The question in the title of this post is prompted by this latest new article in the on-going terrific Law.com/Legaltech News series unpacking modern risk assessment tools. The headline and full subheadline of the piece reveals why it prompts this question: "Constitutional Brawl Looms Over How Risk Assessment Tools Account for Gender: Researchers say that scoring men and women differently is essential to account for risk assessment tools’ inherent gender bias. But it’s an open question whether these adjustments are violating state or constitutional law." I recommend the piece in full, and here are excerpts:
While there’s been a lot of focus on how risk assessment tools treat different racial demographics, little attention has been paid to another issue that may be just as problematic: how gender factors into risk scores. Researchers say accounting for the differences in gender ensures that risk assessments are more accurate, but exactly how they do so may run into constitutional challenges.
Legaltech News found one gender-specific risk assessment tool currently implemented in at least two states: the Women’s Risk and Needs Assessment (WRNA), which like the the Ohio Risk Assessment System (ORAS), was created by the University of Cincinnati. Kansas uses the WRNA to assess parolees’ risk in a women’s prison in Topeka, while Montana deploys it for women on probation or parole throughout its Department of Corrections....
The reason WRNA is needed in the first place is because most risk assessment instruments are validated (i.e. created) on a population that is majority male, in large part due to current gender imbalance in the criminal justice system (i.e. more men than women commit crimes and become incarcerated).
Dr. Teresa May, department director of the Community Supervision & Corrections Department in Harris County, Texas, also notes that on a whole, men have higher recidivism risk than women. “What we know is when you look at gender, almost always—and in fact I don’t know of an exception—the average rearrest rate [for women] is always much lower than men.” Without accounting for these differences, a risk assessment could end up scoring women as higher risk than they actually are....
There’s an ongoing debate over whether using gender as a risk factor, or assigning different cutoff risk levels to both males and females, violates the 14th Amendment. “Basically the Supreme Court of the U.S. has pursued what’s called an anti-classification approach to the equal protection law, which prohibits explicit use of factors like gender and race in making decisions,” says Christopher Slobogin, professor of law at the Vanderbilt University Law School. He adds, “It is permissible, constitutionally, to use race or gender if there is a compelling state interest in doing so. But generally speaking, the use of race and gender is unconstitutional to discriminate between groups.”
However, in Slobogin’s own opinion, he does not think the “Constitution is violated simply because a risk assessment arrives at different results for similarly situated men and women.” He argues that a tool that uses gender as a risk factor and one that has different cutoff scores for genders are functionally the same, adding in those adjustments makes the instruments more accurate.
But others see it differently. Sonja Starr, professor of law at the University of Michigan Law School, for example, recently told the Philadelphia Inquirer that “use of gender as a risk factor plainly violates the Constitution.”
Comments
Post a Comment