Doctoral Dissertation Research in Economics: Do rehabilitative prison policies work? Evidence from a natural experiment in the Dominican Republic

University Of California-Berkeley, Berkeley CA

Investigators

Abstract

High crime rates among those released from prison is of great concern among policy makers because of high social cost and loss of human capital. Despite these concerns, economists do not understand the sources of increased recidivism. Understanding the sources of recidivism will therefore contribute to economic science as well as the formulation of policies to reduce recidivism. This doctoral dissertation research project will investigate whether rehabilitative prisons reduce the rates of recidivism and labor-market participation of prison releasees. The project will also investigate how prisoners, once transformed by their prison conditions, fit into, and affect their communities when they are released. The research will use experimental methods and data from two alternative prison systems---the traditional punitive system and a modern rehabilitative system---to provide empirical evidence on these questions. The PIs have access to unique data sets on prisoners released from traditional prisons and rehabilitative prisons. The results of this research project will contribute to the development of policies to make the U.S. correctional system more efficient and therefore establish the US as a global leader in penal reforms. This doctoral dissertation research project will study the effects of rehabilitative prisons on recidivism rates, labor market attachment of prison releasees, and how releasees contribute to their societies. The PIs have access data sets of prisoners released from traditional and rehabilitative prison. The research will use two quasi-experimental research design projects to identify causal parameters. The first project uses instrumental variables approach to compare individuals randomly assigned to judges with varying preferences for incarceration and prison type to identify the causal effect of incarceration in punitive prisons or rehabilitative prisons on recidivism, and the difference between them. The second project uses a differences-in-differences estimation strategy that exploits the differential rollout of rehabilitative prisons to identify the long-term effects of additional rehabilitative time. Individuals introduced into rehabilitative prisons at the same time with remaining short sentences serve as a counterfactual to those with long remaining sentences because the prisons sort individuals into homogeneous groups across many factors, but not time. Both sets of estimates will provide rigorous evidence on the effects of large-scale prison reform on recidivism. This information is important for US policymakers and prison official as they respond to large-scale incarcerated population and the perceived failure of the current system. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.