371 Introduction What is the nature of the relationship between dyadic alliance formation and conﬂict initiation? Answers to that question in existing research about the alliance– conﬂict relationship are complicated. One possible reason for the variety among the ﬁndings is misspeciﬁcation of the relation- ship between alliance formation and conﬂict, and, in fact, most scholars agree that the relationship is complex. The literature treats alliances as a potential cause of conﬂict as if alliances arise from a process largely divorced from the process that leads to conﬂict. In fact, it is almost certainly the case that the decision to ally and the decision to ﬁght arise from many of the same causal factors. When two states decide to ally, the very things that made them ally also shape the chances they ﬁght, but probably diminish those chances for the very reason that they committed to an alliance. I develop an argument and model linking alliances and conﬂict in this fashion based on three claims: (1) alliance formation decisions and conﬂict initiation decisions are not independent of one another, (2) alliance formation and conﬂict initiation are caused by some of the same © 2006 Journal of Peace Research, vol. 43, no. 4, 2006, pp. 371–389 Sage Publications (London, Thousand Oaks, CA and New Delhi) http://jpr.sagepub.com DOI 10.1177/0022343306064816 Alliance Formation and Conﬂict Initiation: The Missing Link* ANESSA L. KIMBALL Département de science politique, L’Université Laval Existing research on the connection between alliance formation and conﬂict initiation has explicitly focused on the direct effect of alliances on conﬂict by including some measure of alliance behavior as an independent variable in models of conﬂict behavior. Existing research misspeciﬁes the relationship between alliances and conﬂict, because alliance formation and conﬂict initiation are shaped by many of the same factors (in particular, regime type and capabilities), and alliance formation decisions are endogenous to conﬂict initiation decisions. Thus, alliance formation and conﬂict initiation should be modeled in a system of equations where a set of variables shapes alliance formation and conﬂict directly, and indirectly affects conﬂict through the decision to ally. The author estimates a two-equation probit model that accounts for the endogenous nature of alliance formation decisions and, thus, for the indi- rect effects of variables like regime and power on conﬂict. Results suggest that the effect of regime on alliance behavior differs across time periods. Finally, the model provides evidence that the total effects of variables like power and regime on conﬂict are, in fact, mediated by how those variables inﬂuence the decision to ally. * Special thanks go to Dave Clark, Brett Ashley Leeds, Will Moore, Sarah Croco, Jacques Hymans, Bill Reed, and the anonymous reviewers for their helpful comments on earlier versions of this manuscript. This project was pre- sented at the 2003 Annual Meeting of the Midwest Politi- cal Science Association, the 2004 Merriam Junior Master Class in Formal Theory, and the World Politics Workshop at Binghamton University. The data used here and a web appendix are available at http://www.prio.no/jpr/datasets and bingweb.binghamton.edu/~akimbal1. Correspondence: Anessa.Kimball@pol.ulaval.ca.