While traffic stop interactions with the police may be shrugged off as brief inconveniences for whites, for black Americans, they can lead to humiliationviolence, and even death. This has become clear over the last few years, as videos have surfaced, hashtags have trended, and reports have been released—opening up the black box of negative interactions between the police and drivers of color for the world to see.

A forthcoming book, “Suspect Citizen: What 20 Million Traffic Stops Tells Us About Policing And Race” adds to that conversation, taking an unprecedented, granular look at the traffic stops in one state.

In 1999, North Carolina became one of the first jurisdictions in the country to mandate data collection at traffic stops. The expressed goal was to suss out disparities in policing. The resulting dataset, which includes information about the demographics of the driver, the offense for which they were stopped, where they were stopped, and the outcome of the stop, was made public. But the state never actually released a comprehensive analysis of this information.

That’s where Frank Baumgartner and Kelsey Shoub at the University of North Carolina, Chapel Hill, and Derek Epp, now at the University of Texas, stepped in. They took on this unfulfilled mission, analyzing data going all the way back to 2002 when the data-collection mandate expanded to include almost all police stops in the state. ”It is pretty much a census of every traffic stop,” Baumgartner said.

In the book, he and his colleagues lay out stark disparity in policing at North Carolina’s traffic stops, and unpack the reasons behind the trends they observe. CityLab caught up with Baumgartner to discuss these findings:

So who was being stopped?

There’s somewhere between a million and 1.6 million traffic stops in North Carolina each year and we [the state] have a population of about 10 million people. That gives us a baseline chance of being pulled over of about 10-15 percent per year. But we found the odds were significantly higher for blacks, than for white and even Hispanic drivers, compared to their respective population shares.

We also looked at a city-by-city comparison of the proportion of whites, blacks, and Hispanics who live in that town to the proportion that they represent in the traffic stop data. Again, we do this with caution, but still it shows that, on average, black drivers are much more disproportionately represented—about 60 or 70 percent more likely to be in that traffic stops data than in the population of that city.

But according to the [U.S.] Department of Transportation, white Americans are more likely to own or have access to a car than black or Hispanic Americans. So we think that the comparison of these populations is not really very accurate—it’s actually more likely to lead to an underestimate of any racial bias.

What about searches?

Our main focus in the book is who gets searched after a traffic stop because being searched is sign that the officer views you with suspicion. Hence the title of the book “Suspect Citizens.” I’m a white, middle-aged college professor, so the last time I was actually stopped for a traffic violation was 40 years ago, in 1974 … and I’ve never had my car searched after a traffic stop in my life. These things are quite rare for people of, for example, my demographic but they’re quite common generally.

We controlled for why you get pulled over, what time of day it was, what day of the week was that, what police agency was it, what month of the year—all of those things. We still saw these very, very significant, robust findings that young people, males, and people of color are much more likely to be searched after a traffic stop.