Scope and limitations of the review

This review significantly updates to the end of 2011 the review of statistics and research first prepared for the PCA in May 2010. It therefore takes account of developments since the start of the coalition government. Broadly similar headings are covered, although now including sections on ‘Payment by Results’ and some of the outputs of the Offender Engagement Programme. The coverage of re-offending has been completely re-written because of fundamental changes to the data reports. This also applies to parts of the section on sentencing.

This review should not be seen as a comprehensive treatment of ‘what works’. The largest number of additions and changes has followed from changes in statistical series, rather than effectiveness research. There are some gaps: the review does not cover, for example, statistics and research on specific groups of offenders (e.g. women), or the full range of types of intervention. Nor does it comprehensively cover all aspects of the effectiveness evidence review presented for Breaking the Cycle. A particular gap is prevention and early intervention. However, with the change towards a more practitioner-led model of development, the expectation of a single comprehensive summary of research and analysis should be reduced.

This report is footnoted with web links to full reports or abstracts. Similarly to the 2010 document, each section is presented under the headings of: Key Sources; Key Points; and In the Pipeline. It is intended as a resource to be easily mined for further information and research, as well as verification of the information and ideas it contains. References have now been consolidated in a single list at the end. Page footnotes have been retained in order to more easily support data-mining and verification.

The report is accompanied by a spreadsheet of sources, including many consulted but not specifically referenced in this report. These are colour-coded and labelled in an attempt to distinguish between: statistics and research or analysis; trendable series and one-off reports; and reports which have been peer-reviewed or independently quality-assured in some way and those which have not. These distinctions are not always easy to make accurately, and the codings should not be taken as totally clear-cut. Additionally, some websites have also now been included.

The review has been conducted almost entirely through freely available internet sources, either in full text or abstract form. This provides good coverage of published material, including government reports. However, it does not provide complete coverage of all research and analysis conducted within Probation Trusts, or other organisations involved in rehabilitation delivery. While NOMS OEP newsletters sometimes give a window on this world, typically they are lacking in key citation information.  Access to this ‘grey’ literature may become more difficult in the world of Payment by Results.

Overview of sections II to VI

 The most recent Home Office data on crime trends show a recent flattening of the over falling trend since the mid 1990s. Despite some recent increases for specific crimes, there is – up to the time of writing – insufficient data to conclude that trends have now started to increase. Speculations about rising crime in a recession have not so far been substantiated by the best available data, which always lags somewhat behind events.  Reports due in January 2012[1] will be the first to include the period of the August 2011 riots.

The reviews of national crime statistics which have taken place under both Labour and Coalition governments – while not bringing the reliability and professionalism of national crime data into question – have pointed to the need for better presentation and transparency. In order to improve trust in impartiality, the production of crime data is to be moved away from the Home Office and located in the Office for National Statistics. There have also been moves towards increasing public access to local data. Local crime websites now use police data to enable public access to statistics at local levels, including constituencies, police forces and streets. Perceptions of crime vary from actual historical trends, and continue to be influenced by the media, despite this increasing local access to information.

Alongside falling crime rates, the prison population is still growing. Since 2008, up to the riots of August 2011, there was a slowing down of this growth. The summer disorder contributed to a 2% growth from the previous year to a total of 87,501 in September. On 2011 projections, it is uncertain whether the overall population will fall or continue to grow over the next 5 years. The most recent Offender Management Statistics confirms the overall decline of use of Community Orders and a recent decline in use of Suspended Sentence Orders, following steep increases since their introduction in 2005. The overall rise in the probation caseload was driven by SSOs and pre and post custody supervision. Strengthening of Community Orders appears not to have worked as a means of reducing the size of the prison population.

Recent data on the impact of sentencing on reducing re-offending provide a mixed picture. The ineffectiveness of militaristic regimes, organised prison visits intended to deter, and short prison sentences (particularly without rehabilitative interventions) remains well-supported. However, the 2011 MOJ Compendium of statistics makes a range of further comparisons which are less favourable to community supervision (including comparison with longer prison sentences, fines and conditional discharges). This data should be interpreted carefully. Such comparisons are always hampered by the difficulty of making true like-for-like comparisons. International comparisons do not support the view that increased imprisonment has been a cause of overall crime reduction.

Restorative justice and reparation for crime have become key elements of government policy. While the underlying evidence base does not appear to have substantially changed since the last review, recent analysis and interpretation of data – cited in the MOJ evidence review for Breaking the Cycle – appears to be now more positive.

Prison remains the most costly sentencing option. Moves in the direction of cost-benefit analysis for interventions have been made by WSIPP in the US, and Matrix in Britain. However, economic tools for assessing costs and benefits of different modes of intervention are still not fully developed. The Home Office has launched a tool for local value for money assessment to support Integrated Offender Management projects. The international evidence base for appropriately targeted cognitive-behavioural interventions remains strong and new analysis of UK reconviction data supports their use.

Since the last review, there have been major changes to the way re-offending statistics are calculated and presented. The six previous measures have been combined into a measure of proven reoffending. This includes both juveniles and adults, and cohorts are followed for 12 months plus 6 months for processing. Adjustments have been made to allow for changes in cohort composition, making geographical and other comparisons more feasible. The changes mean that the overall rates appear lower than previously, but the underlying pattern remains the same, showing a continuing overall falling trend.

Age patterns continue to show that reoffending peaks in the early to mid teens then falls. Recent falls in overall reoffending have been most marked for younger age groups (up to 29), while the most recent 12 months saw a rise for older offenders. Data on prolific offenders (more than 25 previous offences) show that the concentration of offending in this group is much more marked for adults (accounting for two thirds of offences) than juveniles (accounting for just under one fifth of offences). Well-known comparisons in reoffending between burglars (highest) and sex offenders against children (lowest) still hold. The prisoner-based cohort study presented in the 2010 Compendium shows relationships between reoffending and patterns of needs, including problems in early life, family and schooling and drugs and alcohol use.

The new statistics allow side-by-side presentation of reoffending data by type of sentence, but caveats for effectiveness comparisons remain. More in depth analysis is presented in Ministry of Justice Compendiums for 2010 and 2011, but warnings of the limitations of data matching and current administrative data are still included.  From these comparisons, community supervision seems to be more effective than less than 12 month prison sentences. The other comparisons presented do not show advantages of Community Orders over other sentences. Early estimates of proven reoffending (using short follow up of 3 months plus 3 months processing) are now used for management information, presented by Probation Trust, Prolific Offender schemes, Drug Action Teams and Young Offender Teams.

The limitations of current proven reoffending statistics to measure Payment by Results are noted in the UK Statistics Authority statement of compliance of the new statistics. Various PbR schemes are now underway. Early evaluation lessons indicate problems of: attributing causes of effectiveness (particularly in a multi-agency environment); cherry-picking; establishing sensible comparisons once there is national roll out; and the means by which any savings might translate into less use of custody. The government response to the Justice Committee report on the role of the Probation service, locates responsibility for resolving outcome measurement issues with the wider research community, rather than solely MOJ.

The government’s Green Paper Evidence for Breaking the Cycle, cites desistance as one of four elements of evidence supporting its rehabilitation policy. This review does not add substantial new research findings to the 2010 version. However, new compilations of evidence, supporting the integration of research into practice, have become available through the NOMS Offender Engagement Programme (OEP). The writings of McNeill and others, which have been a focus for the OEP, provide some underlying principles for practice development, but do not provide cut-and-dried prescriptions for practice.

NOMS at times presents the OEP as a hypothesis-testing piece of research, capable of providing ‘proof’ of the effectiveness of offender engagement. Although external research, which may deliver useful findings including reconviction analyses, has been commissioned by NOMS, the OEP should not – in itself – be seen as a research programme. It is probably better seen as (in part) a professional development initiative aimed at changing the manner in which research-informed thinking is integrated into front-line practice.

  • [1] This report is now available. It does not suggest a change from the previous flat trend in overall crime