AN ECONOMETRIC STUDY ON THE DETERRENT IMPACT OF PROBATION - CORRECTING SELECTION AND CENSORING BIASES

A drug offenders data set, recently collected in Florida, is used to e stimate the impact of probation, vis-a-vis imprisonment, on recidivism . Because the sample period is characterized as ''getting tough on dru gs,'' a possible connection between the drug policy and the deterrent impact of probation is explored. Econometrically, the main concern of this article is to correct not only the bias related to the selection process, but the bias resulting from interruption of the duration vari able of those who have not returned to prison by the end of the survey . After employing methodologies to correct for these biases, there rem ains no evidence that probation per se has any significant impact on t he duration of recidivism.