CHICAGO WORKING PAPER IN LAW AND ECONOMICS 10 the prior regularity that they were bound to follow as a matter of law More broadly, the traditional account does not explain why CiL changes track the interests of powerful nations, or why technological changes and other exogenous factors often produce significant hanges in the content of cil The traditional account also cannot explain several pervasive features of the way nations perceive and use CIL. It cannot account for the fact that nations frequently change their views about the content of CIL, often during very short periods of time. 28 Nor relatedly, can it explain why courts and politicians almost always apply a conception of Cil that is in the nations best interest. 29It also does not explain why nations often say that they will abide by a particular norm of CIL, and then violate their promises. 30 Finally, the traditional account does not explain why nations comply with CiL. Why would a nation ever comply with CIL when it is not in its interest to do so? the traditional account assumes that this is what nations do when they appear to act in accordance with CIL. But as noted above opinio juris begs the question of why nations feel obliged to obey CIL. Moreover, the traditional theory does not explain why, if nations obey CIL from a sense of lega obligation, they ever violate CIL There are numerous more general theories about why nations obey international law. 31 The large majority of these theories focus exclusively on, and have relevance only for, treaties. 32 But some 28 29 30 31 For a comprehensive survey, see Harold Hongju Koh, Why Do Nations Obey International Law, 106 Yale L.J. 2608(1997); see also Oscar Schacter, Towards Theory of International Obligation, 8 Va J. Int'l L. 300, 301(1968)(earlier survey noting"In]o single theory [of international obligation] has received general reement and sometimes it seems as though there are as many theories or at least formulations as there are scholars"). Most of the literature canvassed in Koh supra, limits itself to treaty compliance 32 To take a prominent recent example, the Chayes'management theory only purports to account for international regulatory regimes established by treaty. See Abraham Chayes and Antonia Chayes, The New Sovereignty: Compliance with International Regulatory Agreements (1995). In addition, all of the recent
CHICAGO WORKING PAPER IN LAW AND ECONOMICS 10 the prior regularity that they were bound to follow as a matter of law. More broadly, the traditional account does not explain why CIL changes track the interests of powerful nations, or why technological changes and other exogenous factors often produce significant changes in the content of CIL. The traditional account also cannot explain several pervasive features of the way nations perceive and use CIL. It cannot account for the fact that nations frequently change their views about the content of CIL, often during very short periods of time.28 Nor, relatedly, can it explain why courts and politicians almost always apply a conception of CIL that is in the nation’s best interest.29 It also does not explain why nations often say that they will abide by a particular norm of CIL, and then violate their promises.30 Finally, the traditional account does not explain why nations comply with CIL. Why would a nation ever comply with CIL when it is not in its interest to do so? The traditional account assumes that this is what nations do when they appear to act in accordance with CIL. But as noted above, opinio juris begs the question of why nations feel obliged to obey CIL. Moreover, the traditional theory does not explain why, if nations obey CIL from a sense of legal obligation, they ever violate CIL. There are numerous more general theories about why nations obey international law.31 The large majority of these theories focus exclusively on, and have relevance only for, treaties.32 But some 28 For examples, see infra __. 29 For examples, see infra __. 30 For examples, see infra __. 31 For a comprehensive survey, see Harold Hongju Koh, Why Do Nations Obey International Law?, 106 Yale L. J. 2608 (1997); see also Oscar Schacter, Towards a Theory of International Obligation , 8 Va. J. Int’l L. 300, 301 (1968) (earlier survey noting “[n]o single theory [of international obligation] has received general agreement and sometimes it seems as though there are as many theories or at least formulations as there are scholars”). Most of the literature canvassed in Koh, supra, limits itself to treaty compliance. 32 To take a prominent recent example, the Chayes’ management theory only purports to account for international regulatory regimes established by treaty. See Abraham Chayes and Antonia Chayes, The New Sovereignty: Compliance with International Regulatory Agreements (1995). In addition, all of the recent
11 CUSTOMARY INTERNATIONAL LAW purport to apply to treaties and CIl alike. Some positivist theorists argue that nations obey international law-including CIL-because they consent to it. 33 But as many have noted, this position begs the question of why nations abide by the international rules to which they have consented prominent theory in the natural lay tradition contends that nations abide by cil because"they perceive the rule and its institutional penumbra to have a high degree of legitimacy, where legitimacy is understood as"a property of a rule or rule-making institution which itself exerts a pull toward compliance on those addressed normatively because those addressed believe that the rule or institution has come into being and operates in accordance with generally accepted principles or right process. "35 Another theory argues that"repeated compliance [with international law] becomes habitual obedience"as international law "penetrates into a domestic legal system, thus becoming part of that nations internal value set. 36 Yet another prominent theory purports to begin from the more rationalistic premise that nations "observe international obligations unless violation promises an important balance of advantage over cost, "but ultimately explaining international compliance on the basis of morality and the "habit and inertia of continued compliance. "37"Right process, " value set," habit, "and"morality "are empty phrases in these theories. the stand in for the concept of opinio juris without explaining what it means empirical work on compliance with international law has focused on treaties rather than CIL. See harold hongju Koh, Why do Nations Obey International Law? 106 Yale L. J. 2599, 2599 n. 2(1997). For our account of why treaties might foster international cooperation more successfully than CIL, see infra 33 J. Brierly, The Law of Nations, 53-56(6th ed. 1963) 34 35 Thomas Franck, Fairness in International Law and Institutions 24-25(1995) 36 Harold Hongju Koh, Why Do Nations Obey International Law?, 106 Yale L. .2599,2603(197) 37 Louis Henkin, How Nations Behave 49, 58-63(2d ed 1979). The rational choice strand of international relations attempts to explain international cooperation without falling back on notions of morality or opinio juris See sources ited supra note__. As we explained in note_, this is the tradition we ar working in, a tradition that has not to date been invoked to account for CIL
11 CUSTOMARY INTERNATIONAL LAW purport to apply to treaties and CIL alike. Some positivist theorists argue that nations obey international law—including CIL—because they consent to it.33 But as many have noted, this position begs the question of why nations abide by the international rules to which they have consented.34 A prominent theory in the natural law tradition contends that nations abide by CIL because “they perceive the rule and its institutional penumbra to have a high degree of legitimacy,” where legitimacy is understood as “a property of a rule or rule-making institution which itself exerts a pull toward compliance on those addressed normatively because those addressed believe that the rule or institution has come into being and operates in accordance with generally accepted principles or right process.”35 Another theory argues that “repeated compliance [with international law] becomes habitual obedience” as international law “penetrates into a domestic legal system, thus becoming part of that nation’s internal value set.”36 Yet another prominent theory purports to begin from the more rationalistic premise that nations “observe international obligations unless violation promises an important balance of advantage over cost,” but ultimately explaining international compliance on the basis of morality and the “habit and inertia of continued compliance.”37 “Right process,” “value set,” “habit,” and “morality” are empty phrases in these theories. They stand in for the concept of opinio juris without explaining what it means. empirical work on compliance with international law has focused on treaties rather than CIL. See Harold Hongju Koh, Why Do Nations Obey International Law?, 106 Yale L. J. 2599, 2599 n. 2 (1997). For our account of why treaties might foster international cooperation more successfully than CIL, see infra __. 33 See J. Brierly, The Law of Nations, 53-56 (6th ed. 1963). 34 See id. 35 Thomas Franck, Fairness in International Law and Institutions 24-25 (1995). 36 Harold Hongju Koh, Why Do Nations Obey International Law?, 106 Yale L. J. 2599, 2603 (1997). 37 Louis Henkin, How Nations Behave 49, 58-63 (2d ed 1979). The rational choice strand of international relations attempts to explain international cooperation without falling back on notions of morality or opinio juris. See sources cited supra note __. As we explained in note __, this is the tradition we are working in, a tradition that has not to date been invoked to account for CIL
CHICAGO WORKING PAPER IN LAW AND ECONOMICS There are scores of other theories of international law compliance. 38 This is not the place to catalog these theories'many shortcomings, at least as applied to cIl. most of them suffer from the difficulties described above. They provide accounts of international behavioral regularities at very high levels of generality they view international law as an exogenous influence on international behavior rather than explaining it or why nations obey it; they do not account for when or why nations violate the international law and they do not explain how international law originates or changes II. A Revisionist Theory This Section sets forth our revisionist theory of CIL. The theory uses simple game theoretical concepts to explain international behavioral regularities as a function of nations pursuing self interest. 39 We argue that nations pursuing self-interest produce behavioral regularities in four strategic situations: coincidence of interest, coercion, the bilateral iterated prisoners dilemma, and coordination. All of the international behaviors subsumed by the label CIL are variations on one of these four strategic forms. In contrast to the traditional understanding of CIL, the theory rejects the notion that international behavioral regularities result from compliance with a norm that a nation feels legally obliged to follow. It claims that the direction of causality is the reverse. It is not the case that an exogenous, reified entity known as CIL causes nations to act in certain ways; rather, CiL is the label people attach to behavioral egularities that arise endogenously from the interaction of nations pursuing their self-interest 38 Schacter, supra note, at 301, lists thirteen theories. Koh, supra note 39 Discussions of the game theoretic concepts we use can be found in standard G. Baird, Robert H. Gertner, and Randal C. Picker, Game Theory and the Law (1994), which shows how game theory can be used to understand law(but not international law), and James D. Morrow, Game Theory for Political Scientists (1994), which shows how game theory can be used to understand international relations. Another useful reference is Oye, supra note
CHICAGO WORKING PAPER IN LAW AND ECONOMICS 12 There are scores of other theories of international law compliance.38 This is not the place to catalog these theories’ many shortcomings, at least as applied to CIL. Most of them suffer from the difficulties described above. They provide accounts of international behavioral regularities at very high levels of generality; they view international law as an exogenous influence on international behavior rather than explaining it or why nations obey it; they do not account for when or why nations violate the international law; and they do not explain how international law originates or changes. II. A Revisionist Theory This Section sets forth our revisionist theory of CIL. The theory uses simple game theoretical concepts to explain international behavioral regularities as a function of nations pursuing selfinterest.39 We argue that nations pursuing self-interest produce behavioral regularities in four strategic situations: coincidence of interest, coercion, the bilateral iterated prisoner’s dilemma, and coordination. All of the international behaviors subsumed by the label CIL are variations on one of these four strategic forms. In contrast to the traditional understanding of CIL, the theory rejects the notion that international behavioral regularities result from compliance with a norm that a nation feels legally obliged to follow. It claims that the direction of causality is the reverse. It is not the case that an exogenous, reified entity known as CIL causes nations to act in certain ways; rather, CIL is the label people attach to behavioral regularities that arise endogenously from the interaction of nations pursuing their self-interest. 38 Schacter, supra note __, at 301, lists thirteen theories. Koh, supra note __, canvasses dozens. 39 Discussions of the game theoretic concepts we use can be found in standard game theory textbooks. Two particularly lucid and relevant treatments are Douglas G. Baird, Robert H. Gertner, and Randal C. Picker, Game Theory and the Law (1994), which shows how game theory can be used to understand law (but not international law), and James D. Morrow, Game Theory for Political Scientists (1994), which shows how game theory can be used to understand international relations. Another useful reference is Oye, supra note __
CUSTOMARY INTERNATIONAL LAW a word of caution is in order at the outset concerning ur use o concepts from game theory. We use these concepts to organize our ideas and intuitions and to clarify the assumptions made by us and those we criticize. We do not claim that the axioms of game theory accurately represent the decision-making process of a"state"in all its complexity. 40 Because the premises of the theory are relatively crude the theory's predictions lack nuance and subtlety. But a theory is successful if it provides a more coherent and plausible account of behavior than rival theories do. and if it allows one to see old problems in new and fruitful ways. 4I The success of our argument, then, depends on both its theoretical plausibility(the subject of this Section) and its empirical implications(the subject of the next Section A. The basic model What courts and scholars call cil refers to certain behavioral regularities that emerge in international games played among states. In this Section we describe the four strategic positions that we believe capture the behavioral regularities thought to constitute CIL. For expository clarity, we initially discuss interactions between two states: then we discuss the extent to which the conclusions of this discussion can be extended to interactions among more than two states. We then explain how the basic model differs from the traditional conception of Cil I Coincidence of interest The first position is one of coincidence of interest, where states engage in behavioral regularities simply because each obtains private advantages from the same action irrespective of the action of the other Table 1 illustrates such a situation Table 1 might describe the position of two belligerent states nat have navies that patrol a body of water also used by civilian fishing boats from both states. A state's naval vessels are expensive to 40 We explore this issue further in SectionIVA Al For a discussion of the advantages and limitations of using game theory to analyze international relations, see Duncan Snidal, The Game Theory of nternational Politics, in Oye, supra note_ and for more general critical David M. Kreps, Game Theory and Economic Modelling ch 5 (1990
13 CUSTOMARY INTERNATIONAL LAW A word of caution is in order at the outset concerning our use of concepts from game theory. We use these concepts to organize our ideas and intuitions and to clarify the assumptions made by us and those we criticize. We do not claim that the axioms of game theory accurately represent the decision-making process of a “state” in all its complexity.40 Because the premises of the theory are relatively crude, the theory’s predictions lack nuance and subtlety. But a theory is successful if it provides a more coherent and plausible account of behavior than rival theories do, and if it allows one to see old problems in new and fruitful ways.41 The success of our argument, then, depends on both its theoretical plausibility (the subject of this Section) and its empirical implications (the subject of the next Section). A. The Basic Model What courts and scholars call CIL refers to certain behavioral regularities that emerge in international games played among states. In this Section we describe the four strategic positions that we believe capture the behavioral regularities thought to constitute CIL. For expository clarity, we initially discuss interactions between two states; then, we discuss the extent to which the conclusions of this discussion can be extended to interactions among more than two states. We then explain how the basic model differs from the traditional conception of CIL. 1. Coincidence of Interest The first position is one of coincidence of interest, where states engage in behavioral regularities simply because each obtains private advantages from the same action irrespective of the action of the other. Table 1 illustrates such a situation. Table 1 might describe the position of two belligerent states that have navies that patrol a body of water also used by civilian fishing boats from both states. A state’s naval vessels are expensive to 40 We explore this issue further in Section IVA. 41 For a discussion of the advantages and limitations of using game theory to analyze international relations, see Duncan Snidal, The Game Theory of International Politics, in Oye, supra note __; and for more general critical comments, see David M. Kreps, Game Theory and Economic Modelling ch. 5 (1990)
CHICAGO WORKING PAPER IN LAW AND ECONOMICS operate and have important uses(such as protecting the state from invasion), and the fishing boats are not worth very much. Payoffs in Table 1 are based on the assumptions that a state enjoys a payoff of 3 if it neither attacks the boats of the other state nor is subject to an attack; the state incurs a cost of l in order to seize the fishing boats of the other state; and the state loses 1 if its fishing boats are attacked Table 1 attack ignore attack 2.-2 1,2 nore 2.-1 3.3 To determine the equilibrium of the game, assume first that one player ("state j)attacks the boats of the other player(state i"). State i obtains a higher payoff(2)if its navy does not attack the fishing boats of state j, than it obtains if it does attack and seize these boats (-2). Now assume that statej does not attack the boats of state i State i obtains a higher payoff (3)if it ignores than if it attacks (-1) Accordingly, state i ignores state j's boats regardless of state js behavior. Because state js payoffs are the same as state is, state ignores state is boats as well. Thus, in equilibrium each state ignores the boats of the other state. By anequilibrium, we mean that the states will continue engaging in this behavior as long as payoffs do not change Thus, when an equilibrium occurs, one would observe a behavior regularity-in this case, a behavioral regularity consisting of each state ignoring the boats of the other This behavioral regularity is one possible explanation for what is referred to as CIL. 42 Notice that in equilibrium the states act 42 Cf. Kenneth A Oye, Explaining Cooperation under Anarchy: Hypotheses and Strategies 6, in Oye, supra note_. Oye argues that states often obtain mutual gains by acting independently, and refers to the classical liberal defense of free trade, according to which every state does best if it eliminates tariffs regardless of whether other states do. The example is slightly misleading, because gains exist only against the implicit baseline of protectionism. Our example below, of states not sinking their own ships is formally identical, except self-interest leads te
CHICAGO WORKING PAPER IN LAW AND ECONOMICS 14 operate and have important uses (such as protecting the state from invasion), and the fishing boats are not worth very much. Payoffs in Table 1 are based on the assumptions that a state enjoys a payoff of 3 if it neither attacks the boats of the other state nor is subject to an attack; the state incurs a cost of 1 in order to seize the fishing boats of the other state; and the state loses 1 if its fishing boats are attacked. Table 1 attack ignore attack -2, -2 -1, 2 ignore 2, -1 3, 3 To determine the equilibrium of the game, assume first that one player (“state j”) attacks the boats of the other player (“state i”). State i obtains a higher payoff (2) if its navy does not attack the fishing boats of state j, than it obtains if it does attack and seize these boats (-2). Now assume that state j does not attack the boats of state i. State i obtains a higher payoff (3) if it ignores than if it attacks (-1). Accordingly, state i ignores state j’s boats regardless of state j’s behavior. Because state j’s payoffs are the same as state i’s, state j ignores state i’s boats as well. Thus, in equilibrium each state ignores the boats of the other state. By an “equilibrium,” we mean that the states will continue engaging in this behavior as long as payoffs do not change. Thus, when an equilibrium occurs, one would observe a behavior regularity—in this case, a behavioral regularity consisting of each state ignoring the boats of the other. This behavioral regularity is one possible explanation for what is referred to as CIL.42 Notice that in equilibrium the states act 42 Cf. Kenneth A. Oye, Explaining Cooperation under Anarchy: Hypotheses and Strategies 6, in Oye, supra note __. Oye argues that states often obtain mutual gains by acting independently, and refers to the classical liberal defense of free trade, according to which every state does best if it eliminates tariffs regardless of whether other states do. The example is slightly misleading, because gains exist only against the implicit baseline of protectionism. Our example, below, of states not sinking their own ships is formally identical, except self-interest leads to