The effects of aftercare/resettlement services on crime and violence in children and youth: A systematic review
Jennifer S Wong,Chelsey Lee,Natalie Beck
DOI: https://doi.org/10.1002/cl2.1404
2024-05-25
Abstract:Background: High rates of youth re-offending indicate that young custody-leavers face challenges when reintegrating into their communities. Aftercare and resettlement programs can occur pre-, during, and post-release and generally provide multiple forms of support services to address youths' transitional needs. Objectives: The present review examines (1) the impact of youth aftercare/resettlement programs on crime-related outcomes, (2) how treatment effect is moderated by participant, program, and study characteristics, (3) whether some types of interventions are more effective than others, (4) barriers/facilitators to effective program implementation, (5) the theory of change underlying resettlement interventions, and (6) available research on intervention cost. Search methods: A comprehensive set of keywords and synonyms was combined in a Boolean search across 26 electronic databases. Multiple gray literature sources were also searched, including 23 journals, 4 meeting archives, 11 organization websites, 3 open access journal websites, and the CVs of 8 well-known researchers in the field. The search was completed in January 2023. Selection criteria: For objectives 1-3, studies were included if they utilized a randomized controlled design or quasi-experimental comparison group design in which participants were matched on at least some baseline variables and included at least one quantitative individual-measure of crime. For objective 4, included studies presented process evaluations of aftercare/reentry programs, clearly stated their research goals, and used qualitative methods in an appropriate way to answer the stated research question. For objectives 5 and 6, no specific methods were required; any study meeting the criteria for objectives 1-4 which presented findings on theory of change or cost data were included. For all outcomes, only studies conducted in a westernized country, and published after 1991 in English, French, or German were considered. Data collection and analysis: Two coders conducted primary data extraction for the included studies. Data were entered into a Microsoft Excel database. After data extraction, the two coders validated the coding by cross-checking the database with each research report. Discrepancies between coders were discussed until consensus was reached. Where consensus could not be reached, a third coder was consulted. Study risk of bias was addressed using the ROBINS-I (Sterne et al., 2016), ROB-2 (Higgins et al., 2019), and the critical appraisal skills programme (CASP, 2018). Objectives 1-3 were addressed by synthesizing quantitative outcomes from rigorous impact evaluations of aftercare interventions using random effects models and meta-regression. Thematic and narrative analysis was conducted to address objectives 4-6. Results: The search resulted in 15 impact studies, representing 4,718 participants across 21 program sites, and 35 effect sizes. The 21 impact evaluations were rated as having either low/moderate bias (k = 11) or serious bias (k = 10). The synthesis of 15 impact studies found no significant effects for arrest (k = 14; OR = 1.044, 95% prediction interval [0.527, 2.075], t = 0.335) or incarceration (k = 8, OR = 0.806, 95% prediction interval [2.203, 1.433], t = -1.674). A significant pooled effect was found for conviction (k = 13, OR = 1.209, 95% prediction interval [1.000, 1.462], t = 2.256), but results were highly sensitive to the inclusion of specific studies. No meaningful pattern of results emerged in moderator analyses with respect to study, sample, program component, or program delivery characteristics. The 19 process studies were rated as either high quality (k = 12) or moderate quality (k = 7). Thematic synthesis of the process evaluations revealed 15 themes related to the strengths/challenges of program implementation. The assessment of program cost (k = 7) determined a lack of data within the literature, preventing any summative analysis. Authors' conclusions: Current evidence is promising with respect to conviction outcomes but overall does not find that aftercare/resettlement interventions have a reliably positive impact on crime-related outcomes for young people who have offended. High variability across outcomes and reported data resulted in small sample sizes per outcome and limited moderator analyses. Multiple challenges for program implementation exist; additional rigorous research is sorely needed to further investigate the nuances of the program effects.