Using Random Judge Assignments to Estimate the Effects of Incarceration and Probation on Recidivism Among Drug Offenders

Abstract

Most prior studies of recidivism have used observational data to estimate the causal effect of imprisonment or probation on the probability that a convicted individual is re-arrested after release. Very few studies have taken advantage of the fact that in some jurisdictions, defendants are randomly assigned to judges who vary in sentencing tendencies. The present study investigates whether defendants who are randomly assigned to more punitive judges have different recidivism probabilities than defendants who are assigned to relatively lenient judges. We track 1,003 defendants charged with drug-related offenses (and no non-drug-related offenses) who were randomly assigned to nine judicial calendars between June 1, 2002 and May 9, 2003. Judges on these calendars meted out sentences that varied substantially in terms of prison and probation time. We tracked defendants using court records over a four-year period following the disposition of their cases in order to determine whether they were subsequently re-arrested. Our results indicate that randomly-assigned variations in prison and probation time have no detectable effect on rates of re-arrest. The findings suggest that, at least among those facing drug-related charges, incarceration and supervision seem not to deter subsequent criminal behavior.

Keywords: Recidivism, sentencing, specific deterrence, drug crime, natural experiments

JEL Classification: K14, K42