Incarceration, Recidivism, and Employment

Manudeep Bhuller: University of Oslo, Statistics Norway, Institute for Labor Economics, CESifo
Gordon B. Dahl: University of California San Diego, University of Bergen, National Bureau of Economic Research, Institute for Labor Economics, CESifo
Katrine V. Løken: Norwegian School of Economics, Statistics Norway, University of Bergen, Institute for Labor Economics, Center for Economic and Policy Research, CESifo
Magne Mogstad: University of Chicago, Statistics Norway, University of Bergen, National Bureau of Economic Research, Institute for Labor Economics, CESifo

Utilizing a random judge design and panel data from Norway, findings indicate that imprisonment reduces further criminal behavior with re-offense probabilities dropping by 29 percentage points and criminal charges decreasing by 11 over a 5-year span.
The ordinary least squares (OLS) method reveals a potential misinterpretation, suggesting the opposite conclusion, influenced mainly by individuals not engaged in work prior to incarceration.
Individuals who were employed before imprisonment show lasting negative effects on future employment. The study emphasizes that focused rehabilitation during incarceration can be preventive for significant portions of the criminal population.

Over recent decades, incarceration rates have increased markedly; in the US, from 220 per 100,000 in 1980 to 700 in 2012, whereas Western Europe saw a rise from 62 to 112 per 100,000.
This trend raises concerns regarding the reintegration of ex-convicts into society and the potential for recidivism.
The paper examines whether time in prison rehabilitates offenders or poisons opportunities through human capital depreciation and exposure to hardened criminals.
Challenges in studying the causal effects of incarceration are attributed to data availability and correlated unobservables, making causal inferences scarce.

Focused on Norway's criminal justice system, the paper links various administrative data sources to create a comprehensive panel dataset that includes every Norwegian's criminal history and labor market outcomes.
The identification threat was addressed using randomly assigned judges whose varying stringency offered a reliable measure of incarceration effects without bias from case characteristics.
The randomized assignment of criminal cases to judges revealed that judge stringency can significantly impact recidivism rates and labor market outcomes.

Recidivism and Incarceration
- Imprisonment substantially discourages further criminal behavior; post-release, the probability of reoffending decreased notably.
- Insightful OLS vs. Instrumental Variable (IV) findings: OLS indicated positive correlations between incarceration and crime, while IV estimates exhibited a strong preventive effect.
Impact of Work Status Before Incarceration
- Individuals not engaged in employment pre-incarceration exhibited improvements in labor market participation and reduced recidivism.
- A 35 percentage point increase in job training engagement among the unemployed was noted, alongside rising employment rates post-release.
- The previously employed group experienced a 30 percentage point drop in employment during incarceration, with no significant change in recidivism outcomes.
Broader Implications
- The study's findings foster discussions on the effectiveness of imprisonment rates and rehabilitation policies, highlighting potential cost-benefit advantages in the Norwegian context.

Norway's prison system appears effective at reducing recidivism and improving employment via rehabilitation-focused practices, primarily benefiting those who were unemployed prior to incarceration.
The paper urges further examination of varied factors influencing recidivism and employment within different social contexts for broader implications.