Is there evidence to suggest that CBT interventions work

This evaluative report critically reviews research based evidence on the effectiveness of cognitive behavioural therapy (CBT) interventions in the context of Criminal Justice. This involves evaluation of a number of studies which are part of the so-called ‘What Works?’ agenda (McGuire, 2005) which aims to reduce reoffending and support probation supervision. It is found that the evidence for CBT is supportive of its use but that its effectiveness can be dependent on several factors such as relationship between the probationer and worker, while CBT programmes may mask the real life difficulties which probationers experience.

Keywords

Probation, offender, desistance, social context, research, what works, evidence-based practice

Introduction

This evaluative report critically reviews research based evidence on the effectiveness of cognitive behavioural therapy (CBT) interventions in the context of Criminal Justice. The principle of CBT as an intervention is to change the offender’s attitudes and behaviour by using cognitive behavioural techniques. These methods focus on considering and changing thinking in order to avoid patterns of behaviour that can lead to re-offending (Andrews, 1995). I am currently a Criminal Justice worker and have chosen the role of CBT as a practice intervention in probation for several reasons.

The longstanding approach to criminal behaviour across criminal justice systems has been imposition via the courts of punishment such as financial penalty or loss of liberty. Whilst this delivers retribution, punishment does not necessarily reduce levels of crime (Hollin, 2002) and an alternative is offender rehabilitation for which there is room for development. This approach seeks to bring about individual change in the offender and to reduce the likliehood of committing further crimes.

Studies reviewed are part of the so-called ‘What Works?’ research agenda which aims to investigate how to reduce reoffending and support probation supervision. The ‘What Works’ approach to offender rehabilitation has been written about since the 1980’s and outcome evaluations indicate that this approach may have a small but positive effect in reducing reoffending (McGuire and Priestley, 1995). The “What Works” research literature indicates that cognitive behavioural interventions produce the greatest effect (Vennard, Hedderman and Sugg, 1997; Hollin, 1999; Cooke and Philip, 2000; Redondo, Sanchez-Meca and Garrido, 1999).

In policy terms, probation in Scotland differs greatly from that in England and Wales. This has been due to political, legislative and cultural changes which have occurred in the last thirty years and more recently in 1991, the introduction of 100% funding and National Objectives and Standards (NOS).

In light of these issues/factors, the specific question guiding the search for literature is ‘Is there evidence to suggest that CBT interventions work in probationary practice?’

The search for evidence to address this question involved the use of databases ASSIA, Planex Cambridge and Discover. Specific journal databases were also consulted using word searches. These included The Howard Journal, Probation Journal, Criminology and Criminal Justice Journal and British Journal of Social Work. This search yielded the most appropriate and relevant research studies pertinent to the question.

Literature Review

This section critically reviews 6 studies. One reports on a Scottish study within which service users considered their ideal model of the social worker (Barry, 2000). Three studies were qualitative (Rex, 1999; McNeill, 2000; and McCulloch, 2005) while the remaining two were reviews of literature (Vanstone, 2000; and Stanley, 2009). The evidence is reviewed under several key themes.

What contributes to the effectiveness of CBT?

As an approach to working with offenders, cognitive behavioural modification presumes that people who offend are shaped by their environment and have failed to acquire certain cognitive skills or have learned inappropriate ways of behaving, thinking is impulsive and rigid as opposed to flexible. Cognitive difficulties are learned rather than innate in people who offend.

The pursuit of ‘effectiveness’ in the delivery and development of community based disposals has been observed over the last three decades. Although, there has been progress in what is effective in supporting desistance from crime through researchers, practitioners and government ministers to bring theoretical knowledge to practice research, policy and practice highlights that knowledge and success still remains in development. Recognising the breadth of academic and empirical literature in exploring this subject, the review is deliberately limited and draws on the effectiveness research and desistance literature.

Each of the studies describe a brief history of ‘What Works’ in probation. The premise was that cognitive behavioural programmes would deliver reductions of between 5 and 10 per cent in reoffending. The evidence for this proposition was based largely on American and Canadian research and meta-analyses of relatively small scale studies (Lipsey, 1992). Research studies suggested that not only was prison ineffective in reducing recidivism, but also that welfare approaches to crime reduction was ineffectual, intrusive and repressive (Newburn, 1995: Her Majesty’s Inspectorate of Probation, 1998).

Prior to the introduction of the NOS in the early 1990’s, little research into offenders’ views of supervision had been undertaken prior to the introduction of these Standards, which were informed more by the findings from research into what works from the academic or practitioner’s point of view than by concern with the under addressed notion of what works from the offender’s point of view. However, the ‘what works’ principles (which inform NOS) are being adopted throughout the various strands of work with offenders. Three of these principles are problematic when it comes to offenders’ views about supervision which are: risk classification, criminogenic needs and responsivity. Several of the studies were chosen as part of this literature review as they researched workers and offender’s views and it will be argued that a more flexible approach needs to be adopted in relation to these principles before offenders themselves can fully benefit from supervision.

With regards to the four studies carried out three involved workers being interviewed through semi-structured interviews (Rex, 1999; McNeill, 2000; and McCulloch, 2005) and one solely with probationers (Barry, 2000). Rex and McCulloch’s studies also involved semi-structured interviews with probationers as well as with workers. The two reviews of literature discuss what research has already been done.

The research explored the characteristics of probationers on supervision, documented the services of such supervision in meeting offenders’ needs and reducing the risk of further offending. In order to promote ‘analytic generalisability’ a random sample was constructed (Robson, 1993, pp.138-139) to enable proportionate representation in terms of gender, age and geography and two of the studies specified that all the respondents were white. The studies varied in size: six workers and 6 probationers (McCulloch, 2005), twelve workers (McNeill, 2000), sixty-five probationers (Barry, 2000) and sixty probationers and twenty-one workers (Rex, 1999). The two reviews of literature looked at numerous national studies. This allowed the writer to have a wide and varied consensus of effectiveness outcomes.

Offenders Perspective

Across all of the studies, the majority of probationers understood the purpose of probation and their officers were seeking, through a variety of means, to reduce the likliehood of re-offending. Few probationers regarded probation as purely a monitoring exercise, although the majority view was to keep them from offending over half expected their social workers to be proactive in helping them to address their problems and for probation to be more constructive than punishing. In relation to those who wanted help with problems or issues, many inferred that not re-offending was a precondition to receiving such help. (Rex, 1999; Barry, 2000; and McCulloch, 2005).

There was a general consensus that social workers could not stop people from offending but could only help to minimize the risks through support of a more tangible and constructive nature. Even for those respondents who considered re-offending to be a possibility, there was a majority view that it was up to the individual to refrain from future offending. However, probationers attributed changes in their behaviour to their supervisory experiences was the active and participatory nature of those experiences which was one of the features of effective community programmes previously identified by McGuire (1995).

Probationers seemed aware of the efforts required to sustain a decision to stop offending and were more willing to embark where they felt a commitment to and positively engaged in the supervisory relationship. The research carried out is conflicting with regards to personal engagement by probation officers. Over 3/4 of probationers referred to the need for probation officers to demonstrate empathy, and 1/2 reported that their supervisors’ ability to listen, show interest and understanding enabled them to open up. Yet, 1/2 also said that they appreciated more formal aspects of the supervisory relationship, 1/3 found a certain amount of distance helpful and 1/2 wanted to be treated with respect in the sense of not being judged or patronised.

In showing respect, experience and knowledge probationers viewed their probation officer as taking them seriously, were able to disclose sensitive information and took on board what their probation officers said. Nearly 1/2 of respondents reported that the quality of the relationship with their social worker would not affect their success or failure within the order, however, an equal number felt that a poor or negative relationship with their social worker could have had an impact of their ability to complete their order successfully. There were no significant differences in toleration levels between young and adult offenders. The majority stated that if they had not got on well with their social worker, they would not have turned up for appointments or co-operated in other ways and would have had tokenistic involvement.

The majority of probation respondents stressed the need to get to know the individual, background, interests, problems and/or issues and aspirations to be able to actively help resolve any difficulties. The most common thread in the studies reviewed was finding out what the problems were in an attempt to get to the root cause of the offending. Most of the respondents saw their offending behaviour as a consequence of other social or personal difficulties and that the cause itself e.g. unemployment was addressed as opposed to just the effects (offending behaviour).

Offending behaviour is prominent on the minds of people who have been on community based supervision and who are asked about their views about its effectiveness, and this has been demonstrated in the studies where offenders’ views of supervision has been sought. However, there are other problems which equally, or more importantly, need to change through social work support which must come from the offenders’ themselves.

Workers perspective

Workers who were interviewed both individually and in groups (Rex, 1999; McNeill, 2000; and McCulloch, 2005) were offered a range of potential measures of effectiveness and rate them accordingly to which they considered most relevant in judging effectiveness in probation cases. These measures were selected to encompass a wide range of possibilities based on effectiveness agenda, probation policy and literature and reflect value bases for probation (Nellis, 1995). Some workers expressed that they saw managers as concerned only with efficiency (numbers game), sentencers as sometimes caring, especially when vulnerability had been identified, policy makers as interested in value for money, the public protection agenda, reduced offending and the desire to avoid adverse publicity. In general workers identified success with outcomes such as supervision action plans fully and successfully implemented, achieving reduction or cessation in offending by tackling underlying needs and/or problems and improvements. (Rex, 1999; McNeil, 2000; Vanstone, 2000; McCulloch, 2005).

Outcome measures such as reducing reoffending, changing attitudes, increasing victim empathy and reducing probationer’s needs were more important to Scottish workers than order compliance in contrast to an earlier study undertaken by Humphrey and Pease (1992) however, these findings have to be treated with caution, given the small scale of the samples and range of scores for each possible measure shows a lack of consensus amongst workers. There was dissatisfaction with forms of monitoring which, in the past, had focused on quantitative measures which was also indentified in Humphrey and Pease’s (1992) study.

As discussed earlier, most probationers saw their offending behaviour as a consequence of other social or personal difficulties and this was also recognised by workers and several of the studies provide a clear rationale for attending to the broader social contexts in which offenders live and change. The studies underline the inter-relationship between these key contexts and the need for workers to support the complex process of desistance by addressing all of these areas (McNeill, 2000; Rex, 1999). It is also suggested that there is a reticence amongst offence focused probation managers and staff to directly address family related obstacles, social and environmental factors or other underlying issues (McNeill, 2000; Barry, 2000; McCulloch; 2005).

Rex (1999) identifies the relational element of supervision and the importance of probationers feeling valued and engaged in the supervision process although most probationers valued ‘guidance’ and ‘advice’ on how they might resolve social and personal problems rather than direct practical assistance. In contrast, Barry (2000) interpreted focus on probationers’ social problems as a weakness as this is seen to reflect a welfare as opposed to an offence focussed approach to intervention.

Although workers believed that probation should address social problems which mirrored that of probationers’, a minority expressed ambivalence on this issue relating it to organisational pressures, challenges of addressing social problems and a more robust view of what probation was about (McNeill, 2000; McCulloch, 2005). This view corresponds with the current professional climate of narrow targets, high caseloads and changing professional values and this is neither surprising or difficult to defend. However, in light of what probationers identify as important in desisting from crime it poses a problem if probation hopes to demonstrate effectiveness in assisting probationers to achieve and sustain desistance from offending.

Methods used to address probationers’ social problems generally reflect the findings from wider probation research which identify the use of talking methods, referrals to other agencies and direct support as the most common methods to address probationers’ problems (Rex, 2009; McCulloch, 2005).

What seemed evident from all of the studies was a construction of effectiveness which, while aiming for both offence and need-related outcomes, lays significance on both process and progress.

‘What Works’ and Practice

The supervision of offenders is increasingly adopting a groupwork approach using predominantly cognitive behavioural methods which are seen to effectively reducing offending behaviour. The appeal is based on research into North American groupwork programmes (Lipsey, 1992). This in turn informed the guiding principles of NOS. It can be argued that groupwork programmes often operate in a vacuum and ignores the wider social problems affecting offenders and does not differentiate between individual group members’ risk levels and needs. However, not all offenders have any or similar, problems other than their offending behaviour and even the causes and consequences of, and future risk associated with, that behaviour can be interpreted differently between offenders. Community supervision devoid of a wider focus could be viewed as a component of change from assimilation and incorporation to separation and exclusion and this criticism is aimed more at the implementers than the authors of the model who do stress the importance of addressing broader structural problems (Young, 1998 cited in Vanstone, 2000).

Changes in the way that probation officers are trained have increased fears about de-professionalisation and cognitive behavioural programmes equally are seen to contribute to the degeneration of the role into the application of techniques devised by someone else (Pitts, 1992 cited in Vanstone, 2000) although it could be argued that this contributes to a form of professionalism already missing. This practice could also undermine anti-discriminatory practice as it reduces sensitivity to the needs of women and ethnic minorities. Similar to other community supervision initiatives, the development of these programmes has only taken into account the needs of men and not those of women. Therefore, an awareness of what should change should be taken into account to ensure thoughtful and reflective diversity.

Overall, the effectiveness of cognitive behavioural programmes is based on completion effects. Programmes are long and complex with rules about participants dropping in and restarting and this in theory, means that only the full programme will be effective in bringing about cognitive and behavioural change. In contrast, those who do not complete the full programme are less likely to change. The only review that specifically addressed group work programmes was Stanley (2009) and he reviewed other reviews (undertaken in England and Wales) to come to his conclusions. He judged that there was positive conclusions for cognitive behaviour programmes and drug treatment to reduce re-offending but there was weak evidence for the effectiveness of domestic abuse programmes, employment and basic skills training, intensive supervision and unpaid work and they were unlikely to have a positive effect on re-offending. For four types of intervention the National Audit Office (NAO) judged, there was insufficient evidence to suggest that although there might be a positive impact the evidence was not strong enough , mainly due to the low quality of research to show or validate strong positive effects (Davis et al., 2008)

A main piece, although flawed, evidence is an analysis by National Offender Management Systems Research Development and Statistics Directorate (NOMS RDS) of outcomes of predicted and actual reconviction rates for accredited programmes by the probation service in 2004 (Hollis, 2007). This was based on 25, 255 cases but 6,000 had to be excluded due to poor quality data. It would be fair to say, that there was bias (although not reported) within the analysis as both offender compliance and quality of programme delivery are positive factors the bias in the selection was towards successful programmes and outcomes. There was also no comparison group and this was a major weakness in the report.

The report compared actual with predicted reconvictions for offenders engaged in the accredited programmes. It is possible that the amount of cases sentenced to accredited programmes but failed to start were not recorded and the report understates the rate of dropout. Only 40 per cent completed the programme which is nearly 1/2 of the target of 70 per cent set in the original What Works strategy but those who completed were considerably less likely to re-offend than expected and did better than those who did not complete. Allowing for the biases, this looks a positive result and demonstrates the impact and effectiveness of accredited programmes. However, not all programmes were equally successful such as Domestic Violence programmes which in general failed to show a significant reduction in reconviction. Conversely, anger management, sex offender and general offending behaviour programmes showed in excess of 10 per cent reduction in reconviction of cases analysed.

There is an assumed opposition between What Works and individual work with offenders. Work by Chapman and Hough (1998) placed cognitive behavioural programmes within the context of desistance and individual casework. The use of targets could be seen to undermine as an unintended consequence as it has led to an emphasis on completing the group element in programmes at the expense of the quality of engagement with individual offenders.

Conclusion

The findings of these studies confirm a considerable amount of evidence from both desistance and effectiveness literature about the circumstances offenders may be encouraged to cease re-offending. A recurring theme to emerge from probationers’ accounts is the difficulty sustaining a life free from crime. This explains why the professional agenda in addressing offending behaviour and minimizing risk needs to be balanced with the offenders’ agenda of addressing practical concerns. Equally important is the need to develop a close and trusting relationship between the service user and worker as is possible within the continual shift of Criminal Justice policy. If this relationship is not accepted as influential and the offender’s own agenda recognised, then social work supervision could appear to service users to be alienating and irrelevant and consequently open to criticism, non-cooperation and failure.

One major flaw of the ‘What Works’ agenda is the narrowing of practice in cognitive behavioural approaches and the limited utility of standardised programmes with minority groups of offenders. There is an assumption that female and non-white offenders are fitted into programmes that have been demonstrated to be effective with white males.

It is increasingly evident that ‘What Works’ operationalised through accredited programmes is not the only or entire solution and the challenge is to develop the strengths that are there, build on new and diverse understanding of offenders and link these to the literature on intervention effects to enhance knowledge of what works when, for whom and under what circumstances.

Appendix

Paper

Aims

Methods used to collect data

Findings

Relevance of evidence to answering search question

Rex, S (1999) Desistance from Offending: Experiences of Probation. The Howard Journal, 38:4, pp.266-383

To make connections between two different research fields: the ‘What Works?’ literature on the effectiveness of community programmes in bringing about a reduction in reoffending; and the criminal careers literature, the discussion of the wider social processes by which people themselves come to stop offending.

Qualitative – Semi-structured interviews with a group of probationers.

Reports on the extent to which probationers who were interviewed linked their experiences to a reduction in their offending and considers the implications for how probation officers might be able to motivate and assist moves towards law-abiding conduct.

Examines desistance from the point of view of a group of probationers and their supervisors.

McNeill, F (2000) Defining Effective Probation: Frontline Perspectives. The Howard Journal, 39:4, pp. 382-397

The research study attempted to explore workers’ definitions of effective probation.

Qualitative – semi-structured individual and group interviews.

Evidence of considerable diversity of opinion. Generally workers preferred to measure effectiveness in terms.

Cautious optimism about the capacity, opportunity and motivation of criminal justice social workers to rise to the challenge that the effectiveness agenda represents.

Barry, M. (2000) The Mentor/Monitor Debates in Criminal Justice: What Works for Offenders. British Journal of Social Work, 30, pp.575-595

Views from probationers and ex-prisoners about social work supervision in England, Scotland and Wales.

Literature Review drawing on a range of studies from last five years.

Robinson, G (2001) Power, Knowledge and ‘What Works’ in Probation. The Howard Journal, 40: 3, pp.235-254

To examine the To To examine the implications of ‘what works’ in the context of probation, both at the level of professional practice and the broader level of the power base as a whole

McCulloch, T (2005) Probation, Social Context and Desistance: Retracing the Relationship

A small scale Scottish study which draws on participant perspectives and explore the attention given to probationers’ social contexts in supporting desistance from crime.

Qualitative -small scale research study

Stanley, S (2009) What Works in 2009: Progress or Stagnation? Probation Journal, 56: 2, pp153

Reviews evidence produced on the effectiveness of evidence-based interventions (What Works) and in particular cognitive behavioural programmes.

Order Now