The Indianapolis Restorative Justice Experiment
In 1996, the Hudson Institute, a public policy research organization in Indianapolis, IN, began working with the Indianapolis police department, sheriff's department, juvenile court, prosecutor's office, and mayor on a project involving the use of Australian-style restorative justice conferences as an alternative response to juvenile offending. Encouraged by research from other jurisdictions—yet seeking clearer answers about the effects of conferences—Juvenile Court Judge James Payne and Marion County Prosecutor Scott Newman agreed to work with the Hudson Institute's research team to implement an experimental design. The experiment was initiated in September 1997, and this Bulletin presents what the research team refers to as the "Stage One" results of the ongoing experiment.
Program eligibility. Indianapolis justice officials decided to begin using restorative justice conferences with young, first-time offenders. This population was considered the most appropriate both because such youth were not seen as posing an immediate risk to the community and because officials recognized the need to identify more effective early interventions for these youth. The research team hoped that conferences might provide a more effective tool to prevent young, first-time offenders from becoming deeply entrenched in delinquent behaviors.
Consequently, to be eligible for the first phase of the Indianapolis experiment, an offender had to meet the following criteria:
- Be no older than 14 years of age.
- Be a first-time offender (i.e., have no prior adjudications).
- Have committed a nonserious, nonviolent offense.
- Have no other pending charges.
- Admit responsibility for the offense.6
With the exception of the age criterion, these requirements are essentially the same as those that apply to juvenile court diversion programs. If deemed eligible for such a program, an offender is diverted from court and charges are not filed, pending his or her successful completion of the assigned diversion program.
Random assignment procedure. Formal implementation of the Restorative Justice Conferencing Experiment began on September 1, 1997. Court intake officers screened youth for eligibility. Eligible youth were selected for the program through a random assignment procedure. Specifically, when the intake officer determined that a juvenile offender met the program's eligibility criteria, he or she drew an envelope from a stack prepared by the research team. Each envelope in the stack contained one of two possible responses: "yes" or "no." If the intake officer selected a "yes," the youth was assigned to the restorative justice program and the case was turned over to the county coordinator. A "no" selection indicated normal processing, and the youth was assigned to 1 of 23 other diversion programs.
Sample characteristics. From September 1, 1997, to September 30, 1999, 458 youthful offenders participated in the Indianapolis Restorative Justice Conferencing Experiment. Of these, 232 were assigned to the restorative justice treatment group and the remaining 226 to the "control group." Tables 1 through 3 provide descriptive characteristics of both groups.
Table 1, which reports the racial composition of the two groups, shows that the control group included slightly more nonwhite youth (63 percent) than the restorative justice group (58 percent), though the difference was not statistically significant. These percentages are consistent with the racial composition of the general population of Indianapolis youth adjudicated delinquent in 1998—62 percent of whom were nonwhite (Marion Superior Court Probation Department, 1999).
The percentages of male and female offenders in the two groups also indicate that the sample was representative of the general population of juveniles adjudicated delinquent in Indianapolis. For example, approximately 65 percent of adjudicated juveniles in Marion County in 1998 were male, compared with 63 percent of those in the experimental sample (conference and control group combined) (see table 2). The restorative justice group, however, included more males (68 percent) than the control group (57 percent). Although in early analyses researchers were concerned about overrepresentation of males in the restorative justice group, the relative distribution became more even between the two groups as the sample size increased, suggesting that the randomization process is "smoothing out" the initially uneven distribution.
The median age of youth in both groups was 13.0 years. The age distributions of youth in the restorative justice and control groups were also quite similar. Approximately 32 percent were age 14, just over 26 percent were age 13, and approximately 40 percent were age 12 or younger. Previous research has suggested that these young age groups have high rates of reoffending (Snyder and Sickmund, 1995).
Table 3 reports the frequency of primary offenses committed by youth in the restorative justice and control groups. As indicated in the table, conversion (shoplifting) was the most common offense, followed by battery, theft, and criminal mischief. The control group included slightly more youth whose primary offense was conversion, whereas the restorative justice group included more youth charged with theft. Percentages of youth in the two categories combined, however, are almost equivalent for the two groups. Battery (assault) charges accounted for one-quarter of youth in both groups.
Measures. The study had process and outcome measures, including conference observations conducted by trained researchers using an observational checklist; interviews of offending youth, their parents or guardians, and victims; and checks of court records to determine whether participating youth had been rearrested for subsequent offenses.
|
Case Study: A New Approach to Juvenile Offending
An Opportunity To Speak When asked what he was thinking at
the time of the incident, Jason replied,
"Nothing, just that I saw the speakers
and wanted them." Jason struggled
when asked who had been affected by
his actions, telling the group that he
had been affected—by being taken to
"juvenile." "What about the owner of
the car?" asked the coordinator. "Well,
I guess because she got her speakers
messed up, she was affected." Pausing
for a moment, Jason looked at his
mother and whispered that she too
had been affected by his behavior.
Jason's friend Michael gave his account
of what happened, admitting
that he wasn't thinking at the time and
now knows he made a big mistake.
The person most disappointed in
Michael, he explained, was his
younger brother, and that was the
worst part of all this—losing his brother's
trust.
Rhonda next described the incident,
explaining that she heard the two boys
in the parking lot and ran out to see
what was happening. "I saw the one
boy in my car holding the speaker—I
yelled at him to stop and he dropped it
and ran." When asked what she wanted
to receive from the conference,
Rhonda said she wanted to know why
the boys had attempted to steal her
speakers. She also wanted the boys to
understand how she felt and asked
them how they would feel if someone
took their possessions.
Moving around the circle, the conference
coordinator asked the boys'
mothers how the incident had affected
them. Jason's mother said that at first
she was shocked and had a hard time
believing her son would be involved
in something like this. Jason, she explained,
has money from an allowance
and doesn't need to steal anything.
Michael's mother told the group how
disappointed she was that her son had
participated in the incident. She had
always tried to raise her boys to know
the difference between right and
wrong, and it would take a while to
restore her trust in Michael.
Drafting a Contract When asked if there was anything else
she wanted to add to the contract,
Rhonda explained that because the
speakers were replaced and her car
had no permanent damage, restitution
was not necessary. She suggested,
however, that the boys perform community
service work. Following Rhonda's
suggestion, the conference participants
joined in and traded ideas on what type
of work would be appropriate and how
many hours would be fair. The boys
were asked whether they would agree
to community service and whether they
knew of any work that was needed
around their neighborhood. Finally, the
participants agreed that Jason and
Michael would perform 20 hours of
service at a community center to earn
money to pay their court fees. The coordinator
wrote up the contract, and
all of the participants signed it, putting
a formal end to the incident. Benefits of the Process |
Observations of conferences. In observing restorative justice conferences, researchers examined the length of the proceeding; the role of the conference coordinator; the involvement of the offender, youth supporter(s), victim(s), and victim supporter(s); expressions of shame, apology, and acceptance of responsibility by the offender; and elements included in the reparation agreement. Between September 1, 1997, and September 30, 1999, 182 conferences were conducted. Of these, 157 conferences (86 percent) were observed by 1 of 15 trained observers.
Length of proceeding. Restorative justice conferences lasted an average of 43 minutes. The reintegration ceremony, during which conference participants mingled informally and shared refreshments, averaged 10 minutes from the close of the conference.
Role of conference coordinator. Generally, conference coordinators followed the principles of restorative justice conferencing. Observers noted that coordinators maintained a distinction between the offending youth and his or her behavior (i.e., treating him or her as a valued member of the community while condemning the act). Coordinators also focused the discussion on the incident and rarely lectured the offending youth. Coordinators were seen as doing an effective job of eliciting the involvement of all conference participants.
Involvement of offender, victim, and supporters. Observers reported that all conference participants tended to display respect toward the offending youth. In a large majority of conferences, the offending youth also was seen as conveying respect toward the victim. In approximately 22 percent of conferences, observers did not believe the offending youth had been respectful of the victim.
In nearly all conferences, group participants expressed disapproval of the offense. In more than 80 percent of the conferences, observers reported that the youth had apologized to his or her victim, and in half of the conferences, the youth apologized to his or her own supporters. Observers also noted that most offending youth expressed remorse (76 percent) and understood the injury or harm they had caused (66 percent). Although observers could not tell with certainty whether a victim and other group participants had forgiven an offender, observers reported that more than 80 percent of the conferences appeared to include the victim and the group forgiving the offending youth. In three-quarters of the conferences, the observer reported a strong sense of reintegration at the conference close.
In all of the conferences, every participant signed the reparation agreement. Victims appeared satisfied in more than 80 percent of the conferences, and observers described 77 percent of the conferences as positive. Observers also reported that in more than 80 percent of the conferences, a volunteer was appointed to hold the youth accountable to the terms of the reparation agreement. That is, rather than have a court official monitor the agreement, the group designated someone from the community of support to hold the youth accountable. This person was then contacted by the Marion County Restorative Justice Coordinator to verify the youth's completion of the agreement.
Elements of reparation agreement. Apology was the most common element included in reparation agreements (62 percent). To some extent, however, this percentage underrepresents the frequency of apologies. Because many conferences had already included an apology, it may not have been written into the formal agreement. Other common elements included monetary restitution, personal service, and community service. More than half of the reparation agreements included still other elements (typically activities that the group had tailored to the specific circumstances involved). Examples included imposing a nightly curfew and requiring that the youth improve his or her grades and school attendance or participate in afterschool programs.
Interviews of conference participants. A significant part of the Indianapolis restorative justice study was assessing how victims, offenders, and supporters felt about restorative justice conferencing as an alternative to traditional court-ordered programs. The goal was to collect data on participants' attitudes and beliefs about how their cases were handled and on their sense of justice.
Initially, the Hudson Institute encountered delays in implementing the interview procedures. Consequently, the sample size for the interviews is smaller than that of the total sample of conference and control group cases.7 Thus, the results from the interviews come principally from cases occurring during late 1998 and 1999. Given the small sample sizes, the researchers report descriptive findings without assessing the statistical significance of the findings. More detailed assessments will be included in the second stage of the project.
Satisfaction. When respondents were asked how satisfied they were with the way their cases were handled, a significant difference emerged between victims in the control group and victims in the conference group. More than 90 percent of victims in the conference group "strongly agreed" or "agreed" that they were satisfied, compared with 68 percent of victims in the control group (see figure 1). Satisfaction levels of youth and parents in both groups were similar. Overall, both groups expressed high levels of satisfaction, but youth and parents in the control group were slightly more likely to express satisfaction. This difference may reflect the extra demands (e.g., time, accountability) that conferences place on youth and parents.
In measuring participant satisfaction, the study also examined whether participants would recommend the program to a friend involved in a similar situation. Again, the greatest difference between the control and conference groups was for victims. Nearly all victims involved in conferences (98 percent) said that they would recommend the approach, compared with 24 percent of victims in the control group. Offending youth in the conference group were also more likely to recommend the approach (85 percent, compared with 38 percent of youth in the control group). The study found no significant difference between parents in the two groups for this item (see figure 1).
Another indication of participants' satisfaction is whether they would recommend discontinuing the program. Most participants did not recommend stopping the conferences or the control group programs. Conference participants, however, were most likely to endorse continuation of the conferencing program. For example, no victims in the conference group recommended discontinuation. Just over one-fifth of victims in the control group, however, agreed that the program should be stopped. Similarly, 19 percent of youth in the conference group recommended discontinuing the program (compared with 36 percent of those in the control group), and 17 percent of conference parents recommended discontinuation (compared with 25 percent of control group parents) (see figure 1).
The final indicator of participant satisfaction examined was whether participants believed the program was a "good" way to address certain kinds of juvenile crime. Here, both conferences and other court programs received strong endorsements. For victims and parents, the study found little difference between conference and control group participants. Youth in the control group were more likely than those in the conference group (85 percent versus 71 percent) to agree that the program they participated in was a good one (see figure 1).
Perceptions of respect and involvement. Participants in both conference and control group programs felt they had been treated with respect. The study found no significant differences between participants (victims, youth, and parents) in the treatment and control groups in terms of perceptions of respect (see figure 2).
None of the victims in the conference group reported feeling pushed around. However, approximately 20 percent of youth and 15 percent of parents in the conference group felt they had been pushed around. These percentages are lower than those reported by youth and parents in the control group (44 and 38 percent, respectively).8
The study found differences in the two groups' feelings of having been involved in the process. Restorative justice conferences are built on the principle that affected parties should participate in the process, and results indicate that this principle is being achieved in the Indianapolis experiment. Nearly all victims in the conference group (97 percent) agreed they had been involved, compared with 38 percent of victims in the control group. Offending youth in the conference group were also much more likely than those in the control group to feel they had been involved (84 percent versus 47 percent). Nearly 80 percent of parents in the conference group agreed they had been involved, compared with 40 percent of parents in the control group (see figure 2).
Participants in the conference group were also more likely to report having had an opportunity to express their views. For example, 95 percent of victims in the conference group agreed they had such an opportunity, compared with 56 percent of victims in the control group. Similarly, 86 percent of offending youth and 90 percent of parents in the conference group agreed they had the opportunity to express their views, compared with 55 percent of youth and 68 percent of parents in the control group (see figure 2).
Perception of outcomes. A large majority of participants in both the conference group and the control group believed the outcome of their case was fair (see figure 2). Victims in the conference group were more likely than their control group counterparts to describe the outcome as lenient (36 percent and 14 percent, respectively). Conference group youth were slightly less likely than control group youth to describe the outcome as lenient, whereas conference group parents were somewhat more likely than control group parents to describe the outcome as lenient.
Participants in the conference group were more likely than those in the control group to report that the program had helped to solve problems. More than three-quarters of victims in the conference group reported this benefit, compared with one-half of those in the control group. More than 80 percent of conference group youth and parents reported that the program had helped to solve problems, compared with 57 percent of control group youth and 72 percent of control group parents (see figure 2).
|
Case Study: Better Addressing the Needs of Victims
When setting up the restorative justice conference, the coordinator talked with 17-year-old Richard about the purpose of the meeting. Richard admitted that he had broken into his neighbor Sue's car and taken her tape player and several other items. Richard agreed to participate in the conference and indicated a willingness to make amends. On the day of the conference, however, Richard's attitude seemed to walk into the room in front of him, and the other participants sensed that the conference might not go as expected. Sue, the car's owner, nonetheless wanted to proceed. When the coordinator questioned Richard about the incident, Richard skirted the issue of his responsibility and did not appreciate that so many people had attended the conference to help give him a second chance. When it was Sue's turn to speak, she described how she had felt when she discovered someone had broken into her car and stolen her personal property. Looking directly at Richard, Sue asked him why he had chosen her car. After all, she thought they had been friends. After Richard and Sue described the incident, other participants had an opportunity to speak. Gary, a friend of Sue's attending the conference as a victim supporter, explained how Richard's behavior had affected Sue. One of Richard's neighbors told the group that she had always trusted Richard (e.g., allowing him to work in her yard), but now her trust in him had been broken and she wasn't sure how she felt about Richard. Richard's mom told the group that she hadn't raised her son to steal from others but didn't know how to help him change. Once each participant had spoken, a contract was written. Under the terms of the contract, Richard agreed to pay for damage to Sue's car and replace her personal items. After Richard left the conference, Sue commented that she didn't know if she would ever see the restitution payment. She assured her friends and the conference coordinator, however, that the conference had been worth it to her. The most important part, she explained, was the opportunity to tell Richard face-to-face how he had hurt her—that he had destroyed the trust that she had in him, disrupted her sense of safety, and generally made her life miserable for a while. |
Analysis of program completion data and rearrest records. The results described thus far indicate that restorative conferences were implemented in a fashion consistent with the philosophy and principles of restorative justice, that they were more effective than many other court programs in addressing victim needs, and that both parents and offending youth felt very much involved in the process. For many policymakers, however, the fundamental issue is the program's impact on future offending. To address this issue, the study compared program completion data and recidivism rates of restorative justice conference participants with those of youth in the control group.9 (Recidivism was defined as a rearrest after the initial arrest that brought the youth to the juvenile justice system, and recidivism analysis was conducted for both groups at 6- and 12-month intervals.)
Program completion. Youth participating in restorative justice conferences demonstrated a significantly higher completion rate (82.6 percent) than youth in the control group, who were assigned to other diversion programs (57.7 percent). The majority of the 29 youth in the conference group who failed to complete the program were rearrested before attending the conference. In contrast, most of the 71 control group youth who failed to complete their assigned programs failed because of juvenile waiver from the program. In such cases, juvenile court staff closed the case without requiring the youth to complete the assigned program.
Six-month rearrest rates. Table 4 shows 6-month rearrest rates for all youth who have reached the 6-month stage. As the rates for the full sample reflect, the restorative conference group included fewer recidivists than the control group by a margin of 13.5 percent. This statistically significant difference represents a 40-percent reduction in rates of rearrest.10 (The reduction was calculated by dividing the difference between the control and the treatment group rates by the control group rate: (33.9–20.4)/33.9=39.8.)
Researchers also conducted an analysis limited to youth who had successfully completed a treatment (either the restorative conference program or one of the control group diversion programs). Because, as noted above, youth in the conference group were significantly more likely to complete their program than youth in the control group (many of whom were waived out of their programs), the portion of high-risk youth remaining among program completers presumably was higher for the conference group than the control group. In other words, the higher dropout rate for youth in the control group likely resulted in a group of lower risk youth among those who actually completed the program. Thus, limiting the recidivism analysis to program completers provides a conservative estimate of the conference program's effectiveness. This analysis also found a significant difference in rearrest rates for conference and control groups: 12.3 percent and 22.7 percent, respectively. This statistically significant difference represents a 46-percent reduction in rates of rearrest.11
Twelve-month rearrest rates. Table 4 shows 12-month rearrest rates for all youth who have reached the 12-month stage. Of the full sample of youth participating in the restorative conference program, 30.1 percent had been rearrested within 12 months, compared with 42.3 percent of youth in the control group. This statistically significant difference represents a 29-percent reduction in recidivism.12
When researchers examined rearrest rates at 12 months for only those youth who had successfully completed a program, they found a pattern that was consistent with their other results, but the difference in rearrest rates for the conference and control groups did not achieve statistical significance. Specifically, 23.2 percent of youth who successfully completed the restorative conference program had been rearrested at 12 months, compared with 29 percent of youth who successfully completed another diversion program. This represents a 20-percent reduction in rearrest rates, which is not statistically significant.
The lack of statistical significance probably is attributable to two factors: (1) implementation problems in the earliest phase of the experiment, which frequently caused delays in scheduling conferences; and (2) the small number of program completers, particularly in the control group, included in the 12-month analysis. The Hudson Institute continues to monitor these findings to determine whether 12-month rearrest differences for program completers reach statistical significance when the sample size is larger.13Rearrest rates by offense, sex, and race.
Researchers conducted limited analyses
of 6-month rearrest rates for selected subgroups
of offenders.14 Youth who committed
offenses against property had lower
rearrest rates than youth who committed
offenses against persons, and this difference
was comparable for conference and
control group youth. Both males and females
in the conference group experienced
lower rearrest rates than their
counterparts in the control groups; the
difference was greater for females than
for males. There were no racial differences
in rearrest rates for conference and
control group youth, and the overall reduction
in rearrest rates found for conference
group youth was the same for
whites and nonwhites. These findings,
although preliminary, suggest that the
effects of conferences appear consistent
for youth across groups based on offense,
sex, and race. These results should be
considered preliminary, however, until
further analyses based on larger sample
sizes can verify findings.
| Previous | Contents | Next |
 |
|
| Restorative Justice
Conferences as an Early Response to Young Offenders |
|