Criminal Recidivism after Prison and Electronic Monitoring

We study the re-arrest rates for two groups: individuals formerly in prison and individuals formerly under electronic monitoring (EM). We find that the recidivism rate of former prisoners is 22% while that for those 'treated' with electronic monitoring is 13% (40% lower). We convince ourselves that the estimates are causal using peculiarities of the Argentine setting. For example, we have almost as much information as the judges have when deciding on the allocation of EM; the program is rationed to only some offenders; and some institutional features (such as bad prison conditions) convert ideological differences across judges (to which detainees are randomly matched) into very large differences in the allocation of electronic monitoring.