This paper investigates how the devolution of oil windfalls affects the likelihood of political violence. It shows that transferring large shares of oil wealth can prevent conflict, while transferring small shares can trigger it. Among the different transfer schemes, fiscal transfers (to subnational governments) yield the highest levels of consumption, but direct transfers (to people) are the most effective in preventing conflict. By averting conflict, transfers can improve ex ante welfare
however, only a subset of the ex ante welfare optimal transfers is optimal ex post and thus self-enforcing. Among them, those that avert conflict by reinforcing repressive regimes are of particular policy interest.