Being the second part of my reply to Eric Posner …
From reading a more recent post that he has written, Eric seems to be misinterpreting my criticisms of his account as a criticism of rational choice approaches tout court, which they are not. I have no problem either with rational choice or with power based accounts of politics, both of which I have used extensively in my own work. I’ll respond in greater detail to the rest of Eric’s most recent reply at the end of this post, but first of all want to underline what I do think is wrong with his account. It is not that it is rational choice – but that it is a form of rational choice that is too vague and general to tell us very much about international politics or international law.
The fundamental question that I think Eric’s book leaves begging is straightforward – what constitutes state interests? Eric has a lot of fun poking holes in global legalists’ accounts of the importance of sub-state actors and claims that states are best considered as unitary actors. However he himself never comes up with a satisfactory alternative account of what states’ interests are. The two positive statements that I can find (perhaps there are more, or a more developed account in Eric’s work elsewhere):
Begin with a theory that states act rationally in their national interest, which may be understood as a combination of security, wealth and similar goods (p. 44)
In order to define a state’s interest for the purpose of analyzing international behavior, then, one must rely on a theory, even if a very-rough-and-ready theory, about how the interests of citizens are translated by the political process into government policy. One might assume, for example, that democratic governments roughly express the preferences of the public at large, while authoritarian governments express the preferences of an elite. Whatever the case, it seems sensible to assume that trade law reflects state interests in advancing the prosperity of exporters and import-competers, human rights law reflects people’s altruism, the law of the sea reflects merchant and other commercial interests, and so forth. In general, states seek to maximize wealth and security of their people (or elites), and this general policy manifests itself in particular trade, human rights, security, and other foreign policies. (p. 78)
This account of state interests permits nearly everything. It poses no constraints either on the fashion in which states seek to achieve their interests, or on those interests themselves. First, consider Eric’s claim that states ‘act rationally in their national interest.’ This sounds nearly as straightforward as, say, the claim that firms act rationally to maximize their corporate profit (rational actor models of state behaviour are indeed, more often than not, tarted up economic models of firm behaviour – the main difference is the descriptive text surrounding the algebra). And indeed it is nearly as straightforward as the claim that firms act rationally to maximize their corporate profit. The problem is that this claim is not straightforward at all. To quote Daniel Davies
if you allow strategic and reputational issues to be given weight in managerial decisions, then it is very hard indeed to think of something that can’t be justified as being in the best interests of maximising shareholder value over the long term. Paying above-market wages? Efficiency wage argument, maximises shareholder value. Donating to charity? Part of the marketing budget, don’tcha know. Voluntarily refusing to sell violent video games to children? Forestalls the danger of much more punitive government regulation down the line. Etc etc.
As Daniel notes more pungently in an earlier post on state behaviour, this springs from one of the basic dilemmas of game theoretic accounts of politics – that in decently interesting models that take account of iterated relationships, there are too many damn equilibria that can be achieved:
The Folk Theorem in game theory states that any outcome of a repeated game can be sustained as an equilibrium if the minimax condition for both players is satisfied. In plain language, it can be summarised as stating that “if we take strategic considerations into account, there is a game-theoretic rationale for practically anything”. This formulation leads on to my contribution, the Davies-Folk Theorem, which states that “if we take strategic considerations into account,there is a game-theoretic rationale for practically fucking anything” (it’s a fairly simple corollary; proof available from author on request).
Rational pursuit of self-interest in indefinitely iterated relationships can cover a multitude of virtues, including, I suspect, most of the kinds of international cooperation that global legalists would like to see happening. Respect for human rights (perfectly rational if you suspect you will get punished for abusing em) – check. Obeying international rulings that substantially inconvenience you (quite rational, as long as you believe that you will benefit over the long run from being perceived as a good international citizen) – check. And so on – I don’t believe it is necessary to bore the reader with repeated demonstrations of the obvious. Once you allow that reputation and similar forces can help explain rational behaviour, you admit that nearly everything can be described as being rational.
One way to at least mitigate this problem a bit is to suggest that states only care about one thing. If they care only about security, as hardcore realists sometimes suggest, then it could be that many forms of cooperation are implausible (because they may help the other state more than they help you, and the other state can turn this into military advantage which may hurt you over the long run. However, Eric seems to disallow this. In his argument, sometimes state interests revolve around security. Sometimes they turn on economic prosperity. Sometimes they turn on altruism towards others. This means that it is very hard to see what (apart from an utterly unrealistic cosmopolitanism) might be excluded from state interests. And, by extension, it means that it is hard to see what could be excluded from a theory of international law that is based on state interests.
This leads to a degree of argumentative slippage that I find somewhat dissatisfying. As I read it, the book veers between two quite different argumentative goals, neither of which ends up being well served in the end. Sometimes, Eric seems to be interested in arguing that international law, as a discipline, should be remade along more rigorous lines, hewing more closely to the empirical results of political science and rational choice theory. While these two goals strike me as not entirely coherent (there is an accumulating body of empirical work out there suggesting that rational choice doesn’t explain everything at all in international politics), it seems to me to be a perfectly respectable, and indeed laudable project. Rational choice and power relations explain a lot as long as you are careful in defining your theoretical priors (viz., not coming up with circular definitions of power etc). I have no doubt that international law scholars could learn a lot from both this evidence and these theories. I understand that some international law scholars are upset at this style of argument, feeling that it has negative normative consequences (if international law is good, we should not be undermining it). I don’t find this particular style of criticism compelling at all – if Eric is right on the substance, he is right, and if he is wrong, he is wrong. He has no duty as a scholar to shut up and not criticize arrangements that others find normatively attractive.
Yet it is also quite clear that Eric would like to do more than to remake the discipline of international law on better empirical and theoretical foundations without any regard for its innate content. He persistently argues that international law has a modest role in international relations. Throughout the book, he is at pains to argue against accounts (including accounts with a fair degree of empirical evidence to support them) which suggest that international law can have generalizable and substantial consequences for the behaviour of powerful states. Some scholars (by no means an overwhelming majority as best as I can tell) would agree with him. But they do so on far narrower accounts of state preferences and state strategic incentives than Eric is willing to argue on, and are consequently more vulnerable to reasoned criticism if others can show that preferences are not as these scholars have claimed, or that states face different strategic problems than those that they have identified.
Eric (and this is really where I think that the book has a serious flaw), argues in contrast for a vague and broad notion of what preferences are, and a nearly completely unspecified account of how states seek to act upon them (again – perhaps he speaks to this more in work elsewhere that I am unaware of). The result is that nearly any result can be assimilated to his theory. If your basic claim is that powerful states do what powerful states want to do (without telling us much about what states want to do in the first place), then it is very easy to assimilate any result to ‘prove’ the correctness of your theory. States invariably end up doing something – therefore you can argue that they have done what they wanted to do in the first place (this is the revealed preferences problem – in spades). By shoving everything that is potentially interesting about international behaviour into the black box of state preferences, the theory neatly avoids pretty well any possible falsification, and also (which is worse, in my book) avoids saying anything that is very useful. What is really interesting about international relations are the processes through which preferences can be changed (which Eric seems largely uninterested in discussing), and the processes of strategic construction through which the options available to actors may change over time. For example: the European ‘preference’ for institutions that are partially binding and involve some independent systems of accountability, would, if it were realized among, say, the body of rich democratic states, result in a very different kind of international law governing these states than the kind we have today. Eric may find this to be an implausible outcome – but his theory seems to me too vague and general to provide much in the way of an argument as to why it is implausible.
So – what I would have liked in an ideal world would have been two separate books. The first would have been a general account of why international legal scholars need to pay attention to debates over empirical evidence and rational choice theory (even more ideally, it would canvas other theories too – but that may be asking a bit much). The second – and quite different one – would have taken a particular stance within these theoretical and empirical debates, with a much tighter and more clearly specified framework (a formal model or two would be nice), and use of evidence to test this theory and expose both its strengths and its weaknesses. The current book reads to me as a somewhat unhappy amalgam of the two, with the result that it neither provides a good general foundation to reformulate the study of international law, nor yet good evidence and theory to convince us that international law is tightly constrained in the ways that Eric believes that it is. Indeed, the two goals are mutually undercutting. I share with Eric the belief that the study of international law would be a lot better if it took proper account of the body of empirical work that is out there as well as the powerful role of self-interest in explaining behaviour. I am also skeptical of some at least of the arguments of international law scholars about its delicious awesomeness. But I do not think that a general account of the former, if properly done, really provides a specific critique of the latter – for that you need a very different, and much more tightly specified, set of arguments.
Now, onto his most recent post. Eric complains that:
Henry reads too much into the papers. The papers do not show that the Commission caused any countries to improve their human rights practices.
But as the quotes Eric provides show, I never made the claim that Eric appears to be imputing to me. I merely argued that the Commission influenced sanctions, and appeared to be far more responsive to actual human rights concerns than Eric has acknowledged in the past (for example, he briefly complains in the book that “the UN Commission on Human Rights … was dominated by human-rights abusing states and was heavily criticized for focusing on Israel and ignoring the rest of the world. The Human Rights Council has, so far, no better a record.”). Jim and Erik’s articles suggest that the UNCHR – which would appear on its face to present about as unpromising a venue for actual normative respect for human rights as one could imagine – actually showed growing respect for real human rights issues, and appeared to have material consequences for states with poor human rights records. These are findings that have substantial consequences for Eric’s earlier claims. The question of compliance is a far muddier one, and one that I specifically avoided pronouncing upon (my broad understanding is that as best as we can tell, given measurement issues etc, international human rights laws do have substantial consequences for compliance, contra Eric, but that these consequences work through more complicated mechanisms than those that many human rights lawyers would identify).1 More generally, Eric seems to provide two possible explanations of the anomaly that this presents for his particular arguments. First – that the UNCHR was a failure:
Still, the UNCHR was ultimately a failure. That is why states disbanded it and replaced it with the Council. It may well be the case that it targeted abusers more than non-abusers, as Lebovic and Voeten document, and perhaps even put more pressure on abusers than non-abusers. But this was not good enough. Imagine a criminal justice system that is more likely to punish actual criminals than non-criminals, but still punishes some (relative) innocents and lets most criminals get away. Such a system will enjoy little support; even if the sanctions actually sting, they would still not necessarily improve behavior (because the states that improve their human rights records at great cost to themselves must contend with the possibility that they will continue to be criticized).
Second – that the UNCHR was too much of a success.
An even more alarming possibility is that the Commission was disbanded not because it failed but because it succeeded too well. Tired of being criticized by their peers, members of the Council have instituted a number of review mechanisms (universal periodic review that prohibits fact-finding and group judgments, “thematic” review that emphasizes general problems rather than particular governments) that ensure that particular states will rarely be explicitly criticized by the Council as a body, which is what has happened. These activities clog the agenda, leaving no time to discuss Sudan, Egypt, or North Korea. The free rider problem and continuing significant disagreement about which rights are human rights and what can justify violating them have doomed the enterprise.
These possible explanations obviously cannot both be true at the same time. Arguing in the alternative may perhaps be an accepted approach in the legal academy (I am not, obviously, a legal scholar), but it doesn’t seem to me to be a good way of dismissing the merits of empirical claims (it can, of course, be useful, at the earlier stage of hypothesis formulation).
Update: Eric points out correctly via email that I mischaracterized Erik and Jim’s article in one important respect. I say that they argue that the UNCHR seems to have had “material consequences for states with poor human rights records.” Erik and Jim find no evidence that a country’s human rights record directly affects its aid receipts – instead, they argue that there is a relationship between UNCHR resolutions and multilateral aid receipts. This is not of course the same thing – and to suggest that it was reflects sloppy wording on my part.
Eric also replies here – since the debate seems to be running out of steam (he titles his post ‘part n’), I will keep my own reply short. Briefly, he suggests that the major problem that others have found with his (and Jack Goldsmith’s work) is not that it is falsifiable, but that it is false. Fair enough, up to a point – but my reading of his work is that he himself clearly doesn’t consider the more cogent of these criticisms to be valid tests of the falsifiability of his argument. For example, he considers the enforcement of rulings within the European Union to “have limited relevance to international law” because Europe “has been undergoing a process of unification for several decades” (p.49). Furthermore – as I have noted already – the relationship between his particular claims (about the stark limits of international law) and his underlying theory (about states as rational actors with any one of a number of possible preferences in a given context) is somewhat obscure. The place where he perhaps comes closest to making a strong claim that rational choice precludes the kinds of cooperation foreseen by international law is in his early discussion of collective action theory – but here he makes much stronger claims about the consequences of collective action theory than do the theorists of collective action with whom I am familiar.2
1 I do note that Eric’s claim that “even significant sanctions, far worse than the refusal to extend loans or aid, rarely cause states to change their behavior” seems to run in the face, say, of Emilie Hafner Burton’s work on the human rights consequences of preferential trade arrangements.
2 For example, Eric claims that “A state can be bound by a treaty only if it consents to it; thus a treaty that will solve a global collective action problem requires the consent of all states, or all states that contribute to that problem” (p. 29). This may simply be careless wording on Eric’s part, but the bit after “thus” is not supported by collective action theory as I understand it. There is nothing in collective action theory that precludes an actor which really cares about a given collective action problem from individually making up for free riders by contributing more to cover the share that they ‘should’ have covered. Russell Hardin discusses this extensively in Collective Action; in international relations, it is the basis of theories of hegemonic provision of public goods. Collective underprovision of public goods is surely a problem for international action, but one cannot say that it is an insuperable one on the basis of collective action theory alone.