The Effect of Police Use of Lethal Force on Murder Levels in American Cities Before and After Ferguson

Abstract

We assess the effects of police use of lethal force on subsequent murders by victim race and armed status before and after the August 2014 death of Michael Brown in Ferguson, MO. We regress monthly murder levels on instances of police use of lethal force by race and armed status controlling for fixed effects, population, unemployment, and murders in the prior month using city-level data for 93 cities from 2013 to 2015. For 2013–2015, we find that a police lethal force incident predicts a 1.8% increase in murders 2 months following the incident. However, prior to Ferguson, a police lethal force incident increases murders by 4.5% (after 2 months) and we are unable to find any evidence of differential responses to police use of lethal force based on victim race. However, we find evidence of differential responses to police use of lethal force based on victim armed status. Post-Ferguson, police use of lethal force is associated with significant differential responses based on victim race. In addition, we see a shift in the aggregate-level response to police use of lethal force. Post-Ferguson, a police lethal force incident decreases the murder level by 3.8% 3 months following the incident. In addition, lethal force incident involving a non-black victim decreases the number of murders by 4.3% while an incident involving a black victim increases the number of murders by 2.1% (after a 2-month lag). Changes in policing following the events of Ferguson had a generally positive but uneven effect.