Letter to the editor: Bias study had too many flaws
In an article in the Addison Independent on Feb. 2, “Local Chiefs Contest Racial Bias Report,” Police Chief George Merkel of Vergennes and Police Chief Tom Hanley of Middlebury, express some misgivings about the results of a recent racial bias study of local and state policing. I have similar misgivings.
The study was conducted by Stephanie Seguino of UVM and Nancy Brooks of Cornell University, and titled “Driving While Black and Brown in Vermont.” It can be found online. This kind of a study is critically important for us to be able to understand race bias and color of skin prejudice in our communities.
Racial bias is very personal to those on the receiving end, and there is no need for statistics for them to understand how they are being hurt. However, for the wider community statistics are important to be able to understand the magnitude and source of the problem, especially for those who are white or light and not the subject of discrimination.
I grew up in an integrated environment and elementary school; my best childhood friend and still even today, is Black; I went to integrated summer camps; taught in integrated schools; was a Peace Corps volunteer in the Philippines; and continue to work in an integrated situation.
I feel at least some understanding of what bias and prejudice are and how people are hurt. I see racial prejudice in Vermont, both of individuals and institutionally. We need to work to eliminate racism, bias, prejudice, on personal levels and all reaches of our society, and to embrace all of the diversity found in Vermont. And that is why this study is so important.
However, there are many aspects of this report that are troubling. First, the criticisms by the two police chiefs should be listened to closely. I don’t know about Chief Merkel, I do know that Chief Hanley has been working to eliminate bias from Middlebury Police Department for years and I think has been pretty successful. Their point that the reasons and conditions and situations for traffic stops should part any of such study are valid.
As much as I think this sort of study is needed I think that it does more harm than good. The author’s basic comparisons of drivers being stopped are based on false assumptions.
The study attempts to compare “stop rates by race compared to racial shares of the population.” Their question is, “are Black or Hispanic drivers stopped at a higher rate than other members of the population?” The other members of the “population,” in the studies’ terms are, “white, Asian, or Native American.” The study finds that Black and Hispanic drivers are stopped at a higher rate. But the problem is that the population they are compared to is the general population of the geographic area a particular police department is located in, and not the actual population of drivers on the road.
The study uses the American Community Survey, conducted by the U.S. Census Bureau to ascertain racial makeup. This survey simply records the make-up of the population of specific communities, people self-identified by race. No questions about, do respondents to the survey have a drivers license? So this cannot determine the race of drivers on local or major highways of whatever community is being studied.
In addition, the study uses data from the Vermont Department of Motor Vehicles, “not-at-fault drivers in accidents to determine race of drivers.” The authors don’t explain the logic of this, but do admit, “37 percent of the data about race is missing from officers’ accident reports.” This cannot be a reliable way to determine the race of drivers on Vermont roads.
For the most basic questions to be answered the actual driving population must be known. You cannot just simply compare to the general population of the town, county or area of the traffic stop.
Unfortunately the study uses these two methods to establish the race or ethnicity of drivers. Both methods are flawed and the authors admit this, but carry on with their study anyway. I submit that without reliable data about the actual population of drivers, it is not possible to accurately ascertain the race of drivers stopped compared to the actual driving population. Therefore I don’t think the authors’ conclusions are accurate, that “Blacks and Hispanics are stopped at a higher rate than white Asian or Native American.” This question remains unanswered.
A high percentage of drivers on major roads through communities, such as 22A through Vergennes or Route 7 through Middlebury, are drivers from outside the community, from diverse places, drivers passing through, and most likely a very diverse population of people that does not accurately reflect the population of a town or Addison County.
To compound the reliability of the study, police officers are asked, by new Vermont law, to determine the race of drivers who are stopped. This means when an officer is standing on the side of a busy road, with traffic zooming by, and it’s raining, and it’s nighttime, the officer is expected to make an accurate determination of the driver’s race. I don’t think it is realistic, even under the best of circumstances, to accurately determine a person’s race, that is white, Black, Hispanic, Asian, or Native American. Vermont driver’s licenses do not have race listed on them.
When stops are made, in how many cases does the police officer accurately know the race or ethnicity of the driver before the stop is made? It stands to reason that there are many incidences where race of a driver is not known such as, at night, inclement weather, when detected by radar from a long distance, when following a car violating law, and others. This issue was not taken into account by the study.
Unfortunately the study makes it sound like there is a great deal of bias in Vermont policing. There probably is some, but the study doesn’t measure it. I think it incumbent upon the authors to get a better idea of the actual identities of actual drivers so their comparisons are valid.
The authors also need more data. The current study is based on only one or two years of data, which seems quite small for the far-reaching conclusions drawn.
The study goes on to look at what happens after a stop, to arrest rates, whether or not contraband was found and gender of stopped drivers. But it does not research why there are differences in these, or the circumstances, or the outcomes. The tone of the study is only of dry statistics. The are many variables of any stop that any study must take into consideration when analyzing whether or not race or bias is involved.
I hope the authors will come forward to tell us of the studies’ limitations and inaccurate assumptions and tell us they will correct these deficiencies. And I hope they continue their work because we need factual and accurate information in order to move our society toward inclusion, reduce bias, and be standing on the side of love of our fellow travelers.
Paul A. Stone
MIDDLEBURY — Bernard D. Kimball, 76, passed away in Bennington Hospital on Jan. 10, 2023. … (read more)
The Fresh Air Fund, initiated in 1877 to give kids from New York City the opportunity to e … (read more)