Restorative justice innovations in Canada Behavioral Sciences and the Law Behav. Sci. Law 20: 363–380 (2002) Published online in Wiley Interscience (www.interscience.wiley.com). DOI: 10.1002/bsl.498 Restorative Justice Innovations in Canada Robin J. Wilson, Ph.D., C.Psych.*, Bria Huculak, LL.M., Ph.D. Candidate and Andrew McWhinnie, M.A. As many jurisdictions move towards more retributive measures as a means to address public discontent with crime, a parallel movement has developed in regard to restorative justice. This article presents three restorative initiatives currently in use in Canada. Each initiative addresses offender behavior and community engagement at a different point in the justice continuum. The use of Sentencing Circles is an example of how restorative justice principles can be instituted at the front end, prior to an offender becoming lodged in the system. The Restorative Justice Options to Parole Suspension project demonstrates how community engagement can assist in preventing offenders from being returned to the system once they have achieved conditional release. The Circles of Support and Accountability project has enlisted the support of professionally supported volunteers in the community reintegration of high-risk sexual offenders. These initia- tives are presented within a framework of effective cor- rectional interventions and increased empowerment for a variety of stakeholders. Copyright # 2002 John Wiley & Sons, Ltd. It is clear that the Canadian justice system is unsatisfactory for both victims and offenders. Many jurisdictions in Canada and the United States, among others, have converted to the ‘get tough on crime’ zeitgeist. This call to get tough is a demand for more official punishment, to inflict more pain, to separate and sever social bonds and, in some countries, to execute more citizens. These official responses to crime create an illusion of safety and allow some to feel good about our official treatment of offenders. While the retributive responses above have garnered favorable attention Copyright # 2002 John Wiley & Sons, Ltd. *Correspondence to: Dr Robin J. Wilson, Sexual Behaviour Program, Central Ontario District (Parole), Correctional Service of Canada, 330 Keele Street, Main Floor, Toronto, Ontario, M6P 2K7 Canada. E-mail: wilsonrj@csc-scc.gc.ca Dr Wilson is Director of the Sexual Behaviour Program, Central Ontario District, the Correctional Service of Canada. Judge Huculak is a judge on the Provincial Court of Saskatchewan in Saskatoon. Mr McWhinnie is Project Manager of The Restorative Justice Options to Parole Suspension Project in Victoria, B.C., operating from the Victoria Parole Office, the Correctional Service of Canada. The views presented in this article are those of the authors, and do not necessarily represent those of the Governments of Canada or Saskatchewan, or their constituent departments. We wish to thank Hugh Kirkegaard for his assistance with elements of this article. in the media and the community at large, they are not inclusive of the communities they purport to protect. In response, many have become disillusioned with the criminal justice system, and have contributed to a parallel movement in regard to restorative justice. Canada has a longstanding tradition of restorative justice innovation, with many of those initiatives originating in the faith community. In particular, the Mennonite community has a long tradition of using restorative techniques in addressing victim and offender needs. For instance, Yantzi (1998) documents the Victim Offender Reconciliation Program (VORP), a groundbreaking Mennonite initiative that began in Canada in the 1970s and quickly gained favor in the United States. The work of Andrews and Bonta (1998) shows very clearly that sanction alone will not curb crime. Rather, it is the provision of appropriate human service that underscores effective correctional intervention. This article presents three current Canadian restorative justice initiatives. (i) The Sentencing Circles program in use in Saskatchewan demonstrates how restorative principles may be implemented at the sentencing phase of the criminal justice process.(ii) The Restorative Justice Options to Parole Suspension program currently under way in Victoria, British Columbia, shows how such principles may be applied while an offender is on conditional release during the course of a sentence. (iii) The Circles of Support and Accountability (COSA) pilot project in the Greater Toronto Area has been initiated to assist high risk sexual offenders in community reintegra- tion following the completion of their sentences, when no formal assistance (i.e. correctional supervision) is typically available. SENTENCING CIRCLES The concept of Sentencing Circles (SCs) entered the Canadian lexicon in 1992, with the Yukon Territory and the Province of Saskatchewan becoming the focus of the circle process (Church Council on Justice and Corrections, 1996). This section will deal primarily with the SC process in Saskatchewan, where it has been used in both rural–aboriginal and urban settings, with a variety of offenders and offense types. Although it shares common ground with restorative justice theory, the SC process did not develop from an overarching theoretical framework. The practice came first. The Saskatchewan experience initially involved three Northern judges and several communities. The general process combined respect for due process, mediation and interest-based negotiation, and consensus-based decision-making with aboriginal healing and peacekeeping concepts. While the model draws much from certain aboriginal traditions, it is important to note that SCs are not inherently ‘aboriginal’. Aside from implementation criteria developed by the judges involved in the process, there was no initial attempt to delineate exactly how this process would occur in each community. With input from lawyers in the involved districts, the judges determined the procedure and criteria. This has caused difficulties, parti- cularly as the legal profession continues to ‘adapt the conflict to the process, rather than the process to the conflict’, the latter being more closely the agenda of SCs and the community (Hon. Barry Stuart, Yukon Territorial Court, personal 364 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) communication). The implementation of SCs has suffered from this apparent lack of formal structure, resulting in a degree of confusion as to how SCs should be conducted. Subsequently, this has led some to believe that they do not work. However, it appears that it is not SCs that are not working, rather it is the ad hoc or, often, haphazard way in which they are formed that leads to diminished results. For instance, each potential SC should be preceded by a rigorous evaluation of the circumstances in order to ascertain the appropriateness of using such a process. As Sentencing Circles proliferate in Canada, there has been some opposition. First, not all aboriginal bands have embraced SCs, most often because they do not see the process as being consistent with their particular traditional beliefs regarding justice (e.g., the Kaska and Tlingit peoples of the Yukon and Alaska). Second, the legal community has also experienced growing pains, particularly in regard to the Court of Appeal. To date, the appellate court has appreciated the concept but, perhaps, not the dynamics of the SC process (Hon. Barry Stuart, personal com- munication; R. v. Morin, Saskatchewan Court of Appeal, 1995). In a recent decision, the Alberta Court of Appeal (R. v. B. L., 2002) addressed issues surrounding the use of SCs. Specifically, the decision spoke to means of establishing fact when employing SCs, as well as dealing with the issue of appro- priateness of sentence. While the appeal was dismissed in R. v. B. L., the court found that the facts of the case were never established according to the normal process of law. Although the offender pleaded guilty to the charge of aggravated assault, the Crown was not afforded the opportunity to present the facts, nor was it afforded the opportunity to prove any aggravating circumstances. In summary, the facts should be established before a circle hearing, either by admission or by proof in the ordinary course. Those facts should be presented to the circle. The circle is not the appropriate forum in which to prove disputed facts. The sentencing circle deals with the possibility of a restorative sentence based, in part, on the facts constituting the offence. No doubt additional facts will frequently arise during the course of the hearing, but, if challenged, disputed facts must be proven as outlined above. . . . The conflicting comments the judge made in this case make it difficult to ascertain which facts he accepted. But for the agreed facts provided at this appeal, we would return this matter for rehearing (R. v. B. L., para. 44, emphasis added). The Court also addressed the issue of composition of this circle, specifically, in regard to absence of support for the victim and the failure to include meaningful representation from the community. Further, the victim’s wife was not present, even though she witnessed the attack on her husband and, due to the severity of her husband’s injuries (resulting in permanent disability), the Court noted that ‘she was yet another victim’ of the assault. Process Participation in the SC process is entirely voluntary for all concerned. In general, SCs have tended to be used in response to serious offences, particularly those where there is an identifiable principal victim (e.g., robbery, criminal negligence causing Innovations in Canada 365 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) death, and assault). Many of the offenders involved in such crimes would expect to receive a lengthy period of incarceration. Although avoidance of incarceration is not always the result, generally, a non-custodial sentence is recommended. The focus in taking the SC route is on accountability, rehabilitation, and community integration. Offender motivation to change is often evident in facing the victim and the people most important to him/her. Many offenders experience the SC process as significantly more personally challenging than facing a judge in the usual court sentencing procedure, in which there is minimal discomfort, a lawyer speaking for them, and less personal investment of time. The families of both the victim and offender, as well as the community, are often an important part of the rehabilitation aspect and a necessary part of the community integration. In the SC process, an effort is made to distinguish the criminal act from the offender. Although it is made clear that the behavior is not acceptable, the offender is given the opportunity to correct the criminal behavior. Because the SC process puts a face on the stakeholders, an offender’s failure to comply with the resultant terms and conditions of the sentence results in letting down family, friends, the community, and oneself. Also, the offender may face incarceration if the terms of the SC are breached. Consequently, there is considerable social support for the offender to comply. Circles often take several hours in gathering information, responding to it, and in developing a resultant sentencing plan. At the end of much discussion, a successful SC will result in a resolution satisfactory to all stakeholders, and recommendations are made to the judge. Judges rarely refuse to follow the recommendations, which may include elements of reparation and restitution to the victim, as well as a plan for rehabilitation and integration of the offender. Treatment is also an important component of the sentencing plan, as many offenders have significant clinical problems, including addictions. Social control through social support can often be more effective than incarcera- tion which, for many offenders, has generally failed to deter or rehabilitate them. By strengthening connections to family and community, obligations are created that seem to operate both as deterrent and as rehabilitation. The resources of the com- munity are frequently called upon, including individuals such as aboriginal elders. The victim is also supported through the SC process. They are able to express the damage done to them, to understand how and why this victimization occurred, and to hear the offender, and often his family, express genuine sorrow and regret for the criminal behavior and remorse for the harm caused. The power of publicly hearing the words ‘I’m sorry’ is sometimes underestimated. The SC process provides an opportunity for the offender to reinvent aspects of himself, to detach himself from the criminal appellation through various technologies (e.g., treatment), and to utilize the power of the community and its resources to assist in his rehabilitation. Although social control through help and support is an important aspect of the SC process, healing of the offender, victim, and community are equally important. Indeed, the opportunity to provide a measure of healing in regard to a specific criminal event is a sufficient justification for the circle process regardless of whether the offender reoffends. This critical importance of ‘healing’ has been officially recognized by the courts in Saskatchewan. A variety of SC initiatives now exist in Saskatchewan: (i) pre-charge and post-charge diversion of youth directly to a community/healing circle; (ii) 366 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) pre-sentence circles, where the result of the circle deliberations is presented to the judge; and (iii) an SC process which involves a judge, counsel for the offender, the prosecutor, and the community. Since 1992, hundreds of SCs have been held in Canada. The majority of SCs in Saskatchewan have been implemented in the aboriginal community, which is largely rural in nature; however, there are numerous examples of SCs undertaken in urban settings. In general, several core themes are reflected in the SC process: community, dialogue, reparation, rehabilitation, participation, respect, responsibility, empathy, consensus, healing, empowerment, transformation, and hope. Although the accused generally asks for the SC, the judge may also suggest one. Indeed, the first SC in Saskatchewan was initiated by a judge. Further, in R. v. B. L. (2002) the Alberta Court of Appeal suggested that laws surrounding the use of SCs should develop incrementally, stating, ‘in our view, the trial courts are the appropriate forum for this process to mature’. Sentencing Circles in Rural Communities In small, Northern Aboriginal communities where most inhabitants know each other, Sentencing Circles are usually held in a band hall or a church. The setting is informal, with participants (including the judge) in casual attire. The space is free from the ritualistic legal language and furniture of the regular courtroom. Partici- pants usually include the offender, victim, their families, friends, and support systems, others from the community (e.g., alcohol/drug treatment specialists), as well as the judge, prosecutor, and defense. Basically, the process is sufficiently open to include anyone with an interest in participation. The judge acts as a facilitator, in a non-directive way. In some communities, someone other than the judge may facilitate or a co-facilitation model is employed. In aboriginal communities, elders are an integral part of the process. Once all concerned parties have gathered, participants arrange themselves in a circle of chairs. The inner circle includes those who wish to participate and speak, while an outer circle includes those who are interested as observers and who are not as directly connected or do not express an interest in speaking. In aboriginal communities, the process opens with a prayer, often in the local native language as well as in English. This sets the tone for the important nature of what is to occur. Introductions are made and the connections between the parties are explained. The circumstances of the offense are read out by the Crown Attorney, and the offender acknowledges guilt and takes direct responsibility for his/her actions. The offender, his lawyer, and the victim are then given the opportunity to speak. Frequently, the offender opens with an apology, which the victim may accept, provided that genuine remorse is expressed. Each person in the room is given many opportunities to speak while the other participants listen respectfully. The process is transparent, open, flexible, and unscripted in the sense of limiting the nature of the dialogue or limiting who can participate. There is no pressure to speak, and interruptions are not tolerated. There is attention to context, relationships, and dialogue. Sentencing Circles often include a great deal of emotion, pain, and tears; however, participants also express positive emotions. The SC closes with a prayer, shaking of hands and, often, hugs. Innovations in Canada 367 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) Sentencing Circles in Urban Settings Sentencing Circles in urban centers resemble those undertaken in smaller commu- nities, except that the concept of ‘community’ becomes a relational one not based on a particular geographic area. Instead, the participants are involved by the nature of their connection to the offender, victim, or the circumstances of the offense. For example, in a dangerous driving case where, fortunately, no one was injured, participants included Mothers Against Drunk Driving, an emergency room doctor who routinely treated accident victims, a former accident victim who was perma- nently paralyzed, an addictions counselor, and a representative from an insurance company. In urban SCs, the process is one of constructing ‘community’ whereby the process creates a ‘community of purpose’ for a particular individual and a particular circumstance. Although this larger community usually disperses after the circle is complete, important connections are made, including connections to family mem- bers and other members of the community at large. Offers to help are often made by community members. For many individuals, the contact continues or becomes part of a probation order. Urban SCs tend to be large, with the inner circle of participants comprised of as many as 20–30 people. The urban SC process can take much longer than its rural partner, mostly because the first couple of hours involve dialogue intended to construct ‘commu- nity’ where it may not have previously existed. This construction process is necessary in order to build a foundation on which participants can move to the next phase of determining an appropriate response to the crime in question. Indeed, in a ‘constructed’ circle in Saskatoon, one particular young offender made many intimate and distressing disclosures. This youthful male offender disclosed that he had been sexually victimized by a relative, and that he had had thoughts of suicide while imprisoned for his last offense. These disclosures came out in the context of an environment where the participants had no pre-existing relationships. One can speculate that a foundation of trust must have been established in order to permit these disclosures. Certainly, the sharing of such information took immense courage on the part of the offender, and this courage was not lost on the other participants. Conclusions There are many examples of communities, aboriginal and otherwise, adopting the SC approach to justice. In each instance the community was engaged in a direct way, such that dialogue, participatory democratic decision-making, power-sharing, and sharing of responsibility occurred. The uniqueness of the SC process is an ever- present reminder of a different way of responding to crime, particularly, in a public atmosphere bent on ‘getting tough on crime’. The SC process directly challenges the foundational underpinnings of the criminal law construct of the autonomous, rational ‘chooser’ in favor of the historically situated subject as a socially con- structed, relationally connected person. Sentencing Circles contest the reason– emotion dualism, as well as the concept of impartiality, by standing in opposition to a rule-based, ‘objective’, hierarchical system. The Canadian Supreme Court recognized recently recognized this process (R. v. Gladue, 1999) in expressing 368 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) concern for the appalling number of incarcerated aboriginal offenders. The court pointed to sentencing and healing circles as a way of diverting aboriginal offenders from jail. This decision was particularly important in recognizing that poverty, substance abuse, lack of employment, and poor educational opportunities are critical factors in the crisis facing many aboriginal people. Although the court was interpreting amendments to the Criminal Code of Canada in 1996, which directed judges to consider the circumstances of the aboriginal offender in sentencing, the court’s comments can be applied generally to all offenders. That is, sentencing should take into account the structural antecedents of crime. This is consistent with the circle approach, which has been open to both aboriginal and non-aboriginal offenders. In some respects, it may seem impossible that a relatively simple process of bringing people together, sitting in a circle talking, listening, and feeling, could have such a profound impact. Nonetheless, the powerful collective energy produced through open expression of emotion, storytelling, a willingness to connect with others despite differences, and the ability to empathize, coupled with a willingness to rise to a higher level in asking for and receiving forgiveness, all have the potential to occur in the SC process. A Sentencing Circle is a space where the impossibility of separating cognition and emotion exists, where reason meets feeling. What makes this process so powerful is that we would not normally expect all these events to occur in a situation where the principal reason for meeting is a breach of the law and the infliction of harm on another. The Sentencing Circle process is a way of judging that is situated, embodied, and interested as opposed to judging that is unsituated, disembodied, and disinterested. Such contextual reasoning includes perspectives stretched by dialogue and em- pathy. Empathy is a key component in the process, and may be described as attention to the particularized other. It allows for the ability to imaginatively step inside the beliefs, feelings, motives, and thoughts of others in order to reach shared understandings. Empathy for the victim by the offender opens a space for remorse and permits understanding of the offender’s conduct by the victim. Empathy opens the space for reconciliation and forgiveness, and forms the foundation of social change and transformation. It provides a means for bridging differences and enhancing reasoning. Empathy inspires action and a willingness to help others; active empathy calls on all participants to contribute to constructive consensus- based solutions and mutual responsibility. The SC process challenges us to rethink the concept of blame and responsibility. The structural underpinnings of crime challenge us to see crime from differing perspectives in the context of an attitude of caring — the willingness to respond to needs. It is impossible to participate fully in the circle process and not be affected. Sentencing Circles are an approach to crime that addresses victim, offender, and community needs more fully than the traditional court process. The challenge is to link the personal transformative potential to the structural transformation. While the above generally paints SCs in a positive light, fairness requires that we recognize that not all views of this process are favorable. One particular criticism, raised in 1996 by Roberts and LaPrairie, but still largely unanswered, is the need for empirical evaluation of the SC process. Certainly, many of the variables are measurable and, thus, applicable to scientific inquiry. Unfortunately, much of the process has remained ad hoc, perhaps as a consequence of a continued failure to fully Innovations in Canada 369 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) operationalize SCs, or to generate full and consistent support of the model, even within a single jurisdiction. Others have cited the need for extreme caution in the use of SCs, particularly where the power differential between victims and offenders is great. For example, caution is warranted in cases of sexual assault, spousal assault, and in crimes where the victims are children and the offenders are adult. Cayley (1998) cites examples where the assumptions of the community should be closely examined before considering whether or not to employ a SC. Citing Ottawa lawyer Mary Crnkovich and her testimony before the House Justice Committee on behalf of the Inuit Women’s Association (Pauktutit), Cayley noted that an assumption of community homogeneity can drown out the identity and voice of victims in a collective. Instead of a true hearing of the wounds endured, the emotions and the fears of victims who seek justice, and healing through offender accountability and responsibility, victims can be ‘persuaded to comply with the consensual ‘community’ interest rather than . . . ’ their own (Cayley, 1998, p. 207). Regarding R. v. Gladue (1999), cited above, in which the Supreme Court of Canada acknowledged the unsuitability of punitive sentences in the adjudication of cases involving aboriginals, caution in the interpretation of that ruling for the purposes of employing SC in the determination of sentences must be exercised. In R. v. B. L. (2002), for instance, the Alberta Court of Appeal opined that in handing down a suspended sentence, the trial judge ‘erred in his understanding of the breadth of R. v. Gladue (1999). . . . A suspended sentence in the circumstances of this case fails to properly take into account deterrence and denunciation’. The court went on to say While no category of offence is automatically excluded from consideration for a restorative sentence, a sentencing judge must weigh the offender and the offence, properly understanding the role that denunciation and deterrence play in sentencing (R. v. B. L., para. 69). These comments are echoed by Cayley (1998), who notes the words of Yukon Territorial Judge Barry Stuart: ‘Charter rights, statutory provisions, and all the fundamental principles of justice must be retained’ in the SC process (p. 189). Finally, the Alberta Court of Appeal, while not seeking to place strictures on the process involved in mounting a SC, nevertheless offered the following as commen- tary on the case of R. v. B. L.: As noted, the most serious problems with sentencing circles arose with respect to the facts, the lack of input by the parties, the composition of the circle, specifically the wife’s absence in view of [her husband, the victim’s] disability and the absence of objective community members. All these issues worked to undermine the circle’s conclusions (para. 59). RESTORATIVE JUSTICE OPTIONS TO PAROLE SUSPENSION In late 1999, staff working in the Victoria Parole Office of the Correctional Service of Canada (CSC) on Vancouver Island observed how the principles of restorative justice (RJ) were being innovatively applied at the front end of the Canadian justice 370 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) system. Sentencing Circles (see above), along with Community Justice Forums supported by the Royal Canadian Mounted Police (RCMP), stood out as examples of how criminal behavior was being effectively addressed restoratively before offenders became entrenched in the traditional justice system. For Victoria Parole, the question was: Could the same restorative principles that guide ‘front-end’ innovations be applied at the ‘back end’ of the system? The resultant attempt at answering this question has since come to be known as the Restorative Justice Options to Parole Suspension project (RJ Options) (see also Correctional Service of Canada, 1999–2000; McWhinnie & Brown, 1999). Parole Suspension Process In Canada, the former British North America Act established federal jurisdiction over prisoners sentenced to two or more years. Under the rules of the Criminal Code of Canada and the Corrections and Conditional Release Act, some prisoners become eligible to serve portions of their sentence under supervision in the community. The decision to release a prisoner under such supervision is made by the National Parole Board (NPB), which also stipulates conditions of release. However, supervision of these conditions is the responsibility of CSC, a body separate from the NPB. As a major part of their job, community parole officers (POs) supervise offenders in the community, and must continually gauge the level of risk the offender poses to community safety. If the PO feels that risk is escalating (e.g., a parolee breaches a condition to abstain from alcohol), he or she can suspend the offender’s release, resulting in temporary detention in custody pending review. Within 30 days, the PO may return the offender to the community, provided that steps have been taken to ameliorate the risk to community safety (e.g., a commit- ment not to violate the abstain condition, and enrolment in a substance abuse relapse prevention program). This basic example of the ‘parole suspension’ process demonstrates where the RJ Options project might be used — when risk begins to escalate but has not reached the point where community safety is jeopardized. The primary objective of the RJ Options project was to initiate a process using commu- nity group conferences for offenders facing possible parole suspension while maintaining and enhancing community safety. Project Implementation Paul McCold (1998) noted that the best applications of restorative justice are found in practices that employ models such as circles, mediation, and family group conferencing. Community group conferencing is a model developed in Wagga Wagga, New South Wales, Australia, as a way of dealing with youth crime. It involves a tightly scripted method for bringing together the offender and his/her victim(s) in a supportive circle comprised of community members and justice officials. The offender is asked to address the victim(s), take full responsibility for his/her criminal acts, experience remorse and shame, and offer reparations for the damage done, either symbolically, actually, or both. In this example, the offender and victim engage in a process of recognizing that more than just a law was Innovations in Canada 371 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) violated — that real damage happened to real people. In this process, the com- munity is engaged in helping to achieve tangible justice for victims, while assisting the offender in accepting responsibility for his/her part in the violation of a fellow citizen. Community group conferences address Howard Zehr’s (1995) twin principles of harm and engagement by bringing to bear all of an offender’s criminal behavior and the effects of victimization resulting from that behavior (harm). In Canada, the RCMP has facilitated the creation of ‘Community Justice Forums’, which utilize an adaptation of the concept of community group conferencing in addressing some forms of youth crime. This initiative was undertaken with extensive consultation with those involved in the Wagga Wagga project. In February 1999, staff and volunteers of the Victoria Parole were trained as conference facilitators by the Westshore RCMP detachment. Victoria Parole adopted the community group conferencing model and, with its own trained and accredited facilitators, held its first RJ Options session in August 1999. Utilization of a Best Practices Approach The RJ Options project directly taps the two most powerful factors known to influence the risk of criminal offending: antisocial attitudes and antisocial peer affiliations (Andrews & Bonta, 1998; Motiuk & Serin, 1998). The RJ Options project serves to pull together and strengthen prosocial influences present in an offender’s community while targeting antisocial values and beliefs. When an offender is confronted by police, prison officials, treatment staff, or his parole officer, he/she may or may not be able to assess a personal need to make positive changes. However, the power of being confronted by one’s brother, sister, child, spouse, employer, pastor, or even a best friend is quite another matter. Very often, these individuals can tell him what his involvement in crime has meant to them in ways no justice official could. The collective power of the (traditional) justice system, when combined with that added by family and community members, is exponential. Hence, the ability to influence criminal thinking styles (e.g., the belief that no one other than the direct victim is influenced by an offender’s criminal behavior) is also increased. Since offenders referred to the RJ Options project have already been tried and sentenced, the challenge is no longer one of matching a suitable punishment with a particular crime. Rather, the challenge becomes one of repairing harms done. This is an opportunity for the offender to join a particular community and to eventually live at peace. The RJ Options project gathers members of that community around an offender in an effort to resolve problems that, unattended, might escalate to the point where the offender must be removed from the commu- nity. The RJ Options project provides healing and restoration while enhancing public safety. In our experience, community members are more than eager to become engaged in such a process of justice-making. When they do, community members become personally invested in their own safety. Justice is no longer a commodity delivered by dispassionate professionals, but becomes a locally crafted action plan made in collaboration with those same professionals. This plan of action then becomes intrinsically linked with community safety. 372 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) The RJ Options initiative has been presented to a number of organizations (e.g., faith groups, Rotarians, Legionnaires, Universities, and local community groups). In many of these venues, the response has been, ‘It’s about time!’. This sentiment reflects the community’s interest in assuming some of the responsibility for helping to shape what happens with offenders in their midst. This is an appealing concept to many community-minded people. Essentially, it affords communities the opportu- nity to ‘own the conflict’ (Christie, 1977), and, in so doing, to rediscover something of what it means to be a community. Offenders asked to become part of the process (participation is voluntary for everyone) do not find the experience an easy one. For them, two responses are common. (i) ‘This is huge.’ By this, they are acknowledging that this undertaking may be more involved and emotionally taxing than even the court and sentencing process they have already experienced. (ii) ‘I had no idea there were so many people involved in my life.’ This realization is a positive one which usually comes during or after a conference. While some critics are tempted to see the RJ Options project as being ‘soft’ on criminals, it is a difficult, draining, yet ultimately rewarding experience for everyone involved. For many offenders, there is nothing more foreign than sitting in a group comprised of one’s family, friends, community members, victims’ representatives, and correctional workers, and saying what has happened to bring them to this point. Family members, friends, and community all hear the same story, which is simultaneously heard by the parole officer (who may add or clarify information as required). Family members add context and depth to the story. The ‘Unsung Victim’ In implementing the philosophies of restorative justice, there is much concern about victims. Some have said that unless an intervention trying to qualify itself as a restorative practice includes the actual victims of an actual crime with the actual perpetrator, it is only ‘wannabe’ restorative justice. Initially, the RJ Options project was criticized as a such. However, there is due consideration of victim roles and issues in family and community group conferences. This is evident in the story of a mother betrayed by her son’s sexual offences, or that of a grandparent bullied and robbed of her television set for drug money by her cocaine-addicted grandson, or even a twelve-year-old who is bullied on the school ground because his older brother is a ‘con’. Similarly, to hear a woman’s story of being beaten by her lover’s street friends when he was in jail, or that of a young mother and wife fending off drug-debt collectors at her apartment door after her husband was returned to jail, drives home recognition of the integral part that victim perspectives take in this process. Family members of offenders are victims of crime in every sense of the word. However, they are often overlooked in discussions of RJ. Yet, when an offender returns to his home community, these are the people he or she will be with the most. Family members are often frightened, ashamed, angry, and shunned by their own community and other family members. They are also victims of the offender’s crime(s), but often feel obligated by blood ties to bring their victimizer back into their lives, something ‘official’ victims would never be asked to do. Family members Innovations in Canada 373 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) have the same questions, fears and, often, similar trauma issues as do the ‘official’ victims. However, they frequently have nobody to turn to for help. In some cases, families represent more of an antisocial than a prosocial influence on their offending member. This is frequently true with alcoholic families. In one case, the father of an offender insisted that his son meet him at the local tavern, even though the offender was forbidden to imbibe as a condition of his parole. As part of the conference, two things happened. (i) The father learned the true nature of his son’s criminal behavior, including the fact that trouble was more likely to occur while he was drunk. (ii) The father was confronted by his son about his own drinking patterns. At the very least, the pressure to drink with his father (a violation of his condition) was removed. At best, the alcoholism in the family was exposed, leaving room for repair and healing. Does It Work? To date, 22 referrals have been received by the RJ Options project. Of these, seven were not conferenced — four individuals declined participation, the other three had their parole automatically revoked by the NPB before a conference could be held (owing to further criminal charges, which later became convictions). A total of 119 persons participated in the resultant conferences, not including the offenders, with an average of eight community participants per conference. In the context of this project, recidivism is defined as conviction for another criminal offence occurring after referral to the project. The recidivism rate for those not con- ferenced was 43%. On average, those offenders not conferenced spent 53 days in the community. Of the 15 offenders who were conferenced, the recidivism rate is 27%, with the average time spent in the community being 119 days, a difference of just over two months (66 days). The cost of maintaining an offender in the community on supervision by CSC may be estimated as approximately $14 000 per annum, while the cost to maintain that same offender in detention is approximately $40 000. The obvious monetary gains of safely maintaining offen- ders in the community through the RJ Options project should be seen even more positively against the backdrop of increased community engagement with the criminal justice system. While the results above are very promising, they do not represent a formal evaluation. An evaluation is currently under way through the Department of the Solicitor General Canada, Corrections Research Division, with results expected in early 2003. Nonetheless, the current results move thoroughly in the right direction and are certainly in line with what is known to work in correctional practice today (see Andrews & Bonta, 1998). The RJ Options project is also principled, in that it adheres to the best practices of both RJ (harm and engagement) and sound correctional practice (risk, need, and responsivity). The current outcome requires thorough evaluation and scrutiny to determine the project’s true worth. These initial results, however, are encouraging and seem to demonstrate that the principles of restorative justice may indeed be applied at ‘the end of the system’ before the expiration of a sentence, as well as the ‘front end’ of the system before a sentence begins. 374 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) CIRCLES OF SUPPORT AND ACCOUNTABILITY Persons who commit sexual crimes are among the least understood and most publicly feared of all offenders. Public sentiments and legislative attempts in Canada and other countries clearly reflect a general intolerance of sexual offenders. Many individuals espouse a ‘lock them up and throw away the key’ philosophy, with the tabloid media frequently ‘fanning the flames’. A recent sex offender release in Southern Ontario was met with headlines reading ‘A Community Lives in Fear’, ‘Perv’s Neighbours in Scary, Ugly Mood’, and ‘Neighbours: We’ve Done Our Jobs’. The latter headline trumpets the neighborhood’s success in driving the offender from the community. The big question is, however, to where? It is a simple truth that most sexual offenders receive determinate sentences of incarceration. In Canada, higher-risk offenders are frequently detained until the very end of their sentences, resulting in many being returned to the community with little or no community support. Legislation has been enacted in both the provincial and federal domains to aid in long-term risk management, as well as to assuage public fears regarding the return of sexual offenders to community living. However, police services are still adjusting to the increasing demand for public accountability. As such, the police have sought to involve their community partners in managing difficult cases. This section will outline a model for community-based risk management, with a particular focus on the practical aspects of providing support to (and ensuring continued accountability in) high-risk sexual offenders within a professionally supported, volunteer framework. Issues of recruitment, liability, ethics, and the overall efficacy of the project as a means to enhance risk management will also be discussed. Circles of Support and Accountability (COSA), as a means to address offender needs and public concerns, began in 1994 with a group of people from a local church who gathered around a pedophile in a mid-sized Canadian city. These volunteers assisted him in finding a place to live, helped him get settled in the community, and dealt with police, media, and angry community activists. On a daily basis, members of this group visited with the released offender, both supporting him and holding him accountable for his attitudes and actions in the community. Their creative response to this convicted pedophile returning to the community became the template for a similar response, a few months later, in another community. In both of these initiatives, the response balanced the fear of the community on the one hand and the needs of the released offender on the other. The result was the evolution of a model that has come to be known as Circles of Support and Accountability. Motivated by a desire to take the safety concerns of the community seriously, community volunteers create a supportive and responsive community around an offender (known as the core member). The primary concern of all is that there be no more victims. The COSA model recognizes the humanity of both the offender and the victims of their offenses, and challenges the community to work with the offender in ways that promote healing and responsible living. As this work evolved in a more intentional manner, the Mennonite Central Committee (MCC), with its history of pioneering restorative justice initiatives, agreed to sponsor a COSA pilot project focused on the re-integration of warrant expiry sex offenders. Innovations in Canada 375 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) While no formal mandate existed for government involvement, funding and support were provided by the Correctional Service of Canada, based on moral responsibility. The model that emerged from the initial experiences outlined above was a community-based approach, volunteer driven and professionally supported, that gathered four to six volunteers in a circle around an offender as he returned to the community (see Figure 1). As such, the model represents a shift away from the traditional reliance on professional expertise to resolve the problems and issues of sexual offenders and their behavior in the community. In place of this professionally driven model for problem solving, COSA offers a collaborative approach which makes appropriate use of professional resources in ways which re-empower com- munities to respond to these challenges creatively and responsibly, drawing upon the latent skills of community volunteers. Hence, police and other professionals can join circle volunteers by sitting in on circles, on either a consistent or an as-needed basis. The work of the circle happens in daily contacts between individual circle volunteers and the core member, in coffee shops and the wider community, and in weekly meetings to address issues. Everything from the practical concerns of finding appropriate housing to observations that the core member may be moving into his offense cycle is discussed in the circle. The goal of the circle is not to be therapeutic, but to provide the support and accountability of well trained, profession- ally supported community volunteers. To date, the majority of COSA volunteers have come from churches involved in work with offenders, refugees, the devel- opmentally delayed, and other groups traditionally marginalized in society. Volun- teers are trained in a number of areas, including group dynamics, patterns of sexual offending, related legal issues, and restorative justice principles (see below). These volunteers commit to working with the core member of the COSA and, in turn, the core member commits to working with them. While involvement in a COSA by the core member is voluntary, these commitments are formalized in a shared under- standing of expectations. The COSA interacts regularly with professionals involved with the core member, including police representatives, psychologists, and physicians in ways that both enhance the ability of the volunteers to support the core member and hold him accountable, as well as to strengthen professional understandings of the core member. Where necessary, the COSA also advocates on behalf of the core member with these professionals and others (such as landlords). It confronts him about attitudes and behaviors that could lead to his re-offending. It mediates in situations of conflict with the community and others, including family members and even past Stability in the community Known in the community (references checked) Maturity Healthy boundaries Availability and commitment Balance in . . . &Lifestyle &Viewpoint Figure 1. Volunteer recruitment. 376 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) victims. The COSA walks with the core member through problems and crisis situations and celebrates with him the various anniversaries and milestones in his reintegration to society. In short, the COSA is an attempt to re-create community in practical and realistic ways, around one who, by his own actions, has fallen out of community. Training and Support While Circles of Support and Accountability draw strength from their existence as a community-driven initiative, we acknowledge that volunteers do require training and support. As support for this type of approach has grown, several other Canadian communities have developed COSA initiatives. In response, training materials and resources have been developed nationally as a result of several national gatherings. It was subsequently decided that a National Standards Training and Resource manual should be prepared. To that end, an orientation and training regime is nearing completion. Portions of this training package have been piloted at several locations across Canada. This process assists volunteers to become more comfortable in working with offenders, as well as to educate them regarding sexual offending, risk factors, and means by which they can assist in risk reduction in their communities. The training outline shown in Figure 2 has been developed; a similar model is used in the original pilot project in South-Central Ontario. Preliminary Outcome Data: Recidivism While the Circles of Support and Accountability concept appeared to make sense from a philosophical perspective, it has been necessary to formally evaluate the effect Topics of Training Overview of the Criminal Justice System Restorative Justice Needs of Survivors The Circle Model Effects of Institutionalization Human Sexuality and Sexual Deviance Risk Assessment Boundaries and Borders Conflict Resolution Group Dynamics Building Group Cohesion Circle Functions Crisis Response and Preparing for Critical Incident Stress Working with Correctional Officials, Police, News Media and other Community Professionals Needs Assessment Building a Covenant Court Orders Closing a Circle Figure 2. Volunteer training (4.5 days or 36 hours total). Innovations in Canada 377 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) the initiative has had on its constituent groups, particularly ex-offenders and the community at large. At present, evaluation has been completed in regard to the reoffense rate of core members in the pilot project (Wilson & Prinzo, in press). However, it will be necessary to augment these data with more process-oriented data, that is, information that speaks to the individual involvement of stakeholders and the resultant changes that may or may not have occurred in the community as a whole. This latter component of the evaluation of the pilot project is as yet unfinished, and will be reported at a later date. To date, 30 circles have been undertaken in Southern Ontario (i.e., the pilot project administered by MCC). Mean time at risk was 36 months (also mean length of involvement in the program), with the range being from nine months to almost seven years. Average STATIC-99 (Hanson & Thornton, 1999, 2000) scores confirm the high-risk status of offenders involved in the program. In comparing the expected recidivism rate with the observed recidivism rate, using STATIC-99 survival data, we have found that recidivism has been reduced by more than 50% (eight expected versus three observed instances of sexual recidivism). Further, from a harm reduction perspective (see Laws, 1996; Marlatt, 1998), each incident of sexual recidivism was categorically less invasive and less severe than the offense for which the offender had most recently been incarcerated. It should be noted, however, that the actuarial projections provided by the STATIC-99 standardization sample appear to overestimate risk in this group. In a soon-to-be-released evaluation of the COSA pilot project, offenders receiving assistance via a circle reoffended at a lower rate incrementally in comparison with a matched control sample (Wilson, Picheca, & Serin, 2001). CONCLUSIONS We are firmly convinced that prospects for offender success in the community are significantly increased when the offender has a supportive network and a real accountability framework, especially when both are derived from community involvement sensitive to both victim and offender needs. While this is certainly intuitive, it is also supported by the data reported here, as well as by studies demonstrating the benefits of collaborative, community-based approaches (Wilson, Stewart, Stirpe, Barrett, & Cripps, 2000). We hope that further investigative reviews, currently under way, will lend additional credence to this proposition. From the interim data presented above, it would appear that these initiatives have assisted the management of offender risk in the community while maintaining offender accountability, through attention to the principles of risk, need, and responsivity (Andrews & Bonta, 1998). However, data regarding the process of offering and being involved in such endeavors are required to complete the evaluation picture. This component of evaluation will need to solicit process information from offenders and their families/friends, community volunteers and their families/friends, professional and other adjunct persons, and members of the community at large. The intent must be to qualitatively evaluate the effect of involvement in such projects on the various subgroups of stakeholders, while also soliciting information from community persons not specifically involved in the project (as a means to gauge the education and public information component). 378 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) Presently, support for restorative justice initiatives has come head to head with support for the ‘get tough on crime’ agenda. Both of these perspectives have increased community safety at heart but, clearly, the means by which that increased safety is to be achieved is radically different. The research literature (e.g., Andrews & Bonta, 1998) suggests that initiatives which increase community involvement in holding offenders accountable is more likely to result in decreased recidivism. In those communities where restorative techniques have been employed, stakeholders report feeling more knowledgeable about the system and, generally, feel more empowered in dealing with risk. However, most examples result from relatively tight-knit circumstances where, perhaps, ‘community’ means something different than it does in larger, urban settings with a greater degree of anomie. Notwithstanding the important gains we have made with our, admittedly select, groups of offenders, the real key to long-term success seems to be in promoting community development. That is, communities require assistance in getting to the point where they will tolerate initiatives such as those described in this article. We believe that tolerance will increase as links grow between various community members. The examples presented here have garnered support from community agencies and leaders in relatively tight-knit communities, but the blessing of the average citizen has been given with many caveats and misgivings, if given at all. Larger communities appear to be at a distinct disadvantage when it comes to hands-on, community-based risk management. Clearly, there is still much to be accomplished in appropriately assisting our communities to address the realities of risk and risk management inherent in returning offenders to our midst. REFERENCES Andrews, D. A., & Bonta, J. (1998). The psychology of criminal conduct (2nd ed.). Cincinnati, OH: Anderson. Cayley, D. (1998). The expanding prison: The crisis in crime and punishment and the search for alternatives. Toronto: Anansi. Christie, N. (1977). Conflicts as property. British Journal of Criminology, 17, 1–4. Church Council on Justice and Corrections. (1996). Satisfying justice: Safe community options that attempt to repair harm from crime and reduce the use or length of imprisonment. Ottawa, ON: Author. Correctional Service of Canada. (1999–2000). Alternatives to revocation: Safely sustaining offenders in the community. Strategic Planning and Policy Branch (available from the authors). Hanson, R. K., & Thornton, D. (1999). Static-99: Improving actuarial risk assessments for sex offenders. (User Report 1999–02). Ottawa, ON: Department of the Solicitor General of Canada. www.sgc.gc.ca. Hanson, R. K., & Thornton, D. (2000). Improving risk assessments for sex offenders: A comparison of three actuarial scales. Law and Human Behavior, 24, 119–136. Laws, D. R. (1996). Relapse prevention or harm reduction? Sexual Abuse, 8, 243–247. Marlatt, G. A. (1998). Harm reduction: Pragmatic strategies for managing high-risk behaviors. New York: Guilford. McCold, P. (1998). Restorative Justice Handbook. Corrections Compendium, 23, 20–28. McWhinnie, A. J., & Brown R. E. (1999). Restorative justice alternatives to parole suspension. Journal of the American Probation and Parole Association, 23, 44–46. Motiuk, L., & Serin, R. (1998). Situating risk assessment in the reintegration potential framework. Forum on Corrections Research, 10, 19–22 (also available at www.csc-scc.gc.ca). R. v. B. L. (2002) A. R. 44. R. v. Gladue (1999) S.C.R. 688. R. v. Morin (1995) Sask. R. 120. Roberts, J. V., & LaPrairie, C. (1996). Sentencing circles: some unanswered questions. Criminal Law Quarterly, 39, 69–83. Innovations in Canada 379 Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002) Wilson, R. J., Picheca, J. E., & Serin, R. C. (2001, November). Circles of Support & Accountability: Evaluating the efficacy of professionally-supported volunteerism in the long-term reintegration of high-risk sexual offenders. Paper presented at the 20th Annual Conference of the Association for the Treatment of Sexual Abusers, San Antonio, TX. Wilson, R. J., & Prinzo, M. (In press). Circles of support: a restorative justice initiative. Journal of Psychology & Human Sexuality. Wilson, R. J., Stewart, L., Stirpe, T., Barrett, M., & Cripps, J. E. (2000). Community-based sex offender management: Combining parole supervision and treatment to reduce recidivism. Canadian Journal of Criminology, 42, 177–188. Yantzi, M. (1998). Sexual offending and restoration. Waterloo, ON: Herald. Zehr, H. (1995). Changing lenses: A new focus for crime and justice. Waterloo, ON: Herald. 380 R. J. Wilson et al. Copyright # 2002 John Wiley & Sons, Ltd. Behav. Sci. Law 20: 363–380 (2002)