Introduction: Legalization and World Politics Judith Goldstein, Miles Kahler, Robert O. Keohane, and Anne-Marie Slaughter In many issue-areas, the world is witnessing a move to law. As the century turned, governments and individuals faced the following international legal actions. The European Court of Human Rights ruled that Britain's ban on homosexuals in the armed forces violates the right to privacy, contravening Article S of the European Convention on Human Rights.1 The International Criminal Tribunal for the Former Yugoslavia indicted Yugoslav president Slobodan Milosevic during a NATO bombing campaign to force Yugoslav forces out of Kosovo.2 Milosevic remains in place in Belgrade, but Austrian police, bearing a secret indictment from the International Criminal Tribunal, arrested a Bosnian Serb general who was attending a conference in Vienna.3 In economic affairs the World Trade Organization (WTO) Appellate Body found in favor of the United Slates and against the European Union (EU) regarding European discrimination against certain Latin American banana exporters.4 A U.S. district court upheld the constitutionality of the North American Free Trade Agreement (NAFTA) against claims that its dispute-resolution provisions violated U. S, sovereignty.' In a notable environmental judgment, the new Law of the Sea Tribunal ordered the Japanese to cease all fishing for southern bluefin tuna for the rest of the year* 1. See Lustig-Prean and Becken v. The United Kingdom. App .Nos. 314I7A6 and 32377S6 (27 Sept. 1999); Smith and Grady v. The United Kingdom, App. Nos. 339BSA6 and 33986fl6 (Eur. O. H.R.) (27 Sept, I999)(availableal ). 2. S« Roger Cohen, Warrants Served for Serbs' Leader and 4 Assistants, New York Timet, 28 May 1999, Al; and Raymond Bonner, Despite Indictment. Politici ans and Diplomais Control Milosevic's Fu-liire, New York Times, 28 May 1999, A13. 3. See Marlise Simons, Top Bosnian Officer Arrested for U.N. Tribunal, New Yoik Times. 26 August 1999.A10. 4. See European Communities—Regime for the Importation, Sale, and Distribution of Bananas— Recourse to Arbitration by (he European Communities Under Article 22.6 of the DSU, Decision by the Aibitralors. WTO Doc. WT/D527/ARB (9 Apr. 1999} (available at ). 5. $tc Made in the USA Foundation v. United Slates. 56 F. Supp.2d 1226(n.d.Ala. 1999). 6. See Southern Bluefin Tuna Cases (New Zealand u Japan; Austrália v. Japan) (order of 27 August 1999) (Infi Trib. for the Law of Ihe Sea) (available at ). IniemationaiOrganizationSi, 1, Summer 2000.pp. 385-399 C 2000 by The 10 Foundation and the Massachusetts Institute of Technology 386 International Organization These actions, taken in the course of a single year, were representative of a longer term trend: some international institutions are becoming increasingly legalized. The discourse and institutionfiormally associated with domestic legal systems have become common in world politics. This move to law is not limited to the actions of international tribunals. Legally binding environmental treaties have proliferated in recent years. These agreements often trace their lineage to hortatory political pronouncements but often become closer to hard law over time. The Montreal Protocol on Substances Depleting the Ozone Layer, for instance, is now a legally binding and precise agreement with a related system of implementation review involving third parties.7 Arms control agreements display increasing precision and elaboration in their commitments and in the scale of their implementing bureaucracies. The proliferation of nuclear weapons and the possession and deployment of entire classes of other weapons (among them chemical weapons and landmines) are now subject to detailed legal conventions. Despite these prominent examples, however, the move to law is hardly uniform. Compliance with the judgments of international tribunals and WTO panels remains uneven. Military intervention, both unilateral and multilateral, continues to occur without clear international legal authority. The NATO bombing of Serbia in 1999 was only one recent example of a decline in the precision of rules governing armed intervention, as challenges to the old norms of territorial sovereignty have mounted. Major arms control treaties are stalled by domestic political opposition. Neither exchange rates nor the provision of muldlateral financial aid are subject to precise legal rules. An important environmental initiative, the Framework Convention on Climate Change, imposes only vague obligations, although the recent Kyoto Protocol has substantially increased the density of the norms in this regime. Other regimes, such as that for international whaling, have maintained the same degree of legalization over several decades. Many Asian nations have explicitly rejected legalized institutions, preferring a model that eschews formal legal obligations. Latin American nations have been similarly cautious about pooling sovereignty in independent institutions such as those that characterize the EU, despite repeated efforts to promote economic integration of various parts of the continent The goal of this special issue of 10 is a better understanding of this variation in the use and consequences of law in international politics. Legalization, a particular form of institutionalization, represents the decision in different issue-areas to impose international legal constraints on governments. This issue of 10 defines legalization and elaborates different types of legalization. It charts the extent of legalization and its variation across issue-areas and regions. It explains why actors choose to create legalized institutions.lt investigates the consequences of legalization on participants, on political and legal processes, and on the international system. Finally, it explores how international politics within legalized institutions differs from politics in non-legalized institutions. 7. Viaorl998. Introduction 387 By developing a framework for the study of legalization, we are able to unite perspectives developed by political scientists and international legal scholars and engage in a genuinely collaborative venture. We view law as deeply embedded in politics: affected by political interests, power, and institutions. As generations of international lawyers and political scientists have observed, international law cannot be understood in isolation from politics.8 Conversely, law and legalization affect political processes and political outcomes. The relationship between law and politics is reciprocal, mediated by institutions. In this introduction we map the way in which legalization, its causes, and its consequences will be assessed. We provide a summary of the definition of legalization adopted in this issue and as developed in the article by Kenneth Abbott, Robert 0. Keohane, Andrew Moravcsik, Anne-Marie Slaughter, and Duncan Snidal. We describe the different perspectiveson legalization coming from intemationallaw and political science and outline the theoretical puzzles that legalization poses for international relations theory. Finally, we summarize the articles in the context of the broader research problems faced by each of the contributors to this special issue. Legalization and International Institutions Inlernational institutions—enduring sets of rules, norms, and decision-making procedures that shape the expectations, interests, and behavior of actors—vary on many dimensions. The WTO and the international regime for the protection of polar bears are both institutions, but they differ according to the scope of their rules, the resources available to the formal organizations, and their degree of bureaucratic differentiation. In general, greater institution a lizatbn implies that institutional rules govern more of the behavior of important actors—more in the sense that behavior previously outside the scope of particular rules is now within that scope or that behavior that was previously regulated is now more deeply regulated. Substantial institutionalization can be demonstrated to exist in world politics, but legalization represents a specific set of dimensions along which institutions vary. The definition of legalization adopted in this issue contains three criteria: the degree to which rules are obligatory, the precision of those rules, and the delegation of some functions of interpretation, monitoring, and implementation to a third party. Fully legalized institutions bind states through law: their behavior is subject to scrutiny under the general rules, procedures, and discourse of international law and, often, domestic law. Legalized institutions also demonstrate a high degree of precision, meaning that their rules unambiguously define the conduct they require, authorize, or proscribe. Finally, legal agreements delegate broad authority to a neutral entity for implementation of the agreed rules, including their interpretation, dispute settlement, and (possibly) further rule making. B. See Henkln t979;Schacter 1991; ml Claude 1966. 388 International Organization Each of these three dimensions can vary from high to low, and each can vary independently of the others. Abbott, Keohane, Moravcsik, Slaughter, and Snidal construct a typology of legalized institutions that varies from an ideal-type of complete legalization in which all three properties are high through various forms of partial legalization to institutions without legalization. Abbott and Snidal, in their article, elaborate this typology along a spectrum from hard to soft law, and develop a rich and innovative set of candidate explanations for this variation. They draw on a wide range of actual examples of more or less legalized regimes within the categories set forth by the typology. This definition does not portray legalization as a superior form of institutionalization. Nor do the contributors to this special issue adopt a teleological view that increased legalization in international relations is natural or inevitable. In using the concept of legalization to guide our collective analysis, we consider softer variants (lower legalization) to be of equal interest to hard law. Why actors move from one form of legalized institution to another is also central. Such moves include the formalization of an informal understanding or customary practice, the adoption of systematic rules to crystallize and codify practices as they evolve, and the strengthening of delegation to increasingly independent and powerful third-party tribunals. The Uneven Expansion of Legalization Legalization has expanded in contemporary world politics, but that expansion is uneven. In this section we review the approach of recent international legal scholarship, which has chronicled and categorized this "move to law" but has largely failed to evaluate or challenge it. Approaches from political science should be more helpful in explaining the puzzle of uneven legalization; the remainder of the section describes the insights and shortcomings of the political science literature. Variation in Legalization The revival of the use of law in international politics has not gone unnoticed, although legal scholarship has analyzed it through different lenses. The actual or arguable power of law and courts has been one central perspective in the study of the EU and the WTO, for example. Many specialists in law accept that European economic integration has depended on the construction of an effective EU legal system. The European Court of Justice (ECJ) is depicted as a central player in this process, propounding the legally binding character of both the Treaty of Rome and EU directives and ensuring that EU law was enforceable through third-party adjudication, often initiated by private individuals and firms.9 Recently, political scientists and lawyers have challenged thiscanonical narrative, emphasizing congruence between ECJ judg- 9. See Stein 1981; Weil« 1991; and Burley andMallli 1993. Introduction 389 ments and state interests and variation in the relationship between the EC J and different national courts."1 In the legal literature, these debates now fall under the larger rubric of the "constitutionalization* of the Treaty of Rome," a view that emphasizes the role of courts in a unique effort to construct a supranational European polity rather than the legalization of an intergovernmental regime. Constitutional rhetoric and an emphasis on a growing role for law and third-party adjudication also figure in research on the evolution of the world trading system, now increasingly understood as the emergence of an international economic constitution.12 Constitutionalization, however, is a very broad brush, sweeping in foundational problems of social order. More narrowly focused studies of trade regimes and of the commercial arbitration agreements that facilitate foreign direct investment instead describe the phenomenon of "judicializa-tion" of dispute-settlement processes, essentially emphasizing our third variable: delegation.13 Many scholars have analyzed the judicialization of the General Agreement on Tariffs and Trade (GATT) dispute-settlement system, chronicling the long-running struggle between trade "legalists," those seeking third-party adjudication of trade disputes under clear legal nil es, and trade "pragmatists," those supporting nonbind-ing forms of dispute resolution that allowed more scope for power and diplomacy.14 The evolution of GATT dispute-settlement procedures in the direction of a more legalized regime under the WTO represents a victory for the legalists. Contemporary WTO panels are conducted in accord with legal norms. Lawyers present detailed legal arguments that require a response from all parties; panel members construct their decisions with the assistance of a legal secretariat that helps them to resolve legal issues rather than broker a political compromise.15 International commercial arbitrationhasexperienceda similar evolution,as arbitrators become more like judges and arbitration procedure becomes more like judicial procedure.16 An emphasis on courts as both creators and guarantors of an international rule of law extends beyond economic regimes. International lawyers have drawn attention to the proliferation of international and supranational tribunals in such issue-areas as human rights, the law of the sea, intellectual property, and international environmental protection.17 These tribunals range from courts with direct jurisdiction over individual claims and enforcement powers over national governments to much less ambitious "noncompliance bodies" designed to oversee implementation of various environmental agreements.18 An entire subspecialty has developed in international 10. See Garrett and Weingasl 1993; Gatreu, Kelemen, and Schulz 1998; Slaughter. Slone Sweet, and Waler 1998; and Aller 199Ba,b. 11. See Weiler l999;andSloneSweei 1998. 12. See Jackson 1998; and Petersmann 1991. 13. SeeHudecl992;andSloneSweeil999and2000. 14- See Trimble 1985; Jackson, Louis.andMalsushiia I984;and Jackson and Davey 1986. 15. Hudec 1992and 1999. 16. SeeLilliehandBrowerl994;andSloneSweeil999. 17. Romano 1999. 18. Ibid. 390 International Organization criminal law, driven by the creation of international war crimes tribunals, such as the Bosnia and Rwanda tribunals, proposed hybrid arrangements involvingboth national and international judges in Cambodia and possibly East Timor, and a nascent Intema-tional Criminal Court.19 The legal literature typically describes these tribunals, analyzes and evaluates their decisions, and theorizes their relationship to one another in terms of a global legal system.2" A growing number of scholars also seek to evaluate their effectiveness based on a set of criteria developed across issue-areas.21 What the legal literature omits—and what this special issue includes—is an explanation for government decisions to establish such tribunals. Somewhat paradoxically, legal scholars also often fail to analyze courts in the larger context of legalization. The broader definition of legalization that informs this special issue includes judicialization, but it also emphasizes the importance of rule precision and degree of obligation. For example, certain international agreements, including in part the UN Conference on the Law of the Sea Convention, are expressly designed as instruments of codification. Other agreements, such as the transformation of voluntary guidelines of the UN Food and Agricultural Organization and the UN Environment Program regarding hazardous pesticides and chemicals into legally binding treaties are changes in the obligatory rather than the formal qualities of law.22 The rules themselves can become more specific, highly elaborated, and technical, qualities captured in the concept of "juridification" in some areas of domestic law.23 Legalization as defined here also allows capture of a wider range of variation in international institutions. A singular focus on high-profile international tribunals can obscure a broader spectrum of delegation. The World Bank, for example, under its general mandate to promote development, has developed operational standards for environmental impact assessment, treatment of indigenous peoples, and participation of nongovernmental organizations in project planning. These policies become legally binding on borrower states when incorporated in loan documents, and they are enforced by the World Bank's Inspection Panel. In other instances, states avoid such delegation. Merit Jano w notes, for example, that "even the most legalistic of APEC's members [the United States, Canada, and Australia] have not called for the creation of an expanded APEC bureaucracy or the development of a new supra-national authority to administer or develop APEC-wide rules."24 Even within particular treaty regimes, levels of delegation can vary. NAFTA contains a wide variety of dispute-settlement mechanisms: binational panels that produce resolutions enforceable through national courts, trans govern mental commissions appointed to oversee specific issue-areas, and interstate bargaining with a nonbinding panel decision as a focal point.23 19. See Bassiounil999;Morris and Scharf I995and I998;and Rainer and Abrams 1997. 20. See Romano 1999; and Chamey 1999. 21. See Heifer and Slaughter 1997; Romano 1999; Pan 1999; Noyes 1998; Knox 1999; and Heifer 1998. 22. Mekouarl998. 23. Teubnerl987. 24. Janow 1996/1997. 25. Morales 1997. Introduction 391 Just as many existing treatments of the move to law have concentrated on the dimension of delegation and the creation of new judicial or quasi-judicial tribunals, so examination of the consequences of legalization has emphasized its effects on domestic legal institutions. NAFTA's domestic legal effects have varied to the same degree as its legalization. The binational panels established under NAFTA Chapter 19 have changed U.S. administrative behavior in the determination of unfair trading practices that warrant the imposition of countervailing duties.2* Gilbert Winham describes the NAFTA Chapter 19 achievement as the possible "beginning of an internationalized form of administrative law."17 A Mexican legal scholar argues that increased competition from international tribunals promotes domestic judicial reform in Mexico.18 On the other hand, NAFTA's transgovemmental Commission on Environmental Cooperation "is too weak to create the pressures necessary to cause substantial redrafting of environmental legislation" and is useful largely as a device for disseminating information about effective domestic environmental law.29 What is again often missing in these accounts is an exploration of the wider political context. Explaining Legalization The expansion of legalization into new domains and the unevenness of that expansion raise theoretical puzzles that have remained largely unexplored. Why and when do states choose legalized institutional forms when their autonomy would be less constrained by avoidinglegalization?How do legalized constraints operate to change government behavior, if they do? Are efforts to legalize certain issue-areas in world politics realistic attempts to facilitate cooperation or misguided attempts to construct a stable order on the basis of fragile norms rather than the realities of power politics? Although contemporary theoretical perspectives in international relations can suggest tentative answers to these puzzles (or at least suggest where to look for the answers), in recent decades few international relations scholars have directly tackled the question of legalization and its consequences. Realist arguments regard international legal constraints as either nonexistent or weak. For those propounding an anarchic international politics based on national self-help, the central puzzles are why states devote so much attention to constructing legalized institutions that are bound to have so little effect and why states accept as credible pledges to obey legal rules that could effectively bind them to act in ways that might be antithetical to their interests. To the degree that legalization represents rules that do bind at least some governments, the realist explanation is clear: legal rules emanate from dominant powers and represent their interests. Legal rules that "work" bind the weaker members of the system; enforcement of those rules ultimately depends on a willingness by stronger powers to bear the costs of enforcement. 26. Goldstein 1996. 27. Winham 1998. 28. Fii-FierroandLopes-Ayllon 1997. 29. KaustiaJa 1996. 392 International Organization Normatively, realist political scientists have been skeptical of the value of international legal advances. E. H. Carr famously criticized legalism divorced from power and politics,-30 George F. Kennan attacked what he saw as the "legalism-moralism" of U.S. foreign policy.11 Legalismas an approach to world politics—if not legalization as defined here—provides comforting and delusional justification for policies that are inconsistent with the realities of interest and power. Those policies are likely to collapse underpressure, often with catastrophic consequences. In this view, legalism may constrain intelligent diplomatic accommodation. Institutionalist theory offers a different, though only partial, answer to these puzzles. Functionalist theories of international institutions, which stress the role of institutions in reducing uncertainty and transaction costs, have seldom dealt directly with the distinction between legalized and non I egal i zed agreements.32 In certain respects the study of international institutions in political science has been directed to demonstrating that informal institutions—not legalized and lacking any centralized enforcement—could still be effective. On the basis of institutionalist theory one should expect frequent informal agreements, some formal rules, and loopholes that provide flexibility in response to political exigencies. Institutionalist theory accommodates such an t [legalist realities as the neglect of dispute-settlement mechanisms in GATT during the 1970s, at the same time that voluntary export restraints and other "gray-area" measures proliferated outside the GAIT legal structure. Institutionalist theory has explained how cooperation endures without legalization, but it has not explained legalization. Abbott and Snídal extend institutionalist reasoning to an explanation of legalization and its variations in their article in this issue. Although liberal approaches to world politics have been most closely associated with a promotion of international law and legalized institutions (for example, the International Court of Justice), liberalism's theoretical contribution to the investigation of legalization lies more in its emphasis on the importance of domestic politics: the preferences of domestic groups and their mobilization and representation in domestic and transnational political institutions. The attention paid by liberal theorists to the relationship between domestic politics and international institutions produces hypotheses regarding the causes and consequences of legalization. Certain forms of international legalization—in particular those that fall under the category that Keo-hane, Moravcsik, and Slaughter term "transnational legalization"—may be designed to grant domestic actors direct access to international tribunals. This de facto shift in the institutional representation for social actors provides a unique form of representation for many social actors—one that reduces the cost of political action, thereby increasing the flow of internationally directed legal action and hence the likelihood of further development of legal rules. 30. Out 1939. 31. Kennan 1984. 32. Charles Lipson's research on informal agreements is a partial exception,emphasizing lhal informal agreements can promote cooperation and seeking to »plain how decentralized systems of incentives can help lo make agreements effective without a formal legal lysiem. Lipson 1991. Introduction 393 Governments and domestic groups may also deliberately employ international legalization as a means to bind themselves or their successors in the future. In other words, international legalization may have the aim of imposing constraints on domestic political behavior Finally, liberal theory suggests that the primary site for the enforcement of international law is ultimately domestic. International legal norms are most effectively enforced when they are embedded in autonomous domestic "rule of law" legal systems through legal incorporation, judicial acceptance, or acceptance by lawyers and litigants. The more important incentives for compliance are ultimately domestic.*3 Construct! v i sts have called attention to the basis for international identities and institutions in shared norms and beliefs, but they have not explained the distinctiveness of legal norms or why actors sometimes prefer to reinforce normative consensus with legalized institutions. Some constructivist scholars explicitly integrate their approach with the study of law.34 Others emphasize the particular role that law can play as the "crystallization of stale expectations," suggestinga dynamic process of "hardening" norms over time.35 Within legal scholarship, Thomas Franck's theory of legal legitimacy, which explains the "compliance pull" of legal norms through theirdeter-minacy, pedigree, adherence, and coherence, also fits easily within a constructivist frame of analysis.36 Overall, constructivist explanations of legalization are likely to hinge less on functionalist and interest-driven accounts and more on historically contingent narratives regarding the emergence of a particular legal understanding." If legal scholars have provided a rich and valuable account of one dimension of contemporary international legalization—increasing delegation to judicial and quasi-judicial institutions—international relations theorists have pointed to several alternative explanations for the variable appearance of legalization and likely sites for examining the consequences of legalization. Each of these approaches finds expression in this special issue, as a more detailed description of the articles will demonstrate. Organization of the Special Issue The first three articles present a theoretical frame that each of the subsequent articles deploys. Abbott, Keohane, Moravcsik, Slaughter, and Snídal set forth and elaborate the definition of legalization. A typology of legalization, presented in Table 1 of their article, provides a classificatory structure for comparative analysis. The case studies in the remaining articles (except the article by Abbott and Snidal) can be located and compared with one another in the terms set by this table. Abbott and Snidal develop the spectrum of hard and soft legal arrangements. They provide explanations for decisions to make agreements "harder" or "softer" on one 33. Sec Slaughter 1995a; and Matlli and Slaughter J99Bb. 34. Finnemore 1996.139-43. 35- Finnemoreand5ilikinkl998. 36. Fianck 1990. 37. Fiiwemore and Sikkink 1998. 394 International Organization or more dimensions that focus on contracting and transaction costs theory as well as normative considerations. In their view legalization can help states and other actors resolve the commitment problems that are pervasive in international politics, reduce transaction costs, and expand the groundsfor compromise. These benefits stem from both interest-based and norm-based processes, and they accrue to interest-based and norm-based agreements. But legalization also entails contracting costs of its own, as well as imposing constraints on government action (autonomy costs). Under different conditions, including different levels of uncertainty and different time horizons among actors, hard and soft law will imply different ratios of costs and benefits. Hence, it should be possible, Abbott and Snidal argue, to account for variations in legalization by identifyinghow ins titutional arrangements involving greater or lesser degrees of obligation, precision, and delegation generate particular patterns of costs and benefits. Their approach to legalization can be deployed to explain government behavior as well as the behavior of groups that evaluate their interests for or against legalization. Keohane, Moravcsik, and Slaughter introduce an explanatory variable thai ties legalization to domestic politics more directly: whether individuals have direct access to the dispute-resolution process (transnational dispute resolution) or whether those processes are limited formally to governments (interstate dispute resolution). They describe the different politics that surround these two types of dispute resolution and argue that the two have very different results for the expansion of legalization. Since members of civil society may assess the autonomy costs of legalization very differently than governments, transnational legalization has much greater potential for expansion than traditional interstate legalization. The remaining contributors examine specific institutional arrangements, arrayed by region (Karen Alter; Frederick Abbott; and Miles Kahler) and by issue-area (Beth Simmons; Judith Goldstein and Lisa Martin; and Ellen L. Lutz and Kathryn Sikkink). These contributors evaluate the extent of legalization in each case, providing explanations for the pattern of legalization and what effect legalization has had on participants and policies. Alter concentrate s on the effects of high legalization in the EU on EU policies. She discovers, in accord with the model of transnational dispute resolution presented by Keohane, Moravcsik, and Slaughter, that the politics surrounding European law and legal institutions are different from a strictly interstate model. However, she also argues that the effects of this highly legalized system are highly dependent on mediation by domestic political and judicial institutions. The dynamic of increasing legalization is far from automatic or inevitable. She suggests conditions under which that dynamic is likely to be more or less powerful, or to be reversed in the face of national backlash. Frederick Abbott describes another, less legalized regional agreement, NAFTA, that embodies a high degree of obligation and precision but a much lower degree of delegation than one finds in the EU. NAFTA's design ensures that political leaders will continue to make key decisions. Nevertheless, where delegation has occurred, such as the delegation of midlevel decisions to binational panels, the degree of imple- Introduction 395 mentation of these decisions has been very high. In assessing NAFTA's implementation, Abbott notes that the agreement has not been tested in hard times: political commitment to implementation on the part of member governments has been high. Whether the constraints of NAFTA are binding in the longer term will only be revealed when a member government is dissatisfied with major provisions of the agreement. The case of regionalism in the Asia-Pacific region, examined by Kahler in his article, is the principal example of an explicit choice in favor of low legalization. Kahler demonstrates this choice on the part of Asian governments in a comparison of three different regional organizations: the Association of Southeast Asian Nations (ASEAN), the ASEAN Regional Forum, and Asia-Pacific Economic Cooperation (APEC). Kahler presents institutionalization in the near-absence of legalization. Recent differentiation among these regional organizations, however, suggests not only that the level of legalization could change in the future, but that it is explained by the strategic choice of governments in the face of a changing international and regional environment. The final three articles examine legalization in three key issue-areas: international monetary affairs, trade, and human rights. Simmons explains both the willingness of governments to accept binding legal obligations in international monetary affairs and theircompliancewith those commitments. In an internationalregime that has demonstrated wide swings in legalization over time, she concentrates on Article VIII of the International Monetary Fund Articles of Agreement, which requires members to keep current account transactions free from exchange restrictions. Her analysis illuminates both international and domestic explanations for these choices in favor of legalized commitments. Most provocatively, she calls into question any easy connection between democracy and the choice for legalization (or compliance with legal obligations). Goldstein and Martin analyze legalization of the international trade regime, usually accepted as one of the most legalized global economic regimes. Despite the regime's relatively high level of legalization, Goldstein and Martin emphasize that legalization operates through politics, by changing processes of decision making, interpretation,and implementation.However, their assessment of the domestic political effects of legalization under GATT and the WTO leads to skepticism regarding the positive effects of legalization on national compliance and international cooperation. Overall, the impact of legalization on trade liberalization will depend on its effects on the incentives for political mobilization at home. Lutz and Sikkink examine human rights law in Latin America through the evolution of norms and international legal obligations regarding democratization and two human rights abuses, torture and disappearance. Levels of legalization vary considerably in these three issue-areas, but all three demonstrated marked improvement in the 1980s and 1990s. Lutz and Sikkink propose an alternative explanation based on transnational politics and the evolution of international norms to explain improved compliance in the absence (in some cases) of higher legalization. 396 International Organization Kahler, in his conclusion to the special issue, couples the authors' empirical findings with the theoretical frameworks presented in the first three articles. He emphasizes that variation found in the pattern of legalization across issue-area and region can be explained only through understanding both interstate strategic calculations (will other governments accept legal obligations and higher levels of delegation?) and the domestic politics of participating states. The domestic politics of legalization, in turn, involve both choices for or against legalization by domestic groups and the often unforeseen consequences of legalization on the structure and processes of domestic politics. Legalization and World Politics: Common Assumptions and Working Hypotheses The contributors to this special issue have different disciplinary perspectives and investigate cases that vary by region and issue-area. Nevertheless, they share a common set of assumptions and working hypotheses. Legalization is a specific form of institutionalization. The contributors to this issue all adopt the conceptualization of legalization set forth in "The Concept of Legalization" by Abbott, Keohane, Moravcsik, Slaughter, and Snidal: a fully legalized instilutionis one with high levels of obligation,precision, and delegation. Moving away from this ideal-type of "hard law," institution s can be identified as partially legalized and legalized on one or more of these dimensions. Although the form of legalization may vary, as represented in Table 1 of the article by Abbott et al., throughout the special issue legalization is understood as a particular and distinctive form of institutionalization. All legalized regimes are institutionalized (they have durable rules); but not all institutionalized regime s are legalized. Legalized institutions incorporate relatively precise substantive rules or obligations, though they may also contain the procedural rules and largely hortatory obligations that are characteristic of nonlegalized international institutions. With respect to international agreements that are not highly legalized, the interpretation of rules occurs in national capitals; legal institutions,on the other hand, delegate this function to third parties. Legalized institutions can be explained in terms of their functional value, the preferences and incentives of domestic political actors, and the embodiment of particular international norms. In addition to describing variation in legalized institutions, the authors offer explanations for the choice of a legal form of interstate cooperation. The special issue is not committed to the view that legalized institutions are a "better" or more efficient form of organization. Rather, the authors aim to explain why actors choose legalized forms. The explanations advanced cluster into three broad categories. The first group of explanations is based on the anticipated consequences of a "legal" agreement. The incentive for nations to agree to more, rather than less, legalized accords derives from the functions performed by "harder" agreements. Using Introduction 397 this mode of explanation, Abbolt and Snidal emphasize in particular the benefits of legalization in forging credible commitments and reducing transactions costs. Those benefits, however, may be wholly or partially offset by negotiating costs and the added constraints that legalization imposes on government decision-making autonomy. In this functional view legalized institutions provide a different set of prospective benefits than non legalized institutions. The second group of explanations modifies this model of calculation by unitary actors by adding the calculations of domestic political actors. These influential constituencies will have diverse preferences over the move to legalization. The simplest modification of the first explanatory model finds the source of government preferences in the aggregated preferences of influential groups. Individuals and groups will favor or oppose legalization based on their assessments of whether the outcomes will further their interests. More complexity is added, however, when those domestic political actors make strategic calculations designed to constrain not only other governments in their international behavior but also domestic actors, including their own government. In the eyes of some domestic actors, estimates of the relevant consequences of legalization are not only international but also domestic. Since prospective agreements differentially affect domestic actors, they change domestic politics by mobilizing some acton and giving them greater access to policymaking. In particular, precise agreements with binding arbitration introduce new actors into politics, such as domestic courts and lawyers, and they also change the venues in which disputes are handled, away from national capitals and into court-like forums. A third ex planati on for the choice of more legalization lies in normative evolution. Legalization can change domestic normative discourse regarding the efficacy of the rule of law. Some actors favor law not only because it serves their interests but also because they believe decisions taken according to legal precepts are superior to other forms of governance. Belief in law as a "good" is not evenly distributed in the population or across regions. Certainly, lawyers more often believe in the use of law than other occupational groups, though it is hard to separate the normative from the material basis of their support. "Rule of law" societies, given a precise definition by Simmons, appear to have a different record of compliance with legal obligations than other societies. Understanding the interaction of norms and legalization is a challenge for social science. In this special issue, only Lutz and Sikkink explicitly take up the task. However, other contributors also suggest an interaction between an interest-based and a norms-based explanation for regime creation. A key consequence of legalization for international cooperation lies in its effects on compliance with international obligations. Many of the contributors examine the relationship between legalization and compliance. Without legalization, compliance is a difficult concept to define or measure, since no authoritative body exists to interpret its meaning or apply it to particular cases.38 Legalization has its principal effect on compliance and international cooperation through the mobilization of indi- 38. Simmons 1998,78. 398 International Organization vidualsand groups in domestic politics—the compliance (or noncompliance)constituencies discussed in Kahl er's conclusion. As Keohane, Morave sik, and Slaughter argue, these compliance constituencies can have a transnational as well as a national character. Different forms of legalization, particularly interstate and transnational legalization, generate a different political dynamic, leading to variation in compliance with international obligations. All of the contributors agree that legalization has led to some behavior change, although the magnitude and direction of these effects on compliance and international cooperation vary. Perhaps the most prominent cases are provided by the EU and the WTO, in which delegation to courts and quasi-judicial bodies has led to more fundamental changes in the locus of decisionmaking. Compliance with obligations, institutional effectiveness, and increased international cooperation may not coincide, in part because of the domestic effects of legalization. The contributors do not base their investigations on any assumption that legalization, as compared to other forms of institutionalization, will necessarily enhance international cooperation. The increased certainty produced by legalization could, in principle, reduce the risks of agreement and therefore enhance cooperation.39 For the functional reasons already stated, legalized agreements and institutions may induce more long-lasting agreements. The changes in domestic politics generated by legalization may also reinforce the position of interests that favor enhanced international cooperation. On the other hand, because of their increased certainty and precision, legal agreements engender high negotiating costs, both across societies and within them: ex ante bargaining is likely therefore to be more prolonged and difficult.4" As Goldstein and Martin argue, clarity of obligation in an agreement may reduce interest in the agreement itself, since its anticipated distributional consequences are clarified as well. A systematic evaluation of the net effects of legalization on cooperation is beyond the reach of this special issue. It would require both valid measures of cooperation in particular domains and an ability to distinguish the effects of legalization from other causal variables, such as increases in economic integration, shifts in domestic interests, and normative evolution. Since legalization, like institutionalization, is to some extent endogenous to such factors, such an assessment would be daunting.4' The effects of legalization on world politics in the long run will depend on its continuing uneven spread. Its spread will depend on the evolution of international norms, its consequences for domestic and transnational politics, and its perceived benefits for key actors. It is possible that legalization at the end of the century will be remembered as a temporary aberration, as it was in the 1920s. The 39. Bilder 1981. 40. Fearon 1998. 4L Keohane and Martin 1999. Introduction 399 articles that follow suggest at least three ways in which legalization could have longer lasting effects on international politics. First, legalization may have a direct effect on the evolution of international norms. Legalization is more explicitly principled than speciňc diplomatic bargains, even if those bargains implicitly incorporate norms. The strengthening of norms helps explain patterns of compliance and the expansion of legalized forms into new issue-areas; that is, beliefs can be self-reinforcing. Second, legalization may permanently change the nature of domestic and transnational politics in participating countries. International law can become internalized.13 Differential access by groups, the expansion of the role of the courts, and the delegation of authority to third parties could lead domestic actors to change their expectation s and behavior and promote an expansion of legalization. Legalizationcouldalsocre ate transnational communities of support for legalized agreements in specific issue-areas. This pro-law epistemic EU would protect international agreements from retrenchment, and its members would serve as transnational advocates for its expansion. Finally, the more often legal agreements are signed and reap cooperative outcomes, the more often actors will return to this institutional form as a model for future agreements. International cooperation is difficult. Actors rely on focal points, not only for the distribution of gains from cooperation but also for models of organization to assure joint gains. Whatever the reasons for success, success may well become associated with legalized forms. (Just as the failure of some legalized institutions in the interwar decades provoked a negative reaction.) Success or failure, whether due to the institutional form or not, may be the most important determinant of whether the "hard" law model becomes more widespread across regions and issue-areas. Legalization and World Politics: Aims and Expectations The editors and authors of this special issue do not claim to have provided a coherent new theory to explain the differentiated phenomenon that we have defined as legalization. Our interest is principally to open some conceptual and analytical doors to a more sustained and explicitly theoretical analysis of the connections between law and politics in contemporary world politics. Overall, the authors express considerable skepticism about the significance and contingency of the international and domestic effects of legalization. No assumptionis made that legalization is a wave of the future. We did not begin with a normative stance, and we make few hard predictions about the future. Interstate legalization, as reflected in the jurisprudence of the International Court of Justice, has not transformed world politics. Likewise, although significant changes have occurred in such areas as trade and human rights, not all of these changes may be causally associated with legalization. We write neither to praise nor to bury legalization, but to analyze its dimensions, its sources, and its current and prospective effects. 42. Koh 1998.