You can’t watch or read the news without encountering a story about police violence toward people of color, complete with graphic video footage. Often these violent encounters spark large protests and sometimes riots in the cities where the violence occurs. This week, especially, the topic of race and law enforcement is at the forefront of our national consciousness because jurors voted to convict Minneapolis police officer Derek Chauvin guilty of killing George Floyd.
Activists have been calling for police reform for years. Their proposals offer a variety of changes, including more officer training, establishing new law enforcement tactics, defunding and even abolishing police departments.
But for public policy researchers and government officials who will design and implement these changes, there is a looming problem: there is a lack of comprehensive, bias-free data on police shootings in the U.S.
The U.S. Federal Bureau of Investigations (FBI) began collecting data on police officers’ use of force against civilians in a national database in 2019. But participation is voluntary; law enforcement agencies are not required to report use-of-force incidents. In its first year collecting data, the FBI reported that participating agencies represented about 40 percent of police officers in the nation.
Because of a lack of federal data, media outlets and non-profit organizations have begun collecting their own national data on police shootings. The Washington Post began a crowd-sourced database in 2015 that categorizes police violence by age, race, weapon and geographic location. Another project, Mapping Police Violence, compiles information from three other crowd-sourced databases. But crowd-sourcing any type of data can lead to confirmation bias, when people allow their personal opinions – either consciously or unconsciously – to influence what they report. It also leads to sampling errors. (The data are not likely to be representative of the entire country, in this case.)
Based on the best data available, researchers estimate that Black Americans are two-and-a-half to three times more likely to be killed by police compared to White Americans. A study conducted by Princeton statistical researchers found Black people fatally shot by police were twice as likely as White people to be unarmed.
The limited data available point to a serious problem with policing in America, but they don’t provide a clear picture of what’s happening in U.S. cities every day. For example, they don’t quantify situations where police officers use unwarranted force but the civilian is not seriously physically harmed or killed, or whether police officers unfairly target minorities.
In a study published this year, researchers at Princeton University demonstrated that studies often underestimate racial bias in policing or even mask discrimination completely. In other words, the problem may be much worse than it appears in these studies.
Other researchers have approached the topic from different perspectives. For example, a working paper published earlier this year by the National Bureau of Economic Research analyzed more than 2 million 911 calls made in two American cities. They found that White officers were 55 to 65 percent more likely to use force compared to Black officers, when responding to similar calls. They also found that White officers increased their use of force significantly when they were sent to primarily Black neighborhoods. (Black officers increased their use of force moderately in these neighborhoods, but not as much as White officers.)
On the whole, the authors found that interacting with an opposite-race police officer resulted in a 30 to 60 percent increase in the use of force.
When it comes to use of firearms, White officers were twice as likely to use gun force overall compared to Black officers. Opposite-race police officers dispatched to a 911 call were five times more likely to fire their guns in these incidents.
The authors point to multiple limitations in their study. First, they only studied 911 calls, not other interactions between officers and civilians. In addition, in the cities included in the study, a disproportionate amount of crime occurs in minority neighborhoods, which may skew the results.
The take-home message: the data on racial bias in law enforcement and on police shootings in American are incomplete. However, the limited data available on this complex topic point to a serious problem in how law enforcement officers react to and treat racial minorities.
Visit Cornell University’s Bronfenbrenner Center for Translational Research’s website for more information on our work.
Speak Your Mind