How do economic conditions and finding a job affect criminal recidivism? Decomposition of policy and treatment effects with duration outcomes

• In Event: From Cradle to Grave: Crime Across the Entire Life Course

Abstract

The effect of employment of crime is of major interest in the criminological literature. Most empirical studies are using observational data, with obvious limitations concerning estimating causal effects. In this paper, we aim at estimating causal effects of employment on recidivism after an initial release from prison by considering variation in the local labor market. We use an econometric framework which identifies the causal effects of a treatment policy regime when the outcome of interest is a duration variable, and agents are randomized to a policy regime upon entering an initial state which prescribes a stochastic propensity to future treatment. Thereafter, at different moments in time and depending upon their policy regime, surviving agents are randomized to actual treatment. We consider a situation where inmates are released to their municipality of residence at different points in time with varying unemployment rates. We use administrative data from Norway which follows ex-convicts who exited prison for the first time between 2002 and 2008 (N=13,769). We utilize both aggregate employment rates and the individual's actual recorded employment at monthly intervals. We then decompose the causal effects of economic conditions and reemployment on time to recidivism.

Authors

• Stephen P. Kastoryano, University of Mannheim

• Jad Beyhum, University of Mannheim

• Torbjorn Skardhamar, University of Oslo