Imprisionment and Reoffending

Carnegie Mellon University, Pittsburgh PA

Investigators

Abstract

The aim of this research is to provide scientifically sound evidence on the effect of the experience of imprisonment compared to a noncustodial sanction on reoffending. Review of the empirical literature on imprisonment effects on reoffending indicates that, on balance, the evidence points to a criminogenic rather than a preventive effect. However, such a review also indicates that the evidentiary base for this conclusion is weak. The most important statistical weakness of much of the existing research is its vulnerability to "selection bias." That is, even with extensive statistical controls for measured case and person characteristics, persons sent to prison may remain systematically different from those convicted but not imprisoned on unmeasured characteristics related to recidivism. To overcome the selection problem we propose to capitalize on the random assignment of cases to judges in the criminal courts of Pennsylvania and the District of Columbia. Because of random assignment, there will be no systematic difference in case characteristics across judges. Cross-judge variation in their likelihood of imposing a prison sentence will be used as the basis for inferring the effect of incarceration on reoffending. Imprisonment is a sanction that is financially costly, exacts an economic and psychological toll on family members of the incarcerated, particularly children, and raises issues of social justice due to its differential impact on minority members and communities. Thus, beyond being an instrument for meting out just desserts, prison's justification must rest heavily on its demonstrated utility to protect the social order. Understanding the effects of incarceration on reoffending should thus be a key consideration in formulating correctional policy.