Evaluating a Community-Based Restorative Justice Pilot in Urban Districts

Thesis Abstract

Community-based restorative justice (RJ) programs have gained prominence as alternatives to formal criminal justice responses in urban contexts, yet empirical evidence on their effectiveness, pathways, and contextual constraints remains limited. This study addresses the gap by evaluating a pilot RJ initiative deployed across five demographically diverse urban districts, aiming to determine its impact on recidivism, perceptions of fairness, victim satisfaction, and community cohesion. The central aim is to assess whether the pilot reduces reoffending rates among participating offenders, enhances victim trust, and strengthens community responsiveness to conflict, while identifying operational success factors and barriers to scale. Specific objectives are to (1) measure changes in recidivism over a 24-month follow-up period for participants in the RJ process compared with a matched conventional-case cohort; (2) examine shifts in perceived procedural justice, legitimacy of the program, and victim satisfaction using validated scales; (3) evaluate changes in community cohesion indicators, such as social trust and bystander willingness to intervene; (4) identify process indicators related to facilitator training, case triage, and stakeholder collaboration; and (5) develop a theory-informed framework for scaling RJ in similarly structured urban environments. The study adopts a mixed-methods design anchored in a quasi-experimental approach complemented by qualitative process evaluation. The population comprises offenders referred to the RJ pilot (n = 260) and a matched comparison group who progressed through traditional court-based resolutions (n = 260) drawn from district court records and probation databases. Purposive sampling ensures representation across age, gender, offense type, and district. Data collection instruments include (i) offender and victim surveys adapted from the European Forum for Restorative Justice instruments, (ii) a district-level community survey to gauge social capital indices, (iii) program administrative records for case processing timelines and completion rates, and (iv) semi-structured interviews with facilitators, probation officers, prosecutors, defense counsel, and community leaders. Validity and reliability are aided by pilot testing of surveys (n = 60) and triangulation across quantitative and qualitative data. The analysis employs (a) propensity score matching to create a robust comparison group, (b) difference-in-differences (DiD) estimation and multilevel logistic regression to evaluate recidivism and perceived justice outcomes, (c) repeated-measures ANOVA to assess changes in victim and community perceptions over time, and (d) thematic analysis of interview transcripts guided by the grounded theory approach to identify implementation determinants and contextual factors. A cross-case synthesis across the five districts informs the conceptual model of RJ effectiveness in urban settings. Key expected findings include a statistically significant reduction in 24-month reoffending rates among RJ participants relative to the matched control, accompanied by improved victim-reported satisfaction with process outcomes and heightened perceived legitimacy of the RJ approach. It is anticipated that districts with higher facilitator fidelity, stronger stakeholder engagement, and timely case completion will exhibit greater improvements in community cohesion and proactive bystander behavior. The study also expects to uncover barriers such as resource constraints, cultural and linguistic diversity, and inconsistent referral pathways that attenuate program effectiveness. Theoretically, findings are expected to support the applicability of restorative justice theory and procedural justice theory in urban, mixed-ethnicity settings, while extending elaboration likelihood theory to illuminate information processing about RJ among diverse communities. The contribution to knowledge lies in providing rigorous, contextually grounded evidence on the effectiveness and operational conditions of a community-based RJ pilot in metropolitan environments, informing policy design, resource allocation, and scalable implementation models. The study culminates in a practical framework outlining minimum standards for facilitator training, stakeholder collaboration, data governance, and monitoring-and-evaluation protocols to guide replication in comparable urban districts. Policy implications emphasize aligning RJ practices with proportionality, accessibility, and victim-centered objectives, while recommendations advocate for ongoing fidelity monitoring, adaptive governance structures, and a phased expansion plan contingent on demonstrable outcomes. A cautious note of conclusion suggests RJ can reduce reoffending and bolster community trust under conditions of consistent implementation, robust referrals, and sustained community engagement, with scaling contingent on resolving resource, cultural, and coordination challenges identified through the mixed-methods assessment.