the issues are. At the most specific levels of the outline, the issues are broken down into fairly fine distinctions (e.g., alleging a factual error involving a very specific incident and a defendant’s claim that at this particular town on this particular day, he Table 7. Appeals Chamber Judges’ Votes on Appeals Issues by Issues, 1995–2014

All Appeals Votes

OTP Appeals

Defense Appeals

Issue

Percent

Number

Percent

Number

Percent

Number

Procedural

0.157

5895

0.265

400

0.154

5340

Command & Control

0.182

2315

0.231

750

0.085

1545

Joint Criminal Enterprise

0.203

1038

0.334

203

0.155

760

War Crimes

0.103

1465

0.545

0.085

1410

Crimes Against Humanity

0.236

1310

0.495

115

0.193

1150

Genocide

0.208

225

0.184

0.218

160

Sentencing

0.171

2825

0.264

435

0.155

2375

6. See http://www.humanrightsdata.org/documentation/ciri_variables_short_descriptions.pdf for further details on the component measures that make up the 0–8 scale.

MEERNIK, ALOISI, KING, AND MCKAY

did not engage in the type of conduct for which he has been found guilty because he wishes to question the veracity of a particular witness), which are not always as useful to delineate the more fundamental issues at stake. Additionally, the judges often do not rule on such specific claims but do rule on a broader category to which these specific grounds of appeal attach. While we code each and every level of the issues in the outlines, we utilize the issue coding at the second and third levels to categorize the issues we use in this analysis. In all, we find approximately 150 issues raised on appeal, which we have categorized for this analysis into several larger groups. The issues we examine below are 1) procedural issues, 2) command and control, 3) joint criminal enterprise, 4) war crimes, 5) crimes against humanity, 6) genocide, and 7) sentencing. Across these broad issue categories, we find significant differences in the judges’ votes to grant appeals. Appeals are least likely to be granted in general on issues pertaining to war crimes (10.3%) and most likely to be granted on appeals involving crimes against humanity (23.6%). Interestingly, in terms of their legal “age,” war crimes have the lengthiest and most extensive history of adjudication, with international law dating back to the Hague Conventions of 1899 and 1907 and the more recent and well-known Geneva Conventions of the post–World War II period. Thus, such law might be considered relatively more stable and developed in comparison to jurisprudence on crimes against humanity, which has never been codified, with the exception of the tribunals’ statutes. The prosecution, however, is apt to find significant success on both such issues. The OTP wins on war crimes appeals 54.5% of the time and succeeds on crimes against humanity issues 49.5% of the time. They are least likely to be met with success on genocide related appeals (18.4%) and command and control (23.1%) issues. For the defense, their best chances of success are on issues involving genocide and crimes against humanity (21.8% and 19.3%, respectively). It would seem the defense is most likely to win an argument in those areas of international law that have not been adjudicated as often at the international level. Indeed, ICTR handed down the first-ever conviction for genocide at an international court in 1998. Defendants are least likely to be granted an appeal when the issue pertains to either command and control or war crimes (8.5%). The more settled the issue, the more likely the prosecution wins, while the defense is generally much less likely to win on appeal in general. When the defense does win, it tends to be on laws that have been less extensively adjudicated. Conclusion Our goal with these data is to begin an in-depth research agenda regarding judicial decision making at the international tribunals and the development of international law. The first step in this agenda is to more fully explore judicial ideology and whether differences in the voting behavior of judges can be explained with reference to the characteristics of their home states and the nature of the issues under dispute. In so doing, our first goal in this first part of the research agenda is to determine the extent to which such judicial decision-making dimensions exist. We believe the findings presented here provide significant evidence to suggest that judges’ voting patterns can be described as 1) being more likely to defer to the trial chamber (what we termed an “institutionalist” dimension), 2) being more likely to uphold the appeals of the OTP (what we termed as “deference to the prosecution”), and 3) being more likely to uphold the appeals of the defense (what we termed “concern for defendants’ rights”). While most judges tended to cluster around the average in each category, there were outliers in either direction in all three dimensions. Our next steps in this research agenda are to continue development of the theoretical expectations regarding these three dimensions and to determine through multilevel equation modeling how the judges array along these dimensions across the various issue categories. This will help us better understand the nature of the differences in opinion on legal matters among the judges and evaluate the extent to which support for the trial chambers, prosecution, or defense may depend on the issue-centered ideological disputes.

JUDICIAL VOTING BEHAVIOR AT THE APPEALS CHAMBERS

We also believe there is merit to further exploring how judges’ backgrounds may influence their voting behavior. This has proven to be a fruitful area of research in general involving international courts (e.g., Voeten 2007). We believe our finding that judges from more repressive states and those from nations where there is less judicial independence and democracy indicate that governing practices may shape judicial ideology. Future research should further investigate whether professional background, election mechanisms, and other such extra-judicial factors influence decision making. Our next step is to develop a theoretically informed model of judicial decision making. The results presented here offer some tantalizing clues regarding the potential determinants of judicial voting, such as the influence of the political system and the freedoms in the home state. These potentially relevant factors must be justified with reference to literature existing in other areas of international and domestic judicial decision making. Additionally, we must also consider the impact of other factors, such as judges’ professional experience, appeal court dynamics, coalition building, and the role of the presiding judge. As well, we must determine what are the truly important issues raised in appeals that have the potential to significantly alter the ultimate outcomes of verdicts and punishments. Fascinating and substantial work lies ahead as the ad hoc tribunals begin to finish their business and attention shifts more to the International Criminal Court. REFERENCES Darcy, Shane and Joseph Powderly (eds.). 2010. Judicial Creativity at the International Criminal Tribunals. Oxford: Oxford University Press. Drumbl, Mark A. 2007. Atrocity, Punishment and International Law. Cambridge, UK: Cambridge University Press. Ellis, Mark. 1997. “Achieving Justice Before the International War Crimes Tribunal: Challenges for the Defense Counsel.” Duke Journal of Comparative and International Law 7(Spring): 519–37. Jodoin, Sebastien. 2010. “Understanding the Behavior of International Courts.” Journal of International Law and International Relations 6:1–34. Johnson, Scott T. 1998. “On the Road to Disaster: The Rights of the Accused and the International Criminal Tribunal for the Former Yugoslavia.” International Legal Perspectives 10(Spring):111–92. King, Kimi and James Meernik 2011. “A Distant Court: Assessing the Impact of the International Criminal Tribunal for the Former Yugoslavia” in The Legacy of the International Criminal Tribunal for the Former Yugoslavia, edited by Göran Sluiter, Bert Swart, and Alexander Zahar. Oxford: Oxford University Press. Meernik, James. 2011. “Sentencing Rationales and Judicial Decision-Making at the International Criminal Tribunals.” Social Science Quarterly 92(3):588–08. _____. 2007. “The Evolving Application of International Law: Insights from the Appeals Chamber of the Ad Hoc International Tribunals.” International Criminal Law Review 7:117–32. Meernik, James D. 2003. “Equality of Arms? The Individual vs. the International Community in War Crimes Tribunals.” Judicature 86(6): 312–19. Meernik, James, Rosa Aloisi, and Jennifer Ding. 2012. “The Influence of Home Country Characteristics on Judicial Decision Making at the ICTY.” Judicature. 95(4) 171–76. Meernik, James, Kimi King and Geoffrey Dancy. 2005. “Judicial Decision Making and International Tribunals: Assessing the Impact of Individual, National and International Factors.” Social Science Quarterly 86:683–703. Posner, Eric and Miguel de Figueiredo. (2005) “Is the International Court of Justice Biased?” Journal of Legal Studies: 34.

MEERNIK, ALOISI, KING, AND MCKAY

Segal, Jeffrey A. and Harold J. Spaeth. 2002. The Supreme Court and the Attitudinal Model Revisited. Cambridge: Cambridge University Press. Segal, Jeffrey A. and Harold J. Spaeth. 1993. The Supreme Court and the Attitudinal Model. Cambridge: Cambridge University Press. Sikkink, Kathryn. 2011. The Justice Cascade. New York: W.W. Norton & Company. Sluiter, Göran, Bert Swart, and Alexander Zahar (eds.). 2011. The Legacy of the International Criminal Tribunal for the Former Yugoslavia. Oxford: Oxford University Press. Terris, Daniel, Cesare Romano and Leigh Swigart. 2007. The International Judge: An Introduction to the Men and Women Who Decide the World’s Cases. Brandeis University Press. Voeten, Eric. 2008. “The Impartiality of International Judges: Evidence from the European Court of Human Rights.” American Political Science Review 102(4): 417–33.

Plurilateralism: The Key to Saving the World Trade Organization and Boosting Global Trade Liberalization? by Michel Dupuy, University of Bordeaux In1 the field of international trade negotiations, the key aspect of the decades 2000 and 2010 has been the failure of the Doha cycle and the continued expansion of bilateral and regional trade agreements. This article shows, in light of past and present experiences with plurilateral agreements, that plurilateralism, now in the throes of rehabilitation, may be an effective way to boost international trade relations. At the end of our analysis, however, the clear fact is that at the present time, not all the necessary conditions are there and ready for plurilateralism to be taken on board and shared by all members of WTO. Future plurilateral agreements will, in particular, need to focus on themes likely to interest the lesser developed countries, most notably the poorest, which to date have been sidelined from this form of organization in international trade. More generally, the access of these countries to plurilateral agreements will have to be made easier.

Introduction After more than ten years of global trade negotiations conducted as part of the Doha cycle, the 160-member countries of the World Trade Organization (WTO), meeting at the ninth Ministerial Conference held in Bali in December 2013, signed an initial wide-ranging agreement. The â&#x20AC;&#x153;Bali packageâ&#x20AC;? comprises three sections: agriculture, with the authorization to build public stocks for food security purposes; the facilitation of trade aimed at simplifying and speeding up customs procedures; and measures for Least Developed Countries (LDCs), including preferential treatment and market access. However, we should not overestimate the scope of this agreement, which ultimately represents only 10 percent of the original Doha program. Indeed, the fundamental questions relating to freer trade for goods and services remain. Faced with the difficulties encountered with global trade negotiations, many countries have now deferred Regional Trade Agreements (RTAs), e.g., NAFTA, MERCOSUR, ASEAN, etc. As of June 2014, WTO had recorded 585 RTAs, of which 379 were in effect. On average, each WTO member country was party to more than three RTAs. The most recent initiatives include the draft Transatlantic Trade and Investment Partnership (TTIP) and negotiations for the Trans-Pacific Partnership (Table 1). Even though RTAs are recognized by WTO, regionalism nonetheless remains a threat to liberalizing global trade in that it could lead to fragmenting the whole international trading system (Bhagwati 1995). In this context, the signing of plurilateral agreements that assembles a limited number of WTO member countries is today considered by some as one of the ways forward to emerge from the impasse of WTO trade negotiations while at the same time form a defense against regionalism. In the field of international trade, a plurilateral agreement designates more specifically an agreement bringing together a limited number of countries (at least three) that 1. I wish to thank the two anonymous reviewers for helpful comments on a previous draft. All remaining shortcomings and errors are my own.

JIOS, VOL. 6, ISSUE 1, 2015

DUPUY

Table 1. Projects of Mega-Regional Trading Blocs Project

Countries Taking Part in Negotiations

Objectives

Trans-Pacific Partnership (TPP)

Brunei, Australia, Canada, Chile, United States, Malaysia, Mexico, Japan, New Zealand, Peru, Singapore, Vietnam

To facilitate market access by scrapping tariff and non-tariff barriers to trade and investment. To facilitate the development of production and supply chains between member countries. To encourage trade and investment in the sector of innovative products and services.

Transatlantic Trade and Investment Partnership (TTIP)

United States, European Union

To fully suppress customs duties, with the exception of “sensitive products” that are essentially agricultural commodities. To reduce non-tariff trade barriers. To harmonize technical standards and rules. To provide greater liberalization of services. To provide easier access to public sector markets. To liberalize investments.

Source: author

have agreed on common rules in order to facilitate and work for fairer trade in a particular field.2 According to this definition, plurilateralism is akin to a form of “mini-lateralism.” It should also be made clear that plurilateralism sets itself apart from firstly universalism, advocated by the WTO, which is charged with organizing the global liberalization of trade, and secondly regionalism; plurilateral trade agreements (PTAs) are sectorial or issue-based while regional trade agreements are country-based (Nakatomi 2013). An RTA covers many fields, ranging from the trade of goods and services to investment, from intellectual property rights to technical standards and rules, whereas a plurilateral agreement concerns one specific field. The return to favor of plurilateralism starkly contrasts with its deletion prior to the signing of the Uruguay Round Agreements. As we were reminded by Jean-Marc Siroën, (2010:13), “In the name of multilateralism, Uruguay Round negotiators were keen to reduce, not to say delete, plurilateral agreements that engaged signatory countries only.” After the conclusion of the Uruguay Round cycle, there remained, nonetheless, four plurilateral agreements inherited from the Tokyo Round (cf. appendix): the Agreement on Trade in Civil Aircraft, the Agreement on Government Procurement (GPA), the International Dairy Agreement, and the International Bovine Meat Agreement. These plurilateral agreements were figured in the annex to the WTO establishing agreement to conform with a provision of Article X-9, which stipulates that “the Ministerial Conference, upon the request of the Members party to a trade agreement, may decide exclusively by consensus to add that agreement to Annex 4.”3 A fifth plurilateral agreement, the Information Technology Agreement (ITA), was adopted prior to the Ministerial Conference of Singapore in 1996. Today, only three plurilateral agreements remain: the Agreement on Trade in Civil Aircraft, the GPA, and the ITA. The other two agreements were revoked in 1997 in conformance with a provision of Article X-9 of the agreement establishing the WTO, which stipulates that “the Ministerial Conference, upon the request of the Members party to a Plurilateral Trade Agreement, may decide to delete that Agreement from Annex 4.” At the time, the countries that had signed these agreements felt the sectors concerned would be handled more effectively within a global framework. 2. More generally, plurilateralism may be defined as “a shared interest among a limited number of governments that brings these together for interconnection” (Cerny 1993). 3. This provision makes reference only to plurilateral agreements in the strict sense of the term. Alongside these agreements, we find plurilateral commitments focusing on the multilateral framework of the WTO. These include the “Financial Services Agreement” and the “Basic Telecommunications Agreement.”

PLURILATERALISM: THE KEY TO SAVING THE WORLD TRADE ORGANIZATION

A number of recent reports and studies call for plurilateralism (Sutherland Commission 2006; Warwick Commission 2007; the Bhagwati-Sutherland Report 2011; WTO 2013). Thus, the Warwick Commission proposed to ease the obligation-related single undertaking, which stipulates that every item of a global round of negotiations conducted by WTO is part of a whole and indivisible package and cannot be agreed upon separately (“nothing is agreed until everything is agreed”). The commission also proposed to introduce additional elements of “variable geometry,” authorizing the implementation of partial trade agreements on a plurilateral basis between several WTO member countries. For its part, the stance defended in the report by Jagdish Bhagwati and Peter Sutherland is slightly more qualified: “While tariff reductions and the dismantling of non-tariff barriers can of course be achieved in other negotiating formats, the multiplier effect of a multilateral agreement is considerably higher” (Bhagwati and Sutherland 2011: 10).4 Recent academic literature on this subject endeavors primarily to make a general cost-benefit analysis of plurilateralism. Thus, Raymond Saner (2012), Michitaka Nakatomi (2013), Peter Draper and Memory Dube (2013), Bernard M. Hoekman and Petros C. Mavroidis (2013), and Michael Trebilcock (2014) produce a cost-benefit analysis of plurilateralism compared with other forms of trade organization, i.e., universalism and regionalism. Other authors (Scott and Wilkinson 2012; Woolcock, 2013) showcase the benefits and the drawbacks of plurilateralism for emerging countries. The principal conclusion from these works is that plurilateralism may constitute a credible alternative in international trade organization, but it is not a panacea. As an extension to earlier works, the present article, in light of past and present experiences with plurilateral agreements, sets out to ascertain the extent to which the plurilateral path of trade liberalization is likely to give a new boost to global trade liberalization. The ongoing debate over plurilateralism is not solely academic, as it also focuses on concrete projects, including the liberalization of the trading of services and the liberalization of environmental goods. About twenty WTO member countries, the “Really good friends of Services,” effectively opened trade negotiations in the spring of 2013 with the aim of reaching a plurilateral services agreement (known as the Trade in Services Agreement or TISA). More recently, fourteen WTO member countries began negotiations with a view to liberalize environmental goods. This article is placed in a historical perspective to study the links between plurilateralism and the global liberalization of trade. From my study of the main plurilateral agreements, current and past, it emerges that the signing of nondiscriminatory plurilateral agreements, ratified by WTO and involving a large number of countries (a “critical mass” of signatories), boosts international trade. I show, however, that the development of plurilateralism is not without risk for the system of international trade. I then conclude that future plurilateral agreements will need to focus on themes likely to interest the greatest number of countries, including lesser developed countries, whose involvement will have to be encouraged and facilitated by, among other things, administrative and financial assistance. This paper is organized as follows: The first section presents the forms of plurilateralism. The second section shows plurilateralism can be an effective instrument to boost global trade liberalization and save WTO. The third and final section examines the potential risks linked to the development of plurilateralism. The Forms of Plurilateralism Plurilateral trade agreements come in forms that fan out along a number of criteria. The first relates back to the positioning of these agreements with respect to WTO. A plurilateral 4. The plurilateralism debate falls within a larger debate concerning the reform of WTO governance. Amongst the proposals to modify the current rule of consensus is a plan whereby certain WTO members, as an extension to the Sutherland report (2006), may elect to contract more or fewer obligations. In a way, this measure would be tantamount to officialising “de facto variable geometry” resulting from the forming of coalitions within WTO. According to Lawrence (2006), the option consisting in creating “clubs” within WTO (the so-called “club of clubs” option) would help provide an effective response to the diverging interests of WTO members.

DUPUY

agreement may be reached within or outside the framework of WTO. The advantages of the first option are to facilitate the future universalization of the agreement and to better protect the rights of WTO member countries that do not sign the agreement, thanks mostly to the application of WTO dispute settlement procedures. Article II.3 of the agreement establishing WTO states that “plurilateral trade agreements do not create either obligations or rights for members that have not accepted them.” The agreements inherited from the Tokyo round and the ITA are agreements of this type (cf. annex). The second option is less constrictive. A plurilateral agreement reached outside the framework of WTO cannot claim to become an “international standard.” The legal and political clout of plurilateral agreements reached outside WTO is weaker than that of agreements concluded within the organization. In other words, the former come under nonbinding international law and the latter under binding international law. But plurilateral agreements not drawn up in the framework of WTO are less transparent, inasmuch as WTO members who are not contracting parties to an agreement of this type do not take part in negotiations. The anti-counterfeiting trade agreement (ACTA) is to date the only plurilateral agreement drawn up outside the framework of WTO (cf. annex). The parties now engaged in discussions over a plurilateral agreement for services will have to choose one of the two options mentioned above. We should note that the option of negotiating the agreement within the WTO framework would comply with Article 5 of the GATS.5 The second criterion for the differentiation of plurilateral agreements relates to the application, or not, of the MFN (most-favored nation) clause. A plurilateral trade agreement reached within the framework of WTO, under which the commitments subscribed to by signatory nations benefit all WTO members, including those who have not signed it, helps to avoid trade distortions. The ITA is an example of agreements based on the MFN clause. Inversely, a plurilateral agreement negotiated within WTO and which does not extend its benefits to countries that have not signed up helps avoid the problem of free-riders. The GPA, the International Dairy Agreement, and the International Bovine Meat Agreement are examples of this type of agreement. The experience gained with plurilateral trade agreements reached to date shows how extremely diverse they can be. Given this experience, we see that certain issues, such as the integration of plurilateral agreements within the scope of WTO, the application of the MFN clause, or the utilization of the dispute settlement procedure, are crucial when seen in the perspective of making plurilateralism an exit route from the deadlock in which current global trade negotiations now find themselves. Plurilateralism: An Effective Instrument to Boost Global Trade Liberalization The theory according to which plurilateralism might constitute “a solution to the impasse of WTO/DDA as well as a basis for future trade agreements within the WTO context” (Saner 2012: 25) is gaining ground, especially in academic circles (WTO 2013). The study of plurilateral agreements, past and present, will help to validate this thesis. This study will show more precisely that several of these agreements have promoted or continue to promote the global liberalization of trade. This is particularly true with agreements signed within the framework of WTO or its predecessor, the GATT. To date, the most compelling experience regarding plurilateral agreements that have spawned the adoption of universal trade rules is that of the Tokyo Round Codes. Set up in 1980, these were codes of conduct of a plurilateral nature engaging primarily OECD countries and aimed at lowering nontariff barriers (subsidies, anti-dumping duties, import licenses, technical standards, and rules) and at liberalizing certain sectors (bovine meat, dairy sector, civilian aircraft and public sector markets). At the Uruguay Cycle, several of these codes were transformed into universal trade rules accepted by all WTO members. Only four of them remained plurilateral (cf. page 48). 5. Article 5 stipulates that the GATS ”shall not prevent any of its Members from being a party to or entering into an agreement liberalizing trade in services between or among the parties to such an agreement.”

PLURILATERALISM: THE KEY TO SAVING THE WORLD TRADE ORGANIZATION

The ITA and GPA experience has also shown that with the beneficial effects of plurilateral agreements (increased trade between signatory member countries and potentially between agreement members and third-party countries whenever the MFN clause is applicable, improved competitiveness, or stimulation of economic growth), there is a galvanizing effect on the global liberalization of trade, a dynamic which, more often than not, results in new members joining existing agreements (here we find the “domino effect” evidenced by Baldwin [1993] to characterize the dynamic of regional trade agreements). Thus, the ITA, originally signed by twenty-nine countries, now has seventy-four members. Likewise, the GPA, which was initially signed by twenty-two countries, now includes forty-three countries. Using the experience of the GPA and the ITA, we are going to evidence more specifically the mechanisms through which the beneficial effects of plurilateral agreements translate to an increase in trade worldwide. The main objective of the GPA is the mutual opening of government procurement between agreement members. After several rewrites of the initial agreement dating back to 1981, the forty-three current members have opened procurement activities to international competition for an estimated annual value of $1,700 billion, i.e., 2.3 percent of world GNP. Given the GPA’s discriminatory nature, only the suppliers from signatory countries enjoy the benefits of the agreement. In opening up government procurement to international competition, the GPA has stimulated not only the trading of goods and services between members but also the sales of foreign-affiliate firms located in each member country. This fact applies in particular to European countries. For instance, a study conducted by the European Commission (2011) shows that in EU countries, all members of the GPA, direct cross-border government procurement (contracts awarded by the government procurement agents of one EU member country to companies from another member country), represent 3.6 percent of the total value of European government procurement open to tenderers from countries that are party to the GPA, that is nearly €26 billion p.a. in the form of intra-EU trade in goods and services. According to this same study, the percentage of cross-border government procurement rises to 17.5 percent when considering indirect cross-border government procurement, contracts won by foreignbased firms through the intermediary of subsidiaries operating in EU member countries, and even to over 30 percent when taking into account imports from local distributors or agents who work on behalf of an adjudicating entity. These results clearly show that the GPA has facilitated access to procurement “not merely in terms of cross-border sales by foreign-based companies but also, very much, in terms of the sales of foreign-affiliated firms located within the territories of GPA Parties” (Anderson et al. 2011: 20). Furthermore, the GPA has helped to enhance the transparency of government procurement and has strengthened competition in these markets, leading to lower prices and higher quality. The result has been savings and/or additional public investments thereby promoting economic growth and employment. Stronger economic growth has then stimulated trade by all members of the GPA via an increase in their imports. In the end, we have seen the set-up of a virtuous circle between economic growth and international trade. In this way, the implementation of European directives affecting government procurement, which, let us remember, integrate the prescriptions of the GPA, have enabled EU public entities to achieve savings and/ or additional government spending of over €20 billion out of the €420 billion of government procurement published annually at the European level. This has generated increases in terms of GNP, employment, and commercial trade of between 0.08 and 0.25 percent after one decade (European Commission 2011). Econometric studies have confirmed that the GPA has boosted all trade in goods and services among member countries. In a study focusing on twenty OECD countries (seventeen of whom are GPA members) between 1996 and 2008, Hejing Chen and John Whalley (2011) use

DUPUY

a gravity model6 to evidence the effects of the GPA on bilateral trade for these countries. The findings show that GPA exerts a positive influence on the total trading of goods and services for the countries under consideration. For example, bilateral exports of merchandise between OECD countries are 31 percent higher than they would be without the GPA, while the bilateral exports of services are 145 percent higher than they would be if the GPA did not exist. Actually, GPA has a stronger effect on the trading of services than on the trading of merchandise. One possible explanation is that a significant share of government procurement in GPA member countries is concentrated on services. We have just seen that GPA promotes the globalization of trade in goods and services. Now what about the ITA? In abolishing tariffs on information and technology (IT) products, ITA has promoted the trade of these goods not only between member countries but also on a world scale. For instance, a study made by WTO (2012) for the fifteenth anniversary of ITA indicates the existence of a positive correlation between increased trading in IT products—in 2010, ITA participants accounted for 96 percent of exports worldwide amounting to $1,400 billion, almost three times their 1996 value—and the massive elimination of tariffs resulting from the agreement. Through the application of the MFN clause, many emerging countries that are not members of ITA have also benefited from this plurilateral agreement. This is particularly the case with Mexico and Brazil. For instance, Mexican exports of IT products rose by 295 percent between 1996 and 2010, and Mexico has now become the ninth leading worldwide exporter of IT products. Additionally, on the importation front, Mexico has unilaterally set up an “ITA Plus,” which has abolished tariffs on inputs and final goods in the electronics and IT sectors. The fall in price of imported inputs has now given Mexican producers and exporters a distinct competitive advantage. Lastly, Mexico has also benefited from concessions granted under the NAFTA, in the knowledge that the major part of the country’s exports of IT products are bound for the North American market. Over the same reference period of 1996–2010, Brazilian exports of IT products have almost quadrupled. Over and above observing the correlation between scrapping tariffs for IT products under the effect of ITA and growing the worldwide trade of these items, we ought to specify the mechanisms through which this trade agreement has promoted world trade (not only IT products but also all goods and services). We should point out, as WTO noted in its World Trade Report of 2007, that it is unfortunately impossible to estimate exactly the overall incidence of ITA on the world trade of IT products. This is on account of there being no specific starting date marking the full implementation of all commitments made under ITA and because of the significant varying exchange rates and prices of IT products. (In order to neutralize price effects, trade flows must be expressed in real terms. However, for this product category we still do not dispose of data concerning real commercial flows.) ITA has influenced international trade through direct and indirect effects. The direct effect relates to the increasing trade of IT products among agreement signatories resulting from scrapping tariffs. To estimate the scope of this direct effect, we have used a simple method based on price elasticity with respect to demand for imports. The average rate of tariffs applied by all the original participants prior to ITA was 6 percent. At the end of 1996, the imports of IT products among future ITA members came to $520 billion. And price elasticity with respect to demand for imports with ITA member countries was 1.3. If we consider that tariffs were scrapped in one fell swoop, the increase in the initial demand for the importation of IT products with ITA member countries reached almost $41 billion (1.3 x 6 percent x 520). The results of estimations of effects of ITA on the trading of IT products amongst member countries, made using gravity models, confirm that ITA has triggered the creation of traffic (Bora and Liu 2006; 6. Inspired by Newton’s law of gravity, the gravitational model implies the idea that bilateral trade flows are proportional to the product of the GNPs of the two countries considered and inversely proportional to the geographic distance between them. Control variables for the effects of regional and/or plurilateral agreements are usually added.

PLURILATERALISM: THE KEY TO SAVING THE WORLD TRADE ORGANIZATION

Mann and Liu 2007; Sato 2014). Thus, according to Catherine L. Mann and Xuepeng Liu, a non-ITA WTO member imports on average 14 percent more from WTO members if it joins ITA (the effect is greater when the country adhering to ITA is a developing country).7 Alongside the previous effect, ITA has also exerted indirect effects on international trade via, the diffusion of IT goods that have become more affordable owing to the scrapping of tariffs and to the fall in the cost of imported IT inputs. These two phenomena have then helped to increase the productivity and competitiveness of firms in ITA member countries, hence stronger economic growth and an increase in the general exports of goods and services. Empirical studies have shown that the contribution of information and communication technologies (ICTs) to the growth of the global productivity of factors and to economic growth is high in most countries. Thus, according to Khuong Vu (2005), a rise of 1 percent in ICT capital stock would, on average, add around 0.45 of a percentage point to growth in GNP. From the previous analysis, the understanding emerges most clearly that GPA and ITA have contributed to expanding world trade for goods and services. To finish, we should note that, as opposed to previous agreements, plurilateral agreements reached outside WTO constitute a threat for the global trading system. In this way, ACTA (the goal of which is to set up international standards in the fight against counterfeiting in order to upgrade the protection of intellectual property) comprises a certain number of provisions incompatible with the multilateral agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS), the WTO accord that guarantees intellectual property rights. Thus, implementing ACTA restricts the diffusion of generic drugs via the reinforcement of frontier controls (some generic drugs might effectively be assimilated with counterfeit products owing to the relatively broad definition of the notion of counterfeit used by ACTA), whereas the TRIPS agreement enables poor countries to import, under certain conditions, generic versions of patent drugs. At the end of our analysis, we may conclude that plurilateral agreements signed under the aegis of WTO and extending their benefits to WTO members who are not party to these agreements promote the global liberalization of trade. The path to be taken at the negotiating table for future plurilateral agreements is clearly marked out. It is the path taken, in particular, by WTO members engaged in discussing an agreement over liberalizing services planning to reach a plurilateral agreement but with a global purpose. To this end, a good number of these countries feel that GATS should form the bedrock of the future plurilateral agreement. As emphasized by the European Commission (2013:2), “By staying close to the GATS, it could be easier to convince countries that were active in the DDA negotiations to join the initiative, either during the negotiations or later on.” Besides services, several other areas may be at the center of plurilateral agreements. We can identify more precisely six such themes: currency undervaluation, climate and energy, “zero-for-zero” tariffs,8 environmental goods, world value chains, and state-owned and statesupported enterprises. Table 2 presents the major characteristics of projects and recent proposals for the creation of plurilateral trade agreements. Negotiations to conclude a plurilateral agreement for environmental goods began in July 2014. In view of the previous analysis, it is clear that well-led and well-conceived plurilateral agreements can help strengthen global trade liberalization. Open plurilateralism can serve the cause of universalism, although its development is not without risk for the system of international trade. The Potential Risks of Plurilateralism Notwithstanding the previous benefits, plurilateralism has several drawbacks. The first relates to distributing gains between the countries that sign up to plurilateral agreements. 7. The customs duties applied by developed countries to IT products were low even before the implementation of the ITA. For these countries, the scrapping of customs duties when the agreement came into effect did not lead to any high increase in the importation of IT products. 8. The expression “zero-for-zero” means that the signatories of the agreement agree to totally do away with tariffs (i.e., zero consolidation).

DUPUY

Table 2. Major characteristics of projects and proposals for the creation of plurilateral agreements Subjects Services

Situation end December 2014 Negotiations ongoing between the 21 “Really good friends of services”

Objectives To reach a plurilateral agreement with multilateral designs (the agreement would be compatible with AGCS) aimed at fully liberalizing the trading of services between members. A desire to reach a comprehensive agreement; negotiations will focus on all services, including IT and trade services, financial services, and corporate services. Exceptions would concern solely services linked to the exercising of national sovereignty (military contracts, for example). Safeguard measures would be provided for whenever a home service sector might be seriously affected by foreign competition.

Environmental Goods

Negotiations ongoing between 14 WTO member countries

To liberalize the trade of environmental goods: —Lowering customs duties for environmental goods (54 goods pertaining to the areas of green growth have been identified) —Reducing nontariff barriers

Climate/Energy

A proposal likely to interest the United States, Canada, Mexico, the EU, Japan, Australia, and New Zealand

To take account of questions relative to the link between international trade and the environment (Hufbauer 2012): —Setting rules for authorized subsidies to encourage green energies and a reduction in greenhouse gas emissions —Setting rules for carbon adjustments at frontiers —Limitation of local content obligations for green energy projects financed directly or indirectly by the state —The elimination of barriers on the importation of goods and services that help to reduce greenhouse gas emissions —An obligation to adopt standards for greenhouse gas emissions that will not prove discriminatory toward the importation of goods and services

Zero-for-Zero Tariffs

A proposal likely to interest around 20 countries

To pursue liberalization on a reciprocal basis regarding tariffs for the following products: chemicals, electronic goods, and environmental goods.

State-Owned and State-Supported Enterprises

Proposal

To improve the transparency of operations conducted by public corporations and government-supported companies: —The publication of annual accounts in compliance with international standards (these accounts should clearly indicate subsidies and other forms of aid paid by governments) —An obligation to respect rules concerning the opening of publicsector markets as scheduled by WTO’s GPA

World Value Chains

Proposition

The necessity to support the globalization of economic activity: —Faced with the proliferation of RTAs and the resultant overlapping of the original trading rules, there is a need to harmonize and coordinate the procedures to which companies within world value chains are submitted —A plurilateral agreement could serve as a basis for the definition of multilateral rules.

Currency undervaluation

Proposition

To make good the inadequacies of the multilateral systems proposed by IMF and WTO (WTO has no competence over exchange rates and surveillance by IMF has been weak to date). To judge the merits of accusations of monetary dumping and, if relevant, define the appropriate measures that countries victims of this unfair trading practice could adopt.

Source: author

From existing plurilateral agreements, there effectively emerges the fact that gains are unfairly distributed between signatory countries. The unfair distribution of gains between the member countries of plurilateral agreements, in keeping with everything learned from

PLURILATERALISM: THE KEY TO SAVING THE WORLD TRADE ORGANIZATION

theories of international specialization, stems from differences in comparative advantages and size of country. For instance, according to the results of our statistical study aimed at estimating the revealed comparative advantages for the main ITA member countries (Table 3), the main beneficiaries of the agreement in terms of exports were the countries holding the greatest comparative advantages in this field (China, Korea, Malaysia, and Singapore). By contrast, the countries with comparative disadvantages generally recorded a drop in exports. This is particularly the case of the European Union (over the period between 2005 and 2010) and Canada. In the light of previous results and in keeping with the theory of comparative advantages, we may consider that the U.S., the EU, and Japan, which today boast significant comparative advantages in the services sector, should therefore be the main winners from the signing of a plurilateral agreement in this sector (according to Hufbauer, Jensen, and Stephenson [2012], with the said entities recuperating 60 percent of the total rise in exports, table 4). Likewise, emerging countries, which have comparative advantages in the production of electronic goods, should profit largely from a plurilateral agreement in that particular area (Table 5). Table 3. Comparative advantages in the production of IT products and growth in the exportation of the same from the main exporting member countries of the ITA Country

Revealed Comparative Advantage (Balassa Method) 1996, 2005, 2010

Percentage of Variation in Exportations 1996–2010, 2005–2010

Canada China EU Japan Korea Malaysia USA Singapore

0.59, 0.36, 0.28 0.80, 2.46, 2.96 0.69, 0.69, 0.53 1.86, 1.56, 1.25 1.80, 3.16, 2.40 2.56, 3.85, 3.53 1.57, 1.10, 1.00 2.75, 4.01, 3.55

-22.6, -28.9 +3320, +106.9 +57.3, -19.7 +3.2, -14.4 +282.4, +25 +178.8, +7.7 +23, 0 +221.5, +17.9

Note: Balassa index= (Country IT exports/Country total exports)/ (World IT exports/world total exports) If the index is superior (respectively inferior) to 1, the country has a comparative advantage (respectively disadvantage) in the production of IT goods. Source: the author’s own calculations based on data supplied by WTO

Table 4. Plurilateral Services Agreement: listed exports and estimated increases ($billions) Exporting Countries

Service Exports to other countries taking part in negotiations, 2008

Total Services Exports, 2008

Estimated Rise in Exports to Other Countries Signing the Agreement*

Australia Canada Chile Colombia European Union Hong Kong Japan Korea Mexico New Zealand Norway Pakistan United States Singapore Switzerland Taiwan Total

26 57 3 0 454 58 116 50 18 5 38 3 339 35 90 20 1312

45 68 11 4 764 92 148 91 18 9 45 4 532 100 78 37 2045

3 6 0 0 21 1 13 9 5 0 0 1 14 1 2 1 78

Effect of a 50 percent reduction in tariff equivalent barriers between ISA countries. Source: G.C. Hufbauer, J.B. Brandford, and S. Stephenson (2012) *

DUPUY

The second risk, and certainly the most sensitive at the present time, is that the development of plurilateralism results in a growth in asymmetries in international economic relations, excluding or marginalizing the least-developed countries (Scott and Wilkinson 2012; Narlikar 2012). Plurilateralism may then promote the emergence of a multitiered trading system. Experience in plurilateral agreements effectively shows that the areas on which plurilateral trade negotiations tend to focus are chosen by the countries with the greatest clout in world trade. Areas such as services or trading in IT products are of deep interest to these countries but are hardly relevant for the vast majority of lesser developed countries. Likewise, the low number of countries that joined the Dairy Agreement, the International Bovine Meat Agreement, and to a lesser extent GPA, shows the unwillingness of nonmember countries to apply to join existing plurilateral trade agreements. For example, the lack of appeal seen with GPA for emerging countries stems on the one hand from the fact that opening up public markets on a reciprocal basis, such as that conceived by the agreement, places these countries in an unfavorable situation given the “asymmetries in the size of markets and ability to supply procurement markets” (Woolcock 2013: 4) and on the other hand from the absence of specific provisions for Special and Differentiated Treatment (emerging countries being particularly keen to maintain preferential procurement programs in order to support their infant industries).9 From past experiences, we might then conclude that the universalization of plurilateral agreements is all the easier in that the subjects concerned by plurilateral negotiations interest, a priori, the greatest number of countries. Given the evolution of the balance of power within WTO, in particular rising coalitions of developing countries (cf. notably the G21), draft plurilateral agreements today can no longer emanate from developed countries alone. As stressed by Woolcock (2013: 6), “The experience with the GPA shows clearly that leadership of like-minded OECD countries in negotiating a high standard agreement by no means guarantees there will be followers.” Table 5. The increase in trade for twenty-two countries applying zero-for-zero tariffs in three sectors ($billions) Country/Region

Chemicals (Imports, Exports)

Electronic and Electrical Goods (Imports, Exports)

Environmental Goods (Imports, Exports)

22 countries Developed countries Emerging countries European Union Japan United States Brazil China India

15.4, 12.8 4.2, 8.1 11.2, 4.8 1.4, 3.3 0.2, 2.2 2.3, 2.1 1.0, 0.1 4.5, 1.3 0.8, 0.3

35.4, 33.5 6.6, 16.3 28.8, 17.2 3.0, 5.7 *, 6.5 2.6, 3.4 3.9, 0.1 11.3, 6.7 1.7, 0.2

6.3, 4.5 1.2, 3.1 5.1, 1.4 0.3, 1.4 *, 0.9 0.6, 0.6 0.5, * 1.7, 0.7 0.8, 0.1

*gains lower than $0.05 billion N.B.: For each country, commercial gains reflect the increase in its imports from the rest of the world and the increase in its exports to the other twenty-one countries in the sample. Source: Hufbauer, Schott and Wong (2010)

One might object to the previous argument by saying that least-developed countries will have free latitude to join existing plurilateral trade agreements when their economies have grown stronger. But as James Scott and Rorden Wilkinson so rightly point out (2012), it is extremely difficult to sign up to a trade agreement after the event if a country refused to join initially. Countries that are not around the original negotiating table are, de facto, not in a position to push and win ground for their own interests. If they decide to join the 9. Thus, negotiations for China to join the GPA failed for want of an agreement as to the reciprocal commitments of all partners. The revised GPA, which came into effect in 2014, let emerging countries wishing to access the agreement benefit from specific provisions in terms of Special and Differentiated Treatment. This measure should make it easier for emerging countries to join the GPA.

PLURILATERALISM: THE KEY TO SAVING THE WORLD TRADE ORGANIZATION

agreement at a later date, the cost will be high, because their own interests will not have been taken into account. It should also be emphasized that the provision whereby a plurilateral agreement should assemble a “critical mass” of participants (90 percent according to the rule adopted by the WTO),10 based exclusively on share in international trading, is a powerful incentive for WTO member countries to reach agreements that, in principle, exclude the lesser developed countries. As Amrita Narlikar notes (2012: 38), “The biggest criticism of a pure critical-mass approach is that it risks re-inventing the Principal Supplier Principle, and may result in the permanent and institutionalized marginalization of a great majority of countries that could never constitute the critical mass.” Thus, under the WTO umbrella, 90 percent of trading in IT products worldwide is made by approximately thirty countries. To justify their hostility to plurilateralism, developing countries finally point to the economic cost. In the first place, sectorial liberalization leads to a loss of tariff revenues. In a number of developing countries, especially the poorest, tariff revenue is still a significant item for their public finances. For instance, in Africa, the Middle East, and Asia, taxes on trade account for around 20 percent of tax revenues. However, the scope of the argument claiming loss of tax revenues should not be overstated. In reality, the loss of tariff revenues due to the plurilateral agreement scrapping customs duties on imports in the sector concerned tends to be more than offset by additional tariff revenues coming firstly from the increase in importing goods in this sector and secondly by the rising imports in other sectors due to the induced effects of the plurilateral agreement (cf. above). In the second place, the liberalizing trade by plurilateral agreements would threaten the survival of local firms in liberalized sectors. This argument is used by many developing countries to justify their refusal to join GPA and ITA. For this reason, several of them, including Argentina and Brazil, have imposed high customs duties on imports of IT goods in order to protect domestic industries in the IT and communication sector. For the moment, these protectionist policies have had the sole effect of raising the price of ICT for domestic usages and of hampering the diffusion of ICT to the economy as a whole, thereby slowing down the growth rate of the global productivity of factors. The earlier experience seen with India effectively shows the adoption of protectionist measures in the ICT sector has no effect other than to lower the competitiveness of home companies and stymie potential growth. Thus, according to P.D. Kaushik and Nirvikar Singh (2002), for every $1 of tariffs India imposed on imported ICT products (in the years before it joined the ITA), the Indian economy suffered a loss of $1.30 due to lower productivity. The growth in world value chains in the information technologies and communications sector (i.e., the growing fragmentation of the production process for IT goods in successive phases on sites located in different countries) explains in part why protectionist policies have not had the desired effects for national firms operating in this sector. As Stephen J. Ezell notes (2012: 10), “High tariffs on ICT parts and products simply compel ICT firms to bypass these (protectionist) countries entirely in their global supply chains and manufacture and assemble elsewhere.” In the same way, it has been shown that the reluctance of developing countries to open up government procurement by joining GPA is not always founded. The argument whereby foreign suppliers would squeeze out the developing countries’ less competitive domestic suppliers has limited scope. In reality, when a contract is won by a foreign-based supplier, there is positive fallout for local firms in the form of sub-contracts awarded by the foreign company and transfers of technology that help to improve pricing competitiveness (Anderson et al. 2011). More generally, we see that foreign companies win public sector contracts local firms are not equipped to win themselves and vice versa. Thus, the study conducted by the African 10. To avoid the problem of free-riders and incite as many countries as possible to sign up, the MFN clause can be applied only when the number of participants in plurilateral agreements reaches a critical size, i.e., the signatory countries must conduct a significant share of world trade for the goods or services in question. It is the 90 percent rule that has been applied under the frame of the ITA.

DUPUY

Development Bank (2009) revealed that suppliers from regional member countries won 88 percent in terms of number of contacts for 43 percent by value of contracts awarded, while suppliers from nonregional member countries won 55 percent by value and 12 percent in number of contracts. Finally, the rare studies focusing on the effects of developing countries joining GPA, including primarily those concerning the experiences of Korea and Chinese Taipei, show that opening up government procurement has helped to strengthen competition, governance, and the efficiency of domestic markets without causing any significant increase in the rate of import penetration (Choi 2003; Lo 2011). In all, the majority of developing countries feel that plurilateral agreements run contrary to the principles of universality, inclusion, and transparency. Additionally, in their opinion, plurilateral initiatives would turn the attention of WTO member countries away from the question of development, which can be addressed only at a global level. Adjustments are needed in order to overcome the hostility of the lesser-developed countries to plurilateral agreements. The first is aimed at replacing or at complementing the traditional critical mass approach. Several options can be considered. Some authors have proposed using a definition of critical mass based on several criteria; aside from share in international trading for goods and services of interest to us here, consideration could be given, for example, to a requirement for a minimum number of countries or the obligation that plurilateral agreements concern a significant share of the world’s population (Narlikar 2012). In the same vein, the Warwick Commission believes that notions such as increased global welfare, the protection of the principle of nondiscrimination, and the income distribution effects should be considered when defining a critical mass. As emphasized by Christophe Bellmann et al. (2012: 11), this multi-criteria approach would “ensure a certain degree of legitimacy for the critical mass, in addition to economic relevance.” Other authors propose using rules that are more or less far-removed from the notion of critical mass but which offer a majority of countries the possibility to be a part of plurilateral agreements. For instance, the report by Bhagwati and Sutherland (2011) suggests that countries representing less than 1 percent of world trade in a given sector could sign a plurilateral agreement in that sector but would effectively be a part of it only when their share of trading exceeded the 1 percent mark. However, as pointed out by Bellmann et al. (2012: 11–12), “most propositions relative to the critical mass approach basically concern access to markets.” At a time when plurilateral agreements are focusing more and more on the regulation and standardization of trade (cf. Table 2), this sort of critical mass concept seems ill-suited. It is hard to see how we might define a critical mass for negotiations having a bearing, for example on the climate or world value chains. The second adjustment under consideration concerns the terms and conditions of participation for countries wishing to join existing plurilateral agreements. When joining, candidate countries have to be in a position where they can gain prevalence for their own interests, implying that the terms and conditions of currently applicable agreements should be renegotiated (Harbinson 2009). Such a move would forestall the risk of the more powerful countries using plurilateral agreements to impose their viewpoints on developing countries when it comes to sensitive subjects such as those labeled “WTO Plus” (investment, protection of intellectual property, environmental standards, social standards, etc.). Thirdly, developing countries wishing to join a plurilateral trade agreement must have the ability to benefit from specific provisions in respect of Special Differentiated Treatment (SDT). These provisions may, for example, provide for exemptions from certain obligations for technical assistance and capacity support to implement the agreement under consideration. As seen earlier, the decision of developing countries to join GPA is heavily conditioned in particular by adopting special measures that are differentiated in their favor. More generally, in order to reassure developing countries, specific provisions of concern should

PLURILATERALISM: THE KEY TO SAVING THE WORLD TRADE ORGANIZATION

figure in a “code of conduct” that would govern plurilateral trade negotiations (Draper and Dube 2013). Finally, plurilateralism presents one last risk that currently affects business firms, particularly small and medium-sized businesses (SMBs). Expanding plurilateral agreements effectively engenders a “risk of divergence in commercial policies, inducing higher transaction costs for exporters and, in the first place, SMBs, which at the legal level seldom have the right and proper resources.” (CCIP 2012:1–2). Conclusion Faced with the failure of the Doha cycle, plurilateralism may constitute an exit route to give a new boost to global trade liberalization, enabling certain WTO countries to reach agreements in areas where there is a consensus while offering other members the chance to join these agreements when they wish and/or are ready. However, the exit route is a narrow one. The development of plurilateralism cannot be allowed to result in the fragmentation of the international trading system. Safeguards have to be adopted in order to achieve the goal of greater liberalization for worldwide trade. Future plurilateral agreements will, in particular, need to allow for conditions that promote access for the lesser-developed countries, especially the least-favored nations, which to date have been sidelined from this form of organizing international trade. To this end, there is a need to come up with a definition of “critical mass” that can account the largest number of countries and not only those that do the major share of trading in goods and services. There is a second need to shift plurilateral negotiations over to areas that present a real interest for lesser-developed countries. At a more global level, the acceptance of plurilateralism by as many countries as possible will be feasible only when clear and specific answers have been given to the questions surrounding the 1) end-purpose of plurilateralism, 2) the content of plurilateral agreements (clear demarcation between what is included in each agreement and what is not), and 3) interactions between different plurilateral agreements. REFERENCES African Development Bank (2009): Procurement Summary by Year 2004 to 2009, mimeo. Anderson, Robert D., Philippe Pelletier, Kodjo Osei-Lah and Anna Caroline Müller (2011): “Assessing the value of future accessions to the WTO agreement on government procurement (GPA): some new data sources, provisional estimates, and an evaluative framework for individual WTO members considering accession,” WTO Staff Working Paper ERSD-2011-15. Baldwin, Richard (1993): A Domino Theory of Regionalism, NBER Working paper 4465. Bellmann, Christophe, Jonathan Hepburn and Marie Wilke (2012): “Le système commercial multilatéral face aux défis des politiques publiques globales,” Revue internationale de politique de développement, N° 3. Bhagwati, Jagdish (1995): “U.S. Trade Policy: The Infatuation with Free Trade Agreements” in Jagdish Bhagwati and Anne O. Krueger, The Dangerous Drift to Preferential Trade Agreements, AEI Press. Bhagwati, Jagdish and Peter Sutherland (2011): The Doha Round: Setting a Deadline, Defining a Final Deal, Interim Report. Bora, Bijit and Xuepeng Liu (2006): Evaluating the impact of the WTO Information Technology Agreement, Draft. Cerny, Philip (1993): ”Plurilateralism: Structural Differentiation and Functional Conflict in the PostCold War Order,” Journal of International Studies, 22 (1): 27–51. Chambre de Commerce et d’Industrie de Paris (2012): ”Multilatéralisme vs. pluri/bilatéralisme: vers un accord plurilatéral sur les services?” Brèves commentées du commerce international, n° 55: 1–2.

DUPUY

Chen, Hejing and John Whalley (2011): “The WTO procurement agreement and its impacts on trade,” NBER Working Paper 17365. Choi, Inbom (2003): “The Long and Winding Road to the Government Procurement Agreement: Korea’s Accession Experience” in Will Martin and Mari Pangestu, Options for Global Trade Reform: A View from the Asia-Pacific, Cambridge University Press, Chapter 11: 249–69. Draper Peter and Memory Dube (2013): “Plurilaterals and the multilateral trading system,” in Strengthening the multilateral trading system, International Centre for Trade and Sustainable Development. European Commission (2013): Negotiations for a Plurilateral Agreement on Trade in Services, memo. European Commission (2011): EU public procurement legislation: delivering results summary of evaluation report. Ezell Stephen J. (2012): “The benefits of ITA expansion for developing countries,” ITIF Working Paper. Geslin, Albane (2010): “Les traités plurilatéraux: quelle(s) utilité(s) dans le système commercial multilatéral?” in Les sources et normes dans le droit de l’OMC, Nice. Harbinson, Stuart and Bart De Meester (2012): Analysis of WTO-consistent approaches to plurilateral and non-MFN trade agreements, National Foreign Trade Council. Harbinson, Stuart (2009): The Doha Round: “Death-Defying” Agenda or “Don’t Do it Again?” ECIPE Working Paper n° 10. Hoekman, Bernard M. et Petros C. Mavroidis (2013): WTO “à la carte” or WTO “menu du jour”? Assessing the case for Plurilateral Agreements, European University Institute Working Papers. Hufbauer, Gary C. (2012): “The Institutional Crisis at the WTO and Possible Solutions,” in The Future of the WTO and the International Trading System, Directorate-General for External Policies, European Parliament, pp. 22–33. Hufbauer, Gary C, Jeffrey J. Schott and Woan F. Wong (2010): “Figuring out the Doha Round,” Policy Analyses in International Economics, 91, Peterson Institute for International Economics. Hufbauer, Gary C., Jensen Branford and Sherry Stephenson (2012): “Framework for the International Services Agreement,” Policy Brief 12, Peterson Institute for International Economics. Kaushik P.D. and Nirvikar Singh (2002): “Information Technology and Broad–Based Development: Preliminary Lessons from North India,” UC Santa Cruz Economics, Working Paper n° 522. Laborde, David (2011): “Sectorial Initiatives in the Doha Round, miméo. Lawrence, Robert Z. (2006): “Rule-making amidst growing diversity: a club of club approach to WTO Reform and new issue selection,” Journal of International Economic Law, Vol. 9, Issue 4: 823–35. Lo, Changfa (2011): “The benefits for developing countries of accession to the Agreement on Government Procurement: the case of Chinese Taipei” in Arrowsmith and Anderson, The WTO Regime on Government Procurement: Challenge and Reform, Cambridge University Press, Chapter 5: 140–48. Mann, Catherine L. and Xuepeng Liu (2007): “The Information Technology Agreement: Sui Generis or Model Stepping Stone, Conference on Multilateralising Regionalism,” WTO and CPER. Nakatomi, Michitaka (2013): Plurilateral Agreements: A Viable Alternative to the World Trade Organization? Asian Development Bank Institute Working Paper Series, N° 439. Narlikar, Amrita (2012): “The Role of the WTO in a Rapidly Changing International Economic Order” in The Future of the WTO and the International Trading System, Directorate-General for External Policies, European Parliament: 34–42. Rodriguez-Mendoza, Miguel (2012): “Toward ‘Plurilateral Plus’ Agreements,” in The Future and the WTO: Confronting the Challenges. A Collection of Short Essays, ICTSD.

PLURILATERALISM: THE KEY TO SAVING THE WORLD TRADE ORGANIZATION

Saner, Raymond (2012): Plurilateral Agreements: Key to solving the impasse of WTO/Doha Round and basis for future trade agreements within the WTO context, CSEND Policy Brief, n° 7. Sato, Hitoshi (2014): “Does MFN Free Riding Plague the Information Technology Agreement?” RIETI Discussion Paper Series 14-E-003. Scott, James and Rorden Wilkinson (2012): “The Politics and Perils of Plurilaterals,” Economic & Political Weekly, Vol. XLVI n°43: 16–19. Siroën, Jean-Marc (2010): “L’OMC face à la crise des négociations multilaterals,” 4ème Biennale internationale de la négociation commerciale Penser les échanges commerciaux autrement. Sutherland, Peter (2006): “Addressing institutional challenges in the new millennium, WTO.” Trebilcock, Michael (2014): “Between Theories of Trade and Development: the Future of the World Trading System,” University Toronto Law Working Paper Series, n° 2014–10. Vu, Khuong (2005): “Measuring the Impact of ICT Investments on Economic Growth,” Working paper, Harward Kennedy School of Government. Warwick Commission (2007): The Multilateral Trade Regime: Which Way Forward? The University of Warwick. Woolcock, Steve (2013): Getting Past the WTO Deadlock: the Plurilateral Option? Robert Schuman Centre for Advanced Studies Policy Paper, N° 2013/08. WTO (2013), “Is multilateralism in crisis,” WTO Public Forum 2012. WTO (2012), 15 Years of the Information Technology Agreement: Trade, innovation and global production networks. Zartman, William (1994): “International Multilateral Negotiation: Approaches to the Management of Complexity,” Jossey-Bass, San Francisco.

1 January 1981 (initial agreement) 1 January 2014 (new agreement)

1 January 1980 Scrapped end of 1997

1 July 1997

Signed 1 October 2011

Agreement on Government Procurement (GPA)

International Dairy Agreement

International Bovine Meat Agreement

Information Technology Agreement (ITA)

Anti-Counterfeiting Trade Agreement (ACTA)

Source: author

1 January 1980

Agreement on Trade in Civil Aircraft

Yes

Within the WTO Framework

Yes

Application of the MFN clause

Yes

Application of the WTO conflict settlement procedure

To set up international standards in the fight against counterfeiting in order to upgrade the protection of intellectual property.

To scrap tariffs on IT products (for certain sensitive products, an extension of the implementation period may be obtained).

To grow and liberalize world trade for beef.

To grow and liberalize world trade for dairy products.

To open up public-sector markets to international competition, ensuring that national laws and regulations are transparent and do not protect national products and suppliers to the detriment of their foreign counterparts (nondiscrimination).

To scrap tariffs for all aircraft other than military aircraft, and for all spare parts and components for civilian aircraft.

Objectives

The Major Characteristics of Existing Plurilateral Agreements

Number of Member Countries

APPENDIX

Date of First Application

Agreement

61 DUPUY

REVIEWS International Organizations and Internal Conditionality: Making Norms Matter by Heidi Hardt, University of California–Irvine International Organizations and Internal Conditionality: Making Norms Matter by Rick Fawn, New York: Palgrave Macmillan, 2013, ISBN: 9781137305480, 335 pages. This book examines a problem that all international organizations face: how to engage with member states that disagree with aims of the organization. Rick Fawn studies this important question in the context of the Council of Europe (CoE) and the Organization for Security and Co-operation in Europe (OSCE). Unlike other European organizations, these two organizations lack the material benefits or formal enforcement mechanisms that would otherwise motivate states to adopt their norms, and this complicates compliance. The author uses the concept of internal conditionality to identify the processes through which states and organizations navigate the issues over which their values clash. Taking a bottomup approach, Fawn defines internal conditionality as the processes used by international organizations to “meet the challenges” of recalcitrant member states (p. 7). Fawn’s interpretation of internal conditionality differs from external conditionality, which is exercised toward new member states during the accession process. Thus, he builds on previous scholarly accounts of internal conditionality as “institutional strengthening demanded by new circumstances”1 or states’ capacity to honor existing norms2 to show that institutional norms can sometimes be effective at influencing even those states that typically disagree with them. In this sense, while the book ties together scholarship on norms, European studies, and practices of international organizations, in essence, it is a study of norm diffusion. The book then offers five case studies, each examining an issue on which the values of a member state and those of an international organization conflict. Fawn assesses each case along six factors. The first three concern the importance of values to the organization and the recalcitrant member states, while the last three factors are used to evaluate the effectiveness of incentives and sanctions. In the case studies, one learns about the costs and benefits considered by states evaluating the extent to which their policies should match those of the respective organization. For example, by internalizing the norm of abolishing the death penalty, states could increase their prestige and opportunities for economic cooperation (chapter 4). However, when norms conflict with its national interests (chapter 5), the state may choose to pursue domestic policies regardless of norm diffusion within the international organization. Evidence includes some two hundred interviews with practitioners from the CoE and the OSCE, and the book contributes to the emerging literature on the practices of international 1. Gianni Bonvicini and Stefano Silvestri, “The New ‘Arc of Crisis’ and the European Community,” The International Spectator 27, no. 2 (November 26, 1992): 31–43. 2. Iavor Rangelov, Bulgaria’s Struggle to Make Sense of EU Human Rights Criteria (Open Society Institute, 22 February 2010), 2.

JIOS, VOL. 6, ISSUE 1, 2015

HARDT

organizations.3 Through its in-depth case studies, the reader has the opportunity to understand how different mechanisms translate or fail to translate norms into policy shifts. Another strength lies in the book’s definition of what Fawn refers to as “target” or “recalcitrant” states. Previous scholarship has referred to states whose values do not align with the organization as “bad actors”4 or “bully states.”5 In places, the book inappropriately lumps post–Communist states together as recalcitrant states, even though variation exists in their respect for human rights or the extent to which they hold free and fair elections. Nevertheless, Fawn provides insights into the motivations and actions of these states within the CoE and within the OSCE. The author’s descriptions of their behavior could be better categorized as offensive or defensive. For example, states may act to undermine the international organization or act to defend their own set of values against a perceived imposition of Western-created values. Another strength of the book comes from investigating the informality of practices within international organizations. The lack of formalization offers more opportunities for an agreement that is more than the sum of its parts (p. 242). The book concludes with a call for further research into the “balance between legality and flexibility” (p. 241). The author uses interviews to identify how, even in situations that led to disappointing outcomes, states chose through informal processes to permit adjudication at the international level when target states have neither the will nor resources to do so (e.g., Russia’s participation in the European Court of Human Rights, p. 161). The attention to the two-level game is critical to understanding the conditions under which states cooperate or not; however, more discussion of domestic politics could have further advanced the debate on each of these cases. In addition, the author praises the CoE and the OSCE for their inclusive approach to membership without considering the costs of such an approach, including states that flout the organizations’ rules diminishes the organizations’ perceived legitimacy. Other international organizations have reacted to such situations by suspending members, such as the recent G8 suspension of Russia for its involvement in the Ukraine conflict. Inclusion may guarantee dialogue among states, but it is less obvious why a state like Kazakhstan, which “had not fulfilled the most basic democratic and human-rights commitments of the Organization” (p. 195), was awarded such a prestigious position in the international community as the OSCE chairmanship. One should also question the assumption that inclusion will inevitably have positive effects on states’ domestic policies as some of the cases presented led to policy shifts while others did not. From a methodological point of view, the book would have benefited from a clearer justification of case selection and process tracing. The author only notes that cases were selected to demonstrate the “resourcefulness and resilience” in different issue areas (p. 233). But the five cases, ranging from OSCE election monitoring to the CoE’s efforts to abolish the death penalty, seem difficult to compare. It also would have been helpful to see a breakdown of the results of the two hundred interviews (p. 14), such as the use of discourse analysis across subjects’ responses. Nonetheless, some of the lessons summarized in the conclusion could surely apply to other international organizations beyond Europe. Existing and emerging scholarship on the African Union and the Organization of American States is applicable in that some of the member states exhibit similarly deviant behavior. ASEAN would offer a particularly interesting comparison given its design as a more informal organization. Ultimately, Fawn’s book offers insights into the ways in which two understudied organizations function in practice. The book reminds us that norm diffusion can and does continue 3. Vincent Pouliot, International Security in Practice: the Politics of NATO-Russia Diplomacy (Cambridge: Cambridge University Press, 2010); Emanuel Adler and Vincent Pouliot, International Practices, (Cambridge, UK: Cambridge University Press, 2011); Heidi Hardt, Time to React: the Efficiency of International Organizations in Crisis Response (New York: Oxford University Press, 2014). 4. Ryan Goodman and Derek Jinks, “How to Influence States: Socialization and International Human Rights Law,” Duke University School of Law 54, no. 3 (3 April 2008): 629. 5. Hardt, Time to React: the Efficiency of International Organizations in Crisis Response, 184.

INTERNATIONAL ORGANIZATIONS AND INTERNAL CONDITIONALITY

to occur within international organizations even in the face of recalcitrant states. The CoE and the OSCE were founded on the principle of equality. As a result, Fawn’s conceptualization of internal conditionality indicates that all member states should be subject to the same accountability, whether they are “well-behaved” or not (p. 7).

Conflict Resolution and the Everyday Politics of International Intervention by Costantino Pischedda, University of Miami (starting fall 2015) Peaceland: Conflict Resolution and the Everyday Politics of International Intervention by Séverine Autesserre, New York: Cambridge University Press, 2014, ISBN: 9781107632042, 360 pages. Professor Séverine Autesserre’s new book offers an original and insightful critique of the subculture of the expatriate community involved in peacebuilding efforts on behalf of international and nongovernmental organizations throughout the globe (the author defines peacebuilding as “actions aimed at creating, strengthening, and solidifying peace,” p. 21). The book’s central claim is that many of the practices, habits, and narratives that shape peacebuilding efforts on the ground—everyday elements that interveners come to take for granted—are inefficient and even counterproductive. In particular, the peacebuilding community’s tendency to value technical and thematic expertise at the expense of country-specific knowledge for hiring and promotion leads to the deployment of personnel who knows next to nothing about the country of operation. This in turn fosters reliance on simplistic narratives about causes of violence and the use of universal policy templates (like the “fads” the author characterizes as power-sharing agreements and early post-conflict elections) without adaptation to local contexts. As a consequence, international interveners often clumsily promote initiatives that are at odds with socio-political conditions on the ground and thus are likely to yield only limited benefits or even backfire. For example, in the Democratic Republic of Congo (DRC), the international community’s single-minded focus on enhancing statecapacity as a solution to the problem of persistent violence in 2010–11 may have made it worse by empowering an exploitative and repressive state security apparatus. Moreover, peacebuilders’ daily routines, such as following strict security protocols and socializing only in the narrow circle of international interveners, create and maintain boundaries separating them from the communities they are meant to help and from the local actors they are supposed to work with. These modes of operation tend to “prevent local authorship and decrease local ownership” and to provoke “widespread feelings of humiliation and resentment among host populations, encouraging local stakeholders to evade, adapt, or resist international programs” (p. 250). Autesserre presents a rich and persuasive ethnography of the peacebuilding community and its modus operandi, marshaling an impressive amount of information from her extensive fieldwork (in particular in the DRC) and direct experience as an intervener in several countries. The book, however, falls short when it moves from description of “Peaceland” to causal arguments about peacebuilding effectiveness (i.e., about the negative consequences of international interveners’ dominant practices, habits and narratives, summarized on pp. 36–37). As the author herself carefully and repeatedly notes, many of the practices that have undesirable effects are adopted to address real obstacles to peacebuilding initiatives. Arrangements that constrain local actors’ influence on peacebuilding programs are meant to reduce the risk of “policy capture” by government officials, elites and civil society organizations whose interests are not necessarily aligned with those of ordinary citizens (p. 66). Moreover, as several of Autesserre’s interviewees point out, interveners often find themselves in a catch-22 JIOS, VOL. 6, ISSUE 1, 2015

PISCHEDDA

when deciding about the degree of local stakeholders’ involvement in program implementation: “Either they did capacity building, which took enormous amount of time and resulted in poorly executed programs—while intended beneficiaries continued to suffer from the continuation of violence—or they implemented the program themselves, which was more effective in the short term but unsustainable in the long term and created dependency” (p. 90). Analogously, socialization in the peacebuilding “bubble” is, at least in part, a functional response to the need of preserving psychological and emotional balance for individuals working under challenging security and social circumstances. Sharp economic, cultural, and social differences act as powerful obstacles to integration (for example, interveners often complain about being seen by the local population as “walking money-bags whom one could always ask for donations or handouts,” p. 178). Full acceptance in the local social fabric remains elusive even for peacebuilders that “go native” (i.e., move permanently to the country, study local languages and cultures, and marry locals), as these exceptional individuals often feel that very few people fully accept them and deserve their confidence (pp. 179–80). Autesserre’s claim that peacebuilding effectiveness would increase if the practices leading to the undesired effects were abandoned holds only under the assumption that the negative effects those practices mitigate are not as serious. However, the author does not attempt a systematic assessment of these competing negative effects and thus leaves the reader wondering whether some of the “medicines” she prescribes are indeed better than the “disease” they are meant to cure. This is not to say that in its current approach to peacebuilding the international community has hit the sweet spot between competing risks and benefits. To be sure, there are practices for which the case for reform is straightforward (for example, the trend toward the “bunkerization” of peacebuilders’ offices even in contexts of low security threats). But broader reform initiatives need to rest on a hardnosed investigation of the risks of unintended effects and inevitable tradeoffs. Autesserre’s book is likely to remain for a long time a must-read for those willing to engage in this difficult and important analytical task.

The UN and Human Rights: Who Guards the Guardians? by Francesco Rindi, United Nations Organisation Stabilisation Mission in the Democratic Republic of Congo The UN and Human Rights: Who Guards the Guardians? by Gugliemo Verdirame, Cambridge: Cambridge University Press, 2011, ISBN 978-0-521-84190-0, 448 pages. This topical treatise1 addresses the thorny issue of the United Nations’ (UN) accountability for human-rights violations. Does the UN violate human rights? If so, is it in any case bound to observe human-rights obligations? What are the measures placed to ensure the UN’s accountability for human-rights violations? These are some of the key questions raised in this monograph. After explaining in chapter one the concepts and legal definitions used in the book, Gugliemo Verdirame demonstrates in the second chapter that both the UN Charter and customary law bind the organization and its agencies to the observance of human rights law. The third chapter argues that, although the International Law Commission’s draft articles on the responsibility of international organizations represent a step forward in configuring the legal framework for determining the responsibility of international organizations for human-rights violations, they do not solve the problem of the enforcement of the UN’s human-rights obligations (p. 143). The following chapters (four to seven) detail a consistent pattern of human-rights violations by the UN in four of its areas of work, namely humanitarian relief, peacekeeping operations, international administrations, and the implementation of sanctions. According to the author, these activities carry the greatest risk of human-rights abuses, because the UN exercises extensive and direct power over individuals (other than its own employees). Chapter four shows how, during the relief operations in Afghanistan, the UN ended up incorporating into its country assistance the regime of sexual apartheid put in place by the Taliban dictatorship. Likewise, chapter five gives an account of human rights violations in a number of peacekeeping operations. In chapter six, Verdirame introduces a distinction between de jure international territorial administrations, established through a Security Council mandate, and de facto territorial administrations, most notably refugee camps, where the UN controls inhabited territory in the absence of such a mandate. In regards to de jure international administrations, he examines the cases of Kosovo, East Timor and Bosnia, and Herzegovina, arguing that the UN committed human-rights violations mostly by failing to prevent local actors from perpetrating crimes and by shielding the decisions of the UN authorities from any review by local courts (pp. 241–66). The second part of the chapter (pp. 269–98), examines the United Nations High Commissioner for Refugees’ (UNHCR) administration of refugee camps, arguing that those camps where refugees remain for a prolonged period of time, as opposed to temporary ones established as a “short term emergency response to a refugee crisis” (p. 240), are inherently illegal, because they necessarily entail a restriction of the freedom of movement of refugees. Verdirame further argues that other human-rights abuses by UNHCR in refugee camps are not infrequent (pp. 281–98). Chapter seven analyzes the human rights impact of UN sanctions. The comprehensive sanctions regime, in use until the end of the Cold War, targeted a whole country indiscriminately, for 1. The views expressed herein are those of the author alone and do not represent the views of the United Nations Organisation Stabilisation Mission in the Democratic Republic of Congo or of the United Nations

JIOS, VOL. 6, ISSUE 1, 2015

RINDI

example, by blocking its trade, as in the 1990 embargo on Iraq. To avoid the disastrous humanitarian consequences these sanctions had on innocent people, the UN shifted to a regime of targeted sanctions addressing specific groups, individuals, or commodities (pp. 306–10). However, because listed individuals and entities are not accorded an effective and transparent appeals process against the decision of sanction committees, the author also argues that these sanctions undermine human rights, especially the right to a fair trial and due process (pp. 311–19). The final chapter analyses the issue of accountability. The author observes that, in the face of such concrete risks of human rights violations, the prospects for victims of bringing a claim against the UN are minimal. Internal administrative controls are inadequate, patchy, and, generally, do not offer redress mechanisms for victims. Moreover the control exercised by other UN organs over Security Council decisions is scarce and, so far, UN member states have not scrutinized its decisions as much as they could (pp. 321–43). Yet, Verdirame observes that certain domestic and international courts in Europe have developed ways of resisting the implementation of international organizations’ decisions that breach the fundamental individual rights guaranteed in national constitutions. Indeed, according to the doctrine of equivalent protection, when the level of human rights protection offered by an international organization is inferior to the one available nationally, a domestic constitutional court can invalidate an act implementing a decision of an international organization. However, this remains only an indirect form of scrutiny, since the organization’s breach itself does not come under examination (pp. 359–86). Verdirame concludes that the UN suffers from a serious accountability deficit. The expansion of UN authority has not been matched by a corresponding reflection on how to guarantee the rights of individuals affected by its behavior. Even though some progress has been made, the UN, the “guardian” mandated to protect human rights from state abuses, still needs to devise a credible and effective system to ensure accountability for its own violations. Notwithstanding its relevance, studies on this subject are rare. This comprehensive, wellresearched, and meticulous monograph is thus pioneering. It also is an eye-opener for all those practitioners and academics who are accustomed to think that, because of its noble principles, the UN is inherently virtuous. At a time when, with the establishment of an Intervention Brigade, the UN has stepped up its powers within the peacekeeping mission in the Democratic Republic of Congo, Verdirame’s treaty acquires even greater significance. This being said, the section on the UNHCR’s de facto administration of refugee camps (pp. 269–98) can stimulate further discussion, underscoring the need for more comprehensive field research. Whereas the author’s argument on the UNHCR’s potential for human-rights violations over refugees is clear, many of the criticisms of the UN agency and examples of violations appear misplaced. It is debatable that UNHCR, rather than the country of asylum, is responsible for restricting the freedom of movement of refugees in camps. It is also arguable that practices such as “head-counting” refugees constitute human-rights breaches (pp. 282–85). The author describes this process as a form of cruel and degrading treatment, among other things, because UNHCR “forces refugees in enclosures like cattle,” where they are left to wait for many hours often “under a scorching sun,” and because it practices body markings on refugees (p. 284, emphasis added). The process of registering refugees is a necessary procedure to streamline the provision of aid and identify vulnerable persons, ensuring that they receive adequate protection. In such cases, refugees may have to wait several hours to be registered, and the UN agency, partly due to its limited resources, may have to use temporary body markings to speed up the process. While not an ideal solution, it is hard to imagine how UNHCR could provide assistance often to thousands of refugees without having an exact idea of the composition of its population. It is clear, in sum, that more critical thinking on the issue of the UN’s accountability for human rights violations is needed. Professor Verdirame’s monograph can stimulate further debate among academics, practitioners, and, hopefully, within the UN itself.

NATO in Afghanistan: Fighting Together, Fighting Alone by Romain Malejacq, Centre for International Conflict Analysis and Management (CICAM), Radboud University NATO in Afghanistan: Fighting Together, Fighting Alone by David P. Auerswald and Stephen Saideman, Princeton and Oxford: Princeton University Press, 2014, ISBN: 9780691159386, 280 pages. By the end of December 2014, most foreign combat troops had left Afghanistan. NATO in Afghanistan, David Auerswald and Stephen Saideman’s bold attempt to analyze the challenges and difficulties of coalition warfare, could not be more topical. This ambitious book provides a detailed and balanced account of NATO-led combat operations in Afghanistan between 2003 and 2010. Some readers might take issue with the fact that the study does not cover the 2010–14 period and, therefore, fails to analyze NATO’s pullout. But the book does not aim to document the successes and failures of NATO’s International Security Assistance Force (ISAF), nor does it pretend to explain decisions to engage and disengage at the alliance level. Rather, it seeks to point out and demonstrate the complexity of cooperating with others when conducting multilateral operations. The book is based on the simple but crucial observation that, in the current international environment, “fighting alongside other countries” is both challenging and necessary (p. 13). In Winston Churchill’s words, “There is at least one thing worse than fighting with allies—and that is to fight without them” (p. 1). But “why is coalition warfare so hard?” (p. 2). This is what the authors of this book attempt to explain in light of the NATO experience. Auerswald and Saideman’s argument is straightforward and not necessarily counterintuitive. Coalition warfare is difficult because of the primacy of national decisions in multilateral interventions. While this simple fact contradicts NATO’s efforts of creating a centralized and integrated organization, it is something most people familiar with the war in Afghanistan already know from talking to military officers or hearing and reading American officials complain about other countries’ caveats. However, this book provides a sophisticated argument that explains why domestic politics prevail and why this will remain the case in the future. The authors build on principal-agent theory to make a well-documented, well-illustrated, and convincing case: The primacy of national behavior over alliance decisions is a structural problem. In their words, “NATO’s structure and processes establish . . . a hybrid principal-agent relationship between the multiple entities delegating authority to deployed military units and those military units themselves,” which favors the authority of the member states (p. 13). This book not only demonstrates why national control trumps alliance decisions in coalition warfare but also explicates “why there is variation in control mechanisms used by countries involved in multilateral interventions and by some countries over time” (p. 12). This is what makes this book an important one. According to the authors, the key variable for explaining variation in national behavior is the allied country’s political system. Presidential systems, single-party governments, and coalition governments all act differently when it comes to multilateral operations. One of the book’s main achievements is to combine this institutional argument with political considerations and individual preferences of key decision makers to leverage a relatively simple JIOS, VOL. 6, ISSUE 1, 2015

MALEJACQ

hypothesis and develop a sophisticated argument. Auerswald and Saideman convincingly show that the type of political system determines what tools principals use to direct their agents. This hypothesis is in turn demonstrated through rich empirical case studies (although some are analyzed more in-depth than others). In presidential systems (such as the U.S., France, and Poland), “individuals matter a great deal” (p. 113) and can use a variety of tools to control their agents (agent selection, incentives, oversight, restrictions, etc.). For the most part, single-party governments (such as the U.K. and Canada) work in similar ways, since they also empower individual decision makers who act as single principals. Most variation is observed in the behaviors of coalition governments, which the authors explain by introducing two more variables, ideological dispersion and ideological leaning. NATO in Afghanistan is an ambitious study. As such, it should be praised for offering a multi-causal theoretical model that makes sense of a complicated issue, which is one of the book’s many strengths. Some political scientists might find issue with the high number of cases and variables. Readers more comfortable with mono-causal explanations might find the multiplicity of explanatory variables a little unsettling, as it sometimes makes it difficult to understand which factors have primacy and occasionally seems to be reintroducing some of the alternative explanations that the authors had dismissed early in the book (realism, public opinion, and strategic culture). Yet, this is a minor criticism compared to the outstanding qualities of this work. Maybe one way around it could have been to discuss alternative theories more in-depth to see how they can play out in this multi-causal model. Where does national interest come into play? Can we really assume all the countries that have been involved in ISAF in the past thirteen years have shared interests? If not, how does it impact their behavior within the coalition? NATO in Afghanistan is both timely and important. It makes a meaningful contribution to the “research on civil-military relations, and the impact of domestic politics upon foreign policy” (p. 218). It is not only essential reading for scholars of international organizations, international security, and world politics (who will find it both theoretically stimulating and empirically compelling) but also a must-read for policy makers. The implications of this work are broad, significant, and thought provoking. Since domestic politics always prevail, specialization and military cooperation will never cease to present major difficulties and pitfalls even for “the most interoperable and effective multilateral security organization in the world” to make their point (p. 31). Allies will never act the way one wants them to: “[They] do not always show up when needed or they show up but are not able to do what is needed” (p. 232). One can only hope, following Auerswald and Saideman, “that the next set of leaders to deploy troops are aware of the inevitable trade-offs when allies go to war” (p. 236).

Peace Diplomacy, Global Justice and International Agency: Rethinking Human Security and Ethics in the Spirit of Dag Hammarskjöld by Kent J. Kille, College of Wooster

Stahn, Carsten and Henning Melber (Eds.) (2014) Peace Diplomacy, Global Justice and International Agency: Rethinking Human Security and Ethics in the Spirit of Dag Hammarskjöld, Cambridge: Cambridge University Press, ISBN 978-1-10703720-5, 617 pages. The extensive volume Peace Diplomacy, Global Justice and International Agency: Rethinking Human Security and Ethics in the Spirit of Dag Hammarskjöld, edited by Carsten Stahn (professor of international criminal law and global justice at Leiden University and program director of the Grotius Centre for International Legal Studies) and Henning Melber (senior advisor and director emeritus of the Dag Hammarskjöld Foundation), was initiated in part at a 2011 conference and reexamines the second secretary-general of the United Nations (UN) and contextualizes the relevance of his approach and principles for the modern era. The contributors include not only a number of Hammarskjöld experts but also an array of new voices and perspectives from across disciplines. After an overview introduction by the editors, the twenty-four chapters are organized into five parts: 1) personal reflections on Hammarskjöld, 2) his intellectual legacy and leadership, 3) Hammarskjöld and the UN in the Congo, 4) the role of the UN secretary-general, and 5) re-thinking internationalism, before concluding with a reprinting of Hammarskjöld’s final annual report to the General Assembly and his last address to the staff, given shortly before his untimely death in a plane crash in September 1961. The volume is designed to provide a detailed consideration of Hammarskjöld’s legacy as well as his present-day significance. While often imparting very positive reviews of his place in history, the book represents, as the editors express, both a “tribute” and a “critical review” (p. 10) that includes the limits to Hammarskjöld’s efforts and ideals. This approach allows for a stronger base of inquiry into the relevance of Hammarskjöld’s activities and principles for contemporary international affairs. Hammarskjöld is noted as an inspiration for the UN secretaries-general that followed. Kofi Annan, seventh secretary-general, states in his contribution, “I believe the Secretary-General has no choice. He has to follow in the footsteps of Hammarskjöld” (p. 38). Hammarskjöld is also described as an influence on a range of actors and institutions within and beyond the UN system More broadly, the book promotes a clearer understanding of the question, in the words of Pieter Kooijmans, “Why is it that we are so eager to know what Hammarskjöld’s views and position on our present world problems would have been?” (p. 53) across a range of venues and issues. Including different author perspectives is a strength of the volume. Along with diverse backgrounds, the contributors support their arguments with varied forms of material. Demonstrating the ongoing power of Hammarskjöld’s writing, a number of contributors rely primarily on close readings of Hammarskjöld’s own words—his speeches, annual reports, and personal reflections particularly employed in Markings—or, in the case of Roger Lipsey’s chapter, ponJIOS, VOL. 6, ISSUE 1, 2015

KILLE

dering the form and implications of what Hammarskjöld would have written. Analyses of these texts holds clear value, but it is also inherently limited since this only explores Hammarskjöld’s outlook on issues. So the inclusion of chapters using other material (such as deeper archive material and legal analysis) and viewpoints (especially in relation to the Congo and broader theme of decolonization) offers a valuable counterbalance and understanding of the broader story in which Hammarskjöld was enmeshed. At the same time, many of the accounts offer a compelling case for the need to fully understand Hammarskjöld as an individual who brought to the office certain views, ethics, traits, and experiences. The dedicated effort to not just look back at Hammarskjöld but to embed the analysis of his ideas and actions in different historical contexts and modern perspectives also makes the volume more relevant to understanding international relations. Many of the authors are careful to dictate both where the world is similar and different from Hammarskjöld’s time as they consider the related potential relevance of his principles and approach to global events. The coverage of the Congo, both in the distinct section and in material ranging through the book, does a particularly good job of justifying and bridging the situation under Hammarskjöld to the ongoing UN engagement in the country. For building a broader understanding of the role of the UN secretary-general beyond Hammarskjöld, the analysis is limited to mostly comparing with Annan. But there are also some initial interesting in-roads elsewhere in Aoife O’Donaghue’s chapter pointing briefly to the Commonwealth and World Trade Organization and especially in Steven Blockmans’ exploration of implications for the European Union. The book provides an impressive set of material but serves more as a collection of chapters than a fully analytically linked volume. The editors could have pushed the critical value of the book further by providing a more conceptually based, instead of overview focused, introduction and by reflecting across the contributions in a concluding chapter. For example, the reader might be jarred when moving from part two, which often emphasizes Hammarskjöld’s ethical importance, into part three, which in places raises some serious questions regarding the degree to which Hammarskjöld might have been ill-informed or lacking in judgment in relation to the Congo. This represents the value of having such divergent author views but utilizing an editorial voice to assist with making relative sense of these approaches. Melber tackles this to a degree in relation to the Congo in his individual chapter, but an extension of such comparative reflection is encouraged. Analyzing across the chapters could also highlight and consider the overlap of ideas and findings as well as how these reinforce the overarching picture of Hammarskjöld across history and current day. For example, Anne Orford delivers a compelling account of the importance of Hammarskjold’s economist background and thinking, so how does her coverage in relation to the Congo and decolonization mesh with the examination of these areas across other chapters? Since the book is rather long, the selection of what to include could have been more judicious. For example, several chapters, including most of part one, were previously published elsewhere and could be left aside to focus on the new contributions. In addition, there is a tendency to repeat certain basic material or facts across the chapters. Reducing the first section would open up the space to supply the reader with central facts and a timeline, allowing the substantive chapters to focus on analyzing instead of a degree of reiterating. The chapters are well supported through citation via voluminous footnotes, but the book does not have an overarching bibliography. As the latest addition to a sizeable literature on Hammarskjöld, bringing together the preceding publications in this fashion would better demonstrate the relative placement of this volume. Whether for Hammarskjöld diehards or more general readers, the value this volume adds and the degree to which the book acts as the most up-to-date reference on the subject must be closely considered in relation to other works. For example, the comparable collection of Sten Ask and Annan MarkJungvist’s (2005) The Adventures of Peace: Dag Hammarskjöld and the Future of the UN.

PEACE DIPLOMACY, GLOBAL JUSTICE AND INTERNATIONAL AGENCY

Overall, Stahn and Melber have achieved their “hope” to “provide a modest contribution to keep Hammarskjöld’s principles and legacy alive in their relevance for today” (p. xv), and it will be interesting see how Hammarskjöld’s ongoing significance continues to shape our understanding of and approach to world affairs.