An Implementation Evaluation of Community Corrections

This article describes an evaluation of the implementation of a community diversion program for nonviolent offenders in Virginia. The evaluation used a unique application of logit analysis to focus on the extent to which the program was serving the intended population, in this case, offenders who would have otherwise been incarcerated. Results indicate that at least 46% of the divertees would have been incarcerated if the program had not been available. Usefulness of a statistical technique for evaluating implementation and providing concrete results to policymakers is discussed.