25-year outcomes of an in-prison therapeutic community in Texas

Abstract

This study evaluated whether in-prison therapeutic community treatment, risk classification (High vs. Low), and aftercare completion were prospectively associated with time to first arrest following release from prison. Moreover, this project established a predictive algorithm of reincarceration using multilevel modeling to identify variables that can be targeted during treatment to decrease reincarceration rates. Based on the Risk-Need-Responsivity model, measures of the criminogenic needs (i.e., family relationships, pro-criminal attitudes, antisocial personality pattern, substance use, and social supports for crime) were compared to static risk factors in terms of their added predictive utility in modeling long-term recidivism trajectories. Results can be used to inform the field’s current understanding of in-prison therapeutic communities on recidivism and ways to optimize client selection and the treatment continuum are discussed.