U.S. police officers pull over tens of thousands of drivers every day, making traffic stops the most common way people interact with law enforcement. They’re also a frequent source of tension: Black drivers are stopped and searched at higher rates than White drivers, and Black drivers are more likely than White drivers to say they were stopped for illegitimate reasons. Traffic stops can result in fines or the loss of a driver’s license even in the best cases, and can escalate to violence in the worst.
Many advocates argue that illegal discrimination drives the stark racial disparities in traffic stops — that police simply stop Black drivers more often than White drivers who are behaving similarly. Other observers claim that the patterns come from different underlying rates of traffic violations. These possible causes are hard to disentangle, since we usually only know if someone has broken a traffic law when they’re stopped by the police.
In a new study, we and our co-authors — Justin Kaashoek, Lisa Pinals and Samuel Madden — investigated this question by using data on the movements of drivers collected from mobile phones. We found that drivers spent about the same amount of time speeding in neighborhoods where most residents were White as in neighborhoods where most residents were not. But in all 10 of the cities we looked at, officers focused speeding enforcement on geographically small — and often racially unrepresentative — areas. Individual cities differed, but on average across all 10, speeders were stopped more often in communities of color than in White neighborhoods.
Police don’t always stop drivers where they speed
To understand whom the police ticketed for speeding, we used records of traffic stops that we collected as part of the Open Policing Project. We narrowed the data to cities that were large enough and had detailed enough traffic stop data for us to meaningfully compare how often police pulled drivers over for speeding in different neighborhoods. This gave us 10 cities: Aurora, Colo.; Chicago; Houston; Madison, Wis.; Mesa, Ariz.; Oklahoma City; Plano, Tex.; San Antonio; Tulsa; and Wichita.
We then crossed our police-stop data with the real-time driving data from drivers in those 10 cities. We used anonymous and aggregated information from Cambridge Mobile Telematics on the second-by-second speed and location of hundreds of thousands of drivers who, in total, took tens of millions of car trips in 2019 and early 2020, before the pandemic reduced driving. This unique data set let us determine where and how much drivers in our sample actually sped, regardless of whether they were ticketed by the police.
Combining these data sources let us compare where people were speeding with where police were stopping speeding drivers.
In Mesa, for instance, police pulled over speeding drivers in the city’s western quarter, where 50 percent of residents are White, around three times more often than in the easternmost part of the city, where 75 percent of residents are White. You can see this in the figure below, where the red neighborhoods represent places where police pulled over more speeders relative to the amount of time drivers there actually spent speeding. These high-enforcement neighborhoods are concentrated in the western part of the city, where many of Mesa’s minority residents live.
More of the cities we examined looked like Mesa than didn’t. For instance, despite the fact that rates of speeding were roughly the same all around Chicago, police there pulled over more speeders on the city’s south and west sides, where most residents are Black or Hispanic, than on the city’s north side, where more residents are White.
But the pattern of stricter speeding enforcement in neighborhoods with more racial minorities didn’t hold everywhere. In some cities, such as Houston, police enforced speeding violations more strictly in Whiter neighborhoods.
Why does Houston seem so different from Mesa and Chicago? When we looked more closely at where police pulled motorists over, we found that speeding stops were concentrated in small geographical areas. Across the 10 cities we studied, more than 50 percent, and in some cases more than 75 percent, of speeding stops took place in areas where only 10 percent of the population lived.
Police departments zeroed-in on a comparatively small number of locations to catch speeding drivers. Because American cities tend to be quite segregated, focusing enforcement on a handful of speed traps can lead to racial inequities, even if officers are not intentionally discriminating.
The approach we’ve taken helps us connect behavior and enforcement in ways that weren’t possible even a few years ago. But our results have limits. The 10 cities we looked at might differ from other U.S. cities in important ways. And the drivers in our sample might be safer than average.
In particular, our sample came from drivers who had opted to share driving data with their car insurance companies as a way of measuring how safe they are on the road — and they might not want to share their data unless they thought they were good drivers. Fortunately, we found that our results are largely the same if we use driving data collected from various mobile phone tools like mapping apps that likely don’t have this self-selection problem.
Speed traps and speed bumps
Researchers have suggested many possible strategies to reduce the racial disparities in traffic enforcement. For example, rather than relying on a small number of speed traps, police could spread out their enforcement efforts more evenly across a city, or even adopt randomized policing strategies. Or cities could limit policing traffic violations — and the social and financial costs that come with them — by deterring unsafe driving with speed bumps, roundabouts and other physical infrastructure that slow driver speeds and save lives.
Our research shows that speeding is largely disconnected from race. Policymakers may wish to ensure that the ways their officers prevent speeding don’t have a disproportionate racial impact either.
Johann Gaebler (@jgaeb1) is a PhD student at Harvard University.
William Cai (@iamwillcai) is a PhD student at Stanford University.
Sharad Goel (@5harad) is a professor of public policy at the Harvard Kennedy School.