We compared police department use of force policies with police killings data for 91 of the 100 largest police departments to see if there was a relationship between the two. We found that police departments with policies that place clear restrictions on when and how officers use force had significantly fewer killings than those that did not have these restrictions in place.


For this analysis, we used police killings data from The Guardian's The Counted database, from January 1, 2015 - July 15, 2016. As shown by the chart below, there was wide variation in rates of police killings among America's largest city police departments.

Then we examined the extent to which killings by these police departments were related to the number of restrictive use of force policies these departments had, as well as other factors including the number of arrests made by the department, size of the police force, racial demographics of each city, number of assaults on officers, and the median income and level of inequality in each city.



For each of the 8 policies examinedpolice departments that had implemented the policy were less likely to kill people than police departments that had not.

Police departments with four or more of these restrictive use of force policies had the fewest killings per population and per arrest. After taking into account other factors, each additional use of force policy was associated with a 15% reduction in killings by police. According to our analysis, the average police department would have 54% fewer killings and a police department with none of these policies currently in place would have 72% fewer killings by implementing all eight of these policies.


These results indicate that while the chances of killing a civilian increases the more arrests a police department makes, that likelihood is shaped by the department’s policies governing how and when police can use force during those encounters. This suggests that advocacy efforts pushing police department to adopt more restrictive use of force policies - and the accountability structures to enforce them - can substantially reduce the number of people killed by police in America. And while this analysis was limited to examining rates of deadly force, these policies may also be associated with reductions in other forms of police violence as well.

Despite their potential impact, efforts to push for these changes have often been opposed by police organizations that claim more restrictive use of force policies “endanger officers” (See herehere, and here). We find that these assumptions are not supported by the data. Officers in police departments with more restrictive policies in place are actually less likely to be killed in the line of duty, less likely to be assaulted, and have similar likelihood of sustaining an injury during an assault. 

In short, a commitment to protect and preserve life necessitates the immediate adoption of more restrictive policies governing when and how officers use force in our communities.