Survival Patterns and Predictors of Recidivism Among Youth Released from Alabama’s Juvenile Justice System

Background and Significance: Youth involved in the juvenile justice system face significant challenges when reentering their communities, with recidivism posing a major barrier to long-term rehabilitation and public safety. Understanding when reoffending occurs and what factors are associated with it is essential for developing effective, evidence-based interventions. Previous research has identified individual, offense-related, and placement-related characteristics linked to recidivism, though findings remain mixed and vary by context. There is a continued need to explore recidivism patterns over time, particularly during the high-risk period following release. This study addresses two research questions: (1) What is the estimated time to recidivism among youth discharged from the juvenile justice system in Alabama? and (2) Which factors are associated with the likelihood of recidivism over time in this population?

Methods: De-identified youth data were obtained from the Student Information Management System maintained by the Alabama Department of Youth Services (DYS), and recidivism data came from the Administrative Office of Courts. The sample included youth released from DYS custody in 2018 (N = 959), ages 12 to 20 at the time of release, representing 64 of Alabama’s 67 counties. The dependent variable, time to recidivism, was defined as the number of days between release and adjudication for a new offense. Youth who recidivated during the observation period (6 months to 3 years) had time of recidivism; youth who did not recidivate had censored time. Independent variables included age at release, race, gender, release type, custody length, risk and needs scores, offense type, residence area, placement type, and domestic violence history. Kaplan-Meier estimation and Cox regression were conducted using SPSS version 30.

Results: A Kaplan-Meier estimation showed that at 180 days post-release, the survival probability was approximately 79%, meaning 21% of youth had reoffended. By 365 days, it declined to 68%, and by 1,430 days (nearly four years), it was about 47.5%, indicating that over half the sample had reoffended. The median survival time—when 50% had reoffended—was 1,163 days. The steepest decline occurred within the first year, marking it as the highest-risk period; the curve flattened after 400 days, suggesting a reduced risk over time. Longest time-to-recidivism was censored time (1430 days), indicating that some youth remained offense-free. In the Cox regression, older age and violent offense were associated with lower hazard, while secure placement was associated with a higher hazard of recidivism.

Conclusion and Implications: Findings highlight the first year following release as a critical window for recidivism prevention, with risk declining significantly after this period. That over half of youth reoffended within four years underscores the long-term vulnerability of this population. The association between secure placement and higher hazard of recidivism suggests the need to reevaluate intensive confinement strategies, while the protective roles of older age and violent offense point to the complexity of risk profiles. These insights support the development of targeted, time-sensitive interventions and post-release support tailored to youth characteristics. Policymakers and practitioners should prioritize early, community-based strategies to reduce reoffending and support sustained reintegration.