ABSTRACT
New research in the field of developmental criminology has led researchers to reconceptualize desistance as a behavioral process that unfolds over the life course. This approach puts more emphasis on the pathways by which people reach the state of non-offending, and less emphasis on the state of non-offending itself. This reconceptualization has implications for how we measure desistance in longitudinal data. In this paper, we suggest that the traditional measurement approach is inconsistent with this view, and we present an alternative measurement approach based on the premises of developmental criminology. Although not perfect, we argue that the dynamic measure better describes the key elements of the process of desistance. Both approaches are implemented using data from the Rochester Youth Development Study, a longitudinal study of youthful offenders. We demonstrate that the two approaches identify different people as desistors. Moreover, we argue that the dynamic definition of desistance has more promise for providing insight into the changes that are the behavioral focus of the desistance process.
