1
What Works in Correctional Rehabilitation 10 Years On

Leam A. Craig1,2,3,4, Louise Dixon5, and Theresa A. Gannon6

1Forensic Psychology Practice Ltd, The Willows Clinic, UK

2Centre for Applied Psychology, University of Birmingham, UK

3Department of Psychology, Birmingham City University, UK

4School of Psychology, University of Lincoln, UK

5Glasgow Caledonian University, UK

6CORE‐FP, School of Psychology, University of Kent, UK

Introduction

The last 10 years since the first edition of this volume, the literature base of the What Works movement has witnessed seismic changes in policy, practice, and research into correctional rehabilitation, not least of all how we refer to people caught up in correctional rehabilitation. Since Gwenda Willis published her 2018 paper highlighting the ethical concerns and incongruence of labelling people based on criminal convictions, there has been increased emphasis within the research literature for person‐first language and the avoidance or de‐emphasis of offence‐based labels (e.g., ‘sexual offender’) to describe people. In several international journals covering issues of psychology and the law, authors are encouraged to be thoughtful about the connotations of language used in their manuscripts to describe persons or groups. Person‐first language (e.g., ‘persons with sexual offence histories’, ‘individual who has been adjudicated for…’ and ‘child/adolescent with sexual behaviour problems’) is generally preferred because it is often more accurate and less pejorative than terms like ‘sex offender’. Terms like ‘sex offender’ imply an ongoing tendency to commit sex offences, which of course, is inaccurate. Person‐first language is also consistent with American Psychological Association (APA, 2020) style guidelines for reducing bias in written language and for more inclusive language (APA, 2023). Similarly, the use of an offence label like ‘offender’ contradicts the aim of What Works in rehabilitation (Why call someone by what we don't want them to be? Willis, 2018). It would be all the more appealing if clinical practice modelled developments in research in how we refer to the people at the centre of our research and clinical work. It is for this reason that this second edition of the What Works book published in 2013 has been renamed The Wiley Handbook of What Works in Correctional Rehabilitation: An Evidence‐Based Approach to Theory, Assessment and Treatment, 2nd Edition.

Assessment: 10 years on

There have been developments in research and practice since the last edition of this book which provide an opportunity for mapping a service delivery response that stands some chance of meeting the needs of those in crisis. The What Works literature has expanded from the, still key, objective of reducing recidivism to include improving other aspects of the lives of those who perpetrate and those who are affected by crime. Following the 1974 Nothing Works, there was a ‘get tough’ era from the mid‐1970s until approximately 2010 where correctional reform became commonplace amid the background of a slow change in public policy on incarceration. As Jonson et al. (2024, Chapter 2) note, the pendulum is beginning to swing back to a more progressive approach to responding to crime. Amid a backdrop of mounting evidence that prisons are not an effective deterrent (Nagin et al., 2009; Petersilia & Cullen, 2015; Petrich et al., 2021), policymakers are amenable to considering new rehabilitative solutions to the crime problem being endorsed (Listwan et al., 2008).

An emphasis throughout this volume is that of individualised analysis of risk and need to complement any rehabilitative efforts in changing behaviour. Several authors highlight the importance of understanding the aetiology of the person's history, as part of a case formulation (CF) approach, before considering how and which criminogenic factors to target in order to reduce re‐offending (see Freestone & Kuester, 2024; Logan et al., 2024, Rettenberger & Craig, 2024). These processes are embedded within the risk–need–responsivity (RNR) model (Andrews & Bonta, 2010; Bonta & Andrews, 2017) which emerged as a core framework for correctional practice nearly 30 years ago to reinvigorate rehabilitation as an operating framework for agencies responsible for incarcerating and/or supervising individuals as part of a sanction (Taxman, 2024: Chapter 4). Essentially, the risk principle involves identifying who to treat and by what dosage based on the assessment of that person's risk: low‐risk people should receive less intensive treatment whereas high‐risk people should participate in moderate‐ to high‐intensity interventions. The need principle, assessed as dynamic factors, or those attributes that are malleable, focusses on those risk factors which contributed to the behaviour. The responsivity principle refers to the use of social learning and cognitive behavioural strategies to facilitate growth and development, as well as to tailor responses to individual characteristics such as maturation, motivation, learning styles, learning disabilities, mental illness, and other factors that affect receptivity and response to correctional control and services. Taxman (2024) observes that while the RNR framework is implemented across the world, embedded within the simplicity of the RNR framework are contested ideas that prevent the RNR model from realising its true rehabilitative purpose and from being implemented in routine decisions at all levels of the organisation. Taxman highlights that studies examining the RNR framework emphasise the examination of latent classes, or the clustering of risk–needs together. In doing this, a sole emphasis on risk has emerged which does not take into consideration the collective factors that result in how a person presents in terms of overall risk for recidivism. Taxman notes that most agencies acknowledge that risk and need assessment tools are seldom used to inform case plans and make several recommendations as to how contested concepts of the RNR framework can be addressed.

Linked to this is how practitioners assess risk in forensic settings in cases of young people who have offended (Bowers & McKeown, 2024: Chapter 6), people with violent offence convictions (Sebalo, 2024: Chapter 7), people with sexual offence convictions (Thornton, 2024: Chapter 8; Schweighofer et al., 2024: Chapter 9), and in the assessment of stalking threat (Coupland & Storey, 2024: Chapter 10). Common across these topics is the use of actuarial risk assessment instruments (ARAIs) and/or structured professional judgment (SPJ) risk frameworks as standardised and empirically robust methods to assess recidivism risk. The authors emphasise the importance of structured, and in the case of actuarial scales, normed data. There is also an emphasis on taking an individualistic approach to risk assessment, combining an idiographic and nomothetic approach to risk assessment (see Craig et al., 2004a). While actuarial style scales have been criticised for lacking specificity resulting in meaningless confidence intervals (Cooke & Michie, 2014), there is a great deal of research which supports the use of ARAIs, particularly in the use of persons with sexual offence convictions (Thornton, 2024). However, as Craig et al. (2004b) observed 20 years ago, the limitations of ARAIs means that, by their nature, these scales can only reveal part of a much bigger picture of risk.

An important emerging consideration in recent years relates to cultural issues in relation to forensic risk assessment. As Riemer, Olver, and Stockdale (Chapter 30) note, ethnic minorities are overrepresented in correctional services that have emerged amidst a history of colonisation. And, such minorities have a tendency to score higher on actuarial measures of sexual risk. Riemer and colleagues point out that culturally responsive assessment of risk is critical since systemic racism continues to entrench key factors theorised to increase criminal risk (e.g., work and education) and yet ethnic minorities may not inevitably be at increased risk of criminal behaviour. Consequently, thoughtful consideration and contextualisation of risk are required to ensure that risk assessment is fair and does not permeate injustices already experienced by ethnic minority groups.

Linked to the idea of thoughtful consideration of forensic risk is the importance of CF, both in the assessment and in identifying the interventions part of rehabilitation efforts. As Bowers and McKeown (2024) note, for young people in contact with the criminal justice system, their histories are typically complex and often include attachment issues, developmental trauma, as well as exclusion and discrimination, which will only be understood after a more detailed analysis of that person's history, the method of choice in most cases being CF. In a similar vein, Moore (Chapter 24) notes that a working CF is vital for informing complex inpatient work with high‐risk forensic patients. The process of understanding a person's history and organising information taken from that history to a narrative has long been recognised as a cornerstone to effective assessment and diagnoses and is regarded as one of the core...