Over the past few years, we’ve seen how a minor traffic violation can turn deadly. The deaths of Sandra Bland in 2014 and Philando Castile last year, for example, were both preceded by fairly routine police traffic stops.
These instances, as egregious as they are, only represent a small portion of the over 20 million traffic stops made each year nationwide. They make the news because they end with the worst-case scenario. But what about traffic enforcement in general? How does racial bias play out with things like traffic tickets, searches, and arrests?
To answer this question, a group of researchers at Stanford University have spent the past couple of years gathering and analyzing state-level records for traffic stops from all 50 states. After analyzing the data, the team built the site Open Policing to house not only their findings but also the entirety of the data they collected—which consists of over 100 million stops from across 31 states. An interactive map consolidates the data and displays it per state, so that people can easily access and compare that information in areas across the U.S.
These results are sadly unsurprising; the fact that black people get pulled over more than white people has long been a reality of American life. However, proving that these traffic stops were made because of police bias—rather than appropriate policing, for example—is more difficult. To prove bias, the Stanford researchers had to disentangle discrimination from effective policing through statistical analysis.
Stanford started the project in 2015 by analyzing the rates at which police stop drivers across the country. They found that officers stop black drivers at a higher rate than white drivers, regardless of the driver’s age or gender. They then analyzed the rates at which police ticket, search, and arrest drivers across race and found that black drivers are 20% more likely to get a ticket than white drivers, while Hispanic drivers are 30% more likely to be ticketed than white drivers. Both black and Hispanic drivers are twice as likely to be searched compared to white drivers.
Those results clearly point to a disparity in police practices. They do not, however, indicate discrimination, says Sharad Goel, an assistant professor in management science and engineering at Stanford, in the project video. To test for biases, the researchers needed to look at the rate at which drivers were searched compared to how often the police actually found contraband. They found that if officers search white drivers, there is a 10% chance or greater that they will find contraband. But search black drivers, and there is only a 5% chance or greater that they will find contraband.
This disparity indicates that racial discrimination is a factor in traffic stops. Stanford has released all of the data, as well as the code used to analyze it, on its Open Policing project site. It’s hoping that journalists, policy makers, and data scientists will use their consolidated records to work toward eliminating racial bias in the police force. Access it here.