Column: We need a lot more data on police abuse. Here’s why


When the killing of Michael Brown by a white police officer sparked weeks of unrest in Ferguson, Mo., in 2014, many Americans were shocked to learn for the first time that there was no accurate, comprehensive data available on shootings by police officers around the country.

Despite decades of complaints about police brutality, despite allegations of disproportionate violence against people of color and stories of trigger-happy cops shooting unarmed Black men, the FBI, which compiles crime statistics, did not require local police to provide the bureau with complete information.

That gaping lack of data meant that Americans were in the dark about the scope of the problem — and that police were that much less accountable for their actions.


That’s when the Washington Post, to its credit, stepped in, assigning a team of researchers, reporters and data editors to gather information about every fatal shooting by police, including the race, gender and age of the person killed and the circumstances of the shootings.

Last week, as the Black Lives Matter protests continued and the debates over defunding the police were beginning, that information became painfully relevant again, and the Post updated its analysis of what it has collected in its now 5 ½-year-old database.

There is a lot we know now primarily because of the Post’s efforts: About 1,000 people are shot and killed by the police each year. Black people are far more likely to be fatally shot by the police than white people. Some 30% of those shot since 2015 were unarmed. And those unarmed people were also disproportionately likely to be African American.

We know that white people — who make up 60% of the U.S. population — accounted for 45% of the 5,400 people shot and killed by the police since 2015. Black people, who are 13% of the population, made up 23% of those killed. Latinos are 18% of the population and 16% of those killed.

In fatal police shootings of unarmed people, a Black person is four times likelier to be a victim than a white person.

But there’s also a lot the Post’s numbers don’t tell us. Several other organizations now have data-gathering projects underway. And the Los Angeles Times has gathered information about police killings in L.A. County as part of its ongoing Homicide Report project. But much more data is needed if we hope to fully understand and eradicate police violence in communities of color.


The Post’s database, for instance — as useful as it is — covers only fatal shootings by police. That means it doesn’t include the death of George Floyd, who was not killed by a bullet but by a police officer’s knee to the neck. Or those of Freddie Gray or Eric Garner or Sandra Bland, to name a few, none of whom died in a shooting.

The Post’s database also doesn’t include information on police killings committed off-duty, nonfatal shootings by police or other uses of force that don’t result in death.

The organization Mapping Police Violence and several others try to gather a broader range of information than the Post — but again, lack of reporting by police agencies makes it extremely difficult to compile accurate numbers, especially when you move beyond shootings.

Unsurprisingly, police departments are often unhelpful. Steven Rich, the database editor for investigations at the Post, told me that police agencies often deny public records requests, making it impossible to find out the race of the victim or whether that person was armed or unarmed. More than 520 of the 5,400 deaths in the Post’s database are categorized as “race unknown.”

In order to really understand policing and criminal justice in the United States — and to pinpoint inequities — we need laws that require far more information to be supplied to a central repository, open to the public.

The police reform legislation introduced by Democrats in Congress last week in the wake of the Floyd killing would advance the cause of transparency by requiring local police departments to send data on the use of force to the federal government. But whether the bill can pass in the Republican-controlled Senate remains uncertain.

In addition to data, we need thorough analysis to understand police shooting numbers in a broader context. CNN reported last week, for example, that American police typically shoot, arrest and incarcerate far more people than police in other developed nations. In 2018, the rate of police shootings in the U.S. was 31 per 10 million of population, compared to 1 per 10 million in the U.K. and Germany, the network said. (CNN relied, in part, on the Post‘s database for its analysis.)

Numbers alone can’t explain why unarmed Black people are being shot at disproportionate rates, but they can help. There is a huge body of research showing that police are more likely to stop people of color and more likely to arrest Black suspects than white suspects, according to Justin Nix, a professor of criminology at the University of Nebraska at Omaha. Those extra encounters make lethal outcomes more likely. There’s also a greater police presence in many low-income minority neighborhoods, which can lead to more encounters and more shootings, especially when the nature of the policing is flawed.

Implicit bias and explicit racism, of course, play into all this. We don’t just feel that police are quicker to resort to force and more likely to use more severe levels of force when dealing with a Black person than a white person. There’s a large body of research that proves it.

From the data we have, we know categorically that police here in the U.S. kill more people than police in other countries, and that they kill Black people disproportionately. Now we need the data that will help us fashion the right reforms to end those disparities.