VII. Research and Effectiveness
There is now general agreement that restorative justice practice has shown positive impact on a variety of intermediate and long-term outcomes including victim satisfaction and healing, increased offender empathy, and procedural justice or fairness. In recent years, randomized trials, quasi-experiments, and meta-analyses have also consistently demonstrated positive impact on reoffending. While research on restorative justice is in its early stages, relatively speaking, unlike studies of punitive programs and weak or counterproductive treatment models, no research shows that restorative justice practices make things worse (e.g., increase recidivism).
While only two models of restorative conferencing, FGC and victim–offender mediation, have received consistent and positive ongoing evaluation, other conferencing practices such as neighborhood accountability boards and peacemaking circles have also shown promising early results. In addition, an older body of evaluation research on the aforementioned reparative practices—essentially, restorative obligations or sanctions such as restitution and community service—has consistently found positive impact. Moreover, unlike a wide range of punitive approaches, as well as a large number of treatment programs that consistently report negative outcomes in evaluations and meta-analyses, no studies of reparative practices (e.g., restitution, community service) report negative findings.
Regarding crime victim impact, researchers for more than a decade have been able to make the claim that crime victims who participate in face-to-face restorative justice dialogue processes with offenders experience greater satisfaction than those who participate in court or other adversarial processes. Though “selection effects” leave open the possibility that victims who choose to participate in these processes are predisposed to report greater satisfaction than those who do not, the consistency and strength of these results are nonetheless persuasive. In recent years, experimental research has also verified the effectiveness of restorative conferencing in reducing posttraumatic stress syndrome in crime victims.
Much uncertainty remains, however, about the primary causal factor or specific intervention most responsible for producing the typically higher levels of victim satisfaction in this research. For example, it could be argued that these results are due less to the fact that restorative processes are so effective than to the fact that the court and adversarial process is so harmful. Perhaps restorative dialogue processes simply take advantage of what is often called a “Hawthorne effect,” whereby victims who are simply listened to, treated with dignity and respect, and given a wider array of choices are more satisfied than those who go through court, regardless of the effect of any special face-to-face dialogue with the offender. While this could mean that positive effects of a restorative process are actually a result of what is usually called procedural justice rather than some presumed restorative justice impact, authors of the bulk of restorative research publications tend to view greater satisfaction as itself a restorative justice benefit. Although early studies show independent positive impact of restorative obligations—for example, restitution, community service—unanswered questions remain about what additional positive impact might be attributed to the purely restorative features of the restorative process. Yet, evaluation studies in recent years have generally also shown significant reduction in recidivism—at least some of which has been linked to unique features of the restorative process and its impact, for example, on offender remorse and empathy.