Not Just A Law Firm... A Solution

Studies examine racial bias in risk assessment tools

On Behalf of | Jul 5, 2019 | Criminal Defense

After a person in Georgia is taken into custody in relation to a crime, that person may then appear before a judge who decides whether the person should be released or kept in jail until the trial. There have long been concerns that racial bias plays a part in an assumption that black defendants are more likely than white ones to be repeat offenders. In response, some districts nationwide have introduced algorithms that are supposed to help judges make this decision without bias. However, some critics say that because the algorithms themselves are based on a system that is full of bias, they may still result in unfair assessments.

In 2016, a study by ProPublica examined the use of algorithms for risk assessments in a county in Florida. It found that black people had a rate of false positives for being likely to commit another crime that was almost two times higher than that of whites. White people had a higher likelihood of being deemed low risk and then going on to commit another crime.

The Center for Court Innovation designed a risk assessment tool to use hypothetically on real cases in New York and came to a similar conclusion. It identified 25% of black defendants as high-risk who did not offend again compared to 10% of white defendants.

People who are facing criminal charges may be up against these and other problems that put them at a disadvantage. For example, some people may be facing charges based on evidence that was not obtained legally. An attorney might be able to identify this or other problems that could lead to evidence or a case being dismissed. An attorney may also be able to help with other elements of the case whether a person decides to go to trial or accept a plea bargain.


FindLaw Network