March 13, 2019
A study of nearly 100 million traffic stops across the United States reveals that black drivers were about 20 percent more likely to be stopped than white drivers relative to their share of the residential population.
The study, the largest of its kind, also measured the disparity in stop rates before and after sunset. It found that black drivers made up a smaller share of those stopped at night, when it’s more difficult to discern the race of a driver, which suggests that racial bias may influence stop decisions.
For example, in Texas, about 25 percent of drivers stopped right before sunset were black, compared to about 20 percent just after dusk.
The analysis, released today by the Stanford Open Policing Project, found the same basic pattern across all the stops in aggregate. Overall, the data showed about a 5-10 percent drop in the share of drivers stopped at night who are black.
The study also found that once stopped, black drivers were searched about 1.5 to 2 times as often as white drivers, yet officers found drugs, guns or other illegal contraband on black drivers less often than on white drivers, pointing to a double standard in search decisions.
“Looking at millions of traffic stops across dozens of jurisdictions, we found evidence of widespread racial bias in who is stopped and searched by the police,” said Sharad Goel, an assistant professor at Stanford in the Department of Management Science & Engineering, in the School of Engineering and the executive director of the Stanford Computational Policy Lab.
Goel and Cheryl Phillips, the Lorry I. Lokey Visiting Professor in Professional Journalism at Stanford and director of Stanford’s Big Local News project, co-founded the Stanford Open Policing Project. The project recently trained more than 60 journalists on how to analyze the police stop data in an effort to foster important accountability journalism around policing.
The Open Policing Project filed over 110 individual public records requests to cities and towns across the United States. One of the biggest challenges in compiling these records was the lack of uniform record-keeping. The data for the traffic stops is available at https://openpolicing.stanford.edu/ and is archived at the Stanford Digital Repository.
The researchers also have published a technical report, which goes into detail on a variety of statistical measures for quantifying evidence of discrimination in the traffic stops by both local and state police agencies.
In recent years, incidents related to police stops have captured public attention over possible bias in policing. “This data helps show the evidence behind the anecdote,” Phillips said.
In some cases, it’s clear that police strategies may be geared toward reducing crime but could have unintended consequences of increasing disparate treatment of black and Hispanic drivers.
For example, in an in-depth analysis of traffic stops in Nashville, the team found a high incidence of stops for equipment violations – such as broken tail-lights – which did not appear to reduce crime rates but disproportionately impacted minority drivers.
The project also included collaborators outside Stanford, such as Ravi Shroff, an assistant professor in at New York University, who helped coordinate the collection and processing of stop data from municipal police departments.
For more information about the data and findings, contact Sharad Goel and Cheryl Phillips at open-policing@lists.stanford.edu.