Using data from a field experiment in 500 villages, this paper studies how local institutions affect the quality of governance, as measured by aid distribution outcomes. In villages where elected councils exist and manage distributions, aid targeting improves. However, if the distribution is not clearly assigned to either the council or customary leaders, the creation of elected councils increases embezzlement and makes decision-making less inclusive. Requiring that women manage the distribution jointly with customary leaders also increases embezzlement. Thus, while elected councils can improve governance, overlapping mandates between new and existing institutions may result in increased rent-seeking.