Pennsylvania lawmakers struggled for almost a decade to come up with a risk assessment tool for criminal sentencing that would keep low-risk people out of jail and in probationary programs that could better serve them. The problems that lawmakers keep encountering, though, have been completely predictable: the data-driven tools they keep trying to use would only make racial disparities in sentencing and the prison population more pronounced.
Recently, it seems that they lowered their standards in order to get something off the ground.
Pennsylvania Finally Develops a Risk Assessment Tool for Sentencing
Nine years ago, the Pennsylvania Commission on Sentencing was tasked with the job of creating a sentencing tool that would flag low-risk offenders and shuffle them into probation rather than jail after a conviction. The idea was to keep prison populations in Pennsylvania down, and hopefully alleviate what had become a very apparent disparity between the racial population of the state and the number of minorities in prison.
The Commission finally approved that tool on September 5, 2019. They have explored five different versions of it, but none of them passed muster.
Their sixth went up for a vote on September 5 and got approved by the Commission on a 7-2 vote.
The approval, however, does not signify that the newest version of the risk assessment tool is better than all of the earlier ones the Commission tinkered with. It just means that the Commission bent to political pressure to get something, anything, done.
Why the Risk Assessment Tool Will Only Makes Things Worse
The Commission's risk assessment tool is supposed to tell judges which criminal defendants are likely to commit another crime, in the future. In theory, those odds of recidivism should play a role in how that person is sentenced after a conviction. People who are being sentenced for a crime they committed and who are likely to commit another crime in the future should receive a harsher sentence because they continue to be a risk to society. Low-risk people, on the other hand, should not be put through the prison system because they do not “need” it.
The tool that the Commission approved, however, uses algorithms and computer programs that rely on data and trends from the past to predict what will happen in the future.
The problem with that method is that the data and prior trends are rife with racial disparities and other unfair factors. Using those elements to predict recidivism rates for current defendants will guarantee that they will be measured by those same racially-motivated factors that created the problem, in the first place.
Philadelphia Criminal Defense Lawyer Joseph D. Lento
Law enforcement has jumped at the opportunity to use technology before. But risk assessment tools like this one have blatant design problems that cannot be overcome. The two members of the Commission who voted against its approval have already said they would continue to fight it. Until they succeed, people who have been convicted of a crime in Philadelphia will find their sentences influenced by a computer program that can count their race against them.