You may not have heard about them before, but many courts are turning to risk prediction equations as a measure to make informed decisions regarding sentencing, bail, and early release. A judge’s own opinion on the matter may be overlooked in favor of these increasingly prevalent equations. But how accurate are these equations, and moreover, is there a risk of racial bias? The answer, unfortunately, is that there is a huge racial bias that goes a long way to making an already racist institute (the criminal justice system) even more so.
According to ProPublica, “Researchers found that the formula, and others like it, have been written in a way that guarantees black defendants will be inaccurately identified as future criminals more often than their white counterparts.” An earlier investigation by ProPublica, which spurred more recent research by four other independent groups of scholars that came to similar conclusions, found that black defendants were twice as likely to be given an incorrect high-risk level than whites. These scores are called COMPAS risk assessment scores. The investigation observed a racial bias by following defendants who were not arrested for new crimes but were initially given a high-risk level. Blacks were more likely to be given a high-risk level, while whites who were given a low-risk level COMPAS score were actually more likely to commit a new offense than blacks with comparable risk assessments.
The Office of Justice Programs: New Tool in Philadelphia Shows Great Promise