Last spring, analyses of five years of data revealed clear racial disparities in Vermont State Police traffic stops. But after conversations with the troopers whose stops showed the greatest disparities, state police officials say they’ve found no instances of implicit or explicit racial bias.
The conclusion illustrates the challenge of pinpointing the problem, and advocates say the police need to do more to identify bias in individual officers.
While there were issues with the data that were collected, no one questioned the analyses, which showed a disproportionate number of black and Hispanic drivers were pulled over.
One five-and-a-half-year study was produced by the Institute on Race and Justice and the Center for Criminal Justice Policy Research at Northeastern University.
In another study, University of Vermont economics professor Stephanie Seguino analyzed the same state police data with an emphasis on what happens after a car is pulled over. It’s at that point that an officer can be more certain of the race of the driver.
Seguino found that particularly for African-Americans, but also for Hispanics, there was a significantly higher likelihood of a driver being ticketed, arrested or searched compared to white drivers. Yet, when searches were conducted, white drivers were more likely to have contraband.
The data included a breakdown by individual troopers, who were not identified publicly.
This allowed Capt. Ingrid Jonas, the agency’s fair and impartial policing director, to talk one-on-one with troopers who were what she calls "outliers."
Jonas says her conversations revealed reasonable explanations for the racial disparities in the data on individual troopers.
“I have not walked away from those interviews, those discussions, with a sense that this person might have an unconscious or even a conscious problem with bias,” she says.
The explanations often had to do with problems with the way in which the data were collected.
Jonas says other disparities could be explained by a relatively small number of total stops by a trooper; where stops of one or two minority drivers could show up as a large percentage.
There were other instances where it seemed to her that bias couldn’t have played a role.
In one example, she says a trooper made all his stops at night. He parked along the interstate in Chittenden County and ticketed every driver going above a certain speed.
“I feel like that’s a reasonable enough explanation for 23 of those 200-and-something car stops to have been black operators,” says Jonas.
She says her interviews don’t prove there’s not bias – only that it’s difficult to identify.
Col. Matthew Birmingham, who heads the Vermont State Police, makes the same point.
“What’s important for me is that we identify any trooper in the state police that’s acting on that bias. But it’s a very complicated process trying to identify both implicit and explicit bias, implicit being the most challenging because that’s conduct and behavior that the person themselves don’t even realize they’re doing,” he says.
If there’s a better way to pinpoint bias, Birmingham says he’s open to suggestions.
Lawyer Robert Appel, a former Vermont Defender General, says it takes more than a conversation to reveal bias.
“When confronted with, ‘Is it possible that race was an improper factor in your conduct?’ I don’t think you’re going to get too many people who say, ‘Yes, it was.’” says Appel.
Appel, who is a member of the Fair and Impartial Policing Committee, which advises the state police, says the agency could turn to people who are better qualified to detect bias.
“It may be helpful to have outside consultants, experts assist in that process. Maybe there’s other people who have been engaged in this particular work for a longer period of time that could either be involved with the consultations or coach the consultations to have them have a broader impact.”
Jay Diaz of the Vermont Chapter of the American Civil Liberties Union agrees that detecting implicit bias is difficult and takes training.
“I’m not aware of any qualifications that Capt. Jonas has that would qualify her to make determinations about whether someone has implicit bias,” Diaz says. “In terms of litigation that the ACLU has brought, officers have repeatedly come up with certain justifications or pretexts for why they did something when it was clear — and even courts have ruled — that they acted upon a bias.”
Three years ago, the Vermont ACLU filed a suit against the state police over a traffic stop of an African-American driver made by trooper Lewis Hatch. Hatch is no longer a trooper.
Diaz says the challenges the state police face in identifying bias underscores the need to create an independent racial justice oversight board to monitor the criminal justice system, which is a step being considered by the Legislature.
Diaz and Appel give the state police high marks for acknowledging bias, creating a command level position to deal with it and taking steps to improve recruitment, screening and training.
But they say police also have a responsibility to hold individuals accountable.
“Until we uncover and try to unpack our implicit biases, it’s going to have an impact on the work we do,” says Appel.
Birmingham says that now that some of the problems with collecting data have been addressed, more recent data from 2016 show fewer racial disparities in traffic stops.
He says the focus on impartial policing is a process.
“I’m not going to plant a victory flag when we reach a certain number. This is an ongoing process that never stops,” says Birmingham.
He says supervisors randomly review cruiser dashboard camera footage of traffic stops and he’s pushing to acquire body cameras for troopers, which will help efforts to combat bias.
Birmingham says the agency has also designed a process used by hiring panels to screen for bias in prospective troopers.
As for those already in uniform, "Because they’re already employed, we can’t put them through hiring panels, but we can certainly engage them in conversations," he says. "I believe it’s been very beneficial."