and Centre for the Study of Civil War, PRIO.

Preliminary draft. Please do not cite. Current research on the influence of democracy and democratization on civil war has typically relied on aggregated indices of regime type. This study deviates from this conventional approach by focusing more closely on whether there is a link between elections and conflict behavior. First we study the influence of competitive and non-competitive elections on civil-war onsets at the country level. The study proceeds by disaggregating the analysis to the level of ethnic groups. At this level we examine how ethno-nationalist mobilization and sore-loser effects may provoke post-electoral violence. Our findings reveal that elections can be associated with ethnic civil wars depending on whether the elections are competitive or not and whether the conflict is territorial or governmental