This dissertation examines the impact of patterns of political competition on the promotion of good governance in post-conflict societies. There is a well-established link between robust political competition and governments that govern well (Grindle 2007; Grzymala-Busse 2007; Sen 1999). However, this linkage falters in societies with deep identity divisions, such as those emerging from civil war, where political competition is associated with instability and violence (Horowitz 2000; Rabushka and Shepsle 1972; Snyder 2000). These deep divisions inhibit competition by presenting barriers to voter choice and favoring clientelistic competition over programmatic competition. Competition is further inhibited by political institutions designed to prevent electoral competition from spilling over into violent conflict and a return to civil war. I explore the conditions under which post-conflict societies nevertheless develop programmatic competition and good governance. I employ a qualitative and quantitative within-case empirical strategy based on elite interviews in Bosnia and Herzegovina and a unique dataset matching precinct-level voting data with municipal-level governance outcomes to analyze the patterns of programmatic competition across ethnopolitical electoral contexts and their impact on the provision of good governance. I find that municipalities with strong competition between nationalist and non-nationalist parties, and therefore with meaningful programmatic competition, show significantly better governance than those with only clientelistic competition between ethnic parties, even when the ethnic parties are of the same ethnicity. However, I find significant barriers to the promotion of programmatic competition by external forces owing to fundamental conflict between the attempt to promote programmatic competition and the desire for meaningful multiethnic political institutions. My analysis of ticket-splitting across electoral contexts shows that elections to multiethnic political institutions strongly favor ethnic voting. I conclude with policy recommendations aimed at bridging that gap and indications for future research.