State Capacity and Violence: Evidence from the Rwandan genocide

Exploiting local variation in state capacity within Rwanda I investigate the link between state capacity and violence. Using a disaggregated measure of the intensity of the 1994 Rwandan genocide, I establish that greater local state capacity led to greater conflict intensity. I proxy modern state capacity with its precolonial counterpart, measured by the total time a district was incorporated in the precolonial kingdom. This ‘duration of incorporation’ measures the cumulative effect of the centralizing forces in the kingdom and acts as a proxy for state capacity. Since the kingdom expanded through conquest and consolidated through patronage relations revolving around cattle, I instrument the duration of incorporation with the geographical suitability for cattle. This strategy confirms the main result. State capacity, while usually associated with greater public good provision and higher GDP, played a central role in the mass killings in Rwanda.