Determining how professionals can use data-driven tools to reduce unnecessary incarceration and better predict and prevent re-offending
Risk assessment instruments are data-driven evaluations that characterize the likelihood that a person will commit future crime. Risk assessment has long been an integral part of the justice system, but the use of this technology has expanded in today’s era of justice reform, big data, and machine learning. Jurisdictions across the U.S. are relying on these instruments in efforts to reduce mass incarceration without compromising public safety.
Despite promising results, these practices raise high profile questions. Do risk assessments increase professionals’ accuracy, when they must consider a person’s risk of reoffending? Can risk assessment improve treatment provided through the justice system? Will the use of this technology exacerbate racial and socioeconomic disparities in incarceration by ‘baking in bias’?
We are using diverse methods (from experiments to big data analyses) to provide guidance on such issues — and help policymakers choose approaches that will improve criminal justice practice.