Mario Villafuerte/Getty Images

A new study from Stanford University has shone a light on how different police officers from state to state treat drivers of different racial backgrounds.

One of the most comprehensive studies of its kind, the project looks at over 100 million traffic stops across the United States, and mines that data for insight into racial disparities in police interactions. While the overall results may not surprise anyone vaguely familiar with America, the depth of the data is still enlightening.

I dug into the report and found a few big takeaways.

A “Threshold” of Suspicion

The researchers had to develop a method to infer racial or ethnic discrimination during a traffic stop in a way that’s nuanced and accurate. Their solution? A “threshold test,” which—as Stanford’s site puts it—centers on one key question: “Once a driver is pulled over, what level of suspicion must an officer have to conduct a search, and how does this threshold of suspicion relate to the race or ethnicity of the driver?”


To put it another way, the study doesn’t just look at how many drivers of different backgrounds were pulled over, but assesses how much evidence officers require before they pull someone over. In this way, researchers could compare whether drivers from some backgrounds required less evidence to be stopped than others.

The Worst and the Less-Worst

The states with the highest racial disparities in police searches were Ohio, Florida, and South Carolina. In each of the three states, officer searches of white drivers were far lower than their black and Latinx peers. But who gets searched the most is a different story in each state. Check out the following graphs—white motorists are represented by an orange line, Latinx drivers by a green line, and black motorists with blue:

Screenshot taken from Stanford Open Policing Project report
Screenshot taken from Stanford Open Policing Project report
Screenshot taken from Stanford Open Policing Project report

Black motorists in Florida are searched at a significantly higher rate than either Latinx or white drivers, while in South Carolina, Latinx drivers are more likely to get searched. In Ohio, however, black and Latinx search rates are much more in line. Both groups are searched almost four times more than white drivers.

The one state where search rates are even close? That would be Washington.

Screenshot taken from Stanford Open Policing Project report

Drivers of color are still searched more often by police officers than white motorists in Washington, but of the states represented, Washington’s disparities aren’t quite so stark.

Latinx Drivers Are More Likely to Be Arrested

By a fractional margin, Latinx drivers were more likely to be arrested nationwide than black drivers. However, both their arrest rates far outpaced that of white motorists—only 2.8% of white drivers were arrested, compared to 5.3% of black motorists and 5.5% of Latinx motorists.

Screenshot taken from Stanford Open Policing Project report

Legalizing Weed Doesn’t End Racial Bias

The Stanford study found that states that had legalized marijuana usage—in this case, Colorado and (as you can see in the chart above) Washington—had a dramatic drop in search rates during traffic stops. This isn’t altogether surprising: As the study notes, since most searches are drug-related, it makes sense that legalizing pot would drive down those kinds of searches.


However, while searches dropped overall, there still remained “a persistent gap” in the threshold for searching white versus black and Latinx motorists. In Colorado, for instance, people of color were searched at nearly triple the rate of white drivers.

Stops of Latinx Drivers Outpaced Stops of Black Drivers in Some States

North Carolina and South Carolina both recorded far more traffic stops than one might expect, given their population. North Carolina recorded nearly double the amount of stops that Florida did, despite the fact that Florida has twice the population.

Officers in the Carolinas also searched Latinx drivers more than their black and white counterparts. In South Carolina in 2011, for example, the likelihood of a Latinx driver being searched was three times higher than a black motorist.

Screenshot taken from Stanford Open Policing Project report
Screenshot taken from Stanford Open Policing Project report

It’s worth noting that the Latinx and immigrant population has grown substantially in both states over the past 20 years. Could that rapid growth, and negative perceptions toward immigrants (especially those assumed to be undocumented), have affected officers’ biases toward Latinx drivers?

Officers Make Mistakes Identifying Race

When reviewing the data for Texas, researchers found they had to correct some of the data on drivers’ ethnic backgrounds because officers had misidentified Latinx drivers for white ones.

Screenshot taken from Stanford Open Policing Project report

The researchers caught the mistakes and re-categorized the data, but they highlighted how easy it is to wrongly categorize people’s racial backgrounds. As the study suggested, in order to ensure accuracy in this kind of reporting, police departments should audit their own data to catch for these errors.