![Sample our Law journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5942/TOC-Subject-Sample-2021-Law.jpg"})
Abstract
Restorative justice processes, and family group conferences in particular, have become increasingly common in justice system practices across the globe. There has also been significant scholarly interest. Yet, with several notable exceptions, much of the research has been characterized by relatively weak designs, and questions related to the impact of family group conferences on re‐offending remain unanswered. This research is intended to begin to address these issues. The study is based on an experiment conducted in Indianapolis, Indiana whereby young, first‐time‐offending youths were randomly assigned to either a family group conference or one of a number of court‐ordered diversion programs. Nearly 800 youths participated in the experiment, and the cases were tracked for 24 months following their initial arrest. Survival analysis techniques were used to compare prevalence patterns of re‐offending among the treatment and control groups. The results indicated a significant difference between the two groups with the control group experiencing higher rates of failure (re‐offending). The differences were most pronounced during the period of 3–8 months following the initial arrest. Incidence rates were also compared. Assignment to the treatment group was negatively related to incidence of offending. Given the consistent finding of victim benefits in restorative justice processes, the results suggest that conferences hold promise as an effective intervention, at least for young first‐time offenders, and warrant continued experimentation.
Acknowledgments
The authors wish to thank Duren Banks and Nicholas Corsaro for their advice and assistance with the analyses presented herein. This research was supported by grants from the Donner Foundation, Lilly Endowment, Smith Richardson Foundation, Indiana Criminal Justice Institute, and the US Department of Justice, Office of Juvenile Justice and Delinquency Prevention. Points of view expressed are those of the authors and do not necessarily reflect the opinions of the funding agencies.
Notes
1. Hayes and Daly (Citation2003, pp. 726–728) provide an excellent description of common family group conference processes.
2. See also Paternoster, Brame, Bachman, and Sherman (Citation1997), suggesting that perceptions of procedural justice may influence re‐offending in domestic‐violence cases.
3. The initial Indianapolis results were presented in an interim report based on the first 464 youths in the experiment. The present paper represents the complete experiment based on 782 youths.
4. This began June 1, 1998. For the first 7 months of the experiment, approximately 10 youths were placed in each group.
5. VOM and FGC are similar in the sense of bringing victims and offenders together. Consequently, we conducted a series of analyses of the treatment and control groups including and excluding VOM and direct comparisons of VOM and FGC. None affected the findings presented herein. The analyses were complicated by the very low rate of program completion for the VOM program (35 percent completed). Coupled with the fact that there was no random assignment within control group programs, the present findings should not be interpreted as a test of FGC compared to VOM.
6. Some youths and parents chose not to participate in surveys that were conducted as part of this study.
7. Following the initial arrest, youths were processed at the County's juvenile intake center. They were then assigned to either a family group conference or a diversion program (control group). From that point, they had 90 days to complete the assigned program, or they would be returned to court. Youths failing to complete were treated as “non‐completers” but remained in the treatment and control groups in terms of the re‐offending analysis. See subsequent discussion.
8. Family matters (e.g., custody issues, welfare issues), even though they appear on juvenile histories of offending, were not counted towards recidivism.
9. We attempted to fit the data to a Poisson distribution, but the data are overdispersed (variance > mean). In this case, the estimates will be consistent but inefficient (Long & Freese, Citation2003). This is because as the mean increases, the probability of a zero decreases in the Poisson distribution. However, 49 percent of the juveniles in our population have zero re‐arrests. Consequently, the negative binomial regression model is a more efficient and accurate model for our data.
10. There was significant variation within the various diversion programs in terms of program completion. These ranged from 81 percent for community service to 35 percent for Victim Offender Mediation. Given that within the control group, there was not random assignment, we cannot distinguish selection effects from programmatic effects. Yet the contrast between the treatment group and the control group as a whole should not be affected by selection given the experimental design.
11. Although the Indianapolis conferences lasted on average less than an hour, this reflects the less serious nature of the juvenile offending in the Indianapolis program. Conferences handling more serious offending typically run much longer and may require a series of meetings as opposed to one discrete session.
12. It should be noted that this legal definition of failure is distinct from that used by the research team. For the research team, failure was measured as another re‐arrest within 24 months of the initial arrest.