Evaluation of Community Payback Orders, Criminal Justice Social Work Reports and the Presumption Against Short Sentences

8 Conclusions

8.1 In this final chapter, we summarise some of the key themes emerging from the evaluation. We start by revisiting the logic model for the reforms and arguing for a closer focus on some of the 'hidden' mechanisms that link the implementation of the reforms to some of the intended short-term outcomes. We then review briefly what we have learned about the extent to which the reforms were implemented as intended and progress has been made towards key short and medium-term outcomes. Returning to the theme of the missing or hidden elements of the logic model, we conclude by identifying a number of specific areas where there is scope for further progress or improvement.

The logic model revisited

8.2 A logic model is sometimes referred to as a 'theory of change', in that it articulates the underlying mechanisms through which it is intended or expected that a given set of inputs will lead to specific outputs or activities and to short, medium and long-term outcomes. In designing a logic model, it is conventional to start at the far right hand side - with the long-term goals or ultimate purpose of a policy or intervention. In the case of these reforms, those are specified as being: a reduction in reoffending, increased reintegration and reduced 'churn' in the prison population. For those to be plausible long-term outcomes, the model suggests that one would expect to see a number of specific medium-term outcomes: the judiciary making less use of prison sentences, and more (and more highly tailored) use of community penalties; the availability of timely and high quality CJSWRs; provision of a full range of CPO requirements; and engagement, compliance and behaviour change among offenders. And for those medium term outcomes to be plausible, one would look for evidence of specific short-term outcomes, such as the Judiciary having confidence in community penalties and CJSWRs; CJSW staff having the time, knowledge and skills to produce high quality reports and the capacity to provide the elements of CPOs; and offenders understanding what is required of them, taking CPOs seriously and becoming motivated to change.

8.3 Tracing the 'causal chain' backwards to this point seems reasonable in that one can see how each set of outcomes is a necessary (if not always a sufficient) precondition for the next. Perhaps the weakness of the model, however, is its failure to account for how the all-important short-term outcomes are to be realised. There is no specified or self-evident causal chain between the implementation of the reforms - the introduction of CPOs with a full range of requirements and powers, the introduction of standardised and high quality CJSWRs, and the presumption against short sentences - and critical outcomes such as offenders taking CPOs seriously, or the judiciary having more confidence in community penalties and an intention to reduce their use of short prison sentences.

8.4 In the introductory chapter, we identified a number of other themes and objectives that were associated with the development of the reforms - for example, the calls for community penalties to be simultaneously tough, flexible and relevant, immediate, visible and publicly acceptable - and this is perhaps where those themes come back in. These are, in effect, some of the prerequisites - the earlier links in the causal chain - for some of the short-term outcomes in the model.

8.5 What the evaluation formally set out to do was to examine whether the reforms had, in fact, been implemented as intended, and whether there was evidence of the short-term (and some of the medium-term) objectives being met. In practice, much of the discussion with those closest to the reforms - criminal justice practitioners and individuals subject to a CPO - focused instead on these 'missing' or 'black box' elements of the model, and it is around these that many of our conclusions are centred. First, however, we summarise what we have learned in relation to the core aims of the evaluation.

Were the reforms implemented as intended?

8.6 The introduction of CPOs and the new CJSWR represented a major practical challenge for CJSW and the Scottish criminal justice system more generally. Even though the actual volume of CPOs increased gradually, as offences committed after 1 February 2011 started to enter the courts, of course, a great deal of work needed to be done in advance - for example, in ensuring that all relevant practitioner groups were trained and briefed; in revising the NOS Practice Guidance for CJSW; bringing relevant partner agencies up to speed, and so on. Within CJSW, this coincided with other major developments, including the introduction of LS/CMI and, in many local authorities, restructuring of management and existing service provision. This represented a challenging backdrop to the implementation of the reforms and placed considerable pressure on staff time and training provision.

8.7 It also meant that some of the more visible and familiar elements of the reforms tended to be foregrounded. As a result, some of the processes and requirements were established relatively quickly - for example, UPW and Supervision, which directly mirrored work previously undertaken under CSOs and Probation. Other less familiar and more complex elements, by contrast, received less in the way of immediate focus and attention. This was undoubtedly true of OA, which required CJSW staff to think very differently about the possibilities for supporting offenders undertaking unpaid work. It is clear that, in most areas, CJSW staff simply lacked the time to give an adequate level of attention to this element of the reforms. Other requirements - such as DTRs, ATRs and MHTRs - presented similar challenges.

8.8 As CJSW teams started to get to grips with increased volume of UPW placements (resulting, in particular, from the new Level 1 orders), there was more scope to develop thinking and provision in relation to OA - especially in areas where existing services and resources could be redirected towards OA hours. In areas without such existing provision, the development of OA has been slower.

8.9 A further factor influencing the implementation process was the pressure on training and CJSW staff time posed by the near simultaneous introduction of LS/CMI. This made it more difficult to fully address some of the nuances and complexities around particular requirements. As we have seen, these issues have still not been fully resolved - and indeed have been exacerbated by other issues such as the difficulty of accessing appropriate treatment services. As a result, the requirements of CPOs have been implemented and are all technically available. It is much less clear, however, whether they are always being used appropriately and as intended. These issues are returned to below.

8.10 In relation to CJSWRs, the scope for staff engagement was again limited by the range of other simultaneous developments and, in particular, the introduction of LS/CMI. However, the basic template was up and running in time and there was no evidence of any major disruption to the availability of reports when needed by the courts.

8.11 The introduction of PASS had little in the way of immediate practical consequences for Sheriffs or court staff, though it did introduce a requirement for reasons to be formally recorded for any short sentences passed by the court. The question of whether the simple introduction of the presumption into law actually resulted in a change in sentencing behaviour is returned to below.

What evidence is there that the short and medium term outcomes have been realised?

8.12 We turn now to a very brief summary of progress towards the achievement of key summary outcomes. These have been distilled from the original logic model and, in places, combine short and medium term elements.

The timeliness, consistency and quality of CJSWRs

8.13 The issue of the timeliness with which CJSWRs are submitted to court appears to be relatively uncontroversial. The vast majority of CJSWRs are being submitted on time. CJSW staff understand the importance of doing so and systems seem to be set up to deliver this. Sheriffs and court staff did not feel there were any systematic problems in this area. To the extent that there are any issues with timeliness, they relate to the speed with which requests for CJSWRs are conveyed by the courts and reports allocated to individual staff within CJSW. While problems in these areas were sporadic - and do not seem to be impacting on the overall ability of CJSW to meet their reporting commitments - they do cut into the time available for the preparation of CJSWRs and that may have implications for quality.

8.14 There is no doubt that the new template has delivered more consistent and highly structured reports. Both Sheriffs and CJSW staff generally welcomed the standardisation of the format across local authorities, the clearer emphasis on offending and the improved navigability of the new reports.

8.15 What is less clear is that reports are 'focused and concise' or that they consistently include high quality analysis and recommendations. Many Sheriffs and some CJSW staff feel that reports are often too long or contain duplication or unnecessary detail. This was felt to be an unintended consequence of the template itself, both because it was seen an encouraging CJSW staff to 'write to the boxes' and because of a degree of overlap between different sections. Sheriffs suggested that this made it difficult to identify critical information or arguments, while CJSW staff were concerned that the level of detail might prevent staff from developing compelling and coherent analysis and conclusions.

8.16 There certainly remains scope for improvement in key aspects of CJSWRs, both from the perspective of Sheriffs and CJSW staff involved in the participative audit. The latter identified the analysis of risk, in particular, as being of a variable quality. Sheriffs also had some concerns about this, although this sometimes reflected a degree of scepticism about the utility of the underlying concepts and tools. Sheriffs were also occasionally critical of the quality of the sentencing recommendations offered by CJSW staff, although sometimes acknowledged the challenge for report writers in being largely dependent on offenders' accounts.

The readiness and ability of CJSW staff to produce high quality reports

8.17 Although initial training appears to have been well-received and relatively systematic, follow-up has been more sporadic. It is also clear that training to date has tended to focus more on the transactional aspects of the template than on broader report writing skills.

8.18 Although the new template was not widely seen as creating problems with workloads (beyond the initial implementation period where staff were getting used to it), this may be the result of a drop in the total number of requests for reports. There was a consensus that - because of the emphasis on multiple sources and the depth of analysis required - individual reports are taking longer to prepare than SERs. While this was generally seen as worthwhile in order to achieve improvements in quality and consistency, there was concern about workload assumptions relating to the preparation of CJSWRs and the balance of report writing and direct work with clients.

Judicial confidence in and use of community penalties

8.19 There are signs that, on balance, the introduction of CPOs has improved (and certainly not harmed) judicial confidence in community penalties, but that overall attitudes have not been radically transformed.

8.20 Some Sheriffs were clear that they saw the new framework as simpler, more coherent and more flexible, and the vast majority expressed 'a lot' or 'quite a lot' of confidence in CPOs in terms of the capacity of CJSW and other agencies to offer an appropriate range of programmes and services and in terms of the monitoring of progress and use of breach.

8.21 In the Sheriffs' survey, a majority felt that there had been no change in their use of such disposals, but a minority felt they were using them more often and almost none that they were using them less often.

8.22 There was also widespread evidence of a commitment to use community penalties wherever possible - and even in the face of serial offending or non-compliance - if there was any indication from the CJSWR of scope for constructive engagement. Whether this approach can be directly attributed to the introduction of CPOs (or to other aspects of the reforms) is less clear.

8.23 Sheriffs remain sceptical about the seriousness with which CPOs will be regarded by offenders and their acceptability to the general public. Both factors are potentially important background influences on their own confidence in the appropriateness of the disposal.

Judicial intention to reduce or stop the use of short prison sentences

8.24 The published statistics provide indications that actual use of very short prison sentences (i.e. of three months or less) is falling. Indeed, the number of such sentences imposed in 2012-13 was only 81% of the number imposed in 2010-11. However, that needs to be set against a sharp increase in the year following the reforms in the number of sentences of three to six months, and six months to two years. Although both figures fell again in 2012-13, they remain above the level prior to the introduction of the reforms.

8.25 The extent to which such changes can be attributed to the greater availability of community penalties for low tariff offenders, or indeed the introduction of PASS, is not clear. The decrease in the use of very short sentences and increase in the use of sentences of 3 months or over are both broadly consistent with longer-term trends. And in the qualitative interviews, there was little sign of PASS figuring prominently or explicitly in judicial decision-making; although some Sheriffs did suggest it had been a background factor in avoiding a short prison sentence in a small number of cases. Most, however, considered the presumption to be of little practical consequence both because of an existing commitment to use community penalties wherever possible and because of the 'inevitability' of a short custodial sentence in a small number of cases.

8.26 By contrast, there was some evidence from both the qualitative interviews and the Sheriffs' survey that PASS may be encouraging both greater use of community penalties and increased use of longer custodial sentences - a hypothesis that would be consistent with the statistics summarised above.

8.27 Although there was consensus that a custodial sentence remains a 'last resort', there is clearly variation in when Sheriffs consider that threshold to have been crossed. Most regarded prison as the only option in the face of serial non-compliance and the absence of any indication from the CJSW of the potential for constructive engagement. Examples of such non-compliance were often wilful - offenders simply refusing to engage with the opportunities given to them - though others were seen as rooted in complex and overlapping social problems (such as homelessness and addiction). In such cases, a short sentence, while having little explicitly rehabilitative potential, was sometimes seen as helpful in allowing individuals to 'dry out' or 'wipe the slate clean' in terms of an unmanageable accumulation of fines and community penalties.

The provision and use of a full range of CPO requirements

8.28 As we saw earlier, the full set of CPO requirements is technically available. However, the evaluation raised some important questions about whether all the requirements are being used appropriately and with sufficient frequency. There are two main aspects to this.

8.29 The first is the (increasing) proportion of orders involving an UPWOA requirement - now representing four-fifths of all those imposed. By contrast, and with the sole exception of Supervision, no other requirement was used in 2012-13 in more than 10% of CPOs. In relation to the specific use of UPW, there are indications that placements are being commenced faster than in the past; the timescales for completion of UPWOA hours are also shorter. Level 1 orders are being widely used and are seen by Sheriffs as a welcome addition, especially as an alternative to monetary penalties for low tariff young male offenders and those on benefits.

8.30 There are particular issues around the level of use of the three 'treatment' requirements (ATRs, DTRs and MHTRs). Although the precise nature of these varies by requirement, there are common themes relating to clarity about the intended target group, timescales for assessment and access to treatment, a perceived reluctance on the part of some CJSW staff to recommend multiple requirements and a corresponding preference for undertaking such work under the Supervision Requirement.

8.31 The second important (and related) issue here is the wide variation in the use of different requirements by area. While a degree of variation is to be expected on the basis of differences in the underlying characteristics of local authority areas, the nature and extent of variation in the use of specific requirements suggest that differences in interpretation of the NOS Practice Guidance and other factors (for instance, relating to service availability) may lead to inappropriate diversity of opportunity and outcome for offenders across Scotland.

Offenders' understanding of CPOs, engagement, compliance and behaviour change

8.32 Most offenders have little understanding of the difference between CPOs and previous community penalties. Those interviewed for the evaluation, however, seemed to have a good understanding of what was expected of them as part of the order, and of what might happen should they fail to comply. There was positive feedback about the nature and extent of social work support received under Supervision and evidence of the importance of relationships with individual CJSW staff. Engagement was positively driven by experience of concrete help and support (e.g. in relation to housing or addiction needs), as well as by measures to monitor and ensure compliance.

8.33 Around two-thirds of CPOs in 2012-13 were completed - a figure broadly consistent with that achieved under the previous framework for community penalties. Completion rates were highest for orders involving a single requirement, and lower for orders involving two or more different requirements. There is no evidence that the risk of non-completion increases with each additional requirement beyond this point but the specific combination of requirements appears to be a factor here - for example, CPOs which included a Programme Requirement as well as Supervision and UPWOA had the lowest rates of successful completion overall. Of course, it should be noted that all of these relationships may reflect correlation rather than causation - in other words, offenders who are more likely to breach may be more likely to be given particular combinations of orders, rather than be rendered likely to breach by the combination of requirements they receive.

8.34 Despite the fact that overall completion rates are comparable with those for previous penalties, there is a widespread perception among both CJSW staff and Sheriffs that the arrangements for the monitoring of compliance and enforcement of breach are more robust as a result of the reforms. Extensive use is being made of reviews, as a means of encouraging offenders to take orders more seriously; as a substitute for breach in cases where there were doubts about the appropriateness of the CPO; and to motivate the offender (e.g. by setting targets) and acknowledge successful progress if made.

8.35 Sheriffs generally consider that breach is being used appropriately, although there remain some concerns about perceived delays in initiating proceedings. There is also a continuing issue in some areas about the respective roles of CJSW and UPW staff at key stages in the process.

Where now?: Potential next steps suggested by the results of the evaluation

8.36 We have argued, then, that the reforms have been largely implemented as intended: there is a fully-featured and largely functioning set of CPO requirements; the CJSWR template is operational, offering standardised reports containing analysis and recommendations; and the presumption against short sentences is in place. And yet across all three key domains - the judiciary, CJSW and offenders - the evidence of short and medium term outcomes is somewhat muted. In this final section, drawing on some of the original aspirations for the reforms and on the issues that emerged as important to those closest to them in practice, we suggest some ways in which some of the missing links in the logic model might be reconnected and the longer-term goals ultimately realised.

Beyond the template: improving the quality and usefulness of CJSWRs

8.37 As the logic model acknowledges, the degree of confidence that Sheriffs have in CJSWRs is central to the effective and appropriate use of community penalties. As we have seen, certain features of the new template seem to contribute to such confidence. But, in several respects, the evaluation also highlights the limits of the template in terms of delivering confidence and credibility.

8.38 Regardless of overall judicial confidence in the reports, the fact that Sheriffs feel reports are too long is a concern because it suggests that key points or arguments may be lost.

8.39 Feedback from Sheriffs about the length and relevance of reports could be usefully incorporated into training for CJSWR writers. Previous exercises involving the piloting of summary reports should be revisited, along with the scope for making greater use of oral reports in certain types of less complex cases.

8.40 The evaluation also suggests that Sheriffs find the analysis of risk and suggested sentencing options to be less useful than other aspects of the report. In relation to risk, there is some scepticism about the comprehensive assessments deployed by CJSW - especially in relation to less serious cases. In relation to sentencing options and recommendations, there remains a concern among some Sheriffs that social workers are overly dependent on offenders' accounts and/or that they propose unrealistic community penalties when a custodial sentence is 'unavoidable'. Both factors - the limited evidence base on which CJSWR writers are able to draw and the perception that recommendations are sometimes unrealistic - can serve to undermine the credibility of individual reports and CJSWRs as a whole.

8.41 There would be benefit in additional joint work involving Sheriffs and CJSWR report writers in order to develop shared language and understanding around risk, and agreement about the types of cases in which or more or less thorough assessment is warranted. The Scottish Government may wish to review the scope to increase CJSW access to prosecution summaries and revisit the results of previous pilot exercises.

8.42 The evaluation also makes clear that - especially in smaller courts - Sheriffs do not assess the credibility of individual reports in isolation, but through their knowledge of, and trust in, the particular CJSW staff who have prepared them. In other words, the process can have an important inter-personal dimension, even if the emphasis on the template gives a sense of being highly standardised and impersonal.

8.43 The sense of a single 'model CJSWR' is also undermined by the fact that individual Sheriffs have varying requirements and preferences and ultimately 'consume' the reports in different ways. The character of the information they are looking for also varies according to the nature and complexity of the case. For all these reasons, improvements in the quality and credibility of CJSWRs in the eyes of sentencers will not be driven by simple adherence to a set template but by its intelligent application in relation to the character of particular cases.

8.44 There would be benefit in seeking additional opportunities - at both local and national levels - to bring Sheriffs and report writers together, in order to improve understanding of each other's position and priorities, share concerns and improve the fit between what Sheriffs feel they require and what CJSWR writers provide for different kinds of cases.

8.45 As many CJSW staff acknowledge, the template also brings with it a risk that the more complex and analytical aspects of the process will be overlooked. The implications of this are explored below.

Giving CJSW staff the time, skills and support to produce high quality reports

8.46 Although it is clear that fairly comprehensive initial training was given to relevant CJSW staff in the use of the new template, there was a sense from the interviews that this was over-shadowed by the introduction of CPOs and the LS/CMI and that it tended to focus on the technical aspects of the CJSWR rather than on wider report writing skills. This would certainly be consistent with the suggestion that staff are tending to 'write to the boxes' and are not always offering high quality analysis or drawing out the more important messages around individual cases. Since the initial training, it is also clear that inadequate attention has been given in some areas to follow up training for new starters and refresher training for existing staff.

8.47 There should be regular national oversight of the training provided to CJSWR report writers within individual local authority areas; and that training should explicitly aim to improve the quality of analysis and recommendations, as well as covering the technical requirements of the template.

8.48 It also needs to be acknowledged that high quality CJSWRs - that draw on a range of sources and contain high quality analysis and conclusions - are time-consuming to produce.

8.49 Estimates of the time required per report need to be reviewed and cannot be assumed to be consistent over time. The balance of staff time spent on report-writing and other aspects of casework should also be monitored.

8.50 Although CJSWR writers generally felt well supported through broader line management and ad hoc consultation with colleagues, systematic quality assurance of reports appears to be the exception rather than the rule (except for new staff and students) and opportunities to identify problems and contribute to individual and group learning through retrospective audit are being lost.

8.51 There should be clear best practice guidance about the use of quality assurance as part of routine working practices. Consideration should also be given to regular intra- and inter-local authority audit of samples of reports. The findings of quality assurance and audit exercises should be fed back into training and practice.

Sharing experience and good practice in relation to the use of OA

8.52 OA is another aspect of the reforms that was slightly overshadowed by the extent of initial change and has taken some time to bed in. Now that it is starting to do so, it is clear that there are some creative and effective uses being made of the provision. But there is evidence of a lack of vision or ownership about how OA can be used, and of the impact of funding constraints. Senior managers, both in central government and in local authorities, are largely leaving it to practitioners to come up with ideas. While this has led to some highly creative responses, it means that implementation is patchy, unequal and lacking in strategy overall.

8.53 We suggest that a national review of OA should be carried out by the Scottish Government, based on detailed reporting of its use in each area and the identification of examples of good practice. Insights from this work should be incorporated into any further training on CPOs and taken into account in consideration of funding requirements.

Bringing greater clarity and consistency to key aspects of the reforms

8.54 There is a lack of clarity among practitioners in relation to some key areas and, in particular, about the appropriateness of different requirements in specific circumstances. At times, this appears to result from ambiguity in the wording of the legislation or NOS Practice Guidance; in other places, it seems to be rooted in misconceptions or local interpretation and practice. Whatever the cause, it can be concluded that, in these areas, there is a lack of the simplicity and coherence that were intended to be key features of the reforms.

8.55 Against this backdrop, it is not surprising there is considerable geographic variation in the use of the various CPO requirements and in other important areas, such as the use of breach. As already noted, some variation is to be expected on the basis of local characteristics. Moreover, the principle of judicial independence places limits on the extent to which consistency in sentencing can be imposed. Individual courts - and smaller courts in particular - have their own cultures, sometimes in response to highly 'independently-minded' local Sheriffs. In planning for and delivering community penalties, CJSW staff clearly have to work with and respond to those local cultures and circumstances. Nevertheless, they also have a critical role to play in ensuring that approaches to the use of community penalties are broadly similar across Scotland as a whole.

8.56 The NOS Practice Guidance is very important here, and there is clearly a need to make sure that there is a shared understanding of that guidance across different areas.

8.57 The evaluation suggests there would be value in the Scottish Government reviewing the wording of the guidance itself, and in improving national oversight of how it is implemented.

Creating clarity around the actual and appropriate use of the 'treatment' requirements

8.58 There are particular ambiguities and uncertainties around the use of ATRs, DTRs and MHTRs. Specifically, CJSW staff and other actors lack a shared understanding of who exactly each of the requirements is aimed at, how eligibility is to be determined and what form of treatment is appropriate for inclusion within an order.

8.59 As noted above, we suggest that the wording of the NOS Guidance - and, if necessary, the legislation - should be reviewed in order to identify, on a requirement by requirement basis, issues that seem to be creating uncertainty or misunderstanding.

8.60 However, a more fundamental constraint on the ability of the courts to impose rigorous treatment requirements is simply a lack of available resources (and especially health-related resources). For these requirements to be used more widely, there will need to be explicit engagement from partners within the NHS, both in relation to the provision of (rapid) assessments but also the availability of and access to services. Put simply, it is difficult to see how widespread use of such treatments is possible without offenders being given greater priority in terms of waiting times.

8.61 There is scope for greater clarity about the role and responsibility of partner agencies (especially within the NHS) to provide rapid access to assessment and services. If such involvement cannot be provided or resourced, the implications of this need to be reflected in the wording and guidance for the relevant CPO requirements.

8.62 It may well be the case that - as many interviewees told us - offenders with drug, alcohol and mental health problems are accessing relevant help, support and services through other channels (for example, under the Supervision or Programme Requirement, or via existing service links). The problem, however, is that we simply do not know for sure whether this is happening consistently or how. A separate study would be required to understand not only the character of those cases that are being dealt with under the treatment requirements, but also the volume and character of those cases involving similar issues that are being dealt with under other requirements. The current situation not only means that individual offenders with drug, alcohol or mental health problems may face variation by area in terms of the disposals on offer but that, at both an individual and an aggregate level, there is a lack of transparency about how particular types of cases are being dealt with. This, we would argue, runs counter to the objective of a simpler and more coherent framework for community penalties in general.

8.63 The Scottish Government should seek to develop a fuller understanding of how the treatment requirements are being used - for example, by undertaking a review of cases in which a DTR, ATR or MHTR has been imposed, and an audit of such cases returned to court. Local authorities should perhaps be asked - as part of the annual CPO returns or a bespoke exercise - to provide numbers of offenders subject to a CPO who are in receipt of support or treatment for drug, alcohol or mental health issues via another requirement.

Acknowledging the tensions involved in the widespread use of Supervision and OA

8.64 Some of the original objectives of the reforms are potentially in tension with each other, and such tensions need to be acknowledged if they are to be managed effectively. Perhaps the most fundamental is that which exists between the desire for flexibility and relevance, on the one hand, and transparency, visibility and coherence, on the other. For social workers to be able to work effectively with the diverse needs and circumstances of offenders, they need flexibility of the kind afforded by the Supervision Requirement and the 'other activity' element of the UPWOA Requirement - for example, to offer a range of options to different individuals, or to vary the management of an order in response to progress, emerging issues or changes in circumstances.

8.65 But it also needs to be recognised that how such work is conducted is not only decided at the discretion of CJSW case manager, but also that it is largely invisible to other actors, such as sentencers and victims. In many circumstances, this may be appropriate, but the increasing use of the UPWOA Requirement and the tendency to use Supervision in place of the more structured 'treatment' requirements mean that it is more difficult to sustain the commitment to a simpler, more visible framework for the delivery of community penalties as a whole. It also means there is less accountability around the delivery of such work; that the point at which offenders should be breached is less clear; and that it is more difficult to monitor and quality assure the effectiveness of the interventions delivered.

8.66 The guidance could provide greater clarity about the circumstances in which the use of Supervision, in particular, is appropriate; and there would be clear benefits in improved monitoring of the extent and outcomes of work conducted under these more flexible provisions.

Managing the growing use of Level 1 orders

8.67 The evaluation makes clear that the Level 1 order has been widely welcomed and used by Sheriffs, often as a direct alternative not to prison but to a monetary penalty. There are clearly issues of capacity here, although these do not currently seem to be unmanageable - at least in the four case study areas. Equally important, however, is the question of the absence of a CJSWR in such cases, since it has implications for the effective assessment of the suitability of offenders for UPW and the matching of individuals to relevant and appropriate placements. The management of risk is a key aspect of that.

8.68 There is a need to build on and extend emerging good practice in this area - for example, encouraging Sheriffs to make enquiries at sentencing about individuals' suitability for UPW and making greater use of oral reports where appropriate.

Providing Sheriffs with feedback about outcomes and sentencing practice in other courts

8.69 It is clear that Sheriffs tend to have only a limited sense of how their own sentencing practice relates to that of colleagues in other courts and areas. This is likely to contribute to the variation in the use of the different requirements observed in the four case study areas and elsewhere. Sheriffs also appear to receive little feedback about the outcomes of the sentences they impose. If judicial confidence is to be reinforced and the coherent and effective use of community penalties encouraged, these would seem to be potentially important feedback loops.

8.70 Sheriffs might benefit from regular briefings about local and national use of the various requirements, and easy access to research and statistical evidence about outcomes and effectiveness. There needs to be discussion about how such information might be made available, and by whom.

Encouraging communication between Sheriffs and CJSW

8.71 The evaluation highlights the extent to which Sheriffs vary in terms of their closeness of their relationship with local CJSW practitioners and the perception they have of their role beyond the sentencing process. But regardless of the degree of professional overlap and alignment between CJSW staff and the local bench, for any sentencing framework to operate effectively, there need to be clear expectations and a shared understanding between the two groups. In smaller courts, this is easier to achieve, though should not be assumed. In the larger courts, it is likely to require a conscious attempt to bring together the different perspectives outside the everyday work of the court.

8.72 Court user groups and other cross-professional fora should be encouraged to look specifically at expectations and experiences of the various aspects of the reforms. Possibilities should be explored for the discussion of such issues in informal settings that might be conducive to the effective exchange of information, ideas and concerns.

Engaging the public

8.73 The research suggests that sentencers often remain doubtful about the acceptability of community penalties to members of the general public, and that - at some level - those 'perceptions of perceptions' continue to influence thinking about the appropriateness of particular disposals. It is also clear that the notion of community consultation remains under-developed in most areas. There is a need to manage this process to enable communities, voluntary organisations, partner agencies and others to generate creative ideas which are manageable, capable of being delivered using the optimum amount of resource, and deliver a combination of offender engagement with the community and skills development.

8.74 Local authorities should be required to provide clear plans for the prospective management of community consultation, rather than just retrospective examples. Again, consideration should be given to how - and by whom - Sheriffs might be provided with summaries of research evidence about public attitudes in this area.

An expanded logic model and a commitment to continuous improvement

8.75 It was noted in Chapter 1 that attempts to maximise the potential of community penalties have faced the dual challenge of developing a coherent, flexible and above all effective framework for the implementation of such disposals, and of persuading the public (and sentencers) that they represent an appropriate and acceptable alternative to prison.

8.76 There are signs that in a number of important respects - such as the consistency of CJSWRs, the range and flexibility of disposals on offer, and the monitoring and enforcement of compliance - the reforms have moved community penalties in Scotland in the first of those directions. There are also some indications, outlined in the sections above, of how that direction of travel might be maintained (and indeed the speed of travel increased).

8.77 For that to happen, however, there is a need not only to revisit but to expand the scope of the existing logic model in order to address systematically the causal links between the implementation of the reforms and achievement of some of its critical short-term outcomes. That would involve, for example, an explicit engagement with the bases of judicial confidence in community sentences - one which focuses specifically on issues which appear to limit such confidence, such perceptions of the degree of public support for community penalties, of the seriousness with which they are treated by offenders, and the speed, visibility and rigour with which they are imposed and enforced. It would identify the specific mechanisms through which offenders might be expected to take CPOs seriously or become motivated to change. These would again include an explicit focus on the speed and rigour or punishment, but also the extent to which they are genuinely tailored to individual needs and circumstances. And it would focus on exactly how social workers might be equipped to prepare reports that inform effectively the sentencing process.

8.78 In closing, then, the evaluation suggests that the reforms offer a framework capable of delivering the kind of short, medium and long-term outcomes that were originally hoped for. However, that end point cannot be assumed, and the ultimate objectives of the reforms - in terms of reduced reoffending, increased reintegration and reduced 'churn' in the prison population - will only be realised through careful and ongoing analysis and calibration of exactly how all three elements are working on the ground, coupled with a system-wide commitment to partnership working and continuous improvement.