When you’re accused of a crime you want to know how long your sentence could be, and that’s your bottom line. In that moment you don’t have time to consider how sentence lengths are determined, but maybe you should. New research is finding that one of the major tools that judges use in determining sentence length may have major flaws.
The algorithms used to determine recidivism rates are supposed to be cold, impartial logic: data are punched in, answers are distributed. However, recent studies are showing that the algorithms that courts are using are including a racial bias and that’s a big problem. Fairness is the whole justification for using these mathematically calculated formulas to determine sentence lengths. If they include a racial bias, then they are only making the problem worse.
The flaws in the algorithm
The general flaw in the algorithm comes from the arrest data they are based on. Since people of color are policed more heavily and arrested more frequently, then for any given crime, a person of color is more likely to be arrested than a white person under identical circumstances. In that case, then, the algorithm isn’t measuring the likelihood of whether they would commit another crime, but rather their likelihood of being caught.
This sort of data flaw can be traced to federally mandated sentencing from the “War on Drugs” and initiatives designed to be tough on crime, as well as institutional racism. These efforts did not lead to fewer crimes, but rather rising, expensive prison populations and the over-policing of economically disadvantaged areas. With heavier policing in a given area, data would naturally show that the population of that area was more likely to commit crimes, and the algorithm would be self-enforcing.
North Carolina’s sentencing fairness history
North Carolina has a more than 30-year history devoted to finding the fairest possible sentences. Like other states across the country, they turned to data-reliant methods in the hopes of a better system. With this new information, it seems that fairer sentencing will be more difficult than a simple calculation.