Skip navigation
CLN bookstore
× You have 2 more free articles available this month. Subscribe today.

Risk Assessment Tools Perpetuate Inherent Biases and Prejudices

Most states view Kentucky as the best example of utilization of RA tools in the U.S. In 2011, it was the first state to implement the use of RA in deciding bail. According to a 2018 study, the number of people released pending trial at that time increased by 13%.

By 2016, more than half that gain disappeared. Laws changed and judges were given more leeway to ignore recommendations. No oversight existed on the use or rejection of RAs. Kentucky Center for Economic Policy research director Ashley Spalding said judges tended to ignore the results. “They’re overriding the findings of the risk assessment tool,” she said. “In practice, we’re seeing that it is often disregarded.”

A 2019 study showed that judges were more apt to ignore the recommendation for release of Blacks in the moderate risk category than Whites. “Judges see the moderate risk label, and for white defendants, moderate risk was interpreted as low risk, and for Black defendants, it was interpreted as a signal of higher risk,” stated Megan Stevenson, professor at George Mason University and author of the study.

The Laura and John Arnold Foundation developed the Public Safety Assessment, which Kentucky began using in 2013. It is now the most widely used model in the U.S. Reform advocates say the criteria used to determine a risk category does not appear to be causal but correlative. Data entered by clerks, says The Intercept, are prone to errors and unavailable to the public or defendants for review or rebuttal.

Pilar Weiss, director of the Community Justice Exchange, said RAs have become “a totally political, manipulated, secret process. They’re these flawed, racist, classist tools that purport to be based on science that are how the system is making decisions about people’s freedoms and their liberties.”

Predictive RAs work by seeding an algorithm with as much court data as possible. The data are then correlated with characteristics that indicate potential risk value for rearrest or failure to appear on the assigned court date based on data from other defendants with similar characteristics.

Critics contend that the correlative risk values are compared with already biased data based on past records of rearrest or failure to appear when this data have been inherently racist and biased throughout history. Additionally, relying on correlative data, and not causal data, gives emphasis to odd characteristics.

One RA factor is whether the defendant has an official place of residence. Those who do not are considered high risk. Yet a study conducted in Cook County, Illinois, showed that 99% of defendants who were deemed high risk but released before trial anyway showed up for court without a new arrest — virtually the same percentage as those deemed moderate or low risk.

The Leadership Conference on Civil and Human Rights drafted a letter in 2018, which was endorsed by over 100 other organizations, arguing against the current failing RA practice and listed six recommendations that would help reduce the bias and prejudice inherent in RAs — such as allowing adversarial hearings for those determined “at risk” and automatic release for everyone else, making release instead of imprisonment the default. 


As a digital subscriber to Criminal Legal News, you can access full text and downloads for this and other premium content.

Subscribe today

Already a subscriber? Login



Federal Prison Handbook - Side
PLN Subscribe Now Ad 450x450
CLN Subscribe Now Ad 450x600