In a policy of selective incapacitation, severity of punishment increases with an additional conviction. Repeat offenders face the threat of enhanced punishment, which may have a deterrent effect. In
addition, repeat offenders can be sentenced to enhanced prison terms, which may have an additional incapacitation effect. The rationale for such a policy is that punishment is costly and the offender population heterogeneous in terms of how they respond to punishment. In this paper, we evaluate a policy of selective incapacitation of repeat offenders adopted in 2001 in the Netherlands, using monthly data for 12 urban areas between 1999 and 2007. Upon arrest, offenders who have
shown to be of the prolific type (by being arrested ten times and convicted three times within a five-year period) face a prison term that is some ten times higher than usual. To investigate the impact of this policy on crime, we match every individual offender sentenced under this law to the urban area in which he was criminally active. Variation over time in use of the policy constitutes a natural
experiment as decisions relating to themoment of introduction and frequency of application of the policy were not based on local crime trends. We find the rate of theft from car and burglary to be 30
percent lower by the end of 2007 as a result of selective incapacitation. The benefits of the policy exceed the costs by at least a factor of two. The social returns of the policy are found to be subject to decreasing returns to scale.