Soft Law As Delegation

Fordham International Law Journal, Dec 2008

This article examines one of the most important trends in international legal governance since the end of the Cold War: the rise of “soft law,” or legally non-binding instruments that are given legal effect through domestic law or internationally binding agreements such as treaties. Scholars studying the design of international agreements have long puzzled over why states use soft law. The decision to make an agreement or obligation legally binding is within the control of the states negotiating the content of the legal obligations. Basic contract theory predicts that parties to a contract would want their agreement to be as credible as possible, to ensure optimal incentives to perform. It is therefore odd that states routinely enter into agreements establishing international rules and regulatory standards in a wide range of subject areas, from banking to arms control to the environmental protection, that are not legally binding. I begin by proposing a new definition of soft. Although previous definitions have often distinguished soft law from hard law on the basis that only the latter is legally binding, no one has explained what distinguishes soft law from purely political arrangements. By contrast, this article defines soft law as those obligations that, while not legally binding, are given some legal effect through separate legal instruments. I then argue that states use soft law as a way to delegate authority over the content of legal rules and regulations to states that possess a particularly strong interest in those rules. Making an agreement non-binding lowers the penalty associated with deviating from the existing legal rules, and thus encourages states with a significant interest in the content of legal rules to unilaterally innovate. This oligopolistic approach to the evolution of legal rules can, under certain circumstances, lead to more efficient legal rules by avoiding the hold-up problem involved in renegotiating contracts in which every state effectively exercises a veto over potentially efficient amendments. As such, the choice between soft law and hard law implicates a tradeoff between the procedural equity inherent in the doctrine of sovereign equality, and the efficiency of legal rules. Soft law privileges the latter over the former.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

Soft Law As Delegation

FORDHAMINTERNATIONAL LAWJOURNAL Fordham International Law Journal As Delegation - 2008 Article 4 Timothy Meyer Copyright c 2008 by the authors. Fordham International Law Journal is produced by The Berkeley Electronic Press (bepress). Soft Law As Delegation Timothy Meyer This article examines one of the most important trends in international legal governance since the end of the Cold War: the rise of “soft law,” or legally non-binding instruments that are given legal effect through domestic law or internationally binding agreements such as treaties. Scholars studying the design of international agreements have long puzzled over why states use soft law. The decision to make an agreement or obligation legally binding is within the control of the states negotiating the content of the legal obligations. Basic contract theory predicts that parties to a contract would want their agreement to be as credible as possible, to ensure optimal incentives to perform. It is therefore odd that states routinely enter into agreements establishing international rules and regulatory standards in a wide range of subject areas, from banking to arms control to the environmental protection, that are not legally binding. I begin by proposing a new definition of soft. Although previous definitions have often distinguished soft law from hard law on the basis that only the latter is legally binding, no one has explained what distinguishes soft law from purely political arrangements. By contrast, this article defines soft law as those obligations that, while not legally binding, are given some legal effect through separate legal instruments. I then argue that states use soft law as a way to delegate authority over the content of legal rules and regulations to states that possess a particularly strong interest in those rules. Making an agreement non-binding lowers the penalty associated with deviating from the existing legal rules, and thus encourages states with a significant interest in the content of legal rules to unilaterally innovate. This oligopolistic approach to the evolution of legal rules can, under certain circumstances, lead to more efficient legal rules by avoiding the hold-up problem involved in renegotiating contracts in which every state effectively exercises a veto over potentially efficient amendments. As such, the choice between soft law and hard law implicates a tradeoff between the procedural equity inherent in the doctrine of sovereign equality, and the efficiency of legal rules. Soft law privileges the latter over the former. Timothy Meyer* I. INTRODUCTION On March 16, 2008, James Dimon, the head of the global investment bank J.P. Morgan Chase, announced that his bank was acquiring the beleaguered investment bank Bear Stearns for a price of roughly two dollars a share.' On Friday, a mere two days earlier, Bear Stearns' closing price was thirty dollars a share,2 and three weeks earlier Bear Stearns was trading for eighty dollars a share.' The sudden demise of Bear Stearns, an event that heralded the end of the modern investment bank, wiped out the life savings of hundreds of employees and cost many more their jobs. And while the Securities and Exchange Commission blamed Bear Stearns' collapse on an old-fashioned bank run, others suggested that the problem lay in the way government agencies the world over regulate the amount of capital that banks have to keep on hand in the event of a downturn.4 Because competition for financial resources is fierce between countries, there is enormous domestic pressure on governments to loosen regulatory restrictions on their banks.5 In order to combat this "race to the bottom," governments have agreed on a set of capital adequacy standards that govern how national regulators do their jobs.6 In other words, international regulatory * Public Policy and Nuclear Threats Fellow, Institute for Global Conflict and Cooperation and the National Science Foundation; Ph.D., Jurisprudence and Social Policy, University of California, Berkeley; J.D., University of California, Berkeley School of Law. I am indebted to Brian Broughman, Robert Cooter, Michael Gilbert, Andrew Guzman, Anne Joseph O'Connell, Robert Powell, and Martin Shapiro for helpful comments at various stages of this project. I would also like to thank the Institute for Global Conflict and Cooperation and the National Science Foundation for funding this project. I am deeply grateful to Willow and Wyatt for their patience. 1. See Eric Dash, Rallying the House of Morgan, N.Y. TIMES, Mar. 18, 2008, at C1. 2. See id. 3. See Floyd Norris, Rescue Plans That Alarm Shareholders,N.Y. TIMES, Mar. 21, 2008, at Cl. 4. See Floyd Norris, The PatientIs Healthy But Dead, N.Y. TIMES, Apr. 4, 2008, at Cl. 5. See id. 6. See generally BASEL COMMITrEE ON BANKING SUPERVISION, INTERNATIONAL CONVERGENCE OF CAPITAL MEASUREMENT AND CAPITAL STANDARDS (1988), bcbs04a.htm [hereinafter Basel Accord I]; BASEL COMMITTEE ON BANKING SUPERVISION, INTERNATIONAL CONVERGENCE OF CAPITAL MEASUREMENT AND CAPITAL STANDARDS: A REstandards circumscribe the freedom of national regulators in national markets. The source of these regulations: two legally nonbinding agreements known as the Basel Accords.7 The Basel Accords (or Basel I and Basel II) are the work product of the Basel Committee on Banking Supervision, an institution created in 1974 by the central bank governors of the world's ten most financially developed nations.' The rise of institutions such as the Basel Committee and the significanceboth in terms of subject matter and scope-of the non-binding regulatory standards produced by these institutions is arguably the most important trend in international legal governance since the end of the Cold War. Soft international law-quasilegal international agreements that do not directly create binding legal obligations-pervades the international regulatory landscape, establishing international standards in areas such as banking,9 trade,1 ° arms control," the environment, 12 and human rights.1" There are free-standing soft law agreements, such as the Basel Accords or the Missile Technology Control Regime, 4 and there are soft law agreements that supplement or implement important sections of prior multilateral treaties, such as the Australia Group' 5 or the Zangger Committee. 6 Yet despite the pervasiveness and importance of soft law agreements, their role in international law and the international regulatory landscape remains under-theorized. Scholars on one side of the soft law debate argue that soft law, because it is not legally binding, is really non-law. 7 Scholars on the other side believe that soft law does establish quasi-legal rules that can affect state behavior, but have failed to fully distinguish the decision to make an agreement non-binding from a variety of other legal devices that increase what is often termed the "flexibility" in an agreement."8 These scholars generally presume that states wish to make their agreements more credible in order to maximize the joint surplus from cooperation.19 Under this premise, the existence of soft law is a puzzle that is usually explained by arguing that commitment is sometimes too costly, that states sometimes prefer more flexible commitments, or that soft law is cheaper to produce as a matter of domestic or constitutional law. While each of these theories adds to our understanding of soft law, none fully explains why states-repeatedly and in areas of vital importance-choose to create international regulatory standards that have no independent legal effect. This Article offers a theory of soft law that explains a range of soft law agreements left unexplained by prior theories. I begin by offering a more precise definition of soft law than is generally employed by scholars: soft legal obligations are those international obligations that, while not legally binding themselves, are created with the expectation that they will be given some indirect legal effect through related binding obligations under either international or domestic law. Most existing definitions, by contrast, either eschew any attempt to reconcile soft law with the doctrinal distinction between binding and non-binding obligations or, at the opposite extreme, focus exclusively on the doctrinal distinction. The doctrinal definition, which I adopt to explain the difference between hard and soft law, is incomplete because it neglects to explain what is "legal" about soft law, or in other words, what distinguishes soft law from purely political agreements.2" While international law and politics are surely re17. See, e.g., Anthony D'Amato, International Soft Law, Hard Law, and Coherence (Northwestern Public Research Paper No. 08-01, 2008), available at abstract= 1103915. 18. See, e.g., Andrew T. Guzman, The Design of InternationalAgreements, 16 EURO. J. INT'L L. 579, 591 (2005). 19. See id. at 588. 20. See infta Part H.C. lated, they are not coextensive, and as lawyers we must be careful to police the analytical boundaries between the purely political and the legal and quasi-legal. My definition of soft law does just this by identifying the chief "legal" characteristic of soft law: the expectation that a non-binding rule will be incorporated into a binding agreement, either as an interpretation of an existing binding rule (at either the domestic or international level), or through the promulgation of a new set of binding rules (again, at either the domestic or international level) based on the nonbinding rules. Having resolved the definitional issue, I proceed to identify one of soft law's heretofore overlooked features: the fact that soft legal rules can more easily evolve in response to individual states' unilateral actions. That this attribute can sometimes be advantageous becomes obvious when one considers that few legal obligations would remain optimal if left unchanged for extended periods of time. Instead, certain rules are appropriate for certain conditions. As conditions change, states will wish to change the rules governing their behavior. Thus, states are constantly reevaluating the adequacy of the existing legal rules and seeking to create more favorable rules. Although it seems obvious, the fact that states may respond to adverse circumstances or changed conditions by seeking to change legal rules is often overlooked by international legal scholars. Legal scholars tend to treat international law as having two discrete phases: (1) states first act as legislators, crafting legal rules that will govern their future conduct,21 and then (2) states, acting as subjects of international law, must decide going forward whether to comply with the rules to which they have agreed.22 In reality, however, when faced with unpalatable rules states do not forego their roles as legislators. They retain the option of seeking to change international rules, just as they can choose to put themselves into temporary noncompliance with international rules. And precisely because states are aware of the possibility that they will wish to amend international rules, they will pay careful attention to the way in which the design of international agreements affects the costs of renegotiating and amending legal rules. 21. See Anthea E. Roberts, Traditional and Modern Approaches to Customary International Law: A Re-Conciliation, 95 AM. J. INT'L L. 757, 784 (2001). 22. See id. Soft law's advantage, I argue, lies in its ability under certain conditions to enhance cooperation among states by allowing states to forego renegotiation in favor of a unilateral or marketleadership method of amending legal rules. The difficulty with renegotiation is that it is costly and frequently fails even when welfare-enhancing amendments are available. Renegotiation, after all, requires a Pareto-improving amendment to the agreement. And though in principle there always exist transfers sufficient to convert any agreement that increases overall welfare into a Pareto-improving arrangement, in practice such transfers may be frustrated by a variety of transaction costs, such as informational asymmetries or hold-outs. 2' Binding international agreements (i.e., treaties) often exacerbate the difficulties with renegotiation because unless a treaty provides otherwise, every party to a treaty exercises a veto over amendments. And while voting rules that permit amendment with less-than-unanimous consent can reduce such transaction costs, they cannot eliminate them. Moreover, majoritarian voting rules can create the possibility of cycling majorities, itself a transaction cost that can offset benefits from non-unanimous voting by making it difficult to settle on any single welfare-improving amendment. 24 By making legal rules non-binding, soft law makes it more likely that states with a major interest in the content of a specific legal rule will unilaterally alter their behavior in response to changed circumstances; states can, in effect, more cheaply exit the existing soft law agreement. Such a unilateral departure from existing legal rules or norms differs from the mine run of violations of international law because it is not taken for opportunistic reasons or for immediate gain. Instead, it reflects a legal policy decision that a state's individual interests are best served by adhering to a completely different set of rules or norms. When a single state (or group of states) unilaterally announces a new legal policy pursuant to an international agreement, a rein23. See Andrew T. Guzman & Beth A. Simmons, To Settle orEmpanel? An Empirical Analysis of LitigationandSettlement at the World Trade Organization,31 J. LEGAL STUD. 205, 206 (2002) (demonstrating that states find it more difficult to settle trade disputes when the issue in dispute has an "all-or-nothing" feature than when it is more of a continuum, suggesting that states find it difficult to construct compensatory transfers). 24. See Neil Siegel, Intransitivities Protect Minorities: Interpreting Madison's Theory of the Extended Republic (2001) (unpublished Ph.D. dissertation, University of California, Berkeley) (on file with UMI Dissertation Services). terpretation of existing legal rules, or a new set of international regulatory standards to which it will adhere, other states must then choose whether or not to adjust their behavior and legal expectations accordingly.25 Under certain conditions, the result is that individual states with a particularly significant interest in the content of a legal rule or rules can drive the evolution of those rules in a globally optimal way. And by choosing to reduce the costs of unilateral action by making an agreement soft, other states have effectively delegated authority for updating the content of the relevant legal rules to the state (or perhaps several states) that they expect to act unilaterally. Thus, in contrast to many theories of flexibility in international agreements, under this theory the flexibility that is desirable is not necessarily flexibility in one's own commitments. Rather, flexibility in the commitments of others can increase the expected value of an agreement. It does so by allowing the agreement to evolve over time to take account of new circumstances without the costs associated with explicit renegotiation. In effect, the procedural equality in the creation of legal obligations that arises from the doctrine of sovereign equality is contracted away in order to capture the benefits that flow from the political reality that all states are not equally positioned with regard to political, economic, and geographic factors. Put differently, under certain conditions allowing states that are especially powerful or interested in a given issue area the flexibility to escape their obligations can actually work to the benefit of the relatively weaker or disinterested states. This unilateral mechanism for driving the evolution of legal rules has costs. States can attempt to unilaterally change rules for their own benefit when such a change is not welfare-enhancing.26 Furthermore, in trying to coordinate on legal rules a pure market-driven approach can result in the breakdown of coopera25. In some cases, this type of amendment will result in an actual amendment of the text of an agreement; in other cases, it will not. Because state expectations are the essence of legal agreements, it is the shifting of expectations that matters. See Andrew T. Guzman, Saving Customary InternationalLaw, 27 MICH. J. INT'L L. 115, 161-63 (2006). Changing the text of an international agreement can help change expectations, but because multiple plausible interpretations of an agreement exist, announcing and implementing a new interpretation may in some circumstances be sufficient to shift expectations. See, e.g., Laurence R. Helfer, Symposium, PublicInternationalLaw andEconomics: Nonconsensual InternationalLawmaking, 2008 U. ILL. L. REv. 71, 75 (2008). 26. See Guzman, supra note 18, at 592. tion because each state may commit to its own preferred rules.2 7 Finally, soft law can reduce, relative to hard law, the incentive of states to invest in reliance on their partners' compliance with legal rules, thereby reducing the total value of the legal commitment.28 In choosing the form of an international agreement, states will trade the expected benefits from more efficient legal rules over time against these costs associated with greater flexibility. In considering whether permitting a more market-driven approach to the evolution of legal rules is efficient, states will consider three variables: (1) uncertainty over the future efficiency of legal rules; (2) the expected benefits from being able to more accurately rely on compliance; and (3) whether a single state or group of states enjoys a comparative advantage in terms of setting and enforcing legal rules. Where uncertainty is high, the benefits from investing in reliance are modest, and a single state or small group of states can act as a focal point for unilaterally adjusting legal and behavioral expectations, states are more likely to choose soft law over hard law. I illustrate the theory with examples drawn from the area of arms control. Since the end of the Cold War, legal scholars have largely relegated arms control to the back burner.29 Events of the last decade have placed arms control, and its legal regulation, firmly back in the political and legal spotlight. Beginning with the discovery at the end of the First Persian Gulf War that Iraq had a nuclear program far more advanced than experts had thought, ° the nuclear nonproliferation regime has been under attack as inadequate and ineffective.3 ' Subsequent events have reinforced this view: in 1998, India and Pakistan both conducted nuclear tests; 32 in 2003, North Korea withdrew from the Nuclear 2009] Nonproliferation Treaty ("NPT")3 3 and in 2006 tested a nuclear weapon;3 4 in 2003, the United States invaded Iraq with the stated justification of enforcing Iraq's nonproliferation obligations; 35 and the United States and Iran are embroiled in a diplomatic contest over Iran's nuclear aspirations that has at times threatened to spiral into armed conflict.36 Additionally, in the wake of the September l1th attacks fears of nuclear terrorism carried out by sub-state actors have captured the imagination of policymakers and the public at large.3 7 In exploring how the theory of soft law presented in this Article explains the design of the nuclear nonproliferation regime and its evolution in response to the events of the last fifteen years, I also hope to refocus the attention of legal scholars on the importance of international legal and regulatory regimes to controlling the spread of weapons of mass destruction ("WMD"). Lastly, a brief word should be said about methodology. This Article employs an institutionalist framework and standard rational choice assumptions to explain states' decisions to employ hard law or soft law. State preferences are assumed to be exogenous and fixed, and states are assumed to be interested only in maximizing their own utility. States thus have no inherent preference for compliance with international law. These assumptions are standard in the international relations literature, and have become increasingly common in international legal scholries Out Nuclear Tests, N.Y. TIMES, May 29, 1998, at A6; John F. Burns, Nuclear Anxiety in India, N.Y. TIMES, May 16, 1998, at A5. 33. See Seth Mydans, Threats and Responses: Nuclear Standoff-North Korea Says it is Withdrawingfrom Arms Treaty, N.Y. TIMES, Jan. 10, 2003, at Al. 34. See David E. Sanger, The North Korean Challenge, N.Y. TIMES, Oct. 10, 2006, at Al. 35. See George W. Bush, U.S. President, 2003 State of the Union Address (Jan. 28, 2003) (transcript available at 20030128-19.html) . 36. See, e.g., Roxana Vatanparast, InternationalLaw Versus The Preemptive Use ofForce: Racing to Confront the Specter of a Nuclear Iran, 31 HASTINGS INT'L & COMP. L. REv. 783, 784 (2008) (supporting the proposition that conflict could ensue by highlighting President Bush's January 2008 statement that Iran was a sponsor of terrorism and that the United States was "rallying friends around the world to confront this danger before it is too late," and U.S. National Security Adviser StephenJ. Hadley's March 2006 statement that "[t]he doctrine of preemption remains sound .... We do not rule out the use of force before an attack occurs.") . 37. See, e.g., Seasons 1, 2, 4, 6, 24 (Fox Network). For example, seasons 2, 4, and 6 of the hit show 24 deal with nuclear terrorist attacks on U.S. soil. All but the first season deal with WMD-related terrorism. See id. arship.3 s While adopting a rational choice approach has costs in terms of the level of detail incorporated into the theory, these costs in the present context are outweighed by benefits in terms of being able to generate a parsimonious theory. As always, rational choice models can and should be complemented by other approaches to similar problems. The Article proceeds in six parts. Part II reviews the existing literature in both law and political science on flexibility in international agreements. Part III explains the positive theory of soft law outlined above. Part IV illustrates the theory with examples drawn from the nuclear nonproliferation regime. Specifically, I examine how nuclear export restraint, a broad and illdefined obligation under the NPT, has been implemented through a series of soft law agreements, and how those agreements have evolved in response to changed circumstances. Part V contemplates the ramifications of this theory of soft law for traditional views of international law. I argue that the theory of soft law presented in this Article has ramifications for a dynamic versus a static interpretation of compliance with international law, as well as for the tradeoff between equity and efficiency. Part VI concludes. II. FLEXIBILITY IN INTERNATIONAL AGREEMENTS Scholars working in an institutionalist framework have begun in the last decade to examine the design elements that states incorporate into international agreements. 9 States have at their disposal a variety of ways in which to increase the flexibility of international agreements, as well as tools to increase the credibility of their substantive commitments. Examples of the former include escape clauses, 40 sunset provisions or planned renegotiation,4 1 and soft law.4 2 Examples of the latter include hard law, dispute resolution, 4' and monitoring and enforcement mechanisms. 44 I contend that different types of flexibility-creating mechanisms respond to different concerns. The purposes of creating flexibility and flexibility's effects on agreement implementation are thus not the same across different design mechanisms. The flexibility created by soft law is unique because it allows legal rules to evolve more easily in response to political realities and changed circumstances. This Section briefly will review the existing treatment of flexibility in international agreements. INT'L ORG. 289 (2001); Alan 0. Sykes, Protectionism as a "Safeguarde" A PositiveAnalysis of the GATT "EscapeClause" with Normative Speculations, 58 U. CHI. L. REV. 255 (1991). 40. See generally B. Peter Rosendorff & Helen V. Milner, The OptimalDesign of International Trade Institutions: Uncertainty and Escape, 55 INT'L ORG. 829 (2001) (acknowledging that escape clauses increase flexibility of an international trade agreement by adding discretion for national policymakers); see also Sykes, supra note 39, at 282 (noting an agreement with an escape clause is more flexible). 41. See Koremenos, supra note 39, at 293 (explaining that planned renegotiation and realignment reduces ex ante uncertainty); see also Helen V. Milner, Peter B. Rosendorff & Edward Mansfield, International Trade and Domestic Politics: The Domestic Sources of InternationalTradeAgreements and Institutions, in THE IMPACT Or INTERNATIONAL LAw ON INTERNATIONAL COOPERATION: THEORETICAL PERSPECTIVES 216, 220 (Eyal Benvenisti & Moshe Hi rsch eds., 2004 ) (highlighting that flexibility-enhancing devices which are often used include sunset provisions or anticipated renegotiations). 42. See Guzman, supra note 18, at 584 (noting that soft law has less of an impact on behavior than hard law); Daniel E. Ho, Compliance and InternationalSoft Law: Why Do CountriesImplement the BaselAccord?, 5J. INT'L ECON. L. 647, 650 (2002) (recognizing the difficulty of measuring compliance and enforcement of soft law because it is difficult to judge when there has been a violation). 43. See Scor & STEPHAN, supra note 38, at 31 (pointing out that enforcement of international law can be subject to enforcement mechanisms the parties design, like alternative dispute settlement procedures); Guzman, supra note 18, at 585 (adopting the conventional view of dispute resolution which assumes that it increases compliance incentives because it provides a mechanism for identifying violations and could specify formal sanctions). 44. See ScoTT & STEPHAN, supra note 38, at 31-39 (discussing types of enforcement mechanisms, including contractual and dispute resolution remedies); Kenneth W. Abbott, "Trust, but Verify:" The Production of Information in Arms Control Treaties and Other InternationalAgreements, 26 CORNELL INT'L L.J. 1, 18, 24-25 (1993) (highlighting that parties might think it necessary to arrange for monitoring procedures and to provide assurance of compliance through enforcement mechanisms). A. Flexibility and Credibility One approach to explaining flexibility in international agreements emphasizes the difficulties of creating credible (i.e., inflexible) commitments in the international system, rather than the affirmative benefits of flexibility. Basic contract theory predicts that two parties negotiating over an agreement should maximize the value of their agreement by increasing the credibility of their commitments up until the point at which greater incentives to perform are inefficient.4 5 Optimal contractual credibility can be created by imposing expectation damages on a breaching party.4 6 Under a perfectly implemented rule of expectation damages, parties fully internalize the costs of breaching an agreement, and thus only breach when the individual benefits of breaching outweigh the global costs.4 7 Credibility enhances the value of an agreement because it permits the parties to invest in reliance on the other party's performance. Uncertainty about whether the other party will perform reduces the expected return to investing in reliance, and thus reduces the overall incentive to create the agreement.4" Thus, even while parties might wish at some point in time to violate an existing agreement, ex ante they wish to make their commitments credible so as to induce Pareto-improving agreements. Third party enforcement, in the form of, for example, court-imposed expectation damages, permits parties to solve this problem of "dynamic inconsistency," i.e., the problem of ensuring that at some 45. See Guzman, supra note 18, at 587 (stating that parties will negotiate an efficient contract, one which generates the maximum possible joint surplus). 46. See RicHARD A. POSNER, ECONOMIC ANALYSIS OF LAW 117-26 (4th ed. 1992); Steven Shavell, Damage Measuresfor Breach of Contract,11 BELLJ. ECON. 466, 471 (1980); John H. Barton, The Economic Basis ofDamagesfor Breach of Contract, 1 J. LEG. STUD. 277, 283-89 (1972). 47. See Shavell, supra note 46, at 471-72. 48. See POSNER, supra note 46, at 117-28. In fact, a damages regime that shifted the risk of loss from investing in reliance from the relying party to the cooperating party would in many circumstances be inefficient because the relying party would have an incentive to over-invest in the cooperative venture. The efficiency of such a regime depends on the likelihood that the cooperating party will actually perform, or in other words on the credibility of the commitment. However, where the likelihood of performance is low ex ante, the party bearing the risk of loss from investing in reliance, whether it is the relying party or the cooperating party, stands to gain less from the agreement in expectation. That party, whichever it is, therefore is less likely to enter into the agreement in the first place. See id. future point in time a party has the proper incentives to perform its part of a bargain. Of course, as is familiar to all students of international law and international relations, international law lacks a third party enforcement mechanism.49 While in the domestic setting, the administrative or procedural costs of dispute resolution and imposition of damages are subsidized by the state, and the state provides judges, law enforcement personnel, and indeed laws themselves, at minimal additional expense to those seeking to invoke the protection of the laws,5 ° the creation and enforcement of international legal obligations is largely a self-help process. With no third party enforcer to sanction violators, states must either rely on direct sanctions, such as retaliation or reciprocal withdrawal of benefits, or on reputational sanctions, to induce compliance.5" The imposition of direct sanctions will usually be costly to the state imposing the sanctions as well as to the target state, so states will be reluctant to use direct sanctions.5 2 Furthermore, where the violation occurs in the context of a multilateral regime, all states benefit from the imposition of sanctions by one state. In such situations, retaliation is a public good, and as such is subject to a free-rider problem. The result is that sanctions are undersupplied. Alternatively, states can use decentralized mechanisms that bring a state's reputation for compliance with international law into play to increase the expected cost of violation.54 Such mechanisms include dispute resolution provisions, monitoring provisions, and enforcement provisions. Because a reputation for compliance is valuable in the absence of third party enforcement, being perceived as violating international law is costly to 49. See, e.g., Robert E. Scott & Paul B. Stephan, Symposium, Self-Enforcing InternationalAgreements and the Limits of Coe rcion, 2004 Wis. L. REv. 551, 599 (2004) (highlighting the difficulty with using third-party coercion to enforce international agreements because of the lack of international entities capable of carrying out sanctions against those who violate international law). 50. Court filing fees, for example, do not even come close to approximating the costs the state bears in operating a court. See Lawrence B. Solum, ProceduralJustice,78 S. CAL. L. RFv. 181, 254 (2004). 51. See Guzman, supra note 18, at 595-97. 52. See id. 53. See generally MANCUR OLSON, THE LoGic OF COLLECTrVE AcrION (1971) (offering this insight about public goods). 54. See Guzman, supra note 18, at 596. states. 55 Unlike direct sanctions, reputational sanctions are costless to impose.5 6 They merely require observing states to update their estimate of the circumstances under which the violating state will comply with international law.5 7 Nevertheless, like direct sanctions, reputational sanctions are negative sum, and are thus costly to include in international agreements.5" Even though they are not costly to impose, neither are they compensation to the breached-against party, and they are costly to the violating party.59 Professor Guzman has suggested that in designing international agreements states will set the marginal expected sanction from violation equal to the benefit from the marginally deterred violation.6" In his view, this tradeoff explains why states often fail to include credibility-enhancing devices, such as dispute resolution or monitoring provisions, in international agreements, and why they choose to structure some agreements as soft law rather than hard law.6 At some point the cost to the parties of making an agreement more credible, and thus less flexible, outweighs the benefits. Under this theory, flexibility is a byproduct of the costliness of creating credibility in an environment in which sanctions are negative sum. As such, soft law might be construed as a second best alternative to hard law. B. Flexibility as a Device to PromoteAgreement Other scholars have suggested that a variety of flexibility-enhancing devices are available to states to promote agreement in situations in which the parties might otherwise be unable to reach an accord.62 Like soft law, many of these flexibility-enhancing devices reduce the penalty for unilaterally altering or deviating from one's legal commitments. In this sense, flexibility-enhancing devices can be thought of as mechanisms that make law more responsive to political realities. They provide insurance to some relevant actor, either the states themselves or the policymakers and politicians negotiating the agreement. By reducing the costs to changing one's legal obligations, states recognize that shifting political conditions will play a greater role in determining not only compliance, but also the evolution of the content of legal obligations.63 1. Escape Clauses Several scholars have suggested that escape clauses are used to expand the number of possible agreements available to states.6 4 Agreements that include escape clauses, such as Article XIX of the General Agreement on Tariffs and Trade ("GATT"),65 give domestic political actors negotiating an agreement a greater incentive to agree by giving them a legal way to avoid their obligations in future situations in which domestic pressures would dictate a violation. For example, if two states were negotiating a free trade agreement, uncertainty over domestic political pressure might preclude an agreement if one side was worried that at some point in the future one of its domestic industries, say, steel, would demand protection from outside competition. By agreeing at the outset that the state can escape its obligations in such a situation, the parties reduce the expected costs associated with an agreement by eliminating the costs of violation in the specified circumstances. 66 Of course, where states are able to anticipate the specific industry or lobbying group that will demand protection, they can engage in contingent contracting. In effect, they can design legal obligations specifically for the situations that they see arising in the future. In this example, the agreement might contain an exemption for the steel industry triggered by certain economic conditions. Where the uncertainty is more general, however, the inability to contract over all of the possible contingencies would preclude some agreements absent some insurance 63. See, e.g., Rosendorff & Milner, supra note 40; Koremenos, supra note 39; Sykes, supra note 39. 64. See Rosendorff & Milner, supranote 40, at 843, 850-52; Sykes, supra note 39, at 66. See Sykes, supra note 39, at 259. They also reduce the ex post value of the agreement, but that reduction in ex post value is necessary to create an ex ante agreement. See id. 283. 283. 65. See Rosendorff & Milner, supranote 40, at 843, 850-52; Sykes, supra note 39, at mechanism for the political actors negotiating the agreements. Thus, while escape clauses may lower the value of an agreement relative to some hypothetical agreement that did not include an escape clause, in practice they are a necessary concession to the domestic uncertainty faced by public officials.6 7 They trade ex post inefficiency for ex ante agreement.6 8 2. Withdrawal Clauses Similar to escape clauses, withdrawal clauses permit states to escape an agreement that has become excessively onerous. Unlike escape clauses, however, which are designed to accommodate only temporary changes in circumstance, withdrawal clauses represent a relatively permanent change in legal status. While certain states have attempted to use withdrawal provisions as a way to unilaterally modify their legal commitments, as when Trinidad and Tobago withdrew from, and then re-acceded to, the First Optional Protocol to the International Convention on Civil and Political Rights ("ICCPR") 6 9 with a reservation that would have eliminated the jurisdiction of the Human Rights Committee to review petitions from capital defendants,7 ° generally withdrawal clauses can only be exercised to permanently alter a nation's legal status.7 As such, exercising a right to withdraw is a strong signal about future behavior. It indicates an intention not to comply with the terms of the agreement being withdrawn from for the foreseeable future, and can result in a variety of costs to the withdrawing state, including loss of access to information sharing, dispute resolution mechanisms, and 2009] monitoring mechanisms.72 And while in some cases the withdrawing state's intentions may have been clear anyway, in other cases, such as North Korea's withdrawal from the NPT or the American withdrawal from the Anti-Ballistic Missile Treaty,73 this signal can have some value to the remaining parties by providing additional information about the withdrawing party's intentions.7T1 An additional benefit of withdrawal clauses is that they permit states to bring their legal obligations in line with their actions. Because, as noted above, sanctions are negative sum, when negotiating an agreement states will wish to create a mechanism by which to abrogate a legal commitment which they no longer intend to follow, and to which they have little intention of returning in the future. In effect, withdrawal clauses are a legal device by which the costs associated with undeterrable violations of law are reduced. While the legalization of withdrawal probably increases the number of withdrawals on the margin, where a withdrawal clause is present the expected benefits from permitting states to unilaterally align their legal obligations with their intentions outweighs the costs of permitting a few more withdrawals than might otherwise have occurred. 3. Planned Renegotiation Another method to increase flexibility is to incorporate a planned renegotiation into an agreement.75 Planned renegotiation typically comes in the form of duration or sunset provisions.7 6 Agreements such as the NPT have initially included duration provisions that allow the parties to escape the agreement at the end of some specified time. Professor Koremenos has argued that the purpose of duration provisions or planned renegotiation provisions in treaties such as the Antarctic Treaty is to deal with uncertainty over the distribution of benefits under a specific cooperative policy.77 At the time of initial negotiation, 72. See Helfer, supra note 39, at 1614. 73. See id. at 1608-09, 1614. 74. See id. The value of this informational signal can, of course, be overstated. Neither North Korea nor the United States was really attempting to hide its actions taken contrary to the substance of the agreements from which they were withdrawing. Id. at 1609. 75. See generally Koremenos, supra note 39. 76. See id. at 293. 77. See id. at 291. the parties expect that the cooperative policy enshrined in the legal agreement they are negotiating will have a certain distribution, but they are unsure about whether in practice the distribution of benefits will mirror their expectations. Planned renegotiation, either in the form of sunset provisions or amendment procedures that reduce the cost of amending an agreement (by permitting a party to demand renegotiation, or by imposing an amendment rule other than unanimity), allows states to learn over time about the actual distribution of benefits under their chosen policy.78 Planned renegotiation, however, can be of limited value. Treaties with planned renegotiations do not permit states to address inefficiencies in legal rules during the time period in which an agreement is in force. While states can still demand a renegotiation, doing so outside of the rules for amendment prescribed by the agreement can be costly.79 Under planned renegotiation, then, legal rules evolve according to a schedule or set of circumstances chosen ex ante. In some instances, adjusting legal rules according to an ex ante schedule will force states to abide by sub-optimal legal rules between opportunities to renegotiate. Furthermore, simply introducing a planned or on-demand renegotiation does not guarantee that renegotiation will be possible. Renegotiation costs are affected not only by the legal rules governing amendment, but by the number of parties and the degree of tension that exists between their interests. Key member-states may hold de facto veto power, either because they are the median member between revisionist and status quo groups under a majoritarian amendment system, or because the agreement in practice cannot work without that particular state's participation. In other words, renegotiation can fail for all the same reasons that initial negotiations might fail. 4. Reservations Reservations are also a device that introduces some flexibility into agreements. Reservations are a mechanism by which states can unilaterally tailor their legal commitm ents after nego2009 ] tiation of the agreement has been completed, subject to the limitations on reservations imposed by the Vienna Convention on the Law of Treaties and any specific limitations included in the agreement8. 0 As such, the flexibility introduced by reservations is more directly related to the initial allocation of legal obligations, rather than to the subsequent desirability of those obligations. Reservations permit states to convey information about their interpretation of their legal obligations, the circumstances under which they will comply with their agreements, and they can also permit agreement in situations in which officials negotiating an agreement underestimate the degree of domestic opposition to certain provisions.8' However, because reservations generally must be submitted with ratification of a treaty, reservations do not permit states to deal with subsequent changes in conditions that make welfare-enhancing amendments available. C. Soft Law Soft law is another institutional element that is available to states to increase flexibility. While various scholars have analyzed the reasons that states employ or fail to employ soft law,82 none have offered a theory that explains the full range of motivations states have for creating soft legal obligations. 1. What is Soft Law? At the outset, there is a definitional issue. Soft law is sometimes used to refer to agreements with imprecise obligations.83 Agreements with imprecise obligations are soft in the sense that a wide range of behavior can be considered compliant. Similarly, soft law is sometimes defined as those agreements that constrain states to a lesser degree either because the obligations are imprecise, or because authority to interpret and implement the law is not delegated to an international organization or dispute 2009] increasingly favor strong states. Where tension exists between the parties, each time the powerful state amends the legal rules, it has an opportunity to move the legal rules closer to its preferred rule. In some cases, this fact could lead relatively weaker states to press for a binding agreement. The marginal sanction associated with a treaty could constrain this type of opportunistic amendment on the part of powerful states. Notice, however, that even in soft agreements there is a built-in check on the opportunism of the powerful. Revisions to the legal status quo must also be less costly to non-revisionist states than their alternative (either complete abrogation of the existing legal regime, renegotiation, or attempting to force the revisionist party to accept the status quo). If the non-revisionist state does worse under the unilaterally amended legal rules than under renegotiation, the state will refuse to recalibrate its legal expectations to fit the unilateral amendment. Thus, unilateral amendment is not coercive from an ex ante perspective. Weak states agree to soft law agreements in situations in which a de facto delegation to powerful states is a superior outcome for the weaker states. Where the ex ante tension between the parties is so great that unilateral adjustment of the rules will unacceptably affect the non-revisionist state, the parties should instead opt for a treaty. By doing so, they can capture any reliance premium, as well as reduce the costs from a powerful state mistakenly believing that it can offer a unilateral amendment to which the other state will agree. The overall result is that soft law will be superior to hard law in situations in which delegating limited responsibility for the substance of legal rules to the more powerful members of an agreement is more efficient from a global perspective than forcing states to renegotiate as welfare-enhancing amendments become available. In short, the benefits in terms of welfare-enhancing legal rules outweigh the costs in terms of opportunism by the powerful in certain situations. IV. THE NUCLEAR NONPROLIFERATION REGIME The nonproliferation regime, with its collection of treaties, soft law agreements, and counterproliferation initiatives, 126 illusrorism. 133 trates the wide variety of formality that international law can take, as well as the ways that soft law can complement hard law. After the end of the Cold War, legal scholars by and large turned their attention away from the nonproliferation regime and its importance to the international legal and political landscape. 127 However, recent events have placed nonproliferation back in the spotlight. Alleged possession of WMD's in violation of international law served as the justification for the American invasion of Iraq in 2003.128 In that same year, North Korea withdrew from the NPT and announced that it was in possession of a nuclear weapon.129 In 2006, North Korea tested a nuclear device, 3 ° the first such test since India and Pakistan tested weapons in 1998.' Diplomatic attempts to induce North Korea to disarm have absorbed much of the Bush administration's efforts during its second term. 132 Iranian attempts to develop a nuclear weapon have greatly increased tensions in the Middle East. Perhaps of most concern, the proliferation of nuclear material to these so-called "rogue" states has increased fears of nuclear or WMD-related ter This Section describes the evolution of the legal rules associated with the nuclear nonproliferation regime. Specifically, the section describes the development of two soft law regimes, the NSG134 and the Proliferation Security Initiative ("PSI"), in response to the political developments of the last decade. These two regimes are "export control regimes," meaning that they seek to regulate trade in nuclear technology, material, and equipment. As such, the developm ent of these legal regimes im2009 ] plicates a fundamental tradeoff between national security and international trade. While there are a variety of soft law export control arrangements including the Missile Technology Control Regime,"3 6 the Coordinating Committee ("COCOM"), which was supplanted by the Wassenaar Arrangement, 13 7 the Australia Group, 138 PSI, and others, I focus on the nuclear export controls for reasons of expositional clarity, as well as the high degree of political salience. A. Overview In 1968, the Nuclear Non-Proliferation Treaty opened for signature. 13 9 The NPT was an attempt by the world's nuclear powers to arrest the spread of nuclear weapons to nations beyond the five that tested a weapon prior to January 1, 1967.140 The major nuclear powers expected that in the coming years a variety of industrially developed nations could develop the technology necessary to make a nuclear weapon. The NPT was a response to this concern, designed to legally deter states from using their technological capabilities to pursue a nuclear device. 4 ' Initially set to expire after twenty-five years, the NPT was indefinitely extended in 1995.142 With Cuba's accession in 2002, only three countries (Israel, Pakistan, and India) remained outside the treaty.1 43 Along with a series of other multilateral treaties, including the Limited Test Ban Treaty,14 4 the Comprehensive Test Ban Treaty,14 5 the series of treaties creating nuclear weapons-free zones,' 4 6 the Chemical Weapons Convention, 1 47 and the Biological Weapons Convention, 48 the NPT sought to make illegal the acquisition or use of WMD's. The NPT creates a number of legal rules; most notably, it limits the right to possess nuclear weapons to the United States, the United Kingdom, the Soviet Union (now Russia), France, and China, and explicitly prohibits nuclear weapons from being transferred to or received by non-nuclear weapons states.' 49 Amendment of the NPT's rules is governed by Article VIII of the NPT, 15 0 which requires that each party to the treaty ratify the amendment before it becomes binding on that party. 5 ' Thus, formal amendment can effectively only occur with unanimous consent. This accorded with the belief of many major powers that there were no conditions under which the acquisition of nuclear weapons by additional states would be beneficial. 5 2 In terms of the theory presented in this Article, uncertainty about the future desirability of a legal ban on the acquisition (or assistance in acquiring) a nuclear weapon was quite low. By contrast, the NPT deferred the implementation of concrete export control rules to related soft law agreem ents.' Arti2009 ] cle III of the NPT provides the legal basis for nuclear export controls.154 It provides that parties shall not transfer nuclear equipment or material that can be used to produce or process fissionable material unless the transfer occurs under safeguards. 15 5 However, the technology necessary to support a nuclear weapons program could not be expected to remain static, and thus greater flexibility in amending the rules governing technology transfers was desirable. 156 Enshrining concrete rules for export control in the NPT itself would subject those rules to Article VIII's onerous amendment procedures. 157 The soft law agreements that implement Article III of the NPT include the Zangger Committee and the NSG,' 58 both of which provide export control guidelines to states that export nuclear technology. These regimes, which date from the 1970s, might be termed the first generation of export control measures. The second generation of export controls, active counterproliferation measures, evolved to enforce nonproliferation norms in the wake of the end of the Cold War and the fears about the acquisition of WMD by rogue states and substate actors. 59 To the extent that the two ideas are distinct, counterproliferation can be seen as different from nonproliferation in that counterproliferation involves the active enforcement of nonproliferation obligations. Nonproliferation obligations, in contrast, are passive obligations to refrain from proliferating. The centerpiece of cooperative international counterproliferation efforts is PSI.16 ° While PSI has been criticized by some as a departure from principles of legality in the regulation of the spread of WMD,1 6 1 PSI is more appropriately viewed in a dynamic legal context as a challenge to existing legal rules that make counterproliferation more difficult.'1 2 B. India and the Nuclear Suppliers Group In 1974, India detonated a nuclear device in what it termed a "peaceful nuclear explosion."' 6 3 Following this event, the major nuclear supplier countries determined that export controls, which at the time were covered by the Zangger Committee's "Trigger List" of items the transfer of which required safeguards, 164 needed to be strengthened. The result was the creation of the NSG, which currently has forty-five members.' 6 5 The NSG promulgates guidelines for member states that govern trade in nuclear-related technology,' 6 6 and operates by consensus, meaning that negotiating an amendment to the NSG guidelines is not necessarily any easier than amending a forty-five State treaty that requires unanimity.1 67 160. See Daniel H. Joyner, The Proliferation Security Initiative: Nonproliferation, Counterproliferation,and InternationalLaw, 30 YALE J. INT'L L. 507, 509-11 (2005). 161. See generally Samuel E. Logan, The ProliferationSecurity Initiative: Navigatingthe Legal Challenges, 14J. TRANSNAT'L L. & POL'Y 253 (2005). 162. SeeJoyner, supra note 160, at 546. 163. Nobuyasu Abe, Existing and EmergingApproaches to Nuclear Counter-Proliferation in the Twenty-First Century, 39 N.Y.U. J. INT'L L. & POL. 929, 931 (2007). 164. See Int'l Atomic Energy Agency, Communications Received from CertainMember States Regarding Guidelinesfor the Export of Nuclear Material, Equipment and Technology, IAEA Doc. INFCIRC/254/Rev.6/Part 1 (May 16, 2003) . Safeguards are procedures that are put in place through an agreement between a state and the International Atomic Energy Agency to ensure that nuclear and nuclear-related materials are not diverted from peaceful uses into a weapons program. See id. 1 3. Article III of the NPT creates an obligation for non-nuclear weapons states to accept safeguards. Nuclear weapons state may voluntarily accept safeguards, but are under no obligation to do so. See NPT, supra note 139, art. III. 165. See Nuclear Suppliers Group, supra note 11. 166. See Int'l Atomic Energy Agency, Communications Received from CertainMember States Regarding Guidelines for the Export of Nuclear Material, Equipment and Technology, IAEA Doc. INFCIRC/254/Rev.8/Part 1 (Mar. 20, 2006) [hereinafter Guidelines for Nuclear Transfers Part I]; Int'l Atomic Energy Agency, CommunicationsReceived from Certain Member States Regarding Guidelinesfor the Export of NuclearMaterial,Equipment and Technology, IAEA Doc. INFCIRC/254/Rev.7/Part 2 (Mar. 20, 2006) [hereinafter Guidelines for Nuclear Transfers Part II]. 167. See Wable, supra note 94, at 749. 2009] Following the discovery of Iraq's advanced nuclear program in 1992, the NSG expanded its guidelines to cover dual-use equipment, material, and technology.16 This extension was a recognition of the fact that the technological environment heading into the twenty-first century was not the same as the environment that the architects of the NPT had contemplated. Most nuclear technology could now be used either for civilian or military purposes.169 One effect of the controls placed on dual-use technology was to preclude trade in civilian nuclear technology with countries that operated outside the NPT. 17 ° This prohibition follows from the NSG Guidelines, which require states not to authorize transfers to a non-nuclear weapons state for use in an "unsafeguarded nuclear fuel cycle activity." 1'71 Thus, civilian nuclear cooperation with nations such as India and Pakistan, states that are outside the NPT and thus under no legal obligation to accept safeguards, is prohibited by the NSG guidelines. 17 2 These soft legal prohibitions initially made at the international level resulted in considerably more rigorous law and regulation at the domestic level. In the United States, for example, the Atomic Energy Act, 173 the Arms Export Control Act,174 and associated regulations implement the export control guidelines that the United States has agreed to at the international level.175 Despite the existence of these soft legal rules that had been incorporated into American domestic law, on March 2, 2006, President Bush announced that the United States and India had reached an agreement that would permit the United States to cooperate with India in the area of civilian nuclear technology.1 76 The agreement followed a Joint Statement of July 18, 2005, which had provided a framework for negotiations. 177 Because the United States had implemented the relevant parts of the NSG guidelines as part of its domestic law, the U.S.-India Nuclear Cooperation Pact required changes to the Atomic Energy Act. 178 On December 18, 2006, President Bush signed into law a bill that would permit the President to waive with respect to India relevant restrictions contained in the Atomic Energy Act upon a presidential determination that several conditions had been met. 179 Among these conditions was that the NSG amend its guidelines to permit the transfer of nuclear technology.181 On September 6, 2008, the NSG did indeed amend its rules to allow nuclear trade with India. 8' The deal represented a direct challenge to the status quo soft legal rules. In effect, the U.S.-India nuclear pact, 8 2 and the subsequent American legislation making the agreement possible, 183 represented a proposed amendment by the U.S. to the NSG guidelines. The U.S. sought to make India an exception to the NSG rules that transfers can only be made to non-nuclear weapons states that have accepted safeguards on all of their nuclear operations (India, like the nuclear weapons states, would not put safeguards on its military nuclear operations).84 The method of the proposed amendment is what is unconventional. Rather than first negotiating for a proposed change within the institutional confines of the NSG, the United States chose to act unilaterally in an attempt to force the NSG to change its guide2009] lines.'8 It reached an agreement with India before making serious efforts to change the NSG Guidelines, and it altered its domestic laws to make nuclear trade with India possible from a domestic legal standpoint. In acting unilaterally, the United States no doubt incurred, at the very least, a reputational cost. Other members of the NSG were initially very critical of the United States, both in the specific context of India, and more generally of the Bush administration's penchant for "going it alone." 186 Countries such as Sweden, Denmark, Ireland, and Austria expressed reservations about an exception to the NSG Guidelines for India."8 7 China, too, expressed concerns about the U.S.-India Pact.' Yet this cost was not as high as it could have been if the legal rules in question had been "hard" legal rules. Under a reputational theory, the reputational sanction for so directly challenging the legal status quo would be higher for a treaty, precisely because a treaty represents a relatively greater pledge of reputational capital.' 8 9 Treaties are also more likely to come with sanctioning mechanisms, such as dispute resolution or provisions permitting retaliatory sanctions or reciprocal withdrawal of benefits that are not called into play by the NSG. The deliberate decision not to include these mechanisms in the NSG reduces the cost of opportunistic violation, but it also reduces the cost to unilaterally attempting to revise the legal status quo. In effect, it permits both a market leadership mechanism through which legal change can occur-the tactic employed by the United States'-as well as a more conventional negotiation-based platform for amending legal rules. Research in economics has found that such a relatively thinly institutionalized approach to achieving regulatory harmonization can in fact be superior to both a pure market mechanism, and a purely negotiation-driven method of achieving cooperation. 19' In the case of India and the NSG guidelines, the American example was quickly followed by France, which concluded a similar civilian nuclear cooperation agreement with India in 2006,192 and eventually by the NSG itself, which, as noted, amended its Guidelines to comport with U.S.-India Pact.'93 The comparative advantage of soft law, then, lies in the fact that the reduced sanction for non-compliance permits states to unilaterally amend legal rules in a range of situations in which amendment might have been precluded under a harder regime. This advantage suggests that soft law will be appropriate in situations in which a market for legal rules or a hybrid market-negotiation model outperforms a strictly unanimity-based or consensus-based approach to developing international law. Where states creating an agreement expect a single state to have a comparative advantage in setting legal rules, it may be beneficial for all states to want to reduce the cost to that single state of offering unilateral amendments. Such unilateral amendments can be welfare-enhancing in a global sense, even if they are not Paretoimproving. To return to the example of the NSG, export control rules implicate a fundamental tradeoff between security and trade. Restricting the sale of nuclear material and technology reduces a state's security, as well as other states' security, but it also provides economic benefits to the state trading. 194 Because the sale of weapons technology causes a negative externality (the harm to other states' security interests), it is in other states' interest to cooperate on the legal restriction of arms sales. However, the restriction of arms sales does little good if there are suppliers 2009] outside of the legal regime that are capable of providing the same technology that is available from those inside the legal regime. India's existence as a stable nuclear power outside of the legal nonproliferation regime created just such a supply of technology, material, and equipment. The rise of a nuclear India suggested a possible need to amend the legal rules governing international nuclear trade. In effect, the magnitude and source of the security externality had increased, and the legal rules needed to be adapted to deal with this difficulty. However, as the NPT's experience attests, modifying a legal regime requiring unanimity can be extremely difficult precisely because states' interests are in some tension with each other. China, for example, is considerably more wary of a nuclear India than is Germany because of its history as a neighbor and rival of India.195 This tension in interests can create costs to amendment that preclude welfare-enhancing amendment because they are not Pareto-improving (and because states are unable to devise transfers sufficient to convert the amendment into one that is Pareto-improving). A more market-driven amendment process, however, means that amendments do not necessarily have to be Pareto-improving. Unilateral action can secure global improvements in welfare in cases in which "harder" legal institutions would have blocked the changes. In the case of India and the NSG, the importance of the United States to the success of the nonproliferation regime generally, and export controls specifically, mitigated the cost to the United States of challenging the legal status quo. The United States is the world's largest exporter of arms, 196 and so to be successful, any restrictions on trade would have to be restrictions by which the United States is willing to abide. American economic power in this area thus alters the American cost/ benefit ratio to unilateral legal revisions, and makes the U.S.preferred rule (trade with India) a natural focal point for cooperation. C. ProliferationSecurity Initiative Just shortly after the NPT reached its peak in membership in 2002, President Bush announced the creation of the PSI.197 PSI was created following an incident in 2002 in which Spanish Special Forces, acting at the behest of American naval and intelligence personnel, boarded an un-flagged North Korean ship, the So San, carrying missile parts to Yemen.' 98 American military personnel subsequently boarded the vessel as well to verify that the equipment being transported was for use in SCUD missiles. 99 Following a formal protest by the Yemeni government, the United States deliberated for two days before determining that there was no justification under international law for seizing the missile parts. Without any legal authority, the United States released the ship with its cargo, 20 0 but began searching for a way to expand its legal authority to conduct similar interdiction operations. The United States' answer, PSI, was specifically designed to avoid the constraints of multilateral treaties such as the NPT.2 °1 States commit to a "Statement of Interdiction Principles," pledging in effect to harness their national political and legal apparatus to halt the proliferation of illicit materials.20 2 However, unlike the multilateral treaties that preceded it, adherence to the "Statement of Interdiction Principles" is voluntary. A commitment to support PSI, and indeed active support for PSI's objectives in the form of participation in PSI conferences or joint interdiction exercises, 203 do not create binding future legal obligations. PSI has no associated international organization, and thus no delegation whatsoever to a supranational body. PSI members have conducted joint interdiction operations, 2° aimed at stopping the spread of WMD-related material and technology, but because the "Statement of Interdiction Principles" is so vague, activities undertaken pursuant to PSI are organized on an ad hoc basis.205 This stands in contrast to the NSG Guidelines, which while technically not legally binding, are nevertheless quite detailed. PSI is thus an even more thinly institutionalized effort at export control than is the NSG. Yet it would be a mistake to conclude that PSI does not have important legal aspects to it. First, as an export control regime, PSI's purpose is to enforce international nonproliferation obligations by making it more difficult to acquire sensitive technology, material, and equipment. Second, PSI includes a number of subsidiary treaties called reciprocal boarding agreements. 20 6 These treaties, between the United States and states such as Panama and Liberia that have large and unregulated shipping registries, 2 0 7 permit the United States to board and search ships flying the flag of the other nation.20 8 These treaties, in effect, internationalize to some extent national jurisdiction over ships in international waters. More precisely, they resemble a delegation from states that cannot seriously police ships flying their flag to a more powerful state, the United States, to enforce international legal obligations. Third, PSI is a challenge to international maritime law. It is an attempt by PSI members to limit the legal effects of national sovereignty on the high seas and in international airspace. This reduction in sovereignty reduces the legal costs to counterproliferating nations of enforcing nonproliferation norms by waiving background legal rules that counterproliferating nations would otherwise have to violate in order to pursue their objectives. Previous analysis of PSI has focused on existing legal justifications for interdictions conducted under the umbrella of PSI. 2°9 Such analyses have concluded that there is authority for interdiction activity conducted in areas of exclusive national sovereignty and territorial sea areas, but that interdictions in international waters and airspace are on considerably narrower legal footing. Generally, the 1982 Law of the Sea Convention ("LOS Convention"), as well as customary international law, prohibit the search and seizure of vessels at sea, unless the state in which the ship is flagged has approved the search.21 ° Some have speculated that Article 110 of the LOS Convention, which provides for interdiction authority "where acts of interference derive from powers conferred by treaty," 21 1 may provide justification for PSIrelated interdictions.2 12 Similarly, United Nations ("U.N.") Security Council Resolution 1540, binding on all U.N. members, creates export control obligations for states, and admonishes states to take cooperative action to prohibit the illicit transfer of WMD-related material. 21" These obligations, however, are sufficiently vague that scholars have continued to doubt whether there is legal authority for PSI-related activities conducted in international territory. Such analyses of the legality of PSI are illuminating in the sense that they provide a way to think about the legality of state action at any given point in time. As American observance of international law in the case of the So San highlights, international law can effectively restrain even the most powerful states. The United States certainly had the ability to seize the cargo of the So San, but chose not to because of its legal obligations. At the same time, such static analyses also miss the dynamic aspect of international law. Creating conflictual legal norms can be 209. See, e.g., Joyner, supra note 160; see generally Michael Byers, Policing the High Seas: The ProliferationSecurity Initiative,98 AM. J. INT'L L. 526 (2004). 210. See U.N. Convention on the Law of the Sea art. 110, Dec. 10, 1982, 1833 U.N.T.S. 397. Article 110 provides for several exceptions to exclusive flag jurisdiction, including reasonable grounds for suspecting that the ship is engaged in piracy, the slave trade, or unauthorized broadcasting. Id. 211. See id. 212. SeeJoyner, supra note 160, at 536-37. 213. See S.C. Res. 1540, U.N. Doc. S/RES/1540 (Apr. 28, 2004). 2009] used to change international law over time. 214 One way soft law can do this is by creating state practice in support of an emerging norm of customary international law.215 A second way is by causing states' views of their existing legal obligations to change. Soft law, by challenging existing law or providing a framework for interpreting vague provisions of existing law, can alter the substantive meaning of international legal obligations. Within the context of the theory of soft law advanced in this Article, PSI displays the three characteristics that make counterproliferation an ideal area for soft legal governance. As the incident with the So San highlights, 216 international legal rules can become less valuable in the face of new challenges. The American decision to abide by international law217 underscored the less than optimal nature of maritime laws that privileged national sovereignty at the expense of the enforcement of nonproliferation obligations. States intent on trafficking in WMDrelated materials were able to hide violations of their nonproliferation obligations by resorting to claims of national sovereignty in international territory. 218 Similarly, smugglers were able to avoid detection by hiding behind the national sovereignty of states that lacked the ability to monitor the trade being conducted by the ships in their national registries. 219 Rules designed to address these developments had the possibility of creating welfare enhancements. Second, the reliance benefits from having an additional sanction associated with counterproliferation obligations undertaken pursuant to PSI are relatively small. While there is some tension among the interests of PSI members, generally all PSI members are interested in pursuing the objectives of counterproliferation. Thus, the cooperative gains from better legal rules regarding counterproliferation, and more effective coordination of counterproliferation resources, outweigh any increase in the 214. See, e.g., Kal Raustiala & David G. Victor, The Regime Complex for Plant Genetic Resources, 58 INr'L ORG. 277 (2004). 215. See Byers, supra note 209, at 528. 216. SeeJoyner, supra note 160, at 508. 217. See id. at 509. 218. SeeJaYoung Elizabeth Kim, The Agreement After the Six-Party Talks: Are There No Alternatives to the "Modified" Version of the 1994 Geneva Agreed Framework?-An Analysis of the Newly-Adopted Framework and Its SignificanceFor the NuclearProliferationIssues Relating to North Korea, 21 TEMP. INT'L & COMP. LJ. 177, 200 (2007). 219. See generally THE PROLIFERATION SECURITY INITIATIVE, supra note 201. supply of counterproliferation that would come as a result of greater sanctions for violating PSI's soft obligation to cooperate. What tension there is in states' interests comes in two forms. The first is that certain states may not see specific interdiction activities or revisions of maritime law as in their interest. An example is South Korea's decision to participate in PSI only as an observer. 220 This reluctance to fully commit is motivated by its special relationship with North Korea, one of PSI's primary targets. The second source of tension is a familiar collective action problem. All member states would prefer to free ride on the enforcement activities of other states, rather than pay the costs of participation themselves. In some cases, specific PSI activities may not be worth contributing to for certain members. Thus, European States may be hesitant to commit resources to interdiction efforts in the Far East, as opposed to the Middle East. This tension is alleviated by the third factor, the presence of a relatively more powerful state, the United States, that can more efficiently update and coordinate counterproliferation obligations. The United States, as a global superpower, has the incentive to contribute to counterproliferation activities around the world. In effect, because it internalizes a larger share of the benefits from counterproliferation generally than do regional powers, it has an incentive to update the rules and supply counterproliferation. The United States clearly enjoys a comparative advantage in terms of national apparatus available for enforcement of nonproliferation obligations. Similarly, the United States may enjoy an advantage in terms of information about the transport of illicit materials derived from its intelligence services. These types of comparative advantages suggest that the United States is the ideal nation to act as a focal point for counterproliferation norms. Moreover, within limits it is in the interest of all other counterproliferating states to delegate de facto rulemaking and coordinating authority to the United States. Because the United States will be the largest contributor to counterproliferation efforts, rules that are closer to the U.S.-preferred rule will lead to an increase in the American contribution to counterproliferation. This increase in the supply of counterproliferation is a 220. See generallyPark Song-wu, South Korea to Observe PSIExercise,KoREA TIMES, Jan. 25, 2006. public good that in some measure offsets the cost to other states of accepting their less-preferred counterproliferation rules. D. The ABM Treaty and OpportunisticLegal Revisionism Of course, unilateral legal revisionism can be deployed opportunistically. A reduction in the cost to states of challenging the legal regime also increases the ability of states to do so for their own personal gain. In choosing the form of a legal agreement, it is this tradeoff between encouraging welfare-enhancing legal innovation and deterring opportunistic revisionism that will determine whether states choose a hard legal regime or a soft one. An example of the opportunistic use of unilateral legal revisionism is the Reagan administration's decision to proceed with its Strategic Defense Initiative22' ("SDI" or "Star Wars") despite the fact that SDI was rather plainly prohibited by the Anti-Ballistic Missile ("ABM") Treaty. 222 Between the conclusion of the ABM Treaty in 1972 and the Reagan administration's decision to proceed with SDI, there had been a dramatic shift in power between the United States and the Soviet Union.223 Gone were the days of detente. The Soviet Union was rotting under the weight of its centrally-planned economy and decaying economic infrastructure, and in less than a decade would cease to exist. This shift in power meant that the prevailing legal regime no longer mirrored the underlying distribution of bargaining power. Rather than attempting to renegotiate the ABM Treaty, the United States simply announced what might be termed an "interpretative amendment. '' 2 2 1 Itannounced, in other words, that its plans for SDI were consistent with its interpretation of the treaty. 225 While the Soviet Union protested that SDI was not in fact consistent with the ABM Treaty, 2 26 there was little that the Soviet Union could, do short of exiting the treaty completely. They defacto acquiesced to the American reinterpretation of the ABM Treaty, and the ABM Treaty operated for almost another two decades with legal expectations having been readjusted by the American action. Notice that the ABM Treaty involved the reinterpretation of a treaty, rather than a soft law regime. This is significant because it illustrates that unilateral challenges to legal rules within an existing legal framework are not confined to soft legal obligations. However, such challenges are considerably more likely to occur in soft legal environments precisely because the cost to legal revisionism is lower. In the case of the ABM Treaty, the decay of Soviet power likely meant that there was no possible agreement that could deter the Americans from embarking on their preferred policy course. The American decision to reinterpret the treaty, rather than withdraw from the treaty as the United States did in 2002,227 and the Soviet decision to tolerate this reinterpretation indicate that a de facto amendment to the existing legal framework was viewed by both parties as being superior to scrapping the legal framework entirely. E. Conclusions This section has drawn on examples from the nuclear nonproliferation regime to illustrate the most important elements of the theory of soft law presented in this Article. The examples illustrate the fundamental tradeoff that states face in choosing the form of legal agreement. Hard law will deter opportunistic violations as well as opportunistic legal revisionism, but at the expense of deterring welfare-enhancing legal revisionism. In the area of export control restraint, the rise of additional nuclear suppliers altered the value of existing legal rules. Unilateral American action has sought to change those rules to take into account the new global conditions. Similarly, active export control, or counterproliferation, has required both the creation of loose counterproliferation obligations, as well as the modification of rules that make counterproliferation more costly. Again, 227. See Manuel Perez-Rivas, U.S. Quits ABM Treaty, CNN.coM, Dec. 14, 2001, The United States announced its withdrawal in December 2001, but per Article XV of the ABM Treaty, the withdrawal was not effective for six months. See id. SOT LAW AS DELEGATION the United States has been out in front acting as a legal entrepreneur. American enforcement power makes U.S.-preferred policy a natural focal point for cooperation. V. SOFT LAW AND GLOBAL WELFARE Up to this point, this Article has focused on analyzing the variables that affect the decision to choose soft law over hard law. The enterprise has been primarily positive, describing the processes through which these variables can affect states' decision to structure an agreement as soft as opposed to hard. However, the positive theory has implications for how scholars, policymakers, and critics think about soft law and its role in promoting international cooperation. Scholars and activists have at times decried incidents in which states have opted for non-binding accords rather than formal agreements. 228 The assumption underlying this claim is that a less formal commitment means that the cooperative project is necessarily less deep or effective than it would otherwise be. By contrast, Professor Raustiala has argued that this desire on the part of domestic constituencies to create binding agreements can actually reduce the depth of cooperation.229 In making an agreement binding, states reduce the depth of their commitment in order to offset the increased formality. 2 0 This can lead to an undersupply (in the economic sense of an inefficiently low amount) of more detailed, deeper soft law agreements. The theory presented in this Article builds on this work by suggesting a specific mechanism through which soft law can have beneficial effects on cooperation over time. Similarly to Professor Raustiala, I argue that soft law can under certain conditions actually enhance cooperation relative to binding legal agreements. In contrast, however, I argue that the mechanism through which soft law enhances cooperation is dynamic. Soft law is superior to hard law in cases in which soft law leads to more valuable legal rules over time. 228. See Richard H. Speier, A Nuclear NonproliferationTreaty for Missiles?, in FIGHTING PROLIFERATION (Henry Sokolski ed., 1996). For example, commentators have occasionally advocated turning the Missile Technology Control Regime ("MTCR") into a binding treaty because the perception is that overall compliance would be increased by making the MTCR's obligations legally binding. See id. 229. See Raustiala, supra note 82, at 603. 230. See id. This dynamic view of international law is important to how scholars think about international law and departures from settled legal norms. The consensual nature of international law under traditional formulations, in which state consent is required to create or adjust international obligations, can block the evolution of international law in response to changing political circumstances. A single necessary state can block welfare-enhancing changes to international law if that state personally does not benefit, or if it wishes to hold out for a larger share of the surplus from cooperation. This can even be the case when states negotiate rules that create non-unanimous procedures for amending legal rules. In such situations, intransitivity or holdouts can still prevent welfare-enhancing changes to legal rules. Unilateral action that causes states to re-coordinate their legal expectations can solve this problem. Viewed dynamically, challenges to the legal status quo can actually be beneficial over time because they lead to a more efficient process of amendment and welfare-enhancing changes in legal rules. This conclusion suggests that in critiquing departures from legal norms, scholars should take not only a static approach-analyzing state action under existing law-but also a dynamic approach, looking for possible beneficial changes to the legal regime as a result of a market leadership approach to coordinating legal expectations. 22331 In particular, where the legal status quo being challenged is codified in a non-binding accord, scholars should consider the possibility that soft law was chosen precisely to encourage such legal innovation. Of course, because unilateral action can be used opportunistically, commentators should consider the specific welfare effects of new legal rules. While the costs associated with deviating even from soft legal obligations will deter some opportunistic revisionism, they will permit more than under hard law. In addition to the distinction between assessing state action under a dynamic or a static view of international law, this paper also has implications for thinking about the efficiency of international law as measured against the equity. Specifically, this Article has argued that where a certain state enjoys a comparative 231. This argument has been made as applied to customary international law. See, e.g., Roberts, supra note 21, at 784. This Article's contribution is to extend this analysis to explicit agreements and the violation thereof. advantage, such as being more powerful than its partners on some relevant dimension, that state may have an incentive to unilaterally innovate, creating a focal point for re-coordinating legal expectations. Where the parties' interests are not in too much tension, sacrificing procedural equity in terms of legal modifications can result in increases in global welfare. In such situations, the de facto inequality of states can actually serve ex ante to benefit not only those who are comparatively advantaged, but those not comparatively advantaged as well. In other words, legal institutions can be designed to harness the benefits of differences in state power, and ameliorate the negative effects when states' interests are in too much tension. This is not to argue that power differentials between states are a net positive for international law. More extensive study of how power affects both the design, and particularly the implementation, of international law is necessary. This Article does suggest, however, at least one instance in which power differentials can be beneficial to the operation of international law: when power differentials reduce the number of possible legal innovators in a soft law regime, thus allowing states to capture some of the benefits, in terms of credibility, of legalizing cooperative rules as well as the welfare gains from having a market leadership mechanism through which legal rules can be modified over time. VI. CONCLUSION I have argued that non-binding agreements (or soft law) can be created by states in response to a specific set of circumstances: (1) when states are uncertain about the future efficiency of legal rules; (2) when the marginal benefit in terms of compliance from moving from soft to hard law is modest; and (3) when a single state or small group of states enjoys a comparative advantage, such as being more powerful on some relevant dimension, that allows it to act as a focal point for re-coordinating legal expectations by acting unilaterally (i.e., as a market leader). This theory of soft law contributes to the growing literature on the design of international agreements. Specifically, international agreements can incorporate flexibility in a variety of ways and for a variety of reasons. I argue that soft law can be a response to the three conditions above, and creates a specific type of flexibility that can lead to the evolution of international law in response to changing political circumstances. Finally, I have illustrated the theory with examples from the area of arms control. Arms control has resumed a central place in today's geopolitical landscape, and it is time that international legal scholars once again begin to examine how law operates in this vital area. VISED FRAMEWORK ( 2004 ), [hereinafter Basel Ac- cord II]. 7. See generally Basel Accord I, supra note 6; Basel Accord II, supra note 6 . 8. See BASEL COMMITTEE ON BANKING SUPERVISION, HISTORY OF THE BASEL COMMIT- TEE AND ITS MEMBERSHIP 1 ( 2007 ), 9. See, e.g., Basel Accord I, supra note 6; Basel Accord II supra note 6 . 10. See , e.g., Asia-Pacific Economic Cooperation ("APEC" ), (last visited Jan . 21 , 2009 ). 11 . See , e.g., Nuclear Suppliers Group,; en/index.html (last visited Jan . 21 , 2009 ). 12 . See , e.g., Report of the United Nations Conference on Environment and Devel- opment , Rio de Janeiro, Braz., June 3-14, 1992 , Non-Legally Binding Authoritative State- Development of All Types of Forests, U.N. Doc . A/CONF. 151/26 (Aug. 14 , 1992 ). 13 . See , e.g., Conference on Security and Co-operation in Europe: Final Act , Aug. 1, 1975 , 14 I.L.M. 1292 ( 1975 ). 14 . See generally Basel Accord I, supra note 6; Basel Accord II, supra note 6; Missile Technology and Control Regime , http://www.mtcr. info (last visited Jan . 21 , 2009 ). 15 . See generally Australia Group, supra note 11 . 16. The Zangger Committee, http://www.zanggercommittee. org (last visited Jan. 21, 2009 ). 27 . See id. at 591 . 28. See id. at 588 . 29. Some notable exceptions include: Jack M. Beard , The ShortcomingsofIndetermi- nacy in Arms Control Regimes: The Case of the Biological Weapons Convention , 101 Am. J. INT'L L . 271 ( 2007 ); David L. Sloss, ForcibleArms Control: Preemptive Attacks on Nuclear Facilities , 4 CHI. J. INT'L L . 39 ( 2003 ); Richard L. Williamson, Hard Law, Soft Law, and Non-Law in MultilateralArms Control: Some ComplianceHypotheses, 4 CHI . J. INT'L L. 59 ( 2003 ). 30 . See Michael Wines , U.S. Is Building Up a Picture Of Vast Iraqi Atom Program , N.Y. TIMES , Sept. 27 , 1991 , at A8. 31. See , e.g., Orde F. Kittrie , Averting Catastrophe: Why the Nuclear Nonproliferation Treaty is Losing its DeterrenceCapacity andHow to RestoreIt, 28 MICH . J. INT'L L . 337 ( 2007 ). 32 . See John F. Burns , NuclearAnxiety: The Overview; Pakistan,Answering India, Car67. See id. 68. See id. 69. See Helfer, supra note 39 , at 1642 . 70. See id. This maneuver was ultimately unsuccessful . The Human Rights Com- leading Trinidad and Tobago to denounce the Optional Protocol completely . See id. 71. See id. at 1583 , 1620, 1626 . I qualify the permanence of the withdrawal because Still Matters-Non-StateActors , Treaties, and the Changing Sources of InternationalLaw, 23 BERKELEYJ. INT'L L . 137 , 169 ( 2005 ) ; see generally ContemporaryPracticeof the United States Tentative Understandingon North Korea'sNuclearProgram,99 AM. J . INT'L. L . 914 ( 2005 ). 78 . See id. at 295 . 79. See generally Barbara Koremenos, Can CooperationSurvive Changes in Bargaining Power? The Case of Coffee, 31 J. LEGAL STUD . 259 ( 2002 ). 80 . See Swaine, supra note 39, at 307. 81. See id. at 311 . 82. See generally Guzman, supranote 18 ; Kal Raustiala , Formand Substancein Interna- tional Agreements , 99 AM. J. INT'L L . 581 ( 2005 ); Kenneth W. Abbott & Duncan Snidal, Hard andSoft Law in InternationalGovernance,54 INT'L ORG . 421 ( 2000 ); Charles Lipson, Why Are Some InternationalAgreements Informal ?, 45 INT'L ORG . 495 ( 1991 ). 83 . See Prosper Weil, Towards Relative Normativity in InternationalLaw?, 77 AM. J. INT'L L . 413 , 414 ( 1983 ). 126 . See Wei Luo, Research Guide to Export Controland WMD Non-ProliferationLaw , 35 INT'LJ. LEG. INFO . 447 , 472 - 73 ( 2007 ). 127 . But see, e.g., Beard, supra note 29, at 271-72 . 128 . See Michael N. Schmitt , U.S. Security Strategies: A LegalAssessment, 27 HARV. J.L. & PUB. POL' , 737 , 752 - 54 ( 2004 ). 129 . See Brent Snowcroft & Arnold Kanter , A SurprisingSuccess On North Korea , N.Y. TIMES , May 1, 2003 , at A35. 130. See David E. Sanger, North Korea Says It Tested a NuclearDevice Underground,N.Y. TIMES , Oct. 9 , 2006 , at Al. 131 . See id. 132 . See , e.g., Mark Mazzetti & William J. Broad , The Right ConfrontsRice Over North KoreaPolicy , N.Y. TIMES, Oct. 25 , 2007 , at A10; David E. Sanger, U.S. Said to Weigh A New Approach On North Korea , N.Y. TIMES , May 18 , 2006 , at Al. 133 . See Remarks by Stephen Hadley, National Security Advisor to the Center for International Security and Cooperation, Stanford, Cal. (Feb. 8 , 2008 ), available athttp:/ / 6 .html. 134. See generally Nuclear Suppliers Group, supra note 11 . 135. See generally CRS REPORT FOR CONGRESS: THE PROLIFERATION SECURITY INITIA- TIVE ( 2006 ), 136. See generally Missile Technology Control Regime , glish/index.html (last visited Jan . 21 , 2009 ). 137 . See generally Wassenar Arrangement, (last visited Jan. 21 , 2009 ). 138 . See generally Australia Group, supra note 11. 139. See Treaty on the Non-Proliferation of Nuclear Weapons, July 1 , 1968 , 21 U.S .T. 483 , 729 U.N.T.S. 161 [ hereinafter NPT ]. 140 . See id. art. IX(3) (defining a nuclear weapons state as "one which has manu- factured and exploded a nuclear weapon or other nuclear explosive device prior to 1 January , 1967 . " ). 141. See generally id. 142. See Jonas,supra note 103, at 421. 143. See id. at 422 . North Korea subsequently withdrew . See id. at 429 . 144. See Treaty Banning Nuclear Weapons Tests in the Atmosphere , in Outer Space and Under Water , Aug. 5 , 1963 , 14 U.S.T. 1313 , 480 U.N.T.S. 43 . 145. See Comprehensive Test Ban-Treaty, Sept . 24 , 1996 , 35 I.L.M. 1439 . 146. See generally Treaty on the Southeast Asia Nuclear Weapon-Free Zone , Dec. 15 1995 , 1981 U.N.T.S. 129 [hereinafter Treaty of Bangkok]; Treaty of Rarotonga, supra note 106; Treaty of Tlateloco, supra note 106 . 147. See generally Convention on the Prohibition of the Development, Production, Stockpiling and Use of Chemical Weapons and on Their Destruction , Jan. 13 , 1993 , 32 I.L.M. 800 . 148. See generally Convention on the Prohibition of the Development, Production, struction , Mar. 26 , 1975 , 26 U.S.T. 583 , 1015 U.N.T.S. 163 . 149. See NPT , supra note 139, arts. 1- 111 . 150 . See id. art. VIII. 151 . See id. art. VIII(2). Article VIII of the Treaty on the Non-Proliferation of Nu- These arrangements virtually preclude amendment) . Id. 152. But see KENNETH WALTZ, THEORY OF INTERNATIONAL POLITICS 180-83 ( 1979 ). 153 . But see Richard L. Williamson Jr., International Regulation of Land Mines, in TIONAL LEGAL SYsTEM 517 (Dinah Shelton ed ., 2000 ) (reasoning that export control is a WMD CoerciveDisarmamentMeasures and TheirLegal Implications , 29 MICH. J. INT'L L . 197 , 285 ( 2008 ). 154 . See NPT , supra note 139, art . III. Article 111 ( 2) of the NPT provides: See id. 155. See id. art . 111 ( 2 ). 156. See id. art. IV. 157. See id. art. VIII. 158 . See David Sloss, Do InternationalNorms Influence State Behavior?, 38 CEO. WASH. INT'L L. REv . 159 , 188 - 89 ( 2006 ). 159 . See generally Jack 1 . Garvey , The InternationalInstitutional Imperativefor Coun- 10J. CONFLICT & SEC . L. 125 ( 2005 ). 168 . See MICHAEL D. BECK ET AL., To SUPPLY OR TO DENY: COMPARING NONPROLIFER- ATION EXPORT CONTROLS IN FIvE KEY 1COUNTRIES 5 ( 2003 ). 169 . See id., at 2 . 170. See ForeignRelations Law-NuclearNonproliferation-CongressAuthorizes the Presi- dent to Waive Restrictions on Nuclear Exports to India, 120 HARV. L. REv. 2020 , 2024 - 25 ( 2007 ). 171 . See Guidelines for Nuclear Transfers Part II , supra note 171 , 1 1 . 172. See Joseph G. Silver , The Global Partnership: The FinalBlow to the Nuclear Non- proliferationRegime, 21 N.Y. INT'L L. REv . 69 , 92 ( 2008 ). 173. See The Atomic Energy Act of 1954 , 42 U.S.C. § 2011 ( 2000 ). 174. See Control of Arms Exports and Imports , 22 U.S.C. § 2778 ( 2000 ) 175. See The Atomic Energy Act of 1954 § 3(e); Control of Arms Exports and Im- ports § (a)(2 ). 176. See Jim VandeHei & Dafna Linzer , U.S., India Reach Deal on Nuclear Coopera- tion, WASHINGTON POST, Mar. 3 , 2006 , at Al. 177 . SeeJoint Statement Between George W. Bush and Prime Minister Manmohan Singh , Office of the Press Secretary, The White House (July 18 , 2005 ), http:// 6 .html. 178. See Steven R. Weisman , Q & A: U.S.-India Nuclear Pact , N.Y. TIMES , July 20 , 2005 . 179. See M.M. Ali , Bush Signs U.S.-India NuclearAgreement , WASH. REP. ON MIDDLE E. AFF. Mar. 1 , 2007 , at 39 . 180. See United States and India Nuclear Cooperation, 22 U.S.C.A. 8001 ( Dec . 18, 2006 ). 181 . See Somini Sengupta & Mark Mazzetti , Atomic Club Votes to End Restrictions on India , N.Y. TIMES , Sep. 7 , 2008 , at A6. Following Nuclear Suppliers Group ("NSG") TIMES , Oct. 2 , 2008 . 182. See generally VandeHei & Linzer, supra note 176. 183. See generally The Atomic Energy Act of 1954 , 42 U.S.C. § 2011 ( 2000 ). 184 . See generally Weisman, supra note 178 . 185. See generally S. Nihal Singh , Manmohan's Foreign Policy Coup Almost Rivals In- dira's,THE ASIAN AGE , Nov . 13 , 2008 . 186. See generally Strobe Talbott, Good Day for India, Bad for Nonproliferation,DAILY TIMES , July 25 , 2005 . 187. See , e.g., Mark Hibbs , U.S. to FaceSome Opposition ifit Seeks Consensus NSG Rule on India , NUCLEONICS WEEK , Sept . 29 , 2005 , noted in CRS REPORT FOR CONGRESS , U.S. NUCLEAR COOPERATION WITH INDIA: ISSUES FOR CONGRESS 16 (Jan . 12, 2006 ). 188 . See China Hopes US-India Nuclear Cooperation Abides by Non-ProliferationRules, PEOPLE'S DAILY ONLINE , July 29 , 2006 , at I (quoting Chinese Foreign Ministry spokes- regulations") . 189 . See generally Guzman, supra note 18 . 190. See Daniel H.Joyner , Restructuringthe MultilateralExport Control Regime System , 9 J. CONFLICT & SEC . L. 181 , 206 - 07 ( 2004 ). 191 . See generally Joseph Farrell & Garth Saloner , Coordination Through Committees and Markets , 19 RANDJ. ECON. 235 ( 1988 ). 192 . See France and India in Nuclear Deal, BBC NEWS , Feb . 20 , 2006 , http:// Like the earlier U.S. announce- pending the outcome of U.S.-India negotiations . See id. 193 . See generally Varghese K. George , After 34 Yrs, India to Buy FirstForeignReactors, From France, HINDUSTAN TiMES, Sept. 30 , 2008 . 194. See Alan 0 . Sykes , The Economics of Public InternationalLaw 23-31 (Chicago John M. Olin Law & Econ., Working Paper No. 216 , 2004 ), availableat http://papers 564383 . 195. See generally Siddarth Srivastava, Beijing Blusters Over India's Nuclear Deal,As" A TIMES , Nov. 5 , 2005 . 196. See Stephen D. Goose & Frank Smyth , Arming Genocide in Rwanda, 73 FOREIGN AFF. 86 , 95 ( 1994 ). 197 . See Elisabeth Bumiller , The President in Europe: Alliance; PraisingAlliance, Bush Asks Europe to Work with U .S., N.Y. TIMES , June 1, 2003 , at 11 . 198. SeeJoyner, supra note 160, at 508. 199. See id. 200. See id. at 509 . 201. See generally U.S. DEPARTMENT OF STATE, THE PROLIFERATION SECURITY INITIA- TIVE ( 2006 ), available at 202. See generally U.S. DEPARTMENT OF STATE, PROLIFERATION SECURITY INITIATIVE: STATEMENT OF INTERDICTION PRINCIPLES ( 2003 ), availableat rls/fs/23764.htm. 203. See , e.g., U.S. DEPARTMENT OF STATE, SINGAPORE HOSTS PROLIFERATION SECUR- ITY INITIATIVE (PSI) INTERDICTION EXERCISE (DEEP SABRE ) ( 2005 ), availableat http:// 204. See U.S. DEPARTMENT OF STATE, THE PROLIFERATION SECURITY INITIATIVE (PSI): A RECORD OF SUCCESS ( 2005 ), available at 47715.htm. 205. See generally THE PROLIFERATION SECURITY INITIATVE , supra note 201 . 206. See , e.g., Proliferation Security Initiative Ship Boarding Agreement with Libe- ria , U.S.-Liber., Feb . 11 , 2004 , availableat 207. See generally Proliferation Security Initiative Ship Boarding Agreement with Liberia , supra note 206; The United States and Panama Proliferation Security Initiative Ship Boarding Agreement , U.S.-Pan., May 12 , 2004 , availableat r/pa/prs/ps/2004/32414.htm. 208. See generally Proliferation Security Initiative Ship Boarding Agreement with Liberia , supra note 206; The United States and Panama Proliferation Security Initiative Ship Boarding Agreement , supra note 207 . As the name suggests, the obligation is actu- Security Initiative Ship Boarding Agreement , supra note 207 . 221. See Gary M. Buechler , ConstitutionalLimits on the President'sPower to Interpret Representations , 78 GEO. L.J. 1983 , 1983 ( 1990 ). 222 . See id. 223 . See generally Carmel Davis, Power vs . Threat: Explanations of United States balancing against the Soviet Union after 1976 (Sept . 2004 ) (unpublished Ph.D. disserta- tion, University of Pennsylvania) (on file with the University of Pennsylvania) . 224 . See Harold Hongju Koh et al., Symposium , Foreign Affairs and the Constitution: The Roles of Congress, the President,and the Courts, 43 U . MIAMI L. REV. 101 , 102 ( 1988 ). 225 . See id. 226 . See id.

This is a preview of a remote PDF:

Timothy Meyer. Soft Law As Delegation, Fordham International Law Journal, 2008,