Withdrawing from international custom.

• Author: Bradley, Curtis A.

ARTICLE CONTENTS INTRODUCTION I. MODERN UNDERSTANDING OF CIL II. HISTORICAL SUPPORT FOR A DEFAULT VIEW OF CIL III. SHIFT TO THE MANDATORY VIEW A. Timing of and Reasons for the Shift B. Persistent Objector Doctrine 1. Asylum and Fisheries Cases 2. Academic Commentary 3. Implications for the Mandatory View IV. FUNCTIONAL ASSESSMENT OF THE MANDATORY VIEW A. Process of CIL Formation B. Sticldness Concerns C. Holdout Problems D. Adaptability to Change E. Professor Guzman's Proposal V. THE AFFIRMATIVE CASE FOR RESTRICTING OPT-OUT A. Reliance B. Rule of Law and Legitimacy C. Externalities and Agency Problems 1. Externalities 2. Agency Problems D. Penalty Default E. Differences Between Treaties and CIL F. Developing a Typology CONCLUSION INTRODUCTION

There are two basic types of international law--treaties and customary international law (CIL). Treaties are negotiated, usually written down, and often subject to cumbersome domestic ratification processes. Nonetheless, nations often have the right to withdraw unilaterally from them. Many treaties expressly provide for a right of withdrawal, often with a notice requirement. As Professor Helfer has documented, "Treaty clauses that authorize exit are pervasive." (1) Indeed, these clauses exist even in treaties that reflect core principles of international public policy, such as the Geneva Conventions and the Nuclear Non-Proliferation Treaty. (2) Moreover, treaties that do not address the issue of withdrawal may be found to allow implicitly for withdrawal, based, for example, on their subject matter. (3) Even when a treaty does not generally permit withdrawal, nations may still have a right to withdraw in the event of a fundamental change of circumstances. (4)

Unlike treaties, the rules of CIL do not arise from express negotiation, and they do not require any domestic act of ratification to become binding. Although these differences might suggest that nations should have greater flexibility to withdraw from rules of CIL than from treaties, the conventional wisdom is precisely the opposite. According to most international law scholars, a nation may have some ability to opt out of a CIL rule by persistent objection to the rule before the time of its formation (although even that proposition is contested), but once the rule becomes established, nations that are subject to it never have the right to withdraw unilaterally from it. Rather, if a nation wants to engage in a practice contrary to an established CIL rule, it must either violate the rule or enter into a treaty that overrides the rule as between the parties to the treaty. As explained by a committee of experts in its study of CIL for the International Law Association:

There is fairly widespread agreement that, even if there is a persistent objector rule in international law, it applies only when the customary rule is in the process of emerging. It does not, therefore, benefit States which came into existence only after the rule matured, or which became involved in the activity in question only at a later stage. Still less can it be invoked by those who existed at the time and were already engaged in the activity which is the subject of the rule, but failed to object at that stage. In other words, there is no "subsequent objector" rule. (5) We will refer to this as the "Mandatory View" of CIL.

The contemporary international law literature contains almost no explanation for the Mandatory View. For the most part, the Mandatory View is simply stated as being canonical. It is not obvious, however, why it should be easier to exit from treaties than from CIL, especially given the significant regulatory overlap that exists today between treaties and CIL. (6) Moreover, modern claims about the content of CIL often rely heavily on the content of treaties, especially multilateral treaties, even though many of these same treaties contain withdrawal clauses. (7) There are also functional reasons to doubt the desirability of a mandatory conception of CIL, at least when applied across the board to all CIL issues. (8) In addition, the Mandatory View is in tension with other aspects of CIL doctrine. In particular, if unilateral withdrawal from a CIL rule is so problematic, why are individual nations allowed to opt out of the rule, before it becomes established, through persistent objection? Conversely, if the persistent objector doctrine is needed in order to ensure that CIL is consensual, why does that consent rationale not also require the allowance of opt-out through subsequent objection?

The Mandatory View of CIL, it turns out, was not always as canonical as it is today. Rather, a number of prominent international law publicists of the eighteenth and nineteenth centuries thought that CIL rules were at least sometimes subject to unilateral withdrawal. This was also the view of the early U.S. Supreme Court in some of its most famous CIL decisions. That is, publicists and the Court had in mind what we will call a "Default View" of CIL. Few contemporary scholars appear to be aware of this fact, and not one of the leading treatises or casebooks on international law mentions it. The Mandatory View did not come to dominate international law commentary until sometime in the twentieth century. Even then, there were substantial uncertainties about how the Mandatory View would operate, and CIL evolved in ways that are difficult to explain under that View. Modern commentators, however, invoke the Mandatory View without a sense of its intellectual history and without any effort to explain its justifications.

Furthermore, there are reasons to question the normative underpinnings of the shift in the literature from the Default View to the Mandatory View. This shift appears to have occurred in the late nineteenth and early twentieth centuries, at a time when imperialism was at its height and most of Asia and Africa were under the control of the European powers. The "family of nations" was being expanded, but control over law creation was still largely in the hands of the so-called civilized nations, which meant primarily nations in Western Europe. Viewed in this context, the Mandatory View may have evolved as part of an effort to bind new nations and former colonies to international law rules that had already been worked out by a handful of powerful states. (9) Later, when the new entrants into the system gained strength in numbers and the former powers became concerned that they might lose control of the process of law creation in the international arena, the Mandatory View was modified to allow for opting out of a CIL rule prior to the time that it became established. (10) This right of opt-out, however, was crafted in such a way as to disallow new nations from opting out of any of the CIL rules that had developed before they came into the system. This history sits uneasily with contemporary assumptions in international law about sovereign equality and universality. (11)

This Article considers the intellectual history and functional desirability of the Mandatory View. In Part I, we describe the modern understanding of CIL and explain how the Mandatory View fits within this understanding. In Part II, we document that a number of the leading international law publicists of the eighteenth and nineteenth centuries assumed that nations had the right to opt out of at least some CIL rules and that early decisions by the U.S. Supreme Court also reflected this view. In Part III, we trace the intellectual shift in the twentieth century toward the Mandatory View, and we further describe the substantial uncertainties that persisted about how CIL would operate under this View. Finally, in Parts IV and V, we consider the functional desirability of the Mandatory View and conclude that it is difficult to justify that View today--at least for all of CIL. In these parts, we draw on theories developed with respect to contract law, corporate law, voting rules, and constitutional design.

If international law theory made room for some withdrawal rights under CIL, a number of consequences would follow. Nations that found that a CIL rule no longer addressed their needs would have an option other than simply violating it or attempting to bargain around it. Any withdrawal rights would need to be invoked publicly, presumably with reference to the purported CIL rule at issue and also perhaps to the reasons for withdrawal, thereby potentially increasing transparency and dialogue. Such withdrawal rights would also reduce the danger that multilateral treaties could create CIL that, unlike the underlying treaties, would lack an exit option. The potential "sovereignty costs" of establishing and joining such treaties would thereby be lower. (12) At the same time, by allowing for differentiation in withdrawal rights among CIL rules, this approach would allow for the possibility that some CIL rules could be more mandatory than under the current one-size-fits-all approach (for example, by disallowing opt-out through prior persistent objection for some CIL rules).

This is an important time to be thinking about CIL theory. The now-widespread use of multilateral treaties as vehicles for international legislation raises questions about what role CIL should play in the international legal system and how this role relates to the one played by treaties. In addition, whereas historic claims about the content of CIL relied heavily on inductions from national practice, many modern claims rely instead on deductions from international pronouncements, (13) a shift that may have implications for a variety of issues, including withdrawal rights. At the same time, there have been increasing challenges to CIL's legitimacy and effectiveness, with some critics arguing that CIL is no longer useful as a body of international law, (14) and others arguing that it does not act as a constraint on national behavior. (15) Unlike these critics, our goal is not to challenge CIL's legitimacy, but...