Create your own Brochure

[Skip to content]

Search our Site
Monday 23 April 2018

An Evaluation of the Diamond Initiative: Year Two Findings

Version: KR/RF/3/1.0

Authors: London Criminal Justice Partnership


The Diamond Initiative, which became operational at the beginning of 2009, was an imaginative criminal justice policy innovation set up by the London Criminal Justice Partnership. It had at its heart two key features. First, in England and Wales short-term prisoners (those with sentences of less than twelve months) receive no statutory supervision on release, so Diamond offered them resettlement help that they would otherwise not have received. Second, studies of the geography of crime have consistently shown that deprived urban areas suffer disproportionately from criminal victimization, and they also contain among their residents a higher than average proportion of registered offenders. Focusing a fresh crime reduction initiative on deprived urban areas – as Diamond did – therefore makes good strategic sense.

But, as we pointed out in the Foreword to the Interim Research Report on the Diamond Initiative, crime reduction initiatives do not always work, and it is vital that they are rigorously researched. The London Criminal Justice Partnership is therefore to be congratulated, not only for its courage in setting up this new Initiative, but also for its determination to evaluate the initiative fully. That task has been undertaken by the Strategy, Research and Analysis Unit (SRAU) of the Metropolitan Police, supported by an independent Academic Reference Group (ARG) of which we have been the members.

At the time of the Interim Report, no definite conclusions could be drawn about the results of the Diamond Initiative, but there were a number of reasons for cautious optimism. It was of course right, in the interests of accountability and transparency, to state these reasons in the Interim Report. But we live in a complex world where crime issues are of great political interest, and a 24-hour media culture is voracious in its search for stories. So publication of the Interim Report had the effect of increasing the already high political profile of the Diamond Initiative. Under pressure from the media and others to declare Diamond an unqualified success, SRAU researchers and members of the ARG found themselves having resolutely to insist that the research was not yet completed, and that it would be inappropriate to act prematurely on interim results.

This Final Report on the results of the research justifies the earlier caution. After a very careful and thorough evaluation, SRAU researchers have concluded that the assessment of the crimereductive potential of the Diamond Initiative must be less encouraging than appeared to be the case from the data available at the time of the Interim Report.

This report therefore carries within it the potential for a reverse danger from that of the Interim Report. Whereas the earlier report encouraged some to think in terms of unqualified success, the more disappointing results now presented may lead to careless talk of unqualified failure. Neither of these reactions is justified. In reality, the results are complex – reflecting the complexity of the real world – and they need careful interpretation to ensure that the right conclusions are drawn from them. In particular, it is important to emphasise that the results presented in this report do not mean either that the Diamond Initiative was inappropriately mounted, or that the principles of what is often called ‘integrated offender management’ are undermined by the research.

The Ministry of Justice has recently been speaking about a ‘rehabilitation revolution’. However, those who are familiar with the history of research evaluations into rehabilitative work with offenders – a history that now stretches back for half a century – tend to be uncomfortable with this kind of language. On the basis of this body of research work, several conclusions can be drawn with some certainty, but they lead to a policy stance of gradual reformism rather than revolution.

What are these conclusions? There are four. First, helping people to turn their backs on a life of crime is a slow and uncertain process. It is frequently a case of “two steps forward, one step back”, because of the complexity of offenders’ lives (including, very often, matters such as a lack of qualifications and work experience; drug and alcohol use; a tendency to revert to previous patterns of behaviour in a crisis, and so on). But second, and encouragingly, most people who become heavily involved in offending also eventually desist from crime – completely or largely – at some stage in their lives. Despite their criminal records, the evidence is clear that most of them do not wish to continue with a life of crime, and gradually start to turn away from it. Third, the most thorough research evaluations of treatment initiatives, when they show positive results, tend to report fairly small effects (in small-scale ‘demonstration’ projects run by trained specialists, stronger results are sometimes obtained, but it has proved very difficult to replicate such results when a similar programme is ‘rolled out’ on a larger scale). The main implication of these studies is that programmes of work with offenders can accelerate the desistance process, but viewed in the round, they tend to have only a modest impact. This is probably because the work of police, prison and probation officers are only one of many influences on offenders’ lives. Hence, there are several useful tools in the rehabilitation armoury – but no ‘silver bullets’. Fourth, rehabilitation programmes can easily become derailed, especially when they involve complex partnerships between several different agencies; the successful implementation of fresh and apparently promising approaches is a real challenge.

The Diamond Initiative put into practice a set of principles for which there is considerable empirical support. If (as is the case) most repeat offenders would, in adulthood, like to move away from crime, it makes obvious sense to try to support their aspirations by providing appropriate help in matters such as housing, employment, and drugs treatment. In developing such support packages, it is also obviously sensible for the police and probation services to work together, in close partnership with other relevant agencies such as housing departments and job centres. Additionally, there are very good strategic reasons to focus on providing such support to people leaving prison after short sentences: we know that such offenders tend to have extensive criminal records, and a high headline reconviction rate, so this is a group for whom the biggest returns can be expected from a successful intervention.

Given all this, the Academic Reference Group that oversaw this evaluation were surprised by the principal finding of the SRAU research – that the extent and nature of known reoffending in the experimental group was very similar to that in a matched control group. How is one to make sense of this result? The first possibility is that the project had to move forward too far too quickly, and that the scheme did not have enough time to find its feet. There is indeed some supporting evidence of implementation problems. For example, the police were able to deploy resources to Diamond more rapidly than the probation service and other partners – a finding that is not uncommon with partnership projects – with the result that the ‘prospectus’ offered to early participants failed to deliver fully on its promises of multi-agency support. Moreover, as the report states, the Initiative never fully reconciled some working tensions between the police and other partners. A further (and related) possibility is that the timescales for measuring an impact were too short. We know that it takes time – and often many attempts – for people to change ingrained but self-destructive habits such as smoking, drinking, drug use and overwork. Offending is no different, and the processes of change do not always suit the timetables of organisations operating in a highly politicised environment, and looking for quick results. The Academic Reference Group obviously asked itself whether the reasons for the lack of impact might lie in the research rather than the programme. For example, the main outcome measures, reconviction rates, might have been artificially inflated in the experimental group precisely because they were kept under closer surveillance by official agencies than the control group. Alternatively, the assumption that reconviction rates are closely related to reoffending rates may not hold up in the high-crime areas in which the Diamond Initiative was set. Possibly, too, the process of selecting a control group matched to the experimental one could have been flawed. These are certainly possibilities, but the Group’s overall conclusion was that the evaluation was conducted to a higher than usual standard of rigour, and that the lack of effect, relative to the control group, is real rather than artefactual.

However, the report also contains a more positive set of findings that we think it is important to highlight. Diamond’s support was offered on a voluntary basis to offenders serving sentences of imprisonment. They were in no way obliged to agree to the offer and in deciding whether to accept or not they were aware that many of those staffing the Initiative were police officers. Moreover, these were not neophyte offenders; on average, they had been in court and convicted on eleven separate occasions. Of course, some suspicions were expressed. But no fewer than 60% of those who received an offer of support from Diamond accepted it; and those who accepted tended to have a higher risk of reconviction and more social needs than those who did not. There can be no better testimony as to why this initiative was worth attempting, and why it should now be built upon. To adapt a memorable saying of Samuel Beckett: ‘No matter. Try again. Try better.’

What, then, are the main practical implications to draw from this evaluation? Perhaps the most important is that helping people to desist from crime involves a long-term commitment; if the police, probation service and their partners - or indeed the government - expect a return on their investment in the space of a year or less, they will very likely be disappointed. Second, the study has certainly demonstrated the feasibility of involving police officers in rehabilitative work with offenders; and we would expect the long term effects of joint working to be beneficial both for the police and for their partners in the probation service and elsewhere. Third, there is an obvious value in information-sharing across custody and community settings, yet also some serious challenges in making this happen as smoothly as it should.

Although the headline results of this report will disappoint staff on the Diamond Initiative, there is nothing here that calls into question the principle of multi-agency teams designed to plan and deliver social provision that is inclusive of and helpful to offenders. There is a great deal in this report that may help the Metropolitan Police, the London Probation Service and their partners make this vision a reality in the future. They are to be congratulated for their investment in this Initiative and this evaluation – and the SRAU are to be congratulated on the rigour with which the evaluation has been completed.

Sir Anthony Bottoms

Emeritus Wolfson Professor of Criminology, University of Cambridge (Chair)

Rob Allen

Formerly Director, International Centre for Prison Studies, King’s College, London

Roger Bowles

Formerly Professor, now External Affiliate, Department of Economics and Related Studies, University of York

Brian Francis

Professor of Social Statistics, Lancaster University

Mike Hough

Professor of Criminal Policy, Institute for Criminal Policy Research, Birkbeck College, London

Gill McIvor

Professor of Criminology, University of Stirling and Co-Director, Scottish Centre for Crime and Justice Research

Nerys Thomas

Analysis and Science Programme Manager, National Police Improvement Agency



You can download the individual PDF version of this webpage below.