On Tuesday, South Carolina police officer Michael Slager was charged with murder in the shooting death of Walter L. Scott, after video evidence clearly showed Slager firing eight shots into Scott’s back as he fled. That’s extremely uncommon. Last month, The State newspaper in Columbia, South Carolina, identified 209 suspects in the state in the past five years who were fired at by police; 79 died. The State could find only three police officers who were charged with a crime in connection with any of the 209 shootings.
What isn’t unusual is the race of the victim. Of those 79 people in South Carolina who were fatally shot by police, 43 percent (34 people) were African-American. That’s a higher proportion than the African-American share of the state’s population, which was 29 percent in the 2013 census population estimates.
Assessing those disparities is difficult without knowing the racial breakdown of people involved in violent interactions with police, or while police are present — and therefore are at higher risk of being hurt by police. Whatever the popular impression of that racial breakdown may be, there are no statistics that measure violent confrontations with the police by race. We do know, however, that the proportion of arrests that are of African-Americans is close to the proportion of killings by police that are of African-Americans. (According to the FBI, 28 percent of people arrested in the U.S. in 2013 were African-American. To compare, according to data from a project called Mapping Police Violence, 30 percent of people killed by police in the U.S. from August 2013 through March 2015 were African-American. The percentage of people arrested in South Carolina in 2012 who were African-American was 44 percent, compared to 40 percent of those killed.)
But arrest rates reflect police decisions on enforcement as well as underlying crime rates and do not mean that blacks are disproportionately likely to commit crimes and put themselves in harm’s way. For example, African-Americans are far more likely than white Americans to be arrested for marijuana possession — 3.7 times more likely in 2010, according to the American Civil Liberties Union — even though the groups use it at similar rates.
Mapping Police Violence can help us get at racial disparities in police killings another way. It breaks down killings by whether the victim was armed or unarmed at the time of killing. In 2014 and March of 2015, Mapping Police Violence counted 297 people killed by police around the country who were unarmed.1 Of those people, 117 were African-American, 167 were not, and the project couldn’t identify race for 13. That means 41 percent of unarmed people killed by police during that time in the database (with an identified race) were African-American, far out of proportion in a country that was 14 percent African-American in 2013. Among people who were armed when killed by police and for whom researchers had race data, 25 percent were African-American.
That disparity can be seen in a variety of statistics about police fatalities.
The data collected by Mapping Police Violence shows that 31 people were killed by South Carolina officers from August 2013 through March 2015; 30 had an identified race. Twelve victims of the 30, or 40 percent, were African-American, 11 percentage points higher than the portion of the state that was African-American in 2013.2 That puts South Carolina squarely in the middle of other states in terms of the gap between the percentage of police shooting victims who were black and the African-American share of the population, as this table shows.
The Mapping Police Violence project arrived at these numbers by merging and checking several databases, primarily the one called “Killed By Police,” which we audited last year and found to be a mostly reliable aggregator of news reports. Researchers with Mapping Police Violence check each report and exclude those that are ambiguous.
The project counted 1,913 total deaths between August 2013 and March 2015, an annual rate of 1,150 deaths, similar to a recent estimate by other researchers for the annual rate in 2003 through 2009 and 2011. For about 60 percent of the 1,913 deaths, the databases used by the Mapping Police Violence researchers had information on the victim’s race. They managed to find race information for another approximately 30 percent of killings from obituaries, social media, news reports and criminal records, according to Sam Sinyangwe, a statistician, researcher and advocate with Mapping Police Violence.
That leaves about 10 percent of people killed by police without an identified race; they are excluded from the table above. These victims aren’t distributed evenly across states, which could skew the percentages. For instance, if a city with a high African-American population doesn’t report the race of people its police agency has killed, the proportion of victims in that state who are African-American may be artificially low in the data.
“Where you live very much determines your likelihood of being killed by police,” Sinyangwe said in a telephone interview. He said the data raises the question, “What are the underlying policies and practices, or the culture, of police departments that are allowing for this and facilitating these higher levels of police violence? And the converse is, what is going well in these other states, and what can we learn from that?”