Incarceration experience at older ages. Does employment protect against recidivism?

We investigate whether employment protects against and/or postpones recidivism among males who committed their first crime late in life. We use administrative data on 34,401 individuals, 44% of whom were employed at least once during the analysed period. We apply a multi-state model and difference in-differences approach. The results of the multi-state model indicate that employment reduced the risk of recidivism by 7% for those who were at risk of a second incarceration. On the other hand, employment increased the probability of subsequent recidivism by 8-10% for those who were at risk of a third or fourth incarceration. Our results suggest that institutional interventions should seek to increase the labour market attachment of former prisoners, and, in particular, should focus on ex-offenders with only one conviction. Being attached to the labour force seems to matter more, the fewer imprisonment spells an individual has experienced.