Scottish Government

a Risk Assessment and Management of Serious Violent and Sexual Offenders: A review of current issues

Hazel Kemshall

This document is also available in pdf format (104k)

Main Findings

Introduction and Scope of the Report

Recent years have seen various research, policy and legislative initiatives on sexual and violent offenders. In the Scottish context the most significant for this review are:

In the United Kingdom the 1990s saw a growing preoccupation with public protection from the serious harm posed by sexual and violent offenders, and also a desire to respond more effectively to the growing risk presented by paedophiles (Grubin 1998). The cumulative effect has been the pursuit of both legislation and policy for the preventative sentencing and selective incapacitation of 'serious' sexual and violent offenders, and more robust systems for the monitoring and surveillance of such offenders in the community. The review draws on current research literature and presents an overview of:

Risk Assessment

There are two basic approaches to risk assessment: clinical and actuarial.

Clinical and actuarial assessment methods each have advantages and disadvantages. Clinical methods have lower levels of accuracy and are open to the subjective bias of the assessor but have much to contribute in understanding behaviours, environmental stressors, and in establishing treatability and management plans. Actuarial methods have greater predictive accuracy, but can be flawed by the 'statistical fallacy' and low incidence of risky behaviours in the population as a whole. Combined methods are increasingly advocated as the means to increasing the defensibility of risk decisions and formalised inter-agency working is increasingly seen as beneficial to information sharing on risk.

Risk Assessment Tools for Sexual and Violent Offenders

Risk assessment tools for sex offenders have been developed and evaluated primarily in America, Canada and the UK. In England the SACJ has been the most used. It has assisted the prison service in assessing and targeting sex offenders for treatment groups, and the police service in their assessments of sex offenders subject to the sex offender register. Recently the SACJ has been refined into MATRIX 2000, providing greater accuracy without compromising the dynamic aspect of the SACJ.

Numerous tools inform the risk assessment of violent offenders. The key requirements are reliability and accuracy, but also information that would assist case managers in drawing up and implementing risk management plans. The VRAG is the most commonly used and has the widest applicability. The PCL-R has a more restricted purpose targeted at psychopathy, but has been usefully integrated into other assessment tools as appropriate such as the VRAG. The HCR-20 offers additional clinical and risk management information to case managers tasked with treatment or case planning, and is increasingly preferred by criminal justice personnel because of this 'value-added' component.

More recent approaches offer an interesting combination of tools and classification trees, but remain at a largely evaluative stage and do not capture the medium risk classification of concern to practitioners. It is likely that assessment tools will continue to develop and that further tools will be introduced in due course as both research and practice develop. It is suggested that MATRIX 2000 and the HCR-20 are likely to have the most relevance to practice and duties arising from the MacLean recommendations and recent legislation in Scotland.

Risk Management

Cognitive-behavioural programmes have been the most successfully evaluated for the effective treatment of both sexual and violent offenders. Appropriate targeting and matching is also emphasised, and the integration of such programmes into broader strategies of risk management is advocated. Strategies that emphasise the promotion of internal controls, with the imposition of clear external ones are increasingly stressed as the key to the successful risk management of high-risk offenders in the community. Intensive supervision, comprising monitoring, surveillance, and enforcement of rules and sanctions coupled with cognitive behavioural intervention programmes are the features of such high-risk management strategies. A recent initiative is 'Circles of Support'. This approach recognises that many sex offenders are social isolates and literally provides a circle of support in the community for the offender once released from prison or treatment centre. Such circles are made up of volunteers with whom the offender will have significant contact (for example church leaders and mentors), and in addition to social support the volunteers are trained to identify 'warning signals' of relapse as well as informing the statutory authorities should the offender's behaviour warrant it. At present there are two pilots in the UK and long-term evaluation is awaited.

Conclusion

In selecting any method is it is essential that consideration be given to:

In addition, consideration should be given to the tool's use in guiding risk management plans, and in assisting any subsequent review process. For sex offenders, the newly introduced MATRIX 2000 has out-performed other tools and is subject to on-going evaluation. For violent offenders, the VRAG is the most accurate and most widely used, although the HCR-20 provides added value in terms of identifying dynamic factors for case intervention and treatment.

In terms of risk management, integrated risk management strategies, combining both cognitive-behavioural programmes and intensive community supervision, have the best outcomes.