For the first time, traffic stop information for Vermont's local police and sheriff’s departments has been collected and posted online.
One year ago, there was little information available on traffic stops by police in Vermont and the extent of racial disparities in stops, tickets issued, searches and arrests.
There’s long been anecdotal information pointing to racial bias in traffic stops, and the data, posted in response to a legislative mandate, are supposed to help the public, law enforcement and policymakers document and address what’s happening.
A few police and sheriff’s departments have yet to submit their information, but data for all the rest – more than 70 law enforcement agencies – is now posted at the website of the non-profit Crime Research Group for everyone to see.
However, it won’t tell you much.
“A layperson would have trouble making sense of it all or knowing where to go with it,” says Robin Joy, director of research at Crime Research Group in Montpelier. “I think it’s a start for people to get a sense of what’s happening in their jurisdiction, in their town, if they’re interested. But for analysis, it's very limited.”
Joy will be working with a handful of larger agencies to analyze their traffic stop data.
She says for the smaller departments – and there are many – it may take years to collect enough data to analyze for bias.
Joy will focus on traffic stops by Burlington, Rutland, Barre and Bennington police departments in addition to the State Police and Department of Motor Vehicles. Vermont State Police has been collecting its data for a number of years, and it was first made public in 2016, showing clear racial disparities.
To determine if there’s bias, and to what extent it exists, two things are necessary.
First is accurate information about the race of who is stopped and whether they were ticketed or searched.
Second is something to compare that to, in order to determine whether a disproportionate number of minority drivers are being targeted.
In some cases that’s been population data. In other cases researchers estimate the racial makeup of the driving population by looking at accident data. Joy will be using several methodologies to analyze the data, including census data that shows the racial makeup of the commuting population.
In some cases, stops made after dark during commuting hours will be compared to those made during the same hours at the time of year when it is still light; a method called "veil of darkness." Presumably, officers cannot tell the race of a driver before a stop is made during the after-dark hours, and more stops of minority drivers during daylight hours would indicate bias.
Joy says the methodology is important.
“How do we build a consensus of, ‘These results are valid’? That’s very, very important. When you have research that’s out there that nobody buys into, nothing gets changed,” she says.
The method used to analyze data is just part of the equation. The other part is the data itself, which up to now have had problems.
There’s information missing from many of the forms submitted by police departments. If multiple tickets are issued at a single stop, they show up as multiple traffic stops. Non-discretionary stops, such as checkpoints, have also been included when they shouldn’t be. Crime Research Group will work with agencies to correct these discrepancies.
In addition, officers were never instructed on how to collect and enter the information. Karen Gennette, also of Crime Research Group, says compiling data has also been a challenge for police departments.
“It was incredibly challenging for some agencies to extract the data [and] not an easy process for most of the law enforcement agencies” she says.
Even when it’s been collected properly, the data the Legislature requires isn’t necessarily what analysts are looking for. It’s not what Crime Research Group is using in its analysis.
Gennette says her organization is working with law enforcement agencies to improve how traffic stop information is collected make it easier for them to submit it.
Takeaways and next steps
Gennette and Joy stress that despite the problems, the data are valid.
University of Vermont professor of economics Stephanie Seguino agrees. In January, she released her analysis of traffic stop data collected by 29 Vermont police agencies.
“The data are very robust to me, and it tells me that slight changes in measurement or improvement in the quality of the data are probably not going to change the bottom line,” says Seguino.
The bottom line in Seguino’s study: Black drivers in Vermont are twice as likely to be arrested after a traffic stop than white drivers. And black and Hispanic drivers are three to four times more likely to be searched after a traffic stop than white drivers, even though white drivers are more likely to be found with contraband.
Seguino looks closely at actions taken after a stop is made, once an officer knows the race of the driver.
The analysis isn’t the final step in the process. State officials and law enforcement agencies will have to decide what action to take based on the findings.
Seguino says several police agencies are already taking steps to examine their practices based on her findings. But that’s not the case with everyone.
“What I hear from some chiefs is that they do not accept that the data reflect actual racial disparities. A number have asserted that the disparities are justified, based on context, and so they don’t show any indication to me of a willingness to look any deeper into the data and causes of the disparities,” she says.
A move by lawmakers this session to establish a Racial Justice Oversight Board is designed to make sure steps are taken to address bias in the criminal justice system.