Some bias is evident when ticketing speeders in Burlington, Vermont

By Michael Brenneis

The negative safety effects of speeding are well established. The enforcement of speed limits is justified to reduce crashes. But does officer discretion when giving tickets result in bias against one group or another? The results of an analysis of speeding stops in Burlington, VT, show that young drivers, male drivers, and drivers belonging to what the researchers termed a non-white “minority” group are more likely to receive a speeding ticket, rather than a warning.

Traffic stop data from the Burlington Police Department (BPD) was collected for a six-year period between 2012 and 2017. Of 33,874 traffic stops, 4,089 stops for speeding included data regarding age, race, gender, speeding range, and geographic coordinates. Some aggregation was done over age ranges (25 and under, between 25 and 65, and 65 and older), and with speeds (0-10, 11-20, and 21+ mph over the limit), and the researchers merged the 12 percent of speeding stops involving a non-white driver into a “minority” category. Eighty-eight percent of the drivers stopped were white. Twice as many male drivers were stopped for speeding than female drivers. Most drivers fell into the young or middle age categories. Most speeders were clocked going 0 to 20 mph over the limit.

According to the analysis, drivers belonging to the “minority” category were 42.6 percent more likely to be ticketed than white drivers stopped for speeding. Drivers 25 or younger were about 1.5 times more likely to be ticketed than those over 25. Of the 122 drivers stopped for exceeding the limit by more than 20 mph, 40.2 percent were 25 or younger, while 55.7 percent were between 25 and 65.

Stops made between 8 a.m. and 9 a.m. were more likely to result in a ticket than a warning, and those stops that occurred in December were more likely to result in a ticket than stops made in other months. We’re left to draw our own conclusions, but there may be some resource management logic in concentrating enforcement when the most drivers will be on the road; and one wonders if BPD may have a year-end target to meet?

The researchers point out that their results may be dependent on a region’s demographics and geography, and won’t necessarily be directly applicable to other places

The autologistic regression model used in this analysis identifies spatial hot-spots in the ticketing data, which, perhaps with further study, could be used to identify the road characteristics that contribute to the propensity to speed. Hot-spot analysis could be useful locally to guide road redesign and speed enforcement.

Analyzing traffic stops by location through an equity lens could help guide policy, road design, and law enforcement training. Automated enforcement of speeding and red light running has been shown to be an effective deterrent, and potentially reduces systematic bias. Roads designed to move cars as rapidly as possible lull drivers into a speeding complacency. Many communities are adopting practices that put safety, and accessibility for all modes, ahead of speed. Studies like this one could be used to prioritize locations for road redesign. And while we wait for that to happen, we can at least do a better job of enforcing speeding infractions equitably.

Michael Brenneis is an Associate Researcher at SSTI.