Testing competing institutional explanations of the democratic peace: The case of dispute duration

Drawing upon two alternative versions of the institutional explanation for the democratic peace, we suggest competing hypotheses about the relationship between democratic political institutions and the length of dispute participation. One set of hypotheses originates in the argument that because of the bargaining arrangements internal to democratic states, disputes between democratic states will necessarily be drawn out, so that in the time that it takes to secure the domestic political base for war, diplomats have time to find nonwar solutions. A second set of hypotheses, derived from the selectorate argument about how institutions shape the behavior of leaders who want to remain in power, leads to the expectation that selection effects over which disputes to participate in make disputes between democracies shorter than disputes between pairs of other types of states. Using a Weibull survival model we analyze data on the length of Militarized Interstate Disputes during the period 1816 to 1992; we find clear support for the selectorate explanation. Two ancillary hypotheses from the selectorate argument are also tested and supported by the data.