House of Commons - Home Affairs

Select Committee on Home Affairs Written Evidence

CHAPTER 2 "WHAT WORKS"

2.1 "WHAT WORKS"

"What Works" principles

2.1.1 The Prison Service increasingly draws upon research about "What Works" to develop a more evidence-based approach to tackling offending behaviour to assist the rehabilitation of offenders.

2.1.2 "What Works" is a set of principles that have emerged from a large number of international research studies. Adhering to these principles can reduce re-offending on release. The main principles are:

— A well-designed programme or activity: there should be evidence to show why it will work.

— Targeting the need: the activity should change factors that are linked to the offending of those taking part.

— Assessment: accurate identification of prisoners who have a need which can be addressed by that activity.

— Delivery: there should be a balanced and proportionate approach to the way that the activity is delivered.

— Monitoring and evaluation: a commitment to monitoring the quality of delivery and long-term evaluation of outcomes.

2.1.3 The What Works approach applies to all resettlement activities including work, vocational training, education, offending behaviour programmes, drug treatment programmes, therapeutic communities and preparation for release. What Works is already well embedded in accredited offending behaviour programmes, which are available in most establishments and are being expanded.

Prison Service What Works Strategy

2.1.4 The Prison Service What Works Strategy, launched in February 2002, provides a clear direction on how to develop constructive regimes based on "What Works" principles to reduce the risk of re-offending. The purpose of the What Works Strategy is threefold:

— To provide a clear direction for the Prison Service for the further development of constructive regimes to reduce re-offending based on the What Works approach.

— To provide a framework for driving central, area and establishment planning.

— To ensure that a range of interventions meet the diverse needs of the prisoner population.

Correctional Services Accreditation Panel

2.1.5 The CSAP is an Advisory Non-Departmental Public Body, established in 1999 as the Joint Prison/Probation Accreditation Panel, to advise the Home Secretary on the accreditation of programmes aimed at reducing offending. The CSAP consists of a Chair, independent experts, and representatives of the Prison Service, the National Probation Service, and the Home Office. Appointments are made in accordance with the code of Practice on Public Appointments. The CSAP reports annually.

2.1.6 Offending behaviour programmes and drug rehabilitation programmes are accredited through the Correctional Services Accreditation Panel (CSAP). The CSAP has stringent criteria for accreditation including work on the theory behind the interventions, and evaluation. These areas are key to identifying, through evidence, that they will work. The CSAP has been successful in increasing the number of programmes accredited since 1999.

2.1.7 In order to effectively manage what offending behaviour programmes are being developed, a programme development portfolio exists. All interventions seeking accreditation from the CSAP must first be added to the portfolio. Management of this is undertaken by the What Works in Prison Strategy Board, which has a number of senior operational and policy members. The Prison Service Management Board approves changes/additions to the portfolio annually.

Other regime interventions

2.1.8 In support of What Works principles, an Effective Regime Interventions Prison Service Order (PSO) was published in late 2002. This document aims to instil basic What Works principles into other interventions being run in the Prison Service that are not accredited through the Correctional Services Accreditation Panel or by outside organisations such as the National Open College Network.

Awareness of What Works

2.1.9 A key aspect of a successful rehabilitation regime is the awareness of all staff of policy and procedure. The What Works in Prison Strategy, published in February 2002, outlined what had been achieved to date and mapped where we want to be in the future. This document was circulated to all establishments and in Headquarters for staff to refer to. A series of leaflets on the key chapters of the strategy were produced and distributed to all staff in the Prison Service. The series continues to grow, offering bite-sized pieces of key What Works information to staff.

2.1.10 A recent review of the strategy has identified that most of the proposals have been implemented. As a result work has started on a joint What Works strategy with the National Probation Service. This highlights how far we have come, and our increasing joint approach to developing new policy with the National Probation Service.

2.1.11 In all areas of resettlement, the development and delivery of interventions is informed by supporting evidence of their effectiveness. This has led to significant, ongoing evaluation of existing interventions, the results of which will inform the use of these particular interventions in the future.

2.2 OFFENDER ASSESSMENT SYSTEM—OASYS

2.2.1 The new national Offender Assessment System (OASys) is a risk and needs assessment tool. It identifies and classifies offending related needs, such as a lack of accommodation, poor educational and employment skills, substance misuse, relationship problems, and problems with thinking and attitudes. It also assesses the risk of harm offenders pose to themselves and others. From these assessments, sentence plans are drawn up within OASys to manage and reduce these risks. OASys helps to target interventions, ensuring that the individual offender receives those that are most appropriate, thus making them more effective and reducing re-offending and risk to the public.

2.2.2 OASys has been developed jointly by the Prison and Probation Services. It has been designed to provide a "seamless" system for assessment of offenders throughout sentence, whether in the community or prison. The content of OASys is based on research evidence about What Works in reducing re-offending. The OASys system was extensively piloted in the community and in prisons before implementation.

2.2.3 The key outcomes sought from OASys are:

— improved prisoner management so as to protect the public and reduce re-offending; and

— seamless sentence management through closer working with the Probation Service.

2.2.4 OASys will replace the current paper-based sentence planning system. The prisoners (aged 18 and over) for whom OASys applies are the same as those currently covered by sentence planning. The timescales for the preparation and review of sentence plans are similar.

2.2.5 In 2001, when the development of OASys was completed, the National Probation Service (NPS) decided to move to early implementation of the new tool as a paper-based system, but the Prison Service decided not to implement OASys until their IT version was ready. NPS is now rolling out an IT version (e-OASys) and 26 probation areas will have completed e-OASys implementation by the end of December 2003.

2.2.6 It was always recognised that the full benefits of OASys would not be achieved until both Services were running IT versions of OASys and were able to exchange OASys assessments electronically. Connectivity between the NPS and Prison Service OASys IT systems is due to be achieved by early 2004. At that point it is intended that all 42 probation areas will be linked to those Prison Service establishments which by then have gone live on OASys. (The Prison Service rollout will continue to late 2004).

2.2.7 Connectivity will mean that the relevant probation office has access to the latest OASys assessment in good time, prior to a prisoner's release on licence. This will make it easier to make the necessary preparations, for example in relation to accommodation or post-release drug treatment, or public protection arrangements where the prisoner presents a significant risk of harm. The OASys assessment will be reviewed and updated periodically until the end of the licence period.

2.2.8 In the Prison Service, the OASys IT application was delivered to the pilot site, HMP Preston, on 28 April 2003 and is now being rolled out across the Prison Service. Rollout is complete in the first three Prison Service Areas—the North West, South West and Wales—and is now proceeding in the high security estate and women's prisons.

2.3 OFFENDING BEHAVIOUR PROGRAMMES

I. The place of programmes

General

2.3.1 Imprisonment itself is very expensive and at present more than half of all prisoners are reconvicted within two years from release. The Government is determined that prison should not just punish offenders but should be used to reduce their re-offending. Effective programmes are a central part of this strategy alongside a range of other regime activities for prisoners such as basic skills education and drugs treatment. The interventions available for prisoners have been selected according to research evidence and these are subject to ongoing evaluation. This allows the service to develop and improve provision.

Accreditation

2.3.2 All programmes are accredited and audited by an independent panel of international experts (the Correctional Services Accreditation Panel or CSAP). The panel is sponsored jointly by the prison and probation services to advise on interventions likely (on the basis of research evidence) to reduce the risk of re-offending. On-going evaluation of effectiveness is one of the requirements for accreditation and if there is insufficient evidence of effectiveness accreditation can be withdrawn.

2.3.3 To be accredited, a programme must demonstrate to the Panel that it meets the following 10 criteria:

i. A Clear Model of Change

There must be an explicit model to explain how the programme is intended to bring about relevant change in offenders.

ii & iii. Selection of Offenders

There must be a clear specification of the types of offender for whom the programme is intended, and the methods used to select them.

iv. Effective Methods

There must be evidence to show that the treatment methods used are likely to have an impact on the targeted dynamic risk factors.

v. Skills Orientated

The programme must facilitate the learning of skills that will assist participants in avoiding criminal activities and facilitate their involvement in legitimate pursuits.

vi. Sequencing, Intensity and Duration

The amount of treatment provided must be linked to the needs of programme participants, with the introduction of different treatment components timed so that they complement each other.

vii. Engagement and Motivation

The programme must be structured to maximise the engagement of participants and to sustain their motivation throughout.

viii. Continuity of Programmes and Services

There must be clear links between the programme and the overall management of the offender, both during a prison sentence and in the context of community supervision.

ix. Maintaining Integrity

There must be provision to monitor how well the programme functions, and a system to modify aspects of it that do not perform as expected.

x. Ongoing evaluation

There must be provision to evaluate the efficacy of the programme.

Prison Service Offending Behaviour Programmes Unit

2.3.4 The Prison Service's Offending Behaviour Programmes Unit comprises 60 staff. Its role is to ensure that the Prison Service is able to provide high quality offending behaviour programmes to prisoners that will reduce re-offending. This requires it to:

— develop, identify and adapt as may be necessary high quality programmes that will meet the accreditation criteria laid down by the Correctional Services Accreditation Panel that show every likelihood of reducing re-offending;

— develop methods of training to deliver programmes [standards; competencies] and management of programmes that will ensure that the programmes are delivered consistently as designed to the high standards necessary to achieve success;

— deliver, or cause to be delivered, training for staff to manage, deliver and support programmes to the high standards necessary to achieve success;

— audit, via video monitoring, the clinical standards of delivery of programmes in, and the support of that delivery provided by, prisons to ensure that delivery is being achieved as effectively as possible;

— develop methods to monitor the progress that programme participants make during and after programmes as required by the accreditation criteria, carry out that monitoring, analyse the results, and ensure that these are (a) fed back into the future development of the programmes monitored, future programmes and the current and future delivery and management methods; and (b) are disseminated for critical review inside and outside the Prison Service; and

— ensure that long-term evaluation of reconviction evidence is conducted in the ways that will carry all interested parties, including the public, Ministers, the academic community and the Prison Service.

2.3.5 All programmes are delivered within prisons by prison staff, including in some cases, prison psychologists.

II. Funding

Adequacy of provision

2.3.6 Funding has been provided through the Comprehensive Spending Review and SR2000 to expand the delivery of accredited offending behaviour programmes (OBPs) from a target of 3,000 KPI completions in 1998-99 to 8,900 in 2003-04 and beyond, and to increase the repertoire of accredited programmes available to prisoners. The tables below show the investment and expansion in the number of programmes.


Year/£ million	1995-96	1996-97	1997-98	1998-99	1999-2000	2000-01	2001-02	2002-03	2003-04

CSR	-	-	-	-	7.3	8.4	8.9	8.9	8.9
CRP	-	-	-	-	0.7	1.1	1.5	-	-
SR2000	-	-	-	-	-	-	-	3.6	11.6

CSR—Comprehensive Spending Review; CRP—Crime Reduction Programme; SR—Spending Review

In addition for all years there is an element of baseline funding for OBPs which cannot be accurately disaggregated from establishment budgets.

England and Wales


Males and females						Number of completed programmes

	Sex Offender
Financial years	Treatment Programme (All types)		All other Programmes(2)		All accredited programmes
	Completed	KPI(1)	Completed	KPI(1)	Completed	KPI(1)
1993-94	439	-	109		548	-
1994-95	554	-	285	-	839	-
1995-96	439	-	746	-	1,185	-
1996-97	680	680	770	699	1,450	1,379
1997-98	736	671	1,918	1,569	2,654	2,240
1998-99	664	589	2,837	2,540	3,501	3,129
1999-2000	669	585	4,355	4,079	5,024	4,664
2000-01	851	786	5,665	5,200	6,516	5,986
2001-02	867	839	5,842	5,566	6,709	6,405
2002-03	948	879	6,419	6,227	7,367	7,106

(1) KPI completions—these are the actual completions that have been adjusted by the Implementation Quality Rating To give the figure which can be counted against the KPI.

(2) other accredited programmes is made up of reasoning and rehabilitation, Think First Inside (also known as problem solving), thinking skills/enhanced thinking skills, CALM, and CSCP.

III. Programmes

2.3.7 The Prison Service has steadily increased its portfolio of offending behaviour programmes. These are designed and delivered in accordance with evidence-based research as to what is likely to be effective in reducing reconviction.

2.3.8 The repertoire includes two general programmes, Enhanced Thinking Skills (ETS) and Reasoning and Rehabilitation (R&R) which are designed to help offenders develop skills in effective problem solving, social perspective taking and moral reasoning. There is a family of five programmes targeting sexual offending according to level of risk, stage in sentence and intellectual ability. More recent additions to the portfolio are the Controlling Anger and Learning to Manage it (CALM) programme, targeting offending associated with emotional management, and the Cognitive Self Change programme (CSCP) which targets high risk violent adult offenders. Further details on the programmes are set out in an annex at the back of this memorandum.

IV. Distribution

2.3.9 A key objective of both the OBP strategy, and the overarching What Works strategy is to ensure that programmes are distributed on the basis of the needs of the prisoner population and the priorities of the Service. In the longer term OASys will provide information on the risk/need profile of prisoners to inform such resource planning and allocation. In the meantime a sub group of the WWPSB in conjunction with Area Psychologists has developed an alternative approach to developing an information base for the strategic allocation of programmes.

2.3.10 In Summer 2001 a basic needs analysis was carried out in relation to each establishment, using a proforma developed by OBPU and Area Psychologists. This gathered information on offence type, sentence length and length of time in the establishment. Area Psychologists collated the data and, using guidance provided by OBPU, converted it into estimated need for programmes. These data were then put alongside centrally held data about the provision of programmes and completions, and analysed on an area basis, nationally, and by type of establishment. Information about other factors affecting the distribution of programmes (logistics, resource constraints, training capacity etc) was also incorporated. The needs analysis was necessarily simple and based on a number of assumptions; however, it provided some broad indications of need, provision, gaps and inconsistencies.

2.3.11 The resulting analysis provided for the first time an evidence base for a strategic approach to the allocation of funding which is being used to inform allocation of the SR2000 funding from April 2002. On the basis of the widespread need for cognitive skills programmes, they have been the focus of the greatest expansion.

2.3.12 Prisoners in 108 establishments completed accredited offending behaviour programmes in 2002-03. A list of the establishments and the number of programmes completed is attached at annexes at the end of this memorandum, but are summarised below.


2002-3	ETS	R & R	CALM	CSCP	TFI	SOTP

Gross completions	5,028	1,060	218	36	77	948
KPI completions	4,922	1,043	158	31	73	879
No of sites	92	25	12	6	1	26
No of estabs	86	25	12	6	1	26

* there may be more that one "site" in an establishment, eg a separate unit for female prisoners.

V. Expanding Range

Programmes aimed at "vulnerable and difficult groups"

2.3.13 We are also developing programmes for short-term prisoners, domestic violence offenders, a juvenile version of ETS, an emotional management programme for violent women offenders, psychopaths, as well as additions to the sex offender portfolio. Details of the programmes are at an annex at the end of this memorandum.

VI. Research

Effectiveness in reducing re-offending

Cognitive Skills Evaluations

2.3.14 The latest evaluation of cognitive skills programmes, Home Office Research Finding (206) found no difference in the reconviction rates for prisoners who had participated in either an ETS or R&R programme between 1996 and 1998 and a matched comparison group. This second study was in contrast to an earlier study, Home Office Research Finding 161 (2002), which showed a reduction in reconviction for prisoners who had participated in a programme between 1994-96. Both these studies are in relation to adult male prisoners. Possible explanations for findings reported in the second study are:

— These results may merely reflect expected variation—international experience mirrors the variable reduction in reconviction rates found so far in the evaluation of prison-based cognitive skills programmes.

— The positive results of the earlier evaluation may have arisen because staff running the programme and prisoners that participated in the programme were highly motivated.

— The second evaluation relates to a period when programmes were rapidly expanded and this may have affected quality of delivery.

— The treatment and comparison group members could have differed on dynamic risk factors which were not assessed in the course of this study (such as motivation to change).

— There is evidence, from the previous study to the current study, of a shift in programme targeting towards lower risk offenders. Whilst this does not explain the results, it does suggest that there is a drift in selecting prisoners for programmes.

2.3.15 The Prison Service has been responsible for translating broad What Works principles into practice. There is no specific blueprint and so programmes have been developed on current evidence. Before we can conclude that programmes do not work further research is needed to bridge the gaps in current knowledge. Appropriate research is planned or underway and this is documented in section VIII, "Research to improve our understanding".

2.3.16 Results of a third study of cognitive skills programmes completed between 1998 and 2000 are expected to be published by the Home Office at the end of the year. This will compare the two-year reconviction rate of programme participants and matched comparisons. This study looks at programme impact for both adult men and young offenders. As with the previous study matched comparisons were matched to programme participants on static risk factors but again dynamic risk factors were not assessed for either group.

Sex Offender Treatment Programme (SOTP)

2.3.17 The results of an evaluation of the pre-accredited SOTP was published in Home Office research Finding 205 (2003). This study assessed reconviction outcomes for a prison-based sample of adult male sexual offenders who had completed a treatment programme and for a comparison group of similar offenders who had not participated in the programme.

2.3.18 SOTP showed a statistically significant reduction in reconviction for sexual and violent offences. The programme's impact appears to generalise so that although the programme focuses explicitly on sexual offending, violent offending is also reduced by programme participation.

2.3.19 Sexual offenders (in the absence of any intervention) reconvict at a very low rate. For this reason it was not possible to evaluate the SOTP using sexual reconviction rates alone. However, combining sexual and violent reconviction rates has been one way to obtain a preliminary assessment as to whether SOTP is effective in terms of reducing reconviction.

VII. Research to Improve Our Understanding

2.3.20 The Prison Service is committed to continuing to learn the lessons from any evaluation of its programmes to ensure that it continues to work effectively to reduce re-offending.

2.3.21 To improve understanding about what works most effectively with whom, the Service is investing in a comprehensive programme of research:

— A qualitative evaluation of accredited programmes, planned in 2002, investigating factors that influence programme delivery, impact and outcomes and identifying non reconviction benefits of programme participation, taking place in 2003.

— Treatment change research. Detailed consultation on the specification of this research, including members of the CSAP Panel, began in 2002. A three-year programme of research is proposed with the aim of exploring which offenders benefit most from treatment and which factors are related to this change. In particular, this research will look at the effect of static and dynamic risk factors on the short-term outcome (including prison behaviour and treatment change using psychometric measures) of programme completers. The role of group and prison/organisational factors on outcome will also be explored.

— Surveying Prisoner Crime Reduction (SPCR). A four-year longitudinal study is proposed to support a multi-modal approach to prison interventions and their evaluation. It is recognised that offenders often have a range of criminogenic needs, such as drug use, accommodation, basic skills, employment and thinking skills. The study will examine post-release outcomes in terms of how interventions might work in combination in addressing these needs.

VIII. Joint Work

2.3.22 Considerable progress has been made in the Prison and Probation Services working collaboratively in the area of programme development and implementation, as well as with the Home Office Mental Health Unit, Department of Health, and the Youth Justice Board.

2.3.23 A joint work plan has been developed between the Prison and Probation Services, the objectives of which are wherever possible to:

— Ensure joint business planning and analysis of work.

— Ensure a co-ordinated approach to the development of programmes.

— Avoid duplication, maximising shared learning.

— Provide common/shared CSAP submissions.

— Assess impact of business change proposals on each other.

— Develop a common approach to training to support flexible use of resources.

— To work towards a joint project team(s).

2.3.24 Joint work between the Prison and Probation Services currently includes:

— joint Sex Offender programme development;

— joint women's programme development strategy;

— joint domestic violence programme development strategy;

— alignment of psychometric tests;

— alignment of video monitoring of accredited programmes for audit purposes;

— joint procurement of supervision skills training;

— ETS/R&R conversion course;

— joint training for programmes (Enhanced Thinking Skills, cognitive booster, anger management programmes (CALM));

— FOR A CHANGE resettlement programme which is delivered in two stages "through the gate";

— motivational group work programme; and

— development of common booster programmes.

2.3.25 The Prison Service along with the Home Office and Department of Health are part of a tri-partite arrangement to develop programmes for those with high levels of psychopathic traits, and work is continuing with the Youth Justice Board to develop programmes for young prisoners.

IX. Further Developments to Optimise Impact of Programmes

2.3.26 The Service has increased its knowledge in implementing programmes and continues to refine its approach. Evidence on key factors impacting on the effectiveness of programmes, such as the supporting conditions for their implementation, and a greater emphasis on throughcare, has been acted upon.

2.3.27 A joint Prison Probation working group is being established to identify policy and practice issues related to optimising the effectiveness of accredited programmes, drawing on research and evidence about factors relevant to impact.

2.3.28 The implementation of OASys by both Prison and Probation Services, which is well underway, will facilitate improved targeting of programmes and other interventions, providing a comprehensive risk/needs profile against which we can match our intervention strategy. We will also be investigating how targeting for specific programmes can be refined.

2.3.29 In response to the evidence of the benefits of reinforcing the learning from cognitive skills programmes the Prison and Probation Services developed the cognitive skills booster programme, which is being implemented in a limited number of sites for the first time this year.

2.3.30 In recognition of the significance of motivation in effectiveness the Prison Service is developing (jointly with NPD) generic packages to enhance the effectiveness of programmes, starting with a motivational package, which can be used both as a pre-programme primer and as a pre-release intervention for short-term prisoners.

2.3.31 We are commissioning a literature review of the empirical evidence about implementation or delivery variables that affect the effectiveness of programmes designed to reduce recidivism.

2.3.32 We will be reviewing the location and distribution of OBPs across the estate, to track their record on delivery, turnover of staff and cost.

2.3.33 We are assessing the discrepancy between high audit scores and poor reconviction results, in consultation with the CSAP.

2.3.34 We are planning a work programme on the interface of OBPs with other interventions.

2.3.35 Overall this package of work should provide much needed information to increase our learning about how to use our investment in programmes to best effect, and tangible ways to enhance their effectiveness.

2.4 DEMOCRATIC THERAPEUTIC COMMUNITIES

Background

2.4.1 Prison Service Democratic Therapeutic Communities (TCs) offer a long-term residential treatment (18 months to two years) for offenders with specific needs relating to self-management, interpersonal relating, anti-social beliefs, values and attitudes, emotional management and functioning, and who have additional psychological and emotional needs which may hinder participation in other treatment programmes.

2.4.2 Democratic TCs provide a constructive living and learning environment in which prisoners explore and deconstruct what it is in their lifestyles that has contributed to serious criminal behaviour. The processes of group therapy, community functioning, education, and work are all contained, co-ordinated and enhanced by the TC environment. Within this setting, individuals are able to develop constructive, pro-social relationships and understand the processes by which such relationships are developed. At the same time, they explore and reconstruct cognition, feelings and behaviour in order to develop an alternative, non-offending lifestyle, including relapse prevention.

2.4.3 Democratic TCs exist in the following prisons:

HMYOI Aylesbury

Aylesbury TC has places for a maximum 22 male young offenders and is based on a separate wing within the main prison.

HMP Dovegate

Dovegate TC comprises four small therapeutic community units of 40 places each, a High Intensity Programme Unit (HIP) of 20 places and an Assessment and Resettlement Unit (ARU) of 20 places.

HMP Gartree

Gartree TC has places for a maximum of 25 male Stage One lifers. The TC is based on a separate wing within the main prison.

HMP Grendon

Grendon TC comprises two TC units of 40 places, three TC units of 42 places and a 25 place Assessment Unit.

HMP Winchester West Hill

The TC at Winchester West Hill is the first female prison TC in the UK. It has 87 places and is scheduled to begin Small Therapy Groups in October 2003, evolving into a fully functioning TC through 2004.

2.4.4 Throughout this section Democratic Therapeutic Communities will be referred to as TCs. Please note that these are different from Concept TCs used as an intervention for offenders with drug misuse problems.

Adequacy of provision

2.4.5 Spaces in TCs are limited and fully utilised. We do not know how many prisoners in total might benefit from joining a TC and OASys will help in identifying this figure.

Effectiveness in reducing re-offending

2.4.6 TCs have been developed and evaluated in both the health care system and the prison system in England, and relevant research evidence has come from both these systems. Research demonstrates effective outcomes in terms of "hard evidence"—reconviction rates and hospital readmissions—although there have also been a number of studies of clinical or symptomatic outcome. Studies of the process of change have also been carried out and have produced findings that relate to a number of the dynamic risk factors, in particular anti-social attitudes and identification with anti-social and pro-social models. Following accreditation by the Correctional Services Accreditation Panel (CSAP)[1] it is proposed that practice to evaluate the impact of TCs on reconviction rates will be standardised across prison TCs.

Reconviction

2.4.7 Much of the evidence has come from studies of two long-established TCs, Grendon Prison and Henderson Hospital. Although within the health care system, Henderson Hospital treats offenders and routinely uses both re-admission and reconviction data as outcome criteria. In reviewing the reconviction data, studies have only been discussed in detail where the control groups have consisted of untreated offenders rather than treatment dropouts.

Reconviction results from Grendon

2.4.8 Grendon prison opened in 1962 and until the opening of Dovegate in 2001 was the only prison in England run solely on TC lines. It comprises an Assessment Unit and five wings each run as a TC for around 40 prisoners. Following early studies that showed no effect on reconviction rates, research evidence of a positive treatment effect on reconviction began to emerge in the early 1990s. Better quality evidence has been published in the last six years. Those who stayed over 18 moths showed reductions in reconviction of a fifth to a quarter. Further details of the research into reconviction rates at Grendon TC are included in an annex at the back of this memorandum.

Vulnerable and difficult groups of prisoners

2.4.9 The treatment model used by Democratic TCs within the prison service provides an 18-month to two-year residential intervention and so is not available to short-term prisoners.

2.4.10 Places are becoming available for women (new TC for women opening over 2003-04) and there are very limited places for young offenders.

2.4.11 TCs aim to provide treatment to a client group with emotional and psychological needs which may hinder participation in other interventions. In summary, prisoners should have needs in the following areas:

— self-management;

— inter-personal relating;

— anti-social beliefs, values and attitudes;

— emotional management/functioning; and

— identification of additional psychological and emotional needs.

2.4.12 However, certain mental health problems will preclude entry into treatment for the reasons set out below.

Female offenders

2.4.17 The democratic TC model itself has a long and successful history of being used to treat women. NHS TCs treat both sexes, with a preponderance of women. Research at the Henderson Hospital (Copas et al, 1984), found that gender produced negligible differences in successful outcomes as measured by subsequent criminal convictions and psychiatric admissions over five years. In fact, the success rate for women was slightly higher than that for men.

Young offenders

2.4.18 Research needs to establish the validity of measures used with young offenders. Whilst a substantial research base exists indicating reliability and validity with adult male offenders (Shine & Hobson 1997), further work needs to be done with young offenders.

2.5 RESEARCH ON THE EFFECTIVENESS OF INTERVENTIONS IN CUSTODY AND THE COMMUNITY

2.5.1 Because the vast majority of offences committed are either not reported or recorded, and only a small proportion of the others lead to formal proceedings, it is extremely difficult to produce reliable statistics on the level of re-offending. Data from court proceedings can be used to gather evidence on whether those discharged from custody or commencing community penalties are reconvicted within a defined period, normally two years.

2.5.2 However, reconvictions are only a proxy measure of re-offending and this must be borne in mind when using and interpreting studies of reconviction. In conjunction with reconviction rates, studies should consider the use of other reconviction outcomes, such as severity, frequency and time to first reconviction.

2.5.3 In addition, some reconvictions relate to offences that were committed prior to the original sentence, known as "pseudo reconvictions". These are usually excluded from reconviction studies, since they are in no way related to effectiveness of sentence.

2.5.4 It is customary to use two-year reconviction rates to allow time for the criminal justice system to deal with re-offences. Most people who re-offend also do so within two years. However, some interventions could and do have a shorter-term impact and generally we know little about the likely duration (and decay) of the crime reduction effects of sentences.

2.5.5 Reconviction rates are often compared to a predicted rate, based on the types of offenders and other factors. Predicted rates make allowance for (i) any changes in the age, sex, offence and criminal history profile of discharged sentenced prisoners and persons commencing community supervision; (ii) the speeding up of justice and reduced delays in achieving a conviction; (iii) the increased proportion of recorded crime where the offender is brought to justice; as well as (iv) pseudo reconvictions.

2.5.6 There are also a number of reasons why raw reconviction rates are unlikely to be a good indicator of trends in re-offending over time. This is because they are affected by changes within the criminal justice system. Changes in police recording, police numbers and detection rates, use of pre-court disposals, charging and prosecution rates, charging standards and the speeding up of justice can all have an impact on reconviction rates, independent of levels of offending. Once the changing mix of offenders is controlled for, reconviction rates are a good indicator of the changes in the impact of custody/community penalties over time.

2.5.7 Reconviction rates in England and Wales are based on convictions for the offences covered in the Offenders Index, standard list offences only. This covers all indictable offences and some of the more serious summary offences. Hence there will be some offenders who are reconvicted, for less serious offences, who will not be considered as reconvicted under this coverage.

2.5.8 The two-year reconviction rate for adults discharged from prison in 1999 was 55.3%. For those commencing community penalties in 1999 the rate was 44.3%. These rates exclude pseudo reconvictions, and the community penalties exclude breach proceedings, that is, where a further penalty is imposed, but no further offence has been committed. The combined prison and community penalty two-year reconviction rate was 48%, a relative improvement of 3.2% on the predicted rate of 49.6%, from a 1997 baseline. Reconviction rates for community penalties are lower than those for custody primarily because they are targeted on low and medium risk offenders. Reconviction rates for both prison and community penalties are showing improved performance against the predicted rates.

2.5.9 As community punishments and custodial sentences are generally targeted on different types of offender, it is not possible to conclude definitively whether one disposal is more effective than the other. The likelihood that offenders will offend again after going through the criminal justice system is primarily determined by their characteristics and criminal history. Because factors other than the most recent sentence play a major part in determining re-offending and hence reconviction, comparing reconviction rates for different sentence types or different sentence lengths would be misleading. Sentencers have taken some of these factors into account in passing sentence.

2.5.10 Reconviction rates for prisoners vary by offence type, number of previous convictions and length of sentence. Those serving sentences for burglary or theft and handling offences are most likely to be reconvicted (around 75%) as opposed to sexual offences (18%). Only 19% with no previous convictions are reconvicted within two years, as opposed to 79% of those with 11 or more previous convictions. Offenders who are released after serving long custodial sentences (over four years) have much lower reconviction rates than those who serve shorter sentences. Only 3% of those released from life sentences are reconvicted within the two-year period. This will partly reflect the characteristics of offenders and offences attracting longer sentences but may also be because longer sentences give more opportunity for the prison service to address offending behaviour.

2.5.11 There is also some variation between the different types of community sentence. 52% of offenders receiving community rehabilitation orders without specified activities were reconvicted within two years, as opposed to 36% for community punishment orders. All of the types of community sentences showed decreasing reconviction rates compared to predicted rates.

2.5.12 It is, therefore, extremely difficult to measure the deterrent effect or effectiveness of sentences by using reconviction rates. Claims that one sentence is more effective than another cannot be substantiated without taking account of the differences in the offenders given each sentence. Information on the most significant ways in which offenders (and their risks of re-offending) differ is often not readily available to RDS.

2.5.13 The only way to assess effectiveness definitively would be to carry out a randomised control trial, where offenders convicted are allocated randomly to different disposals. The ethics of this render this impractical.

2.5.14 Reconviction rates can be used to evaluate sentencing policy, to provide a baseline against which to assess different management and treatment regimes (or interventions) for offenders, to assess the potential impact of changes in criminal policy on reconviction rates, and to provide the basis for risk prediction.

2.5.15 The effectiveness of specific interventions can be assessed by studying a group of offenders receiving the same initial disposal. The Prison and Probation Services have both introduced new offending behaviour and drug treatment programmes. For instance, as part of their sentence some offenders go on accredited programmes to improve their literacy and hence increase their prospects of obtaining employment, thereby reducing the likelihood of them returning to a criminal lifestyle. To determine the impact that programmes such as these are having on offenders' risk of re-offending it is necessary to conduct studies that control for the characteristics and previous criminal histories of offenders. In these circumstances, reconviction rates are an important indicator of the efficacy of programmes.

2.5.16 In order to assess the effect of an intervention or treatment, a sufficient sample size of offenders is necessary. To detect a "real" reduction of 10 percentage points in reconviction, a minimum of 325 are needed in each of the control and treatment groups. A 5-percentage point reduction requires a minimum of almost 1,300. Hence many reconviction studies can be hampered by the lack of availability of cases sufficient to detect real improvement.

2.5.17 Reconviction can be reasonably predicted by dynamic risk factors that give rise to antisocial behaviour. Examples of these are antisocial attitudes, skills deficits, lack of employment, family status, substance abuse, poverty or having criminal associates. Correctional programmes therefore aim to target these criminogenic needs with primary interventions to reduce the risk of reconviction. Though criminal history and other static factors are strong predictors of re-offending, future offences can be prevented only by addressing criminogenic needs or dynamic risk factors.

2.5.18 Some of the above factors are strongly interlinked and addressing one factor may help in one or more others. For example, education and facilitating basic skills may assist in discharged offenders finding employment and alleviating poverty. Some programmes can therefore address more than one factor.

2.5.19 Findings on criminogenic needs of prisoners and probationers were observed in a number of studies:

— Niven and Olangundoye (2002) found that factors linked with better chances of employment after release from custody were stable accommodation, qualifications, not having a drug problem and receiving help and advice with finding work;

— Lewis and Mhlanga (2001) and Liriano and Ramsey (2003) demonstrated that offenders were more likely to lead unstable and disorganised lives, to have low educational achievements, to have previous convictions, drug and alcohol problems and to be involved in illegal activities to earn money;

et al

— May (1999) found that among those on probation, the unemployed were significantly more likely to be reconvicted compared to employed offenders;

et al

2.5.20 Cognitive behavioural programmes have undergone considerable growth in the National Probation Service and HM Prison Service since the system of accreditation introduced in 1996. The Prison Service currently lists 13 accredited programmes, including three drug treatment programmes, and 14 programmes not yet accredited or under development, including five drugs programmes. The Probation Service lists 12 accredited or provisionally accredited programmes and another nine under development, including five that address substance misuse.

2.5.21 There are five programmes common to both Prisons and Probation: three fully accredited for use in both services (ETS, R&R and Think First), one under joint development and provisionally accredited for both services (Cog Skills Booster) and one fully accredited for use in prisons and being piloted for use in the community (CALM). In addition, among drug programmes, P-ASRO, accredited for use in prison, was adapted from ASRO, a programme commissioned by the NPS for delivery in the community.

2.5.22 Accredited programmes are currently being delivered in 112 of 137 Prisons in England and Wales. 54 of these deliver more than one programme, typically a cognitive skills programme, a sex offender treatment programme (SOTP), and in some cases a third programme.

2.5.23 All 42 Probation areas of the National Probation Service deliver a general offending behaviour programme (ie one of four cognitive skills programmes) and at least one other of an SOTP, Aggression Replacement Training (ART), or ASRO (Addressing Substance-Related Offending).

2.5.24 Losel (1995) showed evidence that some types of offending behaviour or cognitive skills interventions can produce reductions in reconviction rates ranging from six to 15 percentage points. Vennard and Hedderman (1998) showed that interventions that adhere to What Works principles of risk, need, responsitivity and programme integrity have seen reductions of up to 20%.

2.5.25 Recent published findings on offending behaviour programmes include:

et al

— Raynor and Vanstone (2001) found no differences in two-year reconviction rates between a community-sentenced treatment group participating in pre-accredited R&R and a prison-sentenced control group;

et al

2.5.26 On community offending behaviour programmes, a fair amount of evidence will be emerging in the next year.

2.5.27 Thus, a growing body of research indicates that individual prison-based interventions can be effective in reducing re-offending. However, some recent reviews suggest that a broader or multi-modal approach, addressing a number of "risk" factors, going beyond any individual intervention, is what works best (Lipsey, 1995; Gaes et al, 1999).

2.5.28 Evidence from studies in the resettlement of prisoners has been positive:

et al

— Porporino and Robinson (1992) found lower rates of re-imprisonment among offenders who completed a prison basic skills programme, compared to those dropping out or released before programme completion;

— Saylor and Gaes (1997) showed participants in prison work or vocational programmes were less likely to be arrested within 12 months after release, compared to a matched group of non-participants. There were similar results for longer term imprisonment rates;

et al

2.5.29 Given some initial positive results in the UK, it is possible to see how a carefully selected portfolio of interventions may benefit the offender in a broader sense.

2.5.30 There is an emerging consensus among researchers that a multi-modal approach to interventions is likely to be the most effective way of treating prisoners (McGuire, 2002). One of the most comprehensive reviews of the prison intervention literature concludes that all of the criminogenic needs of prisoners must be addressed if their propensity towards crime is to be successfully reduced (Gaes et al, 1999). Lipsey also noted the potential advantage of combining different interventions over the effectiveness of individual programmes.

2.5.31 Thus, a multi-modal approach involves providing prisoners with (as necessary) cognitive skills training, drug treatment, sex offender treatment and educational and vocational training and addressing their practical social needs such as family ties, housing and medical treatment. Addressing the range of prisoners' needs is certainly a challenge, and requires appropriate assessment and co-ordination. There are particular difficulties given the limited time in which the majority of prisoners remain in custody. Over 70% of adult male prisoners are currently sentenced for less than 12 months, and typically serve approximately half this time in prison (Home Office, 2002).

2.5.32 However, taking a multi-modal approach brings methodological difficulties. In particular, such studies are complex to evaluate or analyse.

2.5.33 Evidence on the success of interventions that reintegrate probationers into the community is limited. The three main areas for intervention are employment, accommodation, and drugs and alcohol. There is some work in this area but very little has been published. A notable exception is Sarno et al (2003) who found one-year reconviction rates for those attending the ASSET employment scheme 10 percentage points lower than those who were referred but did not attend ASSET.

2.5.34 Reconviction studies do vary in the quality of research design. The evaluation of interventions in the UK should be committed to improving the standards of reconviction studies. Alongside improving standards of reconviction studies there are a number of ways in which reconviction can be refined in order to provide a greater understanding of the type of benefits that can be attributed to interventions.

2.5.35 RDS is building on the evidence of effectiveness and evaluation methodology through:

— randomised control trials eg for Restorative Justice;

et al

— using short-term evaluation to inform programme development;

— developing an integrated model of reconviction using a range of monitoring and evaluation data, including qualitative-based methods;

— developing a longitudinal, multi-modal survey of prisoners during custody and on their release, to shed stronger light on "what works for whom";

— focusing on the way interventions are delivered (eg McGuire, 2002), relationships with staff and other mediating factors; and

et al

1 Grendon A wing's regime was accredited by CSAP's predecessor panel in 1998. In October 2003, CSAP is considering an application for accreditation of a core model that will apply to all prison TCs. Back



© Parliamentary copyright 2005	Prepared 7 January 2005