Keywords

Introduction

The term “sex offender treatment” is generally used to describe psychological programmes delivered to groups of people convicted of sexual offences, in either custody and community settings, for the tertiary prevention of sexual recidivism. Somewhat less commonly, the term is used to describe other approaches such as medication to reduce an individual’s level of sexual arousal (e.g. Beech & Harkins, 2012: Lösel & Schmucker, 2005).

The term treatment, used in the context of rehabilitative efforts with people convicted of sexual offences, arguably misrepresents what rehabilitation is both able to achieve and aims to do. To a lay person, the term treatment could imply that we know how to reduce sexual recidivism with perpetrators of sexual crimes, suggesting that we can as a matter of course identify and address the symptoms of sexual offending and take action to ameliorate these symptoms in the same way it can be possible to administer medical care for an illness. This term also suggests that if people haven’t received any treatment then they are untreated and will continue to offend. However, the extremely low recidivism rates (e.g. 2.2 % sexual recidivism over 2 years; Barnett, Wakeling, & Howard, 2010) for this type of offence suggests otherwise: it seems that, even taking into account the problem of detecting all incidences of sexual offending, most people with sex offence convictions are likely to desist from further offending whether or not they are treated. Another issue with using the term “treatment” is that it risks failing to acknowledge the active role of the treatment participant in the process of change. As the field of desistance research has shown us, treatment programmes do not “make people into non offenders”. They offer opportunities for people to learn new cognitive and behavioural skills, but they do not in themselves change people. The individual himself or herself is the person who achieves change through, for example, using the opportunities provided by the programme to examine and alter their attitudes that support sexually abusive behaviour. If there is no intent to change, a treatment programme will be unlikely to make any difference to offending (Webster, 2005; Webster, Bowers, Mann, & Marshall, 2005). This is not to say it is always the client’s fault if a treatment programme is ineffective—programmes may also be unengaging or may involve ineffective treatment procedures. The attribution of effectiveness, therefore, is a complex endeavour which must go beyond reflexive finger pointing at either the individual or the programme.

The notion of “treatment” can also distract attention from the wider context that can and is necessary to support desistance from offending outside of the treatment room. Pharmaceutical treatment or surgery could arguably often be expected to work regardless of the context in which it is received by the patient, although even with these types of treatment, the social context also assists with success and recovery. Psychological “treatment”, particularly when it is mandated or otherwise not entirely voluntary (as is usually the case with sex offenders), needs to be situated in a context of wider social and professional support. It is important that stigma is actively minimised, and that the sex offender does not fear for his personal safety, which is often the case particularly in prison (Blagden, Winder, & Hames, 2014; Mann, 2016; Mann, Webster, Wakeling, & Keylock, 2013).

For all these reasons, it is widely accepted that the effectiveness of sex offender treatment is difficult to determine. Therefore, we begin this chapter by acknowledging the mixed evaluation findings in this field and the consequent need for most treatment programmes to re-evaluate their curricula and methods. However, our main purpose for this chapter is to look at the evidence for what parts we have got “right”. What features of our current treatment approach should remain in rehabilitative programmes aimed at reducing recidivism with people convicted of a sexual offence? Although our brief for this chapter was to focus on the strengths of treatment, it is important to stress that we are highly cognisant of the typical problems in most current treatment approaches. By focussing on the strengths, however, we hope to clarify the most effective and promising parts of treatment that should successfully reduce sexual recidivism.

Sexual Offending Treatment Could Be More Effective

Evaluating sexual offending programmes is complicated in part by the very thing that makes engaging in rehabilitation with this group so challenging: people who have sexually offended represent an extremely heterogeneous group. This heterogeneity is evident in the type and detail of the sexual offence committed, an individual’s aetiology, what motivated their offending, co-morbidity issues and their level of risk. Not all risk factors that have been identified for sexual offending perpetrators as a group will apply to each individual convicted of a sexual offence, and the motivation for offending and what drives or maintains it can vary significantly. Therefore, evaluations of programmes may need to consider multicomponent programmes with flexible delivery schedules. From an evaluation point of view, this kind of programme is hard to evaluate, and it will be even harder to draw conclusions about “why” it worked or didn’t work.

There are other practical problems in terms of matching treatment and control groups as issues of heterogeneity discussed above would indicate. Those who undertake treatment may differ considerably from comparisons in terms of denial of the offence (maintaining innocence for offending can be a bar to entering treatment) as well as the role that deviance played in offending. Matching is also difficult because international data suggests that offenders selected to undertake treatment are at greater risk of sexual recidivism than routine samples that have not been identified for a programme even after taking static risk into account (Helmus, Hanson, Thornton, Babchishin, & Harris, 2012).

Lastly, although there have been several systematic reviews and meta-analyses of sex offender treatment effectiveness, these reviews mask considerable heterogeneity of treatment approaches. While some differences between programmes can be accounted for in meta-analysis (e.g. theoretical orientation such as psychodynamic or cognitive behavioural), there are many more subtle differences that may not be available to meta-analysts, such as therapist variability, programme context and degree of felt coercion. So, while contemporary treatment programmes across and within jurisdictions share similarities, they can also differ considerably in significant ways that include the methods they use and what they target and the extent by which they constitute evidence-informed practice (McGrath, Cumming, Burchard, Zeoli, & Ellerby, 2010). The variance found across treatment programmes both complicates our ability to reach conclusions about the overall strengths of treatment while paradoxically advancing our understanding. Different practices potentially allow comparisons to be made on what are the more beneficial aspects of treatment against the desired goals, although unfortunately, this kind of research has rarely been undertaken.

Bearing these caveats in mind, the current evidence base for treatment effectiveness is by no means strong. Although evidence from meta-analysis indicates that sex offender treatment programmes can bring about reductions in recidivism (e.g. Hanson, Bourgon, Helmus, & Hodgson, 2009), this is by no means routinely the case. The efficacy of sex offender treatment continues to be debated robustly in the rehabilitative literature (e.g. Crighton & Towl, 2007; Ho & Ross, 2012; Mann, Carter, & Wakeling, 2012; Marshall & Marshall, 2007). Sex offender treatment cannot be regarded as an “evidence based treatment” according to Kazdin’s (2008) definition of “interventions or techniques that have produced therapeutic change in controlled trials” (p. 147). There have been only a few controlled trials of sex offender treatment, and the largest scale of these studies (Marques, Wiederanders, Day, Nelson, & von Ommeren, 2005) found no effect for the treatment group.

Taking an Evidence-Based Approach to Sex Offender Treatment

There is no approach to sex offender treatment that can be regarded as an “evidence-based treatment” (EBT) according to customary definitions (e.g. Kazdin, 2008). Given this situation, our best efforts can only be described as “evidence-informed practice” or “evidence-based practice”, defined by the APA as “the integration of the best available research with clinical expertise in the context of patient characteristics, culture and preferences” (APA Presidential Task Force on Evidence-Based Practice, 2006, p. 273). Taking an evidence-based approach requires that those who deliver treatment are cognisant of the best available research and also that the best available research is at least adequate in its scope and design.

In the sex offender treatment field, arguably more than with other criminal behaviours, there is a strong community of practitioners who take research seriously. Three international organisations—the Association for the Treatment of Sexual Abusers (ATSA), the International Association for the Treatment of Sexual Offenders (IATSO) and the National Organisation for the Treatment of Abusers (NOTA)—exist to support practitioners; each publishes its own research journal and holds regular conferences and training events that repeatedly make the link between research and practice. Although conference programmes can still sometimes reveal a split between research and practitioner interests (e.g. with workshops being labelled as intended for one or the other audience), in our view there is ample opportunity for practitioners working with sex offenders to be exposed to the best available research, and we have experienced little hostility to the notion of evidence-based practice among the many practitioners with whom we have interacted.

In the remainder of this chapter, we will summarise what we consider to be the messages from the “best available research” and the aspects of typical practice to which they refer. As many programmes contain significant components that could be described as evidence based, it is possible that the lack of robust findings of effectiveness are due to one of three reasons. First, it may be that there is simply a lack of robust controlled evaluation designs, as has been argued by many researchers, and that large-scale high-quality studies would reveal effective approaches to sex offender treatment. Second, it is possible that our best available knowledge is still incomplete, and there are treatment targets or techniques still to be evidenced that would improve overall outcomes. Third, it is possible that current treatment approaches do contain evidence-informed components, but the success of these components is offset by poor quality delivery (e.g. inadequately collaborative or supportive) or by a hostile treatment context (e.g. programmes that take place in the unsafe environment of prisons) or wider societal rejection of those convicted of sex offences.

Bearing these issues in mind, we will proceed to discuss what the best available evidence tells us about effective approaches to sex offender treatment. Our assessment of the evidence for effective treatment below relies heavily on four recent comprehensive reviews of treatment outcome studies (Dennis et al., 2012; Hanson et al., 2009; Långström et al., 2013; Schmucker & Lösel, 2009).

We have categorised this discussion according to the three principles of the risk-need-responsivity (RNR) model of correctional rehabilitation (Andrews & Bonta, 2006): who does sex offender treatment seem to work best for, what targets should be addressed by treatment programmes and how should treatment programmes respond to the particular needs of the client group? The RNR model, if followed, has been shown to improve assessment and rehabilitative efforts with those convicted of criminal behaviour. Importantly, it has been demonstrated that RNR is applicable to the assessment and treatment of sexual offenders (Hanson et al., 2009); Hanson et al. were able to demonstrate through reviewing 23 treatment programmes that greater adherence to these principles was met with better reductions in recidivism. Hanson et al. found that those programmes that followed only one or none of the principles had little effect on recidivism while those that followed two or three had the largest impact.

There are other suggestions in the meta-analyses that treatment works better for some individuals than others in terms of other variables than those covered by the RNR principles, but we are not yet at a position where we can say definitively that some people will or will not benefit. For instance, Schmucker and Lösel (2009) found no difference in treatment effect for those who entered treatment voluntarily compared to those who were mandated to treatment. They also found that younger participants fared better, but this finding was confounded with treatment type; younger participants were more likely to have received multisystemic therapy which focuses efforts on improving the offender’s family and social systems; older participants were more likely to have received cognitive behavioural treatment, focusing on changing attitudes and behaviour. One as yet unanswered question is whether treatment is more effective with men who have sexually offended against adults or those who have offended against children. The most recent meta-analysis (Schmucker & Lösel, 2009) found it impossible to perform “a sensible analysis” (p. 23) on this question.

For Whom Does Treatment Work?

The RNR model’s first core principle, the risk principle, directs that sexual offenders will require different levels of intervention depending upon the risk of recidivism that they present. A number of studies have shown that with low-risk non-sex offenders, treatment has either very little impact on recidivism reduction (Andrews & Bonta, 2006; Andrews & Dowden, 2006) or, in some cases where treatment is intensive, recidivism rates can in fact increase (Andrews, Bonta, & Hoge, 1990; Andrews & Dowden, 2006; Bonta, Wallace-Capretta, & Rooney, 2000; Lowenkamp & Latessa, 2002; Lowenkamp, Latessa, & Holsinger, 2006), and it has been argued that this is likely to be true for sex offenders as well (Wakeling, Mann, & Carter, 2012).

Hanson et al. (2009) examined the applicability of the RNR principles to sex offender treatment but found that the odds ratio for those programmes that targeted only higher-risk offenders (defined as “higher risk than average” (p. 871) but not linked to scores on any particular risk tool) was not significantly better than the odds ratio for programmes that targeted all risk groups. They noted that the risk principle was the weakest of the three RNR principles and concluded that “the magnitude of these differences is sufficiently small as to be of little practical value in most settings” (p. 884) but that “noticeable reductions in recidivism are not to be expected among the lowest risk offenders” (p. 886).

A stronger effect by risk was reported by Schmucker and Lösel (2009) who stated that: “… the results revealed a clear picture. The higher the risk for reoffending, the higher the resulting treatment effect. Treatment for low risk participants showed no effect at all” (p. 24). Although there may need to be exceptions, e.g. those low-risk but high-criminogenic need offenders (Carter, 2014), the best available evidence seems to suggest that treatment programmes are best targeted at higher-risk offenders.

Treatment Targets

The RNR principles provide an important framework to help consider the strengths of sex offender treatment. The need principle requires that criminogenic needs (dynamic risk factors that are amenable to change) are assessed then targeted in treatment. The four most recent comprehensive reviews (Dennis et al., 2012; Hanson et al., 2009; Långström et al., 2013; Schmucker & Lösel, 2009) of sex offender treatment programmes all concluded that as a consequence of mixed findings of effectiveness, treatment providers should ensure that their programmes are focused on issues that have been shown to have strong links with recidivism. For example:

Attention to the need principle would motivate the largest changes in the interventions given to sexual offenders…Consequently it would be beneficial for treatment providers to carefully review their programs to ensure that the treatment targets emphasised are those empirically linked to sexual recidivism. (Hanson et al., 2009, p. 886)

It is fortunate that, perhaps more so than with any other type of criminal behaviour, the factors associated with sexual recidivism have been extensively researched, mainly by Karl Hanson and his associates. There have been several high-quality large-scale studies and meta-analyses of the predictors of sexual recidivism (e.g. Hanson & Bussière, 1998; Hanson & Harris, 2000; Hanson & Morton-Bourgon, 2005; Helmus, Hanson, Babchishin, & Mann, 2013; Mann, Hanson, & Thornton, 2010; see Table 7.1 below).

Table 7.1 Psychological risk factors for sexual recidivism (from Mann et al., 2010)
Full size table

The importance of focusing on these factors as priority targets for treatment is well understood by the practitioner community, and in the last 5 years particularly, considerable effort has been made to bring treatment programmes in line with the outcomes of this research. The key elements of programmes from the 1980s through to the 2000s (Mann & Marshall, 2009) comprised (1) encouraging treatment participants to “take responsibility for their offending”, (2) developing victim empathy and (3) relapse prevention, where treatment participants are trained to anticipate high-risk situations and develop plans to avoid or control them (Laws, 1989).

McGrath et al.’s survey of treatment providers in 2010, and several essays critiquing these treatment goals (e.g. Mann & Barnett, 2013; Maruna & Mann, 2006; Ware & Mann, 2012), encourages those providing treatment programmes to place less emphasis on these areas. We do not yet have sufficient research to justify removing these components from treatment programmes altogether, however. Participant feedback studies typically find that treatment participants value these components of treatment more than any other (Levenson, Macgowan, Morin, & Cotter, 2009; Levenson, Prescott, & D’Amora, 2010; Wakeling, Webster, & Mann, 2005) and on this basis, it may be premature to disregard them altogether.

We can conclude that the best available evidence provides clear direction on the treatment targets for an evidence-informed treatment programme. Table 7.1 summarises a review of the risk factor literature by Mann et al. (2010) and presents a list of the best-evidenced risk factors as well as those variables which have been explored in fewer studies but show promise as factors which predict recidivism and those variables which have been shown not to predict recidivism.

Treatment Approaches

High-Level Approach

The responsivity principle consists of what Andrews and Bonta (2010) term general and specific responsivity. General specificity refers to the adoption of cognitive social learning methods as being most effective in bringing about a change in behaviour. Consequently, the responsivity principle states that offenders generally benefit most from programmes that take a cognitive behavioural approach. Cognitive behavioural programmes are by no means homogeneous activities but should be based on a model of teaching both attitudes and new behavioural skills. General responsivity principles for programmes addressing criminal behaviour also stress the importance of the therapeutic relationship between the facilitator and offender as well as the use of prosocial modelling, reinforcement and other appropriate methods to modify change are highlighted. Specific responsivity requires that programmes recognise the individual needs of participants, such as their intellectual ability, cultural background and personal strengths.

With people convicted of sexual offences, three high-level approaches can be described as evidence informed on the basis of the best available evidence: cognitive behavioural programmes, pharmacological therapies such as anti-androgen treatment, and multisystemic therapy for juvenile offenders, which involves expanding the focus of treatment beyond the individual to his family, peers, school and community systems (Borduin, Henggeler, Blaske, & Stein, 1990). These approaches have not been found to consistently reduce recidivism through controlled studies (see earlier discussion), but across the meta-analyses, their outcomes are consistently superior to counselling, psychotherapy and nonbehavioural methods. For example:

Cognitive-behavioral and multisystemic treatment had larger effects than other approaches. (Schmucker & Lösel, 2009, p. 2)

In practice, it is likely that both pharmacological and psychological therapies will need to be used in unison in order to obtain the greatest benefit. (Dennis et al., 2012, p. 28)

Another popular approach to treatment is the Good Lives Model (e.g. Ward, 2002; Ward & Mann, 2004; Ward, Mann, & Gannon, 2007), which could be described as a version of cognitive behavioural treatment but with a focus on building strengths rather than addressing risk factors, which is the more traditional approach to sexual offending treatment. Good Lives programmes have a strong intuitive appeal for many therapists who prefer to take a positive and future-oriented approach to working with clients, but as yet, there has been insufficient empirically robust outcome research to demonstrate a treatment effect for this approach.

As noted above, however, sexual offending programmes that describe themselves as cognitive behavioural are by no means homogeneous approaches (Hanson et al., 2009). A broad variety of treatment techniques can be described as cognitive behavioural, and so it is necessary to consider the evidence for not just the overall model but also the specific techniques used within a programme. There is research to support methods used for some of the treatment targets listed in Table 7.1 but not all, and there has also been considerable research into treatment style. Further, there is considerable complexity in getting the context right for sex offender treatment, which is an important issue in ensuring treatment is responsive. We will consider these issues separately in more depth below.

Specific Treatment Techniques

Following the groupings of risk factors used by Mann et al. (2010) as set in Table 7.1, we will consider effective methods for (a) sexual deviance risk factors, (b) attitudinal risk factors such as pro-sexual offending attitudes and (c) self-management risk factors including the management of emotions and impulsive urges. The quality of research varies across these different issues, and in addition, it is likely that some of these risk factors are harder to change than others. For instance, it is unlikely to be possible to change a deviant sexual preference such as paedophilic preference. However, there is greater cause for optimism in relation to risk factors such as offence supportive attitudes and self-management.

Sexually Deviant Interests

Conditioning theory , which purports that behaviour experienced as pleasurable will be repeated while behaviour that is unpleasant will not, has given rise to a number of methods to address sexually deviant interests (Laws and Marshall 1990). In accordance with conditioning theory, techniques have been employed to modify deviant interests through aversive therapies that aim to negate the enjoyment of fantasies, e.g. olfactory aversion and masturbatory reconditioning techniques, e.g. directed masturbation aimed at raising sexual arousal to suitable stimuli that are nondeviant. There is no large-scale or controlled research into the efficacy of these technique but only case study reports, few of which are recent. While aversion techniques, such as covert and modified covert sensitisation, are unlikely to eradicate deviant sexual interests, some case study reports have described using these methods to help an individual manage sexual arousal related to offending, and in some cases it has been reported that benefits made can be sustained over time, e.g. Earls and Castonguay (1989). The evidence for directed masturbation, a reconditioning technique, was considered as hopeful in 1991 (Laws & Marshall, 1991), but unfortunately, no further evidence has been forthcoming since this time. The position is similar for thematic shift methods, used in conjunction with aversion techniques (Marshall, 1979) and verbal satiation (Laws & Marshall, 1991).

Therefore, although behavioural techniques have been used to attempt to modify sexual interests since the 1960s, there remains an absence of evidence to support the effectiveness of these approaches. The existing studies are characterised by being of poor quality (e.g. Marshall, Anderson, & Fernandez, 1999; Quinsey & Earls, 1990) or involve small samples sizes or single case studies (e.g. Maletzky, 1985). Existing research has also failed to isolate the role of behaviour modification techniques from the range of different treatment approaches an individual can undertake. Therefore, the effectiveness of behavioural conditioning techniques in changing sexual arousal is unclear.

While the focus of this chapter is on psychological treatments, we note that medication, particularly as an adjunct to psychological approaches, is also used to both change the nature and intensity of sexual arousal. These techniques have been shown to have the best results if used alongside psychological therapies such as cognitive behavioural programmes (Beech & Harkins, 2012; Lösel & Schmucker, 2005).

Cognitive Risk Factors

In terms of cognitive risk factors , we have examined the evidence for various treatment methods that have been employed to target cognitive factors such as cognitions about the world (e.g. the world is dangerous and uncontrollable), cognitions about others (e.g. suspiciousness, hostile attributional bias), cognitions about the self (e.g. seeing the self as damaged or disadvantaged), cognitions about sexual offending in general (e.g. sex with children is not harmful) and cognitions about one’s own offending (e.g. my victim was not harmed by the offence). From the review by Beech, Bartels, and Dixon (2013), it can be concluded that several treatment techniques could be considered to have a reasonable evidence base. First, cognitive restructuring has been shown in at least three studies to be associated with a decrease in offence supportive cognitions for child molesters (Bickley & Beech, 2003; Bumby, 1996; Williams, Wakeling, & Webster, 2007), although this technique should not be used to push treatment participants towards “taking responsibility” for their offending, because taking responsibility is not an established risk factor for sexual offending (Ware & Mann, 2012).

Second, schema therapy (e.g. Mann & Shingler, 2005) has been shown to have some success with people convicted of sexual offences in terms of leading to attitude change (e.g. Thornton & Shingler, 2001; Barnett, 2011) although both these studies emanated from HM Prison Service England and Wales and did not examine reoffending as an outcome. Studies in other settings have not yielded positive effects on measures of cognitive change (Tarrier et al., 2010; see also Eccleston & Owen, 2007)—again, reoffending has not been studied as an outcome from schema therapy.

Third, experiential techniques such as role-play of interpersonal situations where the treatment client takes on different roles within the situation can improve perspective taking, which may be effective in future potential offending situations. These techniques have typically been used in sexual offending programmes to develop victim empathy (e.g. Mann, Daniels, & Marshall, 2002; Webster et al., 2005), but Mann and Barnett argued that this use was based on insufficient evidence and carried dangers, and there is no evidence that these techniques have led to a reduction in reoffending.

Techniques to improve self-management, including emotional management and urge management, are well established in criminal justice settings. Cognitive skills training programmes are widely used for people who have been convicted but not for a sexual offence to considerable effect, especially given that they are relatively short and cheap to run. While the tradition for sexual offence perpetrators has been to eschew this type of programme in favour of offence-focused programmes, two studies have shown that cognitive skills programmes alone are associated with reduced recidivism for people convicted of sexual offences. Robinson (1995) studied 4072 prisoners referred to the Reasoning and Rehabilitation programme while in custody and found that the people convicted of sexual offences within this sample who completed the programme showed a 57.8 % drop in recidivism compared to a control group, Similarly, Travers, Mann, and Hollin (2014) studied the effect of another cognitive skills programme, the Enhanced Thinking Skills (ETS) course, on over 21,000 prisoners in England and Wales, examining impact by risk and offence type. Their sample contained about 1800 men convicted of sex offences (589 rapists and 1235 men convicted of sex offences against children), for whom ETS had been the only intervention they received (i.e. they did not participate in any specialised sexual offending treatment).

On average (although there were some differences according to risk level), the rapists who participated in ETS had a reconviction rate about 20 percentage points less than predicted, and the child molesters had a reconviction rate about 10 percentage points less than predicted. In both these studies, only general reconviction was reported, so it cannot be concluded that cognitive skills training reduced sexual recidivism. However, it appears clear from these studies that cognitive skills training is beneficial for people convicted of sexual offences.

Table 7.2 below summarises what the best-documented techniques for the various risk factors shown in Table 7.1. Of these techniques, only cognitive skills training has been shown to have an impact on recidivism. The other techniques in Table 7.2 lack sufficient evidence and can still only be described as experimental, but they do have supportive theoretical models, and they have been reasonably well described in the literature. A programme combining these various techniques, if individualised to participants, could be considered an evidence-informed approach to reducing sexual recidivism, but not an evidence-based approach. It is clear that sexual offending treatment components need to be evaluated more robustly for their impact on reoffending.

Table 7.2 Techniques to address risk factors associated with sexual offences
Full size table

Treatment Style

Alongside the programme theory, our understanding of the nature of the therapy style and the therapist-client relationship needed to encourage change has advanced and changed over the decades. The confrontational approach, originally advocated for sexual offending treatment (Salter, 1988), is now recognised to be detrimental to group cohesion and individual change. Research has shown that the anxiety raised from a confrontational style could impact upon understanding (Beech & Fordham, 1997). Instead, Mann et al. (2002) found that facilitators who showed a warm, honest and direct approach with expressions of empathy and verbal reward were associated with those groups that achieved their aims.

The treatment style that Marshall et al. identified was consistent with the principles of motivational interviewing (MI; Miller & Rollnick, 2002) techniques that were similarly revolutionising substance misuse treatment, another field where confrontational approaches had previously dominated. MI research had already challenged practitioners that confrontation when tackling addiction problems impeded change by causing defensiveness.

Building on practitioner enthusiasm for delivering treatment in a more motivational way, the Good Lives Model (GLM; Ward, 2002) proposed a theoretical framework by which sexual offending treatment can become part of a positive psychology, helping motivate treatment participants to reach the primary goods that all humans want as part of a fulfilling life. The GLM encourages primary goods to be achieved in prosocial ways, e.g. intimacy with age appropriate adults as part of seeking relatedness, rather than simply identifying the things that an offender must avoid, e.g. seeking intimacy with a child. The GLM has been adopted as a unifying framework by many sexual offending programmes. We regard the GLM as a theoretically sound model for treatment programmes which has the potential to radically change the way in which treatment programmes conceptualise their targets and relate to their clients. However, we must add the caution that despite claims of effectiveness, there have as yet been insufficient robust evaluation designs to enable the GLM to be considered an evidence-based treatment approach.

More recently, treatment providers have begun to explore more biologically informed approaches to treatment delivery. A biopsychosocial approach recognises the role that biological and social factors play in making an individual vulnerable to offending (Carter & Mann, in press). In this model, treatment engagement can be enhanced by making treatment accessible to individuals, regardless of their biological vulnerabilities. For example, developing positive and trusting relationships with facilitators and other group members may be extremely challenging for individuals who have deficits in their neurocognitive functioning. Keeping track and understanding verbal arguments, particularly if they require extended periods of concentration, can also be problematic for these individuals. Incorporating visual, auditory and kinaesthetic methods into programmes to allow a more active and less verbally dependent method of engagement in contrast to the more common talking and introspective methods of therapy could help improve responsivity and engagement.

Treatment Context

The culture in which a treatment programme is set, from the narrow culture of the immediate setting, through the wider culture of the system within which treatment is located, to the broadest level of societal culture, can affect the impact of sexual offending treatment. People with convictions for sex offences know that they are universally reviled. It is perhaps unsurprising that in this context, they are wary of treatment professionals, especially when these same professionals also often have the power to dictate or withhold their release from custody or their freedoms in the community. Furthermore, even if someone attends the best treatment programme and is agreed by all to have made excellent progress, these gains can potentially be quickly undone by experiences of public hostility, disgrace and rejection in the community.

There is evidence that treatment programmes are more effective when delivered in the community rather than in prison. Schmucker and Lösel (2009) reported from their meta-analysis that treatment in prisons failed to show a significant effect overall and noted that this finding accorded with earlier studies (Aos, Miller, & Drake, 2006; Hall, 1995; Lösel & Schmucker, 2005) as well as with the general “what works” literature. They suggested that effectiveness of treatment in prisons may be negated by contamination effects (where participants are mixing socially with more deviant peers), difficulty with the delay in transferring learning to the real world or difficulties during resettlement. Also, there can be differences in clients, offence history and therapists in prison programmes compared to community settings. However, Schmucker and Lösel also noted that inpatient hospital treatment was effective, suggesting that the iatrogenic effects may be particular to a prison environment rather than any inpatient setting. This finding might suggest that being treated away from the “real world” is not necessarily ineffective, so perhaps the key issue is that the wider setting is therapeutically rather than punitively oriented.

In the UK, there has been a long-running debate over the desirability of keeping people convicted of sexual offences in prison in separate units from people serving sentences for offences that are not sexual. The available evidence seems to suggest that this kind of separation aids participation in programmes because it removes, to a large extent, fears for physical and psychological safety (Blagden et al., 2014), freeing up “headspace” to focus on rehabilitation. In contrast, when sex offenders are integrated with non-sex offenders, they have to focus their cognitive resources on survival (Schwaebe, 2005). This is not to say that separate units for people convicted of sexual offences are entirely desirable (Mann, 2016), but on balance they may be able to provide more therapeutic environments akin to that of a hospital, as long as the staff are carefully trained.

Other features of a positive organisational context for treatment include having highly trained and well-supervised nontreatment staff who can listen to offenders, understand their perspectives and build constructive relationships; a strategy to identify and counter myths; treatment aims that are strengths-based; sensitive referral-making; clear and transparent information about treatment; use of intrinsic rather than extrinsic motivators; involvement of family members; provision of choice about the nature of treatment; and involvement of men who have completed treatment in the support of those considering it or participating in it (see Mann, 2009 for a more detailed discussion).

Conclusions

Given the significant and lasting harm sexual offending can cause, it is understandable that victims, policymakers and members of the public may expect perpetrators of these crimes to undergo treatment to stop them from offending again. Reasonable as this expectation may be, it fails to address the paradox that treatment to reduce sexual reoffending presents: that is, for many people who have committed a sexual offence, it is probably not necessary; for others it will be insufficient on its own to achieve this goal. That is not to say that effective rehabilitative programmes should be removed from a range of different measures to help with addressing the risk of sexual recidivism. However, recognition of the limitations of sexual offending treatment is important if we are to realise its strengths.

Hanson et al. (2009) concluded that “not all interventions [for sex offenders] reduce recidivism” (p. 881), but their meta-analysis and those by others all indicate that some interventions do reduce recidivism. The important challenge, therefore, is to isolate the effective components of treatment programmes and to differentiate them from the components of programmes that hinder effectiveness. In this chapter, we have set out to propose which parts of treatment that we have got right. In doing so, we have identified treatment components that could be considered as weaknesses to be removed from programmes or require further research to determine if we should continue with them.

The evidence indicates that programmes are most effective when closely bound to the principles of risk, need and responsivity. It seems reasonable to conclude that programmes addressing sexual offending risk should not be provided to those offenders who are at low risk of recidivism. There is clear guidance on what we should target in terms of need, and we have also highlighted those techniques to employ in programmes that have evidence to support their use in relation to these targets.

We have also highlighted the importance of the context that programmes are delivered in and the wider supporting environment. A stronger public acknowledgement that not all people convicted of sexual offences are high risk, and that many will desist from further offending in the absence of any psychological therapy, could help create environments and a society that better support successful reintegration back into the community or do not alienate those convicted of an offence from living in them. By not automatically viewing perpetrators of sexual offences as persistent offenders and highly risky, we may help people to see past the crime and reduce stigma and negative labelling that people who have committed a sexual offence experience. These attitudes can hinder an ex-offender’s ability to play a constructive role in society. We recognise that some perpetrators of sexual offences raise significant concerns about their risk of further offending and will need to be managed very carefully to rightly meet public protection responsibilities. Nevertheless, more accepting and supporting environments will still be of benefit to poeple who have committed offences with the most complex or difficult needs to address who present the greatest challenges to correctional staff.

In this chapter, we have argued that some aspects of current sexual offending programmes could be described as evidence informed, although few can be termed “evidence based”. The mixed findings for treatment effectiveness likely reflect a combination of poor quality studies, programme aims or techniques that lack internal coherence or consistency (e.g. a mixture of rehabilitative and punitive aims), weak adherence to the RNR model and the iatrogenic effect of hostile cultures outside the treatment environment. These issues vary in the extent to which they are in the control of treatment providers. The content of treatment programmes is obviously important but is not the only thing that determines whether a person convicted of a sexual offence is likely to desist from further offending.

Our aim in this chapter was to consider what aspects of current or typical sexual offending programmes can be considered evidence based or, at a minimum, evidence informed. Our summary shows that it is possible to select appropriate targets for a treatment programme based on robust research. However, effective treatment methods and techniques have been less robustly established or, in some cases, not established at all. It is not possible to conclude that treatment programmes for people convicted of sexual offences constitute evidence-based practice, yet, but the jigsaw is being assembled. To ensure that sexual offending treatment is as strong as it can be, we must continue to research our practice, evaluate our efforts and be prepared to adjust our approach when evidence indicates this is necessary.