Hartford — Shortly after a midnight last March, less than a mile from Town Hall, Hartford police stopped a man who was driving on Route 14, right where the Maple Street section runs beneath Interstate 91 and past the Watson Upper Valley Dog Park.
The street would have likely been deserted, with weather records showing a rainy, foggy night at near-freezing temperatures.
The police officer didn’t arrest the driver, or even give him a ticket. The driver was let go with a warning.
But there was something about this particular traffic stop that stood out to a University of Vermont researcher who has reviewed a year’s worth of traffic-stop data for at least 30 police agencies around the state.
Of the 44 people stopped for possible traffic violations by Hartford police during the week of March 13, 43 of the drivers were either ticketed or warned before they were sent on their way. The 44th, the man stopped near the dog park, was the only one who, before he was released, was searched for contraband.
The database of traffic stops used for the research doesn’t list his name or other identifying information, but it does note his race.
He was the only one who was black.
Hartford’s population of people of color is small, but growing, with Census records showing an increase from 3 to 5 percent between 2000 and 2010. In recent weeks, the community has been in the throes of a lively public discussion about how implicit bias affects people of color in the community. Now, Hartford may have just received a prime example of such bias in action: A report by UVM economist Stephanie Seguino has found that white drivers are less likely to be stopped and searched by Hartford police officers than black drivers.
This racial disparity is present in towns throughout the state, but Hartford’s racial gap seems to be higher than the statewide Vermont average documented by Seguino in the report, which she published last month.
Seguino, who testified about her findings before the Vermont House Committee on Judiciary on Feb. 8, also released a report on the same subject for the Vermont State Police last summer. The studies were made possible by a law that required departments to, as of September 2014, begin collecting race data during traffic stops.
She found that, statewide, an estimated 1.6 percent of drivers in Vermont are black, but they account for 2.9 percent of traffic stops. She also found that black people across the state, once stopped, were four times more likely to have their car searched for contraband — 0.9 percent for white people, as compared with 3.6 percent for black people.
In Hartford, the racial disparity is even more pronounced, according to Seguino, who said she updated her data based on numbers recently received from Hartford police.
Here, just 0.9 percent of drivers are black, but they accounted for 3.3 percent of traffic stops during the 12-month period that ended in August 2016. Among those who were stopped, black drivers were at least six times more likely to be searched than white drivers.
When deciding whether to search for contraband, officers can use their own judgment in deciding whether “probable cause” or “reasonable suspicion” exist to justify the action.
In Hartford, black drivers “are searched at a rate that is the highest I have seen for all the agencies I have looked at for Vermont,” said Seguino, though she also said the sample size for searches, but not stops, in Hartford is too small to “make sound statistical inferences.”
“The size of those disparities indicates, however, that they should be monitored closely,” she said.
Seguino found that, despite the state’s leftward political bent, white people across the state are less likely to be stopped, ticketed, searched and arrested.
“As it turns out, racial disparities in policing in Vermont are as high if not higher than in other parts of the country,” said Seguino.
The study also measured rates for Latino drivers, who were generally stopped at a rate higher than white drivers, but lower than black drivers, and Asian drivers, a population whose data on stops was comparable to white drivers.
Seguino said the racial disparity during traffic stops is likely to be an indicator of how police work with the public in other situations.
“Traffic policing is the most frequent contact the police have with citizens, so data from stops provide a window into other police interactions,” she said.
Numbers
The small data size for searches in Hartford was also noted by Hartford Police Chief Phil Kasten, who joined the department in March 2015 from Carroll County Sheriff’s Department in Maryland, a suburban county northwest of Baltimore.
For the entire 12-month period included in the UVM study, Kasten said, only 36 black drivers were stopped, and just four of those were searched. That, Kasten noted, is an average of one black driver search every three months.
However, Kasten said, he values the numbers as part of a broader set of resources that the department can learn from.
“I think it is important that police managers examine information involving contacts between their personnel and the public, including traffic stop statistics,” he said. “The public must trust that we will police ourselves.”
“Overall, I found Chief Kasten’s comments to be very thoughtful and indicative of the fact that he had had an in-depth look at the data,” said Seguino. “This is encouraging and is precisely what we would hope the data would be used for — self-reflection on the part of law enforcement.”
Kasten spoke strongly against Hartford police officers using the race of a member of the public to inform their judgment.
“Profiling and bias-based policing is strictly forbidden by the Hartford Police Department and will not be tolerated,” he said. “We are committed to fair and impartial policing.”
Kasten said the department works to enforce laws designed to protect the public, including people of color.
“Likewise, we will not tolerate the disparate treatment of any resident or guest by another member of the public,” he said. “Those who chose to commit crimes against anyone based on race, ethnicity, gender, disability, religious preference, sexual orientation or other identifiable trait will be met with the full force of the Vermont criminal justice system.”
Still, the two differed on certain points covered in the study.
Kasten said Seguino’s baseline data for black drivers may be off.
“Hartford is unique, at the crossroads of I-89 and I-91, the town offers all-night gas stations at each exit, nine motel/hotels, and the resort village in Quechee that draws vacationers and tourists year round from more diverse areas in Massachusetts, New York and beyond,” he said.
Seguino said the baseline estimate for black drivers comes from DMV accident data on not-at-fault drivers in Windsor County, and said she feels it does apply to Hartford.
Of the 36 stops of black drivers, said Kasten, “the majority of those stops, 24, occurred after dusk and prior to dawn, when lighting conditions make it difficult if not impossible for the driver’s race to have been a factor prior to the stop. The 12 stops of vehicles operated by black drivers during daylight hours is 1 percent — which is more in line with the 0.9 percent ratio of registered black drivers.”
Seguino said conversations with law enforcement officers have taught her that “police frequently set up where they can look at the passing drivers — under city lights or stop lights or even at 90 degrees with their headlights on. ... In any case, these data are an opportunity for police to discuss their practices to ferret out those that may lead to such disparities.”
There are also different ways to look at the search numbers.
One way is to divide the number of stops of black drivers, 36, by the number of searches of those drivers, four, which leads to the conclusion that, once stopped, a black driver has an 11 percent chance of being searched. Using the same method shows that a white driver, once stopped, has a 1.6 percent chance of being searched.
But Seguino came up with her statewide figures using a different process, because a single stop might include two or more incidents, such as a case in which an officer gives a warning for one violation and writes a ticket for another. Seguino used the number of incidents, rather than stops, to reckon that 18 percent of events involving black drivers resulted in a search, as compared to 2.6 percent for white drivers.
By either method, Hartford has among the highest disparities in the state, with black drivers at least six times more likely to be searched than white drivers.
Looking at selective portions of the database of traffic stops can also affect one’s impressions. The black driver who was stopped near the dog park during the week of March 13 was the only driver searched during a 30-day period, but there are also several time periods in which white drivers were searched, or black drivers were not searched.
The study found that implicit racial bias is widespread, but not necessarily universal, in the Upper Valley.
“I think the example of Randolph is a great one — there is no evidence of disparities there, unlike Hartford,” said Seguino. “Why?”
In Randolph, where the sample size is even smaller than in Hartford, the statewide trend is turned on its head: Of 285 stops of white drivers, there were 23 searches, or 8 percent. Of two stops of black drivers, there were zero searches.
Randolph’s acting Police Chief Loretta Stalnaker said she has a reputation as something of a softie within the department, which she joined in 2000.
“I tend to give more warnings than pretty much anyone, anyway,” she said, adding pointedly that repeat offenders get no sympathy from her.
Stalnaker said she couldn’t be sure why Randolph has stood out in this way, and echoed Kasten’s point, that just one or two cases can skew the numbers when looking at the limited data set of a single locality.
But she did say the department officers undergo training for bias-free policiing.
“I’ve been trained to make the decision before you see the operator,” said Stalnaker. “If it’s egregious, you get a ticket. We have to treat everybody the same. We’re a job that just demands that.”
Kasten said Hartford officers go through a four-hour session on fair and impartial policing during basic training, and also went through a refresher training over the summer, in compliance with new state requirements from the Vermont Criminal Justice Training Council. At the same time, the state certified that Hartford has met a long list of requirements in this area, including by, among other things, implementing department-wide policies that consider and guide officers in bias-free policing.
A Bigger Problem
Questions about racial inequality aren’t new in Vermont, or in Hartford — the department has for years grappled with the fallout of a 2010 incident in which Hartford police, responding to an erroneous burglary report that said an unknown man was inside a Wilder home, used a baton to beat Wayne Burwell, who is black and who turned out to be the homeowner.
Burwell filed a federal civil rights lawsuit against police and the town in 2012, and his legal team said in December that they expect him to receive a trial date this year on claims two officers used excessive force.
The town and some officers have been dropped from the lawsuit, as have allegations that Burwell had been discriminated against because of his race.
Questions about how comfortable people of color feel in town came to the fore recently, when a white member of the Hartford Selectboard received calls for his resignation, following his forwarding of an email that included a racist depiction of President Obama and members of his family. In response, the Selectboard formed a committee to develop recommendations on how to combat racial inequality, but the Hartford School Board, on a 3-2 vote, declined to participate.
People of color have said the email, and the vote, reinforce the perception that they are not safe in the community.
Seguino’s testimony to the House Judiciary Committee was part of a broader discussion about how race affects Vermont’s justice system. The committee also heard testimony that black Vermonters are incarcerated at a disproportionate rate, even when other factors, such as socioeconomic status, are accounted for.
During the discussion, the Vermont Criminal Justice Training Council backed 16 different measures recommended by a work group formed by the Attorney General’s Office, many of which sought to increase training for fair and impartial policing for everyone from police academy cadets through department leaders.
“There is a long history of over-surveillance of blacks related to the perception that black men are ‘dangerous,’ a perspective shared with me by more than one officer in Vermont,” said Seguino “Sometimes police are aware of their biases and sometimes not. These data are a good opportunity to interrogate whether such biases influence policing. Chief Kasten’s comments suggests that is precisely what he is doing with the data and that is gratifying.”
One Interaction at a Time
The study doesn’t address whether the racial disparity in traffic stops is the result of officers being too aggressive with black drivers, or too lenient with white drivers. Seguino suggests that black drivers are being held to a higher standard than white drivers.
While the numbers raise serious questions about implicit bias, they’re built on real interactions between individual officers and drivers.
Audrey Devost, a Hartford High School graduate who is now a senior studying psychology at the historically black Howard University in Washington, D.C., said her adoptive parents have always encouraged her to explore her racial identity.
“I never woke up one day and hated the color of my skin,” Devost said. “I always loved being black. My mom made sure I had cultural outlets.”
Devost was also active in school, and knew a couple of police officers through volunteering at a school safety fair.
Still, she said, when she was a high school sophomore, she set her sights on Howard because “I wanted to know what it felt like to be a majority.”
One morning, in 2012, Devost was driving to school, when she got into a fender bender not far from her home.
“That intersection is so busy, and I get scared sometimes,” she said. “It was my fault. I was just freaking out.”
When a Hartford police officer arrived on the scene, she said, contrary to what the study might suggest, “he was really nice. Race was not a factor in that thing at all.”
Another time, Devost said she was pulled over by Hartland police.
“I honestly never felt like I was racially profiled,” she said. “I like to think that I had a good relationship with the police.”
Moving forward, positive interactions like the ones experienced by Devost — both before and during her on-the-road encounters — might be the key to building a relationship of trust between law enforcement, and the broader public.
Devost said that, with the recent high-profile cases of police brutality across the nation, she’s now more aware of the issue than ever before. But she still doesn’t see the study as a reason to avoid her home state.
“Vermont is a great place to be,” she said. “It’s a great place to live.”
Matt Hongoltz-Hetling can be reached at mhonghet@vnews.com or 603-727-3211.