Using two alternative methods to account for model selection uncertainty and the subsequent inference; Bayesian model averaging and a non-Bayesian approach based on bootstrap resampling, this paper identifies the risk factors for recidivism among felony probationers convicted of drug trafficking. A national sample is used and the individual hazard function is assumed to depend on individual and neighborhood characteristics as well as social interactions among probationers. Both methods yield similar results and point to social interactions as one of the most significant factors affecting recidivism.
© 2024 Center for Statistics and the Social Sciences