Research project ‘What Works’ in Prison?

Prison authorities have many tools – e.g. placement policies, treatment, sentence length, electronic monitoring – to rehabilitate offenders and reduce recidivism. But causal evidence on the intended/unintended effects of these tools is limited. This program begins to fill these gaps. Project 1 assesses whether more time in prison causally impacts outcomes of the family members (parents, children, siblings, partners) of inmates. Project 2 studies if the introduction of electronic monitoring has spill-over effects onto the conditions and post-release outcomes of prisoners ineligible for such sanctions. Project 3 assesses how in-prison health care affects post- release health, healthcare utilization, and crime.

We will match registers of offenders who enter the Swedish prison system from 1992 to 2016 to other national registers (e.g. crime, health, education, labor). We will obtain causal estimates by implementing quasi-experimental research designs that utilize exogenous variation arising from institutional features of the prison system. This ambitious four-year program will generate 3-5 papers, including multiple papers on prison healthcare.

Rising crime rates, calls for longer sentences and capacity constrained prisons in Sweden all make this research urgent. Policies to reduce crime will only be effective if they are based on causal evidence of what works in prison, and if there are no unintended harmful effects.