Risk-Assessment, Race and Recidivism: Examining for Potential Predictive Bias

• In Event: Advancing Equal Justice: Issues, Strategies & Obstacles

Abstract

The use of actuarial risk assessments in the field of corrections has increased dramatically in the last two decades. Risk assessments have increasingly replaced the use of professional judgment in making sentencing, correctional placement, and treatment decisions. Research on risk assessments has consistently shown that these tools are superior to professional judgment in predicting recidivism. Furthermore, the use of risk assessments has been viewed as a way to reduce subjectivity and bias in making correctional decisions. However, the past few years have seen some controversy arise around the use of risk assessments. Some scholars and a few politicians have argued that certain domains that are included in risk assessments (i.e., employment status, conditions of accommodations) work against offenders that come from poor and minority backgrounds. The current study seeks to examine potential predictive biases by race in the use of the Level of Service Inventory – Revised (LSI-R), a widely used risk assessment tool. Risk assessment and recidivism data were obtained for a 2010-2012 New England sample of offenders on probation. Follow-up recidivism data was available for three years after release.

Authors

• Christopher Sedelmaier, University of New Haven

• Mirlinda Ndrecka, University of New Haven