Explanations for the democratic peace have developed along two broad lines: those that focus on democratic-specific mechanisms and those that focus on institutional similarity mechanisms. We test these competing mechanisms against each other by examining a range of regime types and a variety of institutional mechanisms, and for the first time, test them in logistic multiple regression models of rivalry onset using three rivalry datasets. Our comparative test of the two approaches largely justifies democratic-specific explanations for the democratic peace. That is, more than any other regime pairing, two democracies are less likely to engage in a rivalry. We also find some evidence that socialist single-party dyads are less likely to experience rivalry than other dyads, but overall the support for a broader institutional similarity peace is quite limited.
KEYWORDS autocratic peace, democratic peace, regime similarity, rivalryIs the democratic peace part of a larger institutional similarity peace? The democratic peace is an empirical observation that democracies rarely fight each other. In light of this observation, theorists initially advanced democratic-specific arguments that explain the democratic peace through features and phenomena unique to democratic societies and institutions (see e.g.