Restorative justice — the broad family of practices that respond to crime by bringing together those most directly affected — the person who caused harm, the person who was harmed, and the community — to address needs, repair relationships, and determine appropriate responses together rather than through the adversarial, state-centered processes of conventional criminal justice — represents both one of the oldest approaches to responding to wrongdoing and one of the most rapidly growing fields in contemporary criminological policy. Its claim against conventional criminal justice is both empirical and philosophical: empirically, that restorative processes produce better outcomes — higher victim satisfaction, better offender accountability, lower reoffending — than adversarial prosecution for many offense types and offender populations; philosophically, that the adversarial model’s focus on punishing offenders crowds out the more fundamental questions of repairing harm, restoring relationships, and addressing the conditions that produced the offense.
Criminology and Public Policy engages restorative justice as a domain where the evidence base has matured substantially over the past three decades, moving from the anecdotal accounts of pioneering practitioners to a rigorous multi-method literature that includes randomized controlled trials, quasi-experimental evaluations, systematic reviews, and meta-analyses. The evidence supports specific and qualified conclusions: restorative justice processes consistently produce high rates of victim satisfaction, consistent reductions in reoffending for participants in well-implemented programs, and cost savings compared to conventional prosecution in contexts where diversion is the appropriate alternative. The policy challenge is translating these findings into institutional reform of a criminal justice system built on fundamentally different principles and serving constituencies with fundamentally different interests.
Introduction
The theoretical foundations of restorative justice draw on multiple intellectual traditions that converge on the critique of conventional criminal justice’s focus on the state versus the offender at the expense of the victim and community whose interests the state nominally represents. Nils Christie’s foundational argument that conflicts are “property” — that the criminal justice system has appropriated from victims and communities the conflicts that properly belong to them, producing bureaucratic responses that serve professional interests rather than the needs of those directly affected — provided the theoretical anchor for restorative justice’s critique of conventional prosecution (Christie, 1977). John Braithwaite’s reintegrative shaming theory proposed that the key distinction in criminal justice responses is between stigmatizing shaming — which excludes the offender from the community and produces defiant criminal identity — and reintegrative shaming — which condemns the offense while reaffirming the offender’s membership in the community — and that restorative processes are uniquely capable of expressing reintegrative shaming in ways that reduce reoffending (Braithwaite, 1989).
Howard Zehr’s victim-centered account proposed that the questions conventional criminal justice asks — what rule was broken, who broke it, what do they deserve? — are the wrong questions for addressing harm, and that restorative justice’s questions — who was hurt, what are their needs, whose obligations are these? — better serve both victims and communities by centering the experience of harm rather than the logic of punishment (Zehr, 1990). These theoretical critiques are not merely normative objections to punishment but empirical propositions about the consequences of different response frameworks — propositions that the evaluation literature has tested with increasing rigor over the past three decades.
The restorative justice landscape encompasses a wide variety of practices that share the core commitment to bringing affected parties together but differ in format, setting, and application: victim-offender mediation (VOM), which brings victims and offenders together with a trained mediator to discuss the offense and negotiate a response; conferencing, which expands the dialogue to include supporters of both victim and offender in addition to the parties themselves; circles, which further expand participation to include community members and may serve sentencing, healing, or community-building functions; and restorative boards and panels, which provide a community-based review process for minor offenses. These practices have been evaluated in diverse settings — juvenile justice, adult criminal courts, schools, corrections, and community settings — with findings that vary by practice type, offense category, and implementation quality.
The Evidence on Restorative Justice Effectiveness
Victim Satisfaction and Victim Experience
The most consistently documented finding in the restorative justice evaluation literature is the superiority of restorative processes over conventional prosecution on measures of victim satisfaction. Systematic reviews and meta-analyses consistently find that victims who participate in restorative justice processes report significantly higher satisfaction than comparable victims processed through conventional courts — with typical effect sizes substantial enough to be practically as well as statistically significant. The mechanisms appear to be multiple: restorative processes give victims a voice in the process and in determining the response; they provide victims with information about why the offense occurred that conventional prosecution rarely provides; they give victims the opportunity to ask offenders the questions they need answered for healing to begin; and they produce agreements that directly address victims’ needs rather than the state’s interest in punishment.
Heather Strang and Lawrence Sherman’s randomized trial of restorative justice conferencing in Canberra, Australia — the Reintegrative Shaming Experiments (RISE) — provided the most rigorous experimental evidence on victim outcomes, finding large and consistent effects on victim satisfaction with the justice process across multiple offense types, alongside reductions in victim post-traumatic stress symptoms that the restorative process facilitated through the opportunity to confront the offender in a safe, controlled setting (Strang et al., 2006). These findings are replicated in the broader victim-centered literature with sufficient consistency to support the conclusion that restorative processes are substantially better than adversarial prosecution for most victim populations across most offense categories.
Table 1. Restorative Justice Outcomes: Evidence Summary by Outcome Domain
| Outcome Domain | Evidence Quality | Effect Direction | Effect Magnitude | Key Studies |
|---|---|---|---|---|
| Victim satisfaction with process | Strong — multiple RCTs and systematic reviews | Consistently positive | Large | RISE; Sherman & Strang meta-analysis |
| Victim post-traumatic stress | Moderate — several controlled studies | Positive (reduces symptoms) | Moderate | RISE; victim impact studies |
| Offender reoffending (overall) | Moderate-strong — meta-analyses | Positive (reduces recidivism) | Small-moderate | Bradshaw et al. meta-analysis; RISE |
| Offender reoffending (violent crime) | Moderate — Sherman & Strang | Positive; larger than for property crime | Moderate | Sherman & Strang; Latimer meta-analysis |
| Offender accountability / agreement completion | Moderate | Positive | Moderate | Multiple comparative evaluations |
| Cost compared to prosecution | Moderate | Positive (cost savings) | Moderate | RAND evaluations; UK studies |
| School discipline applications | Moderate | Positive (reduced suspension) | Moderate | Pittsburgh school RCT; multiple comparisons |
| Community satisfaction | Limited | Positive | Modest | Community conferencing evaluations |
Recidivism Effects
The evidence on restorative justice’s effects on reoffending is more contested than its effects on victim satisfaction — partly because the evidence is of lower average quality (fewer randomized studies), partly because effect sizes are more heterogeneous across program types and populations, and partly because the comparison condition varies across studies in ways that complicate synthesis. The most comprehensive meta-analysis of restorative justice recidivism effects, by Bradshaw and colleagues, found an average recidivism reduction of approximately 26% across studies meeting minimum quality criteria — a substantial effect that, if replicated in real-world implementation, would represent one of the largest recidivism reductions in the criminal justice evaluation literature (Bradshaw et al., 2006).
Lawrence Sherman and Heather Strang’s systematic review for the Campbell Collaboration reached a more nuanced conclusion, finding that restorative justice consistently reduces reoffending for violent offenses — particularly property crimes committed by individuals with personal victim-offender relationships — while showing mixed results for burglary and less consistent effects across offense categories. The strongest evidence for recidivism reduction is for programs that include a genuine encounter between victim and offender rather than offender-only restorative processes — consistent with the theoretical prediction that the accountability mechanism of confronting the person harmed, rather than merely completing a restorative exercise, is the active ingredient in reoffending reduction.
Trauma, Healing, and the Limits of Restorative Justice
When Restorative Justice Is Not Appropriate
The evidence supporting restorative justice does not support its uncritical application in all offense contexts and for all victim-offender configurations. Research and clinical experience have identified several conditions under which restorative processes pose genuine risks to participants — particularly to victims who are pressured into processes they do not genuinely choose, or who face power imbalances with offenders that restorative processes cannot adequately address. Intimate partner violence and family violence cases pose specific risks because the offender’s control mechanisms typically extend to the restorative process, and research has documented cases where mediation in domestic violence contexts reinforced rather than challenged the power dynamics that enabled the violence (Coker, 2002). Most restorative justice programs explicitly exclude active domestic violence cases for this reason, directing them to specialized advocacy and legal processes rather than to restorative alternatives that cannot adequately protect victim safety and autonomy.
Cases where offenders are genuinely remorseless, where victims are not interested in contact with their offenders, or where community accountability for systemic harm — police brutality, corporate environmental damage, institutional abuse — cannot be adequately addressed through individual-level restorative processes all represent limitations on restorative justice’s scope that advocates do not always acknowledge with adequate honesty. Restorative justice is not a panacea for all criminal justice harms but a set of practices that produces better outcomes than adversarial prosecution in contexts where genuine accountability, victim voice, and community participation are achievable — and the intellectual integrity of the restorative justice movement depends on its advocates being as honest about these limits as about the evidence of effectiveness within appropriate applications.
Applications Across Criminal Justice Settings
The New Zealand Family Group Conferencing Model
New Zealand’s Family Group Conference (FGC) model — integrated into the country’s juvenile justice system through the Children, Young Persons, and Their Families Act 1989 and administered by the Oranga Tamariki — Ministry for Children — represents the most comprehensive institutionalization of restorative justice within a national juvenile justice system, providing a model that many other countries have adapted and that criminologists have studied as the most available natural experiment in systemic restorative justice implementation. The FGC model brings together the young person who offended, family members and supporters of both the young person and the victim, social workers, police, and sometimes the victim and their supporters to discuss the offense, its consequences, and an appropriate response — with decisions made by consensus rather than by judicial authority.
Research on the New Zealand FGC model has documented high rates of victim participation and satisfaction, high rates of agreement completion by young offenders, and recidivism rates that compare favorably to both conventional prosecution and pre-FGC juvenile justice outcomes, though methodological limitations of most evaluations — the absence of randomized controls and the difficulty of constructing credible comparison groups for a system-wide intervention — limit the causal confidence that the evidence supports. The New Zealand experience’s most important contribution to the restorative justice evidence base may be the demonstration that systemic restorative justice — not merely pilot programs on the margins of conventional systems — is organizationally feasible and operationally sustainable, a demonstration that is as important for policy as any specific outcome finding.
Juvenile Justice and Youth Crime
Restorative justice has found its most established institutional home in juvenile justice, where its developmental fit — addressing the adolescent identity and peer dynamics that drive juvenile offending while avoiding the labeling consequences of formal processing that developmental criminology predicts will elevate recidivism — and its resource efficiency have made it attractive to juvenile justice systems seeking to reduce incarceration without abandoning accountability. Research on restorative justice diversion in juvenile justice consistently finds lower recidivism among diverted youth than among comparably charged youth processed through conventional court, with the strongest effects for moderate-risk youth for whom conventional processing produces net harm through labeling and criminal network exposure without offsetting deterrence benefits (Wilson et al., 2017).
The Youth Advocate Programs (YAP) and the Balanced and Restorative Justice (BARJ) model have provided implementation frameworks for integrating restorative principles into juvenile probation and diversion in ways that maintain accountability while building the community connections and victim-centered responses that restorative theory prescribes. Research evaluating BARJ-oriented juvenile probation programs has found improvements in victim satisfaction, offender accountability completion, and in some studies recidivism, though methodological quality varies across evaluations and the field would benefit from additional randomized evaluation evidence.
Adult Courts and Serious Crime
The application of restorative justice to adult criminal courts and serious offenses — including violent crime, sexual assault, and even homicide in some victim-initiated programs — challenges the conventional view that restorative justice is appropriate only for minor offenses and young offenders. Research on victim-offender dialogue programs in serious crime contexts — offered to victims as an alternative or supplement to conventional prosecution rather than as a substitute for it — consistently finds that victims who choose to engage in these processes report higher levels of healing, reduced fear and anger, and better understanding of what occurred than comparable victims who did not participate. Mark Umbreit’s research at the University of Minnesota‘s Center for Restorative Justice and Peacemaking has documented victim-initiated dialogue in cases of vehicular homicide, assault, and other serious offenses, finding that the majority of participating victims report the experience as meaningful for their healing process even in the most serious cases.
The Restorative Justice for Oakland Youth (RJOY) program and similar community-based restorative justice organizations have demonstrated that restorative approaches can address serious youth violence in communities most affected by both crime and criminal justice enforcement, providing an alternative to incarceration that addresses community safety needs through the accountability and relationship repair mechanisms of restorative practice rather than the incapacitation and labeling of conventional prosecution. The evidence for these community-based programs is less methodologically rigorous than for structured VOM and conferencing programs, but the theoretical logic and preliminary outcomes data support continued investment and evaluation.
Schools and Community Settings
The application of restorative practices in school settings — as alternatives to exclusionary discipline — has been evaluated with increasing rigor and has shown consistent evidence of reducing suspension and expulsion rates, with some evidence of reductions in school crime and improved school climate. The randomized trial of restorative practices in Pittsburgh Public Schools found significant reductions in suspensions and improvements in school climate in schools implementing restorative practices relative to control schools, providing experimental evidence for what observational studies had consistently found across multiple school districts (Augustine et al., 2018).
The crime prevention significance of restorative school discipline is both direct — reducing the school exclusion that elevates criminal risk through the labeling and peer exposure mechanisms that suspension creates — and indirect — maintaining the school attachment and community connection that social bond theory identifies as among the strongest protective factors against criminal career onset. Schools that implement restorative practices with fidelity and adequate implementation support show larger and more consistent effects than those that adopt the label without the substantive training and organizational change that genuine restorative practice requires — a finding that underscores the implementation science point that program quality determines outcomes regardless of program model.
Restorative Justice and Transitional Justice
The application of restorative justice principles to transitional justice — the processes through which societies address legacies of mass atrocity, systematic human rights violations, and political violence during transitions from authoritarian rule or armed conflict — extends restorative theory to its most demanding applications and connects criminology to international human rights law, political philosophy, and the comparative study of post-conflict societies. Truth and Reconciliation Commissions — most famously the South African Truth and Reconciliation Commission chaired by Archbishop Desmond Tutu following apartheid — represent the most prominent institutional expression of restorative principles at the societal level, prioritizing truth-telling, acknowledgment of harm, and conditional amnesty over prosecution in ways that drew both on restorative justice theory and on pragmatic political calculations about what transitional justice was achievable.
Research on the TRC’s outcomes — measured in terms of victim healing, social reconciliation, and the prevention of renewed conflict — has been mixed: the Commission provided many victims with information about what happened to their loved ones and with public acknowledgment of the crimes committed against them, but the amnesty provisions were experienced by many victims as impunity rather than accountability, and the Commission’s limited power to impose material reparations left economic injustices largely unaddressed. The South African TRC experience illustrates both the potential and the limitations of restorative approaches at the societal level: they can provide truth and acknowledgment that criminal prosecution cannot achieve, while being unable to substitute for the accountability that many victims require and for the structural reform that prevents recurrence. The scholarly literature on transitional justice — developed by scholars including Priscilla Hayner, whose Unspeakable Truths provides the most comprehensive comparative analysis — positions restorative principles as essential complements to retributive mechanisms rather than as alternatives to accountability.
Restorative Justice and Race
The relationship between restorative justice and racial justice is theoretically promising but practically complicated. Restorative justice advocates have argued that restorative processes — by reducing reliance on the racially disparate formal criminal justice system, by centering victim and community voices that adversarial prosecution silences, and by addressing the harms of offenses without the collateral consequences that criminal convictions impose — can serve racial equity alongside the individual accountability and healing goals that restorative theory emphasizes. Research has found that Black and Latino youth who participate in restorative justice diversion show lower recidivism and avoid the labeling consequences of formal processing that racial disparities in enforcement make particularly costly for youth of color.
The concerns about restorative justice and race are also substantive: community conferencing and circle processes that draw on community social capital may reproduce rather than challenge racial hierarchies when community participants share the implicit racial biases documented in the broader justice system; the community values invoked in restorative processes may not reflect the values of marginalized communities who have experienced the criminal justice system primarily as an instrument of control rather than protection; and the individualized, relational focus of restorative processes may inadequately address the structural racism that produces the conditions in which individual offenses occur. Research by Kathleen Daly and others has engaged these tensions carefully, concluding that restorative justice can serve racial equity but only when it is explicitly designed to do so — with facilitators trained in cultural humility and anti-racism, with processes that acknowledge rather than ignore structural context, and with communities that are genuinely representative rather than nominally so.
Corrections and Restorative Justice
Restorative justice within correctional settings — prisons and jails — represents one of the most underexplored applications of restorative theory, despite the theoretical fit between restorative principles and the rehabilitation goals that correctional policy nominally pursues. Victim-offender dialogue programs operating within corrections — where incarcerated individuals who wish to apologize to victims they harmed, or victims who wish to address the people who harmed them, are supported through a structured process facilitated by trained professionals — have been evaluated with consistent findings of positive outcomes for both participants. Research by Mark Umbreit and colleagues at the University of Minnesota has documented that most participants in victim-offender dialogue in serious crime cases report the experience as beneficial for their healing and understanding, despite the severity of the original offenses and the length of time that may have elapsed since them.
Restorative circles within correctional facilities — bringing together incarcerated individuals with community members, family members, and sometimes victims in ongoing circle processes that address accountability, reintegration planning, and community connection — represent a more ambitious application of restorative principles within corrections, building the community relationships and restorative capacity that reentry will require while simultaneously providing the accountability and community engagement that restorative theory prescribes. Organizations including the Insight Prison Project in San Quentin and the Restorative Justice Project in various states have developed circle-based correctional programming that pilot evaluations suggest improves institutional behavior, reentry readiness, and post-release outcomes, though the evidence base remains at an early stage relative to the victim-offender mediation literature.
Institutionalizing Restorative Justice
Policy Infrastructure for Scaling Restorative Approaches
The scaling challenge for restorative justice — moving from successful pilot programs to systematic implementation across the criminal justice system — requires the development of policy infrastructure that conventional criminal justice institutions have not historically provided. Professional training standards for restorative justice facilitators, quality assurance frameworks for program certification, data collection requirements for program monitoring and evaluation, and funding mechanisms that sustain programs across the grant cycles that have historically limited restorative program continuity all represent institutional investments that scaling restorative justice requires.
The International Institute for Restorative Practices (IIRP), the Restorative Justice Council in the United Kingdom, and the European Forum for Restorative Justice have developed training and certification frameworks for restorative justice practitioners that provide minimum quality standards for facilitator competence. The Council of Europe’s Recommendation CM/Rec(2018)8 on restorative justice in criminal matters and the United Nations Basic Principles on the Use of Restorative Justice Programmes in Criminal Matters provide international normative frameworks supporting restorative justice institutionalization that national policies can reference and build upon.
The Cure Violence model and community violence intervention programs more broadly share restorative justice’s emphasis on community processes and credible messenger accountability while operating primarily as interruption and prevention rather than post-offense response — a distinction that reflects the different contexts in which restorative approaches are applied but whose theoretical connections to restorative principles deserve more explicit recognition than the literatures on each tradition typically provide. The development of hybrid models that combine community violence intervention’s proactive conflict management with restorative circle processes for addressing the aftermath of specific violent incidents has been explored by organizations including the Community Justice Exchange and Community Works West in ways that suggest productive connections between the two traditions.
Research on restorative justice’s cost-effectiveness — comparing the total costs of restorative processes (facilitator time, victim advocate time, agreement monitoring) to the costs of conventional prosecution (police time, prosecutor time, court time, incarceration if sentenced) — consistently finds that restorative processes are substantially less expensive per case for low and medium severity offenses, with cost savings estimates ranging from 50% to 80% relative to conventional prosecution in the most rigorous analyses. These cost savings reflect both the reduced bureaucratic transaction costs of restorative processes relative to adversarial courts and the reduced incarceration costs when restorative diversion reduces custodial sentences. For resource-constrained justice systems seeking to reduce costs without compromising accountability or public safety, the cost-effectiveness evidence provides a practical argument for restorative investment that complements the equity and effectiveness arguments.
Research by Heather Strang and Lawrence Sherman comparing the outcomes of restorative justice conferencing and conventional court processing across multiple offense types in their Canberra and London randomized experiments has consistently found that the conferencing conditions show not only higher victim satisfaction but specific reductions in victims’ desire for violent revenge against offenders — an outcome that the conventional justice system’s punitive focus actively cultivates rather than reduces. This finding is among the most theoretically important in the restorative justice literature: it documents that restorative processes do not merely satisfy victims more but actually transform their emotional responses to victimization in ways that reduce the retaliatory impulses that victims’ advocates sometimes invoke to justify harsh punishment as victim-centered justice. A victim who has participated in a restorative conference with the person who harmed them is, on average, less likely to want that person harshly punished and more likely to believe that the response to the offense was fair — findings whose implications for the standard framing of victim-centered justice as synonymous with punitive justice are profound and underappreciated.
Conclusion
The restorative justice evidence base has matured sufficiently to support specific policy conclusions: restorative processes consistently produce higher victim satisfaction than adversarial prosecution; they produce reoffending reductions that are meaningful even if more modest and more heterogeneous than the evidence on some other evidence-based interventions; and they can be implemented in juvenile justice, adult courts, schools, corrections, and community settings with appropriate adaptation to each context. These conclusions justify substantially more investment in restorative justice infrastructure — training, program development, evaluation, and institutional integration — than the criminal justice system has historically provided.
The normative case for restorative justice does not rest exclusively on its evidence of better outcomes — though that evidence substantially strengthens it. Process values matter independently of outcome measurements: the difference between being subjected to a legal process that decides your fate and participating in a human process that addresses the harm you caused or experienced is not captured in any recidivism or satisfaction statistic, however carefully measured. A justice system that treats its participants — victims, offenders, community members — as capable moral agents who can engage with harm, take responsibility, and contribute to repair is not only more likely to produce behavioral change; it is more consistent with the human dignity that justice is supposed to express regardless of its consequences.
The criminological community’s engagement with restorative justice has been productive but incomplete. The evaluation literature has documented effectiveness in specific applications with increasing rigor; the theoretical literature has elaborated the mechanisms through which restorative processes work and the conditions under which they work best; and the practitioner literature has developed the implementation frameworks that bridge theory and practice. What the literature has less systematically addressed is the normative work of integrating restorative justice into a comprehensive account of what criminal justice should be trying to achieve — the philosophical project of articulating how restorative, retributive, rehabilitative, and incapacitative purposes relate to each other and how policy should navigate their tensions when they conflict. That normative work is as important for the field’s long-term development as the empirical work of continuing to evaluate what restorative processes produce.
The normative case for restorative justice does not rest exclusively on its evidence of better outcomes: it also rests on the direct value of giving crime victims a voice in the process that addresses their victimization, on the respect for human dignity that treating all parties — victim, offender, community — as capable of engagement and repair rather than as objects of state processing embodies, and on the democratic aspiration that communities have meaningful roles in responding to the harms that occur within them rather than delegating entirely to professional state actors whose interests and perspectives do not always reflect community needs. The evidence that restorative processes also reduce reoffending and save costs makes the normative case more practically compelling, but the normative case would stand even if the evidence were neutral — because the process values of restorative justice have worth independent of the outcomes they produce.
References
- Braithwaite, J. (1989). Crime, shame and reintegration. Cambridge University Press.
- Bradshaw, W., Roseborough, D., & Umbreit, M. S. (2006). The effect of victim offender mediation on juvenile offender recidivism. Conflict Resolution Quarterly, 24(1), 87–98. https://doi.org/10.1002/crq.159
- Christie, N. (1977). Conflicts as property. British Journal of Criminology, 17(1), 1–15. https://doi.org/10.1093/oxfordjournals.bjc.a046783
- Sherman, L. W., & Strang, H. (2007). Restorative justice: The evidence. Smith Institute. https://www.ifjresearch.org/files/Restorative%20Justice%20-%20the%20Evidence.pdf
- Strang, H., Sherman, L. W., Barnes, G. C., Braithwaite, J., Inkpen, N., & Teh, M. M. (2006). Victim evaluations of face-to-face restorative justice conferences: A quasi-experimental analysis. Journal of Social Issues, 62(2), 281–306. https://doi.org/10.1111/j.1540-4560.2006.00451.x
- Wilson, D. B., Olaghere, A., & Kimbrell, C. S. (2017). Effectiveness of restorative justice principles in juvenile justice: A meta-analysis. Office of Juvenile Justice and Delinquency Prevention. https://www.ojjdp.gov/pubs/250872.pdf
- Zehr, H. (1990). Changing lenses: A new focus for crime and justice. Herald Press.