Chapter 8: Sex Offender Management Strategies

Introduction

Prevention and intervention strategies for sexual offending behavior, including sex offender management, have become increasingly prominent and important in the United States.¹ The concept of sex offender management has been conceptualized under the construct of a Comprehensive Approach to Sex Offender Management (CASOM) by the Center for Sex Offender Management (CSOM). The CASOM model (CSOM, 2007) includes the following —

• Fundamental principles
  • Victim-centered approach
  • Specialized knowledge and training for professionals
  • Public education
  • Monitoring and evaluation of the strategies
  • Multidisciplinary collaboration
• Critical components
  • Investigation, prosecution and disposition
  • Assessment
  • Treatment (For more on treatment, see Chapter 7, "The Effectiveness of Treatment for Adult Sex Offenders," in the Adult section.)
  • Supervision
  • Reentry
  • Registration and community notification

According to the Bureau of Justice Statistics, at yearend 2008 more than 165,000 offenders convicted of rape or sexual assault were in state prisons (Guerino, Harrison & Sabol, 2011). The National Parole Resource Center (2014) indicates 95 percent of these offenders will ultimately be released to communities, at a rate of approximately 10,000–20,000 per year. Additionally, more than 861,800 registered sex offenders currently reside in communities across the United States (National Center for Missing & Exploited Children, 2017). While it is difficult to track national trends over time, there is little question that the number of sex offenders under correctional supervision in the community has increased substantially over the past 20 years. In fact, sex offender management laws have become so prominent in the United States that the issue was identified as the fifth most important area of concern for state legislators (CSOM, 2008). Such laws typically address issues such as incapacitation, retribution/punishment, deterrence and rehabilitation (CSOM, 2008). During the 2007–08 legislative biennium alone, 1,500 bills related to sexual offenders were introduced in 44 states (six states had no legislative session during this timeframe), with 275 of these bills passing into law (Council of State Governments [CSG], 2010).

In a survey of nearly 600 professionals working in the field of sex offender management, nearly a third of whom (30.5 percent) have more than 10 years' experience working with this population, most believed that community supervision and treatment, and to a lesser extent electronic monitoring (GPS) and sex offender registration and notification (SORN), reduce recidivism. In addition, barriers to successful reentry were identified as lack of housing and employment, public sentiment and residence restrictions, among others (Council of State Governments Justice Center, 2015).

Despite the intuitive value of using science to guide decision-making, laws and policies designed to combat sexual offending are often introduced or enacted in absence of empirical support. This dynamic was recently acknowledged and identified as a concern by the national experts — both researchers and practitioners — who participated in the February 2012 SOMAPI forum. The reasons why this occurs are varied and complex, and they will not be explored in this chapter.² However, there is little question that both public safety and the efficient use of public resources would be enhanced if sex offender management strategies were based on evidence of effectiveness rather than other factors. This chapter on sex offender management strategies was developed with this in mind.

This chapter does not discuss the theoretical and sociological explanations for a given policy or place the research within this context. It also does not present an exhaustive review of the research; it focuses on recent studies deemed to be important for understanding the effectiveness of a given strategy. Finally, its primary focus is on the management of adult sexual offenders. Although some research on juveniles who commit sexual offenses is included, the effectiveness of sex offender management strategies with a juvenile population is addressed in the Juvenile section of this publication.

Summary of Research Findings

Specialized Supervision

The development and refinement of specialized legal supervision for sexual offenders has largely occurred over the past 25 years. Specialized supervision frequently involves specially trained probation and parole officers who manage a caseload of sexual offenders using sex-offender-specific supervision strategies that include special conditions of supervision, multidisciplinary collaboration with a treatment provider and, if appropriate and permissible, the use of GPS monitoring and polygraph testing. Based on responses to a 2008 survey of state officials, most states use some form of specialized supervision to manage risk and provide services to sexual offenders in the community; in addition, many states use sex-offender-specific probation or parole caseloads (Daly, 2008). (For a discussion of adult "Sex Offender Risk Assessment," see Chapter 6 In terms of strategies used by specialized supervision officers, a survey of probation and parole supervisors (N = 732) conducted in 1994 found that 85 percent referred offenders to sex-offender-specific counseling and that 30 percent of probation officers and 32 percent of parole officers had specialized caseloads; however, less than 10 percent required polygraph testing (English, Pullen & Jones, 1996). The importance of multidisciplinary collaboration with supervision officers was also supported in a survey of treatment providers from 45 states and the District of Columbia (N = 190), where 90 percent said their rapport with probation officers was excellent or good, 24.2 percent said probation officers attended weekly group sessions and 87.4 percent said communication with probation officers was essential (McGrath, Cumming & Holt, 2002).

This section reviews research on the effectiveness of specialized supervision practices. It is important to note that these are not sex-offender-specific studies. Research relating to the effectiveness of Circles of Support and Accountability (COSA), civil commitment, polygraph and electronic monitoring (including GPS) immediately follows. These studies focus primarily on sex-offender-specific supervision strategies.

Research

Several large-scale studies have assessed the effectiveness of intensive supervision used with criminal offenders. It is not known whether findings from these studies are generalizable to sex offender populations, but the findings provide important insights concerning the effectiveness of intensive supervision overall. In one large-scale systematic review of 291 studies conducted over a 40-year period on various intensive supervision programs used with criminal offenders, the Washington State Institute for Public Policy found no research support for the effectiveness of community-based intensive supervised probation with a primary surveillance orientation in reducing criminal recidivism (n = 24 studies). (For information on "Adult Sex Offender Recidivism," see Chapter 5 in the Adult section.) However, Washington State did find research support for the effectiveness of treatment-oriented intensive supervised probation, which produced an average reduction in criminal recidivism of 21.9 percent (n = 10 studies). Based on these results, Washington concluded that rehabilitation via treatment — not intensive supervision — leads to a reduction in criminal recidivism (Aos, Miller & Drake, 2006). It should be noted that this study was a follow-up to an earlier study by the same state agency, in which the authors concluded that surveillance-oriented intensive supervised probation had a small effect, which was not statistically significant, on reducing criminal offender recidivism (n = 19 studies) (Aos et al., 2001).

There is no research support for intensive supervision without a rehabilitative treatment approach.

A second study on the effectiveness of intensive supervised probation for general criminal offenders was a randomized clinical trial³ conducted between 1986 and 1991 across 14 sites in nine states. In a one-year follow-up, the offenders subject to ISP were rearrested at a rate of 37 percent, while the offenders not subject to intensive supervised probation were rearrested at a rate of 33 percent. Further, those subject to ISP were recommitted to prison at a rate of 27 percent, while the non-intensive supervised probation recommitment rate was 19 percent. In discussing the study results, the researchers concluded, "Despite the experience of hundreds of intensive supervision programs in this country and many studies, albeit few experimental, we still know very little about the effectiveness of these programs to reduce prison overcrowding and… to reduce crime in detectable ways" (Petersilia & Turner, 1993, p. 121).

Questions about the effectiveness of intensive supervision in the absence of treatment have led to the development of intensive supervision programs with a treatment orientation. A specific example is the containment approach, which includes collaboration on specialized supervision of sexual offenders provided by trained supervision personnel, sex-offense-specific treatment and polygraph assessment. Unlike many other sex management strategies that have been implemented over the years, English, Pullen and Jones (1996) developed the containment approach based on their study of best practices in place across the country.

Research on the effectiveness of the containment approach has been completed in a handful of jurisdictions across the country. One study was conducted in Jackson County, Oregon, where the community corrections office integrated treatment, supervision and polygraph assessment in a multidisciplinary collaboration model. The research compared the recidivism rates for sexual offenders who were subject to the containment approach between 1985 and 1995 (N = 601) with those of sexual offenders from 1) a different county (Linn County) who were not subject to the containment approach between 1985 and 1992 (n = 89), and 2) a group of non-sex offenders supervised in Jackson County between 1985 and 1995 (n = 231) in a matched sample. The study used a three- to five-year follow-up period, and recidivism in this study was defined as a new felony conviction. The study found a recidivism rate of 8.8 percent for offenders in the containment group based on a follow-up period of at least one year, while the rates for the comparison groups were 15 percent and 26.7 percent, respectively. This was a statistically significant difference in recidivism for the containment group compared to both comparison groups.⁴ The researchers also noted that sexual offenders subject to the containment approach had a higher recidivism rate than the comparison groups for the first year, possibly due to the increased supervision scrutiny provided by this approach (Aytes et al., 2001).

A second study of the containment approach used with sex offenders in Colorado found that sexual offenders subject to specialized parole supervision following release from prison (n = 1,003), which included requirements for sex-offense-specific treatment and polygraph assessment, had a statistically significant lower recidivism rate (16.1 percent) than sex offenders not subject to parole supervision (29.3 percent; n = 2,040).⁵ Recidivism in this study was measured as a new arrest, court filing, or return to prison (Lowden et al., 2003). 

Finally, a third study undertaken in Virginia compared sexual offenders subject to containment on probation and parole (n = 583) to all sexual offenders on probation and parole between 2000 and 2002 (N = 1,753) using a three- to five-year follow-up period. The results indicated that the containment sexual offenders returned to prison for any crime at a rate of 11.3 percent, and specifically for a sex crime at a rate of 0.5 percent. The comparison group had a similar return-to-prison rate of 9.9 percent for any crime and a rate of 0.6 percent for a new sex crime, a difference that was not statistically significant. The researchers hypothesized that the higher return-to-prison rate for the containment sexual offenders was due to increased surveillance and detection provided by the model. It is also important to note that the comparison group in this study (all sexual offenders released from prison between 2000 and 2002) included the sex offenders subject to containment (Boone et al., 2006).

Three additional research studies on specialized sex offender supervision are worth noting. One study compared sex offenders (n = 195) under specialized supervision and in sex-offense-specific treatment to a matched group of sex offenders who did not have community supervision using a six-year follow-up period. The results indicated that sexual offenders under community supervision had a sexual recidivism rate of 14 percent based on either a new sexual offense charge or a substantiated sexual offense by child protective services. In comparison, those who were not under supervision had a recidivism rate of 35 percent, leading to the conclusion that specialized supervision resulted in a statistically significant reduction in sexual recidivism⁶ (McGrath et al., 2003).

However, in a contradictory study completed in Illinois, no significant difference in sexual recidivism was found between sex offenders subject and not subject to specialized supervision. In this study in Lake County, Illinois, recidivism was defined as a new sex crime arrest over a three- to five-year follow-up period (Stalans, Seng & Yarnold, 2002). The results indicated that sexual offenders subject to specialized supervision (n = 104) had a sexual rearrest rate of 28.8 percent, while sexual offenders not subject to this strategy (n = 104) had a sexual rearrest rate of 25 percent (Stalans, Seng & Yarnold, 2002).

Finally, in a second contradictory study from British Columbia, Canada (N = 269), there was no significant difference in sexual reconviction between sex offenders subject to a specialized intensive community supervision program for high-risk sex offenders (the CHROME program) compared to the control groups, however, the small number of overall sexual recidivists (3 percent overall) limited the ability to distinguish between groups (Lussier et al., 2014).

Limitations

The research on the effectiveness of specialized sex offender supervision in conjunction with treatment (e.g., the containment approach) has a number of limitations. These include a small number of studies, short follow-up periods, small sample sizes, the use of different recidivism measures (making cross-study comparisons challenging), little information about the specific elements of the programs that are found to be successful and problems with the scientific rigor of some of the studies (including one study where the intervention group was part of the comparison group). Finally, general issues related to underreporting of sex crimes leads to the problem typically seen in sex offender management research — that is, a low base rate for sexual recidivism, which limits the ability to achieve significant differences between the intervention and comparison groups.

Research limitations include short follow-up periods, small sample sizes, different recidivism measures and problems with scientific rigor.

On the other hand, the research on the effectiveness and limitations of generalized intensive supervision for all criminal offenders, particularly when combined with a treatment component, is much more extensive based on a number of large-scale research studies.

In terms of future research directions, it is recommended that research using rigorous scientific methods be encouraged and supported. Comparison studies with large sample sizes and longer follow-up periods should be conducted on the effectiveness of specialized supervision in conjunction with treatment for sexual offenders. Finally, it would be beneficial for future research to identify not only the effect of the intervention, but also the program components that appear to be most beneficial and the mechanisms by which successful outcomes are achieved.

Summary

There is empirical support for the use of intensive supervision with criminal offenders in conjunction with a rehabilitative treatment approach, and some preliminary support for specialized sexual offender supervision models (such as the containment approach) that are delivered in conjunction with treatment. However, there is no research support for the use of intensive or specialized supervision either in isolation or without treatment for either population. Given the above, the SOMAPI forum participants recommended that jurisdictions should use specialized supervision with a rehabilitation orientation as one component of an overall sex offender management strategy.

Circles of Support and Accountability

The COSA model is a supervision strategy involving the use of community volunteers to provide support to an individual sex offender. COSA assists offenders in garnering community resources while holding them accountable to their self-monitoring plan, typically following completion of legal supervision. This program was first developed in Canada but has since been implemented in the United States, United Kingdom, Netherlands, Belgium, Catalonia, Bulgaria and Latvia, with a total of 446 circles identified to date in the literature (Clarke et al., 2015). Currently, there are U.S. COSA programs in California, Colorado, Minnesota and Vermont, with additional projects being developed in North Carolina and Washington, among others.

Research

The COSA model has been the subject of several studies, including a number of single studies evaluating COSA's effectiveness in reducing recidivism (one of which was a randomized controlled trial), one systematic review of the research, two cost-benefit analyses and surveys of sexual offender participants and public member volunteers.

Outcome Studies

In the Canadian outcome studies, one evaluation compared the recidivism rates of 60 COSA high-risk sex offenders and 60 non-COSA high-risk sex offenders using a 4.5 year follow-up period. The study found a 5 percent sexual recidivism rate (defined as a new sex crime charge or conviction) for the COSA group and a 16.7 percent recidivism rate for the non-COSA group. The researchers concluded that COSA participation resulted in a statistically significant reduction in sexual recidivism (Wilson, Picheca & Prinzo, 2005).⁷

In a second Canadian study, the recidivism rates for 44 high-risk sex offenders participating in COSA were compared to those for a matched comparison group of 44 high-risk sex offenders who did not participate in COSA using a 35-month follow-up period. The study found that the COSA group sexually recidivated at a 2.3 percent rate while the non-COSA group recidivated at a 13.7 percent rate, a statistically significant difference (Wilson, Cortoni & McWhinnie, 2009).⁸

Duwe (2012) conducted an evaluation of COSA effectiveness in Minnesota (N = 62) using a randomized controlled trial design and found that COSA participants had a lower rate of sexual rearrest than non-participants (0 percent vs. 3.2 percent, respectively), although the difference was not statistically significant. However, COSA participants did have a significantly lower level of any rearrest than non-participants (38.7 percent vs. 64.5 percent).⁹

Finally, a study of the COSA program in the United Kingdom examined recidivism outcomes for program participants (n = 71), as compared to those screened for COSA but who did not enter the program (n = 71). The United Kingdom COSA model is slightly different than the model that has been implemented in Canada, in that sexual offenders in the program are still under legal supervision. Based on a 55-month follow-up period, the study found a significant reduction in sexual contact or violent reconviction for COSA participants compared to nonparticipants (0 percent vs. 14.1 percent).¹⁰ However, there was no significant difference between COSA and non-COSA sex offenders on noncontact sex offense reconviction (4.2 percent vs. 2.8 percent), failure to comply with the sex offender registry (4.2 percent vs. 8.5 percent) or breach of supervision (2.8 percent vs. 0 percent). The authors described the results related to noncontact sex offenses for COSA participants from a harm reduction perspective in that the recidivistic crimes were less harmful and invasive than their pre-program participation offenses (Bates et al., 2014).

Cost-Benefit Analysis

In a cost-benefit study of a COSA program in the United Kingdom, Elliot and Beech (2012) estimated that program participation would reduce recidivism by 50 percent and that, as a result, COSA produced a return on investment of $1.04 in tangible costs (defined as direct costs of the criminal justice process) for every $1 of cost per program participant (costs were converted from pounds for purposes of this review).

Duwe also conducted a cost-benefit analysis as part of a larger evaluation of a COSA program and concluded that for every $1 spent on COSA there would be return on investment of $1.82 (Duwe, 2012).

Systematic Review

In addition to the single studies described above, a systematic review of the research on COSA was completed by Clarke and colleagues in 2015. They noted a total of 15 COSA studies including one randomized controlled trial, three retrospective cohorts with matched controls and 11 case series (case studies) from the United Kingdom (nine), Canada (three), United States (two) and Netherlands (one). The authors summarized the research on COSA as not currently showing a significant reduction in sexual reconviction or recidivism, but demonstrating harm reduction in recidivism with the new offense being less severe than the index offense, a reduction in the time to recidivism, some evidence of general recidivism reduction and having a positive return on investment (Clarke et al., 2015).

Impact on Sex Offenders and the Public

COSA was identified in the Inventory of Promising or Effective Programs in Sex Offender Management as a research-supported program model.

In a survey of COSA participants and public members, 90 percent of sex offenders from Canada who responded described participation in COSA as helpful in refraining from reoffense, while 68 percent of public members who responded said offenders' participation in COSA made them feel safer (Wilson, Picheca & Prinzo, 2005).

Impact on Volunteers

In a review of the results from several surveys of COSA volunteers, Hoing et al. (2014) noted that COSA had a number of tentative impacts on program volunteers. Positive impacts included increased self-esteem, a sense of contributing to community safety and greater support for restorative justice practices and treatment for sex offenders. However, participation in COSA also negatively impacted community volunteers through stress about the offender's risk and challenges related to dealing with the difficult behavior of sex offenders, as well as negative familial and societal reactions about their COSA volunteer work (Hoing et al., 2014).

Limitations

Regarding the outcomes from single studies, while the results regarding COSA effectiveness thus far have been positive, only four studies have employed a comparison group (one randomized controlled trial) and these studies had relatively small sample sizes. In addition, the relatively short follow-up periods are a challenge for these studies. Finally, only one study has been completed in the United States, while the other three studies were in Canada and the United Kingdom, where polygraph is not used routinely to corroborate disclosure and accountability. As a result, generalization to the United States is still in question. Therefore, future research should include larger samples sizes, sex offenders from multiple jurisdictions in the United States and longer follow-up periods. Regarding survey results, limitations include small response rates and sample sizes, leading to possible self-selection bias, and therefore, any conclusions must be considered tentative and in need of confirmation.

Summary

COSA studies thus far have demonstrated some positive results. While further use of the model is encouraged, implementation should occur in conjunction with rigorous evaluation. Far more high-quality research is needed before the efficacy and effectiveness of COSA with sexual offenders can be firmly established. The strength of the model is that it uses community resources for sex offender management and can be used in the absence of court supervision.

Polygraph

The use of polygraph assessment with sexual offenders is a somewhat more controversial management strategy than the others described thus far. (It is important to note that the containment approach — described above — includes polygraph testing as part of a comprehensive supervision and treatment strategy. This approach is premised on the assumption that the information disclosed via polygraph enhances the ability to create an individualized treatment and supervision plan.) Three different types of polygraphs are used with sexual offenders: a specific-incident exam that focuses on the sexual offense conviction or other specific offenses or behaviors, a sexual-history exam that explores the offender's history of sexual offending behavior and a maintenance exam that reviews the offender's compliance with supervision and treatment conditions.

While the extent of polygraph use in the management of sexual offenders is difficult to document, there is some evidence that polygraph use has increased since the mid-1990s. In terms of sex offenders supervised within the federal probation and pretrial service system, one study found that in fiscal years 2004–05 (N = 2,199), 44 percent of those in treatment were subject to polygraph testing (Baerga-Buffler & Johnson, 2006). Similarly, in a survey of state officials (prison, community treatment, reentry and community supervision), less than 50 percent of the respondents reported polygraph use in prison-based treatment (Daly, 2008). This percentage, however, was significantly higher than the percentage reported by English, Pullen and Jones (1996) based on their 1994 national survey of supervision officers (N = 732), in which less than 10 percent required polygraph testing. According to the Council of State Governments (2008), the following states were using polygraph testing in the management of sexual offenders: Colorado, Nevada, New Jersey, New York and Texas. Evidence that polygraph use has increased since the mid-1990s also comes from a survey of U.S. treatment providers (N = 1,307 programs), which found that respondents reporting the use of polygraph increased from 30 percent in 1996 to 79 percent in 2009 (McGrath et al., 2010).

Research

Research on polygraph use can be broken down into the following content areas: impact on disclosure, impact on sexual offender recidivism, impact on supervision professionals, impact on sexual offenders and test validity.

Impact on Disclosure

Results of multiple research studies across various jurisdictions indicate that using polygraphs with sexual offenders leads to additional disclosures. Reported increases in offender disclosure based on polygraph include the number of victims, offenses and offense categories (Ahlmeyer et al., 2000; English et al., 2000; Heil, Ahlmeyer & Simons, 2003; Hindman & Peters, 2001; O'Connell, 1998); high-risk behaviors (Buschman et al., 2010; Gannon et al., 2014; Grubin et al., 2004); and age of onset, duration of offending and frequency (English et al., 2003). One example of such a study is from the Netherlands, where child pornography sexual offenders who received polygraph testing (N = 25) yielded disclosures of high-risk behavior during treatment in the areas of masturbation to fantasies of sexual contact with children (n = 15) (including masturbation while looking at children in public [n = 9]) and masturbation while manipulating children into posing nude during webcam contact (n = 4). In addition, disclosures included cruising in public places for children (n = 14), taking children's pictures (n = 5) and having scripted scenarios to be used to sexually victimize a child if there were an opportunity to do so (n = 5) (Buschman et al., 2010). It should be noted that no comparison group was used in this study; hence, attributing the disclosures directly to the use of the polygraph is problematic. Increases in the number of victims disclosed via polygraph ranged from an initial self-report of one victim to 11–13.6 victims, depending on the study, following polygraph testing (Ahlmeyer et al., 2000; Heil, Ahlmeyer & Simons, 2003; Hindman & Peters, 2001). The rate of polygraph-aided disclosure was higher than the rate for offender self-reports (Gannon et al., 2014; Hindman & Peters, 2001), and was more pronounced for inmates than parolees (Ahlmeyer et al., 2000; Heil, Ahlmeyer & Simons, 2003). Finally, results of polygraph disclosure research indicated a large number of sanctions and changes in the case plan for offenders (English et al., 2000; Gannon et al., 2014; Tubman-Carbone, 2009).

Limitations: Impact on Disclosure

Polygraph disclosure research undertaken to date has been based on relatively small sample sizes. There also was no corroboration of the disclosures made, allowing for the possibility of false admissions and an overstating of the number of victims. However, many polygraph disclosure studies also note that, given the deceptive polygraph results, there is also a possibility that the true incidence of offending behavior is underreported. This makes the interpretation of disclosure research findings difficult. Perhaps most importantly, most of the disclosure studies lacked comparison groups so it is not possible to know with certainty that the polygraph was responsible for the new disclosures. Further, many of these studies are limited to one state or jurisdiction, with only one study encompassing four states, raising questions about the generalizability of findings to other jurisdictions. Finally, the fact that the polygraph was voluntary in one study suggests the possibility that the results may have been different had all offenders completed the assessment. Future research on polygraph disclosures is clearly needed and it should include matched comparison groups and larger samples. Disclosure studies spanning multiple jurisdictions are also needed.

Impact on Sexual Offender Recidivism

As noted in the "Specialized Supervision" section above, the research results for sexual offenders subject to polygraph testing as part of the containment approach typically demonstrated lower levels of recidivism than sexual offenders not subject to this intervention (Aytes et al., 2001; Lowden et al., 2003). However, in a study conducted by McGrath and colleagues (2007), no significant differences in sexual recidivism between polygraphed and nonpolygraphed sex offenders were found. In that study, the recidivism rates of 104 sex offenders subject to polygraph testing were compared with those of a group of 104 matched sex offenders not subject to polygraph testing. The recidivism rate based on sexual recidivism charges was 5.8 percent for the polygraph group and 6.7 percent for the nonpolygraph group, a difference that is not statistically significant. However, there was a significant difference between the two groups in violent recidivism (operationalized as a new violent crime charge). The violent recidivism rate for those offenders subject to polygraph was 2.9 percent, compared to 11.5 percent for the nonpolygraph group.¹¹ It should also be noted that a large percentage of high-risk behaviors were disclosed during the polygraph examinations (McGrath et al., 2007).

Limitations: Impact on Sexual Offender Recidivism

The limitations cited for the specialized supervision research, and in particular the containment approach, hold for the polygraph research as well. Indeed, the only study that specifically looked at recidivism related to sexual offenders subject to polygraph, compared to those who were not, showed no significant difference in the rate of sexual recidivism. However, this study acknowledged that several issues may have confounded the study results, including the small sample size, potential selection bias (in that probation officers decided who would take the polygraph) and the infrequency of polygraph testing. Hence, the study conducted by McGrath and colleagues (2007) should be replicated using a larger sample size, matched comparison groups and program features that reduce the probability of selection bias and maintain the integrity of the polygraph treatment. Studies examining the impact of polygraph testing on recidivism in different jurisdictions are also needed.

Impact on Supervision Professionals

In a 1998 telephone survey of probation and parole supervisors (N = 679), approximately three-fourths believed that polygraph use enhanced disclosure of offender behavior and two-thirds believed it led to better supervision of offenders (Cooley-Towell, Pasini-Hill & Patrick, 2000). In another survey, more than 80 percent of the offender manager respondents reported that the results of a specific polygraph examination were a useful tool for offender supervision (Gannon et al., 2014). Other surveys of service providers have found similar positive results. For example, in one survey, 96 percent of the respondents reported that the polygraph was helpful (McGrath et al., 2007). In another survey, 100 percent of the providers (n = 11) and 90 percent of the parole officers (n = 105) who responded reported that the polygraph was helpful. In the same survey, 80 percent of the providers who responded reported that having one group member take a polygraph test positively impacted other group members (Tubman-Carbone, 2009).

Impact on Sexual Offenders

Research on the perceived impact of the polygraph by sex offenders themselves is extremely limited. One study that examined this was conducted by Kokish, Levenson and Blasingame (2005). The study surveyed 95 sexual offenders and found that 72 percent of those surveyed rated the polygraph as helpful, while 11 percent said the polygraph was harmful (Kokish, Levenson & Blasingame, 2005).

Limitations: Impact on Sexual Offenders

Most of the limitations commonly found with survey data apply to the above studies. (For more on general limitations of sex offender research, see the "Limitations of the Data" section of Chapter 1, "Incidence and Prevalence of Sexual Offending," in the Adult section.) In addition, the answers provided by sex offenders under supervision may be subject to distortion because offenders may try to give a socially desirable response or portray themselves in a sympathetic light.

Test Validity

One of the significant critiques of the polygraph is that it does not produce valid results. While this chapter only addresses the issue of test validity very briefly, readers are directed to the National Research Council report titled The Polygraph and Lie Detection (2003) for additional information. Its key research findings regarding test validity follow:

• A large-scale review of 57 studies on the use of specific-incident polygraph testing with sexual offenders found that such testing demonstrated the ability to discriminate between truth and deception at a rate well above chance.
• Screening tests (sexual history and maintenance) performed with the polygraph showed less of an ability to discriminate between truth and deception.  

Research Summary

Research suggests that polygraph testing increases offender disclosure across multiple offending or behavior categories, including historical and current offending and high-risk behavior. The empirical evidence also suggests that polygraph testing can help reduce sexual recidivism when used in conjunction with specialized supervision and treatment within the containment approach.

Limitations Summary

One of the key limitations in the polygraph research studies reviewed in this chapter is the inability of the research to distinguish the impact of the polygraph from other strategies (treatment and specialized supervision). Small sample sizes are also a problem and jurisdiction-specific approaches may limit the generalizability of research findings. Future research should employ more rigorous methods to better isolate the impact of polygraph testing on both disclosure and recidivism.

Polygraph Summary

Polygraphs should be used as one component of an overall sex offender management strategy. Until more definitive research regarding the validity and impact of polygraph testing is available, the polygraph will continue to be a controversial technique used inconsistently in sex offender management schemes. If polygraph testing is used in the management of sex offenders, it should be implemented as one component of an overall sex offender management strategy. Polygraph disclosure information may be useful for assessment of risk factors and identification of treatment needs, but in some jurisdictions such information may not be used for prosecution or supervision revocation. Given the questions that remain about test validity, it is not recommended that polygraph results be relied on exclusively for sex offender management decision-making.

 

Electronic Monitoring, Including Global Positioning Systems

Another recent trend in sex offender management and supervision has been the use of GPS to monitor sex offenders. GPS is an updated, more technologically advanced form of the electronic monitoring techniques used with criminal offenders in the past. These earlier versions of electronic monitoring were much more passive in nature, and they typically involved the use of a radio transmitter device (worn by offenders) that alerted a home-based receiver and a remote monitoring station whenever the offender was out of range. Offenders could never be tracked or otherwise located once they left their homes. In a significant technological advance, GPS provides real-time tracking of and location data for the offender, and it also is capable of notifying authorities if an offender enters a prohibited area, such as an offender exclusion zone or victim residence. Monitoring by GPS can be either active (viewing an offender’s movement between locations in real time) or passive (data are saved and reviewed later, and notification is only done electronically based on restriction parameters violated). In the United States —

• Six states use lifetime electronic monitoring (Nieto & Jung, 2006).
• Forty-seven states have some form of electronic monitoring legislation, 19 of which require the use of an electronic monitoring tool. The remaining 28 states permit but do not require electronic monitoring (Button, DeMichele & Payne, 2009).
• Thirty-one states introduced electronic monitoring legislation in 2007, with 14 of 109 bills passing (CSG, 2010).

Research

In a systematic review of 12 studies examining the effectiveness of non-GPS electronic monitoring used with criminal offenders overall (not necessarily sex offenders), WSIPP found no significant reduction in criminal recidivism for offenders subject to electronic monitoring techniques (Aos, Miller & Drake, 2006). However, a second large study that examined the effectiveness of passive monitoring devices and GPS used with criminal offenders subject to home confinement in Florida between 1998 and 2002 did find promising results (N = 75,661). Study findings indicated that criminal offenders placed on both passive electronic monitoring devices and GPS had significantly lower levels of revocation for a new criminal offense or for absconding than did offenders subject to home confinement without such monitoring¹² (Padgett, Bales & Blomberg, 2006).

In a study of non-GPS passive electronic monitoring specific to sexual offenders in three of four Canadian provinces that use such a technique, Bonta, Wallace-Capretta and Rooney (2000) compared the recidivism rates of a group of inmates and probationers who were not subject to electronic monitoring with those of a group of offenders who were subject to it. Based on a one-year follow-up period, those on electronic monitoring sexually recidivated (defined as a sex crime reconviction) at a rate of 26.7 percent, compared to 33.3 percent for the probationers who were not monitored and 37.9 percent for the inmates who were not monitored. However, the researchers noted that although there was a statistically significant difference in recidivism between the electronic monitoring and nonelectronic-monitoring groups,¹³ when the results were controlled for risk there was no difference between them. Hence, they concluded that the observed recidivism reductions were due to offender risk dynamics, not program components (Bonta, Wallace-Capretta & Rooney, 2000).

In a study comparing states that have implemented electronic monitoring laws for sexual offenders with those that have not, Button, DeMichele and Payne (2009) found that the states with such laws were no more likely to have rates of violent crime and rape that were higher than the U.S. average than were states without such laws. Additionally, a second study looked at the rates of forcible rape from 2000–2011 Uniform Crime Report data in 10 states¹⁴ that had passed statewide "Jessica's Laws," which included lifetime GPS monitoring. The results found a marginally significant decrease in rape in Georgia related to implementation of GPS,¹⁵ but no significant reduction in rape related to GPS in any of the other nine states. As a result, the authors concluded that there was a null relationship between lifetime GPS and the rate of forcible rape across the 10 states studied (Dierenfeldt & Varriale Carson, 2014).

State agencies in California, Florida, New Jersey and Tennessee, among others, have studied the use of GPS with sexual offenders. Among sexual offenders on GPS in New Jersey (N = 225), there were 19 nonsexual criminal recidivists or technical violators and one sexual recidivist in a one-year follow-up (New Jersey State Parole Board, 2007). It should be noted there was no comparison group for this study. In a Florida study of 705 offenders on electronic monitoring using predominantly active GPS (70 percent of whom were lower-risk offenders and 30 percent of whom were habitual or sexual offenders), offenders on electronic monitoring had a felony recidivism rate of 2.6 percent, while offenders who were not subject to electronic monitoring recidivated at a rate of 6.6 percent in a one-year follow-up. It is not known whether this difference was statistically significant (Office of Program Policy Analysis & Governmental Accountability, 2005). Conversely, a Tennessee study that compared the outcomes of 493 sex offenders on GPS with those of 370 offenders in the same counties prior to the use of GPS found no significant difference between the two groups in the number of technical violations or new charges that occurred or in the number of days before a first technical violation (Tennessee Board of Probation and Parole, 2007).

A California study compared outcomes for 94 GPS offenders and a group of 91 high-risk offenders who were not on GPS. No significant differences in technical violations (which included offenders who committed a new crime) were found between the two groups (39.6 percent vs. 37.2 percent); however, the GPS group was less likely to abscond. This study also included a process evaluation of GPS that showed equipment problems, signal drift, blocked signals and high caseloads impacted effectiveness (Turner et al., 2007). Finally, in a second California study of high-risk sex offenders (N = 516) (half of whom were on GPS while the other half was a matched non-GPS comparison group), no significant differences in sex crime rearrest (2.7 percent for the GPS compared to 5 percent for the non-GPS group) or reconviction (1.9 percent compared to 4.3 percent) were found based on a one-year follow-up period. However, the GPS group had significantly lower levels of sex-related parole violations (5 percent compared to 12.4 percent),¹⁶ general rearrests (14.4 percent compared to 26.4 percent)¹⁷ and return to custody (58.1 percent compared to 58.9 percent)¹⁸ (Gies et al., 2012).

Limitations

In terms of the limits of the GPS-specific studies, the sample sizes were relatively small (ranging from 94–262), with at least one study referencing primarily a lower risk, nonsexual offender group. The follow-up periods employed in many studies were not of sufficient length. The inability of several studies to detect a positive GPS effect may be related to problems using the technology or staffing limitations within the monitoring program.

The efficacy of electronic monitoring techniques such as GPS cannot be established at this time. Additional studies with sufficient sample sizes and follow-up periods, and matched comparison groups, are needed to test the impact of GPS. Technological and staffing problems within monitoring programs also need to be addressed so that impact evaluations can focus on GPS programs that are implemented and delivered with integrity. Finally, while research on non-GPS electronic monitoring provides important insights about the value of monitoring strategies, the technological differences between passive alert systems (non-GPS and passive GPS¹⁹) and an active monitoring system (active GPS) are significant and must be accounted for when assessing the effectiveness of any specific monitoring technique or electronic monitoring strategy as a whole in any research summary.

Summary

GPS should not be used in isolation and should be a part of an overall sex offender management strategy. While GPS may eventually be found to be effective as one strategy in an overall management approach for sexual offenders, empirical evidence does not at this time establish that the strategy is effective when used in isolation. Policymakers and the public should not view GPS as a viable alternative to empirically supported supervision models that incorporate treatment.

Sexual Offender Civil Commitment

At present, 20 states, the District of Columbia and the federal government have enacted legislation allowing for the establishment of sexual offender civil commitment (SOCC) procedures.²⁰ SOCC is predicated on the belief that some offenders will be at continued high risk (in some cases termed "more likely than not") to commit a new sexual offense if they are not preventively detained and offered treatment designed to lower their risk for recidivism. To be subject to civil commitment, most SOCC statutes require the state to demonstrate that a potential candidate for this measure has 1) a history of engaging in criminal sexual behavior and 2) a “mental abnormality” that, without treatment, would preclude him or her from being able to manage his or her criminal sexual propensities in the community. These “criteria” form the principal basis for SOCC, and persons committed as sexually violent persons/predators (SVPs) are held until such time as a court finds they no longer meet the criteria.

Research

More than 40,000 sexual offenders in Florida have been screened for possible referral to the courts for SOCC proceedings since the law came into effect in early 1999. However, approximately 9 percent of those screened offenders were referred for psychological/psychiatric evaluation and only about 3.5 percent have been referred to court for civil trial. Even fewer of those referred for commitment were actually found to be SVPs — in fact, less than half of those referred for trial (1.5 percent of the total considered) were designated as SVPs. This makes those persons found to be SVPs an “elite” group, at least as far as Florida is concerned (Wilson et al., 2013). Similar numbers have been reported for SOCC programs in other states: Texas (1 percent civilly committed according to Boccaccini and colleagues, 2009); New York (3.7 percent civilly committed according to the New York State Office of Mental Health, 2010); and Minnesota (6 percent civilly committed according to Duwe, 2013).

At present, very few civil commitment programs have released sufficient numbers of offenders to allow researchers to study the impact of civil commitment in a meaningful way. Across the 16 SOCC programs reporting data to the annual survey of the Sexual Offender Civil Commitment Programs Network (Jackson, Travia & Schneider, 2010), the average number of releases per program was less than 10. Further, most releases from civil commitment have occurred recently, meaning that follow-up times would be quite short. As such, very little data currently exist regarding rates of reoffending in SVPs following release to the community.

Studies of Sex Offenders Recommended for Civil Commitment But Not Designated SVP

A number of studies looked at the sexual recidivism rates for sex offenders recommended for civil commitment but not ultimately designated as a SVP. One such study that provides some insight into the impact of civil commitment on postrelease offending examined the reoffense rates of 135 “almost SVPs” (persons who were referred for SOCC, but petitions were not filed with the court) in Washington State (Milloy, 2007). With a uniform follow-up period of six years, 23 percent were convicted of new felony sexual offenses — a rate considerably higher than that found in “routine” samples of sexual offenders. Another study of note compared high-risk/need sexual offenders in a Canadian jurisdiction to SVPs in Florida (Wilson et al., 2013). In that study, the Canadian and American offenders were virtually identical on pertinent risk assessment and clinical factors, and their relative rates of sexual reoffending were also remarkably similar (6.1 percent in 5.48 years compared to 3.2 percent in 2.54 years), although the Florida SVP sample size was small (n = 31) and the follow-up period for the Florida SVPs was relatively short.

In another study involving Florida sex offenders recommended for civil commitment but not ultimately designated SVP, the sexual felony reconviction rates for these offenders was 6.8 percent (for those offenders released for a period between five and 10 years) and 6.5 percent (for those offenders released for a period over 10 years and no more than 14 years) respectively, as compared to 3 and 4 percent respectively, for those offenders not recommended for SVP consideration and released over the same time periods (Carr et al., 2013).

In a study of 102 sex offenders in New Jersey considered for SVP but not actually committed, 10.5 percent sexually recidivated over a 6.5 year follow-up (Calkins Mercado et al., 2013). Additionally, in a study of 1,928 Texas offenders considered for SVP but not actually committed, 3.2 percent sexually recidivated over a 2.25-7.5 year follow-up (Boccaccini et al., 2009). Finally, in a study of 1,722 similar sex offenders in New York, 1.6 percent and 2.9 percent were rearrested for a sex crime after a one- and two-year follow-up, respectively (New State Office of Mental Health, 2010).

---

Studies of Civilly Committed Sexual Offenders Under Community Supervision

In a study of 77 New York sex offenders civilly committed via placement under Strict and Intensive Supervision and Treatment (SIST) in the community, 10 percent violated for sexually inappropriate behavior, with 2.6 percent receiving a new sex crime charge (New York State Office of Mental Health, 2010). In addition, a study of 78 Virginia SVPs on conditional release since 2003 found a 3.8 percent rate for a new sex offense charge (Joint Legislative Audit and Review Committee, 2013). Further, in a study of 347 Texas SVPS, 170 of whom were under community supervision in Texas, none had been charged with or convicted of a new sexual crime. The report does note that both out-of-state SVPS (n=2) were subsequently convicted of new sex crimes. It should be noted that the follow-up time was unspecified in the report (Council of Sex Offender Treatment, 2014).

Finally, in a study of 105 sexual offenders civilly committed in Minnesota from 2004–06, Duwe (2013) used an actuarial risk assessment instrument (MnSOST-R) to estimate the recidivism of this population if it had not been civilly committed and reported that civil commitment likely reduced recidivism by 12 percent over four year

Limitations

There has not been adequate empirical study to determine the effectiveness of SOCC as a sex offender management strategy at this time, at least in terms of the impact of SOCC on postrelease offending. The limited number of sexual offenders released from SOCC, the short follow-up periods researchers would inherently have to use, and the lack of adequate comparison groups all contribute to a paucity of research on SOCC effectiveness. Far more offenders will have to be released from SOCC, and these offenders will have to spend far more postrelease time in the community, before the impact of SOCC on postrelease offending can be studied in a meaningful way.

Summary

SOCC strategies are being used by 40 percent of states (20 states). While these programs seek to contain and treat the most dangerous sex offenders, they have significant costs; a 2005 survey found that annual per-resident costs ranged from $12,680 to $109,000, and that more than $224 million was spent annually to operate SOCC facilities nationwide (Lieb & Gookin, 2005).

Sex Offender Registration and Notification

Registration was first used in the 1930s with repeat criminal offenders as well as sex offenders. California became the first state to implement sex offender registration in 1947, while Washington became the first state to implement community notification on sex offenders in 1990. The goals of Sex Offender Registration and Notification (SORN) programs have been summarized as deterring offenders from reoffending, giving law enforcement an investigative tool and increasing public protection (CSOM, 1999).

The federal government first implemented a national registration law with the Jacob Wetterling Crimes Against Children and Sexually Violent Offender Registration Act in 1994. A national notification law was enacted with the Megan’s Law amendment to the Wetterling Act in 1996. Subsequently, all 50 states have implemented SORN systems. The federal government repeatedly refined and expanded the scope of SORN via a series of amendments to the Wetterling Act (the Lychner Act in 1996, the Jacob Wetterling Improvements Act in 1998, the Campus Sex Crimes Prevention Act in 2000 and the PROTECT Act in 2003), and then ultimately set forth a new SORN scheme with the passage of the Title I of the Adam Walsh Child Protection and Safety Act of 2006, the Sex Offender Registration and Notification Act (SORNA), which replaced the Wetterling Act and subsequent amendments. SORNA’s requirements and how they differ from the Wetterling Act have been documented in other sources. The changes include enhanced registration requirements and procedures, increased availability of sex offender registration information to the public, strengthened information sharing and enforcement mechanisms and greater federal assistance in operating and upgrading sex offender registration programs, sharing and disseminating sex offender information and enforcing registration requirements.

Research

SORN requirements arguably have been implemented in the absence of empirical evidence regarding their effectiveness. It has been suggested that SORN may be a specific deterrent for sex offenders in that it facilitates awareness, monitoring and apprehension of sex offenders and helps prevent sex offenses — particularly repeat sex offenses — from occurring. While these hypotheses were not empirically tested prior to the implementation of SORN requirements, a significant body of research using various methods has since examined the impact of SORN, particularly in relation to recidivism.

Interrupted Time Series Analysis Studies

One research method used to assess the effectiveness of SORN for adult sexual offenders is interrupted time series analysis, which essentially examines an outcome of interest using many observations before and after the implementation of a specific intervention. Several interrupted time series analyses assessing SORN have been completed in recent years. In one analysis of state SORN laws, Prescott and Rockoff (2011) found that SORN may have contributed to a decrease in sex crimes. More specifically, the study found that sex offender registration led to a decrease in the rate of victimization of nonstrangers and a reduction in recidivism for identified sex offenders. However, community notification did not appear to reduce recidivism for identified sex offenders (Prescott & Rockoff, 2011).

A similar analysis focused on the impact of SORN on rape in 10 states. Using Uniform Crime Report (UCR) data on rapes reported to the police as the outcome measure, the study found that statistically significant reductions in reported rape occurred following the implementation of SORN in three of the 10 states (Hawaii²², Idaho²³ and Ohio²⁴). In six states (Arkansas, Connecticut, Nebraska, Nevada, Oklahoma and West Virginia), no significant change was observed following SORN implementation, and one state (California) actually had a statistically significant increase in sex crimes following SORN implementation. Based on the varied findings, the authors concluded there was no systematic influence of SORN on the rate of reported rape (Walker et al., 2006). (For more on UCR data, see the "Uniform Crime Report" section of Chapter 1, “Incidence and Prevalence of Sexual Offending,” in the Adult section.)

Presently, 41 states have some kind of registration for juveniles adjudicated delinquent of sex offenses; 30 states either permit or require public website posting for those juveniles, and that the vast majority require registration and public notification for juveniles transferred for trial and convicted as an adult.²⁶ In reviewing UCR sex crime arrest data from 47 states for 1994 through 2009, Holmes (2009) did not find a statistically significant decrease in the rate of sex crime arrest in either juvenile registration states or juvenile notification states (post-SORN).

Several studies have examined the impact of SORN in individual states. For example, in South Carolina, adult sex crimes were compared to nonsexual assault and robbery crimes pre- and post-SORN implementation (N = 194,575, of which 19,060 were sex crime arrests). Data were examined for 1990 through 2005. SORN implementation occurred in 1995. The study found that the sex crime rate declined by 11 percent²⁷ from pre- to post-SORN while the rates of assault and robbery did not, suggesting the possibility that SORN was a deterrent to sex crimes (Letourneau, Levenson, Bandyopadhyay, Armstrong & Sinha, 2010). In another study from New Jersey, a downward trend in the sex assault rate was observed both pre- and post-Megan’s Law (SORN), but the rate of decline increased after Megan’s Law was implemented (Veysey, Zgoba & Dalessandro, 2008).

A number of state studies did not find evidence that SORN implementation positively impacted the rate of sexual offending or recidivism. Interestingly, one of these studies focused on South Carolina, where another study did find evidence of a positive SORN impact (Letourneau, Levenson, Bandyopadhyay, Armstrong and Sinha, 2010).

In the South Carolina study that did not find evidence of a positive SORN effect, recidivism was examined in the context of registration status for 6,064 male offenders convicted of at least one sex crime in that state between 1990 and 2004. The study found that registration status did not predict recidivism (Letourneau, Levenson, Bandyopadhyay, Sinha & Armstrong, 2010). Another state study taking place in New York analyzed sex crime, assault, robbery, burglary and larceny arrests from 1986 through 2006. Study results indicated that the implementation of the state’s sex offender registry did not decrease the rearrest rate for convicted sex offenders, deter nonregistered offenders from offending or decrease the overall rate of sex crimes. It was also noted that 94.1 percent of child molestation arrests were for first-time sex offenders (Sandler, Freeman & Socia, 2008). Finally, an analysis that focused on South Carolina juveniles who committed sexual offenses between 1990 and 2004 (N = 1275) found that 7.5 percent were charged with a new sex offense and 2.5 percent were adjudicated for a new sex offense during a nine-year follow-up period (Letourneau, Bandyopadhyay, Sinha & Armstrong, 2010). More importantly, the researchers found that registration was not associated with recidivism; however, nonsexual, nonassault recidivism (defined as a new charge) significantly decreased for those on the registry²⁸ (Letourneau, Bandyopadhyay, Sinha & Armstrong, 2010).

Limitations: Interrupted Time Series Analysis Studies

One of the primary limitations of the studies cited above is that time series analysis and before/after methods in general are not as capable of isolating intervention effects as a randomized controlled trial. While an interrupted time series analysis based on a sufficient number of observations can produce highly trustworthy findings, outside factors such as changes in supervision, treatment and other sex offender management practices pre- and post-SORN may also be influencing study results. Further, the authors in the New Jersey study cautioned that wide variety across county sex crime rates was noted, and the analysis did not uniformly and consistently demonstrate downward trends, suggesting that the statewide pattern identified might represent a spurious effect and be an aggregation artifact (Veysey, Zgoba & Dalessandro, 2008). Finally, other variables such as sex crime underreporting (which could be aggravated by SORN due to the unwillingness of intrafamilial victims to report because of fears about SORN) and the limitations of official sex crime statistics may be confounding these results.

Studies Employing a Comparison Group

A number of studies have examined the impact of SORN by comparing the outcomes of sex offenders subject to SORN with those not subject to this strategy. These studies have generally produced mixed results.

One study finding a positive effect examined the recidivism of 8,359 sexual offenders in Washington State. Some of those offenders were subject to SORN, while others were not because SORN requirements were not yet in place. The study found that the sex offenders subject to SORN sexually recidivated (defined as a new Washington state conviction for a felony sex crime) at a 2 percent rate, while the pre-SORN group recidivated at a 7 percent rate²⁹ (WSIPP, 2005). Another study finding a positive impact took place in Minnesota. Researchers compared Level III sexual offenders subject to community notification between 1997 and 2002 (n =155) with precommunity notification sexual offenders retrospectively scored as Level III offenders (n = 125), and Level I and II sexual offenders not subject to community notification (n = 155).³⁰ Based on a three-year follow-up period, the community notification group had a statistically significantly lower sexual recidivism rate based on reconviction (3.2 percent), compared to the prenotification group and nonnotification group (32.8 percent and 9.6 percent, respectively)³¹ (Duwe & Donnay, 2008).

On the other hand, several state-level studies have not found evidence of a positive SORN effect. For example, in an Iowa study, a group of sex offenders subject to registry requirement (n = 233) who were also under legal supervision were compared to a matched group of preregistry sex offenders not under supervision (n = 201). In a 4.3-year follow-up, the registry group sexually recidivated (defined as a new sex crime conviction) at a rate of 3 percent, compared to the nonregistry group’s 3.5-percent recidivism rate. This difference was not statistically significant. However, when the recidivism rates of parolees and probationers were compared, the researchers found that registration requirements may have had more of an impact on parolees (Adkins, Huff & Stageberg, 2000).

In New Jersey, researchers compared the recidivism rates of offenders subject to SORN with those of offenders who were not subject to this strategy (n = 550). Based on a 6.5-year follow-up period, offenders subject to SORN recidivated at a rate of 7 percent, compared to 11 percent for offenders who were not subject to SORN; however, these differences were not found to be statistically significant (Zgoba & Bachar, 2009; Zgoba et al., 2008).

In Wisconsin, the recidivism rates of sex offenders subject to registration and extensive notification between 1997 and 1999 (n = 47) were compared with those of sex offenders who had limited notification requirements (n = 166). No statistically significant differences in sex crime rearrest rates over a four-year follow-up period were found, as 19 percent of the extensive notification group sexually recidivated, compared to 12 percent for the limited notification group (Zevitz, 2006). Similar findings were reported in a Washington State study.

Limitations: Studies Employing a Comparison Group

Research on SORN as it relates to offender recidivism has produced mixed results.

The primary limitation of the studies described above is the inability to control for all outside factors and to isolate the effects of SORN requirements on recidivism.

Survey Data

Surveys of stakeholders can provide descriptive data about the impact of SORN on different populations, including the public, sexual offenders and supervision officers.

Impact on the Public

One multistate study (n = 115 from 15 states) of community members found general familiarity with and support for SORN, along with a belief that it prevents offending (Schiavone & Jeglic, 2009). State-level surveys of community members regarding SORN in Florida, Nebraska, Washington and Wisconsin found that the public —

• Was aware of and supported SORN (Anderson & Sample, 2008; Lieb & Nunlist, 2008).
• Thought it was fair (Brannon et al., 2007).
• Believed that it provides safety for their family (Anderson & Sample, 2008; Lieb & Nunlist, 2008; Zevitz & Farkas, 2000a).
• Thought it makes sex offenders follow the law (Phillips, 1998, as cited in CSOM, 2001; Lieb & Nunlist, 2008; Brannon et al., 2007).
• See the benefits of SORN and learning about sex offenders through SORN (Phillips, 1998, as cited in CSOM, 2001; Lieb & Nunlist, 2008).
• Took preventive measures (38 percent)³² based on SORN information (Anderson & Sample, 2008).
• Reported suspicious behavior of offenders (3 percent)³³ (Lieb & Nunlist, 2008).
• Accessed the registry (31 percent),³⁴ but those who did were more likely to be female, to be affluent and to have children (Sample, Evans & Anderson, 2011).

Impact on Offenders

Survey responses indicate that SORN has both negative and positive impact on offenders and that the public is generally supportive of SORN as promoting public safety.

In a review of eight individual surveys on SORN’s impact on sexual offenders subject to it,³⁵ Lasher and McGrath (2012) found that —

• 8 percent of sex offenders reported physical assault or injury
• 14 percent reported property damage
• 20 percent reported being threatened or harassed
• 30 percent reported job loss
• 19 percent reported loss of housing
• 16 percent reported a family member or roommate being harassed or assaulted
• 40–60 percent reported negative psychological consequences

However, more than one-third of adult sex offenders reported communities being safer and approximately three-fourths felt it was a deterrent to offending (Lasher & McGrath, 2012).

A number of studies involving surveys of sexual offenders in states across the country indicate that SORN requirements have a range of negative impacts on sexual offenders. These include negative impacts on sex offenders’ jobs, housing, friends and family (Ackerman, 2009; Levenson, D’Amora & Hern, 2007; Tewksbury, 2004; Vandiver, Dial & Worley, 2008), which results in stress, isolation, loss of hope and shame/embarrassment (Levenson & Cotter, 2005a), and the greater likelihood of living in disadvantaged neighborhoods where services are less available (Hughes & Kadleck, 2008). Studies have also found that 10–13 percent of sex offenders report experiencing violence (Brannon et al., 2007; Levenson, D’Amora & Hern, 2007) and harassment (Vandiver, Dial & Worley, 2008; CSOM, 2001). While many sexual offenders report the belief that SORN would not deter reoffending and was unfair punishment (Ackerman, 2009; Brannon et al., 2007; Levenson, D’Amora & Hern, 2007; Tewksbury & Lees, 2007; Tewksbury, 2004), many also report that SORN requirements motivate them to be successful (Levenson, D’Amora & Hern, 2007; Levenson & Cotter, 2005a).

Impact on Criminal Justice and Law Enforcement Professionals

In a survey of probation and parole officers (n = 77), respondents reported they generally believed community notification served an appropriate goal but had a high cost for corrections in terms of personnel, time and money. They also believed it made sex offender housing difficult to locate (Zevitz & Farkas, 2000b).

In a second survey of criminal justice professionals (N = 1,157), law enforcement officers and prosecutors were more supportive of SORN and residence restrictions, while other corrections professionals (e.g., wardens, parole board members and community corrections personnel) were less supportive. In addition, higher levels of education and experience was inversely related to support for such policies, in that the most experienced and educated personnel were less supportive than their less experienced and educated counterparts (Mustaine et al., 2015).³⁷

Finally, in a survey of law enforcement professionals (N = 1,247) from 49 states and the District of Columbia, including command staff (35 percent), supervisory staff (30 percent) and line sworn and civilian staff (35 percent), the primary purposes and effectiveness of SORN were identified as monitoring offenders and information sharing, and to a lesser extent informing the public and criminal investigation. Deterrence of future sexual offending by registrants was not typically identified as a goal or outcome of SORN. Concerns identified by respondents included public misunderstanding of and inaction as a result of SORN, which provides a false sense of security. In addition, incomplete registry information including the lack of accurate risk information on registrants, the lack of integrated data with other information systems, transient and homeless registrants and lack of court supervision were identified as needs. Law enforcement recommendations included enhanced penalties and prosecution for registry noncompliance and the need for additional resources (Harris et al., 2015).

Impact on Sexual Assault Survivors

In a survey of sexual assault survivors (n = 598), respondents reported support for SORN to be applied to all sexual offenders regardless of the relationship to the victim and did not believe that it impacted reporting by survivors to law enforcement. However, survey results provided less support for SORN as a mechanism to provide for public safety or deter future sexual offending, and indicated concerns related to SORN providing a false sense of security (Craun & Simmons, 2012).

Limitations: Survey Data

The limitations of survey data have previously been identified and are applicable here.

Impact of Failure To Register

Several studies have examined whether sex offenders who fail to comply with registration requirements are more likely to recidivate than offenders who do comply. For example, a Washington State Institute for Public Policy study (2006) found higher recidivism for noncomplying sex offenders compared to their registration-compliant counterparts. Noncomplying sex offenders had a felony sex crime conviction recidivism rate of 4.3 percent, while complying sex offenders had a rate of 2.8 percent. It is unknown whether this difference was statistically significant (WSIPP, 2006). Studies in Minnesota, New Jersey and South Carolina, however, failed to find any significant differences in recidivism between registration-compliant and noncompliant sex offenders. In Minnesota, Duwe and Donnay (2010) compared the recidivism rates of 170 sex offenders who had a failure-to-register charge between 2000 and 2004 with those of 170 nonfailure-to-register sex offenders and found that the noncompliant sex offenders were no more likely to sexually recidivate (defined as a new sex crime arrest or conviction) (Duwe & Donnay, 2010). Similarly, a study focused on sex offenders in South Carolina (N = 2,970) found that those who failed to register were no more likely to sexually recidivate (11 percent) than those not so charged (9 percent) (Levenson et al., 2009). Finally, in a study of New Jersey sex offenders (N = 1,125), 644 of whom failed to register and 481 who did register, researchers again found no significant difference between the two groups in terms of their sexual rearrest rates (18 percent for the failure-to-register group compared to 11 percent for the registering group) (Zgoba & Levenson, 2012).

Limitations: Impact of Failure To Register

Relatively few studies have examined whether noncompliant offenders are more likely to reoffend than compliant offenders, and the studies again suffer from the low base rate for sexual recidivism and limited generalizability.

Accuracy Research

A number of studies have examined the accuracy of sex offender registries. For example, Hughes and Kadleck (2008) reviewed the accuracy of sex offender registries in Nebraska and Oklahoma and found that approximately 90 percent of the Nebraska records were accurate (n = 975), while 56.5 percent of the Oklahoma records were accurate (n = 5,163). In a random sample of New York registry records (n = 200), 37 percent of the records were found to be inaccurate, including 27 percent that did not match driver’s license information and 2.5 percent that had wrong addresses (Office of the New York State Comptroller, 2006). Finally, in a Vermont study of sex offender registry records (n = 57), 75 percent of the records were found to have critical or significant errors (Vermont State Auditor, 2010).

Limitations: Accuracy Research

Audits of sex offender registry records provide important insights about the accuracy and reliability of sex offender registries. The major limitations of these studies are that they often are based on small sample sizes and their generalizability to other jurisdictions remains unknown.

Summary

In summary, research on the effectiveness of SORN remains relatively limited and findings from the studies are somewhat inconclusive. Findings from time series studies are mixed. Some studies find lower rates of sex crimes following SORN implementation, while others do not. Studies based on a comparison of outcomes for sex offenders subject and not subject to SORN also produced mixed findings. An arguable lack of sufficient scientific rigor may further cloud the import of studies in this area. Therefore, the results of SORN research undertaken to date continue to leave open questions about the effects of registration and community notification requirements. Finally, few if any studies to date have examined the multifaceted elements of registration laws, generally, or SORNA, specifically which incorporates requirements, procedures and information sharing and enforcement mechanisms going beyond those prevalent in SORN programs examined in past studies. Research findings on the effectiveness of SORN are mixed, and more high-quality studies with sufficient scientific rigor are needed. Sexual offenders in survey responses claim a range of negative impacts from SORN; however, many see it as a deterrent to committing future crimes. Further research is clearly needed to corroborate these survey findings. Surveys of community members indicate that the public is familiar with SORN laws, and also that they are generally supportive of SORN. Finally, registry accuracy studies have found significant problems with registry records in some states. The need for accurate registry information was recently highlighted by the sex offender management experts who participated in the 2012 SOMAPI forum. Clearly, additional research is needed to help better answer questions about SORN effectiveness and about which aspects of the policy may be beneficial and cost-effective and, conversely, which may not. Given the limitations of existing research regarding SORN, the SOMAPI forum participants recommended that future changes to SORN be studied prior to enactment, particularly in the context of existing knowledge about sexual offender risk and recidivism. Pilot testing prior to full-scale implementation provides one mechanism for examining potential impacts, both positive and negative.

Residence Restrictions

Sex offender residence restrictions that limit where convicted sex offenders may legally live have become more popular across the country. These restrictions typically prevent sex offenders from living within 1,000 to 2,500 feet of schools, daycare centers and other places where children congregate. The first states to adopt residence restrictions were Delaware and Florida in 1995. Currently, 30 states and many more municipalities have residence restriction laws, some in accordance with Jessica’s laws (Meloy, Miller & Curtis, 2008). As with many other sex offender management strategies implemented across the United States, there was no research evidence to support the effectiveness of residence restrictions prior to the enactment of this policy. However, empirical evidence questioning the effectiveness of residence restrictions is becoming available.

Outcome Data

Several studies have looked at sexual offender recidivists to determine whether living in proximity to places where children congregate was a risk factor and whether residence restrictions would have deterred reoffense. In one study commissioned by the Colorado legislature for the purpose of studying the potential impact of residence restrictions prior to implementation (a recommended practice), no significant difference in recidivism (defined as any new criminal conviction) patterns was found based on whether or not an offender lived in proximity to schools and daycare centers (Colorado Department of Public Safety, 2004). In a study of sex offenders subject to residence restrictions in Florida (n = 165), researchers found no significant difference in the distance recidivists (defined as a new sex crime rearrest) and nonrecidivists lived in proximity to schools and daycare centers (Zandbergen, Levenson & Hart, 2010).

In Jacksonville, Florida, researchers investigated the effects of a 2,500-foot residence restriction ordinance on sexual recidivism (defined as a new sex crime arrest) and sex crime arrest rates. No significant differences in recidivism were found pre- and post-policy implementation. Similarly, there was no significant difference in sex crime arrest rates pre- and post-policy implementation. The authors concluded that the residence restriction ordinance did not reduce recidivism or deter sex crimes (Nobles, Levenson & Youstin, 2012).

In a study of county and local residence restrictions in New York (N = 8,928 cases; 144 months of data from each of 62 New York counties), researchers found no significant impact on sexual recidivism against child or adult victims or on arrests for sex crimes against child victims. However, there was a 10 percent decrease in the rate of arrests for sex crimes against adult victims.³⁸ As a result, the researchers concluded that residence restrictions do not appear to deter sexual recidivism or sex crime arrests where the victim was a child, but they may deter sex crimes involving adult victims (Socia, 2012).

The Iowa Department of Criminal and Juvenile Justice Planning studied the effect of Iowa's 2,000-foot residence restriction law, which was implemented in August 2005. The number of charges for sexual assaults involving minor victims was examined for both the 12-month period preceding the law’s implementation and the 24-month period after the law went into effect. The study found no significant downward trend in the number of charges following passage of the law. In fact, sex crime arrests increased steadily over each of the three years (913, 928 and 1,095) of the study (Blood, Watson & Stageberg, 2008).

In a study of the impact of residence restrictions on sex offenders in Michigan and Missouri, the results indicated a decline, but not statistically significant, in sex offenders living in restricted areas following implementation of the restriction (Michigan: 23 percent lived within a restricted area prior to the law, while 22 percent lived in such an area post-law; Missouri: 26 percent lived within a restricted area prior to the law, while 21 percent lived in such an area post-law). This suggests there was not a significant displacement of sex offenders to rural areas as has been found in other studies, although those that did live within a restricted area were more likely to be in a disadvantaged area. In addition, there was no significant difference found in general recidivism after the law was enacted in either state, although recidivists in Michigan were reconvicted more quickly following implementation of the law (the same did not hold true for Missouri). Finally, the study noted that sexual recidivism outcomes for those subject to restrictions could not be analyzed due to the low base rate in the overall sample (less than 3 percent) (Huebner et al., 2014).

Socia (2015) completed a review of Uniform Crime Report data on forcible rape over a 19-year period from 49 states (excluding Illinois) and the District of Columbia. The results indicated that rate of forcible rape was 13 percent higher in states having residence restrictions than those that did not have such restrictions.

One of the more comprehensive studies of residence restrictions occurred in Minnesota. The researchers examined the characteristics of recidivism events for 224 sex offenders who committed a new sex crime and were reincarcerated between 1990 and 2002. The researchers found that 79 percent of these offenders knew the victim prior to the reoffense. Moreover, 85 percent of the reoffenses studied occurred in a residential location and 39 percent occurred outside the home, with 9 percent taking place within 1 mile of the offender’s house. Of these 9 percent, three offenders contacted a victim at a restricted location; two of the offenders were not in proximity to where they lived and the third contacted an adult victim. The researchers concluded that none of the reoffenses would have been deterred by residence restrictions (Duwe, Donnay & Tewksbury, 2008). It is interesting to note that in Minnesota, the Department of Corrections raised concerns about the unintended negative consequences of residence restrictions, including the potential for sex offenders to congregate in rural areas without ties to the community, thereby resulting in social isolation; a lack of work, education and treatment; and being farther away from supervision (Minnesota Department of Corrections, 2003).

Finally, in a convenience sample study of sex offenders randomly selected after being released from prison between 1996 and 2006 (n = 293 child molesters and 112 rapists), researchers found that 76.5 percent of the offenders met their victim in a private location and only 6.8 percent met a victim in proximity to a residence restriction setting. Additionally, 82.2 percent of offenses occurred in a private setting and 9.1 percent of victims were strangers to the offender, with 18.8 percent of rapists and 14.7 percent of child molesters meeting the victim in a public location. Based on this analysis, the researchers suggested that social rather than geographic proximity influenced offending (Columbino, Mercado & Jeglic, 2009).

Limitations: Outcome Data

Limitations of residence restriction outcome studies are similar to those previously identified for other research, including small sample sizes, short follow-up periods, low sexual recidivism rates and the inability of most studies to isolate the impact of residence restrictions from other influences.

Survey Data

Impact on the Public

A number of researchers have studied the impact of residence restrictions on where sex offenders reside in the community. In a Chicago study (N = approximately 4,000), researchers found that sex offenders were more likely to live in disadvantaged neighborhoods (30 percent of sex offenders lived in these areas, which is nearly 5.5 times greater than the number living in affluent areas). In a study of those sex offenders violating the residence restriction law (n = 1,008), 29 percent lived in a disadvantaged neighborhood and 2 percent lived in an affluent neighborhood. Finally, it was noted that 70 percent of the disadvantaged area was off limits to sex offenders, compared to 32 percent of affluent areas. The research suggests that residence restrictions lead to a disproportionate number of sex offenders living in disadvantaged neighborhoods (Hughes & Burchfield, 2008). Similar results concerning the disproportionate impact of residence restrictions have been found in other studies. A Minnesota Department of Corrections study found that more offenders would be relegated to rural areas as a result of residence restrictions (Minnesota Department of Corrections, 2003). In a New Jersey study of three different areas (rural Alpha and Phillipsburg, urban Newark and suburban Bergen County), researchers found that half of the rural area, 93 percent of Newark and 66 percent of Bergen County would be restricted (Mandelstam & Mulford, 2008).

Impact on Offenders

Sex offenders have had to relocate due to the passage of residence restriction laws in states and municipalities (Barnes et al., 2009; Chajewski & Mercado, 2008; Levenson & Cotter, 2005b; Tewksbury & Zgoba, 2010). Residence restrictions further limit the housing options in urban areas (Barnes et al., 2009; Chajewski & Mercado, 2008; Page et al., 2012; Levenson, 2008). This combination led to a report of increased homelessness (Levenson, 2008), loss of family support and financial hardship (Page et al., 2012; Levenson & Cotter, 2005b).

In a study of nonincarcerated registered sex offenders in Florida (n = 23,523), the rate of failure to register (FTR) was approximately 11 percent with fewer than 2 percent having multiple failures to register and about 3 percent listed as having absconded. In addition, 3.2 percent of registrants were identified as transient and the majority of those (84 percent) were no longer under probationary supervision. The study indicated that larger number of transients were found in counties that had more residence restrictions covering larger areas with higher population density and housing costs (Levenson et al., 2015).

Impact on Criminal Justice Professionals

For more information related to the viewpoints of criminal justice professionals on residence restrictions, see Mustaine et al. (2015) in the SORN section.

Limitations: Survey Data

Limitations of survey data have previously been highlighted.

Summary

In summary, there is no empirical support for the effectiveness of residence restrictions. In fact, a number of negative unintended consequences have been empirically identified, including loss of housing, loss of support systems and financial hardship that may aggravate rather than mitigate offender risk. In addition, residence restrictions lead to the displacement and clustering of sex offenders into other areas, particularly rural areas. Given the above, expansion of this policy was not recommended by the group of sex offender management professionals attending the SOMAPI forum.

Summary

This chapter has focused on the effectiveness of a number of prominent sex offender management strategies, including specialized supervision, COSA, polygraph, GPS, civil commitment, SORN and residence restrictions. Specialized supervision, in conjunction with rehabilitation, appears to be effective in reducing recidivism for sexual offenders. However, the use of specialized supervision in the absence of rehabilitation is not supported by research. The few single studies of COSA that have been undertaken thus far have produced encouraging findings, but far more research employing larger samples of offenders and more rigorous designs capable of isolating COSA effects are needed. Nevertheless, given COSA’s ability to facilitate collaboration with members of the community, the SOMAPI forum experts recommend COSA as a sex offender management strategy. Research related to the use of polygraph assessment is somewhat less definitive. Therefore, the polygraph, if used, should only be used in conjunction with a comprehensive supervision and treatment approach.

In terms of SORN, research to date has exhibited mixed results on sex offender crime rates and recidivism. Studies have not adequately controlled for outside factors that might serve as an alternative explanation for the observed study outcomes. Future, more rigorous research on the effects of SORN is needed. Despite these limitations, there is broad public and policymaker support for SORN and a perceived public safety benefit among these groups.

Finally, the evidence is fairly clear that residence restrictions are not effective. In fact, the research suggests that residence restrictions may actually increase offender risk by undermining offender stability and the ability of the offender to obtain housing, work and family support. There is nothing to suggest this policy should be used at this time.

Sex offender management policies are often implemented on a one-size-fits-all basis for all sexual offenders. The merits of using targeted rather than one-size-fits-all strategies were recently acknowledged by participants in the 2012 SOMAPI forum. The SOMAPI forum experts recommend implementation of all of the above-noted policies that show a positive impact, with the caveat that the use of any strategy should always be commensurate with offender risk and need.

Future Directions

The SOMAPI forum experts recommend that sex offender management policymakers strive to use empirically supported strategies. Granted, there are times when new strategies are identified in the absence of research and need to be tested for effectiveness, as innovation in criminal justice practice, including sex offender management, is important. Therefore, it is recommended that future implemented policies should be evidence-generating.

Sex offender management continues to be a priority for the public, policymakers and professionals. Using research to identify what does and does not work can help ensure that the best possible strategies for protecting the public and reducing victimization are in place. Only through objective, systematic study can we definitively know what is and what is not effective. As Patty Wetterling, in whose son’s memory the first SORN system was developed at the federal level, has observed, “People want a silver bullet that will protect children, but there is no silver bullet. There is no simple cure to the very complex problem of sexual violence” (Human Rights Watch, 2007).

Notes

1 While sex offender management has also taken on increasing importance in other countries, the focus of this chapter is on sex offender management strategies in the United States. 2 For an in-depth review, see Levenson and D’Amora (2007). 3 A study consisting of random assignment to either the intervention group or a comparison group, and comparing the outcomes for the two groups. 4 p < .41 for the Linn County sexual offenders and p < .01 for the Jackson County nonsexual offenders. 5 p < .01. 6 p < .001. 7 p < .05. 8 Ibid. 9 p < .05 10 p < .05 11 p Ibid. 12 p < .001. 13 p < .05 14 California, Florida, Georgia, Kansas, Louisiana, Missouri, North Carolina, Oregon, Rhode Island and Wisconsin. 15 p < .10 16 p < .001. 17 Ibid. 18 p < .05. 19 Passive GPS does not allow movement to be viewed in real time (active GPS) and must be downloaded from the device to a computer. 20 Dr. Robin Wilson provided assistance with the development of this section. 21 See, for example, U.S. Dep’t of Justice, The National Guidelines for Sex Offender Registration and Notification, 73 Fed. Reg. 38029, 38044-45, 38047, 38058-61, 38069-70 (July 2, 2008); U.S. Dep’t of Justice, Supplemental Guidelines for Sex Offender Registration and Notification, 76 Fed. Reg. 1630, 1636-38 (Jan. 11, 2011); Harris and Lobanov-Rostovsky, 2010. 22 p < .1. 23 p < .05. 24 Ibid. 25 Ibid. 26 For further details as to each state’s treatment of juveniles adjudicated delinquent of sex offenses and their corresponding registration responsibilities and notification requirements, see ALA. CODE §15-20A-28 (2014), ARIZ. REV. STAT. ANN. § 13-3821(D) (2014), ARK. CODE ANN. § 9-27-356 (2014), CAL. PENAL CODE §290.008(a) (2014), COLO. REV. STAT. §16-22-102(3) (2013), DEL. CODE ANN. tit. 11, §§ 4121(a)(4)(b) & 4123 (2014), FLA. STAT. § 943.0435(a)(1)(d) (2014), IDAHO CODE ANN. § 18-8403 (2014), 730 ILL. COMP. STAT. 150/3-5 (2014), IND. CODE ANN. §11-8-8-4.5(b) (2014), IOWA CODE § 692A.103 (2013), KAN. STAT. ANN. § 22-4902(b)(2) (2013), LA. REV. STAT. ANN. §15:542 (2013), MD. CODE ANN., CRIM. PROC. § 11-704.1 (2014), MASS. GEN. LAWS. ANN. ch. 6, §178K (2014), MICH. COMP. LAWS. SERV. §28.722 (2014), MINN. STAT. §243.166 (2014), MISS. CODE ANN. §45-33-25 (2013), MO. REV. STAT. §§ 211.425 & 589.400 (2014), MONT. CODE ANN. § 46-23-502 (2013), NEB. REV. STAT. ANN. § 29-4003 (2013) (only juveniles relocating from out of state with pre-existing registration requirements are required to register, https://sor.nebraska.gov/FAQ), NEV. REV. STAT. ANN. § 179D.095 (2014), N.H. REV. STAT. ANN. 651-B:1(XI), N.J. REV. STAT. § 2C:7-2 (2014), N.M. STAT. ANN. §29-11A-5.1 (2013), N.C. GEN. STAT. §14-208.26 (2014), N.D. CENT. CODE §12.1-32-15 (2013), OHIO REV. CODE ANN. §2950.01 (2014), OKLA. STAT. tit. 10A, §2-8-102 (2013), OR. REV. STAT. §181.823 & §181.609 (2013), 42 PA. CONS. STAT. §9799.12 (2014), R.I. GEN. LAWS § 11-37.1-2(c)(4) (2014), S.C. CODE ANN. §23-3-430(C) (2013), S.D. CODIFIED LAWS § 22-24B-2 (2014), TENN. CODE ANN. §40-39-202(28) (2014), TEX. CODE CRIM. PROC. ANN. art 62.001 & 62.351 (2014), UTAH CODE ANN. § 77-41-102(9)(f) (2014), VA. CODE ANN. §9.1-902(G) (2014), WASH. REV. CODE ANN. §9A.44.128 (2013), WIS. STAT. §301.45(1g)(a) (2014), WYO. STAT. ANN. §7-19-301 (2014). 27 p < .0001. 28 p < .05. 29 Differences among study periods are statistically significant beyond the .05 probability level. 30 Level III sex offenders (high public risk) are those who score greater than 7 on the MnSOST-R, a state-based risk assessment instrument, and are subject to broad public notification. On the other hand, Level I (low public risk) and Level II (moderate public risk) sex offenders score less than 4 and between 4 and 7, respectively, and are not subject to broad public notification. 31 p < .01 for the prenotification group and p < .05 for the nonnotification group. 32 The percentage of public members who responded to this item was less than for previous items. 33 Ibid. 34 Ibid. 35 Lasher and McGrath (2012, p. 1) reported in the abstract, “Eight quantitative studies that examined the social and psychological impact of community notification on adult sex offenders (N = 1,583) were reviewed. The pattern of results across studies showed considerable similarities despite marked variability in the populations examined, survey methods used and response rates obtained.” 36 p < .01 37 p < .01 38 p < .01.