Customary International Law: An Instrument Choice Perspective

This Article proceeds as follows. Part II begins by considering custom’s design features, which the authors distinguish from the canonical elements of custom (state practice and opinio juris) and the individual doctrines associated with CIL. Specifically, they contend that, as an ideal-type, custom is non-negotiated, unwritten, and universal, three characteristics that distinguish CIL from both treaties and soft law, which are almost always negotiated, written, and rarely universal either in formation or application. These design features help to explain some of custom’s peculiar doctrinal characteristics, and they cut across the doctrinal divide which is said to distinguish “traditional” and “modern” custom. Part III considers the constraints that limit states’ recourse to CIL to particular types of cooperation problems or “domains.” Although custom’s design features make it ill-suited to resolve many transborder public goods or collective action problems, they argue that states can nonetheless generate custom in a range of potentially important contexts. Drawing upon numerous historical and contemporary examples, they show that the design features discussed in Part II facilitate custom’s formation primarily in three situations: when all states benefit from a customary rule with low distributional costs, when powerful nations impose a custom on weaker states, and when states seek to entrench shared normative values. Outside of these three domains—or where there is overlap among them—custom is much less likely to form. Part IV considers custom’s future in an international legal landscape dominated by multilateral treaties and soft law initiatives. The authors argue that states select CIL as their instrument of choice (within the constraints imposed by custom’s domains) when its substantive norms or its design features confer advantages over those of soft law or treaties. For example, states may use custom to “unbundle” certain negotiated aspects of multilateral conventions, especially those that preclude reservations. A state could decline to become a party to such a treaty yet profess that some or most of its provisions (in particular those it favors) are binding as CIL. Drawing from the law of the sea, international economic law, human rights, the laws of war, and other substantive issue areas, Part IV provides numerous contemporary examples in which states prefer custom to treaties or soft law, based on either its substantive norms or design features. Accordingly, custom remains relevant even in the age of soft law and treaties, so long as states act within the constraints imposed by custom’s domains. Part V concludes.