The Translation of Common Sense: A Response to Verdier and Voeten

[A] state may comply because it knows its decision to defect creates a precedent that may undermine a cooperative norm it values . . . Thus, if a state values the continued existence of the cooperative norm and believes its decision to defect will create a precedent that undermines the norm, that state may refrain from defecting despite short-term incentives to do so. But if the state does not value the rule or believes that its defection will not much affect the behavior of others, it may choose to defect and free ride on continued compliance by others.1


Alejandro Lorite Escorihuela*
In their article Precedent, Compliance, and Change in Customary International Law: An Explanatory Theory, Pierre-Hugues Verdier and Erik Voeten propose a new theory of customary international law (CIL), which can be summarized as follows: [A] state may comply because it knows its decision to defect creates a precedent that may undermine a cooperative norm it values . . . Thus, if a state values the continued existence of the cooperative norm and believes its decision to defect will create a precedent that undermines the norm, that state may refrain from defecting despite short-term incentives to do so. But if the state does not value the rule or believes that its defection will not much affect the behavior of others, it may choose to defect and free ride on continued compliance by others. 1 Their theory is valuable, the authors contend, because among other things it does not contradict "traditional" doctrine and scholarship. From this angle, Verdier and Voeten's endeavor is not immediately valuable for its contents, since we will learn nothing new, substantively speaking, about international law. Their contribution is essentially to furnish an "explanation" of the received tradition in the key of rational-choice theorizing. It is thus an exercise in the translation of normative propositions relating to customary law into explanatory and predictive propositions, which traditional theory does not produce. 2 As a translation, the endeavor is founded on a double mischaracterization. First, Verdier and Voeten assume that "traditional" international law is plausibly interested in describing the world, as opposed to implementing a normative worldview. Second, they present traditional international law as if it did not already embrace the foundations of their demonstration-which it does, but as a matter of legal ideology and disciplinary commonsense, not as fact or science.
Aligning itself with traditionalism, their theory is supposed to show how customary law is indeed a factor weighing on states' decisions, contrary to what "existing rational choice theories" 3 have been able to explain. After all, they say, even great powers take precedent seriously in making rational calculations, and, after all, it does make rational sense for states to use customary law as a mechanism for social signaling. Philip Jessup lectured the foreign policy establishment seventy-five years ago on this general idea that international law made considerable political sense. 4 International law, customary or otherwise, he suggested, is a factor in international relations: it has been invoked for hundreds of years, it has been growing for hundreds of years, and instances of its violation may be as deliberate as instances of conformity to its rules. If it is there, simply stated, it is because it serves someone's interest. Very possibly, international law serves the interest of those who make the rules, i.e. states. This may be a reduction of Jessup's complex picture of the life of international law. But it is, without much reduction, the claim made by Verdier and Voeten: states pay attention to precedent, and the process of customary law, including its formation, may therefore constrain them in one way (in the short term), while still reflecting their interest (in the long term).
In "traditional" international law, this idea has functioned as a premise to international legal argument since the Lotus. 5 What is new is that, until 15 years ago, one would not have habitually used a sequence of terms such as "equilibrium," "reputation," and "defection" to express that commonsense and, more specifically, to explain commonsense as a result of mechanical interactions. Verdier and Voeten are participating in a dialogue that would not exist without Jack Goldsmith and Eric Posner's 1999 "rational-choice" intervention in customary law matters 6 and their subsequent The Limits of International Law. 7 Now, the accepted notion is that international law, whether in its creation or its reiterated invocation, needs to be validated by the "rational" proof that it can really exist, regardless of whether it does. This requires, as Verdier and Voeten demonstrate in detail, dressing up received tradition in the linguistic ornaments of science to achieve an "explanation." Their contribution therefore helps reveal, perhaps more than anything else, how the "rational-choice" conversation has evolved over 15 years of its inner life. In its bubble, the argument presented by Verdier and Voeten's article and overall theory is clearly clever and erudite, whereas, in context, their analysis functions mostly to reinforce the core rational-choice program successfully pushed by Goldsmith and Posner, to which they have contributed a moderately internationalist wing.
Whatever the authors' intentions, however, adding an appendix that appears both internationalist and friendly to traditional doctrinalism completes the overall design of the originators of rational-choice scientism in international law. That ideological perspective includes instrumentalizing rational-choice theorizing to delegitimate normative (legal or moral) talk at the international level. As Paul Schiff Berman suggested, The Limits of International Law, an unavoidable reference for Verdier and Voeten and their main interlocutors, did not actually succeed in demonstrating the empirical limits of international law, contrary to what its authors apparently set out to achieve. 8 But that is less surprising, Berman continued, if Limits is properly understood not as a descriptive account, but as "a normative vision advancing an ideology of international relations realism." 9 And yet Verdier and Voeten deal with the whole rational-choice theorizing effort in international law in the key of scientism, in isolation from all social, cultural, scholarly, and political context. Constitutional law scholarship about "foreign affairs law," polemical contributions against international institutions, and, of course, the Office of Legal Counsel-all those contemporaneous elements in the relevant cultural landscape are apparently irrelevant to the discussion of "customary international law" initiated by Goldsmith and Posner. It is even possible to discuss the latter's "CIL" arguments of 1999 without putting them in the context of the "CIL" arguments of 1997. 10 Within that narrow argumentative frame, Verdier and Voeten seem to add nothing to rational choice but a word, "precedent". This serves to describe the spectrum of time frames corresponding to a variety of state interests and simplified into a short-term / long-term dichotomy. The addition allows also for fidelity to traditional doctrines of customary international law. Traditional doctrines, however, are premised on some normative angle on the nature of customary law, whereas accepting the premises of rational-choice theorizing means accepting that there is ultimately no distinction between why states do what they do (e.g. administrative disorganization) and what the source of law's authority really is (e.g. contribution to a world public order of human dignity). Within the literature discussed by Verdier and Voeten, the answer to both is state interest. Hence the permanent risk of tautology. Verdier and Voeten want to both maintain the integrity of traditional doctrine, which is normative, and defend an inductive fact-based notion of customary law reflecting selfinterested behavior. That makes absolutely no sense, in either one of these methodological universes.
Sharing the methodological baseline of the "new international law scholarship" 11 -especially when accepting even partly the metaphysics of political realism (which Verdier and Voeten do) -renders customary law's constraint on state will a tautology. The tautology is that since states are rational, they do not do anything against reason: if law's impact on their behavior is due to the fact that legal behavior is reasonable, then states are following their self-interest (customary law is "endogenous" to behavior) and not "law" (as if law were an "exogenous" influence). And of course, you do not have to prove the former empirically, since it is axiomatic. The only point of such a conversation is thus basically to convince Goldsmith and Posner that, despite the fact that states are rational, cooperation is possible; and, as others have put it, customary law could serve as a focal point for such cooperation. 12 But then, the distinction between endogenous and exogenous should give way. 13 And so necessarily will the distinction between what is really law and what is something else. 14 We are therefore actually performing, as has been done in law-and-economics literature before, an exegesis of the axiom of rationality.
Against that background, Verdier and Voeten's point is that behaving in a way that is contrary to (shortterm) state interest makes rational sense precisely because of states' self-interested concern over precedents. The explanatory part of the theory consists therefore of saying that it is in the interest of states to abide by, or create, norms of customary law because such behavior is in their (long-term) interest. The merging of normative and explanatory is thus achieved axiomatically: states are rational, they use customary law, so customary law is rational, which is explicable by reference to the distinction between short-term and longterm interests.
The ambivalence, or willful blindness, towards the relation between normative claims and scientificdescriptive claims leads, at the level of internal mechanics of the argument, to problems with the examples chosen or the type of conclusions that are drawn. Some examples are based on arbitrary assumptions, which makes the empirical process useless. For instance, states are presented as seeming to ignore their self-interest  12 Trachtman, supra note 8, at 73. 13 Id. at 19, 114 (first noting the logical difficulties of basing law on self-interest if we want to maintain the autonomy of both, and then suggesting that customary law may be considered endogenous to states in the aggregate and exogenous to states taken individually).
14 Guzman, supra note 8, at 192 ("The way I have defined CIL . . . blurs the line between legal rules and 'mere' norms-at least as compared to traditional definitions under which there is a sharp distinction between that which is CIL and that which is not.").
in not extending their jurisdiction to foreign states and instead granting them immunities. That it is not in "their" (short-term) interest is not demonstrated, but slipped in as an assumption alongside state selfishness. But, regardless, some people say that honoring immunities comes from the fear of retaliation, and our authors will say that retaliation and fear for one's reputation do not work consistently as rational explanations. We have thus set up a problem. And the solution is "precedent". Obviously, using "precedent" as a new factor in the explanation of customary international law is made possible by the unexplained disjunction of short-term and longterm interests, which expands the scope of rational behavior. So axiomatic assumptions really do all the work. And astonishment at the result comes only within a frame where there is a presupposed contradiction between a given state action and logic or intuition. But that contradiction is set up in the first place (states should intuitively not grant immunity if they are not crazy), and it is then solved by tautology (it is not crazy to act in such a way, because states are not crazy).
Other examples are wrong empirically. For instance, the authors suggest the case of the 2003 invasion of Iraq as an example of a superpower being cautious not to depart from settled precedent (and thus showing the influence of what we otherwise call the process of formation of customary law). The United States government, they suggest, did not use the argument of preemptive self-defense, but instead argued that Operation Iraqi Freedom was legal based on existing Security Council resolutions. 15 The example is supposed to show a great power sacrificing the benefits of short-term behavior, i.e. speaking of preemptive self-defense, in exchange for the benefit of not having other states use that argument later as a result of shifting understandings of self-defense to which the United States would have blindly contributed. But we know that the United States government was of several minds about the use of arguments of preemptive, or anticipatory, or preventive, self-defense, as well as all other legitimating grounds for the projected invasion (the OLC certainly tried to sell all of them privately), 16 and its most notorious employees continued even after changing jobs. 17 Moreover, if, as Verdier and Voeten note, the argument of past Security Council resolutions prevailed in the choice of public arguments, it is certainly not because of the assumed fear of consolidating a rule about preemptive self-defense. That rule was then, and has continuously remained, part of the official national security strategy of the United States (NSSUS): as a version of the so-called "Bush doctrine" as early as June 2002, 18 and then in the official NSSUS (2002) 19 (not to mention the OLC again), 20 before the actual announcement of the invasion of Iraq by President Bush. 21 It remained in the NSSUS (2006), 22 and still is now in some less ostentatious form as part of the Defense Strategic Guidance of 2010. 23 The invasion of Iraq is therefore arguably the worst example that Verdier and Voeten could have used.
What, then, do we do with that odd example? I am not sure. Since we don't exactly know whose expression of interest counts as the U.S. interest, and we are not sure what the precedent that is being avoided really is, we can make that example tell anything we want. What I am sure of, however, is that, at that time, people like John Yoo 24 or Eric Posner, 25 both separately and jointly adamant pushers of rational-choice theory, 26 deployed arguments in favor of expanding Article 51 to preemptive self-defense in a way that would cover Iraq in positivist, as well as rational-choice, terms. That type of elements of information is, I think, what should matter in the story of rational-choice militancy in the past 15 years, and it could give some context and purpose to that particular example. It amounts to willful blindness that the political context (which is the only way of making sense of the original rational-choice's systematic legal xenophobia towards international law) is invisible even in discussing something like the use of force. But, before all else, it causes examples such as Iraq to make no sense at all. The author's treatment of the Kosovo crisis and the doctrine of humanitarian intervention 27 displays exactly the same empirical and theoretical problems.
But my sensation of missing the point completely, because of the sense that the authors are continuously reinventing the wheel, is mostly reinforced by their central example of domestic adjudication of foreign immunities. I quote their analysis at length to ensure that truncating the argument is not the main reason for my losing the plot: When the German Supreme Court is considering whether to adopt restrictive immunity, it cannot precisely weigh all the costs and benefits described above, but it can do what CIL doctrine prescribes: conduct a survey of state practice and opinio juris. If it finds that many states have already adopted restrictive immunity, the case for doing so will be stronger. In conducting its survey, the court will place greater emphasis on the practice and opinio juris of certain states. Under traditional doctrine, the "most affected states" (…) have greater weight. Not coincidentally, these states are the ones who decisions are most important to the incentives of the remaining states. The court will also consider whether other states apply the rule consistently and universally, thus expressing an understanding that it is CIL, or only on a reciprocal basis, which might indicate they are playing a different game that may call for a different reaction. In a word, our theory suggests that traditional CIL doctrine serves as a tool for actors involved in CIL determinations to assess prevailing practice among other states, which is relevant to their own decisions. 28 The theory adds exactly nothing, whether in terms of description or explanation, to what dry formalist doctrinal accounts of customary law have told us. Whether states are rational or crazy, traditional customary international law doctrine already tells us that we have to imagine them as free actors, i.e. sovereigns: each state establishes for itself its legal situation vis-à-vis other states, as the formula goes. 29 The theory about the formation of customary law is based on states being as free as the Lotus court told us that they are. And, further, to know whether there is a particular rule (or regularity) concerning naval warfare out there, the practice and opinion of major naval powers ("specially affected states") is essential, 30 just as we pay special attention to nuclear states in matters of regulating (or depicting state practice around) nuclear weapons. 31  in short, yes, the theory that Verdier and Voeten propose can accommodate traditional doctrine, because it both describes things in exactly the same way and gives an explanation that is also exactly the same as the one necessarily implied, or actually proffered, by traditional doctrine.
The process of exoticizing disciplinary commonsense rises to more extreme levels still. The authors examine the "two elements" theory of customary law and note that there is uncertainty about the exact contours of both elements: how many states, how long a period, how much homogeneity, and so on. Moreover, Verdier and Voeten continue, states seem to assess those elements opportunistically depending on the context, thereby making it impossible to come up with objective standards for the formation of customary law. Their theory, however, will solve and explain all this: [T]he desirability of upholding that rule depends on a state's expectation that others, too, will uphold it in the future. Therefore, the more that foreign state practice supports such an expectation, the more relevant it will be. This evaluation is necessarily influenced by context and cannot easily be reduced to mechanical rules, but it is nevertheless crucial to the enterprise. 32 What is really added here is that the pursuit of self-interest is in all cases context-specific. Here again, the theory does not seem to have helped in any way, to the point that I am beginning to forget what the theory was to begin with. But they continue: Thus, while traditional legal doctrine appears inconsistent in simultaneously holding that state practice is a central element of CIL formation but also that the required type, volume, duration, and consistency of practice are a matter of appreciation in every case, this situation reflects an inherent element of the process. 33 I fail to see any inherent logical inconsistency in traditional doctrine holding that the assessment of necessary practice and opinio juris is context-, even rule-, specific. What is wrong with positivist formalism is not rigidity, but rather concealed flexibility. Yet Verdier and Voeten's theory here yields repeatedly as a result that the position of states will be context-specific-as if traditional doctrine rejected that notion (which should be obviously put in question by the very notion of "affected states"). In other words, their theory does not produce anything beyond contesting traditional doctrine on views that it does not hold, while proposing instead notions upon which traditional doctrine is actually premised. This, however, is concealed by socialscience lingo: "the various principles articulated by courts and scholars . . . are heuristics that identify situations in which expectations of continued adherence are likely to arise among states, but they likely cannot be articulated more precisely. 34 Somehow a lot of the debate within the rational-choice literature can proceed inside the bubble of factchecking and bean-counting, while dismissing the notion that we are talking about arguments and persuasion. Customary international law and its use by states has never been about anthropomorphized belief or false consciousness, it has always been about signaling-because it has always been about paying attention to the stuff of legitimate arguments. Discovering it again and again as a matter of science is, therefore, becoming a little suspicious.
So, to go back to the beginning, one has to wonder about the meaning of an intensive scholarly enterprise such as that pursued by Verdier and Voeten, so much of which is dedicated to drowning out the diplomatic reality out there, in which international law is an argumentative part of political life. It is obviously appropriate to miss the good old days of the New Haven School, when social science was more transparently taken as part of an actual social process, and legal scholarship as part of a grander cultural process with political stakes. Today, apparently, it is plausible to enter rational-choice theorizing as an ahistorical laboratory in order to prove scientifically that international law exists. What for? To convince those who deny its impact on state behavior in rationalist terms, yet still discuss it in the most traditionalist of ways when needed, for instance to legitimate the forced disappearance of civilians. 35 I would say that rational-choice theorizing and everything that depends upon it in international law is, as demonstrated in the methodological casuistry of customary international law, bizarrely blind to the fact that the use of rational-choice theory itself must be the result of rational choice-for instance, the result of an overarching ideological commitment. The fact that no normative lesson is drawn in Verdier and Voeten's sparse conclusion is therefore the most symptomatic and depressing fact about the nature of their contribution.