Addressing the Race Gap in Incarceration Rates: An Agent Based Model

Abstract

ABSTRACTUsing an agent-based model of incarceration, the authors conduct a series of simulation experiments testing the efficacy of policy interventions designed to address racial disparities in incarceration rates. The first experiment eliminates race-based sentencing disparities, and additional experiments eliminate race-based sentencing disparities and disrupt the clustering of incarcerated individuals in their social networks. Findings suggest that eliminating race-based sentencing disparities will not result in a substantial closing of the incarceration gap because the clustering of ex-offenders contributes to offender recidivism and disparities in incarceration rates. The model suggests that addressing sentencing biases combined with interventions designed to disrupt the clustering of incarcerated individuals would reduce race-based incarceration gaps. Policy implications based on these findings are discussed.