Since George Floyd’s high-profile murder in 2020, the role of police and police departments in the community, as well as their interactions with civilians, have come under intense scrutiny and examination. Through video documentation and extensive social media coverage over the last five years, police violence has been deemed one of the most pressing examples of structural racism in the United States. Incidents of police brutality have sparked conversations on federal defunding and structural reorganization as many citizens, particularly Black Americans, express their feelings of fear and distrust. Yet the fact remains that police departments are highly reliant on community cooperation in identifying and solving crimes. 

Jesse Bruhn, Annenberg Assistant Professor of Education and Assistant Professor of Economics, along with a team of colleagues, conducted a research study on how such significant acts of police violence, particularly high-profile events, affect community trust and engagement with law enforcement.

While there are existing studies on this relationship within the field of criminal justice research, most of the findings are subject to selection bias. The widespread usage of 911 call data, which acts as a measurement of incidents observed in a community and that community’s willingness to report those events to the police, presents the challenge of conflicting data analyses among primary investigators. Researchers typically only observe crimes that have been reported to or directly witnessed by police, but decreases in 911 calls can represent either reduced engagement with police or actual reductions in crime–separate conclusions that encourage two vastly different policy strategies. 

To isolate changes in reporting from actual changes in crime, Bruhn and his co-investigators paired 911 call volume with data on acoustically detected gunshots in order to examine call-to-shot ratios. The gunshot data was acquired through a system of fixed-location microphones that many cities across the country have installed to detect and locate gunfire. This data provides a measurement of local crime rates that is less reliant on human reporting. By combining this dataset with 911 call volume, the team was able to determine how likely a community is to call the police for a given gunshot and to gain a survey understanding of how community trust and attitudes change after acts of police violence.

Bruhn and his colleagues found a sharp rise in gunshots coupled with declining 911 call volume across thirteen major U.S. cities in the immediate aftermath of the murder of George Floyd in May 2020. While gunfire spiked following Floyd’s killing and remained persistently high through the end of 2020, the call-to-shot ratio declined by over 50 percent—an effect that is mirrored across multiple cities and racial groups and that persisted over time. A close examination of data from the National Crime Visualization Survey comparing similar incidents suggests that individuals victimized after George Floyd’s murder were less likely to report these crimes due to mistrust of law enforcement and fear of police harassment. Taken together, these patterns indicate that declines in crime reporting are connected to a decline in police trust over the time period.

The study’s results provide novel insight into the effects that high-profile acts of police violence may have on civilian crime reporting. These effects are large, persistent and widespread. If incidences of police violence degrade citizen trust in law enforcement, then these events can result in reduced police efficacy, increased crime, and additional threats to public safety. The study’s findings will fill a significant gap in scholarship on the causal effects of police violence on civilian-police cooperation as well as offer clarity to the discourse spurred by conflicting data conclusions and studies. Bruhn hopes this paper will “put that debate to bed by being definitive on the answer to this question [of police violence’s effect on community engagement].”