The CDC, Like the FBI, Dramatically Undercounts Deaths at Hands of Police

The media do a better job at keeping track of who the government kills than the government does. Go figure.

By “who the government kills,” I specifically mean who the police kill. A new study released this week shows that the Centers for Disease Control (CDC), which tracks stats on causes of death in the National Vital Statistics System, seriously undercounts how many people in the United States are killed during encounters with police.

The report was put together by researchers from the Harvard School of Public Health, who compared the CDC’s numbers for 2015 to The Guardian‘s database of people killed in police encounters in the United States.

Not only did the CDC undercount the deaths, but it did so dramatically, catching only half of them. The researchers counted 1,166 fatalities that year. The Guardian‘s database caught 1,086 of them. The CDC registered only 643.

Justin Feldman, lead author of the study, explained to Reuters that the CDC relies on diagnosis codes in state death certificates that indicate “legal intervention” was involved in the death. But many medical examiners and coroners fail to mention police involvement on these documents.

This seems to happen most frequently with deaths that don’t involve guns (such as a person dying after getting Tasered) and with deaths in less wealthy counties. As just one example, none of the 30 people killed by police in Oklahoma in 2015 were included in the CDC count.

The CDC isn’t the only agency doing a bad job of tracking these deaths. In fact, the Guardian project—and another by The Washington Post that won a Pulitzer—were devised because the FBI’s efforts are so insufficient.

The FBI does have a program to track police killings, but participation is voluntary and many law enforcement agencies do not participate. In some cases, entire states don’t participate. So the FBI, tasked with tracking national statistics on violence and crime, does not have good numbers. They’ve got amazing stats on how many police are themselves killed or assaulted every year. But they don’t have reliable figures on how many people the police themselves kill.

The FBI announced in 2015 that it will work on improving these statistics to make them more reliable, but the bureau lacks the authority to mandate full participation by law enforcement agencies. Civil rights groups would like the FBI to try to tie federal grants to participation in the program.

If we don’t have the most basic information on how many people the police kill, that makes it all the harder to evaluate the circumstances behind these deaths, the trends they might represent, and whether there are changes that could reduce the risks of someone dying.

“As with any public health outcome or exposure, the only way to understand the magnitude of the problem, and whether it is getting better or worse, requires that data be uniformly, validly, and reliably obtained throughout the U.S.,” notes Nancy Krieger, one of the report’s authors, in a press release.

Check out the report out here at PLOS Medicine.