CHAPTER 8

PILOTING VICTIM-OFFENDER CONFERENCING IN SOUTH AFRICA

Amanda Dissel

Published in Monograph No 111, February 2005

Beyond Retribution
Prospects for Restorative Justice in South Africa

edited by
TRAGGY MAEPA

A victim–offender conferencing (VOC) project was initiated by a consortium of non-governmental organisations (NGOs) in 1999.81 This project operated as a pilot project for one year, after which it was extended for two subsequent years.82 This chapter outlines the motivation, strategies, methodology and outcomes of the project.

The VOC project sought to build on the growing restorative justice movement that has begun to take hold in Canada , the United States and New Zealand. It was conceived as a community based restorative justice approach for dealing with crime through a face-to-face mediated interaction between offenders, victims, and their families or members of other support networks. It aimed to formulate a restorative model more familiar to African customary values, and at the same time empower people to work in partnership with the formal criminal justice system.

Although the VOC project was conceived of as a community based initiative, it was also intended as a diversionary process to relieve the workload of the justice system. It therefore sought to work in close cooperation with the police and justice sectors – primarily those officials based at the magistrate’s courts.

Outline of the project

The project was established in three areas in Gauteng: Dobsonville, Westbury and Alexandra. It was extended to include Odi in the second year. In each of the areas the project partnered with community based organisations that had experience in dealing with the criminal justice system, and some experience of community based mediation or conflict resolution. These organisations, which became known as the implementing ‘sites’, were:

• Conquest for Life in Westbury;

• the West Rand Justice Centre with offices in Roodepoort, and outreach programmes in various areas including Dobsonville in the West Rand;

• the Alexandra Community Law Clinic; and

• Odi Community Law Centre.

The VOC project was conceptualised by a steering committee that also bore overall responsibility for its implementation. In the second and third year, it was run under the auspices of the Restorative Justice Initiative, a consortium of organisations involved in the project. Each implementing site had one member who sat on the steering committee. Each site also recruited approximately 10 mediators from their communities, who received training in mediations skills and restorative justice. One mediator was appointed , and paid a nominal fee, to coordinate mediations.83 A project manager was appointed to oversee the management of the whole project.

Cases were referred to the VOC project by the courts, police and community based organisations. Most cases were referred by prosecutors, who assessed the merits and seriousness of the case and made a recommendation that this be handled by the VOC process. In such cases, the criminal prosecution would be suspended until the VOC process had been completed or the case was resolved. Rather than targeting young offenders exclusively, the VOC project was open to all age groups and types of offenders.

The project aimed to allow victims to express their needs and feelings, and to create an environment for the offender to begin to understand the consequences of his or her actions. This approach allows for the facts and emotions of the dispute or offence to be dealt with in a safe environment. It aims to encourage the parties to move towards reconciliation, redress and restitution through both parties reaching an agreement.

Based on restorative justice, the principles that underpinned the VOC were:

• Acknowledging the injustice: The offender needs to acknowledge responsibility for the offence. The offender has to confront the consequences of his or her action, and see the victim as a person with real feelings and needs. Without this there can be little progress in resolving or reconciling the hurt and damage that has occurred.
  
  

• Restoring the inequity: This involves a delicate process of leveling the power imbalances that exist between the offender and victim as a result of the offence or the nature of the relationship between the parties. It provides a forum where victims and their families are given time to speak and be heard by the offender. They are given the opportunity to express their needs and concerns.
  
  

• Addressing the future: This is the process of developing an appropriate and concrete plan of action accepted by all parties involved. The plan should address symbolic as well as material needs of the victims and must sufficiently spell out the future intentions of the offending parties in order to ensure that revenge or retaliation is not embarked upon.84

Cases referred to the VOC project

A total of 660 cases were recorded as mediated by the sites over the three year period. Of these, 134 were mediated at Odi, 159 in Westbury, 163 in Alexandra, and 204 in West Rand. Additional cases that were not captured in the research process were referred to the sites by the courts, police and community based organisations over the three year period. The discussion below relies on data provided by the implementing sites to one of the organisations that was responsible for analysing the data for research purposes. Additional cases may have been mediated or dealt with that are not reflected here.

Cases were also referred that did not proceed to mediation, or were only partly mediated. Unlike family group conferencing, the project required both the victim and offender to participate. If either one or both parties refused, the mediation was stopped, and the matter referred back to court. When this happened, the reasons were that:

• one or both of the parties could not be found;

• one of the parties did not attend the mediation;

• the victim did not want to participate in mediation;

• the offender did not wish to continue;

• inappropriate referral; or

• the victim withdrew charges before it could be mediated.

Analysis of mediated cases

The analysis below refers only to those cases that did go for mediation. In total 660 cases were mediated.85 There were several cases in which a dispute between the parties resulted in more than one referral. Sometimes these cases were mediated separately, and sometimes together. There were several cases that had been referred by the police, not because a charge had been laid, but because the parties requested assistance in solving their dispute.

Offences

Of the 660 cases mediated by VOC, there were 706 separate charges recorded against offenders. Offenders were charged with a variety of different offences (Table 1). Violent offences against the person were most common, constituting around 62% of the total. The charge of common assault was the single most common type of crime, accounting for 32% of the cases, while assault with intent to commit grievous bodily harm (assault GBH) accounted for 27% of cases.

From the second year of operation the category of ‘domestic violence’ was also included. These were cases that were often referred by the Domestic Violence Units established at the magistrate’s courts, as well as by community based organisations. No formal charges had been laid against the offender in these cases. Nineteen percent of cases referred related to such domestic violence incidents. Because domestic violence is defined very broadly in the Domestic Violence Act (116 of 1998), physical violence was often, but not necessarily, manifest in these cases. In the second and third year of the project there was a deliberate concentration on more serious cases of domestic violence, and this is reflected in the statistics below.

Offences against the dignity of a person included crimen injuria (3%) and defamation of character (1%).88 In the VOC project, the offenders were usually charged with either crimen injuria or defamation of character.

There were also many offences against property. Most of these were charges of malicious damage to property, which accounted for 9% of the cases. The majority of these incidents occurred within the context of a domestic dispute between sexual partners, marital partners and family members. These cases involved broken windows, doors and household furniture.

Theft constituted 6% of cases mediated. Again, most of the theft cases occurred within the context of a domestic relationship, although several occurred within an employment relationship.

Five cases of rape and attempted rape were referred in the second and third years.89 Rape is an extremely serious offence that is increasingly dealt with more severely by the courts. It is unclear whether mediation is a suitable means for resolving these kinds of cases, particularly when it is meant to supplement, rather than complement a criminal trial. It is suggested that mediators should have specific training, and the victim be provided with additional support, before rape cases are mediated in this way.

The participants

In total, 750 offenders and 674 victims were referred to the VOC project in respect of the 660 cases. In several cases there were multiple victims or multiple offenders. The majority of offenders (70%) were male, while the majority of victims (75%) were female. In the majority of cases, for instance in the 1999/2000 year, males perpetrated offences on females in 98 of the 158 (57%) cases. Females perpetrated offences on males in 17 cases. In the remaining 63 cases, both victims and the offenders were of the same gender.

While most diversion and restorative processes in South Africa have so far focused on young offenders, the VOC project accommodated all ages, but most of the participants were adults. The ages of parties varied widely. While the average age of offenders was 33 years, the average age for victims was slightly older, at 35 years. Sixty five (9%) of offenders were younger than 21 years, the youngest being 11, while the oldest offender was 72 years of age.

There were fewer victims (55) than perpetrators under the age of 21 – the youngest victim being nine years. He was the unintentional victim of a stone thrown by a 17 year old boy in his street. Another victim who was 10 years old was assaulted by a woman who accused him of scratching her car as she drove past him. The oldest victim was 79 years of age.

Relationships between victim and offender

Howard Zehr, an influential advocate of restorative justice, says that restorative justice “defines crime as a conflict between persons, putting the individuals and their relationships at centre stage”.90 This is particularly apparent when looking at the circumstances of the disputes, or crimes, referred to the VOC project. Understanding these relationships provides a context for understanding the causes and manifestations, as well as the consequences, for the disputes. It also helps the mediator to assist the parties in arriving at some resolution.

In the majority of cases referred to VOC, there were existing relationships between the victim and offender (Figure 1). A large proportion of the parties (39%) were involved in some form of intimate or sexual relationship with each other. In 19% of these cases, the parties were married to one another.91 Others were dating (7%), co-habiting as a married couple (7%), or involved in a sexual relationship (3%). Three percent of cases involved parties who were divorced or separated from their intimate partners.

One hundred and sixty three cases (24%) involved parties who were in some familial relationship to one another, such as parent and child, aunt and nephew, grandparent and child. In other words, 63% of the cases dealt with disputes that had occurred within a domestic relationship as defined by the Domestic Violence Act of 1998.92 Not all these cases however constituted domestic violence. The Domestic Violence Act defines domestic violence as:

any controlling or abusive behaviour that includes physical, sexual, emotional, verbal and psychological abuse, economic abuse, intimidation, harassment, stalking, damage to property, entry into the victim’s residence without consent, or any other controlling or abusive behaviour (my emphasis).

Clearly this is a very broad definition that could incorporate violent or abusive behaviour occurring within a range of different relationships, and is no longer confined to disputes occurring within a marital, dating or sexual relationship.

Although each VOC case dealt with one particular incident of abuse or violence, the single incident was often part of an ongoing pattern of abuse which included other forms of abusive behaviour such as constant criticism, humiliation, enforced social isolation, physical abuse, shouting and swearing, destruction of possessions, rape and other forms of sexual assault, threats, etc. Not all cases of abuse fitted this pattern, however, with some victims reporting that the incident in question was an isolated one. Less than half of the cases mediated by the VOC project could be classified as domestic violence cases.

The Domestic Violence Act is perhaps the best indication that government intends taking the issue of domestic violence seriously, and seeks to “afford the victims of domestic violence the maximum protection from domestic violence that the law can provide”. The Act provides for the victim to obtain protection orders against the offender, with the threat of sanction for non-compliance. However, it does not deal with criminal prosecution of offenders. These continue to be dealt with by the existing provisions of criminal law.

Given the high number of domestic violence cases, it is open to question whether victim–offender mediation is the most appropriate form of dealing with these cases. Interviews with prosecutors indicated a reluctance to refer domestic violence cases to VOC, particularly in serious cases of abuse, or when the offender had previously been convicted of an offence in a domestic violence case, or when the behaviour formed part of an ongoing pattern.

This, however, was clearly not the case for most VOC incidents. In the 1999/2000 year, 16 of the offenders reported having been convicted of a prior offence, 10 of whom were offenders referred to VOC in terms of a ‘domestic violence’ case. Five had prior convictions for theft; one for drinking and driving; three for common assault, and one for assault GBH. It is not recorded whether any of these convictions related to offences against the same complainant. More offenders had previously been charged with offences but not convicted. Since this project relied on self-reported data, it is not known how many more of the offenders had prior convictions.

Another large percentage of cases (16%) referred was those in which the parties were neighbours.93 Although many of the neighbours were indeed living on separate plots adjacent to one another, several of the disputes also arose between parties living on the same plot of land. The high number of neighbourhood disputes is indicative of the crowded conditions under which people in the selected areas live, particularly Alexandra and Dobsonville. Conditions here are cramped and people often compete over the same limited resources, such as access to washing lines or water. It is not surprising that ‘petty’ irritations become enormous issues for confrontation.

Seventy-two (11%) of the disputes referred were between friends. Forty-five of the cases related to incidents where the parties had some other form of relationship to each other, such as an employment relationship, or where two men, unknown to one another, fought over the same woman.

The offenders were strangers to the victim in only 22 (3%) of the cases. In most victim–offender processes worldwide, the parties are unknown, or slightly known to each other, and it is in this context that the process has been established. It is unusual for offences between strangers to play such a small part of a victim–offender mediation project, and perhaps reflects that prosecutors and police officials perceive victim–offender conferencing to be more suitable when the parties are known to each other.

The mediation process

The main goal of mediation is to:

• encourage the offender to acknowledge responsibility;

• engage in a process of storytelling; and

• formulate a plan of action to deal with the problem of offending.

On average, mediations were completed in two hours and 25 minutes. The largest proportion of cases (39%) took between one and two hours to complete. The mediation times varied significantly from case to case, depending on the complexity of the case, number of people involved, experience of the mediator, and other factors. One case took 30 hours to mediate over a number of different sessions. The shortest mediation time was half an hour.

The mediation aimed to formulate a plan of action to deal with the problem behaviour and develop a plan for the future. If a plan was agreed upon, it was reduced to writing and signed by both parties and the mediator. The agreement would then be forwarded to whatever agency had referred the case to VOC, such as the court, police, or welfare agency.

In certain sites, the court was asked to postpone the trial for a defined period in order to allow the parties to carry out the terms of the agreement. The sites monitored the agreement, and assisted the victim to withdraw charges against the offender if all the conditions in the agreement had been fulfilled. The offender was required to be present in court when the matter was withdrawn. For instance, in the Newlands magistrate’s court, the magistrate would read out the agreement and ascertain the offender’s commitment to it. She would warn the offender of the consequences of non-compliance, and would also warn him/her that stricter action would be taken should he/she commit the same offence again. Sometimes the magistrate elaborated on the agreement, for instance by warning the offender to keep away from the victim.94

An agreement reached through the mediation process was not always a guarantee that the case would be withdrawn from the court role. There were three cases in Alexandra where, although there was an agreement between the parties, the prosecutor refused to withdraw charges. In this regard, the control prosecutor at the Wynberg court stated:

Generally victim–offender conferencing is a good idea. But the problem is that the parties go to VOC, and they settle the problem between themselves. They believe that the criminal case should no longer proceed. The perception I have is that the court is dealing with criminal cases, and the VOC project is dealing with social problems. But it is impacting on the criminal case…I have to assess the facts and make a decision. In many cases, especially the serious ones, I do not withdraw the charges.95

Clearly the prosecutor does bear the ultimate decision as to whether the agreement is sufficient, or whether the cases should proceed to trial. However, failure to accept the agreement can undermine the process when the parties anticipate that their participation in VOC will result in the charges being withdrawn.

No agreement was reached in 52 (8%) of the cases that went to mediation. In these instances, the matter was referred back to court to be dealt with in the usual way.

Agreements between parties

According to the restorative justice paradigm, a crime should be considered in terms of the harm it has caused, and the outcome should not be to punish or rehabilitate, but to repair or compensate for that harm. The VOC project was premised on the notion that all kinds of harm should be considered, not only the harm that is reflected in the criminal charge before the court. The harm could include physical injuries, material losses, psychological consequences, and relational troubles. Restoration can be achieved through diverse means such as restitution, reparation, compensation, apologies and reconciliation. This may be direct, indirect or symbolic in nature. The actions can be addressed directly at the victim, or towards a broader community, or even towards society, such as in community service.96

The importance of the agreements in reducing further offending was alluded to by Morris and Maxwell who conducted research following family group conferences with young offenders.97 They found that reconviction rates are reduced when some of the potentially restorative aspects of the conferences are achieved. Regression analysis suggested that when offenders apologised to their victims, they were three times less likely to be reconvicted than those who had not. Furthermore, offenders were four times less likely to be reconvicted if the victim had attended the conferences.98 A later study found that when the young people had made amends to their victims they were less likely to be reconvicted.99

The agreements arrived at through mediation were part of the restorative process. In the VOC project mediations, there were complex and creative responses to the particular disputes that presented themselves. Although some aspects of the agreement were general, such as an apology, other aspects were very specific, such as forbidding a particular sangoma from coming to the family home. Parties tended to commit themselves to a range of different things that could include an apology by the offender, an undertaking by the victim to withdraw charges, a commitment by the offender to pay restitution, as well as an undertaking to behave in a civil manner towards the victim in future. There were an average of 3.6 different undertakings made in the agreements.

One characteristic of these agreements is that they were often binding on both parties, not only the offender. Both the victim and offender could, and did, make undertakings to one another and to other parties. At times, the victim or offender would also make undertakings that would affect other people, such as family members, friends or children.

Apology

Apology was a vital part of the restorative process and appeared explicitly in 346 (49%) of the agreements. Apology and forgiveness bears a special meaning in African culture. Mafani examines the meaning of these words in South Africa’s African languages.100 In Xhosa and Zulu, the same word is used for forgiveness and apology: “The relationship is such that the offender asks for forgiveness or for peace, while the offended forgives or grants peace, thus giving both inner peace”.101 She writes that in seSotho and Zulu, there is no word for apology, but the process of apologising is actually asking for forgiveness. She argues that forgiveness can be seen as one of the elements in a long term relationship of reconciliation between individuals or groups of people. Thus, in certain cultures, apologies can be one of the most important factors leading to the reconciliation of parties.

Restitution

Restitution appeared to be a less important element in the agreements, arising in 27% (180) of cases. The restitutions took the form of direct replacement for goods damaged or stolen. Another form of restitution was to pay for the cost of medical expenses incurred as a result of an assault, or for the loss of earnings as a result of the victim taking time off from work. In one instance the offender agreed to pay the victim’s legal costs.

Although mediators report that there were several cases where the victim had asked for large amounts in compensation, there were no cases where the restitution or compensation agreed upon was in excess of the direct damage or injury sustained.102 No agreements were made concerning compensation for pain or suffering.

Agreements to address the problem of re-offending

The precipitating factors behind the commission of the offence were varied and numerous. Some of the agreements tried to take this into account when finding ways to discourage the offender from re-offending:

• Drug and alcohol abuse played a substantial role in many of the offences, both on the part of the victim and the offender. Sixty undertakings were made to stop or limit alcohol or drug consumption. Without an obligation to attend counselling or obtain additional support, this is potentially a weak agreement, particularly as many of the parties appeared to be addicted to drugs or alcohol.
  
  

• Offenders agreed to go for counselling in 62 cases. In 154 cases, the offender undertook not to abuse the victim again, in a physical, emotional or verbal manner.
  
  

• Other agreements involved undertakings to resolve disputes more peacefully in future, to improve communication, and to respect one another.

Restoration of relationships

The VOC process itself was aimed at restoring relationships, and mediators reported that the mediations brought people together in order to talk about their grievances.103 Not all the agreements reflected this process, however. Indeed, in 89 cases the parties agreed to either terminate or change the nature of their relationship. Nevertheless, the VOC project provided them with an important and safe mechanism for arriving at this decision. This was also important in cases where one of the parties wished to end the relationship, but the other did not. The VOC process enabled an agreement to be obtained from the reluctant party. Thus mediators reflected that: “VOC is a perfect platform to mend relationships or to dissolve them in a peaceful manner”.104

Practical changes in the lives of the parties

There were 115 agreements in which parties undertook to make changes to their lives that would affect the relationships, and hopefully prevent further disputes arising, or avoid re-offending. The most common changes were in living arrangements, with one party moving out, another party moving into a house, or a particular room, or an undertaking to make the house accessible to all family members. Other practical measures included an undertaking to find a job, to take a particular medication, to request the police to return a confiscated gun, or to not leave the gunlying around the house in future. All of these were specific responses required by each situation.

VOC as a solution for South Africa

Crime is increasingly a concern in South Africa, with levels of serious violent crime reaching unacceptable highs in recent years. The criminal justice system is struggling to turn this tide, resulting in ever increasing delays in the courts and more awaiting trial prisoners held for long periods of time. Government’s ability to investigate and successfully prosecute crimes has not improved and the conviction rate is extremely low. In this context, finding alternative ways of dealing with crime that are accessible and acceptable to all South Africans, should be a priority.

It is tempting to ask whether the victim–offender process is as effective as a prosecution and conviction through the formal criminal justice system. Aside from the difficulty in evaluating ‘success’ or ‘effectiveness’ of either of these processes, it is impossible to prove that one system is better than the other. Of more interest is an examination of the benefits of each system, and to look at the context in which each would be most appropriate.

Part of the VOC project design required the mediators to contact the parties between one and three months after the mediation. The purpose was to determine victim and offender satisfaction with the process, and also to determine fulfillment, or non-fulfillment, of the agreement. Although this was a requirement in the second and third years, only 191 cases (29%) had been followed up.

All participants reported a high level of satisfaction with the process, suggesting that there is a workable system outside of the courts for dealing with offending. This process allows both victim and offender to be fully involved in all aspects of the case and its resolution. Further studies can determine whether the mediations have a lasting impact on offending behaviour, particularly since most of the problems occurred within a long term relationship between the parties.

Another important aspect is that whereas 5,5% of cases reported at the police station result in conviction using the formal criminal justice route, 76% of offenders who were referred to the VOC project acknowledged responsibility for the offence before the cases were referred. While an acknowledgement of responsibility in no way compares with the legal status and moral force of a criminal conviction, it does have meaning to the victim and offender of that particular offence.105 The VOC process results in the offender taking responsibility and acknowledging accountability for his or her wrongdoing in a higher percentage of cases than in those processed through the courts. In addition, 165 victims (86% of those followed up) reported that the agreements had been completely fulfilled. Other victims reported a partial fulfillment of the agreement.106

It is also important to develop a system that is compatible with South African values and identity. A key issue is that in many respects victim–offender conferencing reflects traditional African values, such as ubuntu, which is taken to mean the essence of humanness. Ubuntu has also been described as a philosophy of life, which represents personhood, humanity, humanness and morality.107

This concept has underpinned various traditional African ways of resolving conflict through the Magotlas or Inkundla, where reconciliation, restoration and harmony were seen as the basis for adjudication. The victim, the offender, and the community were placed at the heart of the dispute, and the main purpose of the adjudication was to acknowledge the wrong and to make amends for the harm done. Like restorative justice, these systems emphasised a communal approach to dealing with conflict, and saw the law not as a tool for personal defence, but for the protection of common interests. Justice Mokgoro argues that under this system, the conciliatory character of the adjudication process aims to restore peace and harmony between members, rather than the adversarial approach to litigation. The importance of solidarity requires restoration of peace between litigants, rather than a victor and a loser.108

However, with increasing development and urbanisation, these traditional systems have been eroded. Despite this, traditional values can still be recognised in a modern context. The concepts of ubuntu are aligned with those of the country’s constitution, and with those of restorative justice. We need to look at ways of bringing these concepts together in the development of a justice system that is fair and accessible to all people.

Victim–offender conferencing offers a system that is flexible enough to incorporate the parties’ belief systems. It is highly accommodative of different cultures. Because they are mediated by someone from the same community as the parties, there is a greater potential that the same value system will be shared. The mediators are able to handle the parties with sensitivity, and to assist them in arriving at a resolution that is appropriate to their situation and culture.

It appears that victim–offender conferencing may be one solution to dealing with some crimes in South Africa. The VOC project indicated that it might be effective in dealing with crimes of a more minor nature. Its applicability to more serious crimes still needs to be developed and tested. Restorative justice solutions have been used satisfactorily alongside formal criminal justice processes in other countries,109 which suggests a route for dealing with serious crimes in South Africa as well. The Department of Correctional Services has adopted restorative justice as a strategy, opening the possibility to work with offenders convicted of more serious crimes. The support of government and the courts is essential if this important restorative justice work is to continue. It is now necessary to move beyond the policy phase into implementation of real restorative initiatives.