Why Oil-Rich States are So Violent

New mimeo by Blair: I develop a theory of whether and where civil wars emerge in resource-rich states, based on bargaining within the state over the spoils and on the state's attempts to safeguard the resources. I argue that these processes are shaped by where the resource is discovered. In contrast to existing scholarship which describes oil wars over territorial control of natural resources, I predict that regional economic and security inequalities lead only to conflict in resource-poor regions. To test the theory, I present a new dataset of the timing and location of oil and natural gas discoveries and a novel causal identification strategy. I find that civil war is more likely in states with oil discoveries, but only if it is found on land where local populations can threaten to interrupt production. In stark contrast to scholarly and practitioner consensus, I find that these civil wars take place not in the oil region itself, but elsewhere in the state.