Released sex offenders require social and physical environments that, at a bare minimum, support the process of re-entry and ultimately reintegration. In an ideal world, there would be a seamless relationship between interventions that focus on risk factors causally associated with sexual reoffending and subsequent release of offenders into the community. From this viewpoint, therapists ought to possess a comprehensive understanding of the factors that facilitate successful change and resumption of citizenship roles and seek to build strengths alongside the reduction of criminogenic needs (Ward and Maruna 2007). The available research indicates that if released sex offenders are provided with the resources to access stable housing, establish pro-social support networks, create intimate relationships, and are presented with opportunities for employment, they are less likely to sexually reoffend (e.g., Hanson and Harris 2000; Hanson and Morton-Bourgon 2005; Hepburn and Griffin 2004; Willis and Grace 2008, 2009). The process of ceasing sexual and general offending and becoming a productive member of society is called desistance and has been the subject of intense criminological research over the last eighty years or so (Laws and Ward 2010).

Utilizing desistance literature may well enhance current efforts to reduce sex offender recidivism rates, thereby improving community safety. There are contrasting reports of recidivism rates in sex offenders depending on the level of risk and type of offender sampled by different researchers. Moreover, there is evidence that currently accepted estimates of reoffending rates might markedly underestimate the true extent of sexual and nonsexual crimes by convicted sex offenders. Figures reported from a review of research studies conducted by Hanson and Bussiere (1998) indicated an average sexual recidivism rate of 18.9% for rapists and 12.7% for child molesters. The influence of victim type and previous offences on sexual reoffending rates were examined in a review conducted by Harris and Hanson (2004) of 10 follow-up studies, which had a combined sample of 4,724 adult male sexual offenders followed for periods of up to 15 years. They report that the combined overall recidivism rates for all offenders (14% after 5 years, 20% after 10 years and 24% after 15 years) were similar to rapists (14%, 21% and 24% at 5, 10, and 15 years, respectively) and the combined group of child molesters (13%, 18% and 23%). Significant differences were reported, however, between groups of child molesters over these periods and the rates for extrafamilial boy-victim child molesters were larger (35%). Similarly, those with prior sexual convictions have been consistently found to have higher rates of sexual reoffending than those without previous sexual convictions, for example, 37% over a 15 year period (Harris and Hanson 2004).

The above generally accepted figure of a 24% recidivism rate after 15 years of sexual offending is thrown into doubt by some research and also considerations of risk level and degree of deviancy (Langevin and Curnoe 2006). Once risk is factored in there are significant differences between offender groups with respect to their chances of reoffending, going beyond the 24% figure (Hanson and Morton-Bourgon 2005). High risk offenders have been found to reoffend at much higher rates with Helmus et al. (2009) reporting that the ten year recidivism rate for individuals scoring 9 on the Static-99 (a measure of risk for sexual reconviction ) was 54.3%. Other studies have found life time reoffending rates to be significantly higher than what is commonly cited with Langevin et al. (2004) stating that when undetected sex offences are used approximately 88.3% of their sample of 351 committed further offences over a period of 25 years. Similarly, Friendship and Beech (2005) calculated that taking into account unofficial sources of data multiplied the official sexual conviction rate by a factor of 5.3. Finally, Prentky et al. (1997) in a long term follow up study of offenders discovered that 52% of child molesters and 39% of rapists from a treatment centre were subsequently charged during a 25 year period.

Thus, it is important to note that official reoffending rates are likely to underestimate true reoffending rates considerably and therefore it is reasonable to expect that the true figure will be much higher than those evident in official figures. Rice and Harris (2006) state that “In our considered judgement, it is uncontroversial to conclude that a large proportion of violent and sex offences actually committed go undetected by the criminal justice system” (p. 97). Furthermore, they agree with Langevin et al. (2004) that the true rate of reoffending is greater than that typically reported in the research literature and conclude that “if all subjects have 25 years of opportunity, recidivism exceeds 50%” (p. 98). Keeping in mind the above research data it is apparent that a large number of sexual offenders go on to commit further serious crimes and that it is imperative that correctional management policies and intervention initiatives reflect this fact.

At the heart of desistance theories and research is an ethical assumption that offenders are people like us and deserve the opportunity to live normal lives once they have been punished. Unfortunately the reality is that too often released sex offenders are deprived of such basic yet fundamental social and psychological goods because of emotionally fueled and uninformed public responses to news of those released and legislation such responses have inspired. Sex offender registries, community notification, and residence restrictions are all examples of legislation designed to protect the public from sex offenders living in the community. Given that enactment of community protection legislation is emotionally driven rather than empirically informed, it is not surprising that there is little evidence supporting its assumed effectiveness (Duwe et al. 2008; Letourneau et al. (in press); Levenson et al. 2007b; Sandler et al. 2008; Vasquez et al. 2008; Zandbergen et al. 2010). In fact, paradoxically, the worry is that community protection legislation might be increasing possibility of subsequent reoffending, a risk it is supposed to deter. Residence restrictions, for example, have contributed to social alienation of released sex offenders and severely restricted their housing options (e.g., Levenson and Cotter 2005; Zandbergen and Hart 2006). Furthermore, landlords are unlikely to rent houses to released sex offenders (Clark 2007), and those fortunate enough to find housing run the risk of being driven out of town through community organized pickets, vigils, and evictions (Petrunik and Deutschmann 2008). The picture is equally gloomy when it comes to securing employment, with sex offenders facing significant discrimination by potential employers (e.g., Albright and Denq 1996; Levenson et al. 2007b). Accordingly, addressing the public’s negative attitudes and responses to released sex offenders is of fundamental importance in ensuring that they can be successfully reintegrated into the community.

In view of the crucial role of social acceptance in the reintegration process it is surprising that only a few studies have addressed public attitudes towards sex offenders, and to our knowledge, there has been no research investigating attempts to change public attitudes so that they are conducive to sex offender desistance. The existing literature on attitudes toward sex offenders has focused more attention to attitudes held by professionals working with sex offenders, rather than public attitudes. Without doubt, such research is also important. Professionals holding negative attitudes towards sex offenders risk adopting a punitive, confrontational style in their interactions with them. In the case of therapy staff, such an approach likely compromises quality of the therapeutic relationship between clinicians and offenders, itself paramount in promoting adaptive change. Marshall and his colleagues (e.g., Marshall et al. 2003; Serran et al. 2003) demonstrated that displays of empathy, warmth, and encouragement facilitated treatment change. It can be reasonably assumed that therapists holding positive attitudes toward their clients would be more inclined to show empathy, warmth, and encouragement.

In light of the above research it is evident that effective treatment, re-entry, and reintegration of sex offenders partially hinges on the way they are regarded by mental health professionals and members of the public. In our view, it is unlikely that job offers, educational opportunities, and ultimately close, supportive, and loving relationships will materialize in the absence of goodwill and an affirmation of the intrinsic value of sex offenders. In other words, attitudes and the values they express are partially constitutive of the social and psychological conditions that are conducive to the reduction of reoffending. Our aims in this paper are to review the findings of research on community attitudes about sex offenders and to provide a concise overview of the accepted findings and also identify any lacuna that exists in the literature. More specifically, we set out to (i) advocate for increased attention to addressing community attitudes towards sex offenders in efforts to promote desistance, (ii) discern the state of current knowledge concerning community attitudes, and (iii) provide recommendations for influencing public attitudes and responses so that they promote, rather than hinder, sex offender re-entry and reintegration. We do not intend to systematically address the methodological problems apparent in current research studies and will concentrate instead on tracing the main themes emerging from current research (Willis et al. 2010, for a more comprehensive review of the literature on community attitudes).

The outline of this paper is as follows. First, we provide a synthesis of the current research literature on attitudes towards sex offenders. Second, we consider in more detail those studies that include community member samples. Third, we review interventions aimed at promoting attitude change amongst professionals working with sex offenders and formulate recommendations for promoting positive attitude change amongst the general public. In order to provide a theoretical context for the review we will briefly discuss the concept of desistance and summarize its core ideas.

A Brief Description of Desistance Theory and Research

The notion of desistance has many definitions. It has been described, for example, as a self-reported complete termination of criminal behavior, a cessation of official citations for criminal behavior, a gradual slowing down of criminal behavior, and a marked decrease in the frequency, intensity, and seriousness of criminal behavior. The definitions we find most appealing state that desistance is not an event, but a process replete with lapses, relapses, and recoveries, quite similar to the addiction relapse prevention model originally espoused by Marlatt and Gordon (1985). In criminology we find this position echoed in the work of Maruna (2001) and Laub and Sampson (2001, 2003). Desistance research, which is primarily descriptive, seeks to understand the change processes that are associated with individuals turning away from lives of crime and becoming reintegrated into the community (McNeill et al. 2005). Thus, desistance from criminal behavior is considerably more than simply stopping. As the desistance process advances, there may be intermittency, a combination of pauses, resumptions, indecisiveness, and ambivalence, all of which may finally lead to termination. Desistance is often defined as a termination point, “the last officially recorded or self-reported offense” (Kazemian 2007, p. 9). However, it is more properly seen as a dynamic, ongoing process. In essence, it is the state of stopping and staying stopped that we refer to as “desistance” (Maruna 2001).

Available evidence indicates that there are a number of social and psychological factors that facilitate effective reintegration (Laws and Ward 2010). These events are variously referred to, for example, as “turning points” (Laub and Sampson 2003; Sampson and Laub 1993), “hooks for change” (Giordano et al. 2007), a “change in narrative identity” (McNeill et al. 2005), or “making good” (Maruna 2001). Major desistance influences evident in the literature includes aging (recidivism rates drop as offenders age), marriage, work and job stability, education, cognitive transformation (i.e., creation of a new, more adaptive narrative identity), the pygmalion effect (i.e., social acceptance can increase an offenders chances of positive change), “knifing off” (i.e., severing bonds to criminal past), spirituality, fear of serious assault or death, and sickness and incapacitation. Theories of desistance tend to focus primarily on concepts of social control and agency respectively, and stress the importance of individuals being able to create new social bonds and opportunities for meeting their needs in prosocial and personally meaningful ways (Ward and Maruna 2007).

The rate of recidivism for untreated sex offenders especially underlies the importance of adding to the array of interventions used with this group in order to protect the community. The fact that natural desistance can occur independently of the actions of correctional personnel is of crucial importance and offers practitioners an untapped arena to capitalize on in their attempts to encourage offense-free lives. If offenders relinquish antisocial goals and inclinations because they have found satisfying jobs or become romantically involved with people they care about, then it follows that any social initiatives that make these events more possible are to be encouraged. Some individuals might require more scaffolding than others in acquiring the capacities necessary to construct and put into action a plan for living that is adaptive and meaningful. Sometimes a greater need for professional input is a legacy of offenders living in particularly impoverished social environments with minimal social capital and sometimes it is because they possess few psychological resources of their own. In either of these situations treatment programs can be helpful: the former setting out to instill psychological skills and the latter concentrating on creating social opportunities and supports. What we are suggesting is that differences between natural and professionally assisted desistance may reside in the psychological and social resources available to specific individuals rather than representing qualitatively distinct routes to crime cessation (Laws and Ward 2010). All human beings require help from other people to acquire and utilize the psychological capabilities and social resources necessary to realize their aspirations whether this involves completing job training, participating in social activities, or remaining crime free. Exactly what kind of help is needed or likely to be most useful will be a function of their personal characteristics and situation. The reason positive community attitudes towards sex offenders are critical is that they partially mediate the transition between the treatment arena and life on the outside.

A Synthesis of Empirical Literature on Attitudes Towards Sex Offenders

We will briefly consider some methodological issues relating to the measurement of attitudes towards sex offenders. Several methods have been used to assess attitudes, including multi item scales that provide an overall score reflecting where an individual’s attitude sits on a continuum ranging from very negative to very positive (Church et al. 2008; Hogue 1993; Weekes et al. 1995; Wnuk et al. 2006). They have been predominantly used in studies investigating group differences in attitudes and the effectiveness of interventions designed to influence attitudes. Attitudes Towards Sex Offenders (ATS; Hogue 1993) scale is frequently the measure of choice in studies, and is an adaptation of Attitudes Towards Prisoners (ATP; Melvin et al. 1985) scale. In his construction of the ATS, Hogue omitted the word “prisoner” in each item of the ATP and inserted the word “sex offender.” Thus, although the ATS appears to assess attitudes towards sex offenders, the content of questions was in fact not designed to target sex offender specific stereotypes, but rather stereotypes common to offenders in general. By way of contrast, two recently developed measures, the Attitudes Toward the Treatment of Sex Offenders (ATTSO; Wnuk et al. 2006) scale and the Community Attitudes Towards Sex Offenders (CATSO; Church et al. 2008) scale, incorporate stereotypical beliefs about sex offenders, thus potentially capturing richer information about attitudes towards sex offenders than the ATS. Given their relative newness, however, the ATTSO and CATSO have not yet been widely adopted by researchers (for exceptions see Balow and Conley 2008; Sahlstrom and Jeglic 2008).

Researchers have demonstrated that people’s attitudes towards sex offenders are more negative than their attitudes towards general offenders (Craig 2005; Hogue 1993; Weekes et al. 1995). Nevertheless, despite this main effect the research literature has illustrated that some groups of people have less negative attitudes compared to other groups. To make discussion easier we have summarized the key findings of studies that compare ATS scores across different groups in Table 1. In the first major study of public attitudes, Hogue (1993) administered the ATS to groups of police officers, prison officers (both with and without involvement in sex offender treatment), rehabilitation workers (probation officers and psychologists), and sex offenders. All groups significantly differed in their overall attitudes towards sex offenders. Specifically, police officers demonstrated the most negative attitudes and sex offenders showed the most positive attitudes. Probation officers and psychologists showed more positive attitudes than prison officers involved in treatment, who showed more positive attitudes than prison officers not involved in treatment. That is, employees with lesser contact with sex offenders showed more negative attitudes, whereas employees with greater contact with sex offenders, and sex offenders themselves, showed more positive attitudes.

Table 1 Group differences on the attitudes towards sex offenders scale
Full size table

As evident in Table 1, the pattern identified by Hogue (1993) has been consistently replicated across a number of studies. Students and general community members have endorsed more negative attitudes than employees working with sex offenders, and within the latter group, greater contact with sex offenders has been associated with less negative attitudes. Interpreting such findings, it has been suggested that in the absence of contact with known sex offenders, attitudes might be more heavily influenced by sex offender stereotypes portrayed in the media (Kjelsberg and Loos 2008). Similarly, individuals working with sex offenders might have received specialized education and training, potentially reducing their reliance on stereotypes when forming attitudes (e.g., Lea et al. 1999). However, it should be noted that in the only study to consider the effect of previous training on attitudes towards sex offenders, no effect was found (Hogue and Peebles 1997). Another possible explanation for between-group differences in attitudes towards sex offenders is that higher levels of education required for certain occupations (e.g., psychologists) could result in less negative attitudes compared to other occupations not requiring the same level of education (e.g., police officers). Thus, the extent to which contact with sex offenders, specialized training, or higher educational attainment is associated with less negative attitudes remains unknown.

Considering demographic differences in attitudes towards sex offenders, most studies have found no differences between male and female respondents (Brown 1999; Hogue and Peebles 1997; Johnson et al. 2007; Katz et al. 2008; Kjelsberg and Loos 2008; Sahlstrom and Jeglic 2008), with some exceptions. Ferguson and Ireland (2006) found that females endorsed less negative attitudes than males, whereas other studies have found that females reported more fear about the prospect of a sex offender living nearby and greater agreement with community notification policies (Caputo and Brodsky 2004; Kernsmith et al. 2009; Levenson et al. 2007a; Phillips 1998). Considering the influence of respondent age, two studies showed that increased age was associated with more positive attitudes (Craig 2005; Kjelsberg and Loos 2008), one study reported that younger participants endorsed more positive attitudes than older participants (Brown 1999), and another one study found no effect of respondent age on attitudes about sex offenders (Katz et al. 2008). Phillips (1998) found that the age group most likely to have minor children (30 to 40 year olds) expressed the most serious concerns about a sex offender living nearby. A number of studies have shown that parents and individuals living with children have reported similar attitudes to nonparents and individuals without children in their homes (Craun and Theriot 2009; Katz et al. 2008; Levenson et al. 2007a; Nelson et al. 2002). However, one study found that parents of minor children were more likely than nonparents to opine that community notification was important (Caputo and Brodsky 2004).

Brown (1999) found that higher socio economic status (based on occupation) was associated with more favorable attitudes about sex offender treatability than lower socio economic status. As discussed by Brown, differences in attitudes based on socio economic status might be explained by differences in educational level, whereby less educated people tend to hold more negative attitudes than better educated people. With one exception (Valliant et al. 1994), no studies were found that directly considered the effect of educational level on attitudes towards sex offenders; with Valliant et al. (1994) discovering that there were no differences in attitudes between first and third year psychology students. It must be acknowledged, however, that first and third year university students do not represent substantially disparate groups in terms of educational level. In addition, Brown asked participants which newspapers they most frequently read, and established that participants in higher socio economic groups read more broadsheet newspapers (e.g., The Times, The Guardian) whilst participants in lower brackets read more tabloid newspapers (e.g., The Sun, The Mirror). The added sensationalism of crime stories in tabloid newspapers over and above that common to broadsheet newspapers might further have accounted for differences in attitudes between socio economic groups.

Several studies have considered the impact of personal experience of sexual abuse or closeness to a victim of sexual abuse on attitudes towards sex offenders. Some studies concluded that personal experience and/or closeness to a victim of sexual abuse had no influence on attitudes (Brown 1999; Hogue and Peebles 1997; Katz et al. 2008; Levenson et al. 2007a; Sahlstrom and Jeglic 2008). However, by way of contrast, other researchers have found such factors to have a positive impact on attitudes (i.e., through respondents endorsing less negative attitudes) (Ferguson and Ireland 2006; Nelson et al. 2002). Similarly, one study reported that being a victim of a violent crime was associated with reduced misperceptions about stranger danger (Craun and Theriot 2009). Only one study considered the impact of knowing a sex offender, and discovered that respondents who knew a sex offender viewed sex offenders more favorably than respondents who didn’t know a sex offender, however this difference did not reach statistical significance (Sahlstrom and Jeglic 2008).

In summary, the research literature illustrates an emerging pattern whereby increasing contact with sex offenders, be it in a professional or personal capacity, is associated with less negative attitudes towards them. Thinking about implications of these intuitively unexpected findings, it is plausible that associating with individuals who have engaged in sexually abusive behavior may humanize them and thereby reduce individuals’ reliance on simple sex offender stereotypes when forming beliefs and attitudes (Ferguson and Ireland 2006; Kjelsberg and Loos 2008). In addition, research findings suggest that specialized training and higher educational attainment might contribute to less negative attitudes, however the nature and direction of these relationships are unclear. The next section considers negative public attitudes in greater detail, through reviewing findings from qualitative studies that have included community member samples.

Studies with Community Member Samples

In one of the first and most comprehensive studies using a general public sample, Brown (1999) surveyed 312 community members randomly selected from the electoral roll about their attitudes and anticipated behavior towards sex offenders. Concerning attitudes towards sex offender treatment, the overwhelming majority of respondents (95%) thought sex offenders serving a determinate prison sentence should receive treatment. Despite considerable support for prison based treatment, however, participants were skeptical about the effectiveness of treatment with one quarter stating that treatment could never prevent recidivism. When asked whether they supported the existence of a treatment center in their community, almost two thirds (64%) of participants opposed the idea, while the remaining participants (36%) supported the idea. Most of the participants who opposed the community based treatment center indicated that they would behave in an opposition consistent manner, for example 26% of these participants reported that they would start a campaign and 80% stated that they would sign a protest petition. In contrast, participants who responded in favor of a treatment program in their neighborhood reported that they were less prepared to actively support it. Considering anticipatory behavior towards known sex offenders who had completed their sentences, only 6% of respondents indicated that they would rent housing to sex offenders, and whilst more encouraging, only 30% of respondents indicated that they would be prepared to employ sex offenders. In reviewing her findings, Brown (1999) commented that although community members were supportive of prison based rehabilitation endeavors they were ill prepared to accept the reality that once treated, such offenders required opportunities to “live normal lives in society” (p. 245). Clearly, such attitudes are unlikely to assist the process of desistance and create opportunities for offenders to gain employment or establish supportive and close social networks.

Since Brown’s (1999) study, researchers have investigated the accuracy of community members’ perceptions of sex offenders, which undoubtedly impact on their attitudes. Levenson et al. (2007a) explored the accuracy of public perceptions amongst 193 residents in Florida. They found that community members believed that sex offenders represented a homogenous group with very high recidivism rates, supporting their hypothesis that public beliefs were inconsistent with the research literature. In terms of punishment and treatment, respondents favored lengthy terms of imprisonment (M = 38.8 years), and although endorsing treatment, consistent with Brown (1999), they were skeptical about its benefits. Similarly, in an online survey of 127 adults via an internet messaging board, most respondents believed that most sex offenders will reoffend and that treatment for sex offenders is not effective (Katz et al. 2008). Brown et al. (2008) surveyed 976 community members in the United Kingdom about their perceptions of sex offenders and attitudes towards community reentry. Consistent with earlier findings, respondents tended to overestimate reconviction rates. When asked about accuracy of the media’s depiction of sex offenders, more than 50% of respondents believed that the media was either accurate or underreported the true level of risk posed by sex offenders. In another study, Craun and Theriot (2009) found that 29.7% of respondents expressed more concern about a stranger sexually abusing a child compared to someone known to the child, and 56.7% were equally concerned of sexual assault by a stranger or someone known to the child. Moreover, they found that awareness of a locally residing sex offender was associated with greater endorsement of misconceptions about stranger danger. Thus, as highlighted by Cruan and Theriot, community notification policies might have unintentionally increased public misperceptions about sexual offending.

Saghara and Wilson (2006) employed an experimental method to investigate the endorsement of sex offender stereotypes amongst 60 professionals involved in the treatment of sex offenders and 71 school teachers. The authors presented participants with vignettes about a sexual assault and a description of the alleged perpetrator. Details of the alleged perpetrator were manipulated according to stereotype consistent (builder, dirty man, sexually frustrated, unmarried) and inconsistent (university professor, married man with two children, book shop owner, victim’s father) information, and participants were asked to rate the likelihood that different suspects were guilty. The authors found that participants experienced in working with sex offenders endorsed significantly fewer stereotypes than the inexperienced group. Moreover, they found that knowledge of child sexual abuse (e.g., prevalence rates and perpetrator characteristics) significantly mediated the relationship between endorsement of stereotypes and experience in working with sex offenders. In other words, the less knowledgeable individuals were about child sexual abuse, the more likely they were to endorse stereotypical beliefs about child sex offenders.

It is well known that sex offenders do not represent a homogenous group, yet few studies have accounted for this fact when investigating attitudes towards sex offenders (Ward et al. 2006). Rather, researchers have often classified sex offenders as comprising a single group and participants have been asked to comment on their attitudes towards this group as a whole. In one of the few studies differentiating different types of sex offenders, Ferguson and Ireland (2006) gave participants a vignette that manipulated offender type: stranger rapist, acquaintance rapist, stranger victim pedophile, or familial victim pedophile. Participants were asked to focus their responses to the ATS on their vignette, and results showed no differences in attitudes towards different types of sex offenders. More recently, Kernsmith et al. (2009) conducted a telephone survey with 733 community members residing in Michigan. Respondents were asked to rate how afraid they were at the prospect of sex offenders convicted for different types of offenses (i.e., incest, statutory rape, marital rape, pedophilia, date rape, historical offenses) living in their neighborhood, and their support for registration of different offender types. All sex offenders elicited some fear in participants, which was itself related to support for registration requirements. Pedophiles and incest offenders were most feared, and statutory rapists were least feared. However, 65.1% of respondents still supported registration requirements for statutory rapists. Taken together, these studies suggest that all types of sex offenders evoke a significant threshold level of anxiety amongst the general public.

Several studies have investigated public perceptions about community protection policies, and these studies have consistently demonstrated that the public support the enactment of community protection legislations (Brannon et al. 2007; Brown et al. 2008; Kernsmith et al. 2009; Levenson et al. 2007a; Lieb and Nunlist 2008; Phillips 1998; Salerno et al. 2010; Schiavone and Jeglic 2009). For example, Levenson et al. (2007a) discovered that respondents overwhelmingly favored subjecting all sex offenders to community notification and believed it was a useful measure in reducing recidivism rates. Although residency restrictions did not receive the same level of support as community notification, more than half (58%) of respondents believed that residency restrictions were effective at reducing recidivism. Respondents were asked whether they would support community protection polices in the absence of scientific evidence to support their utility, with 73% of respondents indicating that they would still likely support such policies. Similarly, Schiavone and Jeglic (2009) found that 51% of respondents to an online questionnaire agreed that “low risk” sex offenders should still be subjected to registration and community notification policies, and 20% of respondents believed such polices should also apply to sex offenders posing “no risk.” Other researchers have found that notification can increase citizens’ anxiety due to a lack of education and information about protecting oneself or one’s children from sexual assault (Caputo 2001; Zevitz and Farkas 2000). Finally, consistent with Sanghara and Wilson’s (2006) earlier findings showing increased knowledge about child sexual abuse was associated with less negative attitudes, one study found that less knowledge about child sexual abuse yielded greater support for notification (Redlich 2001).

Investigating attitudes toward juvenile sex offenders, Sahlstrom and Jeglic (2008) gave 208 students three scenarios of a sexual assault. Different independent variables were manipulated in each scenario, specifically perpetrator age (8, 9, 11, or 13 years), then sex, then ethnicity (Caucasian, African-American, Hispanic, or East Indian). Following reading each scenario participants were asked a series of questions including how they thought the case should be handled (e.g., formally through child protection services and/or the Police or informally through parents), and the seriousness of the assault. In general, respondents believed juvenile sexual offending was serious and required formal intervention despite differences in age, sex, and ethnicity of the perpetrator. Consistent with their findings, Salerno et al. (2010) reported that respondents equally supported subjecting both adult and juvenile sex offenders to registration laws, but that registration requirements were less likely supported for individuals convicted of less severe offenses. Interpreting such findings, Salerno et al. highlighted the impact of retributive concerns in addition to utilitarian concerns in people’s justice decision making processes.

In summary it seems that negative public attitudes towards sex offenders are partly a function of moral outrage and disgust towards sex offenders, regardless of their typology or assessed level of recidivism risk, and partly a function of public misperceptions about sexual offending. Applying a social psychological theory of prejudice and stereotyping can enlighten our understanding of perceptions of sex offenders. With other types of social prejudice, it is known that segregation and lack of knowledge about the “out-group” results in greater prejudice and stereotyping, whereas personal knowledge of and experience with a particular group reduces prejudice (Allport 1954; Gaertner et al. 1991). Stereotypes can develop from uninformed perceptions and may serve to reduce fear and manage interactions (Sherif et al. 1988). Thus sex offenders may seem—to those who know only what they see through the media—to be unpredictable, evil, and very dangerous, whereas people with greater knowledge and/or experience with known sex offenders are less likely to rely on media depictions in forming their attitudes. In fact, a study involving interviews with U.S. politicians revealed that their primary source of information about sex crimes was the media, and that this information shaped their legislative proposals (Sample and Kadleck 2008).

In our view this is a problem both for ethical and pragmatic reasons. Ethically, offenders are due the respect owed to all human beings and thus an opportunity to be reconciled with the community once they have been punished (Ward and Salmon 2009). From a pragmatic viewpoint, hostile, mistaken beliefs, and a refusal to actively help sex offenders re-enter and establish themselves within the community may backfire and create greater levels of risk. Accordingly, several researchers (Levenson et al. 2007a) have advocated for the value of providing accurate information to the public in an attempt to positively influence attitudes towards sex offenders. To our knowledge, no studies have investigated interventions designed to influence public attitudes although a number have examined interventions designed to influence attitudes held by professionals working with sex offenders, as reviewed in the next section.

Effect of Training on Professionals’ Attitudes

The above literature review reveals that attitudes towards sex offenders vary amongst members of the public, and to some extent, correctional and mental health practitioners. The range of attitudes arguably gives rise to conflicting ways of conceptualizing sex offenders, expressing a view of offenders as moral strangers or deviants who need to be contained, to “people like us” who deserve a change to redeem themselves and rejoin the community (Laws and Ward 2010). From the viewpoint of desistance theory and research, distorted and relentlessly punitive attitudes constitute obstacles to successful behavioral change, and ultimately, may well result in increased rather than lesser risk of further offending. Therefore, it makes sense to devise interventions to create more accurate and constructive attitudes towards offenders and the people that support them. The only work that has been done in this area that we are aware of are programs designed to positively influence attitudes of professionals working with sex offenders, which we will now discuss.

With one exception (Taylor et al. 2003), the ATS has been used to measure the effectiveness of training programs designed to positively influence attitudes of professionals working with sex offenders. In the first such study, Hogue (1995) administered the ATS scale before and after a training program delivered to British Prisons’ employees (including prison officers, probation officers, and psychologists). The three week training program was designed to prepare staff for facilitation of structured sex offender treatment groups, and included theoretical and practical components. An effect of training was found and participants’ scores on the ATS scale significantly increased post training, indicating more positive attitudes towards sex offenders compared to pre-training.

More recently Craig (2005) administered the ATS scale to residential hostel workers and probation officers working with sex offenders before and after a two day training workshop. The workshop aimed to familiarize participants with theoretical models of sexual offending, challenge common misperceptions about sex offenders, identify and appropriately respond to cognitive distortions, and equip participants with a basic understanding of risk assessment, treatment programs, and relapse prevention strategies. Results revealed that mean ATS scores did not differ post-training compared to pre-training, pointing to a minimal effect on participants’ attitudes towards sex offenders. In discussing this result Craig highlighted that unlike Hogue’s (1995) intervention, his program was shorter and narrower in focus and was not designed to prepare participants for group treatment facilitation. In a study similar to Craig’s, Johnson et al. (2007) investigated the impact of an educational program on probationary police officers’ attitudes towards sex offenders. Contrary to their predictions, participants showed significantly more negative attitudes towards sex offenders post-training compared to pre-training. As suggested by Johnson et al., the focus on factual information without any attention to rehabilitation issues might have partly explained their results. In another study, Kjelsberg and Loos (2008) examined the impact of a two day educational program on prison employees’ attitudes towards sex offenders. Participants were administered the ATS prior to commencement and one year following the program, and consistent with Craig’s results, the authors found no change in ATS scores. Kjelsberg and Loos offered several interpretations of their findings, including that the educational program was indeed ineffective, that the program was effective in the short term but that this effect was not sustained one year later, and that the program was effective but that the ATS was limited in its ability to capture attitude change over time.

Finally, Taylor et al. (2003) evaluated the utility of an educational program for nurses and social workers who worked with sex offenders with learning disabilities. The program included modules relating to sexuality and intellectual disability, motivation for sexual offending, and information relating to the assessment, treatment, and management of sex offenders. Program effectiveness was measured through pre- and post- training administration of the Sex Offender Knowledge and Attitudes Questionnaire (SOKAQ), which was developed specifically for the study. The SOKAQ questions closely reflected the program content (e.g., “when working with sex offender clients, what kind of assessments do we need to administer?” p. 207). Taylor et al. reported that overall, participants’ responses on the SOKAQ improved following the educational program, however whether this was due to attitude change or simple regurgitation of program content was unclear, given the nature of the measure used.

In summary, the available evidence indicates that short educational programs are ineffective, or at worst, harmful in effecting attitude change among professionals working with sex offenders. Integrating these results with earlier findings that some individuals expressed support for community notification policies regardless of offender risk level (Schiavone and Jeglic 2009) or in the absence of evidence for their effectiveness (Levenson et al. 2007a), it seems plausible that educational programs served to increase the salience of inaccurate beliefs about sex offenders and their risk, however small that risk might be. In our view this raises the possibility that moral beliefs centered on themes of retribution, punishment, and unacceptable threat may be causing individuals to lose sight of the fact that offenders are fellow human beings rather than simply sources of risk and contamination (Ward and Salmon 2009). Moreover, educational programs, like community protection policies, might have inadvertently served to perpetuate this image of sex offenders as qualitatively distinct from the rest of us. Furthermore, effectiveness of Hogue’s (1995) training program might be due to its focus on rehabilitation, through instilling in participants the idea that sex offenders are fellow human beings and capable of change. In light of these results and possible interpretations, it seems crucial that careful consideration is given to the content of any educational program designed to influence people’s attitudes towards sex offenders.

Recommendations for Influencing Public Attitudes to Promote Desistance

In this final section we draw upon findings from the literature reviewed to offer recommendations for effecting attitude change amongst the general public. To reiterate, desistance from sexual offending is dependent on specific environmental conditions, for example, stable housing, access to employment opportunities, cognitive transformations, and social support. Currently, community responses towards individuals convicted of sexual offenses severely hinders opportunities for desistance, and therefore addressing public attitudes is an obvious step to promote a safer and more just society. As illustrated in the preceding section, utilization of educational programs to professionals working with individuals who have engaged in sexually abusive behavior has so far largely proven ineffective, and therefore there is no reason to believe that similar programs would prove effective with the general public. Instead, we will offer alternative, somewhat speculative, recommendations for positively effecting change in public attitudes and promoting desistance.

The media represents the predominant means through which individuals acquire information about sexual offending (Brown et al. 2008). Given that the media sensationalizes sexual crime stories and focuses almost exclusively on low base rate crimes such as abduction, murder, and high volume predatory behavior (Pratt 2007), it is hardly surprising that misperceptions about sexual offending are so pervasive in western societies. A content analysis of news articles in the U.S. revealed that individuals who have engaged in sexually abusive behavior were often portrayed as unredeemable, with themes of inevitable recidivism (Sample 2001). Sample also found that over-reporting of single but heinous events (e.g. sexually motivated abduction and murder) gave the impression that these sorts of crimes are widespread despite their statistical improbability. The media represents a powerful engine driving public opinion, and therefore our first recommendation is to encourage academics and other suitably qualified professionals to actively engage with the media. Potential media contributions might include feature articles in newspapers and magazines, dispelling common misperceptions and providing accurate, research-based information. In addition, agreeing to (or even better, volunteering for) media interviews to offer an informed perspective on sexual offending issues represents another opportunity for correcting commonly held misperceptions. Currently, academics often refrain from speaking to the media, whereas other groups such as lobbyists for tougher sentences and victim advocates make themselves readily available for comment (Pratt 2007). Moreover, it would seem pertinent from the studies reviewed that academics and professionals ensure that individuals with histories of sexual offending be portrayed as fellow human beings capable of change, rather than as a deviant and somewhat sinister group who are qualitatively distinct from the rest of us. At the same time, it is vital that academics do not appear oblivious to the abhorrent nature of sexual abuse and its devastating effects on victims and the rest of the community. Rather, the aim is to convey understanding of the severe harm inflicted on innocent members of the community by sexual abuse while also appreciating that holding people accountable also entails giving them a chance at reconciliation following proportionate punishment and, if necessary, specialist treatment.

Second, we recommend that increased attention be given to creating opportunities for individuals with histories of sexual offending to live productive lives in society, thus actively augmenting the natural desistance process. Undoubtedly, this requires collaboration between criminal justice professionals, employers, landlords, and various community groups. A detailed discussion of the practicalities surrounding such collaboration is beyond the scope of this paper, however the Circles of Support and Accountability (COSA; Wilson et al. 2002) model represents an excellent example of collaboration between key stakeholders whereby desistance from sexual offending is promoted through professionally-trained volunteer support networks. Recent COSA evaluations have found that all key stakeholders, including core members (individuals with convictions for sexual offending), community volunteers, affiliated professionals, and community members with no involvement in COSA responded favorably to COSA (Wilson et al. 2007a), and that core members had lower recidivism rates compared to a matched control group (Wilson et al. 2007b). Interestingly, the recently passed “Second Chance Act” (U.S. legislation that emphasizes re-entry and provides funding for reintegration programming) explicitly excludes individuals with convictions for sexual offending from receiving these services. Ostensibly, the idea was to direct funding and services to seemingly more deserving and redeemable offenders with the hopes of enhancing success rates. But the wisdom of diverting rehabilitation and reintegration services away from those perceived to pose the greatest threat to communities is counterintuitive.

Third, we advocate for comprehensive changes to community protection legislation. Current laws mandating public notification may inadvertently promote stereotyping and prejudice. In addition, laws restricting where individuals convicted of sexual offenses can live and work further segregate them from mainstream society and therefore reinforce ostracism and exclusion (Levenson et al. 2010). Interaction and personal relationships, however, can contradict negative expectations and lead to more positive views, greater acceptance, and less intolerance (Wright et al. 1997). In keeping with our theme of promoting desistance, we recommend that community protection legislation is reserved for individuals with stranger victims and a high risk of reoffending, rather than its blanket application which is common in many States and countries.

Finally, as we have done in this section, we recommend that fellow researchers discontinue labeling individuals with a history of sexual offending with words such as “sex offender,” “child molester,” “sexual predator,” and so on. Not only might such labels perpetuate an image of sex offenders as a qualitatively distinct group, but they could also induce a negative pygmalion effect in individuals with a history of sexual offending (Maruna et al. 2004). That is, individuals might start to see themselves as others see them: as inherently dangerous, moral strangers who do not deserve a chance at redemption, and indeed, ought to be quarantined somewhere away from the rest of humanity (Ward and Maruna 2007). We propose, that until such individuals view themselves as fellow human beings capable of change, their chances for desistance are almost certainly compromised.

Conclusion

In this paper we have approached the investigation of attitudes towards individuals who have committed sexual offenses within a desistance framework. An attractive feature of desistance ideas and research is the recognition that ultimately individuals with histories of sexual offending need the help of members of the community alongside the best efforts of practitioners if they are to become productive and caring citizens. An implication of this viewpoint is that public attitudes and responses to sexual offending have profound impacts on the range and quality of opportunities for successful reintegration and desistance amongst persons convicted of sexual offenses. Despite increasing research demonstrating the importance of factors such as stable housing, employment, caring relationships, and social support in preventing sexual recidivism, and likewise also revealing the ineffectiveness of community protection legislation, things have not changed. It makes sense both ethically and pragmatically to share responsibility for the process of change with individuals who have engaged in sexually abusive behavior. They cannot do it on their own and we cannot afford to continue playing a tune that is so discordant, and ultimately, harmful to us all.