国际法相关参考资料：国际法综合理论

Between Power and Principle: An Integrated Theory of International Law Oona A. Hathaway? Forthcoming in the Chicago Law Review (2005) In March of 2003, as American tanks rolled toward Baghdad, international lawyers in the United States and abroad decried the action as a violation of the United Nations Charter. The invasion, some worried, would strip away the last pretense that international law could constrain state action. Others openly questioned whether the increasingly wounded global legal regime was worth saving. If states so openly flouted it, was international law really worth the trouble? The hand-wringing and condemnation were scarcely new. Well before the invasion of Iraq the tide of events had given pause to all but the staunchest believers in international law. Within six short months of entering office, President George W. Bush had withdrawn from the Kyoto global climate accord,1 threatened to unilaterally abrogate the 1972 Anti- Ballistic Missile Treaty, and revoked the U.S. signature on the treaty creating the International Criminal Court. The U.S. president thus looked ready to make good on the promise that Jesse Helms, then-chairman of the Senate Foreign Relations Committee, had made to the UN Security Council only a year earlier to resist any effort to “impose the UN’s power and authority over nation states.”2 ? Associate Professor, Yale Law School. J.D., Yale Law School. I thank the Carnegie Foundation for its generous support of this project through the Carnegie Scholars Program. My thanks also to Craig Estes, Galit Sarfaty, and Alan Schoenfeld for their research assistance and to Ulrich Wagner and especially Alexandra Miltner for their help with compiling and analyzing the datasets used in this Article. I am grateful to Bruce Ackerman, Yochai Benkler, William Bradford, Jutta Brunnée, Steve Charnovitz, Robert Ellickson, Ryan Goodman, Larry Helfer, Rob Howse, Dan Kahan, Alvin Klevorick, Barbara Koremenos, Ariel Lavinbuk, Mike Levine, Jonathan R. Macey, Daniel Markovits, Eric Posner, Kal Raustiala, Roberta Romano, Scott Shapiro, Peter Shuck, Alan Schwartz, Jim Whitman, Tim Wu, Kenji Yoshino, the faculty of the University of Bremen, and participants in the University of Southern California Conference on Compliance with International Law, the International Law Roundtable at Vanderbilt Law School, the Yale World Fellows program, and the University of Toronoto’s workshop on international law for helpful conversations about and comments on earlier drafts. Thanks are also due to Gene Coakley and the rest of the staff of the Yale Law School library for their outstanding assistance. Finally, I owe the greatest debt to Jacob S. Hacker, for his support at every stage of this project. 1 For an interesting discussion that places the decision of the United States to withdraw from the Kyoto Protocol into a broader context by considering the US approach to international environmental law since 1992, see Jutta Brunnee, The United States and International Environmental Law: Living with an Elephant, 15 EJIL 617 (2004). 2 Senator Jesse Helms, Speech to the United Nations Security Council (Jan 19, 2000), online at http://www.sovereignty.net/center/helms.htm (visited Nov 30, 2004). Between Power and Principle 2 Yet the blame for today’s crisis atmosphere cannot be laid in Bush’s lap. While the Bush administration fanned the flames of concern, the issue of what role international law can play in regulating international relations has bedeviled the world community for decades. After World War II, even as the world pressed ahead with the United Nations and other new international institutions, widespread dismay over the failure of earlier institutions to prevent the collapse of order prompted a wave of attacks on the Wilsonian ideal of an international system founded on global legal order.3 As long as there was no sovereign power to manage enforcement, critics argued, international law was meaningless. Regarding it as otherwise was not just unrealistic but dangerous.4 In the face of these attacks, international lawyers have worked assiduously to refine, interpret, and apply international law. But they have not yet done enough to respond to the ever more intense concerns about the field’s validity.5 It is perhaps not surprising, then, that 3 See, for example, Hans J. Morgenthau, Politics Among Nations (Knopf 3d ed 1966) (offering a realist critique of international law); Edward Hallett Carr, The Twenty Years’ Crisis: 1919–1939 (MacMillan 2d ed 1946) (same); Hans J. Morgenthau, Positivism, Functionalism, and International Law, 34 Am J Intl L 260 (1940) (same). 4 See, for example, Raymond Aron, The Anarchical Order of Power, in Stanley Hoffman, ed, Conditions of World Order 25, 47 (Houghton 1968) (concluding that international society is an anarchical order of power in which might is supreme); Charles W. Briggs, The Cloudy Prospects for “Peace Through Law,” 46 ABA J 490, 493–95 (1960) (acknowledging that international law can only be enforced by a world sovereign, but concluding that establishment of a world government is a dream). 5 Louis Henkin writes: “These depreciations of international law challenge much of what the international lawyer does. Indeed, some lawyers seem to despair for international law until there is world government or at least effective international organization.” Louis Henkin, How Nations Behave: Law and Foreign Policy 25–26 (Columbia 2d ed 1979)). For more, see Part II. There are, of course, exceptions. For example, the “new stream” scholarship has long been critical of traditional approaches to international law. See generally Jason Mark Anderman, Note, Swimming the New Stream: The Disjunctions Between and Within Popular and Academic International Law, 6 Duke J Comp & Intl L 293 (1996) (criticizing the traditional “new world order” international law theory and concluding that the “new stream” academic international law theory is logically and morally superior); Nigel Purvis, Critical Legal Studies in Public International Law, 32 Harv Intl L J 81 (1991) (discussing traditional international law theory’s reduction to marginality due to its impossibility); Anthony Carty, Critical International Law: Recent Trends in the Theory of International Law, 2 Eur J Intl L 66 (1991) (providing an analytic presentation of the main themes of four contemporary international law scholars, all of whom advocate nontraditional concepts of international law and criticize traditional international legal theory); Phillip R. Trimble, International Law, World Order, and Critical Legal Studies, 42 Stan L Rev 811 (1990) (reviewing Lung-Chu Chen, An Introduction to Contemporary International Law: A Policy-Oriented Perspective (Yale 1989); Richard A. Falk, Revitalizing International Law (Iowa State 1989); David Kennedy, International Legal Structures (Nomos Verlagsgesellschaft 1987)); Martti Koskenniemi, From Apology to Utopia: The Structure of International Legal Argument (Finnish Lawyers’ 1989); David Kennedy, A New Stream of Between Power and Principle 3 much of the public debate over international law has been polarized and unproductive. Skeptics argue strenuously that international law is mere window dressing. Advocates frequently assume that states abide by their international legal commitments “almost all of the time.”6 More must be done to evaluate critically the role that international law can and does play in shaping state behavior. Legal and political science scholars have begun to meet this challenge,7 yet we still remain remarkably ill-equipped to predict or explain the real-world International Law Scholarship, 7 Wis Intl L J 1 (1988) (criticizing traditional conceptions of international law and advocating a rhetorical and interactive theory); David Kennedy, International Legal Structures (Nomos Verlagsgesellschaft 1987) (criticizing traditional theories of international law and advocating an internal analysis of international law’s doctrinal and rhetorical cohesiveness); James Boyle, Ideals and Things: International Legal Scholarship and the Prison-House of Language, 26 Harv Intl L J 327 (1985) (critiquing attempts to define international law as manifestations of a pervasive reification by legal scholars); David Kennedy, Theses About International Law Discourse, 23 Ger YB Intl L 353 (1980) (identifying a crisis in traditional international legal scholarship and advocating an analytical approach that examines the transformational rules governing discourse for hints about the structure of the international law dilemma). 6 Henkin, How Nations Behave at 25–26 (cited in note 5). 7 In recent years, a significant literature has arisen on the topic, among both political scientists and international legal scholars. See William Bradford, International Legal Compliance: An Annotated Bibliography (unpublished manuscript 2004), online at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=577104 (visited Nov 25, 2004) (examining significant literature that evaluates the role of international law in state behavior). Some notable examples include William Bradford, In the Minds of Men: A Theory of Compliance with the Laws of War (unpublished manuscript on file with author); James Raymond Vreeland, Why Do Governments and the IMF Enter into Agreements?: Statistically Selected Cases, 24 Intl Polit Sci Rev 321 (2003) (examining the national considerations of governments in entering into International Monetary Fund (IMF) agreements through evaluation of the national motivations in two outlying cases); James Raymond Vreeland, Institutional Determinants of IMF Agreements (Dec 11, 2002), online at http://pantheon.yale.edu/~jrv9/Veto.pdf (visited Nov 24, 2004) (examining whether domestic institutions influence decisions to participate in IMF programs); Beth Simmons, Why Commit? Explaining State Acceptance of International Human Rights Obligations (2002) (unpublished manuscript on file with author); Andrew Moravcsik, The Origins of Human Rights Regimes: Democratic Delegation in Postwar Europe, 54 Intl Org 217 (2000) (examining why governments allow an international human rights regime to constrain domestic sovereignty); Beth Simmons, International Law and State Behavior: Commitment and Compliance in International Monetary Affairs, 94 Am Polit Sci Rev 819 (2000) (examining patterns of commitment to and compliance with international monetary law); Martha Finnemore, National Interests in International Society 69–88 (Cornell 1996) (examining the national motivations behind the Geneva Conventions’ rules for warfare). Between Power and Principle 4 impact of the over 50,000 international treaties now in force, covering nearly every aspect of international relations and nearly every facet of state authority.8 This Article offers a theory of international treaty law that helps fill this gap. I call the theory an “integrated” theory of international law because it brings together and builds upon two cross-cutting facets of existing scholarship on international law and politics. First, the theory draws on both political science and legal scholarship, using them together to construct a broader and deeper understanding than is possible working solely within either discipline alone. Second, it operates at the intersection of two distinct theoretical approaches regarding the role of international law—what I term the interest-based and norm-based approaches—that cut across the disciplinary divide between political science and international legal scholarship. The theory—and this Article—is integrated in yet another sense as well. It accepts and seeks to respond to the challenge mounted by the critics of international law. By integrating their insights into an analysis of how international law affects what states actually do. Rather than reject altogether the arguments of international law skeptics—as international lawyers and scholars are sometimes want to do—I seek to show when and why their claims have power and when and why they do not. I begin by confronting the root causes of doubt about the power of international law. I focus attention on the voluntary nature of international treaty law (the fact that countries often choose whether to be bound by it) and on the frequent absence of any central enforcement power. I go on to show how, despite and at times because of these distinctive features, international treaty law can profoundly shape state behavior. The central goal of this Article is to explain the impact of international treaty law on state behavior (that is, on countries’ decisions to comply or not with treaty law). The integrated theory I develop herein begins with the claim that understanding the political interplay between countries’ decisions to commit and to comply is essential to a complete picture of the influence of international treaty law. I thus argue that we must pay attention to the interaction of the domestic and transnational levels at each stage of a country’s entry into an international legal framework: its decision to commit (or not) to international legal rules; and its decision to comply (or not) with them. When we do so, we find that compliance not only depends upon the decision to commit, but commitment also depends upon the decision to comply. Beginning with this interactive mode of thinking about state decisionmaking, the Article proceeds to describe and elaborate the two central ways in which treaties shape what countries do: through legal enforcement of the terms of the treaty and by bringing about collateral consequences for state interests. Both operate at the domestic as well as transnational level. Legal enforcement occurs at the domestic level when domestic actors use the country’s own legal system to enforce the terms of international legal agreements. At the transnational level,9 legal enforcement occurs when international bodies or other states 8 The United Nations Treaty Series currently contains over 50,000 treaties. United Nations Treaty Series Overview (2003), online at http://untreaty.un.org/English/overview.asp (visited Nov 24, 2004). 9 This Article adopts the definition of “transnational” used by Philip Jessup: he wrote that “transnational law” includes “all law which regulates actions or events that transcend Between Power and Principle 5 that are party to the treaty respond to violations of a treaty in ways provided for in the treaty. Collateral consequences, by contrast, arise when domestic and transnational actors premise their actions toward a state on the state’s decision to accept or not accept international legal rules. The reactions of these actors to the state’s actions can affect, among other things, foreign investment, aid donations, international trade, domestic political support, and political contributions, and hence create powerful incentives for states to commit to and comply with treaties. Viewing state decisions to commit and comply in with broader lens makes it possible to understand why countries behave in ways that standard theories miss—failing to join treaties, for instance, that they could easily comply with (because they have little to gain and much to lose), or joining treaties that they have little inclination to obey (because they have much to gain and little to lose). For example, it is often thought that countries with poor human rights records will be reluctant to join treaties that embody higher standards. Yet in fact these countries often have stronger incentives (and weaker disincentives) to join human rights treaties than states with better records—first, because such countries usually have weak rule of law and thus create limited opportunities for domestic legal enforcement; second, because human rights treaties usually lack transnational legal enforcement mechanisms, such as supranational enforcement or credible threats of state-to-state retaliation; and finally, because such countries, by displaying their (sometimes insincere) commitment to human rights, increase their standing among other nations, international bodies, private investors, domestic actors, and others and thereby obtain significant collateral benefits. The integrated theory thus predicts that, holding other factors constant, countries with very poor human rights records can be as likely or even more likely to ratify treaties as countries with better records, but that unlike those with better records, they are unlikely to comply with those commitments—which is in fact the pattern found. In short, the theory not only provides a comprehensive vision of the potential and the limits of international law; it also gives rise to unique (and often counterintuitive) predictions that are consistent with the available evidence. This Article focuses exclusively on international treaty law. For the purposes of this first effort to articulate the theory, I put customary international law to one side.10 The focus in this piece is on state decisions to commit to and comply with international treaty law. It is national frontiers. Both public and private international law are included, as are other rules which do not wholly fit into such standard categories.” Philip C. Jessup, Transnational Law 2 (Yale 1956). Hence “transnational” is used here in its literal sense: it means across nations, as opposed to “international,” which means between nations. The term “transnational” therefore encompasses a larger universe of activity and interactions than does the term “international.” When applied to law, for example, transnational law includes any law that has cross-border effect, whereas international law refers only to treaties or other law that governs interactions between states. 10 This Article is aimed only at explaining international treaty law. Hence references to “international law” are intended to include only international treaty law and not customary international law, jus cogens norms, private transnational legal interactions, or domestic law that extends across borders. Between Power and Principle 6 widely understood that states cannot be bound by a treaty unless they agree thereto.11 Commitment does not operate in the same way for customary law (with limited exceptions, states are regarded to be obligated by international customary law regardless of whether they wish to be).12 As a consequence, although many of the same factors will likely be important 11 See Part II.A. 12 Customary international law does not require the same kind of affirmative act on the part of a state to subject it to the law, but it is generally accepted that states can—with some important exceptions—avoid application of customary international law simply by persistently objecting to it. The persistent objector rule “is an accepted application of the traditional principle that international law essentially depends on the consent of states.” Restatement (Third) of the Foreign Relations Law of the United States § 102, Reporters’ Note 2 (1987). On the persistent objector rule, see Ian Brownlie, Principles of Public International Law 10 (Oxford 5th ed 1998) (concluding that the persistent objector doctrine is well recognized by international tribunals and in the practice of states); Jonathan Charney, Universal International Law, 87 Am J Intl L 529, 538–42 (1993) (noting the existence of the persistent objector rule, but concluding that the rule is open to doubt due to its infrequent use and the existence of several exceptions to it). For an application of the rule, see Fisheries Case (United Kingdom v Norway), 1951 ICJ 116, 139 (holding that the international rule concerning fisheries zones is inapplicable because Norway has always opposed any attempt to apply it to the Norwegian coast). Indeed, some question whether it is even possible for states to be “bound” by, or have an obligation under, international law given their sovereignty. Perhaps the most famous statement of the notion that in a world of independent states, sovereignty resides in the states was made by the Permanent Court of International Justice in 1927 in the S.S. Lotus case: “International law governs relations between independent States. The rules of law binding upon States therefore emanate from their own free will as expressed in conventions or by usages generally accepted as expressing principles of law.” The S.S. Lotus (France v Turkey), 1927 PCIJ (ser A) No 10, 18. Several works have discussed the dilemma inherent in the notion of obligating a sovereign state, including David Kennedy, International Law and the Nineteenth Century: A History of an Illusion, 17 Quinnipiac L Rev 99, 112–31 (1997); Rosalyn Higgins, Problems and Process: International Law and How We Use It 13–16 (Clarendon 1994); Thomas M. Franck, The Power of Legitimacy Among Nations 187–94, 202–07 (Oxford 1990); Richard A. Falk, The Status of Law in International Society 19–23 (Princeton 1970) (concluding that sovereign consent has stalemated attempts to obligate nations to comply with international law derived through consensus, natural law, or consent); J.E.S. Fawcett, The Law of Nations 6–11 (Oxford 1968) (concluding that international law has a moral basis and an arbitrary structure that is accepted because it makes the conduct of states more constant, more predictable, and so more effectual); Oscar Schachter, Towards a Theory of International Obligation, 8 Va J Intl L 300, 307 (1968) (arguing that obligatory norms, established through the same, consistent five processes, are the basis of obligation in international law); W. Friedmann, Legal Theory 574–80 (Columbia 5th ed 1967); James Brierly, The Law of Nations: An Introduction to the International Law of Peace 49–56 (Oxford 6th ed 1963); C. Wilfred Jenks, Law, Freedom and Welfare 83–100 (Stevens & Sons 1963); Hans Kelsen, General Theory of Law and State 341–63 (Russell & Russell 1961) (Anders Wedberg, transl); James Between Power and Principle 7 to explaining state compliance with customary law, the way in which the theory operates necessarily will be quite different. The Article proceeds as follows. I begin in Part I by briefly examining the existing international relations and legal literature on the influence of international law on state behavior. I classify the literature into two broad camps—interest-based models and norm- based models—and seek to trace out in broad outlines the competing explanations offered by each. I conclude this Part by briefly discussing the start of a promising convergence of these two theories upon which this Article seeks to build. Part II begins to delineate my own account of the reach and limits of international law. In this Part, I examine the aspects of international treaty law that have sown the seeds of doubt as to whether it is really “law” at all. I focus particularly on treaty law’s largely voluntary nature and the relative absence of central enforcement mechanisms in international law. This discussion forms the foundation for my integrated theory of international law, which I develop in Part III. Part IV assesses the theory against existing accounts using new empirical evidence on state behavior under key human rights and environmental treaties, as well as existing empirical studies. Finally, Part V concludes with suggestions for future research and for designing international law to harness its real but limited power more effectively. I. Existing Literature The divide between advocates and skeptics of international law is in part the legacy of a gradually disappearing schism in scholarship and teaching between students of law and students of international relations, manifested institutionally as a split between law schools and political science departments. From the close of World War II through the last decade, scholars at law schools who taught and studied international law ignored many of the questions of context and power relations that had become the central concern of international study in political science departments. Political scientists, for their part, tended to dismiss international law altogether. Even within legal academia, international law was, until the last decade, regarded largely as a curiosity—a subject of study truly relevant only to the few who devoted themselves to it. With increased globalization, the isolation of international law has begun to melt away. But what has replaced it is, in many cases, almost equally dismissive. Rather than integrate the work of international legal scholars into the rest of the curriculum, legal academics have instead tended to regard international law merely as an extension of existing areas of domestic law—as simply torts, corporate law, or criminal law that happens to cross Leslie Brierly, The Basis of Obligation in International Law 1–68 (Oxford 1958) (Sir Hersch Lauterpacht and C.H.M. Waldock, eds); H. Lauterpacht, The Function of Law in the International Community 3–4 (Oxford 1933). For a particularly interesting refutation of the assertion that sovereign states cannot be bound by international law, see H.L.A. Hart, The Concept of Law 215–21 (Clarendon 1961). See also Richard L. O’Meara, Note, Applying the Critical Jurisprudence of International Law to the Case Concerning the Military and Paramilitary Activities in and against Nicaragua, 71 Va L Rev 1183, 1203 (1985) (claiming that “a truly independent, sovereign state cannot be bound without its consent”). Between Power and Principle 8 borders.13 This vision, while partially accurate, fails to acknowledge the ways in which international law is fundamentally different from its domestic counterpart.14 Nonetheless, over the past decade, two broad theoretical approaches regarding the role of international law in state behavior have started to cut across the disciplinary divide between political science and international legal scholarship. The first, which I term the interest-based approach, argues that states create and comply with international law only when there is some clear objective reward for doing so; in other words, states follow consequentialist reasoning or what has been termed the “logic of consequences.”15 The second, which I label the norm-based approach, argues that governments create and comply with treaties not only because they expect a reward for doing so, but also because of their commitment (or the commitment of transnational actors that influence them) to the norms or ideas embodied in the treaties. Hence, in this view, states often follow what has been termed the “logic of appropriateness” rather than that of consequences. Moreover, nonstate and substate actors are the focus of much more attention in the norm-centered account than in the interest-centered one, for they play an important role in constructing state preferences.16 As I shall show in the brief review that follows, each of these accounts offers valuable insights into the ways in which states react to international law.17 13 Harold Koh writes: [T]hat reminds me of something a former law school dean told me sixteen years ago when I said I was coming to Yale to teach International Business Transactions: that there is no genuinely transnational body of international business law, because transnational business law is like that famous non- book, The Law of the Horse, which consists of Chapter I: “Contracting for a Horse,” Chapter II: “Owning a Horse;” Chapter III, “Torts by a Horse,” and Chapter IV: “Litigating over a Horse.” Harold Hongju Koh, The Globalization of Freedom, 26 Yale J Intl L 305, 305 (2001). Moreover, international law is rarely a required course in law school; this reflects a widespread belief among legal academics that an understanding of international law is superfluous to a solid legal education. 14 See discussion in Part II.A–B. 15 James G. March and Johan P. Olsen introduce the terms “logic of consequences” and “logic of appropriateness,” which they draw from their study of cognitive psychology in their research on organizations and political institutions. James G. March and Johan P. Olsen, Rediscovering Institutions: The Organizational Basis of Politics 160–62 (Free Press 1989) (comparing the logic of consequences—where behavior is willful, fills subjective desires, and is driven by preferences and expectations about consequences—with the logic of appropriateness—where behavior is intentional, fills the obligations of a role in a situation, and stems from a conception of necessity, rather than preference). See also James G. March and Johan P. Olsen, The Institutional Dynamics of International Political Orders, 52 Intl Org 943, 949–54 (1998); John W. Meyer and Brian Rowan, Institutionalized Organizations: Formal Structure as Myth and Ceremony, 83 Am J Sociology 340 (1977) (critiquing the reliance on consequentialist logic in the social sciences). 16 The line between the interest-based and norm-based approaches is far from absolute. Interest-based models often fall back on normative insights in order to explain Between Power and Principle 9 A. Interest-Based Models The interest-based approach has its roots in the realist view of international cooperation, which became dominant in American political science scholarship in the wake of World War II. In this view, states are seen as rational, unitary actors in pursuit of self- interest. Early realist accounts used this vision of state action to argue that international agreements exist and are enforced only when they serve the interests of the most powerful states.18 More recent scholarship, by contrast, argues that regimes19—including legal otherwise inexplicable state behavior in the human rights arena, and norm-centered accounts do not deny the power of rational self-interest to motivate state behavior. Indeed, Moravcsik has aptly labeled this a “curious convergence” of the two main theoretical accounts in his work on the European Convention on Human Rights. Moravcsik, 54 Intl Org at 224–25 (cited in note 7). Moreover, several scholars have done excellent work at the intersection of interest-based and norm-based accounts. See, for example, Martin Shapiro & Alec Stone Sweet, Governing with Judges: Constitutional Powers in Europe (2000). And much work in the so-called English School of international relations cannot be characterized as falling solely within one camp or the other. The goal here is therefore not to exaggerate the differences between the approaches, but simply to outline the prevailing modes of thought on the role of international law in shaping state behavior. 17 This outline (as well as what follows) necessarily simplifies what are deeply complex theoretical accounts. For a more comprehensive introduction to these theoretical approaches, see Oona A. Hathaway and Harold Hongju Koh, Foundations of International Law and Politics (Foundation Press 2004). See also Oona A. Hathaway, Do Human Rights Treaties Make a Difference?, 111 Yale L J 1935, 1942–62 (2002). 18 Indeed, E.H. Carr and Hans Morgenthau, among others, made the case that states are simply rational unitary actors motivated by their geopolitical interests. Law, in this view, is nothing more than one of a variety of tools used by states to enhance their own power. See generally Morgenthau, Politics Among Nations (cited in note 3); Carr, The Twenty Years Crisis (cited in note 3); Morgenthau, 34 Am J Intl L 260 (cited in note 3). In the 1970s, Kenneth Waltz and others expanded on the realist perspective, arguing that states may pursue a broader set of interests and sometimes make strategic decisions to engage in international cooperation. Nonetheless, even in this “neorealist” approach, states’ primary motivation remains preservation and accumulation of power. As Waltz put it in his classic book, Theory of International Politics, states are “unitary actors who, at a minimum seek their own preservation and, at a maximum, drive for universal domination.” Kenneth N. Waltz, Theory of International Politics 118 (McGraw-Hill 1979). In this view, the most powerful states create and join treaties that advance their power interests and then force weaker states to join and comply by threatening sanctions or promising benefits. 19 Regimes are defined as “sets of implicit or explicit principles, norms, rules, and decision-making procedures around which actors’ expectations converge in a given area of international relations.” Stephen D. Krasner, Structural Causes and Regime Consequences: Regimes as Intervening Variables in Stephen D. Krasner, ed, International Regimes 1, 2 (Cornell 1983). Compare Robert O. Keohane and Joseph S. Nye, Power and Between Power and Principle 10 regimes—can influence the behavior of international actors.20 States create and comply with the requirements of international regimes, these scholars claim, because the regimes allow Interdependence: World Politics in Transition 19 (Little Brown 1977) (defining international regimes as “governing arrangements that affect relationships of interdependence,” or “networks of rules, norms, and procedures that regularize behavior and control its effects”). The term “regimes” is now commonly used interchangeably with “institutions.” See, for example, Ronald B. Mitchell and Patricia M. Keilbach, Situation Structure and Institutional Design: Reciprocity, Coercion, and Exchange, in Barbara Koremenos, Charles Lipson, and Duncan Snidal, eds, The Rational Design of International Institutions 131, 133 (2004) (“We also use the term regime interchangeably with institution.”) (emphasis added). 20 In political science, recent variants of this approach are variously labeled “modified structural realism,” “intergovernmental institutionalism,” “neoliberal institutionalism,” and “new institutionalism.” See, for example, Robert O. Keohane, Theory of World Politics: Structural Realism and Beyond, in Robert O Keohane, ed, Neorealism and Its Critics 158, 190–97 (Columbia 1986) (outlining a research plan for Modified Structural Realism); Friedrich Kratochwil and John Gerard Ruggie, International Organization: A State of the Art on an Art of the State, 40 Intl Org 753, 759–60 (1986) (explaining that international regimes can help clarify why in the 1970s states did not respond to the pressure on institutional arrangements in “beggar-thy-neighbor terms”); Robert O. Keohane, After Hegemony: Cooperation and Discord in the World Political Economy 63 (Princeton 1984) (“Theories of regimes can incorporate realist insights about the role of power and interest, while also indicating the inadequacy of theories that define interests so narrowly that they fail to take the role of institutions into account.”). Modified Structural Realism is the immediate bridge from Structural Realism, often referred to as Neorealism. See, for example, Robert O. Keohane, Realism, Neorealism and the Study of World Politics, in Keohane, ed, Neorealism and Its Critics 1, 17. See, for example, Judith Goldstein, et al, eds, Legalization and World Politics (MIT 2001) (examining the legalization of regimes and the impact of regimes on world politics); Barbara Koremenos, Charles Lipson, and Duncan Snidal, The Rational Design of International Institutions, in Barbara Koremenos, Charles Lipson, and Duncan Snidal, eds, The Rational Design of International Institutions 1, 2, 3, 7 (2004) (referring to various treaties as included among the institutions the volume examines). Legal scholars have contributed to this reconceptualization by using institutionalist approaches to examine state compliance with international law. See Andrew T. Guzman, A Compliance-Based Theory of International Law, 90 Cal L Rev 1823 (2002) (putting forward a theory of international legal compliance in which rational, self-interested states and international law punish violations through reputational and direct sanctions); Jack L. Goldsmith and Eric A. Posner, A Theory of Customary International Law, 66 U Chi L Rev 1113 (1999) (presenting a theory of customary international law that draws on rational choice theory); John K. Setear, An Iterative Perspective on Treaties: A Synthesis of International Relations Theory and International Law, 37 Harv Intl L J 139 (1996) (using an institutional approach to explain the Law of Treaties). Between Power and Principle 11 states to engage in cooperative activity that would otherwise be impossible.21 By allowing states to restrain themselves and others from pursuing short-term interests at the expense of shared long-term goals, regimes make it possible for states to obtain benefits that exceed the costs of membership.22 Yet as different as the many variants of this approach are, they share at least two key assumptions: States engage in consequentialist means-end calculations, and state interests can be deduced from the state’s material characteristics and the objective conditions it faces. Moreover, these models traditionally focused exclusively on state-level interactions, with scholars largely ignoring substate dynamics. Despite increased attention to international law by interest-based scholars, current interest-based approaches remain ill-equipped to explain the existence of—much less state compliance with—treaties that impose costs in return for little or no apparent benefit. Human rights and environmental treaties, for example, impose substantial sovereignty costs on states in return for the collective goods of human dignity and a healthier world environment. In a world where self-interest is the central motivating force of state action, why would states waste time and energy creating treaties that yield little obvious individualized benefit? And why would they ever abide by them? Interest-based theorists commonly argue such treaties are simply “cheap talk”—used by the governments of powerful states to justify actions that are in fact taken for self-interested reasons.23 Proponents of this view, however, give no explanation as to why such cover is valuable—as to why, that is, the great powers feel the need to justify the pursuit of their interests. Moreover, they are at a loss to explain why it is that the most powerful nations commonly refuse to join such treaties. Alternatively, rationalists argue that countries are willing to join treaties that offer little in the way of benefits if they also ask countries to modify their behavior little or not at all and hence impose minimal cost.24 In this view, then, compliance 21 See Robert O. Keohane, The Demand for International Regimes, in Krasner, ed, International Regimes 141, 147 (cited in note 19) (“In general, we expect states to join those regimes in which they expect the benefits of membership to outweigh the costs.”). 22 See Keohane, Demand for International Regimes at 147 (cited in note 21). 23 Waltz, Theory of International Politics at 200 (cited in note 18); Carr, The Twenty Years’ Crisis at 71–75 (cited in note 3) (explaining that states have engaged in “morally discrediting the policy of a potential enemy and morally justifying [their] own” in order to aid their efforts to obtain geopolitical power); Morgenthau, 34 Am J Intl L at 11 (cited in note 3) (“All nations are tempted—and few have been able to resist the temptation for long—to clothe their own particular aspirations and actions in the moral purposes of the universe.”). 24 In this view, states will create treaties in these areas only if the treaties require very little of them. Once the treaties are created, countries with practices consistent with the treaty’s requirements might join, whereas countries with noncomplying practices will not. See George W. Downs, David M. Rocke, and Peter N. Barsoom, Is the Good News About Compliance Good News About Cooperation?, 50 Intl Org 379, 380 (1996) (arguing that “the high level of compliance and the marginality of enforcement result from the fact that most treaties require states to make only modest departures from what they would have done in the absence of an agreement”). Jack Donnelly argues that support for human rights treaty making evaporates when it comes to efforts to create strong human rights treaties. Donnelly explains, “[t]he most important problem . . . was and remains the fact that a Between Power and Principle 12 with international law is widespread, but only because states only join treaties that require them to act no differently than they already do. Yet the empirical evidence shows this claim to be false. Countries frequently commit to treaties with which they cannot easily comply;25 they then, perhaps not so surprisingly, fail to meet those treaty commitments.26 Indeed, countries with the poorest practices—and hence the highest cost of compliance—are sometimes more likely to ratify treaties than those with better practices, all else held equal.27 Traditional interest-based theory thus leaves many unanswered questions: Why do states create and join treaties that provide for sometimes significant intrusions on state sovereignty, particularly when many countries later fail to abide by their requirements? When and why, that is, do countries create and join treaties that, on their face at least, go beyond the aspirational? And, most important for this Article, when and why do they comply or fail to comply with those treaties? B. Norm-Based Models Norm-based models of international law reject rationalist scholars’ contention that the consequentialist pursuit of self-interest alone can explain state behavior. While acknowledging that state behavior is often motivated by self-interest, normative scholars stronger international human rights regime does not rest on any perceived material interest of a state or coalition willing and able to supply it.” Jack Donnelly, International Human Rights: A Regime Analysis, 40 Intl Org 599, 616 (1986). Instead, the existing regime rests on states’ sense of moral interdependence—an interdependence that he suggests is strong enough to sustain a weak human rights regime but not strong enough to lead to the creation of a regime with authoritative decisionmaking powers. A stronger regime does not exist, in other words, because states are reluctant to surrender sovereign authority and do not see a stronger international human rights regime as presenting “a safe prospect of obtaining otherwise unattainable national benefits.” Id at 616–19. 25 See, for example, the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, 1465 UNTS 85 (Dec 10, 1984, entered into force June 26, 1987) (hereinafter Convention Against Torture); American Convention on Human Rights, 1144 UNTS 123 (Nov 22, 1969, entered into force July 18, 1978); International Covenant on Economic, Social, and Cultural Rights, 993 UNTS 3 (Dec 16, 1966, entered into force Jan 3, 1976); Convention on the Political Rights of Women, 193 UNTS 135 (Mar 31, 1953, entered into force July 7, 1954); Convention for the Protection of Human Rights and Fundamental Freedoms, 213 UNTS 221 (Nov 4, 1950, entered into force Sept 3, 1953). Each of these treaties requires international inspection of member countries’ human rights practices. 26 While many of the treaties cited in note 25 have strong requirements on their face, the enforcement and monitoring provisions are for the most part quite weak. Hence countries can and do engage in widespread noncompliance without meeting rebuke. See Hathaway, 111 Yale L J at 1951, 1976–88 (cited in note 17). 27 See, for example, Hathaway, 111 Yale L J 1935; Oona A. Hathaway, Why Do Countries Commit to Human Rights Treaties? (2004) (unpublished manuscript, on file with author). Between Power and Principle 13 contend that it is also motivated by the power of principled ideas—ideas that are not given by nature but are themselves constructed through interaction among individuals, groups, and states. There is a rich normative scholarship in both political science and law. In political science, the norm-based scholarship is built on the insights of “constructivist” theory.28 In this view, interest-based scholars are wrong to assume that states engage only in consequentialist pursuit of objective self-interest. Rather, states internalize norms and act in accordance with them because they understand them to be correct or appropriate.29 Moreover, in contrast with the rationalist approach, the normative approach argues that transnational actors and their interests are not fully formed or unchanging. Rather, they are constituted or “constructed” by and through interaction with one another. In other words, “[t]he international system can change what states want.”30 International law can change state action, in this view, “not by constraining states with a given set of preferences from acting, but by changing their preferences.”31 The legal norm-based scholarship starts with the assumption that nations obey international law “almost all of the time.”32 It also takes international law as a given, for the most part assuming that treaties exist and states join them without seeking to explain why or 28 Some notable examples of constructivist work are as follows: Finnemore, National Interests in International Society at 2 (cited in note 7) (“[S]tates are embedded in dense networks of transnational and international social relations that shape their perceptions of the world and their role in that world. States are socialized to want certain things by the international society in which they and the people in them live.”); Margaret Keck and Kathryn Sikkink, Activists Beyond Borders: Advocacy Networks in International Politics 3 n 6 (Cornell 1998) (noting the constructivist influence on their work); Alexander Wendt, Social Theory of International Politics (Cambridge 1999) (defending an international system constructivist theory that draws especially on structurationist and symbolic interactionist sociology); John Gerard Ruggie, What Makes the World Hang Together? Neo-utilitarianism and the Social Constructivist Challenge, 52 Intl Org 855 (1998) (providing an analytical account of social constructivism in current international relations); Alexander Wendt, Anarchy is What States Make of It: The Social Construction of Power Politics, 46 Intl Org 391 (1992) (developing a constructivist theory of power politics). 29 They may also do so simply out of a sense of habit. See James N. Rosenau, Before Cooperation: Hegemons, Regimes, and Habit-Driven Actors in World Politics, 40 Intl Org 849, 861–74 (1986) (examining countries as habit-driven actors). 30 Finnemore, National Interests in International Society at 5 (cited in note 7) (emphasis in original). 31 Id at 5–6. 32 See, for example, Harold Hongju Koh, Why Do Nations Obey International Law?, 106 Yale L J 2599 (1997) (presenting a theory of compliance that combines managerial and fairness approaches with analyses of interaction, interpretation, and internalization of international legal norms); Abram Chayes and Antonia Handler Chayes, The New Sovereignty: Compliance with International Regulatory Agreements 3–9 (Harvard 1995) (presenting a theory of compliance in which practitioners assume a general propensity of states to comply with international obligations). Between Power and Principle 14 when they do so. Hence these theories focus almost exclusively on the question of compliance: Why, they ask, do states obey international law most of the time? In legal scholarship, two separate variants of normative theory have gained widespread attention—fairness theory and legal process theory. The fairness model primarily associated with Thomas Franck focuses on the perceived fairness of the legal obligations at issue: A “fair” legal obligation exerts a “compliance pull” that leads states to comply with it.33 A second legal normative account instead focuses on legal process, with separate branches of the theory focusing on horizontal and vertical interactions among countries. Abram and Antonia Chayes offer a model of horizontal legal process, which they entitle “managerial legal process.” In this view, states obey international law not because of sanctions,34 but because their prior agreement to do so creates an “obligation of obedience.”35 Harold Koh’s related “transnational legal process theory” focuses less on horizontal ties across states and more on vertical interactions within states and between the international and domestic arenas. In his view, state behavior is influenced by international law through a process of “interaction, interpretation, internalization, and obedience,” by virtue of the efforts of various agents of internalization, including transnational norm entrepreneurs.36 A central step in this process is legal internalization—“when an international norm is incorporated into the domestic legal system through executive action, judicial interpretation, legislative action, or some combination of the three.”37 When that internalization is complete, he argues, states comply. Norm-based scholarship offers important corrective to rationalist theories by focusing attention on the powerful role of ideas in international law. Many norm-based accounts also encourage attention to the role and influence of nonstate actors that are often ignored in traditional interest-based accounts. Yet the central drawback of such theories is that they do not yield much in the way of specific expectations for state behavior. Perhaps the only consistent prediction that emerges is that states can, for the most part, be expected 33 Thomas M. Franck, Fairness in International Law and Institutions 7–9 (Clarendon 1995). Fairness, in Franck’s view, has both a substantive and procedural component: To be fair, rules must lead to distributive justice and they “must be arrived at discursively in accordance with what is accepted by the parties as right process.” Id at 7 (emphasis in original). 34 Abram Chayes and Antonia Handler Chayes, The New Sovereignty: Compliance with International Regulatory Agreements, 2–3, 34–67 (Harvard 1995). 35 Id at 115–16. To deal with the few states that fail to meet this obligation, they argue for “managing” compliance by, among other things, ensuring transparency of legal requirements and of parties’ success or failure in meeting them, creating a mechanism for resolving disputes under the regime, and building capacity for compliance. Together, these processes will persuade noncomplying states to conform their behavior to the requirements of the legal regime. 36 Harold Hongju Koh, The 1998 Franckel Lecture: Bringing International Law Home, 35 Houston L Rev 623, 644–55 (1998). See also Harold Hongju Koh, How Is International Human Rights Law Enforced?, 74 Ind L J 1397, 1413–14 (1999). 37 Koh, 106 Yale L J at 2657 (cited in note 32). Between Power and Principle 15 to join and comply with treaties. Yet this proposition finds mixed empirical support, as detailed in Part IV.38 To be sure, norm-centered theory does allow us to trace after the fact why some laws succeed while others do not, and it offers good reason for thinking that norms matter in the formulation of key areas of international law. But current accounts do not provide a clear guide in advance to which laws will succeed and which will not. Which norms will be internalized and which not, and why? Why are some laws persuasive and others not? Even when a norm-based approach does provide some guidance (fairness theory, for example, tells us that laws that are substantively and procedurally fair will be followed, and managerial theory predicts that laws with clear and specific requirements are more likely to be followed than those that impose vague restrictions), it does not explain why some states commit themselves more readily than do others. Finally, current norm-based theories do little to help us account in advance for state-to-state variation in compliance: What makes some states more likely than others to comply with treaties? C. A Promising Convergence A more recent strand of scholarship in the rationalist vein offers the promise of finding some common ground between interest-based and norm-based approaches. This model, termed the “liberal institutionalist” perspective by its proponents (sometimes also referred to as “institutional liberalism”), addresses some of the shortcomings of existing rationalist accounts. The theory, which has been developed and applied by both legal and political science scholars,39 opens the black box of domestic politics that is largely 38 Normative theories appear to fare no better than rationalist theories when tested against the available empirical evidence. As I show in greater detail in Part IV, the predictions of these theories not only find little support in the evidence, but are in some cases clearly contradicted by it. Most notably, compliance with treaties is much less widespread than normative theories would predict. See Hathaway, 111 Yale L J at 1987 (cited in note 17). Moreover, countries with practices that are more consistent with a given treaty (indicating a stronger ideological commitment to the norms embodied in the treaty) are no more likely than those with less consistent practices to commit to the treaty. See Hathaway, Why Do Countries Commit to Human Rights Treaties? (cited in note 27). 39 This perspective is put forward most prominently in political science scholarship by Andrew Moravcsik. See generally Moravcsik, 54 Intl Org 217 (cited in note 7); Andrew Moravcsik, Taking Preferences Seriously: A Liberal Theory of International Politics, 51 Intl Org 513 (1997) (positing a “nonideological” and “nonutopian” formulation of liberal international relations theory). See also Andrew Moravcsik, Negotiating the Single European Act: National Interests and Conventional Statecraft in the European Community, 45 Intl Org 19, 27 (1991) (“[The theory of] intergovernmental institutionalism . . . locates the sources of regime reform not only in the changing power distribution but also in the changing interests of states. States are not ‘black boxes’; they are entities entrusted to governments, which themselves are responsible to domestic constituencies.”). In international law scholarship, the most prominent advocate is Anne-Marie Slaughter. See, for example, Laurence R. Helfer and Anne-Marie Slaughter, Toward a Theory of Effective Between Power and Principle 16 unexamined by other interest-based scholars, and looks to the political institutions, interest groups, and state actors that shape state preferences to explain state behavior in the international arena.40 In this view, states pursue the aims preferred by “powerful domestic interest groups enfranchised by representative institutions and practices.”41 Hence, state behavior is the result of complex interactions between political players at the domestic level, and cannot be explained as simply resulting from power-maximizing behavior or strategic calculation by a unitary actor.42 This strand of interest-focused thought, like most norm-based explanations of state decisions to join treaties, calls for us to peer inside the state, looking for the individuals and groups that influence governments through political institutions and social practices. The Supranational Adjudication, 107 Yale L J 273, 277–78 (1997) (noting the critical role that domestic government institutions play in securing compliance with supranational adjudications); Anne-Marie Slaughter, International Law in a World of Liberal States, 6 Eur J Intl L 503, 504 (1995) (integrating a theory of international law with liberal international relations theory and its accepted assumption that states’ domestic political structures and ideologies alter the way states behave); Anne-Marie Slaughter, The Liberal Agenda for Peace: International Relations Theory and the Future of the United Nations, 4 Transnatl L & Contemp Probs 377, 397–98 (1995) (articulating a conception of the UN based on liberal international relations theory and its focus on state-society, rather than state-state, relations); Anne-Marie Slaughter Burley, International Law and International Relations Theory: A Dual Agenda, 87 Am J Intl L 205, 207 (1993) (“Liberals focus not on state-to-state interactions, at least not in the first instance, but on an analytically prior set of relationships among states and domestic and transnational civil society.”) (emphasis in original); Anne-Marie Burley, Law Among Liberal States: Liberal Internationalism and the Act of State Doctrine, 92 Colum L Rev 1907, 1920–21 (1992) (examining the transnational legal relations between liberal and nonliberal states and exploring the ways differing domestic political institutions can shape how states conduct their interstate relations). Moravcsik outlines three separate variants of liberalism—ideational liberalism, commercial liberalism, and republican liberalism. Moravcsik, 51 Intl Org at 515. 40 As Andrew Moravcsik puts it: “Societal ideas, interests, and institutions influence state behavior by shaping state preferences, that is, the fundamental social purposes underlying the strategic calculations of governments.” Moravcsik, 51 Intl Org at 513 (cited in note 39). 41 See id at 519–20. 42 In a recent article on the origins of the postwar European human rights regime, for example, Andrew Moravcsik uses the liberal approach—specifically a variant he terms “republican liberalism”—to explain the creation of the Convention for the Protection of Human Rights and Fundamental Freedoms, 213 UNTS 222 (Nov 4, 1950, entered into force Sept 3, 1953. In Moravcsik’s republican liberal perspective, governments in power prefer to maintain maximum discretion, yet they also wish to reduce political uncertainty. Where political uncertainty is great, states may be willing to surrender discretion to human rights regimes, for example, in order to constrain the actions of future governments. International treaties like the ECHR, in other words, offer governments a means of “‘locking in’ particular preferred domestic policies.” Moravcsik, 54 Intl Org at 225–26 (cited in note 7). Between Power and Principle 17 two approaches differ primarily in the assumptions made about the motivations of these actors and the source of their interests. Whereas liberal theory assumes that the relevant actors are motivated by objective self-interest,43 normative theory focuses attention on actors presumably motivated primarily by ideas—ideas that are constructed through interactions between and among states and nonstate actors.44 Norm-based theory also places greater emphasis than does liberal theory on horizontal connections across states and less emphasis on domestic regime type. The two approaches also share similar weaknesses. Liberal theory, which focuses attention on substate dynamics, is better positioned than other interest-based theories to account for the existence of treaties that might detract from, rather than benefit, unified state interests in security, sovereignty, or wealth. Yet liberal theory, like norm-based theories, tends to be descriptive rather than predictive. Both the liberal and normative approaches can be used to construct a persuasive explanation of state action after the fact, but they have tended to be less useful for predicting state behavior in advance. And to the extent that they have yielded predictions, those predictions have generally been limited in scope.45 43 See Andrew Moravcsik, Liberal International Relations Theory, in Colin Elman and Miriam Fendius Elman, eds, Progress in International Relations Theory 159, 161 (MIT 2002) (“The first assumption [of liberal international relations theory] is that the fundamental actors in international politics are rational individuals and private groups, who organize and exchange to promote their interest.”); Moravcsik, 51 Intl Org 513 (cited in note 39) (arguing that in liberal international relations theory, social ideas and institutions act to shape state preferences). There has been an effort to reincorporate idea-driven and interest-driven accounts in political science. Notable among them is the edited volume, Judith Goldstein and Robert O. Keohane, Ideas & Foreign Policy: Beliefs, Institutions, and Political Change (Cornell 1993). Yet even though they give much greater weight to the role of ideas in politics, they retain a central assumption that the relevant actors are motivated by self- interest. They explain: In general, we see ideas in politics as playing a role akin to that enunciated by Max Weber early in this century: “Not ideas, but material and ideal interests, directly govern men’s conduct. Yet very frequently the ‘world images’ that have been created by ideas have, like switchmen, determined the tracks along which action has been pushed by the dynamic of interest.” Judith Goldstein and Robert O. Keohane, Ideas and Foreign Policy: An Analytical Framework, in Goldstein and Keohane, Ideas & Foreign Policy, 1, 11–12 (internal citations omitted). 44 As Martha Finnemore puts it, “Material facts do not speak for themselves, and attempts to make them do so have limited utility.” Finnemore, National Interests in International Society at 6 (cited in note 7). An example in the legal literature can be found in Peter H. Schuck, Citizens, Strangers, and In-Betweens 91–138 (Westview 1998) (exploring the role of ideas in producing the Immigration Act of 1990, Pub L No 101–649, 104 Stat 4978 (1990), codified in relevant part throughout 8 USC). 45 Liberal theory gives insight into why democracies might act differently from nondemocracies, and republican liberal theory gives insight into why less established democracies might act differently from more established democracies. But liberal theories as Between Power and Principle 18 Despite these drawbacks, the overlap between these two perspectives offers a potential starting point for building a coherent integrated theory of state behavior under international law that draws on the insights and strengths of both approaches. This Article begins to do just this. I begin this project in the next Part by first seeking to understand the unique ways in which international treaty law functions. How is international treaty law different from domestic law and how do those differences affect its ability to shape state behavior? As will become clear, answering these questions gets at the crucial features of international treaty law—and is therefore the first step toward better understanding when and how treaties guide states. II. The Nature of International Law Is international law really “law?” Political scientists and legal scholars outside the international law field regularly raise this question. In doing so, they often point to the fact that international law lacks one or more qualities that are usually believed to be essential components of what we call “law.” Their challenge is therefore not simply semantic. It is instead meant to suggest that international law is really of little or no consequence because it has no power independent of the sanctions attached to it. In turn, international law’s defenders generally respond that the suggestion that international law is meaningless flies in the face of daily experience, and that international law carries independent weight much in the same way that domestic law does. Although the debate over whether international law is really “law” and the debate over how best to explain the relationship between international law and state behavior usually appear to be separate, the two are at bottom about much the same thing. In fact, the divide between normative and rationalist scholarship outlined in Part I can be seen, at root, as a disagreement over the nature of international law. Is international law more like a tax law or a gunman’s order to hand over one’s wallet? In other words, does international law create a legitimate legal obligation such that states are compelled to abide by it even if they will not face a penalty for failing to do so or is it simply a command that will be followed only if backed by sanctions?46 The vast majority of norm-centered scholars would say the former.47 The vast majority of interest-centered scholars would say the latter.48 a whole do not help us explain variation within these categories. Why do some nondemocracies commit and comply with treaties and others do not? Why do some weaker democracies commit to and comply with treaties while others do not? Liberal theories do not, as of yet, provide any good answers to these questions. Moreover, the theories remain ill-equipped to explain state action regarding treaties that are poorly enforced; indeed, there is no point to committing to a treaty to bind one’s successors if the treaty does not in fact bind them. 46 To some degree, this debate conflates two separate issues. The first is whether what we call international law is in fact law such that it generates a legitimate legal obligation on the part of those who are its subjects. The second is what leads states to abide or not by such obligations. The norm-based scholarship often appears to assume that if a legal obligation is legitimate (and is perceived as such), states will not only be obligated to comply but indeed will be likely to comply. Of course, this is not necessarily true. For discussions of Between Power and Principle 19 Both sides of this debate err, I believe, in taking an all-or-nothing approach. International law is neither just like domestic law, nor is it inconsequential. Instead, it differs from domestic law in ways that affect—but do not eliminate—its ability to influence state behavior. Two central differences stand out: First, international treaty law is voluntary— states are not bound by it unless they accede to it; and second, international law lacks a single sovereign with the power to enforce the law. I consider these two characteristics in turn, and then move to their implications for my theory of state behavior.49 related issues, see, for example, Ronald Dworkin, Law’s Empire 176–224 (Harvard 1986) (arguing that a legislative principle of political integrity—which asks lawmakers “to try to make the total set of laws morally correct”—is the primary principle in American political practice, from which adjudicative legitimacy flows); Jurgen Habermas, Communication and the Evolution of Society 178–79 (Beacon 1979) (Thomas McCarthy, transl) (“[L]egitimation conflicts flare up only over questions of principle . . . Such conflicts can lead to a temporary withdrawal of legitimation; and this can in certain circumstances have consequences that threaten the continued existence of a regime.”); Max Weber, Economy and Society: An Outline of Interpretive Sociology 31 (Bedminster 1968) (Guenther Roth and Claus Wittich, eds) (arguing that social order perceived to be legitimate is more stable than order based on expediency or habit). 47See, for example, Thomas M. Franck, Legitimacy in the International System, 82 Am J Intl L 705, 706 (1988) (arguing that compliance with international obligations is “secured at least in part by perception of a rule as legitimate by those to whom it is addressed”); Franck, Fairness in International Law at 8 (cited in note 33) (arguing that a belief in the international law’s legitimacy encourages compliance); Koh, 106 Yale L J at 2645–58 (cited in note 32) (discussing voluntary obedience as a preferred enforcement mechanism in international law); Koh, 25 Houston L Rev at 644–55 (cited in note 36) (arguing that internal acceptance of international rules is a four phase process whereby a provoker “binds” a coerced party to obey as part of an internal value set); Koh, 74 Ind L J at 1414 (cited in note 36) (same); Koh, 26 Yale J Intl L at 305 (cited in note 13); Chayes and Chayes, The New Sovereignty (cited in note 33) (describing a cooperative, problem-solving, as opposed to a coercive, approach to international law); Schachter, 8 Va J Intl L at 307 (cited in note 12) (arguing that obligatory norms, established through the same, consistent five processes, are the basis of obligation in international law); Higgins, Problems and Process at 16 (cited in note 12) (arguing that few international lawyers believe effective sanctions are necessary for international law to exist “or [believe] that sanctions predicate the existence of particular norms of international law.”); Jenks, Law, Freedom and Welfare at 83–100 (cited in note 12) (arguing that the basis of states’ obligations in international law is the will of “the world community”). 48 See, for example, Jack L. Goldsmith and Eric A. Posner, The Limits of International Law (2005); Downs, Rocke, and Barsoom, 50 Intl Org at 380 (cited in note 24) (arguing that the depth of cooperation in international agreements is linked to the level of enforcement). 49 The discussion that follows is influenced by, but distinct from, the classic discussion of the legal character of international law in H.L.A. Hart’s chapter on international law in The Concept of Law. Hart, The Concept of Law at 208–31 (cited in note Between Power and Principle 20 A. International Treaty Law Is Voluntary A first defining characteristic of international treaty law is the voluntary nature of the legal obligation it imposes. Treaties operate directly on states, but if a state does not consent to an international treaty, it is clearly not bound by its provisions.50 This aspect of international treaty law stands in stark contrast to domestic law.51 An individual person cannot decide whether, for example, the property, torts, and criminal laws of the state in which she resides apply to her. They apply to her whether she likes it or not. In fact, they apply to her even if she does not know that they exist.52 Of course, she can seek to change a law she views as wrong or illegitimate through the legislative process or she can move to a different jurisdiction and in this way “choose” the laws that apply to her, but as long as she remains in the jurisdiction she must continue to act in accordance with the 12) (exploring the “two principal sources of doubt concerning the legal character of international law”: first, the “adverse comparison of international law with municipal law” and, second, the “obscure belief that states are fundamentally incapable of being subjects of legal obligations.”) Id at 210–11. 50 See Restatement (Third) of Foreign Relations Law of the United States pt I, introductory note at 18 (1987) (“Modern international law is rooted in acceptance by states which constitute the system.”); Vienna Convention on the Law of Treaties, Art 19, 1155 UNTS 331, 341 (1969) (May 23, 1969, entered into force Jan 27, 1980) (“A treaty does not create either obligations or rights for a third State without its consent.”); Curtis A. Bradley and Jack L. Goldsmith, Treaties, Human Rights, and Conditional Consent, 149 U Pa L Rev 399, 436–37 (2000) (“One of the most established principles in international law is that ‘in treaty relations a state cannot be bound without its consent’”), quoting Reservations to the Convention on the Prevention and Punishment of the Crime of Genocide, 78 UNTS 277 (Dec 9, 1948, entered into force Jan 12, 1951); Louis Henkin, International Law: Politics and Values 28 (Kluwer 1995) (“For treaties, consent is essential. No treaty, old or new, whatever its character or subject, is binding on a state unless it has consented to it.”); E. Jane Ellis, International Law and Oily Waters: A Critical Analysis, 6 Colo J Intl Envir L & Pol 31, 38 (1995) (“According to the prevailing view of international law, states cannot be bound to an international convention unless they agree to be so bound and have voluntarily ratified that convention.”). 51 Of course, those who take the view that obligation under domestic law rests on a theory of constructive consent might see this contrast less sharply. Nonetheless, even in that view, international law differs in requiring express case-by-case consent for much of its provisions. 52 At common law, “every one is conclusively presumed to know the law” and hence ignorance of the law is generally not a viable defense. State v Woods, 107 Vt 354, 179 A 1, 2 (1935) (holding that a woman prosecuted for adultery could not assert the defense that she mistakenly believed she was legally divorced). See John Selden, Table Talk 30 (Huntington Library ed 1689) (“Ignorance of the Law excuses no man; not that all Men know the Law, but because ‘tis an excuse every man will plead, and no man can tell how to confute him.”) (emphasis removed). Between Power and Principle 21 requirements of its laws.53 In short, an individual is legally bound by the domestic law of the jurisdiction in which she acts; she has no choice in the matter. The same is frequently not true of the subjects of international law. Voluntariness thus defines an important characteristic of international treaty law that distinguishes it from domestic law. Whereas domestic law binds those within its jurisdiction regardless of their assent,54 international treaties do not apply unless those who would be subject to them agree to be bound. This has important consequences for international treaties’ effects on state behavior. In the next Part, I explore the consequences of the voluntariness of international treaty law for its ability to shape the behavior of those it seeks to govern. First, however, I consider a second characteristic of international law—its frequent lack of enforcement—that even more directly affects countries’ decisions of whether to comply with international treaty law. B. International Law Often Lacks Enforcement International law lacks a central governmental authority that has the power to enforce its commands. This characteristic is the source of much of the doubt about the “legal” nature of international law.55 In the decades immediately following World War II, 53 She can engage in civil disobedience, though if she does so she must recognize that she likely will be subject to a penalty for violating the law, even if she finds the law entirely illegitimate. 54 The exception to this rule is, of course, the state itself. It is commonly argued that the state, as sovereign, cannot be bound by law—domestic or international—unless it permits itself to be so bound. See note 72. One could therefore generalize the argument made herein to any circumstance in which the state commits itself to be bound by law. Indeed, one could see the adoption of domestic rule of law institutions such as an independent court system, a democratic electoral system, and institutional checks and balances as signaling devices that operate on the domestic level much as I argue treaties operate on the international level. See Part III.C. Moreover, those same institutions serve to ensure that the state observes not only its international legal commitments, but also domestic legal limits on its actions as well. See Part III.A. The key difference regarding the voluntariness of law may therefore be between law that operates on sovereign states and law that operates on individuals. Yet, to the extent that a larger proportion of laws at the international level operate on states rather than individuals, the distinction between the domestic and international arenas remains important. 55 Numerous scholars have discussed the issue. See, for example, Franck, Fairness in International Law at 707–13 (cited in note 33); Kelsen, General Theory of Law and State at 328–41 (cited in note 12) (examining whether international law can be described in terms of “rules of law”); Fernando R. Teson, A Philosophy of International Law 16–22 (Westview 1998) (discussing the question raised by the Kantian notion of international order as mirroring domestic order: must a “successful system of international law” be “centralized into a super-state with a world government”?); Alan Watson, The Nature of Law 77–81 (Edinburgh 1977) ; Michael Barkun, Law Without Sanctions: Order in Primitive Societies and the World Community 65 (Yale 1968) (discussing international law as a form of Between Power and Principle 22 enforcement of international law looked almost nothing like law enforcement in an effective domestic legal system. The central adjudicatory body, the International Court of Justice, did not possess compulsory jurisdiction and its judgments were few. The international bodies that most resembled a legislature and executive—the United Nations General Assembly and Secretariat—were considered weak and ineffective. To top it off, the United Nations Charter had outlawed all use of aggressive force by any member state against another, even for the purpose of law enforcement, unless the force was exercised in self-defense or with the express consent of a Security Council that was paralyzed by Cold War animosities.56 Several decades later, the institutions of the international system are far stronger. Although the international community does not have a police force or military that stands ready to penalize violations of international law, the United Nations can, and regularly does, send troops provided by member nations to prevent or quell hostilities. While the international court system remains a diffuse patchwork of overlapping jurisdictions, it has “horizontal law,” defined as law that is not grounded upon the police power of the state); James Brown Scott, The Codification of International Law, 18 Am J Intl L 260, 261–64 (1924) (discussing the roots of international law as a series of rules or usages of civilized states which had “hardened into rules through repeated practices”); Fawcett, The Law of Nations at 11 (cited in note 12) (arguing that international law “has its own sanctions, seldom imposed by force or command . . . natural sanctions, slow to mature and gradual in effect, but made compelling by the growing interdependence of the world”). The question of whether enforcement is necessary to what we call “law” is of course not limited to the international context. Philosophers of law have long debated whether “law” can exist if not backed by sanctions. Undoubtedly a threat of sanction is not sufficient to create law—for if it were, a gunman’s order to a victim to hand over his money on threat of death would create a legal obligation. Hart, The Concept of Law at 80 (cited in note 12). But is it necessary? Many have argued that it is not. Even absent a threat of sanction for its violation, they claim, what we call “law” carries with it a weight—it imposes an obligation on those who are subject to it. While there may be some persons who are concerned with and follow legal rules only “because they judge that unpleasant consequences are likely to follow violation,” id at 88, they are outnumbered in a viable legal order by those who see legal rules as more than that. See Jeffrie G. Murphy and Jules L. Coleman, Philosophy of Law: An Introduction to Jurisprudence 29 (Westview rev’d ed 1990) (“In any viable and stable legal order . . . the majority of citizens (and certainly the majority of those involved in some official capacity in enacting or enforcing the rules) must . . . see [rules] as standards of criticism and justification.”). The majority of those governed by the system must take what Hart calls an “internal” attitude toward the rules of the legal order—they see violations of the law as “not merely a basis for the prediction that a hostile reaction will follow but a reason for hostility.” Hart, The Concept of Law at 88 (emphasis in original). For those who take an internal point of view, legal rules are used “as guides to the conduct of social life, as the basis for claims, demands, admissions, criticism, or punishment, vis., in all the familiar transactions of life according to rules.” Id. In this way, the sense of obligation that attends to legal rules in a functioning legal system gives them a power wholly apart from and independent of the sanctions attached to them. 56 UN Charter Arts 39–51. Between Power and Principle 23 grown exponentially in strength during the post-World War II era.57 The General Assembly and Secretariat of the United Nations have also gradually gained power. In part as a consequence of this emerging system of strengthened international institutions, enforcement of international law has grown markedly in the last half-century. Yet there remain vast domains in which enforcement of international law is nonexistent or, at best, sporadic. Further, international law continues to rely heavily on private and peer-to-peer enforcement. This lies in contrast with law in a functioning domestic legal system. In the domestic context (at least in one considered “lawful”), there exists a central legitimate enforcement authority—or, as in the United States, overlapping federal, state, and local authorities—that monitors and enforces the laws. Though even in the most effective domestic legal systems there will undoubtedly be many instances of illegal conduct that do not meet with a penalty, in functioning domestic legal systems a central body keeps such instances to a minimum. Moreover, the possibility of enforcement is nearly always present, even if actual enforcement is not. The regular absence of central enforcement thus defines a central characteristic of international law. What is the consequence of this difference for the ability of international law to influence state behavior? Interest-based scholars tend to conclude that international law that is not backed by sanctions is not effective.58 Norm-based scholars, by contrast, conclude that international law need not be backed by sanctions to influence state behavior.59 In this latter view, while law may be backed by sanctions that give those governed by it an additional reason to act, the law differs from the gunman’s threat in that an obligation—and hence a reason to act—arises independently from international treaty law. Each of these positions is too extreme. The frequent lack of sanctions in international treaty law does not deprive treaties of all influence. Nor is the sporadic enforcement of international treaty law inconsequential. It does affect, but not obviate, the ability of international law to shape the behavior of those it seeks to govern. Like the voluntariness of international treaties, this aspect of international treaty law has important and poorly understood consequences for international law’s effect on state behavior. The 57 The post-World War II era has seen the rise of the European Court of Human Rights, which has compulsory jurisdiction; the International Tribunal for the Former Yugoslavia and the International Tribunal for Rwanda, which have compulsory, though quite limited, jurisdictions; the World Trade Organization’s Dispute Settlement Body and Appellate Body, which both hold jurisdiction over compulsory dispute resolution for WTO members; and the new International Criminal Court, among others. See Jennifer Martinez, Towards an International Judicial System, 56 Stan L Rev 429, 528 (2003) (characterizing these bodies as elements of an emerging international judicial system and concluding that participants in this system should “promote an institutional framework for cooperation among states, compliance with international law, and the maintenance and development of democratic, rights-respecting national governments.”). Finally, domestic courts are increasingly applying international law. See generally Andrea Bianchi, International Law and US Courts: The Myth of Lohengrin Revisited, 15 Eur J Int’l L 751 (2004) (discussing the use of international law in US Courts). 58 See note 48. 59 See note 47. Between Power and Principle 24 next Part of this Article is devoted to explaining how these qualities of international treaty law influence state behavior. That is, given what we now know about the special nature of international treaty law, how and when does it shape state behavior? III. An Integrated Theory of International Law Two points are clear: International treaty law is voluntary, and it often lacks enforcement. These two points, in turn, have enormous implications for the ways in which international law influences state action. States must voluntarily accept international treaty commitments in order for them to be binding. The voluntary nature of international treaty commitments necessarily has important implications for states’ compliance with those commitments. Moreover, the regular absence of enforcement in international treaty regimes means that once states commit to treaties, compliance must be motivated by more than the threat of external sanctions if it is to have any significant effect on state behavior. Hence any comprehensive theory of international law must provide an account of how these two characteristics influence state compliance with international law. The theory I propose here incorporates and moves beyond existing accounts of compliance with international law. The theory rejects the claim that where transnational legal enforcement is absent, international law cannot change state behavior. Yet it also rejects the claim that enforcement is irrelevant. It instead places such enforcement in a broader framework in which it plays an important, but not exclusive, role in generating compliance with international law. That broader framework includes two forces that create incentives that influence states’ decisions to commit to and comply with treaties: legal enforcement and collateral consequences. Legal enforcement is determined by the terms of the treaty and the enforcement of those terms as specified in the text. For example, the World Trade Organization can authorize trade sanctions against member states if they fail to adhere to the terms of the agreement.60 Collateral consequences arise from the anticipated reactions of individuals, states, and organizations to the state’s decision to commit to the treaty and then to abide or not to abide by its terms—reactions that fall outside the legal framework created by the treaty or its implementing legislation but nonetheless affect the state’s material and other interests. Each of these incentives is generated at both the domestic and transnational levels. At the domestic level, both legal enforcement and collateral consequences are generated by individuals and nongovernmental organizations (NGOs) (where their focus is on affecting the local domestic political process), mediated by domestic political institutions. At the transnational level, they are generated by foreign citizens, other states, NGOs (where the focus is on affecting action across state borders),61 and international organizations such 60 See Understanding on Rules and Procedures Governing the Settlement of Disputes, Art 22.2, WTO Agreement, Annex 2, 33 ILM 1226 (1994) (DSU). 61 See Keck and Sikkink, Activists Beyond Borders (cited in note 28) (describing transnational activist groups); Thomas Risse and Kathryn Sikkink, The Socialization of International Human Rights Norms into Domestic Practices: Introduction, in Thomas Risse and Kathryn Sikkink, eds, The Power of Human Rights 1, 5 (Cambridge 1999) (arguing in part that networks of domestic and transnational actors “challenge norm-violating Between Power and Principle 25 as the United Nations, the World Bank, the World Trade Organization, and the International Monetary Fund. A summary of the four categories of incentives and examples of each appear in Table 1.62 TABLE 1: SUMMARY OF INCENTIVES FOR STATE COMMITMENT AND COMPLIANCE Legal Enforcement Collateral Consequences Domestic Incentives for commitment and compliance arise from expected enforcement of the treaty, where the source of the enforcement is actors inside the state. Incentives for commitment and compliance are created by anticipated reactions of domestic actors to a government’s decision to commit to an international treaty. Transnational Incentives for commitment and compliance arise from expected enforcement of the treaty, where the source of the enforcement is actors outside the state. Incentives for commitment and compliance are created by anticipated reactions of transnational actors to a government’s decision to commit to an international treaty. In understanding how this four-part framework shapes state behavior, it is important to come back to the two central characteristics of international treaty law: its voluntary character and the frequent absence of central enforcement. The first characteristic, the voluntary nature of treaties, means that the effects of international laws are contingent on who agrees to be bound. Who agrees to be bound is, in turn, contingent on the law’s likely effects. The effect of international treaty law on states thus depends upon, and in turn influences, the choice of states to accept international legal commitments in the first place. Put another way, the issue of commitment to international governments by creating a transnational structure pressuring such [norm-violating] regimes ‘from above’ and ‘from below’”) (internal citation omitted). 62 This framework is influenced by the social norms literature. See, for example, Robert C. Ellickson, Order Without Law: How Neighbors Settle Disputes 126–27 (Harvard 1991) (delineating “first-party,” “second-party,” and “third-party” mechanisms for operating a system of social control). See also Richard A. Posner and Eric B. Rasmusen, Creating and Enforcing Norms, with Special Reference to Sanctions, 19 Intl Rev L & Econ 369, 370–72, 372 n 4 (1999) (examining norms and delineating six types of sanctions—automatic sanctions; guilt; shame; informational sanctions; bilateral costly sanctions; and multilateral sanctions—which they map onto Ellickson’s three-part framework). I am grateful to Bob Ellickson for conversations about this scholarship. Between Power and Principle 26 treaty law precedes that of compliance, as there is no issue of state compliance unless the state has committed. At the same time, the causal arrow also runs in the other direction: States considering whether to commit take into account the anticipated costs and benefits of the commitment. The anticipated costs of commitment, in turn, depend to a large degree upon the expected cost of compliance with the treaty commitment—how much, that is, the state expects to alter its behavior if it commits to the treaty. This interplay of commitment and compliance leads to a prediction that might otherwise appear counterintuitive: the more likely a state is to change its behavior to comply with a treaty, the more reluctant it will likely be to commit to it in the first place, all else being equal.63 I return to the implications of this claim in Part V below. The second distinctive characteristic of international treaty law—frequently weak or nonexistent enforcement—clearly undermines its ability to shape state behavior. But the integrated theory shows that international treaties that are not backed by sanctions can nonetheless affect state behavior. Most legal scholars would agree that domestic citizens abide by the law for a complex mix of reasons in addition to legal enforcement—including, among others, fear of retribution by the wronged party and concerns about the their reputation if others learn of their wrongdoing. Simiarly, there are many reasons why states might abide by treaties besides direct legal sanction. Thus, while transnational legal enforcement is the overriding concern of most existing models of state behavior, my approach suggests that it is only one of several factors that determine how international treaty law shapes state behavior. Like all of the existing accounts of state behavior outlined above, this account places the state at the center of the analysis.64 It does so for a simple reason: The state is, in fact, the primary subject of international law.65 It is the state that signs or ratifies a treaty, and it is the state that is subject to the treaty’s requirements once it has done so. In most cases, international treaty law applies not to individuals, nongovernmental organizations, corporations, or other entities, but to states and states alone. This is not to say that states are the only actors influenced by international treaty law. In most states, treaties are binding on entities operating within the state via domestic application of that state’s international commitments. I therefore focus on states not because international law affects only states, but because only states are in a position to accept or reject it and only states are directly bound by it once they accede to it. While the decision to comply with international treaty law is made on behalf of the state as a whole, the political process by which that decision is made involves the aggregation of sometimes competing preferences. Accordingly, the theory offered here includes an account of the role played by domestic political institutions and central domestic political actors. States are a conglomeration of individual actors, and those actors’ ability to influence 63 This relationship may be weaker in states where the governmental leaders’ time horizons are shorter. See note 71. 64 See Part I. 65 See Restatement (Third) of Foreign Relations Law § 101 at 22 (cited in note 12) (defining “international law” as consisting of “rules and principles of general application dealing with the conduct of states and of international organizations and with their relations inter se, as well as with some of their relations with persons, whether natural or juridical”). Between Power and Principle 27 state policy in the international arena is determined in large part by the political institutions that channel the exercise of state authority and power.66 Individual actors, in turn, are motivated by a complex mix of factors. Indeed, actors within a state have competing preferences.67 They can be expected to act in accordance with 66 While this approach disaggregates the state, it does not go so far as to consider a variety of sub-state factors that might play an important role in international law and politics in particular circumstances. For instance, the theory articulated here takes no account of the personalities of individual heads-of-state. That is because, although individual personalities might play an important role in particular instances, it is difficult—if not impossible—to incorporate these traits into a broader theory of state behavior, much less into a theory that claims to have any predictive value. To say this is not to suggest that such traits are unimportant or could not possibly be theorized; there may be room for a study that draws on social psychology or personality theory to systematize the place of these traits in international relations. But this variable is omitted, as are many other details that may be relevant, because including them would undermine the purpose of the project. For the theory presented herein to in fact be a theory, it must of course be more than a description of the relevant phenomena: It must describe a pattern in complex phenomena. It must therefore omit distracting details to serve as a guide to explaining not only what has come before but also what can be expected in the future. See F. A. Hayek, The Theory of Complex Phenomena, in F. A. Hayek, Studies in Philosophy, Politics, and Economics 22-42 (University of Chicago Press 1967) (offering a classic discussion of the role of theory in the social sciences and explaining that, “A theory will always define only a kind (or class) of patterns, and the particular manifestation of the pattern to be expected will depend on the particular circumstances . . . .”); Gary King, Robert O. Keohane & Sidney Verba, Designing Social Inquiry 100-05 (Princeton University Press 1994) (noting that a social scientific theory must be falsifiable). 67 The presence of enforceable international law creates opportunities for some domestic actors to use international law to achieve domestic policy objectives that might otherwise be difficult to achieve or maintain. Such efforts to use international law as leverage are likely to be found where those who have influence or control over foreign policy are different in their policy positions and goals from those who have influence or control over domestic policy. Furthermore, these efforts can be expected to be more pronounced in cases where control of government by one party is tenuous and hence those currently in control of foreign policymaking seek to make international commitments to constrain their successors. For related arguments, see Vreeland, Institutional Determinants of IMF Agreements (cited in note 7) (arguing that governments that are more constrained domestically often seek to use IMF agreements to push through unpopular policies that would otherwise be impossible to achieve); Vreeland, 24 Intl Pol Sci Rev 321 (cited in note 7) (arguing that as the number of veto players increases, executives are more likely to turn to IMF agreements); Keck and Sikkink, Activists Beyond Borders at 13 (cited in note 28) (putting forward a “boomerang” model of international politics); Moravcsik, 54 Intl Org at 225–43 (cited in note 7) (arguing that “international institutional commitments, like domestic institutional commitments, are self-interested means of ‘locking in’ particular preferred domestic policies . . . in the face of future political uncertainty.”); Peter Gourevitch, The Second Image Reversed: The Between Power and Principle 28 self-interest in many circumstances, as, for instance, when a state seeks to formulate trade policy.68 Yet they may also pursue goals that are not closely linked to their own material well- being.69 (This distinguishes my account from interest-based accounts of state behavior that do not accept that such actors can have motivations that are not principally derived from self-interest.)70 For example, government actors are often subject to pressure from groups and individuals who seek improvements in the human rights of peoples in other nations. I do not attempt to propose a theory of how individual and group preferences are formed. I instead accept that self-interest and normative ideals can both influence preference formation (and, indeed, self-interest may dominate under some circumstances and norms in others). I begin only with the assumption that individuals and groups have preferences (however formed) that they pursue in part by bringing pressure to bear on governments. Decisions are ultimately made by those wielding political power, which I refer to here simply as “the government.”71 I argue that country or treaty characteristics that make International Sources of Domestic Politics, 32 Intl Org 881, 911 (1978) (arguing that “[t]he international system is not only a consequence of domestic politics and structures but a cause of them”); Robert Putnam, Diplomacy and Domestic Politics: The Logic of Two- Level Games, 42 Intl Org 427, 457 (1988) (arguing that governments exploit “IMF pressure to facilitate policy moves that [are] otherwise infeasible internally”); Luigi Spaventa, Two Letters of Intent, in John Williamson, ed, IMF Conditionality 441, 463 (Institute for International Economics 1983) (arguing that IMF demands allowed the Italian government and Italian unions to force their constituencies to accept unpopular, but necessary, fiscal programs to help turn around Italy’s economic recession in the mid-1970s). This remains a fruitful area for future research. 68 See, for example, Oona A. Hathaway, Positive Feedback: The Impact of Trade Liberalization on Industry Demands for Protection, 52 Intl Org 575, 597 (1998) (examining the role of domestic interest groups in the formulation of domestic trade policy). 69 Moreover, societal groups may seek government action favoring their views for symbolic as well as more conventional self-interested reasons. As Gussfield puts it, “Affirmation through law and government acts expresses the public worth of one subculture’s norms relative to those of others, demonstrating which cultures have legitimacy and public domination. Accordingly it enhances the social status of groups carrying the affirmed culture and degrades groups carrying that which is condemned as deviant.” Joseph Gussfield, On Legislating Morals: The Symbolic Process of Designating Deviancy, 56 Cal L Rev 54, 58 (1968). 70 For more on the role of ideas in interest-based accounts, see note 41. 71 This designation is intended to refer to the individuals and institutions that make up government in the aggregate. To the extent that the process by which states express consent to a treaty generally requires the involvement and assent of several parts of government, this aggregation does not deviate substantially from reality. (Although in most states the executive branch is charged with negotiating treaties, most states require assent of other governmental actors before the treaty can be considered legally binding.) My ongoing research will explore whether any characteristics of parts of government might affect treaty commitment decisions in a predictable way. For instance, I will examine whether the number of years the executive has remaining in office affects states’ signature and ratification Between Power and Principle 29 commitment to international law less attractive to the government will make the state less likely to commit and, conversely, those characteristics that make commitment to international law more attractive to the government will make the state more likely to commit. The same is true of state decisions to comply with international legal commitments. What, then, makes commitment and compliance more or less attractive to a government? I turn to the four ways in which international law affects state behavior: domestic legal enforcement, transnational legal enforcement, domestic collateral consequences, and transnational collateral consequences. A. Domestic Legal Enforcement How each state will react to international laws (particularly those that are not enforced by international institutions or other transnational actors) depends in important part on its internal institutions for enforcement. Indeed, much of international law is obeyed primarily because domestic institutions create mechanisms for ensuring that states abide by their international legal commitments whether or not particular governmental actors wish it to do so. The internalization of international legal requirements and compliance with them depends on the extent to which those outside the government can be expected to act to enforce the state’s international legal commitments against the government. This in turn depends on what kind of domestic enforcement mechanisms the state possesses. Does it have a strong and independent judiciary to fairly adjudicate claims of litigants who believe that the state has failed to meet its international legal obligations? Are there sufficient protections for civil rights that individuals and groups can bring enforcement actions against the government without fear of reprisals? If a state does have such rule of law institutions in place, it can be expected to engage in domestic legal enforcement,72 even if little or no decisions (one might expect that executives with shorter time horizons might act to take advantage of the benefits of treaty membership, particularly where the costs are likely to accrue after the leader has left office). My preliminary tests do not appear to bear out any significant relationship. Perhaps this is due to the fact that even where a change in regime is expected, governments continue to act with regard for future constraints. A government in power facing an imminent change in regime may not expect itself to be constrained by a treaty commitment in the near future, but it may hope to return to power in the future, at which time it would be constrained. It may also be reluctant to harm its own future political prospects by committing to a treaty that severely constrains the successor government’s ability to act in a way that might be perceived as harmful to state interests. 72 A skeptic may object to this line of reasoning by claiming that it is nonsensical to talk of courts or others “enforcing” a government’s international legal commitments because they cannot do so unless the government allows them to do so. If this is true, as Roger Fisher pointed out nearly a half-century ago, then it is equally nonsensical to talk of enforcing much of constitutional law and criminal law. See Roger Fisher, Bringing Law to Bear on Governments, 74 Harv L Rev 1130, 1133 (1961): Even such hard, positive laws as the criminal and tax laws depend ultimately on compliance with them by the Government, and the general pattern is one of compliance . . . Between Power and Principle 30 transnational legal enforcement occurs.73 In states lacking such a system, however, it is more difficult for domestic actors to force the government to live up to its commitments.74 To argue this is not to argue that domestic interest groups in states with strong domestic rule of law states will always favor enforcement of treaties. Indeed, the opposite may sometimes be true (and powerful domestic interest groups that oppose treaty membership can create strong domestic collateral incentives for the government to avoid committing to the treaty, as will be discussed in more detail below). But once a treaty has been ratified by a state, individuals and groups that favor implementation of that treaty have access to a tool that would otherwise not be available to them to change state behavior in ways that are consistent with the treaty. Hence the existence of domestic enforcement does [This pattern] demonstrates that a pattern of governmental compliance can be secured without a supragovernmental police force. It is undeniably the case that when a state is the subject of domestic laws, those laws are, in a sense, “unenforceable,” for a sovereign state complies with domestic law only because it agrees to do so. This does not lead us to say, however, that domestic laws that apply to the state and its agents are unenforceable. In a state that observes the rule of law, the government commits to observing legal limits on its actions and permits enforcement of those limits, with only well-specified exceptions. 73 Alter describes this process at work in the European Community. See Karen Alter, Establishing the Supremacy of European Law: The Making of an International Rule of Law in Europe (Oxford 2001) (examining why European national courts changed national legal doctrines to accommodate the supremacy of European law). 74 Also important is whether the treaty is self-enforcing or not. Some states adopt a monist view of international law, under which international law is automatically incorporated into the legal system and is directly enforceable as such. In states with a monist view of international law, treaties are presumed to be self-executing unless otherwise specified. By contrast, in states with dualist systems, treaties must be implemented through legislation in order for its requirements to be enforceable. For more on this distinction and for two sides of a debate over how it applies in the United States, see Curtis Bradley, Breard, Our Dualist Constitution, and the Internationalist Conception, 51 Stan L Rev 529, 530 (1999) (“The monist view is that international and domestic law are part of the same legal order, international law is automatically incorporated into each nation's legal system, and international law is supreme over domestic law.”); and Michael P. Van Alstine, Federal Common Law in an Age of Treaties, 89 Cornell L Rev 892, 921-37 (2004) (detailing the over 400 treaties that the United States has ratified that contain self-executing provisions). Moreover, to the extent that states use reservations, understandings, and declarations (“RUDs”) to exempt theselves from elements of a treaty or to require implementing legislation for what would otherwise be a self-enforcing treaty, this reduces the domestic enforcement incentives for compliance with the treaty. This can be extremely problematic, because RUDs are significantly less visible and less well-understood than ratification. Hence a state may ratify a treaty in order to obtain the collateral benefits of membership, and then issue RUDs that make that commitment unenforceable. The availability of RUDs thus makes commitment more likely, but at the cost of weaker incentives for compliance. The desirability of RUDs thus deserves much more critical examination. Between Power and Principle 31 not require more domestic support for the treaty than opposition to it.75 Indeed, when domestic enforcement of a treaty commitment occurs through an independent court system, implementation of the treaty can be countermajoritarian. International law thus creates a more strongly observed obligation in states in which the government is constrained by independent courts that provide access to extragovernmental actors to challenge state action (and hence in which domestic enforcement is significant). Such states can be expected to adhere more closely to the terms of an international legal norm to which they have committed. They tend to take the law as creating a duty that they must satisfy, not as simply defining when and where sanctions will be levied. For some of these states, transnational sanctions may even be unnecessary to compliance—they will comply with the requirements of the law even if such sanctions for noncompliance are minimal or nonexistent. By contrast, states lacking internal institutions that allow those outside government to enforce a state’s legal commitments will be unaffected by domestic enforcement incentives. The legal obligation created by international law therefore can be and often is disregarded.76 The relationship identified here creates powerful (and counterintuitive) incentives that affect government decisions to commit or refuse to commit to international legal constraints. The more a government of a state expects to face domestic enforcement of the state’s international commitments, the more likely it is to expect to be required to change its practices to abide by international law if its practices are not already consistent with that law. And the more likely a state is to change its practices to abide by international law, the more costly and hence less attractive committing to it will appear. States that are more likely to engage in domestic enforcement of the terms of international legal agreements are therefore less likely to commit to them in the first place, all other things held equal. In other words, generally speaking, the more likely an international agreement is to lead to an improvement in a state’s practices, the less likely the state will join it. Of course, this is not to say that states with strong internal enforcement mechanisms never make international legal commitments. To the extent that such domestic enforcement tends to be found in states that also have practices more consistent with the requirements of international law, states with stronger domestic enforcement may be no less likely—and perhaps even more likely—to commit (because the cost of transnational legal enforcement is small or nonexistent). And to the extent that states with strong domestic enforcement systems evince a commitment to the rule of law that extends to the international realm (resulting in positive domestic collateral conseqeunces for committing to treaties), such states may be more likely to support an international treaty system by ratifying treaties. Nonetheless, the feedback effect between commitment and anticipated compliance leads to 75 This lies in contrast to domestic collateral effects discussed in Part III. C. 76 As Hart might put it, some states, but not all, live by the rules seen from the “internal” point of view. Hart, The Concept of Law at 55 (cited in note 12). Indeed, in the year 2000, only 34 out of 157 countries rated a 10 on a 1 to 10 scale of democracy (indicating that the country is very democratic), whereas 42 rated a 0 (indicating that it is not at all democratic). Nonetheless, 85 rated at least a 6 on the scale, and 59 rated at least an 8. For more on the democracy measure used in this Article, see Appendix B. Between Power and Principle 32 a specific prediction: All other things held equal, the more costly domestic enforcement is likely to be, the less likely a state is to commit to a treaty. The dynamic of “domestic enforcement” or “internalization” described here builds upon and enriches both the interest-based and norm-based accounts of the role of international law. For interest-based accounts, it has the effect of broadening the notion of enforcement to include internal enforcement efforts. And it gives norm-centered scholars a more detailed and precise mechanism to account for the process of international legal internalization. Moreover, it provides them with an answer to the charge that they cannot determine when and why some international rules will be more likely to be internalized than others. One need only look to the treaty terms (Is it self-executing? Does it require implementing legislation?) and the domestic institutions of member states (Can actors independent of the government compel it to abide by its international legal commitments?) to make such predictions. B. Transnational Legal Enforcement Whether or not sanctions are essential to law, interest-based theories are undeniably correct that, holding all else constant, international rules backed by transnational legal enforcement (a threat of some form of cross-border legal sanction) are likely to be more effective in changing the behavior of those bound by them than are rules that are never or rarely enforced in this way. International organizations charged with enforcing international laws can make it costly for a state to fail to comply with international laws to which it has subscribed. Moreover, states may receive reciprocal benefits when they abide by international laws that can be legally revoked if states fail to live up to their commitments. Together, these enforcement mechanisms (or their absence) generate incentives for states to comply (or not) with international law. The incentives created by transnational legal enforcement play a particularly important role in states that do not effectively internalize international legal commitments and hence do not accept international law as creating an independent obligation to act.77 For at least these states, sanctions provide a central reason to obey the law, and where enforcement is absent or minimal, compliance among these states suffers. Moreover, even those states that do effectively internalize international legal requirements may find additional motivation for compliance in sanctions and therefore may comply more effectively with international legal commitments that are enforced by transnational actors than with those that are not. As a consequence, international law not backed by such enforcement is less capable of shaping the behavior of those it is intended to govern. 77 These are states that have weak domestic legal enfocement (see Part III. A.) and that experience weak domestic collateral consequences (see Part III. C.). My claim that states that do not successfully internalize their legal commitments require sanctions to force them to comply is reminiscent of Hart’s insight that even in the best legal systems, there are individual actors that, in Hart’s terminology, adopt the “external view” on the legal rules. For such actors, Hart acknowledges, sanctions may be the only means of motivating compliance. Hart, The Concept of Law at 88 (cited in note 12). Between Power and Principle 33 Transnational legal enforcement can take a variety of forms. The transnational legal enforcement of the terms of the treaty can draw states into joining treaties by offering benefits to those who join. For example, a state may join a trade treaty because once it commits, other states will be required by the treaty’s terms to charge lower tariffs on its exported goods. To the extent that powerful government constituents benefit from these actions by other states, the government, too, will benefit and will hence favor adoption of the treaty. Hence transnational legal enforcement can create incentives that draw states into committing to the treaty. Such incentives also push states to comply with the treaty once they have joined, for they otherwise are unlikely to be able to continue receiving the benefits of membership. Transnational legal enforcement may also push governments away from committing to a treaty. Such enforcement can make treaty membership costly by increasing the likelihood that they will comply with its terms, thereby constraining the ability of governments to act in ways that would otherwise be unconstrained. Costs may be generated by enforcement of treaty terms by a treaty body charged with monitoring the terms of the treaty (for example, the World Trade Organization78). Transnational legal enforcement need not come from an international organization. It may also arise from treaty-authorized reciprocal enforcement through retaliation by parties to the treaty for violations of its terms. Many treaties permit members to engage in enforcement of the terms of a treaty to which they belong by engaging in reciprocal defection or unilateral or coalitional enforcement in retaliation for the failure of another member to meet the treaty’s terms. For instance, in the trade context, when a country defects from a treaty by charging higher than permitted tariffs, the harmed parties to the agreement might be authorized to retaliate with similar actions.79 78 The confrontation over United States steel tariffs in 2003 illustrates this point. Under threat of European and Asian trade sanctions authorized by the World Trade Organization under the DSU, 33 ILM 1226 (cited in note 60), President Bush lifted tariffs on imported steel that he had imposed the prior year. See David E. Sanger, Backing Down on Steel Tariffs, U.S. Strengthens Trade Group, NY Times A28 (Dec 5, 2003). 79 This is a form of legally sanctioned “tit-for-tat.” See Robert Axelrod, The Evolution of Cooperation 57–60 (Basic Books 1990) (discussing the implications of the “tit- for-tat” response in business, congressional deal-making, and the history of World War I). If not permitted by the treaty, such retaliatory activity is not legal enforcement but instead a collateral consequence of failure to perform as the terms of the treaty require. See text accompanying notes 85–101. It is worth noting that the incentives generated by the threat of retaliatory behavior may operate differently depending on the country giving and receiving the sanction. Indeed, the power of such enforcement depends to a large degree on the relative power of the countries involved. For example, the threat of retaliation by a country with a large market for a violation of a trade treaty will provide a greater incentive for compliance among its trading partners than the threat of retaliation by a country with a smaller market. Similarly, a country with a larger market may be better positioned to absorb the cost of retaliatory behavior and hence less likely to be motivated to comply by fear of retaliation. Between Power and Principle 34 To better understand decentralized enforcement of treaties, it is worth considering for a moment how international treaties often resemble contracts. As in other contractual contexts, states engaging in treaties may be thought of as arranging their transactions “without the aid of an independent enforcement mechanism whose powers are significantly greater than their own.”80 As Anthony Kronman has demonstrated, where there is no central coercive power (and, to a lesser extent, even where there is), those who wish to enter into a nonsimultaneous exchange must rely for their security on one of several available security- enhancing devices, including what he terms “hostages,”81 “collateral,”82 “hands-tying,”83 and “union” of parties’ self-interest.”84 Such devices, when incorporated into treaties, allow for exchanges to be made and followed through upon even in the absence of sovereign enforcement (and even in the absence of anything resembling a functioning legal system). This is no less true in the international context than in the domestic. States seeking to create agreements in the absence of a central enforcement authority (or in the presence of a weak one, such as the United Nations or the World Trade Organization), may use a variety of techniques to secure their agreements. For example, in the trade arena, states charge lower tariffs and in exchange receive lower tariffs. If a state defects, the other state might retaliate by defecting as well (thus taking back the “collateral” of reciprocal lower tariffs). Treaties that deal with issues of national security may utilize “hostages”—placing one another’s citizens at reciprocal risk to secure the agreement.85 And many treaties attempt to use the negative impact of defection on the state’s reputation as a “hands-tying” mechanism.86 When 80 Anthony T. Kronman, Contract Law and the State of Nature, 1 J L Econ & Org 5 (1985). 81 Id at 12–15 (defining a “hostage” as anything of value to the hostage-giver, but not necessarily the receiver, that is handed over to secure a promise of future performance). 82 Id at 15–18 (defining “collateral” as an asset offered to secure a promise to perform that has value to both the recipient and the giver). 83 Id at 18 (defining “hands-tying” as an action that makes “a promise more credible by putting it out of the promisor’s power to breach without incurring costs he could otherwise have avoided”). 84 Id at 21 (defining “union” as “any arrangement that seeks to reduce divergence [of the self-interest of the parties to an agreement] by promoting a spirit of identification or fellow-feeling between the parties”). 85 Kronman in fact characterizes the Anti-Ballistic Missile Treaty of 1972 and the power possessed by each party to destroy one another’s populations as a mutual exchange of hostages. Id at 12–13. Of course, treaties often utilize more metaphorical types of hostages as well. 86 For more on this use of reputation, see Part IV.B. One could recast some of the literature on the special role of liberal states in international law in light of the observation that “hands-tying” mechanisms conduce to international agreements. Liberal states are arguably better situated to engage in “hands-tying” commitments because they are more likely to abide by the rule of law and have internal controls that limit government’s ability to act. See, for example, Helfer and Slaughter, 107 Yale L J 273 (cited in note 39); Slaughter, 6 Eur J Intl L at 533 (cited in note 39) (“[T]he domestic constraints on liberal governments are more likely to create the conditions in which States entering into an international agreement Between Power and Principle 35 these techniques are used in a treaty, they give rise to transnational legal enforcement that creates incentives for and against commitment to and compliance with a treaty. C. Domestic Collateral Consequences International treaties not only affect state behavior through domestic and transnational legal enforcement. Treaties also give rise to what I call collateral consequences—that is, the anticipated consequences for, among other things, foreign aid and investment, trade, and domestic political support. Collateral consequences arise when domestic and transnational actors premise their actions toward a state on the state’s decision to accept or not accept international legal rules. The category is meant to encompass not just the effect of treaties on countries’ reputations, though this is an important element of it, but also explicit and implicit conditioning of benefits that states value on the state’s decision to ratify and/or abide by the terms of a treaty. While several accounts of international law and politics capture various elements of collateral consequences, none views them as a whole.87 In part as a result, they are unable to persuasively account for why states would fail to comply with treaties that they voluntarily joined.88 States’ concerns about collateral consequences of their choices are salient both when they decide whether to commit to a treaty—whether, that is, to voluntarily accept legal limits on their right to act as they wish—and when they decide whether to comply with it. As with legal enforcement, collateral consequences are generated at both the domestic and transnational level. At the domestic level, a government’s decision to accept a voluntary international legal commitment may generate reactions among domestic actors. Where powerful political constituencies in the state have staked out a clear position on issues related to the treaty, the government knows that domestic political support will be affected by the decision to commit to or refrain from committing to the treaty. In fact, governments that favor policies that underlie treaties embodying popular positions on these issues are presumably more likely to achieve office in such states in the first place; hence the central political actors in such states are already more likely to be committed to those principles. Where political pressure favoring a treaty commitment exists, therefore, governments can be expected to be more likely to join the treaty as a consequence. (And the opposite is also true: Where political pressure opposing treaty commitment exists, governments are less likely to join.) This effect is particularly strong in democratic nations, where nongovernmental advocacy groups are better able to influence the government. In this way, the pull of treaties have reason to believe that their co-parties are equally constrained by domestic courts, such that domestic judicial enforcement would not handicap one party significantly more than another.”). See also Anne-Marie Slaughter, A Liberal Theory of International Law, 94 Am Socy Intl L Proc 240, 246 (2000) (arguing that “the global rule of law depends on the domestic rule of law”). 87 See, for example, sources discussed in note 90, infra (discussing reputation). [new cite] 88 As detailed below, more recent accounts discuss aspects of what I collectively refer to here as “collateral consequences.” See Part III. C and D. Between Power and Principle 36 emphasized by normative scholars in the context of treaty compliance can also play an important role in countries’ prior commitment decisions.89 Whether or not the incentives created by domestic collateral consequences push countries toward compliance with those commitments once they are made depends on a variety of factors. These include whether the actors that pushed for commitment sustain their attention to the issue once commitment has been achieved, whether they are able to exercise power over the governmental actors that determine compliance (rather than just over those that made the commitment decision), and whether compliance with the treaty commitments can be relatively easily monitored (which in turn depends in part on the transparency of the treaty requirements). Where domestic actors who care about the issues covered by the treaty are able to place pressure on the government to commit to a treaty, their actions often push the state toward compliance as well. Some have even advocated a strategy of human rights activism that takes advantage of this dynamic. They argue that human rights activists should encourage states to commit to treaties to satisfy domestic constituencies; those constituencies can then use the treaty commitment as leverage for changes in state practices.90 In those instances, however, where a treaty commitment is not followed by continued pressure from domestic interest groups, or if those placing the pressure are not well enough connected to the parts of the government that make compliance decisions, a treaty commitment may be followed by little or no change in state behavior.91 D. Transnational Collateral Consequences Collateral consequences may also come about by shaping the way in which a country is viewed by the international community, which in turn has identifiable consequences.92 A country’s decision to commit or comply sends signals to transnational actors, signals that can 89 See note 33 and accompanying text. 90 See Koh, 35 Houston L Rev at 646–48 (cited in note 36) (discussing the role of norm entrepreneurs in the process of internalization). See also Douglass Cassel, Does International Human Rights Law Make a Difference?, 2 Chi J Intl L 121, 122 (2001) (“International human rights law also facilitates international and transnational processes that reinforce, stimulate, and monitor these domestic dialogues.”); Risse and Sikkink, Socialization of International Human Rights Norms at 5 (cited in note 61) (arguing that transnational advocacy networks “empower and legitimate the claims of domestic opposition groups against norm-violating governments”). 91 See generally Hathaway, 111 Yale L J 1935 (cited in note 17). 92 As George Downs and Michael A. Jones note, international theorists use the related concept of “reputation” to refer to both “(1) the extent to which a state is considered to be an honorable member of the international community and (2) the degree to which a state reliably upholds its international commitments.” George W. Downs and Michael A. Jones, Reputation, Compliance, and International Law, 31 J Legal Stud 95, 96 n 2 (2002). The focus of their work and that of most prior international theorists is on the second definition. Id. By contrast, the focus of attention in this Article is on both definitions, but primarily on the first. Between Power and Principle 37 generate reactions among other states (both those that are members of the treaty regime and those that are not), nongovernmental organizations (NGOs), investors, and transnational organizations such as the International Monetary Fund or World Bank. The reactions of these actors to the signals can, in turn, augment or harm the government’s ability to obtain what it wants in both domestic and international arenas. Transnational collateral consequences may come about through the explicit or implicit linking of foreign aid, trade, or other transnational relationships to the state’s decision to commit to or comply with an international legal rule. For example, states wishing to join the European Union are now required to first join the European Convention on Human Rights before they will be considered.93 Similarly, the World Bank is increasingly taking account of a wide variety of state practices in determining whether to provide loans.94 And the new guidelines for the European Union’s Generalized System of Preferences, which grants trade preferences to 178 developing nations, requires participants to ratify 27 core human rights and environmental treaties by the end of 2008 in order to remain in the program.95 Similar consequences may come from international nongovernmental organizations that track state practices and penalize failures by states to meet their treaty obligations, private investors and companies that might withdraw or withhold funds from the country, and individual countries that might withhold foreign aid, among others. Transnational collateral consequences can create powerful incentives for states to commit to and comply with treaties—incentives that can sometimes lead states to act in ways that would otherwise be deeply perplexing. Such collateral consequences may motivate states to comply with their legal commitments to demonstrate to other states that they will keep their international agreements, even if the agreements turn out to be unfavorable for 93 See generally Hans Christian Kruger, Reflections Concerning Accession of the European Communities to the European Convention on Human Rights, 21 Penn St Intl L Rev 89 (2002). 94 Indeed, the World Bank is now taking countries’ records of “governance” into account in its lending decisions. The Bank measures governance by estimating six factors: voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, and rule of law. The Bank uses eighteen separate sources for these indicators, including several measures of countries’ human rights practices. See Danny Kaufmann, Aart Kraay, and Massimo Mastruzzi, Governance Matters III: Governance Indicators for 1996–2002, World Bank Policy Research Working Paper 3106 at 4–5 (2003), online at http://worldbank.org/wbi/governance/pubs/govmatters3.html (visited Nov 25, 2004). See also Sigrun I. Skogly, The Human Rights Obligations of the World Bank and the International Monetary Fund 108 (Cavendish 2001) (concluding that nothing in the World Bank’s and IMF’s Articles of Agreements prevents those institutions from considering human rights issues in their operations). 95 See Developing countries: facts and figures on the new EU scheme of trade preferences for 2006-2008, available at http://trade- info.cec.eu.int/doclib/cfm/doclib_results.cfm?action=results1 (last visited January 10, 2005). Between Power and Principle 38 them.96 Thus a state may comply with an agreement that it would rather ignore to demonstrate to other international actors that it can be trusted in international negotiations. Yet the reach of this effect may be limited. First, it is limited by the difficulty posed by the task of monitoring compliance; if states believe that noncompliance will go undetected, the state gains little or no reputational benefit by complying. Moreover, as George Downs and Michael Jones have persuasively argued, although a state’s reputation may be affected by a failure to comply with an agreement, the impact is likely limited to agreements that other states have reason to believe “(1) are affected by the same or similar sources of fluctuating compliance costs (or benefits) and (2) are valued the same or less by the defecting states.”97 In short, the consequences of defection from a treaty for a state’s reputation for agreement- keeping may be limited to other agreements in the same area and of the same level of importance. But collateral consequences can also lead states to join treaties with which they will not comply. Under some circumstances, states may receive collateral benefits from committing with treaties with which they do not comply. How and why might this happen? There are three principle reasons. First, the difficulty of obtaining information about state practices may lead transnational actors to rely on ratification as an apparently cost-effective substitute. It is easy to determine whether a state has ratified a treaty; it is much more difficult to evaluate whether it is complying with it. Indeed, collateral benfits are sometimes made explicitly conditional on treaty ratification and not on compliance with the treaty’s terms for precisely this reason.98 NGOs, which are an important source of information about state practices, have limited resources. They may face pressure to allocate those resources where the need is greatest and hence may turn their attention away from a country once it ratifies a treaty, even if there is no evidence that the treaty has been effectively implemented. Second, transnational actors may accept treaty ratification as an indication of a government’s intentions, even if the state’s current practices are not consistent with the treaty. Transnational actors may thus provide collateral benefits such as foreign aid or lower trade barriers to a country that has ratified a human rights or environmental treaty as a reward for its expressed intention to change the course of the state’s public policy. Only several years later does it become apparent whether this intention was matched by deeds, at which time those benefits may be withdrawn. This time lag between the accrual of collateral benefits from ratification and the appearance of collateral costs for noncompliance can create sufficient incentives for states to commit even if it is uncertain whether they will actuall comply. This is particularly true in states with high turnover in government, where the government making the decision to ratify is unlikely to be in power several years hence when the collateral costs for noncompliance arise. 96 See Guzman, 90 Cal L Rev 1823 (cited in note 20); Keohane, After Hegemony at 105–08 (cited in note 20). 97 Downs and Jones, 31 J Legal Stud at 97 (cited in note 92). 98 For example, it appears that eligibility for the Eureopean Union’s Generalized System of Preferences can be maintained by simply ratifying the identified 27 core treaties by the end of 2008. Developing countries: facts and figures on the new EU scheme of trade preferences for 2006-2008, supra note ? (“In any case, the 17 conventions have to be ratified by the beneficiary countries by 31 December 2008”). [new cite] Between Power and Principle 39 Finally, it is possible that some of the transnational actors that exert pressure on states to commit to treaties care little about whether the country actually complies with those commitments. For instance, a company considering whether to make an investment in a country may wish to have evidence that the country is making an effort to improve its environmental or human rights practices (which it can in turn provide to shareholders or interested media), but it may not particularly care whether there are, in fact, real improvements in the country’s practices.99 Moreover, transnational actors may have more difficulty than domestic actors in obtaining access to information about state practices covered by the treaty. Hence, they may be less able to effectively gauge whether states are complying with their treaty commitments. All these collateral consequences—both domestic and transnational—may have particularly important effects on newer regimes and transitional governments. New regimes generally have few or no past practices to point to in their efforts to convince members of the international community that they can and will serve as good international citizens. Many new regimes rise to power during periods of significant civil unrest during which time human rights violations tend to rise,100 economies fall into turmoil, and institutions of governance come under severe stress. Moreover, regime changes frequently come about in countries that are already unstable. Hence new governments may feel a particularly acute need to distance themselves from the practices of past governments and signal to the international community a break from the past by making an open and public commitment to observe central norms of the international community.101 For such regimes, the reputational value of committing to a treaty regime may therefore be quite high. They may be more likely to commit their states to a treaty as a consequence.102 In short, collateral consequences have important implications for compliance with international treaty law. The explicit linking of aid and other benefits to voluntarily joining a 99 On the other hand, investors may genuinely care about the existence of rule of law institutions, to the extent that the presence of such institutions might provide some protection against expropriation or arbitrary regulation. Indeed, Daniel Farber’s work suggests that investors should also care about human rights practices for a similar reason: “[B]y adopting entrenched legal protection for human rights, countries communicate that they are willing to sacrifice short-term advantages to obtain long-term benefits such as economic growth.” Daniel Farber, Rights as Signals, 31 J Legal Stud 83, 98 (2002). 100 See Hathaway, 111 Yale L J at 2040–42 (cited in note 17). 101 See Oona A. Hathaway, The Cost of Commitment, 55 Stan L Rev 1821, 1854 (2003) (arguing that a new regime’s willingness to commit to treaties to signal a break from a past regime may explain why the level of commitment to treaties with relatively worse human rights ratings exceeds expectations). Of course, in some areas, new governments strive to demonstrate continuity with the old. Most notably, new governments frequently agree to honor the international debts accrued under prior governments in an effort to demonstrate creditworthiness of the country. 102 Moravcsik, 54 Intl Org 217 (cited in note 7), makes the related argument that less established democratic nations were the ones that pushed for a binding human rights treaty in Europe because they—unlike more established democracies—saw the sovereignty costs imposed by the treaty as less significant than the benefits of the treaty. Between Power and Principle 40 treaty increases the pressure for commitment. Whether it increases the chances for compliance depends on whether actors actually monitor and respond to violations. The collateral reputational effects of treaties also create incentives for commitment, but again whether compliance follows depends on whether reputation hinges just on joining a treaty or on actually complying with it. And the monitoring and advocacy work of domestic groups affects both commitment and compliance, but particularly the latter. Because domestic groups can make noncompliance more difficult and costly, commitment decisions often hinge on whether domestic political actors supportive of international law exist and are poised to press for treaty adherence. Taking into account these collateral consequences and the incentives they create for treaty commitment and compliance leads to some counterintuitive conclusions. First and foremost, it helps answer a central puzzle of international law: Why would states fail to comply with treaties they voluntarily joined? The answer lies in the fact that the collateral consequences outlined here can lead states to commit to treaties in order to obtain various material and nonmaterial benefits, but those same incentives do not always conduce to compliance. If treaties are not well enforced, countries may commit to them to obtain the collateral benefits of commitment but then fail to live up to their commitments by complying with the treaty terms (particularly if there is also no domestic enforcement of the treaty commitment). In this way, the voluntary nature of international law, coupled with the frequent lack of enforcement, can produce what at first appear to be perplexing results. Second and related, domestic legal enforcement and collateral consequences, which sometimes draw from a similar source but operate through different channels, can create countervailing incentives both for and against treaty commitment.103 If a government recognizes that the forces pressuring it to commit to a treaty will also pressure it to comply with the treaty, it might be more reluctant to commit than it would otherwise be, as the commitment will be more likely to constrain the government’s ability to act freely in the future. Third, collateral consequences can also create incentives for the states that are most likely to comply with their international legal commitments to avoid ever making them in the first place. Governments that already possess strong reputations in an area covered by a treaty may find that they can obtain little additional reputational benefit by ratifying one more treaty. They may also be more sensitive than those with poor practices to any possible 103 The same forces that create strong domestic enforcement incentives for commitment and compliance—namely, active domestic political interest groups that can exert independent pressure on the government—also tend to generate domestic reputational incentives for compliance. Yet while the actors that create the incentives are largely the same for domestic legal and nonlegal incentives, the channels through which they operate are quite different. Domestic legal enforcement entails direct enforcement of existing international legal obligations by domestic actors acting through domestic rule of law institutions—most notably the judiciary. By contrast, collateral consequences operate on governments more indirectly through the reactions of domestic political interest groups and other interested domestic political actors to shifts in governmental reputation: Governments anticipate the political rewards or penalties they expect to receive from interested domestic parties when they decide whether to act in ways that are likely to affect their reputation. Between Power and Principle 41 failure on their part to meet the treaty’s requirements. Thus, if a country already possesses a strong reputation in the area covered by the treaty, it may rightly be concerned that in joining a treaty it stands to gain little and yet may lose a great deal if it is found to have acted in any way inconsistent with the treaty’s requirements. Even though the likelihood that countries with good practices will engage in actions inconsistent with the treaty may be quite small, the costs of such a revelation may be quite high. And not only is the cost of a discovered violation higher, but the chances that, if one exists, it will be found out may be higher as well. Such countries may also be concerned that they will unintentionally violate the vague provisions of the treaty—most leave significant room for interpretation.104 Governments that possess weak reputations, on the other hand, may find that they have more to gain from the reputational boost offered by treaty membership. Such governments may also have less to lose—those that possess a poor reputation have a shorter distance to fall—even though the likelihood that their actions will violate the terms of the treaty is greater. The costs of failure to meet the terms of the treaty (and the likelihood of any shortcomings being discovered) may be small and hence do little to dissuade those with poor practices from committing to treaties. In other words, there may be diminishing returns to reputations. Those who are already high on the curve have little to gain and much to potentially lose from ratifying a treaty, whereas those who are low on the curve may have a great deal more to gain and less to lose from ratifying the same treaty. Hence those with better practices (who are more likely to comply) may, ceteris paribus, be less likely to join the relevant treaties and those with poorer practices (who are less likely to comply) may be more likely to join. It is worth noting in this regard that recent work in “signaling” theory draws upon similar intuitions.105 Signaling models are used to explain situations where actors engage in behaviors the costs of which appear to outweigh the benefits.106 The signaling model works 104 Indeed, for this reason, one might expect such countries also to be more likely to resort to extensive use of reservations, understandings, and declarations when they do join international treaties. See note 73. 105 Signaling theories were originally developed to describe individual or occasionally corporate behavior and have only recently been applied to the realm of state behavior. See Thomas Ginsburg and Richard McAdams, Adjudicating in Anarchy: An Expressive Theory of International Dispute Resolution, 45 Wm & Mary L Rev 1229, 1272 (2004) (applying signaling theories to help explain how third-party expression can influence the behavior of other nations); David H. Moore, A Signaling Theory of Human Rights Compliance, 97 Nw U L Rev 879, 82 (2003) (arguing that signaling theory helps explain why compliance sometimes occurs "when nations face neither effective domestic pressure nor direct pressure from other states to comply"); Farber, 31 J Legal Stud 83 (cited in note 85); Beth A. Simmons, Money and the Law: Why Comply with the Public International Law of Money, 25 Yale J Intl L 323, 361 (2000) (concluding that legal obligations systematically raise and focus actors’ expectations about government behavior). 106 In such models, actors with good characteristics (who, for example, care more about the future—who have, in economic lingo, a “low discount rate”) engage in costly actions as a way of demonstrating their desirable characteristics. See Eric A. Posner, Law and Social Norms 18–22 (Harvard 2000) (applying signaling theory to law); Eric A. Posner, The Between Power and Principle 42 reasonably well to explain areas of international law in which enforcement and monitoring are strong. In such cases, transnational legal enforcement costs of the international law to poor performing states (the “bad types”) are higher than they are for better performing states (the “good types”) and hence the states are less likely to join. But it works less well where enforcement and monitoring are weak—which, as I argued above,107 includes much of international law.108 By contrast, the model offered here describes state action when enforcement and monitoring are weak as well as when they are strong. The Integrated Theory: A Summary The integrated theory incorporates and builds upon the insights of existing theories, placing emphasis on factors which crucially influence state behavior that are often ignored by existing theoretical accounts. Interest-based accounts provide thorough and detailed explanations of the costs and benefits that arise directly from enforcement of international law by transnational actors, what I call the transnational legal enforcement. Yet their focus on these incentives leaves them at a loss when confronted with international legal rules that—like human rights treaties—have little or no binding power. Norm-based accounts, for their part, have a powerful argument about why rules without centralized enforcement mechanisms still carry force. Yet they tend to err in the opposite direction from interest- based scholars, paying little attention to the role of legal enforcement. I, by contrast, argue that such incentives do matter, but that interest-based theories tend to place too much emphasis on them, as they form only one of four categories of incentives that I argue influence state decisions to comply with international legal rules. As noted in Part I, much of the recent work in both the interest-based and norm- based traditions shares a focus on substate dynamics as a source of state behavior in the international system. It is on this shared insight that the approach presented here builds. I seek to move beyond existing work in part by focusing closer attention on the variable role that domestic institutions play in determining whether and to what extent treaty terms are Strategic Basis of Principled Behavior: A Critique of the Incommensurability Thesis, 146 U Pa L Rev 1185, 1194 (1998). To serve as signals, the acts must be ones that potential partners believe only the better types engage in; the acts must have a cost structure such that the bad types cannot engage in them because doing so would be too costly. Id. Signaling theorists acknowledge that bad types sometimes are able to mimic the signals sent by good types. If they can do so successfully, the good and bad types will converge on a similar message, and the message will cease to serve its signaling purpose. Id at 1194–96. 107 See Part II.B. 108 Defenders of signaling theory might respond that commitment to an international treaty should not be regarded as a true signal unless monitoring and enforcement are strong and hence commitment particularly costly to bad types. To say this, however, is to essentially concede defeat, at least with regard to the large body of international law that contains very little monitoring and enforcement. The theory presented here, by contrast, seeks to provide a more comprehensive account of when and how concerns about reputation can play a role in state decisions to commit to international law at all levels of treaty enforcement. Between Power and Principle 43 actually enforced.109 Specifically, I argue that domestic legal enforcement depends on the extent to which domestic institutions allow nongovernmental actors to enforce the state’s international legal commitments against it. Where such institutions are present, states are more likely to change their behavior to comply with international legal requirements but less likely to voluntarily accept them. In states in which such internal institutions cannot be found, the opposite is true: The states are less likely to change their behavior to comply with international legal requirements but more likely to voluntarily accept them. Furthermore, I argue that many scholars of state behavior have paid insufficient attention to what I term the collateral consequences of treaty commitment and compliance. While various scholars have pointed to aspects of collateral consequences—the role of state reputation, for example, has been a topic of much scholarly attention of late—none has taken the broad scope suggested here or worked to fit the collection of such incentives into a broader theory of state action. By contrast, I argue that a government considering whether to commit to a treaty is influenced not only by the costs and benefits that can be traced to the specific terms of a treaty, but also—and sometimes more so—by the collateral effect of treaty commitment and compliance on the government’s ability to achieve its broader aims in both the domestic and international areas. Moreover, I argue that these incentives operate not in isolation but as part of a broader framework that shapes state action. IV. Assessing an Integrated Theory of International Law A theory of state action is only as good as its predictions are accurate. To assess the worth of the theory introduced here against existing theories, it is thus necessary to consider how well it explains what states actually do. By this measure, as will become clear, the integrated theory offers a substantial advance over other current accounts of state behavior. It more accurately predicts and explains several key findings of the available empirical research across a variety of areas of international law. Here, I use new empirical research combined with the results of existing scholarship that uses large-scale quantitative data—some of which draws from my own earlier work—to assess the causal link between legal commitments and state behavior. The new empirical evidence is drawn from a dataset that records the environmental and human rights practices of more than 160 nations over the course of forty years.110 Although this data provides a suitable starting point, much work obviously remains to be done, both in terms of additional large scale quantitative research and more traditional case study methods. Yet the promising performance of the integrated theory against this first glimpse of the real-world empirical evidence—evidence that not only does not consistently support but often undermines 109 Even the republican liberal account assumes that the sovereignty costs of treaties are uniform across states. See Hathaway, Why Do Countries Commit to Human Rights Treaties? (cited in note 27). 110 For more on the treaties examined and data used in this analysis, see Appendices A and B. Between Power and Principle 44 existing theories—suggests that the theory provides a promising starting point for understanding how states behave under international law.111 The integrated theory leads to several unique claims about state behavior under international law. I will focus here on three: * A tradeoff exists between the strength of enforcement of a treaty and the number of states that commit to it. Where transnatinal enforcement of an international treaty is strong, states that are not already in compliance are less likely to commit to it (holding all else equal). Conversely, where transnational legal enforcement is weak, states that are not already in compliance with the terms of a treaty are often as likely to commit to a treaty as those that are already in compliance. * Domestic enforcement is essential to compliance with much of international law. Once they have committed to a treaty, the governments of nations with strong rule of law are likely to be forced to comply with their commitments as a result of enforcement of their agreements through domestic institutions. The same is not true of nations with weak domestic rule of law. * The collateral consequences of treaty membership can sometimes lead states with poor practices to commit to but not comply with a treaty. Membership in a treaty can bring valuable collateral benefits, such as increased foreign aid or cross-border trade. Where compliance is not well monitored and treaty requirements are not enforced, these collateral benefits of membership can lead states to join treaties with which they will not or cannot comply. A. The Tradeoff Between Enforcement and Commitment While enforcement of international law by international actors is not absolutely essential to effective international law, it is far from irrelevant. Where international legal rules are accompanied by sanctions for their violation, there are several predictable results. Perhaps most obvious, where penalties for noncompliance with an international legal rule are significant, states that are not already in compliance are less likely to commit as a consequence. Hence, unless the sanctions are offset by incentives favoring commitment (such as strong reciprocal benefits to membership), treaties containing such sanctions will gain fewer adherents. In other words, there is a tradeoff between enforcement and commitment: Where transnational legal enforcement of a treaty is stronger, fewer countries will commit (holding the benefits of membership constant), but those fewer adherents will be more likely to comply with the terms of the treaty than they would be if the treaty were less strongly enforced. 111 This section is not intended as a complete examination of the existing literature on commitment to and compliance with international law. The focus is instead on large-scale empirical evidence regarding the relationship between international treaties and state behavior in the areas of human rights, the environment, and, to a lesser extent, trade. I am currently at work on much more extensive studies of each of these areas that will provide a more comprehensive review of the relevant literature as well as more detailed empirical evidence than can be offered here. These studies will appear in both article and book-length form. Between Power and Principle 45 If this claim is correct, one would expect that where transnational legal enforcement is minimal, countries will be more likely to commit to treaties; and where they are more significant, countries will be less likely to commit. Evidence that these predictions hold true abounds. To begin with, patterns of commitment across areas of international law provide some evidence of the predicted tradeoff. Human rights treaties and environmental treaties gain adherents at a much faster pace than do comparable trade treaties. For example, the Convention Against Torture and Other Cruel, Inhuman and Degrading Conduct112 (Convention Against Torture) had 134 parties by the end of 2003, the twentieth year since it opened for signature.113 Similarly, the Vienna Convention on the Protection of the Ozone Layer of 1985114 (Vienna Convention) had 187 parties by the advent of its twentieth year.115 By contrast, the General Agreement on Tariffs and Trade (now superseded by the World Trade Organization), which offered substantial benefits to the members in the form of lower trade barriers among trading partners but which imposed significant costs in return, obtained only 69 members within a comparable period.116 Further evidence that the tradeoff between enforcement and commitment exists can be found by looking at the compliance practices of states. As noted above, if a tradeoff exists, one would expect that where transnational legal enforcement is low, countries will readily join treaties with which they are not already in compliance. As I detail below, my evidence shows that states regularly join human rights and environmental treaties (most of which involve little transnational legal enforcement) with which they are not already in compliance. To begin with, states that join environmental treaties often have practices that are far out of line with the requirements those treaties impose. Data on states’ environmental practices show that many states that are members of various environmental treaties have 112 1465 UNTS 85 (cited in note 25). 113 The Convention Against Torture, 1465 INTS 85 (cited in note 25) opened for signature in December 1984. As of March 1, 2004, there were 134 parties. United Nations, Multilateral Treaties Deposited with the Secretary General, online at http://untreaty.un.org/ENGLISH/bible/englishinternetbible/partI/chapterIV/treaty14.asp (visited Nov 25, 2004). 114 Vienna Convention on the Protection of the Ozone Layer, 1513 UNTS 293 (Mar 22, 1985, entered into force Sept 22, 1988). 115 The Vienna Convention, 1513 UNTS 293 (cited in note 114). As of March 1, 2004, there were 187 parties. United Nations, Multilateral Treaties Deposited with the Secretary General, online at http://untreaty.un.org/ENGLISH/bible/englishinternetbible/partI/chapterXXVII/treaty1 1.asp (visited Nov 24, 2004). 116 Sixty-nine states had ratified the General Agreement on Tariffs and Trade (GATT) (which came into existence in 1947 and was superseded by the World Trade Organization on January 1, 1995) by the end of 1966. Calculated from World Trade Organization, The 128 countries that had ratified the GATT by 1994, available at http://www.wto.org/english/thewto_e/gattmem_e.htm (visited Nov 24, 2004). Of course, one related reason for the smaller membership may be the WTO’s more restrictive method of admitting countries into the organization. Between Power and Principle 46 environmental records that conflict with the requirements of those treaties at the time they adopt the treaties.117 For example, countries that have ratified the Vienna Convention, which established mechanisms for international cooperation into the effects of ozone depleting chemicals on the ozone layer, actually produce more CFCs on average than those that have not. The same is true of the 1987 protocol to the Vienna Convention, the Montreal Protocol on Substances That Deplete the Ozone Layer118 (Montreal Protocol). Looking across all countries and years in the dataset, countries that have ratified the Vienna Convention produce, on average, 1.1 percent of total world CFCs. Yet those that have not ratified (this group includes observations from countries that later go on to ratify as well as from those few countries that never ratify) produce on average only .3 percent of the world share of CFCs. The same is true of the Montreal Protocol. Again looking across the entire dataset, countries that have ratified produce on average 1.1 percent of world CFCs, whereas those that have not produce only .2 percent of world CFCs.119 This does not mean, of course, that the treaties are necessarily ineffective. It is possible that these results are due in part to the effort of countries with the worst environmental problems to use international agreements to leverage internal changes that are difficult for them to obtain otherwise. But it does suggest that countries that ratify environmental treaties often do not have better environmental records than those that do not. My research on human rights treaties produces similar findings. To illustrate, I focus here on the Convention Against Torture. The Convention’s adoption by the United Nations in 1984 culminated an effort to outlaw torture that began in the aftermath of atrocities of the Second World War. Nations that ratify the Convention consent not to intentionally inflict “severe pain or suffering, whether physical or mental” on any person to obtain information 117 See World Economic Forum, 2002 Environmental Sustainability Index: An Initiative of the Global Leaders of Tomorrow, Environment Task Force (2002), online at http://www.ciesin.columbia.edu/indicators/esi/ (visited Nov 24, 2004) (providing an overview of the environmental practices of over 140 nations in a wide variety of areas). The data outlined in Appendix B also support this claim, as detailed below. 118 The Montreal Protocol on Substances That Deplete the Ozone Layer, 26 ILM 1550 (Sept 16, 1987, entered into force Jan 1, 1989). 119 Interestingly, researchers looking at firm self-regulatory environmental programs have found similar results among private firms that seek certification for their environmental management systems from the International Organization of Standardization, a private- sector body. For example, Andrew King and Michael Lenox find evidence that firms whose downside risk is minimal and that face the greatest scrutiny are more likely to be early adopters of standards. See Andrew King and Michael Lenox, "Who Adopts Management Standards Early? An Examination of ISO14001 Certifications (unpublished manuscript) (August, 2001) available at http://faculty.fuqua.duke.edu/~mlenox/papers.htm#iso14k. Similarly, Michael Lenox and Jennifer Nash find adverse selection in self-regulatory environmental programs not unlike what I find for environmental treaties. Michael Lenox and Jennifer Nash, "Industry Self-Regulation and Adverse Selection: A Comparison Across Four Trade Association Programs." (Business Strategy & Environment. Forthcoming 2002) http://faculty.fuqua.duke.edu/~mlenox/papers.htm#iso14k. I am grateful to Edward Swaine for pointing out the similarities. Between Power and Principle 47 or a confession or to punish that person or to intimidate or coerce him or a third person.120 Today the Convention is seen by many as a symbol of the triumph of international order over disorder, of human rights over sovereign privilege. Often ignored in the celebrations of the Convention Against Torture by proponents of international law, however, is the fact that while it is quite strong in substance, it is remarkably weak in enforcement. The central enforcement procedure in the treaty is a requirement that state parties submit reports to the Committee Against Torture, an international body created by the treaty.121 But failure to abide by even this minimal commitment is generally ignored.122 Stronger enforcement procedures are optional: Countries can agree to allow states and individuals to file complaints against them with the Committee Against Torture (through a procedure specified in Articles 21 and 22 of the Treaty), but they are not required to do so in order to join the treaty.123 120 Article 1 states: For the purposes of this Convention, the term “torture” means any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. It does not include pain or suffering arising only from, inherent in or incidental to lawful sanctions. Convention Against Torture, art 1, 1465 UNTS 85 (cited in note 23). 121 The Convention Against Torture, art 19, 1465 UNTS 85 (cited in note 23), reads in part: “The States Parties shall submit to the Committee . . . reports on the measures they have taken to give effect to their undertakings under this Convention.” Similar requirements are found in other human rights treaties. For example, Article 40 of the International Covenant on Civil and Political Rights reads, in part: “The States Parties to the present Covenant undertake to submit reports on measures they have adopted which give effect to the rights recognized herein and on the progress made in the enjoyment of those rights.” International Covenant on Civil and Political Rights, 999 UNTS 171, 181 (Dec 16, 1966, entered into force Mar 23, 1976). 122 As of 2000, 71 percent of all state parties to human rights treaties had overdue reports, and 110 states had five or more overdue reports. ANNE F. BAYEFSKY, THE UN HUMAN RIGHTS TREATY SYSTEM: UNIVERSALITY AT THE CROSSROADS 8 (Transnational Publishers 2001). For descriptions and assessments of the intergovernmental human rights enforcement system, see INTERNATIONAL HUMAN RIGHTS IN CONTEXT 592-704 (Henry J. Steiner & Philip Alston eds., 2d ed. 2000); Philip Alston, Final Report on Enhancing the Long-Term Effectiveness of the United Nations Human Rights Treaty System, U.N. ESCOR, 53d Sess., Agenda Item 15, 37, U.N. Doc. E/CN.4/1997/74 (1996). 123 Convention Against Torture, Arts 21, 22, 1465 UNTS 85 (cited in note 23) provide for these additional enforcement mechanisms. Between Power and Principle 48 The Convention against Torture provides a compelling example of the tradeoff between enforcement and commitment. Although over 130 nations have ratified the Convention, the use of torture is unfortunately not limited to the past even among those who have joined. Indeed, countries with the worst torture ratings and countries with the best torture ratings both ratify the Convention at roughly the same rate—about 40 percent. I find similar results for a wide array of human rights treaties.124 While the states that join human rights treaties usually have better practices than those that do not, the difference between the two groups of nations is much smaller than many would expect.125 Moreover, states with poor human rights practices regularly join human rights treaties, sometimes at a rate similar to that of countries with the very best human rights practices.126 For example, roughly half of countries that have reportedly committed no genocide have ratified the Convention on the Prevention and Punishment of the Crime of Genocide;127 the same is true of countries with the very worst genocide ratings.128 Indeed, holding political and economic factors constant, states with good human rights records are no more likely to commit to human rights treaties than those with poorer records.129 Among nondemocratic nations (which tend to have weaker domestic rule of law institutions),130 the pattern is even 124 See Hathaway, 111 Yale L J at 1935–2042 (cited in note 17); Hathaway, 55 Stan L Rev at 1854 (cited in note 101); Hathaway, Why Do Countries Join Human Rights Treaties? (cited in note 27) (looking at enforcement of human right treaties for the period between 1960 and 1999). See also Linda Camp Keith, The United Nations International Covenant on Civil and Political Rights: Does It Make a Difference in Human Rights Behavior?, 36 J Peace Res 95, 112 (1999) (suggesting that the ICCPR’s “implementation mechanisms are too weak and rely too much upon the goodwill of the party state to effect observable change in actual human rights behavior”). 125 See Hathaway, 111 Yale L J at 1976–88 (cited in note 17); Keith, 36 J Peace Res at 104 (cited in note 124) (showing that the practices of states that are not party to the International Covenant on Civil and Political Rights do not differ substantially from the practices of states that are parties). 126 See Hathaway, 111 Yale L J at 1982–87 (cited in note 17). 127 Convention on the Prevention and Punishment of the Crime of Genocide, adopted Dec 9, 1948, 5 Exec Doc O, 78 UNTS 277 (entered into force Jan 12, 1951). 128 See Hathaway, 111 Yale L J at 1982. 129 See Hathaway, Why Do Countries Commit to Human Rights Treaties? (cited in note 27). 130 The interrelationship between democracy and rule of law has been widely recognized and discussed. In the empirical portions of this paper, I use democracy rather than a direct measure of rule of law. I do so for a few reasons. First, by referring to nations with strong domestic rule of law, I mean nations where an independent legal system serves as a check on state power. The primary empirical measure of “rule of law” used by researchers is produced by the International Country Risk Guide. See www.icrgonline.com. This variable includes both “law”—“an assessment of the strength and impartiality of the legal system”—and “Order”—“an assessment of popular observance of the law.” Id. By contrast, the best available measure of democracy—from the Polity Project—is composed of three interdependent elements: “the presence of institutions and procedures through which Between Power and Principle 49 more striking: Nondemocratic nations with worse reported human rights practices appear more likely to have ratified human rights treaties than those with better reported practices.131 Only in the area of trade—where transnational incentives tend to be higher—does it appear that noncomplying states that are more able to comply with treaty requirements are more likely to commit to a treaty as a consequence. A study by Beth Simmons of certain rules governing financial policies of national governments in the International Monetary Fund’s Articles of Agreement shows that countries with economic indicators that suggest they will find it difficult to comply with the obligations are somewhat less likely to commit to them.132 Thus the empirical evidence supports the integrated theory’s prediction that where transnational legal enforcement is weak, states will be more likely to commit to and less likely to comply with treaties. This appears to be the case in areas—such as human rights and the environment—where transnational legal enforcement is weak. By contrast, where citizens can express effective preferences about alternative policies and leaders”; “the existence of institutionalized constraints on the exercise of power by the executive;” and “the guarantee of civil liberties to all citizens in their daily lives and in acts of political participation.” Monty G. Marshall and Keith Jaggers, Polity IV Project: Political Regime Characteristics and Transitions, 1800–2002, online at http://www.bsos.umd.edu/cidcm/inscr/polity/index.htm (visited Nov 24, 2004) (including a description of variables and a link to the data set). This better reflects the institutional characteristics that are important to my claims, though it admittedly is also not a perfect measure. Interestingly, the two measures are highly correlated. See, for example, Roberto Rigobon & Dani Rodrik, Rule of Law, Democracy, Openness, and Income: Estimating the Interrelationships (NBER Working Paper) (September 2004) (finding that rule of law and democracy tend to be mutually reinforcing). 131 See Hathaway, 55 Stan L Rev at 1854 (cited in note 101). Conversely, democratic nations with worse human rights practices appear less likely to commit to human rights treaties, holding other factors constant. See Hathaway, Why Do Countries Commit to Human Rights Treaties? (cited in note 27). 132 See Simmons, 94 Am Polit Sci Rev at 822, 825–26 (cited in note 7) (noting that the economic controls she includes to test the proposition that states that are more likely to be able to comply will be more likely to commit “basically fulfill expectations, although most fall short of traditional standards of statistical significance”). James Vreeland also conducts empirical studies of the International Monetary Fund, though his primary focus is on the impact of IMF programs on economic growth and on the factors that lead states to enter into IMF agreements. See James Raymond Vreeland, The IMF and Economic Development (Cambridge 2003); Adam Przeworski and James Raymond Vreeland, The Effect of IMF Programs on Economic Growth, 62 J Dev Econ 385, 403 (2000) (suggesting that governments adopt IMF programs when facing foreign reserve crises and when they need to shield themselves from the political costs of fixing budget deficits); Vreeland, The Institutional Determinants of IMF Programs at 1 (cited in note 7) (arguing that “governments are more likely to enter IMF arrangements when there are more veto players in the political system because reform-minded governments often use the leverage of the IMF to push through unpopular policies”). Between Power and Principle 50 transnational legal enforcement is strong, as it often is in the area of trade, the evidence suggests that states are less likely to join if they are not already in compliance with the treaty’s requirements and are more likely to change their actions to comply with the treaty after they have joined. B. Domestic Enforcement of International Law Is Essential to Compliance The integrated theory presented here makes clear that strong domestic institutions are essential not only to domestic rule of law, but also to international rule of law. Where international bodies are less active in enforcement of treaty commitments—as in the areas of human rights and the environment—it falls to domestic institutions to fill the gap. In some states, this reliance on domestic institutions is effective. In others it is less so. In democratic nations, where domestic rule of law and hence enforcement tends to be relatively strong (because the judiciary, media, and political parties are free to operate independent of the executive), states are more likely to abide by international law whether it is externally enforced or not. In less democratic nations, where domestic enforcement can be less effective, states are less likely to abide by international law that is not enforced by transnational bodies. Again, the Convention against Torture provides a helfpul illustration of the argument. As noted above, the Convention enjoys only very weak transnational legal enforcement. The integrated theory highlights the fact that, in such situations of weak transnational legal enforcement, domestic enforcement mechanisms take on added importance as a constraint against treaty noncompliance. Given the interaction it posits between compliance and commitment, the integrated theory thus predicts that more democratic nations, which are more likely to engage in domestic enforcement, are less likely to commit to the Convention if their practices are inconsistent with its requirements than if they are not. The same is not true of nondemocratic nations.133 As outlined in Table 2 below, I find that the predictions of the integrated theory, counterintuitive as some of them are, prove accurate.134 The Table shows the percentage of countries at each level of reported torture practices that had ratified or signed the Convention Against Torture, or agreed to the enforcement provisions outlined in Articles 21 and 22 to the Convention. It demonstrates that democratic nations are more likely, at each level of reported practices, to join the Convention Against Torture.135 For example, 69% of 133 Indeed, when there is little or no transnational legal enforcement, nondemocratic nations (which tend to have weker indepenent rule of law institutions and hence less domesic enforcement of treaties) may be even more likely to commit to treaties when they have worse pracrices because they can obtain collateral benefits from doing so. See supra Part III. C and D. 134 The data used in Table 2, and in the rest of this Article, are described in more detail in Appendix B. Unless otherwise indicated, all calculations are by the author using the data described in this Appendix. 135 Democratic nations are also less likely to torture than are nondemocratic nations. Countries that reportedly torture the least (have a torture rating of “1”) have an average democracy rating of 7.59; countries that reportedly torture the most (have a torture rating of Between Power and Principle 51 democratic nations that have engaged in no torture ratified the Convention, whereas only 6% of nondemocratic nations with similarly excellent records did so. But this gap falls as contries’ practices worsen. Indeed, while democratic nations with worse torture records are somewhat less likely to ratify the Convention than those with the better records, nondemocratic nations with worse records are more likely to ratify the Convention than nondemocratic nations with better records. This is true across the continuum of torture ratings. At almost every level, nondemocracies with worse reported torture practices are more likely to commit to the Convention Against Torture than those with better reported practices. By contrast, democracies with progressively worse torture ratings are often less likely to make legal commitments that prohibit them from engaging in torture, particularly to the stronger enforcement mechanisms provided for in Articles 21 and 22.136 TABLE 2: DEMOCRACIES’ AND NONDEMOCRACIES’ COMMITMENT TO THE CONVENTION AGAINST TORTURE, 1985–1999. Torture Rating: 1 (no torture) 2 3 4 5 (torture is “prevalent” or “widespread”) Ratified Convention: 6% (78) 18% (248) 31% (494) 39% (250) 37% (120) Signed Convention: 9% (78) 25% (248) 40% (494) 52% (250) 53% (120) Nondemocracy Articles 21 and 22: 1% (78) 1% (248) 6% (494) 5% (250) 12% (120) Ratified Convention: 56% (285) 66% (319) 51% (229) 42% (119) 57% (49) Signed Convention: 81% (285) 79% (319) 67% (229) 54% (119) 67% (49) Democracy Articles 21 and 22: 46% (285) 34% (319) 19% (229) 8% (119) 4% (49) Note: The number of total observations appears in parentheses. A similar dynamic appears to be at work in the area of the environment. I focus here in particular on the Vienna Convention and the subsequent Montreal Protocol, which together require parties to the treaty to freeze and eventually phase down their use of “5”) have an average democracy rating of 2.42. For more on the sources of data on torture and democracy, see Appendix B and Hathaway, 111 Yale L J at 1969–72, 2029–30 (cited in note 17). 136 See note 123 and accompanying text. Between Power and Principle 52 specific chemicals and which are often hailed as the most successful international agreements in modern times. I also include in the analysis the Copenhagen Amendment to the Montreal Protocol, which steps up the phase-out schedule for ozone depleting chemicals for countries that separately ratify it. As with the Convention Against Torture, the transnational legal enforcement mechanisms for the agreements are fairly weak, again relying primarily on reporting requirements. The integrated theory predicts that democratic states, which tend to have stronger rule of law and hence better domestic enforcement of international law, will be less likely to commit to the environmental agreements if their practices are inconsistent with the agreements’ requirements than if their practices are already consistent with their requirements. And once they commit, they are substantially more likely, because of those same domestic institutions, to actually comply with the agreements’ requirements than are nondemocratic nations that have committed. The predictions of the integrated theory are once again borne out by the evidence. Table 3 compares the pattern of commitment to the Vienna Convention, Montreal Protocol, and Copenhagen Amendment for nondemocracies and democracies as their share of world consumption of the most common ozone depleting chemicals, chlorofluorocarbons (CFCs), increases. Table 3 shows that democracies are more likely to commit to the treaties at all levels of CFC consumption. For example, of nondemocracies with the lowest share of CFC consumption, 26 percent ratified the Vienna Convention, compared to 77 percent of democracies with a comparable share of CFC consumption. The high level of commitment to the treaty among democratic nations is fairly consistent across all levels of CFC production (perhaps reflecting consistently strong positive domestic collateral incentives for commitment). By contrast, nondemocratic nations with higher world shares of CFC consumption (that is, those that pollute more) are not less likely to ratify, but are actually much more likely to ratify than those with smaller world shares of CFC consumption (those that pollute less). For example, 26 percent of nondemocratic nations with the lowest share of CFC consumption ratified the Vienna Convention, compared with 79 percent of nondemocratic nations with the highest share of CFC consumption. Between Power and Principle 53 TABLE 3: DEMOCRACIES’ AND NONDEMOCRACIES’ RATE OF COMMITMENT TO THE VIENNA CONVENTION, MONTREAL PROTOCOL, & COPENHAGEN AMENDMENT World Share of CFC Consumption: 1 (lowest share of Annex A CFCs) 2 3 4 (highest share of Annex A CFCs) Ratified Vienna Convention: 26% (235) 59% (222) 64% (211) 79% (154) Signed Vienna Convention: 28% (235) 62% (222) 68% (211) 79% (154) Ratified Montreal Protocol: 28% (202) 62% (207) 67% (201) 79% (145) Signed Montreal Protocol: 32% (202) 69% (207) 71% (201) 83% (145) Nondemocracy Ratified Copenhagen Amendment: 0 % (68) 9% (128) 26% (140) 39% (103) Ratified Vienna Convention: 77% (217) 70% (201) 80% (205) 83% (285) Signed Vienna Convention: 78% (217) 71% (201) 83% (205) 88% (285) Ratified Montreal Protocol: 80% (208) 72% (191) 79% (197) 83% (266) Signed Montreal Protocol: 82% (208) 72% (191) 82% (197) 91% (266) Democracy Ratified Copenhagen Amendment: 52% (161) 31% (133) 40% (140) 44% (165) Note: The number of total observations appears in parentheses. Table 4, which examines percentage reductions in CFC consumption of democracies and nondemocracies according to whether or not they have ratified the Convention and Protocol, is even more revealing. It demonstrates that nondemocracies that ratified the Vienna Convention and Montreal Protocols by 1989 (the year the Montreal Protocol went Between Power and Principle 54 into effect) expanded their consumption of CFCs more in the three prior years than those that did not. (This is especially revealing because the Protocol and the Amendments to it require percentage reductions in CFCs from a base year of 1986.) Yet the opposite is true of democracies: Democracies that ratified the Convention and Protocol by 1989 reduced their CFC consumption by more than 20 percent in the three years before, whereas democracies that did not ratify expanded their CFC consumption by more than 13 percent. This supports the integrated theory’s claim that democracies act differently in their decisions to commit to treaties than do nondemocracies: They have stronger domestic enforcement of international law and—all else held equal—they are therefore less likely to commit to treaties with which they anticipate they will be unable or unwilling to comply and they are more likely to comply once they do commit. TABLE 4: PERCENT CHANGE IN CFC CONSUMPTION, 1986–1989, BY REGIME TYPE AND TREATY RATIFICATION Countries that did not ratify by 1989 Countries that ratified by 1989 Vienna Convention 4.7% (36) 9.1 % (19) Nondemocracy Montreal Protocol 8.0 % (38) 8.9 % (17) Vienna Convention 13.9% (20) -20.7 % (24) Democracy Montreal Protocol 13.9 % (22) -20.7 % (22) Note: Percetages are the median percentage change in CFCs contained in Annex A of the Montreal Protocol. The number of total observations appears in parentheses. The Copenhagen Amendment is not included here because it did not enter into effect until 1994. These results are supported by other existing empirical evidence.137 Perhaps the most robust finding in the empirical literature to date is that democratic nations behave differently with regard to international law than do nondemocratic nations. Andrew Moravcsik, who examines the European Convention on Human Rights, argues that international treaties offer states a means of consolidating democratic achievements, “thereby enhancing their 137 See Hathaway, 111 Yale L J 1935 (cited in note 17); Hathaway, Why Do Countries Commit to Human Rights Treaties? (cited in note 27). Between Power and Principle 55 credibility and stability vis-à-vis nondemocratic political threats.”138 He shows that in Europe, potentially unstable democracies—countries in which democracy is established but nondemocratic groups threaten its future—were the strongest advocates of the European Convention, against the opposition of established democratic nations and dictatorships. Those who opposed the Convention did so, he argues, because the sovereignty costs that would be imposed by the binding human rights treaty far outweighed the marginal benefits of enhanced political stability that the treaty could be expected to bring.139 My earlier empirical work also confirms these findings. In particular, my examination of several multilateral and regional human rights treaties explores the role of democracy beyond the European Context—and, in doing so, both reinforces Moravcsik’s argument that democracy matters and challenges the specific spin he puts on the argument. I find that democracies are on the whole more likely to join international human rights treaties than nondemocracies.140 However, they are less likely to join human rights treaties if they have worse human rights practices than if they have better practices (by contrast, nondemocracies are more likely to join if they have worse practices than they are if they have better practices).141 Moreover, only the most democratic states appear to improve their practices after ratifying human rights treaties.142 In sum, the empirical evidence strongly supports the integrated theory’s prediction that where transnational legal enforcement is weak, how states will respond to international law depends in large part on the domestic legal enforcement mechanisms that are in place. As predicted, the evidence shows that democratic nations that engage in actions prohibited by a treaty are less likely to ratify that treaty than democratic nations with better practices. This pattern among democracies lies in stark contrast to the pattern among nondemocratic nations, which actually appear to be substantially more likely to ratify treaties if they have worse records than they are if they have better records. C. The Power of Collateral Consequences Last but not least, the integrated theory highlights the role that collateral consequences play in state decisions to participate in and comply with international law. Several predictions arise out of this new focus, one of which I will emphasize here: Membership in certain treaties can bring collateral benefits that states value. For example, membership may boost a state’s reputation. Sometimes that boost to reputation occurs regardless of whether the member state actually abides by the treaty’s requirements. This is particularly true where the treaty is not well enforced and compliance is not effectively monitored, leaving member states that do not police themselves to face little risk of exposure if they fail to abide by the treaty’s requirements. As a consequence, states that 138 Moravcsik, 54 Intl Org at 220 (cited in note 7). 139 Id at 228–29. 140 See Hathaway, 111 Yale L J 1935 (cited in note 17); Hathaway, Why Do Countries Commit to Human Rights Treaties? (cited in note 27). 141 See Hathaway, 55 Stan L Rev 1821 (cited in note 101); Hathaway, Why Do Countries Commit to Human Rights Treaties? (cited in note 27). 142 See Hathaway, 111 Yale L J 1935 (cited in note 17). Between Power and Principle 56 engage in violations and have weak domestic rule of law institutions have every reason to join treaties that confer reputational benefits, such as human rights and environmental treaties. States that join treaties primarily to obtain such benefits can be expected to not always comply with treaties to which they have committed and sometimes have even worse practices subsequent to commitment. There is significant evidence that states do not act as many would expect after they commit to treaties. Once again, data from state practices under the Vienna Convention and Montreal Protocol provide interesting insights into state practices. If we examine state practices in 1993, the year before the first phase down in CFCs in the Montreal Protocol, we find an interesting pattern. First, all of the democratic countries in the dataset have joined the Montreal Protocol by this date. Second, in this year at least, the rate of reduction in CFC production is greater for countries that have not yet ratified the Vienna Convention or Montreal Protocol than it is for those that have. As reflected in the first set of columns in Table 5, countries that had not ratified the Vienna Convention by 1993 saw their CFC production fall 20.1 percent from the prior year, compared to a reduction of less than 1 percent among those that had ratified. Similarly, among those that had not ratified the Montreal Protocol by 1993, CFCs fell 10 percent from the prior year, compared to 1.7 percent among those that had ratified. Finally, as also outlined in Table 5 below, this pattern is driven largely by the fact that nondemocracies that have ratified by 1993 actually marginally increased their CFC production from the prior year while those that had not ratified by 1993 reduced their CFC production.143 While we cannot read too much into single-year changes in production (particularly given the very small numbers of nonratifiers at this late date), these numbers do suggest that treaty ratification does not always have the expected or intended effect.144 143 Indeed, democracies that acceded to the Copenhagen Amendment show the opposite relationship, with a greater than fifty percent decrease in CFC production from the prior year (see Table 5). This is further evidence of the importance of domestic enforcement noted above in Part IV. B.[check, please]. 144 Existing empirical studies provide further support for the claim that states do not always change their practices in the ways that the treaty’s advocates intend. Consider a controversial study in the Journal of Public Economics by James Murdoch and Todd Sandler. James C. Murdoch and Todd Sandler, The Voluntary Provision of a Pure Public Good: The Case of Reduced CFC Emissions and the Montreal Protocol, 63 J Pub Econ 331 (1997). It claims to find that the Montreal Protocol had virtually no independent impact on countries’ use of ozone-depleting gasses. Murdoch and Sandler argue that the treaty did not change states’ behavior but instead merely codified an existing trend of voluntary cutbacks in emissions. Between Power and Principle 57 TABLE 5: CHANGE IN CFC PRODUCTION, 1992–1993, UNDER THE VIENNA CONVENTION, MONTREAL PROTOCOL, AND COPENHAGEN AMENDMENT Countries that did not ratify by 1993* Countries that ratified by 1993* Vienna Convention -20.1 (3) -.9 (88) Montreal Protocol -10.0 (4) -1.7 (87) All Countries Copenhagen Amendment -.2% (79) -34.9% (23) Vienna Convention -20.1 % (3) 2.7 % (30) Montreal Protocol -10.0 % (4) 2.4 % (29) Nonemocracy Copenhagen Amendment 0.11 % (37) 20.2 % (6) Vienna Convention (0) -11.1 % (58) Montreal Protocol (0) -11.1 % (58) Democracy Copenhagen Amendment -3.3 % (42) -50.4 % (17) *For the Copenhagen Amendment, the dates are one year later, as the Amendment opened for signature on November 1992 and went into effect in June 1994. Note: The number of observations appears in parentheses. Note: Calculated as median percentage change in CFCs contained in Annex A of the Montreal Protocol. Again, my research on human rights treaties produces similar findings.145 Not only do states not appear to improve their practices after ratifying human rights treaties,146 but 145 See Hathaway, 111 Yale L J 1935 (cited in note 17); Hathaway, 55 Stan L Rev at 1821–62 (cited in note 101); Hathaway, Why Do Countries Join Human Rights Treaties? (cited in note 27); Keith, 36 J Peace Res 95 (cited in note 124). 146 See Keith, 36 J Peace Res 95 (cited in note 124) (finding that becoming a party to the International Covenant on Civil and Political Rights does not appear to make a difference in human rights behavior); Hathaway, 111 Yale L J 1935 (cited in note 17). Between Power and Principle 58 some evidence suggests that some of the countries that joined human rights treaties may have worse practices than would be expected had they not joined. Only Simmons’ study of the International Monetary Fund provides more hopeful results.147 She concludes that at least some of the IMF rules are effective. Most notably, she finds that a direct declaration of adherence to Article VIII (which requires states to avoid restrictions on current payments and discriminatory currency practices) consistently has a negative effect on the probability that states will impose prohibited restrictions on their current accounts.148 Yet Simmons also finds that a large number of states fail to comply with their legal (and voluntary) commitment to keep their current account free from restrictions and to maintain unified exchange rates.149 This is again consistent with the integrated theory’s predictions. The IMF rules, unlike the human rights and environmental treaties, carry more significant transnational legal enforcement incentives. Hence the integrated theory would predict that the rules would be more likely to be effective for all parties—democratic and nondemocratic alike—than are treaties with weaker transnational legal enforcement. The new empirical evidence presented here as well as existing studies therefore provide compelling support for the integrated theory’s predictions. The theory predicts that states that have better human rights or environmental records (and better reputations) are not more—and are sometimes even less—likely to join human rights and environmental treaties than states that have worse records (and worse reputations). The empirical evidence shows this to be true, and particularly true among nondemocracies, which do not face the countervailing pressure of internal enforcement discussed above. The theory predicts that countries will use treaties as a substitute for real action when monitoring and enforcement are weak and collateral incentives are great. The empirical evidence shows that this is the case—that the calculated risk that states with poor records (and reputations) take in joining treaties with which they don’t intend to comply may sometimes pay off. Finally, the theory predicts that treaties with stronger transnational legal enforcement will be more effective as a consequence (because states cannot obtain the collateral benefits of membership without accepting the cost of transnational legal enforcement). The empirical evidence on the effectiveness of the IMF rules shows this to be true. The Empirical Evidence: A Summary The claims that arise from the integrated theory find solid support in the real world. First, the evidence shows that where transnational legal enforcement is minimal, how states will respond to international law depends in large part on the domestic legal enforcement, just as the integrated theory predicts. Second, the evidence supports the prediction that where transnational legal enforcement is stronger (as in the area of trade), states are less likely to commit to and more likely to comply with treaties than they are when transnational legal enforcement is weaker (as in the areas of human rights and the environment). Finally, the integrated theory’s counterintuitive predictions regarding the role of collateral consequences also find compelling support. States that have better human rights and 147 See generally Simmons, 94 Am Polit Sci Rev 819 (cited in note 7). 148 Id at 830–31. 149 Id at 827. Between Power and Principle 59 environmental records are not more—and are sometimes even less—likely to join human rights and environmental treaties than states that have worse records. This is particularly true among nondemocracies, again just as the theory predicts. By contrast, the traditional accounts of international law are at a loss to explain most of the empirical results outlined above. States do not only agree to join treaties that require them to do what they are already doing or already intend to do, as interest-based theories contend.150 They actually join treaties that commit them to do something more. Moreover, interest-focused scholars who believe that international law is meaningless cannot help us understand how or why states that ratify certain human rights treaties do not simply have practices that are no different from those that have not ratified, but might sometimes have practices that are worse. And while they help us understand some of the differences in the responses of democratic and nondemocratic nations to international law, they do not help us understand differences among democratic nations, much less among nondemocratic nations.151 Norm-centered theories of international law are also at a loss to explain many of the empirical results. States with poor human rights and environmental records commit so readily to human rights and environmental treaties that it would not be unreasonable to conclude that they do so only because they do not take the commitment all that seriously. More troubling for such advocates of international law, however, is the evidence suggesting that countries that ratify human rights and environmental treaties do not engage in fewer violations of the terms of those treaties as a result. And normative theories are no better able to account for the systematic differences among democratic and nondemocratic nations in their willingness to commit to and comply with international law. V. Lessons for the Future The integrated theory of international law provides an outline of how international treaties shape state behavior. It incorporates and moves beyond both interest-based and norm-based accounts, while highlighting factors that neither account fully captures. And it provides novel empirical predictions that are, without exception, borne out by the existing, albeit sparse, empirical evidence. What, however, are its implications for those who wish to use international law to shape state behavior? Let me close this Article by considering three lessons in particular that flow from the model: * To improve compliance with international law, efforts should be made to mitigate the tradeoff between enforcement of and commitment to international treaties. * Effective domestic enforcement of international legal commitments is essential to their success. International legal compliance can therefore be improved by strengthening domestic rule of law institutions. 150 Except, perhaps, in the area of trade, where treaty commitments are more visible and enforceable. 151 Andrew Moravcsik’s account helps explain why democracies would be more likely to commit to treaties at higher rates, but it is unable to account for the other results regarding differences between democracies and nondemocracies. See generally Moravcsik, 54 Intl Org 217 (cited in note 7). Between Power and Principle 60 * International law can and should take better advantage of states’ regard for collateral consequences to foster behavior that is consistent with international law. A. Mitigate the Tradeoff Between Enforcement and Commitment Where international legal rules are enforced by transnational sanctions, the integrated theory predicts and the empirical evidence shows that states that are not already in compliance are less likely to commit as a consequence. What does this tradeoff between enforcement and commitment mean for international law? It suggests that where international institutions do not put in place effective enforcement mechanisms, there is of necessity greater reliance on other methods of maintaining compliance. Yet we must remember that these other methods do not, as we have seen, always have the intended effects. In particular, the reliance on domestic enforcement to fill the gap left by weak international enforcement can produce a regime that is shunned by precisely those states who would be the best members and the most likely to change their behavior as a consequence of joining. This creates a tradeoff between enforcement of a treaty widespread commitment to it. It also means that treaty membership is least likely to be found where it is most likely to be effective in changing behavior. This tradeoff is not, however, the same everywhere and always. It can be made less severe. There are two ways that the international community could more effectively mediate the conflict between commitment and compliance. The first is to move states incrementally down the path toward stronger international rules with true enforcement provisions. Rather than confront states immediately with a legal regime that couples challenging goals with strong sanctions for failure to meet them, states can be gradually led toward stronger legal rules. This can be accomplished by starting with relatively weak international rules backed by little or no sanctions that all states feel comfortable joining, but then gradually pushing states to accept successively stronger and more challenging requirements. The danger of this approach, however, is that it can stall at any point in the cycle. The creation of weak international rules may frequently serve to offset pressure for stronger rules that would be more effective. Hence this incrementalist strategy must be embarked upon with caution. In fact, if incrementalism is to be successful, it may be necessary to require participants in the regime to make successive steps toward stronger and more enforceable rules. A single treaty that has tiered levels of membership and allows states set periods of time to move from one level to another will likely be more successful at producing positive change than the current strategy of layering successively more challenging treaties upon existing less-challenging treaties without requiring states to move from one to the other. Alternatively, a treaty regime that expels countries that fail to comply with minimum requirements within set periods of time may accomplish the same goal. A second method for overcoming the inherent conflict between commitment and compliance is to find ways to make membership in challenging regimes more advantageous to states. Thus treaties might include benefits that help offset some of the costs that they impose. (Trade treaties already do this: They offer lower barriers to a state’s exports in return for its acceptance of lower barriers to imports.) Similarly, the UN or other international bodies might seek to provide states with assistance that can make them more willing to Between Power and Principle 61 commit and comply. This might include, for example, financial incentives for membership or technical assistance to aid states in bringing their institutions into compliance.152 B. Strengthen Domestic Rule of Law Institutions The integrated theory emphasizes the connection between strong domestic institutions and international rule of law. Understanding this connection between the domestic and international arenas can profoundly affect the way in which we view international law. To begin with, it helps us better understand and explain the intuitions of the liberal strand of international relations and international law literature: Democratic nations are more likely to follow the law for the precise and predictable reason that their internal institutions give their government leaders less leeway to do otherwise. Moreover, the difference between democratic and nondemocratic nations in their willingness to follow international law also follows a predictable pattern: Where transnational legal enforcement is stronger, the willingness of the two sets of states to commit to and comply with international law will differ less, and where transnational legal enforcement is weaker, it will differ more. The focus on this connection between domestic institutions and international law also provides insight into the internalization process that is the centerpiece of many norm- based theories of international law. Internalization does not occur uniformly across nations and across areas of international law. Rather, internalization will occur most rapidly for laws that are enforced by transnational actors and in states that possess domestic institutions that allow individuals and groups to enforce international legal commitments against the government. The lesson that comes from these insights is obvious. Because so much of international law relies so heavily on domestic rule of law institutions, strengthening those institutions could have a profound impact on compliance with international law. This is especially true in areas of international law where transnational legal enforcement is weak. Of course, strengthening rule of law institutions may lead states to be less willing to commit to treaties in the first place, as states with stronger rule of law institutions will face more robust domestic enforcement of their treaty commitments. Yet, while this is undoubtedly a concern, it is mitigated in part by the fact that the countries dissuaded from committing will be those who had previously joined with little or no intention of complying. Moreover, evidence from the human rights arena indicates that, as rule of law institutions gain strength, there can be countervailing pressure on the government to participate more heavily in the international legal arena.153 152 There has been some attempt to do something like this with the United States’ new Millennium Challenge Account. Under this new program, countries will be rated according to various performance indicators and money provided to those that are found to be “ruling justly, investing in their people, and encouraging economic freedom.” President George Bush, Remarks at the Inter-American Development Bank, Washington, D.C. (March 14, 2002), online at http://www.whitehouse.gov/news/releases/2002/03/20020314-7.html (visited Nov 24, 2004). 153 Strengthening the rule of law is a deeply complicated enterprise fraught with controversy and pitfalls. For a careful and critical look at the challenges posed to efforts to Between Power and Principle 62 While few would disagree with the goal of improving domestic rule of law institutions, there is a great deal of disagreement as to how best to achieve that goal. A complete answer to that question is well beyond the scope of this Article. I will only pause here to note that this Article should not be mistaken as arguing in favor of forcible democratization. Although I argue that democratization and the likely attendant improvement in domestic rule of law will lead to better compliance with international law, this is only one of many considerations that must be weighed in contemplating whether and how to pursue democratization. Not least among the other considerations is the likelihood of success of forcible regime change. Indeed, in my own view, forcible democratization is not nearly as likely to be successful in achieving the ultimate end of a stronger rule of law as is a more incremental approach that focuses on gradually building domestic institutional capacity. C. Harness the Power of Collateral Consequences. Last but not least, the integrated theory highlights the role that collateral consequences play in state decisions to participate in and comply with international law. Countries’ concerns for their reputations and for aid, trade, and other benefits that are sometimes linked to treaty commitment and compliance can be used more effectively than they currently are to strengthen the influence of international law. At present, membership in certain treaties can confer a boost to a state’s reputation, often regardless of whether the member state actually abides by the treaty’s requirements. This is possible because the international community does little to police many treaty requirements, leaving member states that do not police themselves to face little risk of exposure if they fail to abide by their requirements.154 As a consequence, states that engage in violations and have weak domestic rule of law institutions have every reason to join treaties that confer reputational benefits, such as human rights and environmental treaties. How might we address this perverse incentive without radically altering these international legal regimes? The integrated theory suggests that besides strengthening internal rule of law, simply monitoring the activities of treaty members more effectively could substantially improve the situation.155 For example, if states’ violations of the terms of engender the rule of law, see Rosa Ehrenreich Brooks, The New Imperialism: Violence, Norms, and the “Rule of Law”, 101 Mich L Rev 2275, 2339 (2003). 154 Nongovernmental organizations, such as Amnesty International, Human Rights Watch, EarthFirst, and the Sierra Club, do serve to monitor state behavior. But NGO resources are often spread thin, making comprehensive monitoring impossible. Moreover, to the extent that NGOs are often associated with particular political positions, the information they provide sometimes carries less credibility than would similar information from an independent international institution. NGOs also frequently do not have the institutional access that would be necessary to provide true monitoring of state practices. 155 In implementing this proposal, it would be important to remain cognizant of the tradeoff between commitment and enforcement, as discussed in Part IV. A. above. Compare Kal Raustiala, Form and Substance in International Agreements (unpublished manuscript) (2004), online at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=505842) (arguing that Between Power and Principle 63 the Convention Against Torture were likely to be investigated and made public by an independent international organization, states that do not intend to abide by the treaty would be substantially less likely to join while those that do intend to abide by its requirements would be no less likely to commit to it. At the very least, those designing treaties should consider the powerful signal that membership in a treaty allows a state to send. Treaty terms should be designed with these reputational consequences in mind. Treaties that are likely to confer reputational benefits on member states should require states to accept some burden in return. Similarly, links between treaty membership and compliance and various benefits of membership in the international community, such as foreign aid, foreign investment, and trade, could be more frequently and effectively used to attract states into joining and complying with treaties. The European Union, for example, currently conditions membership in the Union on membership in the European Convention on Human Rights and acceptance of the compulsory jurisdiction of the European Court.156 It is no coincidence, then, that the Convention enjoys unparalleled participation and compliance. Individual states, too, have made similar attempts to link treaty commitment and compliance to the provision of various benefits. The United States, for example, has conditioned foreign aid on potential recipients’ human rights practices since the 1970s.157 These provide promising models for using the benefits provided by the international community to strengthen the laws that govern it. Conclusion How does international treaty law shape what states do? How can law that is both voluntary and often unenforced have any effect on the behavior of those it aims to govern? In this Article, I have sought to confront and respond to these questions by putting forward an integrated theory of international law that builds on the insights of political science and legal scholarship, on norm-based and interest-based approaches. In it, I seek to emphasize how the distinctive characteristics of international law lead not to legal impotence but instead to unique, predictable behavioral patterns. I have argued that because international treaty law is voluntary, compliance with international treaties cannot be understood without first understanding why states commit to treaties in the first place. Commitment and compliance are interwoven: States commit based on their incentives to comply, and states comply based on their incentives to commit. And because international treaty law is often not enforced by any central body or international actors, we must look to other factors to understand the impact of treaties on state behavior. Although international treaties that incorporate transnational legal enforcement are more likely to be effective, treaties that are not enforced in this way still can have profound effects. To begin with, domestic enforcement of treaty requirements by form and substance are intertwined and that it can be difficult, if not impossible, to change one without changing the other). 156 See note 93. 157 Of course, this condition has not always been perfectly observed, varying markedly among different administrations. Between Power and Principle 64 domestic actors through domestic institutions can serve to ensure that nations abide by their international legal commitments. However, because not all states enjoy robust domestic rule of law institutions, reliance on domestic enforcement of treaties can lead to a troubling pattern: Countries with strong domestic institutions (and hence strong domestic enforcement) are sometimes more reluctant to bind themselves to international treaties than are countries with weaker such institutions, even if their practices are more consistent with the requirements of the treaty. Moreover, collateral consequences of treaties can be just as profound in their effects as transnational legal enforcement. Many of the most powerful effects of international legal rules play out in the collateral responses of actors to the signals treaty membership sends and the possibilities for linkages treaties create. Yet the precise effect of collateral consequences depends upon the ability of actors to monitor and respond to violations and, even more crucially, on the reciprocal effect of the incentives to comply on the state’s decision to commit in the first place. If the integrated theory contains one overarching lesson, it is this: International law is neither as weak as its detractors suggest nor as strong as its advocates claim. It is not mere window dressing nor is its power similar to that of domestic law. To view international law through either lens is to see international law through a glass darkly. That is what too much of the writing on international law thus far has done. By discarding an all-or-nothing approach in favor of a more nuanced understanding of when and how international law can shape what states do, we can find ways to use international law more effectively to bring order to a world that desperately needs it. Between Power and Principle 65 APPENDIX A: LIST OF TREATIES Short Name of Treaty Full Name and Citation of Treaty Convention Against Torture Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, opened for signature Dec 10, 1984, S Treaty Doc No 100-20 (1988), 1465 UNTS 85 (entered into force June 26, 1987) Article 21 Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, opened for signature Dec. 10, 1984, art 21, S Treaty Doc No 100-20, at 26-27 (1988), 1465 UNTS 85, 118-20 (entered into force June 26, 1987) Article 22 Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, opened for signature Dec. 10, 1984, art 22, S Treaty Doc No 100-20, at 27-28 (1988), 1465 UNTS 85, 120 (entered into force June 26, 1987) Vienna Convention Vienna Convention on the Protection of the Ozone Layer, opened for signature Mar 22, 1985, TIAS No 11,097, 1513 UNTS 293 (entered into force Sept 22, 1988) Montreal Protocol The Montreal Protocol on Substances that Deplete the Ozone Layer, opened for signature Sept 16, 1987, 26 ILM 1550 (entered into force Jan 1, 1989) Copenhagen Amendment Copenhagen Amendment to the Montreal Protocol, opened for signature Nov. 23, 1992, 32 ILM 874 (1993) (entered into force June 14, 1994) Between Power and Principle 66 APPENDIX B: DATA SOURCES AND DESCRIPTIONS Torture. I generated the data on torture by coding the sections on torture in the United States Department of State Country Reports on Human Rights.158 The Torture index ranges from 1 to 5. CFCs. I derive the data regarding CFCs from the United Nations Environment Programme’s Secretariat for the Vienna Convention and the Montreal Protocol. The data measure the consumption of ozone-depleting substances covered in Annex A to the Montreal Protocol (Group I: Chlorofluorocarbons (CFCs)). The totals are measured in Ozone Depleting Potential (ODP) Tons.159 Democratic Regime. The definition and measurement of democracy has been the source of a great deal of debate among scholars.160 I use here what is widely recognized to be the best available comprehensive data on democracy—the measure of democracy (DEMOC) in the Polity IV data set.161 The scale is constructed additively using coded data on six separate variables: competitiveness of executive recruitment, openness of executive recruitment, regulation of executive recruitment, constraints on the chief executive, regulation of political participation, and competitiveness of political participation.162 I transform this 11-point scale into a 0/1 variable, with a “1” indicating a “democratic regime” (6 to 10 on the polity scale), and “0” indicating a semi- or nondemocratic regime (0–5 on the polity scale).163 Ratification and Signature. I obtained ratification and signature data on each of the treaties examined here from the database maintained by the Secretary General of the United 158 For more on how I constructed the index, see Hathaway, 111 Yale L J at 1969–72 (cited in note 17). 159 United Nations Environment Programme, Secretariat for the Vienna Convention and the Montreal Protocol (the Ozone Secretariat), Production and Consumption of Ozone Depleting Substances 1986–2000 (2003), available at http://geodata.grid.unep.ch (visited Nov 24, 2004). 160 See Hathaway, 111 Yale L J at 2028–29, n 311 (cited in note 17). 161 This is the best available comprehensive data on democracy. It defines democracy as “general openness of political institutions,” which in turn is defined as an amalgam of competitiveness of executive recruitment, openness of executive recruitment, regulation of executive recruitment, constraints on the chief executive, regulation of political participation, and competitiveness of political participation. See Monty G. Marshall and Keith Jaggers, Polity IV Project: Political Regime Characteristics and Transitions, 1800–2002, online at http://www.bsos.umd.edu/cidcm/inscr/polity/index.htm (visited Nov 24, 2004) (including a description of variables and a link to the data set). Polity IV Dataset Variables List, online at http://www.bsos.umd.edu/cidcm/inscr/polity/index.htm (visited Nov 24, 2004). 162 Id. 163 In addition, I convert codes of -66 and -77 to “0,” and treat -88 as missing, prorating the missing data using surrounding entries, where possible. This is in accordance with the recommendation of the authors of the database. Between Power and Principle 67 Nations in the United Nations Treaty Collection.164 164 United Nations Treaty Collection, online at http://untreaty.un.org/English/treaty.asp (visited Nov 24, 2004).