Prison Work and Convict Rehabilitation

I study the causal pathways that link prison work programs to convict rehabilitation, leveraging administrative data from Italy and combining quasi-experimental and structural econometric methods to achieve both a credible identification and the isolation of mechanisms. Due to competing channels, I find that work in unskilled prison jobs impacts convicts on longer or shorter terms differently. Increasing work time by 16 hours per month reduces by between 3 and 10 percentage points the reincarceration rate, within three years of release, of convicts on terms longer than six months – because prison work counteracts the rapid depreciation of earning ability experienced by these convicts. For those on shorter terms, the analogous increase leads instead to a re-incarceration rate that is up to 9 percentage points higher, because of a liquidity effect that weakens deterrence.