Democracy doesn’t solve war---best models.

Conclusion The dyadic understanding of the democratic peace has become ubiquitous in International Relations. By looking beyond simple dyadic analysis, accounting for the embededness of states in a much more complex network, we found the democratic peace may not be as robust as previously thought. Our results demonstrate that after accounting for the tendency for like-regime states with common enemies not to fight one another, the effect of the democratic peace not only vanishes, but jointly democratic dyads seem to be more conflict prone than mixed dyads. These results are consistent across operationalizations of the outcome variable, our triadic closure predictor, measurements of joint democracy, and a variety of other factors. We believe this explanation for the democratic peace is not a mechanism for understanding the democratic peace, but instead, an alternative. What we have shown here is that conflict between democracies indeed exists and the peaceful relations occasionally found are not necessarily a function of the affinity of democratic states, or intrinsic attributes of democratic states, but instead, a function of the strategic inefficiencies of fighting a state with a shared enemy. While regime type may influence the interests of states, we find that it does not directly influence the probability that any two states fight one another. There are three major implications to our research. First, scholars should be hesitant to consider dyadic conflict in isolation, as there are network dependencies informing whether a state engages or joins a MID. Second, preferences operating in addition to network interdependencies and collaboration explain much of the democratic peace. Third, when studying conflict, scholars and practitioners should consider the cost structure of collaboration, and how these dynamics inform not only conflict initiation, but conflict escalation. Particularly interesting is that the theoretical mechanism at work here is dramatically simpler than any of the established justifications for the democratic peace. We do not rely on arguments about institutions or norms, but just the simple and intuitive proposition that it does not make much sense for two states fighting a third to also fight each other. What the existing literature seems to have missed, usually theoretically and almost always empirically, is that dyadic conflicts do not occur in isolation, but in the context of a complex network of relations.