JPHS 13(3-4) JOURNAL ONLINE.VP Circles of Support: A Restorative Justice Initiative Robin J. Wilson, PhD, CPsych Michelle Prinzo, MA SUMMARY. This paper addresses the increasing difficulties faced in community-based management of sexual offenders in Canada. Those of- fenders at particularly high-risk to re-offend (e.g., sadistic rapists and se- rial child molesters) often receive indeterminate sentences, and are rarely released to the community prior to death or incapacitating illness. However, many other high-risk offenders are released from custody at the end of a determinate sentence, often without the benefit of adequate supervision or treatment. In a restorative justice initiative managed by the Mennonite Central Committee of Ontario, 30 high-risk sexual of- fenders released at sentence completion were provided with community support in the form of Circles of Support and Accountability. A brief overview of the Canadian penal system and its handling of sexual of- fenders is given to provide the social and political framework in which many current restorative justice projects have been undertaken. It is ar- Robin J. Wilson is affiliated with the Correctional Service of Canada, Central Ontario District (Parole), 330 Keele Street, Main Floor, Toronto, Ontario, M6P 2K7 Canada. Michelle Prinzo is affiliated with the Ontario Correctional Institute, Brampton, Ontario, Canada. The authors would like to thank Andrew J. R. Harris for his comments on an earlier version of the manuscript. The views expressed herein do not necessarily represent those of the Correctional Service of Canada, the Ministry of the Solicitor General and Correctional Services, or the Governments of Ontario and Canada. [Haworth co-indexing entry note]: “Circles of Support: A Restorative Justice Initiative.” Wilson, Robin J., and Michelle Prinzo. Co-published simultaneously in Journal of Psychology & Human Sexuality (The Haworth Press, Inc.) Vol. 13, No. 3/4, 2001, pp. 59-77; and: Sex Offender Treatment: Accomplishments, Chal- lenges, and Future Directions (ed: Michael H. Miner, and Eli Coleman) The Haworth Press, Inc., 2001, pp. 59-77. Single or multiple copies of this article are available for a fee from The Haworth Document Delivery Service [1-800-HAWORTH, 9:00 a.m. - 5:00 p.m. (EST). E-mail address: getinfo@haworthpressinc.com].  2001 by The Haworth Press, Inc. All rights reserved. 59 ***DO NOT RUN ALT-A AGAIN *** gued that traditional punitive measures have done little to address risk to the community and that effective interventions in the community must not be limited to time under warrant. The Circles of Support initiative fo- cuses on the need to engage the community in the offender reintegration process. Data are provided regarding recidivism rates in comparison to actuarial projections determined from STATIC-99 (Hanson & Thornton, 1999) survival curves. Recidivism across 30 high-risk offenders, with a mean follow-up time of 36 months, currently stands at less than 40% of that predicted by STATIC-99. [Article copies available for a fee from The Haworth Document Delivery Service: 1-800-HAWORTH. E-mail address: Website: © 2001 by The Haworth Press, Inc. All rights reserved.] KEYWORDS. Sexual offenders, restorative justice, community man- agement, supervision INTRODUCTION The punishment and rehabilitation of offenders has been the subject of much recent debate in popular and scholarly circles in many nations of the world, particularly as the discussion pertains to sex offenders. In- ternational symposia, such as the 6th International Conference on the Treatment of Sex Offenders (during which the following paper was pre- sented), have been convened to stimulate dialogue regarding the safe and ethical risk management of sexual offenders. While the goal of all parties concerned is the prevention of further victimization, the means by which individual nations or jurisdictions strive to achieve that goal have been varied. Although much of the discussion regarding offender risk manage- ment has focused on traditional methods of correctional sanction, reha- bilitative programming, and psychological or psychiatric treatment, there has been a recent trend towards investigation of alternative meth- ods of furthering reintegration of offenders released to the community. Collectively, a group of these alternative methods have come under the heading “restorative justice,” although some might challenge the actual linkages between various initiatives. Overall, the restorative justice movement is one that seeks to facilitate the restoration of offenders to communities by combining education, support, and accountability, for both offenders and members of the community-at-large. The following 60 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions http://www.HaworthPress.com paper discusses a Canadian restorative justice initiative intended to fa- cilitate community reintegration of high-risk sexual offenders. How- ever, we would like to start with a brief history of the management of offenders in Canada, as a means to set the stage for our presentation of the initiative proper. Early attempts to maintain law and order in Canada may be seen as being directly descendent from practices originating in England and France, the two nations which primarily colonized what later became Canada. Prior to the 18th Century, the response to criminal offenders in Canada focused largely on corporal punishment. During the 18th Cen- tury, there was a transition from corporal punishment to imprisonment as the sanction of choice. Imprisonment in workhouses was established as the primary mode of punishment in many provinces. In these work- houses, prisoners were employed to perform labour jobs within the in- stitution. Persons awaiting trial or those awaiting sentencing were held in gaols. Over the period of 1790-1830, major shifts occurred in the perception of crime and the roots of criminal behavior. Crime came to be viewed as a consequence of community disorder and family instability (Griffiths & Cunningham, 2000). The Canadian Penitentiary Act of 1834 emphasized the use of penitentiaries as a general deterrent and as a mechanism for re- forming criminals through hard labour. Reflecting the origin of the word “penitentiary,” religion was a focal point and moral reeducation of in- mates was strongly encouraged. Hard labour was also used. In 1835, Kingston Penitentiary (Canada’s first) opened its doors in Ontario. But by 1840, concerns arose regarding the effectiveness of punishing and reforming offenders. Recidivism rates were high and there was extensive use of corporal punishment inside the facility. The Penitentiary Act of 1851 reflected calls for improvement in institutional conditions and to deal with offender needs in more effective and hu- mane ways. During the 1870s, various prisons were constructed across Canada. The 1889 Act to Permit the Conditional Release of First Offenders in Certain Cases introduced the practice of releasing offenders on their own recognizance rather than imposing a sentence. Probation was first mentioned in revisions to the Criminal Code of Canada in 1892. Changes to the Code in 1921 required offenders to report to an officer of the court. This “probation” became the cornerstone of what has become “community corrections” in Canada. In Canada, jurisdiction regarding supervision of offenders is split be- tween the provinces and the federal government. Unlike the United Robin J. Wilson and Michelle Prinzo 61 States, where jurisdiction is determined by offense type, Canadian of- fenders are routed to provincial or federal supervision based solely on the length of sentence to be served. Sentences of less than two years (i.e., two years less a day, or under) are administered by the province in which the offense occurred, while offenders serving sentences of two years or more fall under federal jurisdiction. All probationary sentences are supervised by the provincial correctional service. Currently, the majority of Canada’s federal offender population achieves conditional release at some point during their sentence, be that Day Parole, Full Parole, or Statutory Release. Parole is roughly equiva- lent to earned remission or “time off for good behavior,” and is granted by the National Parole Board, which is independent from the correc- tional service. Statutory Release (SR) is mandated in law, but may be suspended (i.e., this keeps the offender in prison) for those who meet rigorous criteria demonstrating that some undue risk is posed to com- munity safety. Those offenders who achieve conditional release are su- pervised by a parole officer who ensures adherence to release conditions imposed by the National Parole Board. Although conditional release in Canada has a long history, recent shifts in policy reflect the community’s increasing intolerance for crim- inal behavior, particularly in regard to sexual offenders. The sociopolit- ical pendulum has swung back to punishment, away from rehabilitation. Public outcry has led to political responses such as longer sentences, more stringent criteria for granting conditional release, and measures to enhance supervision in the community. Accordingly, there has been a decline over the past decade in the proportion of federal offenders af- forded community supervision (Correctional Service of Canada, 2000). As a result, fewer federal offenders achieve conditional release prior to their Statutory Release Date, and many SR candidates are being de- tained until sentence completion (or Warrant Expiry Date–WED). This is particularly true for sexual offenders. Although maintaining offenders in prison for longer periods of time may satisfy some of the public’s belief that offenders require punish- ment and a significant period of removal from society, the practice of limiting community supervision is a proverbial double-edged sword. Indeed, research shows that the availability of facilitated community re- integration significantly impacts risk for future involvement in criminal activities, with those offenders receiving community support being at less risk for recidivism (Andrews & Bonta, 1998; Wilson, Stewart, Stirpe, Barrett, & Cripps, 2000). Andrews and Bonta (1998) have clearly shown that imprisonment without appropriate rehabilitative ser- 62 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions vices amounts to a virtual “deep freeze,” in which offenders are returned to the community slightly worse than when they went in. Conversely, in- terventions which attend to risk levels, criminogenic needs, and of- fender responsivity can significantly increase success on release. Traditionally, offenders released at WED have not had the benefit of services (e.g., treatment, advocacy) available to conditionally released offenders. This is in clear contradiction to the research noted above. Sexual Offenders The Criminal Code of Canada has undergone numerous modifica- tions during the 20th century. However, it was not until the 1970s that changes were made regarding the classification of sexual offenses as crimes of violence. These changes reflected the public’s call for clearer definitions of offenses, appropriate sentencing, and rules governing evi- dence and legal proceedings. For example, prior to 1976, Common Law rules allowed evidence pertaining to the woman’s sexual and moral character to be entered as evidence for the defense. As such, the trial process was often biased against the victim, in that it was assumed that the victim had likely consented to the particular act in question if she had previously consented to sexual activity with the accused. In 1976, amendments were made to the Criminal Code of Canada providing that no questions could be asked about the sexual conduct of the complain- ant with other people unless “reasonable notice” had been given to the courts. This has been colloquially referred to as the “rape shield.” In 1978, the Law Reform Commission of Canada (LRCC–see review in Holden, 1999, available from the present authors) published a work- ing paper which gathered the majority of sexual offenses (e.g., rape, in- decent exposure, indecent assault) together under the general crime of “sexual assault.” Sexual Assault, as defined by the LRCC, would be committed if a person had sexual contact with another person without consent. The recommendations of the LRCC were enacted in 1983, and comprised the first major attempt to operationalize the Canadian legal community’s definition and handling of sexual offenses and offenders. Since that time, other legislation has been enacted to further clarify the Criminal Code of Canada sections related to sexual offenses. Subse- quent laws have dealt with protection of victim rights, the reporting and prosecution of sexual assaults, victim impact statements, and defini- tions of consent. Presently, penalties for sexual offenses range from six months to life imprisonment. Substantial changes in regard to correctional policies Robin J. Wilson and Michelle Prinzo 63 were reflected in the Corrections and Conditional Release Act (CCRA) of 1992. The CCRA set out policy and guidelines for correctional ser- vices and parole boards in their dealings with offenders. One specific change was in regard to detention provisions, in that the Act made it easier to detain sex offenders, particularly those who committed crimes against children. One of the two criteria for detention states that if “the offense was a sexual offense involving a child and there are reasonable grounds to believe that the offender is likely to commit a sexual offense involving a child before expiration of the offender’s sentence according to law” then a referral for detention will be considered. Many jurisdictions in Canada (and, indeed, other countries) have ex- perienced the difficulties associated with high profile cases of sexual assault or sex offender recidivism. It would appear that the public’s sen- timents towards sexual offenders are hardening and, as such, policy changes have resulted in more restrictive measures for many sexual of- fenders. In particular, more Canadian sexual offenders are being de- clared “dangerous offenders” (DOs), a legal definition which provides for the imposition of an indeterminate sentence for repeat, serious of- fenders and, in particular, sexual offenders. Experience has shown that few DOs are ever granted conditional release and, thus, serve 100% of their sentences in penitentiaries. For those serious offenders not meet- ing the DO standard, the courts now have the additional option of im- posing Long Term Offender (LTO) designations, which consist of up to 10 years of post-sentence community supervision. Of those offenders not designated as DOs or LTOs, many more are being detained to WED in accordance with provisions of the CCRA. While the result of the policy changes noted above has been that sex- ual offenders are removed from society for longer periods, there are considerable concerns regarding the larger implications of such legisla- tion. In particular, resources in the community (e.g., parole supervision, counselling, psychoeducational programming) are normally linked to a supervising agency, like the Correctional Service of Canada (CSC). Of- fenders released at WED have not traditionally had the benefit of access to community reintegration services. Over the years, community-based management of sexual offenders has become an increasingly public af- fair, to the extent that many jurisdictions now have policies regarding public notification, sex offender registration, and provision of DNA samples. In Canada, sex offenders comprise 24% of all federal inmates. National Parole Board statistics reveal that the percentage of prisoners in the total prison population actually referred for detention increased from 4.3% in 1989-1990 to 10.2% in 1994-1995 (Mennonite Central 64 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions Committee, 1996). The majority of the difference is attributable to in- creased detention rates for sexual offenders. Policy documents of the Correctional Service of Canada define com- munity reintegration as “all activity and programming conducted to prepare an offender to return safely to the community and live as a law-abiding citizen” (Thurber, 1998). CSC personnel use offender-spe- cific information, including static and dynamic risk determinants, in their efforts to reduce risk for re-offense. Comprehensive data manage- ment allows for actuarial assessment, and also serves to better inform parole supervisors about offender functioning in the community. Such data are crucial for program development and implementation, as well as the formulation of correctional plans for individual offenders. These measures help offenders to reintegrate to society while allowing for en- hanced monitoring of offense-specific behavior patterns. Close super- vision of precursors to offending allows for therapeutic suspension of conditional release, and periodic revocation where indicated. If an of- fender is detained until WED, the community loses the ability to moni- tor these variables and, subsequently, the spirit of reintegration is defeated. In the past, decision makers were forced to rely on clinical judgments rendered by professionals; however, the average predictive accuracy of such judgments in predicting sex offence recidivism has been shown to be only slightly better than chance (Hanson & Thornton, 1999). In- creasingly, the decision to detain an offender is determined using em- pirically-derived, standardized procedures and resultant actuarial measures (Hanson, 2000). However, despite advances in this domain (e.g., STATIC-99–Hanson & Thornton, 1999), the procedures used in determining risk are still subject to limitations of reliability and validity. Notwithstanding recent findings and innovations in sex offender risk assessment, much debate remains in the literature regarding how risk should be assessed and managed. Hanson and Bussière (1998) conducted a rigorous meta-analysis of the predictors of sexual recidivism, the results of which suggest that sex offence recidivism is closely related to sexual deviance. The strongest predictors were positive phallometric assessments, a history of sex of- fences, and a history of diverse sexual crimes. They also found that sex- ual recidivism was associated with prior nonsexual offences and a diagnosis of Antisocial Personality Disorder (particularly, those ex- treme cases of APD commonly thought to be reflective of psychopa- thy–Hare, 1993). The meta-analysis also revealed that predictors of general and non-sexual violent recidivism were similar to predictors of Robin J. Wilson and Michelle Prinzo 65 general recidivism among non-sexual criminals. Almost all predictors of sexual offence recidivism were historical or stable (i.e., unlikely to change, regardless of the intervention employed). The most changeable risk factor was motivation for treatment, in that offenders who rejected or performed poorly in treatment were at higher risk to re-offend. The Rise of Restorative Justice Over the last 10 years, “restorative justice” has been widely recog- nized in Canadian public discourse (Roach, 2000). This has happened as many nations grapple with burgeoning prison populations and the difficulties associated with returning previously incarcerated persons to increasingly intolerant communities. Those at the forefront of the re- storative justice vanguard assert that any successful reintegration of of- fenders to the community must include active participation of all community stakeholders. This has been, at times, a bitter pill for fearful citizens to swallow, particularly in regard to sexual offenders. Although restorative justice is not a new concept in Canada, it was never fully embraced until sentencing reforms were made in 1996. Ear- lier committees had acknowledged the potential for use of such princi- ples, but the 1996 reforms were the first to recognize the restorative concept that the provision of reparations to victims and the community are legitimate goals of sentencing (Martin’s Annual Criminal Code, 1999). Since those reforms, restorative justice continued to rise in crimi- nal justice discourse, and can be seen in decisions made by the Supreme Court of Canada (see review in Roach, 2000). The multiple purposes of sentencing, which include rehabilitation, deterrence, incapacitation, and retribution, have resulted in a great deal of subjectivity in regard to sentencing decisions by judges. Adding re- storative justice opportunities to the mix is likely to exacerbate this dif- ficulty but, nonetheless, represents a potential for transforming the way in which we deal with offenders and the goals of sentencing. The con- cept of restorative justice includes important principles such as fairness, equality, accountability, reparation, forgiveness, inclusion, and healing. In traditional criminal justice systems, justice is used to heal victims and requires that all stakeholders have an opportunity to be involved in harm assessment and reduction (Miller & Schacter, 2000). The Law Commission of Canada outlines three fundamental princi- ples of restorative justice: 66 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions 1. crime is a violation of a relationship among victims, offenders, and the community; 2. restoration involves the victim, the offender, and community members; and 3. a consensus approach should be used in the application of justice. (Roach, 2000) The Supreme Court of Canada has ruled that restorative sanctions can achieve the multiple purposes of punishment, including denunciation, de- terrence, proportionality, and rehabilitation (R. v. Gladue; see Roach, 2000). Restorative justice is often defined and used as a means to promote accountability of offenders for crime. Offenders are encouraged to ac- cept responsibility and to face the victims of their crime, as well as other members of the community. Some argue that accountability is a more meaningful response in acknowledging the consequences of crime than are short terms of imprisonment. Accepting responsibility and making amends are key features of restorative justice. These are also particu- larly helpful in establishing workable offender reintegration programs. Restorative justice has also shown that rehabilitation of offenders is an important and achievable goal. The 1970s saw a decline in the sup- port for rehabilitation of offenders, as many subscribed to the “nothing works” philosophy (Martinson, 1974; Roach, 2000). Many continue to adhere to this doctrine, in spite of evidence to the contrary (Gendreau, Smith, & Goggin, in press; Hanson, 2000). Further, the Supreme Court of Canada has defended the efforts of restorative justice and identified that rehabilitation does not end in prison, but that community participa- tion in justice, including treatment, and the participation of families and the community at large, plays a pivotal role. To summarize, restorative justice is defended as a more effective means of crime prevention than punitive approaches (Roach, 2000). The focus is on the specific deterrence of offenders such that an of- fender who is restored, or is on the road to restoration, is less likely to re-offend. Further, restorative justice initiatives act as a general deter- rent in that they serve to increase community involvement regarding, and monitoring of, particular behaviors in particular communities. Re- storative justice in Canada is continuing to grow in its appeal, as com- munity groups, victims, women, and Aboriginal peoples continue to work together to help restorative principles gain acceptance by criminal justice professionals. Robin J. Wilson and Michelle Prinzo 67 Circles of Support and Accountability Notwithstanding the advances reported above, risk assessment con- tinues to be an inexact science and complete accuracy is likely unob- tainable. With that in mind, it is unlikely that the public will tolerate even the comparatively small recidivism rates to which innovations in risk assessment and community supervision and treatment strategies have contributed (Wilson, Kirkegaard, & Heise, 1998). In essence, without the ability to completely eradicate recidivism, communities will continue to call for measures to keep high-risk offenders away. Nonetheless, most sexual offenders are serving determinate sentences, and will face reentry to the community at some point. As noted above, those offenders detained to WED are released with few structured sup- ports to assist in their reintegration, with the end result being that there are no additional safeguards for the public other than those afforded by the police. Such offenders are not bound by any release conditions, are not supervised by parole or probation officers and, at present, are not routinely required to inform law enforcement of their residence or ac- tivities. Efforts to address the latter point, using Megan’s Law in the United States as a model, are presently before the Ontario Legislature. Presently, police may apply to the court to have peace bonds imposed on certain offenders, which allows for the imposition of conditions for remaining in the community. The Community Reintegration Project (CRP–Mennonite Central Committee, 1996) was specifically conceived to address the needs of communities by providing outreach to high-risk sex offenders released at WED and by providing a framework for the formation of Circles of Support and Accountability. The purpose of the CRP is to speak to the fears of victims, to reduce the risk of re-offence, and to ease offenders’ transition from institutional to community settings. Circles of Support are intended to provide the development of a relapse prevention team for an ex-offender. A typical circle is comprised of concerned citizens (volunteers) who have a common interest in helping to prevent further sexual victimization through guidance, advocacy and monitoring of of- fender activities in the community. The circle has the dual responsibil- ity of providing a caring community for the core member (i.e., the offender) while striving to increase community safety. A graphic repre- sentation of a Circle of Support is shown in Figure 1. The primary initiative of the CRP is a pilot project set up for South-Central Ontario involving recruitment of volunteers from the community to form Circles of Support around high-risk/high-profile 68 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions sex offenders. Circles provide intensive support for the core member to assist in a safe, orderly adjustment to everyday life in the community. In order to achieve this, Circle members must fulfil many roles: 1. Circle members function as advocates, working to increase coop- eration with police, victims, treatment providers, and the commu- nity. 2. They confront the core member about those attitudes and behav- iors which put him at risk for relapse. 3. Members are available to assist the core member through emer- gencies. Robin J. Wilson and Michelle Prinzo 69 Community Family and Friends Circle of Support Core Members Police–Special Agencies, Consultants Faith Community and General Services FIGURE 1. Relationships of the Circle Within the Community 4. As part of the greater community, Circle members mediate com- munity concerns. 5. Circle members encourage and join the core member in celebrat- ing anniversaries, milestones, and other victories in their reinte- gration. Optimal circle size is four to six community volunteers, in order to provide a variety of opinions. Members of circles have included mem- bers of faith congregations, police officers, psychologists, family medical practitioners, and community advocates. All Circle volun- teers are screened and informed of the initial one-year commitment. The first two months of any Circle are usually intense, involving one to two contacts with the core member per week. Depending on the of- fender and his needs, daily contact may be required during the period immediately following release. The full Circle meets weekly and once the core member adjusts to his new life in the community, meetings are reduced to twice monthly. To be effective in the reintegration pro- cess, Circles should meet on an as-needed basis. It is expected that some Circles will be maintained for several years, reflecting the real- ity that, for some offenders, reintegration is a long-term process. To date, 30 Circles have been undertaken in South-Central Ontario, as part of the pilot project. Circles have been undertaken in other parts of Canada; however, data regarding their success or failure will not be presented here. METHOD Subjects Subjects were 30 federally-sentenced, male sexual offenders de- tained until their warrant expiry date (demographic data are presented in Table 1). All subjects were identified by institutional staff as being at substantial risk to re-offend following release, hence the detention order by the National Parole Board. Each subject is currently, or has recently been, involved in a Circle of Support in South-Central Ontario, under the management of the Mennonite Central Committee of Ontario. Aver- age time of follow-up, which is an approximation of the length of time of involvement in a Circle, is 36 months with the range being from 16 to 79 months. 70 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions Actuarial Measures The STATIC-99 (Hanson & Thornton, 1999) was created by adding together items from two previously-existing measures, the Rapid Risk Assessment of Sex Offender Recidivism (RRASOR–Hanson, 1997) and the Structured Anchored Clinical Judgement (SACJ–Grubin, 1998). The resultant scale was found to be more accurate in its predic- tive ability than either measure alone. Hanson and Thornton also showed that the STATIC-99 is more accurate than unstructured clinical judgment, and that it possesses moderate accuracy in predicting any vi- olent recidivism among sex offenders. The Sex Offender Risk Assessment Guide (SORAG–Quinsey, Har- ris, Rice, & Cormier, 1998) is an actuarial risk assessment tool devised by its authors to measure risk for recidivism by sexual offenders. It is simi- lar in construction to the Violence Risk Appraisal Guide (VRAG– Quinsey et al., 1998). While the SORAG has no published validation data, the VRAG has demonstrated reasonable predictive validity with respect to potential for violent recidivism, including sexual re-offend- ing. The STATIC-99, RRASOR and the SORAG were completed for each offender after review of the documents and reports on the Offender Management System (OMS), the national offender data bank of the Robin J. Wilson and Michelle Prinzo 71 TABLE 1. Group Demographics (N = 30) M(SD) age in years 41.6 (11.65) % deviant phallometrics 46.7 M(SD) STATIC-99 6.03 (2.17) M(SD) RRASOR 3.17 (1.60) M(SD) SORAG 17.30 (10.64) Victims M number 2.8 % female only 53.3 % familial only 7.0 % children only 46.7 % Predicted survival (STATIC-99) 73.98 % Actual survival 90.0 M(SD) follow-up (months) 36 (14.84) Range (months) 16 - 79 Correctional Service of Canada. Each offender’s level of risk was as- sessed by his scores on the STATIC-99, RRASOR, and SORAG (see Table 1). Once each offender’s level of risk was calculated using the ac- tuarial measures, survival analysis was used to estimate recidivism probabilities. Survival analysis calculates the probability of recidivating for each time period as long as the offender has not yet recidivated (Hanson & Thornton, 2000). Survival analysis was used due to its ad- vantage of being able to estimate month-by-month recidivism projec- tions when follow-up periods vary across offenders. Instances of recidivism were defined as charges or convictions recorded on an of- fender’s criminal record. RESULTS Mean time at risk was 36 months with the range being from 16 months to just over six-and-a-half years. Average scores on the STATIC-99, RRASOR, and SORAG (N = 23) were 6.03, 3.12, and 17.30, respectively. These ratings confirm the higher risk status of of- fenders involved in the study. Using mean time at risk and average STATIC-99 score, in combination with established survival data (Hanson, personal communication), we predicted that approximately seven instances of sexual recidivism should have occurred in this group to date. Average expected survival over the 30 cases was 73.98% using the formula: S (survival probabilities)/N However, only three persons have recidivated to date: one for indecent telephone calls, another for offenses against a child, and the last for sex- ual assault against a female adult. In each case, the offense committed was categorically less severe or invasive than the crime for which the offender had previously served a prison sentence. The difference be- tween the expected survival and the observed survival in our sample was 16.02%. Using a z test with a correction for small sample (Hardyck & Petrinovich, 1969, pp. 167-169) in which the deviation of the observed survival proportion is compared to the expected survival proportion, based on Hanson’s data, we were able to demonstrate a strong trend toward significance (z = 1.79, N = 30, p < .075). 72 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions DISCUSSION This paper discusses a Canadian approach to restorative justice with conditionally-released, high-risk sexual offenders. Since the pilot pro- gram began, many others have “sprung up” around the nation. Further, members of our group have been invited to present the model in numer- ous international settings, including the United States, United King- dom, and South Africa. Generally, the response has been favorable, indicating significant interest in the prospect of adding another dimen- sion to the safe and ethical management of offenders. The Circles of Support initiative uses community volunteers to aid in the management of sex offender risk. While these volunteers are screened and provided with training and support from professionals (e.g., psychologists, psychiatrists, etc.), and many of the volunteers are retired social service workers, the point is well taken that these are aver- age community persons involved in a potentially dangerous enterprise. We argue, however, that the risk management provided by government agencies has fallen short of the needs of the community. This is particu- larly true of offenders released at sentence completion, usually to no treatment and no correctional supervision. Of those services that might be available through other agencies, decreases in funding and the resul- tant longer waiting lists have jeopardized community safety. Therefore, it is our position that the community must take a greater stake in ensur- ing its own safety. By including community members in the risk man- agement process, we increase the level of understanding of the issues at hand, and empower citizens to be informed as to the risks present in their surroundings. While some may argue that this is fatalistic, we be- lieve that an informed community is a safer community. This study recorded scores on three actuarial sexual offense risk as- sessment measures (RRASOR, SORAG and STATIC-99), using the latter to predict estimated sexual recidivism rates. All three measures assessed the subjects to be high risk to re-offend sexually, demonstrat- ing concurrent validity of the measures. This confirms earlier research in which the STATIC-99 predicted sex offense recidivism with similar levels of accuracy to the SORAG (Hanson & Thornton, 2000). The results of comparisons between projected and actual recidivism show that the group of offenders included in this analysis are recidivating at a rate less than 40% of that expected using the STATIC-99 survival data provided by Hanson (personal communication). Of the 30 core members included in this study, only three have recidivated to date, while the probability was that seven would have done so. While it is Robin J. Wilson and Michelle Prinzo 73 possible that more offenders might have recidivated, the likelihood of that having occurred is decreased by the level of involvement of the Cir- cle in the core member’s life. While difficult to determine conclusively, the strong trend toward significance in comparing the observed and expected recidivism pro- portions suggests that involvement in a Circle of Support has aided the offenders reported here in making a smoother transition to the commu- nity. Specifically, being held accountable for their actions, and being af- forded a close community support network, has given them the kind of support that might have been available had they received conditional re- lease. Results of a recently-reported investigation looking at recidivism rates for conditionally-released offenders to the same geographic area (Wilson et al., 2000) also suggested that collaborative facilitation of of- fender reintegration, including a variety of stakeholders, increases the likelihood of success. In that study, recidivism rates were also lower than those reported elsewhere in the literature (Barbaree, Seto, & Maric, 1996; Motiuk & Brown, 1996). At present, the recidivism rates of our subject group have not been compared specifically to an analogous group released without the bene- fit of a Circle of Support. National data reported by Motiuk, Belcourt, and Bonta (1995) suggest that the sexual recidivism rate of detained sexual offenders is in line with, if not slightly higher than, the predic- tions made by the STATIC-99 survival data used here. It would be of in- terest to many to establish whether or not a statistically significant difference exists in regard to recidivism rates. However, Barbaree (1997) has suggested that, due to low base-rates, the demonstration of “treatment” effects is difficult. Nonetheless, this should not be seen as a rationale for discontinuing initiatives such as that detailed here. Indeed, Gendreau and his colleagues (in press) have ably demonstrated that small differences in recidivism rates have large social implications with respect to victim harm and a variety of cost indicators (e.g., trying, in- carcerating, and treating offenders; and providing services for victims). The Circles of Support model might be of particular interest to those ju- risdictions where the costs associated with maintaining offenders have grown unwieldy. The current study used actuarial assessment, primarily of static vari- ables, to predict a recidivism rate which was then compared to an actual recidivism rate. The reader is cautioned that, although certainly better than clinical judgment alone, actuarial assessment is also less than per- fect. Hanson and Thornton acknowledge that the STATIC-99 provides only moderate predictive ability in regard to sexual or violent recidi- 74 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions vism. Recent research (Hanson & Harris, 2000) has suggested that gen- eral risk assessment efforts would be greatly enhanced by consideration of dynamic factors, such as intimacy deficits, treatment failure, and non-cooperation with supervision. Hanson and Harris (2000) have recently introduced the Sex Offender Needs Assessment Rating, or SONAR, which attempts to accomplish the same goal as the STATIC-99, however, in re- gard to dynamic factors. While static indicators of risk have been tradi- tionally viewed as superior in determining risk for re-offense, meta- analyses conducted by Gendreau, Little, and Goggin (1996) demon- strated that offender needs (or dynamic variables) are able to predict re- cidivism with similar strength. Dynamic factors are often more useful to treatment and supervisory personnel in managing offender risk in the community. We argue that Circles of Support are a good first step in in- cluding the community-at-large in risk management, and that they are a viable means to increase public safety. REFERENCES Andrews, D.A. & Bonta, J. (1998). The psychology of criminal conduct, Second Edi- tion. Cincinnati, OH: Anderson. Barbaree, H.E. (1997). Evaluating treatment efficacy with sexual offenders: The insen- sitivity of recidivism studies to treatment effectiveness. Sexual Abuse, 9, 111-128. Barbaree, H.E., Seto, M.C., & Maric, A. (1996). Sex offender characteristics, response to treatment and correctional release decisions at the Warkworth Sexual Behaviour Clinic (Working Papers in Impulsivity Research Report). Toronto, ON: Clarke Insti- tute of Psychiatry. Correctional Service of Canada (2000). The safe return of offenders to the community: Statistical Overview 2000. Ottawa, ON: Research Branch, CSC. (available at www.csc-scc.gc.ca) Gendreau, P., Little, T., & Goggin, C. (1996). A meta-analysis of the predictors of adult offender recidivism: What works? Criminology, 34, 575-607. Gendreau, P., Smith, P., & Goggin, C. (in press). Treatment programs in correc- tions. In: J. Winterdyk (Ed.), Corrections in Canada: Social reaction to crime. Scarborough, ON: Prentice-Hall. Griffiths, C. & Cunningham, A. (2000). Canadian corrections. Scarborough, ON: Nel- son Thomas Learning. Grubin, D. (1998). Sex offending against children: Understanding the risk (Police Re- search Series Paper 99). London: Home Office. Hanson, R.K. (1997). The development of a brief actuarial risk scale for sexual offense recidivism (User Report 1997-04). Ottawa: ON: Department of the Solicitor General of Canada. (available at www.sgc.gc.ca) Hanson, R.K. (2000). Risk assessment: Prepared for the Association for the Treatment of Sex Abusers. Beaverton, OR: ATSA. Robin J. Wilson and Michelle Prinzo 75 Hanson, R.K. (2000, November). The effectiveness of treatment for sexual offenders: Report of the ATSA Collaborative Research Committee. Plenary address made at the 19th Annual Research and Treatment Conference of the Association for the Treat- ment of Sexual Abusers, San Diego, CA. Hanson, R.K. & Bussière, M.T. (1996). Predictors of sexual offender recidivism: A meta-analysis (User Report 1996-04). Ottawa, ON: Department of the Solicitor Gen- eral of Canada. (available at www.sgc.gc.ca) Hanson, R.K. & Bussière, M.T. (1998). Predicting relapse: A meta-analysis of sexual offender recidivism studies. Journal of Consulting and Clinical Psychology, 66, 348-362. Hanson, R.K. & Harris, A.J.R. (2000). The Sex Offender Need Assessment Rating (SONAR): A method for measuring change in risk levels (User Report 2000-1). Ot- tawa, ON: Department of the Solicitor General of Canada. (available at www. sgc.gc.ca) 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. (available at 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. Hardyck, C.D. & Petrinovich, L.F. (1969). Introduction to statistics for the behavioral sciences. Philadelphia, PA: W.B. Saunders. Hare, R. (1993). Without conscience. New York, NY: Guilford. Holden, R. (1999). Legal implications of sexual assault. In: R.J. Wilson & D. Lagace (Eds.), Assessment, treatment, & supervision of sex offenders. Ottawa, ON: Correc- tional Service of Canada. Martin’s Annual Criminal Code (1999). Aurora, ON: Canada Law Book. Martinson, R. (1974). What works? Questions and answers about prison reform. The Public Interest, 35, 22-54. Mennonite Central Committee (1996). Community Reintegration Project Manual. To- ronto, ON: MCC. Miller, S.B. & Schacter, M. (2000). From restorative justice to restorative governance. Canadian Journal of Criminology, 42, 405-420. Motiuk, L.L. & Brown, S.L. (1996). Factors related to recidivism among released fed- eral sex offenders (Research Report # R-49). Ottawa, ON: Research Division, Cor- rectional Service of Canada. (Available at www.csc-scc.gc.ca) Motiuk, L.L., Belcourt, R.L., & Bonta, J. (1995). Managing high risk offenders: A post detention follow-up (Research Report #R-39). Ottawa, ON: Research Branch, Cor- rectional Service of Canada. (available at www.csc-scc.gc.ca) Quinsey, V.L., Harris, G.T., Rice, M.E., & Cormier, C.A. (1998). Violent offenders: Appraising and managing risk. Washington, DC: American Psychological Associa- tion. Roach, K. (2000). Changing punishment at the turn of the century: Restorative justice on the rise. Canadian Journal of Criminology, 42, 249-280. Thurber, A. (1998). Understanding Offender Reintegration. Forum on Corrections Re- search, 10. (available at www.csc-scc.gc.ca) 76 Sex Offender Treatment: Accomplishments, Challenges, and Future Directions Wilson, R.J., Kirkegaard, H., & Heise, E. (1998, September). Community reintegration project for high risk sex offenders. Presentation at the 23rd Annual Training Institute of the American Probation and Parole Association, Norfolk, VA. Wilson, R.J., Stewart, L., Stirpe, T., Barrett, M., & Cripps, J.E. (2000). Commu- nity-based sex offender management: Combining parole supervision and treatment to reduce recidivism. Canadian Journal of Criminology, 41, 177-188. Robin J. Wilson and Michelle Prinzo 77