Lead by Example: The Effects of Police Supervisors on Officer Behavior

American police organizations are bound by a strict chain of command. However, the impact of supervisors on police decision-making remains understudied. This project analyzes whether changing a police officer's supervisor can be an effective policy tool for reducing excessive enforcement. In order to assess the scope of impact for such a policy, the author estimates the effects of police supervisors on their subordinates' use of force and arrests. Utilizing patrol assignment data from 5 years of police calls for service in Dallas, TX, the author constructs a novel dataset that links patrol officers to their supervising sergeants. The empirical design exploits the frequent movement of workers between supervisors in order to estimate individual supervisors' effects on their subordinates' enforcement decisions. Preliminary findings suggest that a one standard deviation increase in the use of force effects of a patrol officer's supervisor increases the likelihood that the officer uses force by 42% of a standard deviation. Next, the author considers how police departments can identify the officers who would make the most effective supervisors by linking hundreds of newly-promoted sergeants to their personnel records, including demographic data, past law enforcement decisions, and promotional exam scores. Results indicate that the officers who perform the best in the promotional exam process are those who induce more force in their subordinates. These initial findings highlight potential benefits to reforming the supervisory selection process in policing, a policy which could prove to be more cost-effective than other large-scale interventions. CA/NCF