Democratization, leader education and growth: firm-level evidence from Indonesia

Abstract Does the economic success of democratization depend on newly elected leaders’ characteristics? We exploit unique Indonesian process, where districts exogenously democratized in different years. In a census manufacturing plants, employment drops by 5% that elect non-college educated mayor, while stays constant under college graduates. Non-college mayors substantially raise taxation but provide less infrastructure, do not spend more social programs, and are often involved corruption cases. Other leader attributes district characteristics, as well tests for pre-treatment trends, selection unobservables, close elections explain away important role education shaping local policies growth.