Examining sex offender recidivism rates and public policy
After reading the sex offender recidivism section on page 150 in Chapter 9 respond to the following questions.
(1) Should we change the way we treat these offenders?
(2) Why do you think the public clamors for more punishment for these offenders?
(3) What stereotypes or myths does this section highlight about these types of offenders?
Sex Offenders and Repeat Crimes (Recidivism)
Commonly held beliefs among the public and policy makers view all sex offenders as a similar group of deviant offenders who prefer to target strangers and have high rates of repeat offending. Research on sex offender characteristics and behavior found that these beliefs are not true. In reality, sex offenders are neither specially deviant, nor specialists in terms of sticking to sex offenses. Before addressing these facts, it is important to recognize that those sex offenders who are available for study typically include only those who have been caught by criminal justice officials or who have sought treatment. As depicted in Figure 9.1, this limits our knowledge of the larger group of sexual offenders. All sex offenders who fall in the dark section of the largest circle, outside the smaller circles, have not been detected by authorities: they are lost in the “dark field.” As a result, available research is limited to those offenders detected by criminal justice authorities or who are engaged in treatment, either voluntarily or under order by the justice system) (e.g., see Becker, Stinson, Tromp, & Messer, 2003). As Figure 9.1 indicates, most sex offenders are unknown and unstudied.
Notwithstanding the dark field, we do know that convicted sex offenders are actually a highly diverse group of offenders with a multitude of different characteristics, victim preferences, and motivations for offending (Robertiello & Terry, 2007); very few have a psychiatric disorder. Studies of sex offenders in correctional and clinical samples provided a number of different models for explaining sexually deviant conduct. For most offenders, these explanations involve a combination of factors, as opposed to the fictional account of a compulsive offender who cannot resist a pathological urge. While mental illness plays a role in some offending, for the majority of offenders, it does not (See, generally, Strassberg, Eastvold, Kenney, & Suchy, 2012; Marshall, 2007).
Most sex offenders are like other criminal offenders: People convicted of sex crimes are typically generalists with other criminal involvements. A large representative sample in 1994 of 38,000 offenders released from prison in 15 states demonstrates what many other studies also ascertained (Miethe, Olson, & Mitchell, 2006). Miethe and colleagues found that the “sex offenders” in their sample (released prisoners who were most recently incarcerated for a sex offense) had criminal records showing far less specialization than other offender types. Rather than “sex offenders,” they were criminals who had a recent sexoffense in their history along with other kinds of offending: The longer the criminal record of the sex offender, the less likely they were to be specialists. This contrasted with prisoners incarcerated for property and public order offenses, who were more likely to specialize in these types of offenses, the longer their criminal career.
Figure 9.1 Limitations to Sex Offender Studies
Another piece of the misunderstanding surrounding sex offenders includes their relationship to victims. Most sexual offense victims are assaulted or abused by someone they know, not attacked or abducted by a stranger. In fact, almost 50% of all sexual abuse against children under the age of six, and approximately 42% of sexual abuse offenses against children between the ages of six and eleven, are committed by a relative of the child (Craun & Theriot, 2009). In terms of sexual offenses involving adult victims, one survey found that nearly 98% of female victims reported knowing the perpetrator (Craun & Theriot, 2009). This particular misperception results in public support for legislation that requires sex offender registration information to be available to the general public, community notification policies, and residency restrictions, all of which may be misleading the public. While a community resident worries about the sex offender who lives two blocks away, they are unaware or ignore the fact that acquaintances and relatives are much more likely to pose the real threat.
Finally, recidivism rates for sex offenders are not as high as the public perceives. Research on sex offender recidivism found that not only do recidivism rates vary by offender characteristics (e.g., type of offense, relationship to victim, etc.), but are relatively low (Sample & Bray, 2006). Despite our gut feeling that sex offenders must be repeat offenders, many decades of empirical research demonstrates that most convicted sex offenders do not commit new crimes (Zimring & Leon, 2008).
A large representative sample of released prisoners were followed for three years, and the results reveal that sex offenders are less likely to reoffend (Langan & Levin, 2002, pp. 6–8). See Figure 9.2, which shows that the majority of those prisoners were arrested at least once within 3 years of release (67.5%, shown by the large white circle). Figure 9.2 also includes two smaller circles which show who committed new sex offenses: The shaded circle shows that 22% of those prisoners who were not called sex offenders (i.e., they were in prison for something else, and thus were not on the registry) subsequently committed a new offense that was a sex offense. This is in striking contrast to the small circle that represents repeat sex crimes: Just 5.3% of those prisoners were called sex offenders (i.e., they were incarcerated for a prior sex offense) and then went on to commit new sex crimes.
Another way to appreciate the low recidivism rate for sex offenders is to compare them with other kinds of offenders. In the same study, offenders with the highest rearrest rates were those who stole motor vehicles (78.8%), those who possessed or sold stolen property (77.4%), and burglars (74.0%) (Langan & Levin, 2002). Those with the lowest rearrest rates for any kind of crime were rapists (46.0%), those who committed sexual assault (41.4%), and those convicted of homicide (40.7%).
his finding of low recidivism is confirmed by studies with longer follow-up periods. A meta-analysis (a study which combines numerous studies and controls for variation in the study design) with an average follow-up of 4 to 5 years found an average sex offense recidivism rate of 13.4% (N = 23,393; 18.9% for 1,839 rapists and 12.7% for 9,603 child molesters) (Hanson & Bussiere, 1998; see also Hanson & Morton-Bourgon, 2004). These results are also evident in state-specific studies. For example, in a report to the Ohio Sentencing Commission in 2005, it was reported that Ohio sex offenders had a recidivism rate of only 8% for a sex-related offense and 14% for a non-sex offense during a 10-year follow-up period (Office of Criminal Justice Services, 2006).
The Impact of Laws on Known and Unknown Offenders
Research conducted in the years following the enactment of current sex offender policies evaluates the impact on reducing sex offender recidivism. Early studies found mixed results for the effectiveness of registration and notification legislation, as well as for residency restrictions (Center for Sex Offender Management, 2008). Some states have reported that offender registration and notification laws have deterred offenders from recidivating and have found significantly lower rates of recidivism in groups that are required to register compared to offender recidivism rates prior to the implementation of registration laws. However, these studies lack methodological rigor, and are also offset by studies from other states that found no such impacts; some studies even indicated that these laws may actually be contributing to increases in offenders’ risk to recidivate (Center for Sex Offender Management, 2008). Research also indicated that sex offender legislation is associated with a false sense of security in citizens (since most offenders victimize someone who is known to them, not strangers in their neighborhood), increased hostility and vigilantism against sex offenders by community members, and obstacles for offenders seeking housing and employment (Center for Sex Offender Management, 2008).
It is evident from this research that commonly held beliefs regarding sex offenders and their behavior are not accurate. However, sex offender legislation and policies continue to use these misperceptions and justifications for laws that impact nearly every sex offender, regardless of their risk to reoffend, creating overly restrictive and punitive conditions of supervision that may result in an array of negative consequences. A report to the Ohio Criminal Sentencing Commission in 2006 listed several unintended consequences of the state's registration and notification laws. These included misleading the public, negatively impacting family members of offenders and victims, hindering offender reentry, increasing displacement, increasing vandalism and retribution, and a reluctance to report offenses (Office of Criminal Justice Services, 2006). Furthermore, the report suggested that some of these issues increase an offender's stress level and sense of isolation, two factors that play a significant role in an offender's risk to reoffend (Office of Criminal Justice Services, 2006). Research on sex offender legislation and policies primarily focused on offender registration and community notification practices and whether these strategies at supervising sex offenders resulted in lower rates of recidivism among convicted sex offenders. An early study in 1995 was one of the first outcome studies examining the effectiveness of community notification laws (Schram & Milloy, 1995). This study was conducted in Washington and examined the rearrest rates of 90 sex offenders who operated under the highest level of notification and 90 matched offenders who were released prior to the enactment of the notification law. At the end of a 54-month follow-up period for both offense groups, the authors report that there were no significant differences in rearrest rates for the two groups. However, the authors state that the findings, while not statistically significant, tend to suggest that the sex offenders in the community notification group were rearrested at a quicker rate than the group that were released prior to the enactment of the law. Despite the limitations of this early study (particularly, the small sample of sex offenders used in analyses), the findings do not support the effectiveness of community notification practices in reducing recidivism.
A more recent study by Sandler, Freeman, and Socia on the effectiveness of New York's registration and notification laws was published in 2008. This study used time-series analyses to examine arrest rates for sexual offenses before and after the implementation of New York's Sex Offender Registration Act (executed in 1996) in order to see whether the laws reduced recidivism rates or deterred first-time offenders. The authors used data from the New York State offender criminal history files provided by the New York State Division of Criminal Justice Services. Twenty-one years of New York State monthly arrest counts were collected beginning in January 1986 (ten years prior to the enactment of the registration and notification law) to December of 2006, eleven years following the enactment of the law (Sandler, Freeman, & Socia, 2008). Results indicated that the enactment of the registration and notification laws had no significant impact on rates of sexual offending (total sex offenses, rapes, and child molestations). More importantly, the study also revealed that approximately 95% of arrests for sexual offenses (both before and after the enactment of the registration and notification law) were first-time sex offenders, not new offenses by previously convicted sex offenders (recall Figure 9.2). Therefore, the vast majority of sex offenses were committed by individuals who were not listed and would never have been on a sex offender registry (following the enactment of the law). The authors concluded that the legislation, which was originally designed to limit convicted sex offenders’ opportunities to recidivate, targets only a very small percentage of those committing sexual offenses and is ineffective at deterring first-time offenders who are committing the vast majority of offenses (Sandler et al., 2008).
Studies that examined the impact of residency restriction ordinances focused on whether these policies reduced the risk of sex offenders coming into contact with potential victims, particularly children. A 2008 study found that only a handful of the offenders in their sample made direct contact with a child victim that was located within one mile of their residence. In addition, of those offenders who did make direct contact with a victim, none of them did so near a school, park, daycare center, or any other location that is typically included in the residency restriction statutes. These findings fail to support the effectiveness of these types of policies at reducing victimization of children (Duwe, Donnay, & Tewksbury, 2008).
Some critics of current sex offender legislation posit that these overly punitive laws may actually lead to an increase in recidivism and sexual violence, citing the tenets of criminological theories (Wakefield, 2006). Wakefield described the harmful effects of registration and notification laws and residency restrictions, in the context of labeling theory and secondary deviance. The author states that these laws lead offenders to feel stigmatized, isolated, and rejected by their community, obstacles that they may feel they can never overcome. The negative feelings associated with these obstacles may lead offenders to engage in a self-fulfilling prophecy and ultimately commit a new offense (Wakefield, 2006).
According to social control theory, social bonds keep individuals from initially engaging in deviant behavior, and may possibly be what prevents offenders from engaging in future criminal acts (Kilmer, 2012). Close ties with family members, engagement in conventional activities like employment, and connectedness to one's community through involvement in community activities (all of which are considered important social bonds) are extremely difficult to secure for sex offenders once they reenter society following incarceration. As a result, offenders may feel isolated, alienated, and frustrated all of which are shown to be linked to an increased risk for recidivism (Ohio Department of Rehabilitation and Correction, 2007; Tewksbury & Zgoba, 2010). Therefore, overly harsh sex offender legislation and policies, particularly residency restrictions, may be preventing sex offenders from developing strong social bonds and ultimately be responsible for increasing the offender's risk to recidivate (Levenson, 2006; Levenson & Cotter, 2005; Levenson & Hern, 2007).
When all sex offenders are grouped together as a homogeneous, high-risk population, legislation is applied to every offender, which potentially reduces the benefits of these laws and could possibly contribute to recidivism. The current sex offender legislation requires law enforcement to monitor and supervise all sex offenders, regardless of their offense type. In addition, offense type, not individualized risk assessment, typically determines the duration and conditions for the offender's registration and contact with law enforcement (Leon et al., 2011).
Therefore, agencies are overburdened with the number of offenders they have to supervise, and officers may not be able to provide the level of supervision that is required, which can result in high-risk offenders slipping through the cracks. Some argue that resources are better used if they are focused only on those offenders at highest risk to recidivate rather than every offender equally (Leon et al., 2011). That way, high-risk offenders are placed under careful supervision, and low-risk offenders can more easily reintegrate back into society, thus lowering their sense of isolation and stress, and ultimately reducing their risk to reoffend (Robbers, 2009). It is therefore important for legislators to realize that sex offenders are a very heterogeneous group with varying motivations for their behavior as well as other unique characteristics that influence their risk to recidivate
Race, Gender, and Class Implications of the Policy
Feminists and advocates are troubled by the lack of progress in holding sex offenders accountable, and this may be partly attributed to the misleading sense of accomplishment which popular sex offender laws perpetuate (Janus, 2006; Leon, 2011a; Schultz, 2005). By focusing on the subgroup of sex offenders who have been caught and convicted (recall Figure 9.1, which illustrates how many offenders remain unknown), we assuage our fears and concerns, but do little to impact future sex crimes or the structural conditions, including views of women and other rape myths that allow sexual violence to continue (Burt, 1998; Frohmann, 1997). Research also shows that sex offender legislation may perpetuate racial and economic inequality. In a study with the general conclusion that the unintended consequences of residency restrictions dramatically overshadow their purported benefit, Socia (2011) demonstrates how offenders are pushed into overpopulated areas, which has particular implications for economically disadvantaged offenders and for impoverished neighborhoods. In an examination of counties in upstate New York that are representative of the typical parcel density found throughout the United States, Socia measured effects on housing availability, affordability, and social disorganization. He found that sex offenders who must move into the least restricted neighborhoods under the most severe residence restrictions have few housing or rental options. If these few sparsely populated neighborhoods do not contain enough housing for all sex offenders who might have to move there, then homelessness likely increases, which further inhibits the ability to manage and support the offender in the community. Several other recent articles further examined the collateral consequences of these laws on particular communities (Hughes & Burchfield, 2008; Hughes & Kadleck, 2008; Mustaine, Tewksbury, & Stengel, 2006; Zevitz, 2004).
Sex Offenses Committed by Youths
When youth under age 18 are caught engaged in sexual contact, including sexual acts involving force, or with much older youth sexually touching younger children, but also including mutually agreed-on sexual contact, those youth can be taken to court and adjudicated delinquent on sex charges (the juvenile court system does not use the word “convicted” in order to avoid stigmatizing youth and to distinguish them from adult proceedings). Prosecutors can also waive these youth to adult court if they feel the charges are serious enough. Approximately 15 to 17% of reported sex offenses are committed by youth (Andrade, Vincent, & Saleh, 2006; Miner, 2002). While some juveniles may offend as a result of motivations similar to their adult counterparts, juveniles may also have unique motivators that are typically not found in adult offender populations and therefore may have different risk factors and treatment needs (Andrade et al., 2006; Miner, 2002).
Importantly, research also shows that young people who commit sex crimes rarely grow up to be adult sex offenders. Most recently, Zimring and his colleagues examined a Philadelphia birth cohort, which included 13,160 boys and 14,000 girls born in 1958 and followed them through age twenty-six (Zimring, Jennings, Piquero, & Hays, 2009). Even though they used a more inclusive measure than is typically used (police contact as opposed to arrest, adjudication, or conviction), the data did not show a connection between juvenile sexual offending and adult sexual offending.
However, many of the legal and clinical responses to juvenile sexual offending mirror responses seen in the adult population. Stemming from the misconceptions of adult sex offending, legislators fear that juvenile sex offending is becoming increasingly widespread and that juveniles have a high risk of recidivism, similar to their adult counterparts (Letourneau & Miner, 2005). Despite the research, youth adjudicated as sex offenders are subjected to more severe and longer sentences than other juvenile offenders and in some states, they are also subjected to registration, community notification, and civil commitment (Chaffin, 2008; Letourneau & Miner, 2005); under the Adam Walsh Act, just like adult offenders, juveniles are included in public registries, although a few states refused to comply (Freeman & Sandler 2010).
How do We Fix It? Suggestions for Policy Reform
Advocates and policy makers face public pressures to “do something” about sexual offending, and this overshadows the use of empirical evidence to reform public policies: Symbolic responses often take precedence over implementation concerns and related unintended consequences. Attempts to reform the laws have to be framed in terms of real advances in addressing sexual violence. This requires an emphasis on the harms done by sweeping all sexual offending into the same category and the related dilution of law enforcement and other resources.
Reform efforts must differentiate. For example, fixes to the current broad laws should be framed as reallocating resources to focus on the truly dangerous. The Jaycee Lee Duggard/Phillip Garrido case is an illustrative one: Had the California probation officers assigned to visit Garrido been able to focus on a small caseload of high-risk offenders, then there is the likelihood that they would have provided more than cursory oversight of his house and would have identified the kidnap and rape victim Jaycee Lee Duggard rather than allowing him to terrorize her for eighteen years (Strobel, 2009).
Reformers who challenge sex offender policies must be wary of supporting any recommended changes that appear to go “soft” on sex offenders. This can be done by emphasizing that most needed reforms come after punishment: Amendments to registration and residency restriction policies do not change the available sentence structure for sex offenses (Leon et al., 2011). Reforms instead address the separate issue of more sensibly managing sex offenders after their punishment to bring policies in line with facts, not emotions.
