Too many people assume, generally without having given any serious thought to its character or its history, that international law is and always has been a sham. Others seem to think that it is a force with inherent strength of its own… Whether the cynic or sciolist is the less helpful is hard to say, but both of them make the same mistake. They both assume that international law is a subject on which anyone can form his opinions intuitively, without taking the trouble, as one has to do with other subjects, to inquire into the relevant facts. —J. L. Brierly
Regime design matters.1 International treaties and regimes have value if and only if they cause people to do things they would not otherwise do. Govern ments spend considerable resources and effort drafting and refining treaty language with the (at least nominal) aim of making treaty compliance and effectiveness more likely. This article demonstrates that whether a treaty elicits compliance or other desired behavioral changes depends upon identifiable characteristics of the regime’s compliance systems.2 As negotiators incorporate certain rules into a regime and exclude others, they are making choices that have crucial implications for whether or not actors will comply. For decades, nations have negotiated treaties with simultaneous hope that those treaties would produce better collective outcomes and skepticism about the ability to influence the way governments or individuals act. Both lawyers and political scientists have theorized about how international legal regimes can influence behavior and why they often do not.3 Interest in issues of compliance and verification has a long history in the field of nuclear arms control.4 More recently, this interest in empirically evaluating how interna tional institutions, regimes, and treaties induce compliance and influence behavior has broadened to include other security areas as well as international trade and finance.5 Concern over the fate of the earth’s environment recently has prompted a further extension into questions of whether and how environ mental treaties can be made more effective at eliciting compliance and achieving their goals.6 Researchers in all these issue-areas face two critical questions.
First, given that power and interests play important roles in determining behavior at the international level, is any of the compliance we observe with international treaties the result of the treaty’s influence? Second, if treaties and regimes can alter behavior, what strategies can those who negotiate and design regimes use to elicit the greatest possible compliance? This article addresses both these questions by empirically evaluating the international regime controlling inten tional oil pollution. Numerous efforts to increase the regime’s initially low levels of compliance provide data for comparing the different strategies for eliciting compliance within a common context that holds many important explanatory variables constant. The goal of the treaties underlying this regime has been to reduce intentional discharges of waste oil by tankers after they deliver their cargoes. Since the late 1970s, these treaties have established two quite different compliance systems, or “subrcgimcs,” to accomplish this goal. One has prohibited tanker operators from discharging oil in excess of specified limits. The other has required tanker owners to install expensive pollution reduction equipment by specified dates.
Treaty parties viewed both subregimes as equally legitimate and equally binding.7 The two subregimes regulated similar behavior by the same nations and tankers over the same time period. The absence of differences in power and interests would suggest that compliance levels with the two subrcgimcs would be quite similar.8 According to collective action theory, these cases are among the least likely to provide support for the hypothesis that regime design matters: subregime provisions required the powerful and concentrated oil industry to incur large pollution control costs to provide diffuse benefits to the public at large.9 Indeed, the lower cost of complying with discharge limits would suggest that compliance would be higher with those limits than with equipment requirements. Not surprisingly, violations of the limits on discharges have occurred frequently, attesting to the ongoing incentives to violate the agreement and confirming the characterization of oil pollution as a difficult collaboration problem.10 A puzzle arises, however, from the fact that contrary to expectation compliance has been all but universal with requirements to install expensive equipment that provided no economic benefits.
The following analysis clearly demonstrates that the significant variance across subregimes can only be explained by specific differences in subregime design. Comparing the two compliance systems shows that the equipment subregime succeeded by ensuring that actors with incentives to comply with, monitor, and enforce the treaty were provided with the practical ability and legal authority to conduct those key implementation tasks. Specifically, the regime elicited compliance when it developed integrated compliance systems that succeeded in increasing transparency, providing for potent and credible sanctions, reducing implemen tation costs to governments by building on existing infrastructures, and preventing violations rather than merely deterring them.
Compliance theory and definitions
Explaining the puzzle of greater compliance with a more expensive and economically inefficient international regulation demands an understanding of existing theories about the sources of compliance in international affairs. Realists have inferred a general inability of international regimes to influence behavior from the fact that the international system is characterized by anarchy and an inability to organize centralized enforcement. In what has been the dominant theoretical view, “considerations of power rather than of law determine compliance.”11 To explain variance in treaty compliance, look for variance in the power of those with incentives to violate it or in the interests of those with the power to violate it.
Treaties arc epiphenomenal: they reflect power and interests but do not shape behavior. This view docs not imply that noncompliance is rare in international affairs. Although nations will violate rules whenever they have both the incentives and ability to do so, as Hans Morgenthau notes, “the great majority of the rules of international law are generally observed by all nations.”12 For the realist, behavior frequently conforms to treaty rules because both the behavior and the rules reflect the interests of powerful states. More specifically, compliance is an artifact of one of three situations: (1) a hegemonic state forces or induces other states to comply; (2) the treaty rules merely codify the parties’ existing behavior or expected future behavior; or (3) the treaty resolves a coordination game in which no part}’ has any incentive to violate the rules once a stable equilibrium has been achieved.13 Treaty rules correlate with but do not cause compliance. Therefore, efforts to improve treaty rules to increase compliance reflect either the changed interests of powerful states or are misguided exercises in futility.
The strength of this view has led to considerable attention being paid to whether rules influence behavior and far less being paid to design features that explain why one rule influences behavior and another docs not. In contrast, international lawyers and institutionalists contend that the anarchic international order need not lead inexorably to nations violating agreements whenever doing so suits them. Other forces—such as transparency, reciprocity, accountability, and rcgimc-mindedness—allow regimes to impose significant constraints on international behavior under the right conditions.14 Implicit in the institutionalist view is the assumption that power and interests alone cannot explain behavior: a given constellation of power and interests leaves room for nations to choose among treaty rules that will elicit significantly different levels of compliance. High compliance levels can be achieved even in difficult collaboration problems in which incentives to violate arc large and ongoing. Treaties can become more effective over time, and regimes may even learn.
Agreeing with Morgenthau that compliance will be quite common, institutionalists do not exclude the possibility that the regime, rather than mere considerations of power, causes some of that compliance.16 In essence, this debate revolves around whether in a realm of behavior covered by an international agreement, that behavior is ever any different than it would have been without the agreement. If we define “treaty-induced compliance” as behavior that conforms to a treaty’s rules because of the treaty’s compliance system, institutionalists view treaty-induced compliance as possible. In contrast, realists see all compliance as “coincidental compliance,” that is, behavior that would have occurred even without the treaty rules. The debate between these theories highlights the demands placed on research that seeks to identify those design characteristics of a regime, if any, that are responsible for observed levels of compliance. I define compliance, the dependent variable, as an actor’s behavior that conforms with an explicit treaty provision. Speaking of compliance with treaty provisions rather than with a treaty captures the fact that parties may well comply with some treaty provisions while violating others.
A study of “treaty compliance” would aggregate violation of one provision with compliance with another, losing valuable empirical information.17 Restricting study to the explicit rules in a treaty-based regime allows the analyst to distinguish compliance from noncom pliance in clear and replicable ways. Obviously, a focus on explicit rules ignores other potential mechanisms of regime influence, such as norms, principles, and processes of knowledge creation.18 However, this restrictive definition has the virtue of bringing the debate to a level at which research on actual treaties and actual compliance can contribute to the intellectual and policy debates. This article evaluates the features of a regime that may determine compli ance by differentiating among three parts of any compliance system: a primary rule system, a compliance information system, and a noncompliance response system. The primary rule system consists of the actors, rules, and processes related to the behavior that is the substantive target of the regime. In the choice of who gets regulated and how, the primary rule system determines the pressures and incentives for compliance and violation. The compliance information system consists of the actors, rules, and processes that collect, analyze, and disseminate information on instances of violations and compli ance.
Self-reporting, independent monitoring, data analysis, and publishing comprise the compliance information system that determines the amount, quality, and uses made of data on compliance and enforcement. The noncom pliance response system consists of the actors, rules, and processes governing the formal and informal responses—the inducements and sanctions employed to induce those in noncompliance to comply. The noncompliance response system determines the type, likelihood, magnitude, and appropriate ness of responses to noncompliancc. These categories provide the framework used in the remainder of this article to evaluate the oil pollution regime’s sources of success and failure in its attempt to elicit compliance.