ROADS DIVERGE: LONG-TERM PATTERNS OF RELAPSE, RECIDIVISM AND DESISTANCE FOR A RE-ENTRY COHORT

The goal of this grant is to inform reentry programming and policy by testing and modifying theoretical models of criminal desistance for drug-involved, minority, and female offenders. Phase I uses extant longitudinal data from a previous KEY/Crest program study of offenders released from Delaware prisons in 1991-1998---enhanced by updated archival criminal arrest, conviction, and incarceration data---to model post-prison relapse and recidivism trajectories and identify patterns (e.g. those who desist, those who relapse and offend regularly, and those whose use and offending changes). Phase II adds data using life calendar interviews to collect self-report information on attitudes, life circumstances, and other things that may affect these patterns; both quantitative and qualitative data analysis will be applied. ca/ncf