Does increased post-release supervision of criminal offenders reduce recidivism? Evidence from a statewide quasi-experiment

Approximately 4.8 million offenders are subject to community supervision in the United States. This paper examines whether a program that assigned different supervision levels based on a risk assessment instrument, had any effect on offenders’ recidivism rates. Using a large statewide sample of adult offenders in Washington State and a regression discontinuity design, I compare offenders whose risk characteristics are similar but who received different levels of post-release supervision. I find that offenders who received more supervision were not less likely to reoffend. The result holds for high-risk and low-risk offenders and for various types of recidivism.