Abstract 

This paper explores the trade-off between police officer behaviors and crime, focusing on the relationship between productive law enforcement, crime reduction, and high-cost behaviors (misconduct and use of force). I use a decade of daily shift data from the Chicago Police Department to estimate the causal effects of individual officers on stops, arrests, use of force, misconduct, and crime. Officer heterogeneity generates significant variation in each outcome. However, officers more prone to costly behaviors are only modestly more productive, as measured by enforcement, and are not better at reducing crime. Furthermore, officer-level activities, such as stops and low-level arrests, have minor direct effects on crime. Finally, I apply social cost estimates to policing activities and crime outcomes. Officers more prone to misconduct and use of force do not prevent enough crime to justify their high-cost behaviors, despite making more arrests and stops. These results challenge the belief that aggressive policing improves public safety.