Holy Trinities,Rivalry Termination,and Conflict

In this article we attempt to correct a number of gaps in the current literature on strategic rivalry. First, we argue that liberal and realist theories of conflict and cooperation have been generally ignored by scholars engaged in rivalry research. Secondly, we argue that current rivalry research fails to disentangle termination processes from conflict within rivalries, mainly due to problems with common operationalizations of rivalry. To bridge these gaps, we test the effects of liberal variables-manifested in the Kantian tripod (democracy, interdependence, and IGO membership)-and what we more loosely term the realist tripod (bipolarity, shared threat, and capability balance) on both rivalry termination and the probability that rivals will engage in a militarized conflict. We conduct this test utilizing a new data set of strategic rivalries compiled by Thompson (1999, 2001) which corrects the tautological operationalization of rivalry commonly used in rivalry research when conflict is the dependent variable. Overall, our results paint a variegated picture, underscoring the importance of rivalry as a special class of dyadic relationship. While realist variables better explain rivalry termination, the direction is opposite that predicted by some realist hypotheses; among liberal variables, only democracy is a robust predictor of both termination and conflict.