The emperor’s new clothes?

There has been a growing, if rather belated, recognition of the extent of sexual offending toward both adults and children in society. Recent examples of this would include the reports into large-scale and systematic child sexual abuse in Rotherham (Jay, 2014) and the serial sexual abuse by Sir Jimmy Savile OBE (Lampard & Marsden, 2015). In turn this has led to the establishment of a series of large-scale police investigations alongside a major independent inquiry in the UK, chaired by Justice Lowell Goddard (Independent Inquiry into Child Sexual Abuse, 2015).

This growing awareness has led to policies designed to reduce the risk of sexual offending that have largely involved incapacitation, deterrence, and rehabilitation. These were matched by the provision of generous and increasing funding, much of which went to correctional settings from the 1980s onwards. The numbers of those detained as a result of convictions for criminal sexual behaviours increased, as did the relative length of time spent in custody. Harsher sentencing guidelines and the use of extended sentences were implemented along with, at least until recently, increased use of various forms of indeterminate sentence. Arguably this served to protect the public from those presenting the highest risk simply as a result of their removal from wider society. The deterrent effects of harsher sentencing remain unclear, as do efforts at rehabilitation in correctional settings.   

Correctional settings (prisons, young offender institutions, secure mental hospitals, and so on) have seen a dramatic growth in their funding and staffing for rehabilitative work with sexual offenders. This growth happened in parallel with the increasing focus on incapacitation and deterrence and the resulting detention of larger numbers of convicted sexual offenders. It also took place at a time when public services were becoming increasingly market based, with the use of 'managerialist' target-driven approaches becoming dominant. Psychologists came to the fore in the rehabilitative efforts with sexual offenders in corrections. The changes in public policy greatly influenced the application of psychology in this area of practice and did so, it has been argued, in ways that have not always best served the public interest (Towl, 2010).

Early efforts at rehabilitation with sexual offenders involved a wide range of approaches, on the basis that there was a very limited understanding of the causes of sexual offending or the clinical needs of those showing such behaviours. Interventions were developed and delivered, generally on a multidisciplinary basis, by those with substantial and extensive practice experience, along with high levels of professional skill and autonomy (Hopkins, 1993). Work in this area was understood as being experimental in nature. Early reviews of these initial trials frequently involved convicted sexual offenders and those detained in correctional settings. The results were seen as being mildly encouraging but by no means definitive, with the evidence base suggesting that around half of the reported trials showed modest positive effects but with the proviso that this was based on a body of studies with methodological weaknesses (Polizzi et al., 1999).

Where are we now?

In line with the mixed research findings, early efforts to work with the subgroup of sexual offenders detained in correctional settings were cautious and were framed in terms of the notion of 'What works?' (Towl, 2010). This began as a question to be addressed but quickly came to be associated with bold and often poorly founded claims about what was 'known' about the efficacy of psychological interventions with sexual offenders. Notions of 'What works?' appealed both in political and policy terms. In much of the literature, the question within the term was prematurely dispensed with, in favour of often inadequately developed assessments and interventions (Crighton & Towl, 2007; Hart et al., 2007).

In turn these notions led, in corrections, to increasingly mechanistic and standardised psychological interventions. Historically these have been designed to modify thoughts and beliefs and sometimes physiological arousal. The overall aim has been the development of alternative behaviours in sexual offenders. There has though been a significant narrowing of the theoretical base, with an increasing predominance of a single modified cognitive behavioural approach and the use of some behaviour modification techniques with a subset of offenders.

A specific technical language involving a curious mixing of psychological and managerial jargon, similar to the Newspeak described by George Orwell in 1984, also developed to describe this process. Sexual crimes became 'offending behaviour' and specific groupwork based psychological interventions became 'programmes'. Medical terminology was often adopted, with these psychological interventions being described as 'treatment' and the number of therapeutic sessions being described as 'dosage'. The term 'criminogenic' was adopted from criminology but the meaning was altered to emphasise the role of individual characteristics, rather than the social, political, economic or other factors associated with sexual offending. These developments had a clear marketing appeal and were resonant with the growing managerialism in, and marketisation of, service provision in corrections. The ability to 'treat' and by implication potentially 'cure' individuals in secure settings, separated from wider society, might fairly be seen as easier than addressing wider public health and other issues that underpin sexual offending (Thomas-Peter, 2006).

Broadly speaking, social democratic principles had dominated public policy in the UK from the immediate post war period until the late 1970s, with a broad consensus around the development of greater welfare provision, along with goals of reducing health, social and economic inequalities. The 1980s onwards saw a clear break in this consensus, with moves to place choice and markets at the heart of the allocation of public resources. Monopolistic state provision of services came to be seen, with some justification, as being inefficient, bureaucratic and unresponsive. Moves towards a mixture of public, private and charitable activity, along with efforts to increase choice and self-help, was advocated as the solution to these concerns.

This new orthodoxy was reflected in the introduction and growth of privately run institutions or services in corrections (Farnham & Horton, 1996). In turn, the role of managers in areas of public sector provision progressively changed, from that of managing institutions and services, to one of managing contracts for the delivery of centrally specified targets. Such changes have been described as the New Public Management (NPM: Nielsen, 2014). This approach can be seen as a rather simplistic application of the principles of market-based businesses to public services. The model used, it could be argued, was more akin to the state-based capitalism seen in Eastern Europe in the 1970s, where central targets were set for a number of units of production and the role of state industries was to meet these planned targets.  

This approach to the management of public services was associated with a period of very rapid growth for psychologists in corrections. Highlighting the asserted efficacy of particular structured assessments and interventions to identify and reduce the risk of reconviction for sexual offending served a number of largely political purposes. It arguably gave a further justification for the increased use of indeterminate sentences, as offenders' levels of risk of reconviction could be assessed, at least in Newspeak terms, and their progress to release could be based on the 'completion' of specific 'treatment' in custody to reduce this assessed risk based on an individual's 'criminogenic factors'. It also provided a clear basis of bids for additional funding based on 'products' that could be delivered as units and measured accordingly as business 'metrics'. The purchasing of such 'products' therefore grew markedly and led to related processes of standardised assessment, allocation and interventions. Targets for completion could be simply set out as part of a regular planning cycle and in turn measured and met. In this manner ever more sexual offenders could be 'rehabilitated' for release back into the community.

This has all resulted in psychology in correctional settings becoming a disproportionately managerial, rather than clinical process, where the aim is to meet the central targets set within the managerial plan. It has also led to a cycle of growth in assessed need and in the provision defined as addressing this. There's an ethical tension too – those responsible for the planning and identification of needs should be clearly separate from the suppliers of the services.

On the whole, NPM approaches in public services have arguably contributed progressively and significantly to a lowering of levels of professional skill requirements and a diminution of professional autonomy. As such work became increasingly procedural, standardised and based on a single theoretical approach, the use of a less trained and skilled workforce could, from an NPM perspective, be justified. In turn this could be used, in combination with growing economies of scale, to appear to drive down costs and drive up numeric outputs. The marginal costs associated with 'treatment' of sexual offending, such as the professional staff involved, could in this way be progressively reduced.

The use of NPM in corrections also raises further ethical and professional issues. The model used for allocation to the 'treatment' is typically one of fitting individuals to standardised, largely group-based, courses. This contrasts with the normative professional practice in psychology, of tailoring interventions closely to individual needs. This may well have adverse impacts on the response to psychological interventions, but this is not captured at all within simple counting of the number of 'treatment completions'. Attendance at a given number of sessions of 'treatment' is what is measured and so becomes, at a political and policy level, what matters.

A number of questions arise from such inflexibility. For example, do we really know that group work is more clinically effective than individual work? Is the response bias for those subject to indeterminate detention simply to repeat what courts and tribunals want to hear, in the hope of release? Within the New Public Management framework, such issues have been largely neglected in favour of considerations of the 'metrics' that drive the process, such as throughput. To date there has, with some notable exceptions, been little in the way of adequate discussion or professional analysis of the effects of this managerial model on psychological practice (Brown et al., 2015; Towl, 2010).

A false sense of security?

Increasingly, though, there has been a growing recognition of the failings of current interventions on sexual reoffending (Crighton & Towl, 2007; Hanson, personal communication cited in Rice & Harris, 2013).

Sexual offenders in correctional settings are a highly selected group and, it might reasonably be argued, are unlikely to be typical of the larger group of all sexual offenders. As a group, though, only a minority of this subgroup will go on to be reconvicted. Reconviction rates are of course a proxy measure of true rates of sexual reoffending. It does, however, seem evident that for many of this subgroup, the deterrent effects of conviction and community supervision appear to have an impact in reducing the risk of reconviction to relatively low levels. This is illustrated in a recent study using Ministry of Justice data. Of those assessed using a structured assessment of risk (the Risk Matrix 2000) it was noted that, based on a sample of 1000 convicted sex offenders, 70 would fall into the very highest-risk grouping. Of these, five would go be reconvicted within two years and the rest would not. In total, 23 out of the sample of 1000 would be reconvicted within this time frame (Cooke & Michie, 2014). Similar findings of relatively low rates of reconviction have been reported by others (Barnett et al., 2010). Given such findings, the use of blanket standardised interventions seems a poor use of scarce resources where the aim is to reduce sexual reoffending.     

Systematic reviews have also highlighted the insufficient methodological quality of much of the evidence base. Despite around half a century of research in this area, the lack of randomised trials of efficacy is most striking and may have contributed to a lack of theoretical progress in this area. Within corrections though there is no compelling evidence of significant reductions in sexual reoffending or sexual harm to others, for those completing rehabilitation work on 'programmes' (Crighton & Towl, 2007; Hanson, personal communication cited in Rice & Harris, 2013).

The most recent systematic reviews of the area have underlined these concerns. Cochrane Collaboration reviews were undertaken in 2003 and again 2012, with the later review concluding that the implementation of sex offender 'treatment' had proceeded with little apparent regard for caveats surrounding the evidence base (Dennis et al., 2012). These would include the observation that sexual offending is both a social and a public health issue, with high incidence and prevalence levels, with a wide acceptance of considerable hidden sexual victimisation and damaging effects on the mental health of victims.

The Cochrane review identified only 10 studies of sufficient methodological quality for inclusion, involving data from 944 men. Five of these involved essentially cognitive behavioural therapy (CBT) and studied 664 men, with four studies including a no-treatment or a waiting-list control condition. One study compared CBT with standard care. The authors noted that for the largest study of 484 men, intervention was compared with a no-treatment condition and the long-term outcome data were reported. The time 'at risk' in the community was similar at 8.3 years for the treated group and 8.4 years for the control group. They note that no difference was found between the groups in terms of the risk of reconviction for sexual offences (Dennis et al., 2012). The Cochrane review team went on to conclude that current approaches could result in the continued use of ineffective and potentially harmful interventions, producing a false sense of security from the misplaced belief that 'treated' individuals are at reduced risk of reoffending. In fact, it is simply not clear that is the case, or that those who do not complete these 'treatments' are at a higher risk of reoffending.

Alternative directions

Rehabilitative work in correctional settings has taken precedence over prevention and work in the community. Correctional settings have been a comparatively straightforward setting to provide work that fits with the New Public Management approach. Group work courses are arguably more straightforward (and efficient) to set up and run, attendance is easier to ensure and course 'completions' are easy to measure and report, in turn meeting annual targets for 'output'. Possibly, partly as a result of this, funding has come to be largely focused on secure prisons and hospitals in general and high-security settings in particular. We would argue though that public protection in high-security settings is chiefly served by means of incapacitation (physical separation and containment) and deterrence, rather than rehabilitation via 'treatment'. It is by no means clear that such settings are the most appropriate to undertake psychological interventions designed to reduce the risk of sexual crimes, especially in high-security prisons where there are other more pressing clinical and public protection priorities.

Given the less than encouraging evidence base and insufficiently developed theoretical understanding of sexual offending, further, but different, experimental approaches to working with sex offenders are urgently warranted to reduce sexual offending. In line with this, randomised trials of a range of approaches appear to be needed, as a means of driving both the development of research and clinical practice.

A large majority of sexual offenders, it can be persuasively argued, are unlikely to ever be detained in correctional settings. Prevention work in the community is therefore critical if the public is to be effectively protected. Efforts at prevention of sexual offending in the first place appear to be far more desirable than punishment and rehabilitation. Prompt interventions with those who exhibit sexually harmful behaviours early in life may also provide a more promising approach to public protection, focusing resources earlier on the developmental pathway of sexual offending. Early interventions in the community also have the advantage of focusing on those in a position to cause harm for the longest period. Shifting resources for prevention and rehabilitation towards community settings would also help address risk in those who present current risks of sexual offending in the community. All in all there is, we argue, a pressing need to move resources and expertise from custodial to community settings.

There are a number of examples of community-based approaches that show potential promise. In the UK the Lucy Faithfull Foundation, the children's charity Barnardo's and others have developed a variety of preventative models (Dennis & Whitehead, 2012). Both organisations have also developed community-based work with those at an early stage of developing sexually inappropriate behaviours. Project Dunkelfeld in Germany provides a well-developed example of an approach based on a public health model of prevention of sexual offending. Here free confidential advice and intervention work based on a prevention model is available to those who have never offended, offenders who have not been caught and previously punished offenders (Beier et al., 2009).

For those who have been punished and returned to the community, the use of support networks, such as the 'Circles of Support' approach, appear to have the potential to reduce risk (Kitson-Boyce, 2014; Thomas et al., 2014). Originally developed in Canada this approach has been adapted for use in the UK by a number of probation services and has shown some encouraging results in reducing sexual offending and also a range of other criminal and antisocial behaviours, although here too randomised studies have been slow to develop.

Changes of the kind outlined above are likely to require fundamental changes to service delivery that do not, on the face of it, fit well with the prevailing management orthodoxy. They suggest a need for a higher and more broad-based level of skill than current practice. The New Public Management stress on centralised target setting, simplification, reduced autonomy and 'metrics' concerned with throughput do not map easily to these alternatives approaches. As psychologists, though, we need to acknowledge the role of current managerial ideologies in structuring our approach, and be prepared to challenge these where necessary. There is also a need to be clear about the poverty of the evidence base and the need to work towards newer and more promising approaches. These need to be driven by population-based prevention approaches rather than simply individual rehabilitation. They need to be based primarily in the most effective, rather than the most (politically) convenient, settings. In turn, this will require a shifting of resources from institutions to community-based services. Most critically, what we do needs to be underpinned by an experimental approach based on rigorous research and independent evaluation.

Meet the authors

'As Head and Deputy Head of Psychological Services for prisons and probation in England and Wales we reconfigured services nationally, increasing the number of psychological staff from around 400 to over 1000 over five years. Over recent years there has been a growing awareness of the prevalence of sexual offending and the harm caused across communities. Psychologists have worked diligently in this challenging area, but there is now a need to look afresh at how to most effectively reduce the onset of sexual offending and reduce the risk from known and unknown perpetrators.'

- Graham Towl is Professor of Forensic Psychology and Pro Vice Chancellor at Durham University. [email protected]

- David Crighton is a Consultant and Honorary Professor of Forensic Psychology at Durham University

References

Barnett, G.D., Wakeling, H.C. & Howard, P.D. (2010). An examination of the predictive validity of the Risk Matrix 2000 in England and Wales. Sexual Abuse: A Journal of Research and Treatment, 22, 443–470.
Beier, K.M., Neutze, J., Mundt, I.A. et al. (2009). Encouraging self-identified pedophiles and hebephiles to seek professional help: First results of the Prevention Project Dunkelfeld (PPD). Child Abuse & Neglect, 33(8), 545–549.
Brown, J., Shell, Y. & Cole, T. (2015). Forensic psychology: Theory, research, policy and practice. London: Sage.
Cooke, D.J. & Michie, C. (2014). The generalizability of the Risk Matrix 2000. Journal of Threat Assessment and Management, 1(1), 42–55. 
Crighton, D. & Towl, G. (2007). Experimental interventions with sex offenders: A brief review of their efficacy. Evidence Based Mental Health, 10(2), 35–37.
Dennis, J.A. Khan, O. Ferriter, M. et al. (2012). Psychological interventions for adults who have sexually offended or are at risk of offending. Cochrane Database of Systematic Reviews, Issue 12. Art. No. CD007507. 
Dennis, D. & Whitehead, H. (2012). Stop it now! UK & Ireland helpline and campaign report 2002–2012. Birmingham: Lucy Faithfull Foundation.
Farnham, D. & Horton, S. (Eds.) (1996). Managing the new public services (2nd edn). London: Macmillan.
Hart, S.D., Michie, C. & Cooke, D.J. (2007). Precision of actuarial risk assessment instruments. British Journal of Psychiatry, 190(49), s60–s65. 
Hopkins, R.E. (1993). An evaluation of social skills groups for sex offenders. In N.K. Clark & G.M. Stephenson (Eds.) Sexual offenders: Context, assessment and treatment (pp.52–59). Leicester: British Psychological Society.
Independent Inquiry into Child Sexual Abuse (2015). Terms of reference. London: Author. Retrieved 25 November 2015 from www.iicsa.org.uk/about-the-inquiry/terms-of-reference. 
Jay, A. (2014). Independent Inquiry into Child Sexual Exploitation in Rotherham 1997–2013. Rotherham: Rotherham Metropolitan Borough Council.
Kitson-Boyce, R. (2014). Prison based Circles of Support and Accountability. Nottingham Trent University Criminology Conference. 
Lampard, K. & Marsden, E. (2015). Themes and lessons learnt from NHS investigations into matters relating to Jimmy Savile independent report for the Secretary of State for Health. London: Department of Health.
Nielsen, P.A. (2014). Learning from performance feedback: Performance information, aspiration levels, and managerial priorities. Public Administration, 92(1), 142–160.
Polizzi, D.M., MacKenzie, D.L. & Hickman, L.J. (1999). What works in adult sex offender treatment? International Journal of Offender Therapy and Comparative Criminology, 43(3), 357–374.
Rice, M.E. & Harris, G.T. (2013). Treatment for adult sexual offenders: May we reject the null hypothesis? In K. Harrison & B. Rainey (Eds.) The Wiley-Blackwell handbook of legal and ethical aspects of sex offender 'treatment' and management. Oxford:Wiley-Blackwell.
Thomas, T., Thompson, D. & Karstedt, K. (2014). Assessing the impact of Circles of Support and Accountability on the reintegration of adults convicted of sexual offences in the community. Centre for Criminal Justice Studies, School of Law, University of Leeds.
Thomas-Peter, B.A. (2006). The modern context of psychology in corrections. In G.J. Towl (Ed.) Psychological research in prisons. Oxford: BPS Blackwell.
Towl, G.J. (2010). Psychology in prisons. In S. Wilson & I. Cumming (Eds.) Psychiatry in prisons: A comprehensive handbook. London: Jessica Kingsley Publishers.