Victims' Experiences with, Expectations and Perceptions of Restorative Justice: A critical Review of the Literature

3. Victims Participating in Restorative Justice Programs (cont'd)

3.2 Court-based Programs (cont'd)

3.2.2 Victim-Offender Mediation Programs


In 1995, Umbreit published an assessment of mediation programs in four Canadian cities: Langley, British Columbia; Calgary, Alberta; Winnipeg, Manitoba; and Ottawa, Ontario. The four programs are very different from one another, including the types of cases they target (juveniles/adults), and the point of case referral (from pre-trial to post-sentencing).

The researchers conducted 323 phone interviews with victims from the four different programs. In all, 183 of the victims interviewed had participated in mediation (the experimental group) and 140 victims had their cases referred to the program but mediation did not take place (the control group). The design is an after-only design, which means we do not know if the two groups were initially equal. This makes it impossible to attribute any observed differences between the research groups to the treatment (mediation). The interviews were conducted two months after either the mediation session or the date when the prosecutor, court or related agency otherwise disposed of the case (control group).


Compared to victims in the non-mediation group, those who had participated in mediation were more likely to indicate that the following concerns were important:

  • receiving answers from the offender,
  • telling the offender the impact of the event,
  • receiving an apology from the offender, and
  • being able to negotiate restitution with the offender.

The researchers reported that victims who participated in mediation were more likely to be satisfied with the manner in which the justice system responded to their case (78%) than victims who were referred to mediation but never participated in mediation (48%). However, due to the weak research design, it is not possible to attribute the observed difference to mediation. The groups may not have been equal before the mediation; thus, the observed difference may be due to some other factor such as offence or offender characteristics. Generally, victims in the mediation programs felt that they participated voluntarily (90%). However, a small but significant group felt that their participation was involuntary.

The researchers reported that victims in the mediation group (11%) were less likely than those in the comparison group (31%) to express fear of being revictimized by the same offender and were less likely to be upset about the crime (53% versus 66%). While the researchers attributed the observed differences between groups to mediation, this conclusion cannot be drawn based on the available data.

Canada and the United States

In response to a general hesitancy to implement restorative justice programs for adults, Umbreit and Bradshaw (1997) examined the experiences of victims in adult versus juvenile programs. They provided information on victim experiences in two different programs: one for juvenile offenders based in Minnesota (United States) and one for adult offenders in Manitoba (Canada).

The design employed in this study used unequal comparison groups. The juvenile project involved property crimes (90%) while the adult project involved significantly more violent crimes (40%). Besides a difference in the types of offences, the program procedures were not identical. This study involved a post-test only design as interviews were conducted with victims in both programs only after the mediation session. The problem with a post-test only design is that victims’ attitudes may have been different before participation in the program. Therefore, any observed differences between groups cannot unequivocally be attributed to the different programs.


Umbreit and Bradshaw (1997) compared victims’ satisfaction with the justice system, their satisfaction with mediation, and victims’ subjective experience of distress. The comparison was performed using t-tests. On most items, no significant differences were found. Victims in the juvenile program were, however, more likely to claim that mediation had helped them to participate in the justice system. This difference may reflect differences in procedures rather than the offenders; in the juvenile program, victims had an extra meeting with the mediator. Victims’ subjective experience of distress was defined in terms of how upset the victim was and their fear that the offender would re-offend. Victims in the juvenile program were less likely to believe that the offender would commit another crime against them or against someone else. While it is impossible to be certain that this difference is due to the fact that the offenders were juveniles, it is probable that victims are more likely to believe in the rehabilitative effects of interventions when the offender is juvenile than when he or she is an adult. There was no significant difference regarding how upset victims were about the crime following mediation.


As part of a cross-national study, Umbreit and Roberts (1996) collected data from two mediation programs operating in England (Leeds and Coventry). Like the study mentioned above, these two programs represented both juvenile and adult offenders and accepted a variety of different types of offences. Unlike the Canadian programs, both the British programs offer either direct or indirect mediation. In both programs, most cases (80%) were handled through indirect mediation. The aim of this study was to conduct a cross-national study using common data collection instruments and analysis across victim-offender mediation projects in three countries. Concretely, the findings would be compared to the above Canadian data and the U.S. data presented in paragraphs 3.2.2 and 3.1.1, respectively.

This study used three research groups. The researchers conducted 19 interviews with victims who had participated in direct mediation (the experimental group1), 25 victims who had participated in indirect mediation (experimental group 2) and 26 victims who had their cases referred to the program but mediation did not take place (control group). The interviews were conducted after either the mediation session (direct or indirect) or the date when the prosecutor, court or related agency otherwise disposed of the case (control group). In all, 70 phone interviews with victims from the two different programs were conducted. As in Umbreit’s study (1995), this study design was an after-only design, which means it is not known if the groups were initially equivalent. Again, this makes it impossible to attribute any observed differences between the research groups to the treatment (mediation). In addition, the small number of cases in each of the three groups limits the value of most statistical methods. Unfortunately, the authors did not specify the statistical methods used in the study.


The researchers compared victim satisfaction with the criminal justice system for victims who participated in mediation (direct or indirect) against those who did not; they found that the two groups did not differ significantly. Most victims were satisfied with the criminal justice system (62% for the two experimental groups versus 58% for the control group). They found no statistically significant differences between the two experimental groups – direct and indirect mediation – when comparing victim satisfaction with the criminal justice system. In both groups, a majority of victims (68% direct-mediation group and 57% indirect-mediation group) were satisfied with the criminal justice system.

The authors reported that most victims were satisfied with the outcome of their mediation session. While more victims in direct mediation were satisfied (84% versus 74%), the authors found no statistically significant differences between the direct and indirect mediation groups.

The study also looked at voluntary participation in mediation. Overall, 95% of the victims who went through direct mediation felt participation was their choice, while 70% of the victims who went through indirect mediation reported participation was voluntary. The authors did not offer a possible explanation for this observed difference.

The authors also compared victims’ concerns. They reported that victims in the mediation groups (direct and indirect) were significantly more likely to believe that it was important to receive answers than victims in the non-mediation group. Similarly, victims in the mediation groups felt that telling the offender the impact of the crime was as important as receiving an apology. Once again, however, it is impossible to know whether these findings reflect general predispositions among the mediation group (hence their willingness to participate) or are a result of their participation in the program.

Victims in the mediation group, and particularly in the direct mediation group, were less likely to claim to fear revictimization by the offender. However, the observed differences were not statistically significant and the design did not allow the researchers to draw causal inferences.


Another study using mediation and adult offenders was conducted by Smith, Blagg and Derricourt (1988). They reported on a scheme in Britain involving cases in which the defendant was remanded on bail. The authors emphasized that the aim of the program was mediation and not reparation, which they equated with compensation. The study is based on interviews with 21 victims: 15 victims met their offenders and 6 did not want to meet their offender. While it is a very small study, the researchers did report some interesting findings.


The first response by the victims when they were approached by the probation officer running the scheme was described by the authors as “puzzled.” Some were angry and initially saw the scheme as a plea on the offender’s behalf.

  • The authors cited the following reasons given by victims for not participating:
  • they did not want to see the offender again (victims who knew offender), it would not achieve anything,
  • an apology would be meaningless (offence was clearly premeditated),
  • the victim’s main interest was financial compensation, and
  • victims did not participate for fear they would be seen as foolish.

Among the 15 victims who met their offender, nine were almost exclusively positive concerning the experience, and four were indifferent or cynical (doubted it made a difference). One victim who was assaulted by her two brothers felt it was positive with one brother but negative with the other. Another victim regretted it, wished that she had never agreed and felt that the outcome was more damaging. This last case involved a victim who felt even more anxious after meeting the offender. She was revictimized twice following the initial victimization and suspected the offender had done it out of revenge. The authors pointed out that the mediator should have followed up on the victim’s emotional state of mind after the mediation.


In their article entitled “Mediation for Reparation: The Victim’s Perspective,” Aertsen and Peters (1998) reported on a mediation program in Leuven, Belgium. The program receives cases from the office of the public prosecutor and all cases are returned to this office following mediation. The program is reserved for cases in which the prosecutor’s office has already decided to prosecute. Typically, these are more serious types of offences. The results of mediation are forwarded to the judge, who can take them into consideration when sentencing the offender. The authors described the program as victim oriented. This is reflected, for example, in the program procedure where the mediator first contacts the victim and then the offender. The aim of the program is to offer the victim reparation for material as well as non-material damages. The program is flexible in that mediation can either be direct or indirect. In an effort to protect the victim, the mediator will not even suggest direct mediation if he or she feels that the victim does not seem ready to meet the offender.


Based on an unknown number of case studies and interviews with victims, the authors reported the following reasons given by victims for their participation in mediation:

  • to confront the offender with his or her responsibility,
  • the expectation of a positive impact on the offender,
  • a longing for restitution and reparation,
  • the need for direct information about the reason and the circumstances of the offence, and
  • the need to pass a message to the offender, sensitizing the person to the consequences of his or her actions.

Regarding the desirability of a face-to-face meeting with the offender, between 30% and 50% of victims confirmed that they would like to make use of the opportunity. The authors compared this finding to that of Loschnig-Gspandl and Kilchling (1997), who reported that 45% of all victims in their study expressed their readiness to meet the offender. Loschnig-Gspandl and Kilchling also reported that 30% to 40% of the victims who manifested initial resistance against the idea of a personal meeting with the offender agreed to an indirect mediation, while only 27% rejected any type of mediation.

According to Aertsen and Peters (1998), objections against meeting the offender were mostly related to the victims’ feelings of fear, anger and scepticism about the possibility of a meaningful interaction with the offender. The authors pointed out that the possibility of indirect mediation must always be kept in mind as this is a less threatening procedure for the victim.


The authors reported that an agreement was reached in 50% of all referred cases. However, the evaluation of the project showed that the proposal of mediation and the communication between parties in itself are appreciated more than the agreement.

Aertsen and Peters emphasized the importance of follow-up activities. They reported that it is a disappointing experience for victims when the agreement is not fulfilled.

Date modified: