Race and Police Killings: Additional Thoughts
Back in October, I went through the FBI's Supplementary Homicide Report (SHR) in search of racial bias in police killings. I found that the overall ratio of blacks to non-Hispanic whites shot by police was 0.82:1, which is high given that blacks are just 13 percent of the population -- but I also found a ratio of 1.5:1 among homicide offenders. My results were similar last month when I focused on data collected from big-city departments that, according to a Wall Street Journal report, actually seem to report police killings in a consistent fashion. In other words, there is no evidence in these data that police officers are targeting blacks for fatal violence; the disparity can be explained by the disparity in rates of serious violent crime.
Here I'd like to highlight two new reports that support my findings.
First, Kevin A. Hassett of the American Enterprise Institute conducted an analysis similar to mine, except that instead of homicide-offending data, he used overall violent-crime data drawn from the National Crime Victimization Survey. Here's the number of violent crimes committed by each group divided by the number of police killings of that group:
Here we have a closer call than we had with my data, but even this milder adjustment (the racial disparity in overall violent crime is less than the disparity in homicide offending) brings the police-killing data into focus, revealing little if any bias.
And today the Washington Post has a report about cases in which police officers are killed. As I wrote last week, this is a rare occurrence, so the numbers can lurch around from year to year -- but a big advantage of these data is that they focus on fatal confrontations with police instead of violent crimes more generally, which could make for a better comparison with data on people killed by police. Here's what the Post found:
There were 511 officers killed in felonious incidents and 540 offenders from 2004 to 2013, according to FBI reports. Among the total offenders, 52 percent were white, and 43 percent were black.
This suggests a black-to-white ratio of 0.83:1, which is a pinch higher than my killed-by-police ratio. Further, due to data limitations, Hispanics are not excluded from the white data, as they were in my analysis. ("Hispanic" is an ethnic category, while "white" and "black" are racial categories -- and they're not mutually exclusive -- so it's a tricky problem for government agencies.) In Census data, about 19 percent of whites are also classified as Hispanic, and the number is similar for whites identified as homicide offenders in the SHR. Making this adjustment brings the ratio up to about 1:1. (Remember, these numbers aren't adjusted for population size. Also, many departments don't report ethnicity data in the SHR, so Hispanics are undercounted there.)
Essentially, once you make any plausible attempt to adjust for violent-crime rates, the disparity in police killings disappears. This has now been shown with homicide data from the Supplementary Homicide Report, violent-crime statistics from the National Crime Victimization Survey, and cop-killer numbers from the FBI.
Robert VerBruggen is editor of RealClearPolicy. Twitter: @RAVerBruggen