Black and young defendants in Virginia have faced increased jail sentences after the state instituted a risk assessment system.
Megan T. Stevenson of the Antonin Scalia Law School at George Mason University and Jennifer L. Doleac of the Department of Economics at Texas A&M University researched and examined tens of thousands of felony convictions in Virginia.
The duo focused on the period between 2000 and 2004, where they measured how judges used a risk assessment algorithm instituted throughout the state in 2002.
They found that defendants younger than 23 were 4 percentage points more likely to be incarcerated after risk assessment was adopted, and their sentences were 12 percent longer than older peers.
“People are getting this very stigmatic label – high risk for violent recidivism – primarily because they’re 19 years old,” Stevenson said in a news release.
In Virginia, 13 points are added to the risk score if a defendant hasn’t reached the age of 30. By contrast, if someone is older than 30, and imprisoned at least five times, the risk score only adds nine points.
While the algorithm doesn’t use race to help calculate a score, African Americans were four percentage points more likely than whites to be incarcerated since the state adopted risk assessment. Additionally, the sentences for African Americans were 17 percent longer, according to the study.
“Despite being adopted with an express decarceral purpose, Virginia’s nonviolent risk assessment did not result in lower incarceration rates or sentencing,” the study authors wrote. “Instead, incarceration was reallocated from those with lower risk scores to those with higher risk scores. Theoretically, this should have reduced recidivism rates. We find no evidence that this was the case.”
They noted that the use of judicial discretion could at least partially explain this.
Although risk assessment use did lead to a relative increase in sentencing for young defendants, the increase was nowhere near as large as it would have been if judges had fully complied with the sentencing recommendations associated with the algorithm, the researchers said.
Judges used their discretion to systematically divert young offenders, despite their higher risk of reoffending, and that’s likely due to longstanding practices of treating youth as a mitigating factor, as opposed to an error in prediction.
Since a young age is one of the most important predictors of reoffending, a preference for leniency for this group will curtail risk assessment’s expected benefits, the authors concluded.
“The disappointing recidivism results may also be partially explained by structural challenges in predicting offending. Even in optimal circumstances, risk predictions explain only a tiny fraction of the variation in reoffending,” Stevenson and Doleac wrote.
Similar algorithms reportedly are used in 28 states.
“Virginia’s nonviolent risk assessment reduced neither incarceration nor recidivism; its use disadvantaged a vulnerable group (the young), and failed to reduce racial disparities,” the study authors said.
To view the complete study, go to https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3489440.