Samenvatting

This paper examines the relationship between imprisonment length and registered recidivism. The data come from a unique longitudinal and nationwide study of Dutch male prisoners, serving an average of 4.1 months of confinement (N=1,467). Ideally an experimental design would be appropriate to examine the influence of different sentence lengths on recidivism. In order to approximate such a design using observational data, we adopt a propensity score methodology to control for selection bias in the dose-response relationship. Using a six-month follow-up, we do not find significant differences in post-release recidivism between men who spent shorter or longer periods of time in confinement. We discuss the pros and cons of the methodology applied as well as potential implications of the findings.