19 Sep IL/IR Symposium: Comment on von Stein
[Tim Meyer is an Assistant Professor of Law at the University of Georgia School of Law]
As is de rigueur in discussions of compliance with international law, von Stein’s chapter quotes in the opening paragraph Louis Henkin’s statement that “almost all nations observe almost all principles of international law and almost all of their obligations almost all of the time” (p. 477) – the claim that launched a thousand journal articles. Appropriately, von Stein’s excellent review of the compliance literature returns to Henkin in conclusion, noting in sum that “we know, for instance, that it is not the case that almost all states respect their obligations almost all the time.” (emphasis in original) (p. 495). In between, von Stein provides a clear, concise, and illuminating review of theories of compliance with international law and the empirical evidence for and against, and empirical challenges in evaluating, these theories. International law seems to drive states to conform to its mandates at least some of the time, and our understandings of the mechanisms at work is improving, if still in need of improvement.
In this comment, I want to suggest one way in which we can deepen our understanding of how international law affects state behavior. Specifically, I want to problematize the notion of compliance as a dependent variable. Von Stein’s essay describes the state of the art in compliance studies, but as Dunoff and Pollack note the IL/IR literature consists overwhelmingly of the application of IR theories to international law. Reconceptualizing how international law affects state behavior is a key way in which law can increasingly inform IL/IR scholarship. To put it simply, compliance – “the degree to which state behavior conforms to what an agreement prescribes or proscribes” (p. 478) – is undoubtedly a useful place to start studying how international law affects behavior, and great strides have been made in this area, but moving forward we need a conception of legal process that more accurately reflects how states actually implement and evaluate compliance with international law.
Such a conception begins with the recognition that written legal rules, standing on their own, are rarely precise enough to know “what an agreement prescribes or proscribes.” How are we to know based on a written treaty alone (von Stein excludes from her study customary international law and compliance with rulings, p. 478) whether a state has accorded a foreign investor “fair and equitable treatment?” Or what constitutes a “like” product for purposes of assessing compliance with most-favored nation or national treatment obligations in trade?
Treaties containing these kinds of imprecise standards are not created with the intention of immediately changing state behavior to conform to the legal standard. They are incomplete contracts that states are constantly negotiating and renegotiating, contesting the meaning of legal rules as much as they are evaluating and sanctioning each other’s conduct. In most cases, state expectations as to what constitutes compliance form and shift on the basis of this contestation, rather than on the basis of the text of a treaty. States may codify customary international law in an effort to change its meaning. They may use decisions by a Conference of the Parties to elaborate vague legal rules, as Bodansky describes the decisions of the parties compiled in the CITES Handbook as doing (Bodansky 2010). They may bring disputes before international tribunals in an effort to develop a jurisprudence on a relevant issue.
Treaties do still constrain states, contestation notwithstanding. Some actions are so far beyond what any state would consider compliant that sanctions (at least of the reputational variety) are predictable enough to provide a deterrent. Moreover, as states’ expectations as to what constitutes compliance coalesce around particular understandings, the instrumental sanctions that von Stein describes (pp. 479-483) will likely become relatively more effective at generating changes in the behavior of noncompliant states. To the extent that state expectations do not coalesce around agreed interpretations of rules, instrumental sanctions are weakened.
To be clear, this critique is different from one based on a claim that compliance is a binary variable. Von Stein is quite clear that she considers compliance to be a spectrum, rather than a dichotomy (p. 478). Further, I agree with von Stein that legal rules provide a good benchmark for measuring the effect of international law on state behavior. Effectiveness – the extent to which a treaty solves an underlying problem as measured against a counterfactual world with no treaty (p. 493) – is difficult to evaluate. Counterfactuals are often hard to evaluate, and international law cannot be studied in an experimental setting.
Nor am I claiming that ambiguity causes noncompliance, as some of the literature argues (p. 486). It would be more precise to say that ambiguity causes disputes. Where ambiguity exists states as a group do not have firm expectations about what behavior is compliant. These expectations may be clarified in the course of a dispute, during which a variety of actors – states, tribunals, international organizations – may take a position on whether the challenged action is compliant. Moreover, these actors may use sanctions as tool to signal their commitment to their interpretation going forward. The outcome of disputes can thus constrain states prospectively, but a finding of noncompliance may tell us relatively little about whether state action violated a pre-dispute shared understanding of what the legal rule required.
Finally, I am not urging a notion of compliance based on process-based theories as von Stein describes them. In her view, process theories view legal rules as generating compliance by shaping identities and social practices and relations. She contrasts this understanding of compliance with her definition (p. 493). Having a process-based understanding of how international law shapes state behavior does not require, however, that one accept that compliance is caused by a process of acculturation or through the shaping of social identities. The act of establishing an imprecise law – a standard rather than a rule – touches off a perfectly rational contest for elaborating the rule. Process-based theories, in other words, need not be linked to an assumption that states have an independent preference for complying with international law, or a theory of how legal process changes social relations. Rather, a process-based theory might view treaties as a form of relational contracting between parties in long-term relationships. On this account, a legal regime’s success or failure hinges not on the immediate compliance outcomes it produces, but on its long-run ability to manage the relationship between the parties successfully.
This account raises questions about how best to operationalize studies of international law’s behavioral effects. I do not mean to suggest in any way that studying compliance as von Stein defines it is inappropriate. To the contrary, studying any phenomenon as complicated as the effect of international law on state behavior necessarily requires some degree of simplification. The question, though, is where we go from here. A few tentative thoughts suggest themselves.
First, one approach would be to expand studies of compliance beyond first-order compliance with legal rules to include shifts in behavior in response to significant interpretative events. Recent studies have had success in looking at tipping points in when states change their behavior in response to changes in international law. A similar approach might be to look for disputes with significant ramifications for a legal regime and look for evidence that the behavior of states other than those party to the dispute changed in response to the dispute’s outcome. Second, as von Stein notes (p. 495), qualitative empirical studies can provide much the texture lost in quantitative studies. Third, it may be that von Stein’s notion of compliance is a more meaningful metric in mature regimes that have had a chance to develop a consensus about how their rule should be applied to common factual situations. Compliance may be a relatively less useful metric in new regimes in which state expectations about what constitutes compliance are not fully formed and/or diverge.