By Li Sian Goh

Despite the widespread implementation of new policies intended at reducing police officers’ use of force in encounters with civilians, evidence about these policies remains patchy. Consisting of a series of three papers, my dissertation seeks to contribute to the body of knowledge on measures to reduce police use of force. The first paper considers the effect of consent decrees and other forms of federal intervention on fatalities resulting from citizen encounters with the police. Using a panel dataset of deaths caused by the police between the years 2000 and 2016, a difference-in-differences analysis found that court-appointed monitorships reduced fatalities by 29.1%. The second paper examines the effect of departmental oversight over body-worn cameras on officer-involved shootings. Using a panel dataset of police departments in 36 large departments across the United States, a difference-in-differences analysis finds that while the presence of body-worn cameras alone did not reduce shootings, restricting officer discretion on when cameras should be activated reduced shootings by 33.3%. The final paper evaluates the effect of a de-escalation training program. A difference-in-differences analysis of individual officers revealed no significant changes in serious force levels between officers who had been trained and officers who had not been trained. However, an analysis comparing the department which implemented the training with other law enforcement agencies in the same state suggested that serious force rates declined by approximately 40%. In totality, the dissertation

Philadelphia: University of Pennsylvania, 2020. 171p.