First-Day criminal recidivism

Abstract

We report that on any given day the number of inmates released from incarceration significantly affects the number of offenses committed this day, and we name this as first-day recidivism. Our estimates of this novel approach to study early recidivism are robust to a variety of alternative model specifications. We then show that first-day recidivism can be eliminated by an increase in the gratuity provided to prisoners at the time of their release. A simple cost-benefit analysis shows that increasing the gratuity at release is a very efficient policy.