Fatal encounters with police should be documented nationally

Police stood guard as demonstrators marked the one-year anniversary of the shooting of Michael Brown, in Ferguson, Mo.
Police stood guard as demonstrators marked the one-year anniversary of the shooting of Michael Brown, in Ferguson, Mo. Scott Olson/Getty Images/File 2015

Just how many people do American police kill? Four years after protesters filled the streets of Ferguson, Mo., to protest police violence, we finally have a pretty good idea: about three a day.

Law enforcement officers are responsible for 1 in every 12 adult male homicides, a number that rises to 1 in 10 in rural areas. That brings the number of people killed by police to about a thousand a year. Last year, 46 officers were murdered in the line of duty, and almost the same number were killed accidentally.

These figures were revealed in a study published last month in the American Journal of Public Health, which concluded that “contact between civilians and law enforcement exposes individuals to a nontrivial risk of premature death.”


But the researchers, led by Cornell University’s Frank Edwards, also concluded that the number of deaths is twice as high as previously estimated.

Why the discrepancy? Until a few years ago, no one kept track of police killings — and to this day, there’s no official count kept anywhere.

The statistics the researchers drew on for their study didn’t come from any government source; instead, they were compiled by a single journalist, D. Brian Burghart. A former Reno News & Review editor, Burghart started tracking police killings on his website, fatalencounters.org, in 2013, after becoming frustrated with what he called “a complete lack of data.”

Burghart compiles his database largely by scraping news reports, but his numbers might be the most accurate that exist. Law enforcement officers are not required to keep a count of people they kill, nor are they required to report those deaths to any federal agency.

Considering the extent of fatalities revealed in the AJPH study, it seems only reasonable that such accounting be mandatory. Police departments should be required to report any deaths caused by their officers to the Department of Justice — and that data should be readily available to police and to the public.


Without accurate data on police killings, there’s no opportunity for police departments to use that data to improve how they handle dangerous situations.

As Burghart writes on his website, “I can think of many ways this information could be used by law enforcement to enhance their interactions with individuals that result in fewer people dying. . . . How does any agency know whether its own use of deadly force is greater or less than any other if there is no way to compare? Are there predictable situations that could keep officers out of harm’s way that could be learned from incident comparisons? And why do different jurisdictions whose officers train at the same police academy have different outcomes to similar situations?”

Effective police reform will remain impossible until accurate data on police killings are easily and widely available.