StreetCred PKIC Data: Traffic Stops

24% of 2015 incidents began on a traffic stop.

Nick Selby

This week, the StreetCred Police Killings in Context (PKIC) dataset launches. You can read about the dataset, its methodology and peer review, and get, work withand create your own analysis products or derivative works from the data, free. This article is one of several deep-dives into the PKIC data.

Traffic stops in 2015 set in motion a substantial minority of incidents in which an unarmed person died after a confrontation with police. While 65% of incidents this year in which an unarmed civilian died after a police encounter began as a response to a 911-call about a violent crime (robbery, carjacking, domestic violence or assault) or property crime (burglary, car theft or vandalism) in progress, and while an additional 7% were incidents in which 911 callers described the decedent as being, “crazy”, “on drugs”, “covered with blood”, “yelling”, or threatening people, it is notable that 31 of 125 (just under 25% of all incidents) began on a traffic stop.

Traffic stops have also been the catalyst for the most controversial of incidents: of five incidents over which indictments were handed down in 2015 involving the death of an unarmed civilian after a police encounter, three began on traffic stops.

We paid particular attention to traffic stops because these incidents were the part of the data in which the police initiated the stops — without request from the community. In almost every other case, police arrived on scene at the request of a citizen.

This was especially important because there were two extremely high-profile incidents in which officers made claims about incidents that occurred on traffic stops that witness video later proved to be untrue. In both those cases, the officer was White, the decedent was Black. Both involved officers have been indicted. The third officer indicted so far this year on an incident stemming from a traffic stop was a white female officer who shot a white male driver.

While many have stated that Black drivers are targeted, some in the press basing this assertion in some dodgy demographic analysis, in fact unarmed Black, White and Hispanic drivers who died after an interaction with police that began on a traffic stop have been stopped in just about equal groups — 10 Black, 10 Hispanic and 9 White.

This may be looked at in a number of ways —while it is very true that, as a group, unarmed Blacks and Hispanics outnumbered unarmed White drivers who were killed by just more than two to one, it is equally true that non-Black drivers outnumbered Black drivers by the same two-to-one ratio.

There are some interesting differences between the White and other drivers: 20% of all deaths of unarmed drivers (regardless of race) involved narcotics (4% methamphetamine). Of the unarmed drivers who died in an encounter with police that began on a traffic stop, White drivers were nearly nine times more likely to have narcotics than Black or Hispanic drivers: narcotics were a factor in 44% (4 of 9) of cases in which unarmed white drivers were killed, compared to one case, or 5%, of Black and Hispanic drivers as a group. Alcohol was a factor in three, or 15%, of cases of Black or Hispanic drivers as a group, compared to one, or 11%, of white drivers.

Of all drivers, 23% had prior convictions for violent crime: 3 of 10 (30%) of unarmed Black drivers, 2 of 9 (22%) unarmed White drivers and 2 of 10 (20%) unarmed Hispanic drivers who were killed by police had a history of convictions for violent crimes.

We do not have data on whether any of the officers involved in White, Black or Hispanic deaths after traffic stops had a history of excessive force complaints. In all, 4 of 125 (3%) officers in incidents in the PKIC dataset had a history of violent behavior. PKIC lowers the bar when making these comparisons: with civilians, we consider prior convictions for violent crime, while for officers, we consider prior accusations. You can read more about our methodology here.

How You Can Help

Disagree with what you’ve read here? Have another take? Download the data here and perform your own analysis. Tell us what you would fix, or how you would do it better.