The Decision To Ally: a Theory and Test

the study of alliance formation. Most of the early work in this area centered on theorizing or speculating about the formation of particular alliances (Liska 1968; Russett 1968; Rosen 1970). More recent work, spurred by the diffusion of time-series analysis into political science as well as the availability of data, has focused on alliance as a systemic phenomenon (McGowan and Rood 1975; Job 1976; Siverson and Duncan 1976). Unfortunately, neither of these approaches has been able to provide us with a theory of alliance formation. This state of affairs would not be so bad if alliances were unimportant. Yet, at least two studies have demonstrated that alliances are of great importance for the initiation and spread of international conflict (Altfeld and Bueno de Mesquita 1979; Bueno de Mesquita 1981). In this paper I will employ some notions of micro-economics first introduced by Bruce Berkowitz (1983) to the study of alliance cohesion in order to develop a simple theory of the process by which national governments choose to form military alliances. I will then proceed to test a necessary but not sufficient condition for alliance formation derived from the theory on the set of major power alliances in the 19th century.