Learning theory approach to the deterrence of criminal recidivism

Data on punishment and criminal arrests in a total birth cohort of men in Denmark (N = 28,879) was used to test the following hypotheses derived from learning theory: (a) the imposition of sanctions reduces rates of subsequent criminal arrest; (b) the more severe the sanction received for an arrest, the lower the rate of recidivism; (c) different types of sanctions have similar effects on recidivism; (d) the higher the proportion of sanctions received for past arrests, the lower the rates of future arrest; (e) continuous sanctions reduce arrest rates more than intermittent sanctions; and (f) discontinuation of punishment results in recovery of criminal arrests. Results support the above hypotheses, except Hypothesis b; our results suggest that sanctions have similar effects on recidivism regardless of their severity.