Elections have been found to increase the risk of conflict recurrence. In this thesis I investigate how institutions that constrain election winners mitigate the destabilizing effects of elections. I propose that post-conflict elections will only increase the chances of conflict recurrences in cases where institutional constraints on elected governments are weak. In these cases the post-conflict commitment problem makes it difficult for election winners to reassure elections losers that the settlement of the conflict will be respected. This makes it less likely that the losing side will be willing to hand over power to an elected government. Where the broader institutional framework is strong enough to constrain election winners after they assume office, elections may help pave the way for durable peace.
I employ a set of Cox regression models on a dataset of all peace spells in the 1972-2005 period to test this proposition empirically. The analysis finds robust support for the interaction between post-conflict elections and institutional constraints on elected governments. Where no such constraints are in place, post-conflict elections significantly increase the risk of conflict recurrence. If these institutions are strong, post-conflict elections are related to durable peace. Thus, whereas existing studies have found either a negative or no effect of post-conflict elections, I demonstrate that the effect of post-conflict elections is conditioned by the broader institutional context. This finding is robust to various model specifications, suggesting that a post-conflict democracy where competitive elections are combined with institutions of checks and balances may make peace more durable.