Elected authorities across Latin America have pursued various policing practices to solve the problems of criminal violence and police ineffectiveness that have seized the region over the last decades. Some prioritized building institutional capacity and encouraging police-community partnerships. Most, however, saw the adoption of a military model as the only viable path forward. This dissertation provides a first attempt at understanding what explains the decision to militarize public safety and its consequences for the institution—the police—elected authorities were trying to reform. To address these questions, it takes a subnational approach and turns to a silent but highly consequential type of police militarization occurring across many Latin American cities: the appointment of military members as police chiefs. Whereas the conventional wisdom points to violence and partisanship as key drivers of police militarization, in the first chapter, I argue that the militarization of police leaders responds to political motives. Drawing on evidence from 5,580 police chief appointments in Mexico, I show that local elected authorities appoint military chiefs when they face coercive pressures and strategic incentives to align with the federal government’s militarized security strategy. In the second chapter, I further argue that military chief appointments provide elected authorities a low-cost opportunity to signal competence in security governance because character-relevant information is conveyed to citizens when politicians militarize public safety. Responses to a mock news report survey experiment lend support to my theory. Latin Americans not only express mayors are more effective in security matters when they appoint military chiefs, but they also think that mayors are improving policing and corruption. The third chapter examines the conditions under which the integration of military members into police departments results in bureaucratic destabilization—or disruptions in the inner workings of the police. Drawing on 300 anonymous interviews and a qualitative comparative analysis of seven police departments in Mexico, I demonstrate theoretically and empirically that destabilization is not only the product of differences in training and mode of operation between the military and police but rather a conjunction between professional differences on the one hand and rank position on the other.