Although the concept of soft law has existed for years, scholars have not reached consensus on why states use soft law or even whether “soft law” is a meaningful analytic category. In part, this confusion reflects a deep diversity both in the types of international agreements that states employ, and in the strategic situations that produce these agreements. In this paper, we advance four complementary explanations for why states use soft law. Our explanations account for a much broader range of state behavior than the existing literature is able to explain.

First, and least significantly, states may use soft law to solve straightforward coordination games in which the existence of a focal point is enough to generate compliance. In such situations, there is no difference between hard and soft law from a compliance standpoint. But because hard law does not deliver any additional compliance benefits, states may choose soft law to avoid even modest costs associated with making an agreement binding.

Second, under what we term the “loss avoidance” theory, moving from soft law to hard generates higher sanctions, which both deter more violations and, because sanctions in the international system are negative sum, increase the net loss to the parties. States will choose soft law when the marginal costs in terms of the expected loss from violations exceed the marginal benefits in terms of deterred violations.

Third, the “delegation theory” predicts that under certain conditions states will choose soft law because, relative to hard law, soft law makes it easier for individual states to renounce existing rules or interpretations of rules and adopt new ones. These renunciations can prompt a larger set of states to coordinate their behavior around the new standard, and thereby drive the evolution of soft law rules in a way that may be superior to formal renegotiation.

Fourth, we introduce the concept of international common law (“ICL”), which we define as a non-binding gloss that international institutions, such as international tribunals, put on binding legal rules. A binding rule generally requires each state to consent to being bound. A rule of ICL, on the other hand, can be made with the input of a subset of those states bound by the underlying legal rule, such as, for example, those states that have consented to the jurisdiction of a tribunal. The rule of ICL nevertheless affects all states bound by the underlying rule because it shapes states’ expectations as to what constitutes compliance with that rule. As such, ICL provides cooperation-minded states with the opportunity to deepen cooperation if they are willing to surrender some measure of control over the content of legal rules to an international institution.

These four explanations of soft law, and in particular the theory of ICL, highlight the importance of soft law in the international legal system. They demonstrate that there are a range of international instruments that are employed precisely because they are non-binding but still produce legal consequences. Moreover, the explanations offered in this paper explain the circumstances under which the quasi-legal nature of soft law will be attractive to states.