Cooperation, Conflict, and the Costs of Anarchy ⇤

I consider a model in which two states choose how much to arm and whether to attack each other in successive periods. Arms are useful not only for deterrence or for taking territory, but also because they influence the resolution of a set of disputed issues. It is shown that states can cooperate on the issues by limiting military competition, but only as far as a “war constraint” allows. Factors determining the tightness of the war constraint imply hypotheses about the international determinants of military e↵ort and thus the costs of anarchy. Panel data on arms spending and force sizes are used to evaluate these implications. States spend less as they and their neighbors become more democratic; they spend less when they grow in aggregate income and more when their neighbors become stronger; and they reduce force levels after they acquire nuclear weapons. Over the past 60 years, there is a pronounced downwards trend in military spending and force levels, which is plausibly linked to implications of the nuclear revolution for major power conflict, major power policies that discourage invasions and annexations by smaller states, and the spread of democracy. The baseline model synthesizes, clarifies, and revises informal arguments about the determinants of interstate cooperation and conflict. Its application suggests that for contemporary world politics, even Realist arguments, rightly understood, imply that the costs of anarchy need not be ‐ and quite possibly are not ‐ particularly large.