This article explores the link between industrial policy and electoral outcomes under dictatorship. Using a difference-in-differences analysis of county-level panel data from 1971-88 in South Korea, it examines whether the industrial policy implemented by an authoritarian government affects constituents' electoral decisions. It finds that counties receiving economic benefits through the construction of industrial complexes cast more votes for the incumbent party in subsequent elections. The effects are larger in elections immediately after the appointment of an industrial complex or at the beginning of its construction compared to elections held after the completion of construction. Furthermore, the study tests and rejects reverse causality and migration effects as possible alternative mechanisms for the changes in electoral outcomes. Finally, to understand a unique feature of authoritarian elections, it tests whether industrial complexes affect electoral fraud. Using a genetic matching methodology, it finds that places with new industrial complexes are less likely to experience electoral fraud.