Where Did It All Go Wrong? 2

Earlier this year I went away for a week and as luck would have it, my absence coincided with an important report published by an outfit called Crest  Where did it all go wrong? A study into the use of community sentences in England and Wales. 

Although I gave it a brief mention here, mostly highlighting the proposals, I regret not having given it a closer look because it rather neatly confirms everything we've previously discussed at some length, including the hugely damaging effect of TR. As this supposedly independent report makes clear, the Probation Service has clearly been thoroughly smashed by utterly inept political involvement over several years and there is now an urgent need for an effective plan 'B' that goes beyond paying the CRCs more money. 

In essence, we have a major and growing crisis in our prisons, coupled with a massive reduction in community sentences and now clear evidence of a significant increase in crime. Interestingly, despite being published only three months ago, the opening line and premise "Despite overall crime falling" has proved to be completely false. 

A long but important read I feel. Here are my selected highlights of the analysis contained in the report - the proposals were covered in the earlier post:- 

Executive Summary

Despite overall crime falling, our criminal justice system remains under pressure. Nowhere is that more apparent than in our prisons, which are, in the words of the former Chief Inspector, “in their worst state for a decade”, with violence, overcrowding and self-harm higher than at any point on record. 

Policymakers have long understood that a key part of the solution to an overstretched prison system lies in a more effective regime of community sentences, able to secure the confidence of magistrates and the public. As far back as 2003, Pat Carter (whose review presaged the creation of the National Offender Management Service) was calling for sentences in the community “to be made more demanding” as a way to re-balance the system. And in November 2016, the Lord Chief Justice called for more offenders to be sentenced to “tough” and “visible” alternatives in the community, in order to reduce the numbers sent to prison. 

The notion that community sentences can be a more effective, cheaper alternative to prison is supported by a strong body of evidence. At their best, sentences served in the community can offer a powerful tool for addressing the root causes of offending behaviour, reducing the rate at which an offender reoffends and thus lowering demand on the system overall. 

Yet despite their obvious potential, community sentences (community orders and suspended sentence orders) are being used less than at any point over the last 15 years. Since 2004, the numbers sentenced to community orders have halved, and overall numbers of sentences served in the community are down 25%, whilst the numbers sentenced to custody have remained relatively stable. Not only is this fuelling unnecessary pressure on our prisons, it is impacting the financial viability of community rehabilitation companies, who are struggling to cope with a lower than anticipated volume of paid work. 

This report is the first systematic attempt in over a decade to understand what lies behind this phenomenon. It reveals that community sentences: 

● are implemented in a way that bears little resemblance to the evidence of what works: they are neither intensive, swift, nor punitive enough to act as a proper deterrent. Most importantly, offenders are not held properly to account for complying with their sentence. The Probation Inspectorate (HMIP) has found that in a third of cases where the offender breached their order, “insufficient effort was made by the CRC responsible officer to re-engage them” 

● are failing to transform lives, acting as little more than a stepping stone on the path to prison: 35% of those sentenced to custody have received at least five previous community sentences

● have lost the confidence of magistrates: a new survey of magistrates commissioned for this report reveals that over a third of magistrates (37%) are not confident that community sentences are an effective alternative to custody, and two thirds (65%) are not confident that community sentences reduce crime.

These problems are the result of long term structural issues relating to the operation of the criminal justice system, which largely pre-date recent changes to the mix of crimes and government policy reforms. In particular, there appears to have been a long term decline in: 

● the quality/depth of advice provided to the court to guide sentencing decisions - in the form of ‘pre-sentence reports’ (PSR): there has been a transition over the past decade from PSRs being detailed, written reports to speedy, short, written and oral reports. Almost half (42%) of reports in 2015 were delivered orally, with no information recorded, compared to just 5% in 2006

● the level of information/training provided to magistrates: meaning they are unable to make the most effective use of community sentences and/or to take into account probation providers’ capacity to deliver. Over a third (36%) of magistrates do not feel that the training has adequately prepared them for dealing with community sentences and their requirements 

● probation’s ability to deliver personalised sentences that address the underlying causes of an offender’s behaviour and hold the offender to account for compliance: our qualitative research has revealed a deep-seated sense of decline amongst probation staff about the quality of services being provided and the ability to enforce breaches, which has been exacerbated by recent government policy changes. Four in ten magistrates (39%) are not confident that community sentences can be tailored to suit the individual needs of an offender.

These proposals come at a time of significant change to the way offenders are managed in the community. Whilst it is too early to be definitive, there is emerging evidence that the government’s flagship reform programme - Transforming Rehabilitation (TR) - will exacerbate the problems identified above, reducing dialogue between probation and the courts, reducing incentives to deal swiftly with breaches and stifling innovation in the delivery of services to prevent reoffending. There is also little doubt that the fiscal context, with funding having declined since 2010 and set to continue falling, will add to the pressures identified in this report. 

Our report seeks to learn the lessons of the recent past, in order to influence the future of sentencing and probation reform. The research was informed by a large number of interviews with police and crime commissioners, magistrates, probation staff, police and policymakers. We also commissioned a new survey of magistrates through the Magistrates’ Association.

Approach

Various explanations have been postulated about the apparently declining relevance of community sentences: 

● A loss of confidence by sentencers and the public in the effectiveness of community sentences 

● Changes in crime patterns, meaning a more serious mix of offenders is coming before the courts, for whom community sentences are less appropriate 

● Risk aversion within the National Probation Service in recommending community sentences in pre-sentence reports 

● The impact of the Transforming Rehabilitation reforms, including the design of Community Rehabilitation Company (CRC) contracts and the introduction of through the gate support for short term prisoners

History and purpose of community sentences 

Chapter summary: 

● The community sentences (community orders and suspended sentence orders) in use today were introduced in 2005, but subsequent policy reforms since 2010 have affected the way offenders are managed in the community and how community sentences are used in practice. 
● Technically, though both sentences are served in the community, COs are non-custodial sentences and SSOs are sentences of imprisonment served in the community. Both COs and SSOs can be comprised of one or more of 13 possible conditions, or ‘requirements’ to enable the sentence to be tailored to the needs of the offender. 
● Whilst in theory, Community Orders (COs) and Suspended Sentence Orders (SSOs) are identical (in terms of the requirements they can impose) SSOs tend to be less onerous, since the threat of incarceration is deemed a punishment in and of itself. 
● In 2016 just over 100,000 COs and 56,000 SSOs were given out, compared to almost 200,000 COs and 22,000 SSOs in 2006.

The development of community sentences 

The community order (CO) and the suspended sentence order (SSO) as they are recognised today became available in 2005, having been introduced by the Criminal Justice Act 2003, and were designed to provide more credible alternatives to short term custodial sentences and a solution to the rising numbers in prison. 

The new CO replaced and standardised the range of pre-existing community sentences (known as community penalties) that had previously developed erratically, and simplified the order by combining all other former community sentences, meaning magistrates and judges could more easily tailor sentences according to the particular nature of the offence and the offender. 

The SSO revived an older style of sentence that was rarely implemented due to stipulations under the former legislation that it could only be used in “exceptional circumstances” . These 17 stipulations were scrapped in the 2003 legislation, giving new life to the suspended sentence. 

Since 2010 the government has introduced a number of reforms to the way we manage offenders in the community. During the last parliament there were three big policy developments:

● 2012 Legal Aid, Sentencing and Punishment of Offenders Act (LASPO), which reformed SSOs 
● 2013 Crime and Courts Act, which reformed COs 
● 2014 Offender Rehabilitation Act, which: a. introduced the Rehabilitation Activity Requirement (displacing supervision requirements); and b. introduced the so-called ‘Transforming Rehabilitation’ reforms to probation

The general thrust of reforms since 2010 has been to make community sentences more robust and rigorous (thus driving up confidence), whilst making it easier for sentencers to craft an appropriate intervention for a particular offender. However, it is far from clear that they have achieved their intended purpose, as the next sections of this report will show.

Trends in the use of community sentences

Chapter summary: 

● The number of community orders has essentially halved over the past decade, with an increasingly sharp decline from 2011 onwards, whilst custody and fines have remained relatively stable. The rising use of suspended sentence orders has not offset this trend 

● The use of community sentences has declined fastest for theft and drugs offences 

● There appears to be little difference between the makeup of offenders on COs and SSOs, based on their harm and reoffending risk evaluations 

● Unpaid work is typically the most common requirement commenced under a community sentence, and since its introduction in 2015, the RAR has become the second most common 

● Delays are common between sentencing and commencing a requirement, and vary between regions and type of requirement 

● The majority (70%) of offenders undertaking a community sentence go on to complete their sentence, however: 

○ 30% of community sentences fail for the committal of an additional offence, or for breach of sentence stipulations ○ reoffending rates have remained relatively flat over the past decade and vary between disposals: custody stands at 45% (rising to 60% for short custodial sentences); COs at 35%; and SSOs at 31% (which has decreased from 37% a decade ago) 

○ reoffending performance varies between CRCs, with interim reoffending rates ranging from 26% to 43% for community sentences ○ prolific offenders (those with 15 or more previous convictions or cautions) make up an increasing proportion of those sentenced; for community the proportion has risen from 15% in 2005 to 25% in 2015. 

○ 75% of offenders sentenced to immediate custody for an indictable offence in 2014 had previously served at least 1 community sentence

Court sentencing

Trends in court sentencing over the last decade paint a stark picture. Against a backdrop of falling sentences overall, the number of offenders sentenced to community sentences has fallen dramatically (essentially halving), particularly since 2011, whilst those sentenced to immediate custody have remained relatively stable. Just over 100,000 community orders and 56,000 suspended sentence orders were given out in 2016, compared to almost 200,000 COs and 22,000 SSOs a decade ago.

These figures are even more dramatic if we look at the last five years: during which time there has been a staggering 78% decline in the number of community sentences. In 2010, a report by the think tank Policy Exchange described the so-called “unprecedented expansion of community sentences” as “the untold story of the criminal justice system over the last twenty years”. Seven years on, it is their almost disappearance that is the untold story of the CJS.

Trends in the composition of community sentences 

Unpaid work has consistently been, and remains today, the most frequently used requirement commenced under a community order, followed by Supervision (now displaced by the new ‘Rehabilitation Activity Requirement’ - see below) and Curfew. Unfortunately, there is very little publicly accessible data on what constitutes ‘unpaid work’ - the kind of work involved, how intensive it is and where it takes place.

As the National Audit Office (NAO) has noted, the fact that ‘accredited programmes’ represent a falling proportion of requirements commenced has significant implications for the financial viability of CRCs, who are paid according to the number of offenders completing such programmes (as opposed to ‘RARs’, for which they are required to pick up the costs), which CRCs have linked to the declining quality of PSRs. (In response, the NPS and CRCs have begun working on an ‘Effective Sentencing Framework’, which will seek to address this issue by guiding report writers toward the appropriate sentence.) 

Looking in detail at the mix of requirements commenced under community sentences it appears that COs became slightly more punitive between 2012 and 2015 (45% of sentenced included requirements deemed as principally punitive in 2015 compared to 41% in 2012) - 47 following changes introduced through the Crime and Courts Act. However, this is still someway short of the government’s intended ambition that every CO includes a punitive element.

Loss of confidence by magistrates 

Interviews conducted with magistrates for this report found that there is a high confidence in the principle of community sentences and what they are designed to achieve. However, a new survey by Crest shows that magistrates lack confidence in the practical effectiveness of community sentences. Our survey shows that:  

● Over a third of magistrates (37%) are not confident that community sentences are an effective alternative to custody 
● Two thirds (65%) are not confident that community sentences reduce or deter crime 
● 45% are not confident that community sentences effectively rehabilitate offenders 
● Three quarters (76%) are not confident that community sentences protect the public 
● Half (49%) are not confident that community sentences can be tailored to suit the individual needs of an offender 
● Just under a third (30%) are not confident that community sentences provide an appropriate response to the range of low-to-medium level offences they cover.

Drivers of decline

Our key conclusions: 

● The primary driver of is declining confidence in community sentences, which relates to long term structural issues to do with the operation of the CJS 
● In particular, the lack of information accessed by magistrates (pre- and post-sentence) and the declining quality of advice pre-sentencing from probation seem to have been key factors 
● It is possible that the number of community sentences may have fallen due to a change in the cohort of offenders, in particular, with a rise in prolific offenders, but this is unlikely to have been the primary driver of changes in sentencing behaviour 
● These trends pre-date recent policy changes, such as TR, though TR is likely to exacerbate the problems. In particular, the split between CRCs and the NPS, and the structure of CRC contracts is in all likelihood going to reduce confidence even further 
● Longer term, the biggest barrier to confidence is likely to be the continuing evidence (and perception) of low effectiveness

The role of probation

Another possible explanation for the decline in community sentences is an increasing risk aversion within the probation service in recommending community sentences in PSRs.

A growing target-driven culture 

Many of the probation staff we interviewed felt that the move to a more target-driven culture in probation (which began around 2003-04) had had significant implications for the provision and delivery of community sentences. Historically, the stated role of the probation service had been to ‘advise, assist and befriend’ offenders until the establishment of NOMS in 2004, when the emphasis shifted toward a focus on punishment, rehabilitation and public protection. Focus group participants were in agreement that the combined impact of these changes had been to drive an emphasis on process, rather than outcomes, leading to greater risk-aversion across the system and, crucially, reducing the amount of time officers spent with offenders. 

Focus group participants suggested that probation has been heavily influenced by changes in political ideology, meaning that the service provision, at least in the time the participants had been in probation (for most, over a decade), was in constant flux. This, they suggested, meant that offenders were now confused by the system - making it less likely that they would complete their requirements and harder for probation officers to fulfil their duties. 

Reduced quality of advice, pre-sentencing 

Before imposing a custodial or community sentence, the court is required to obtain a PSR unless the court is of the opinion that a report is unnecessary in all the circumstances of the case. PSRs are prepared by probation officers for consideration pre-sentencing, both in the magistrates’ courts and in the Crown Courts. They are designed to give information to the sentencer about the circumstances and context of the case and the offender. Some reports can be prepared on the day by NPS court officers. These reports are for cases already known to the probation service, or if the case is straightforward ('Oral' and 'Fast Delivery' reports). 

Others require a fuller assessment carried out by a probation officer ('Standard Delivery Reports'). Nationally, only 69% of the PSRs that proposed a community sentence are followed by a community sentence being given, compared to 81% of PSRs that propose custody.

The types of report prepared for magistrates have also changed dramatically over the last decade. In 2006, standard written reports made up the majority of PSRs (66%) - in 2015 they made up 11%. By the same token, almost half of reports in 2015 were delivered orally, compared to just 5% in 2006.

This change in PSR is indicative of a conscious policy transition from standard PSRs (which can take up to 15 working days to prepare) toward fast oral reports (which can be prepared in 20 minutes). The guidelines published by the sentencing council recommend that "ideally a pre-sentence report should be completed on the same day to avoid adjourning the case". A 69 recent report into efficiency in the criminal justice system recommended a move to dispense with PSRs where possible or replace them with oral reports in cases which do not require a PSR, i.e. a community order which includes a single requirement that does not necessitate the involvement of probation (e.g. a curfew order). 

Our interviews with probation staff supported these findings. Many held the view that the changing nature of PSRs meant that they now lack detail, which could mean that requirements being handed out as a result of report recommendations are not always appropriately tailored to the individual offender. Some participants felt that these effects had been exacerbated by TR, with NPS staff lacking the time/resources to draft PSRs of the requisite quality and detail. As a result, offenders were increasingly being handed a community sentence that would not address the root causes of their behaviour, would not allow probation to effectively do their job, and result in the offender coming back in front of the courts. Whilst efficiency and speed are to be sought after in the delivery of justice, it is not evident that the appropriate balance with quality has been struck.

The impact of recent policy reform

There is little doubt that the problems surrounding community sentences pre-date the roll-out of TR, and until data on reoffending are compiled in late 2017, judgements about the overall performance of probation services must remain partial. However, evidence compiled by the NAO and from our focus groups suggest that TR may have exacerbated long term problems that have been affecting the system in certain ways. These are set out below: 

Communication between probation and the court 

Under the terms of TR, CRC staff are not allowed in the court, despite retaining responsibility for 70% of offenders serving community sentences. There is emerging evidence that this has reduced levels of communication between sentencers and probation, whilst reducing the quality of the PSR. For example, in our interviews with magistrates and probation staff, nearly all participants felt that the split (between CRCs and the NPS) as a result of TR had increased the distance between magistrates and probation staff, as they no longer had any direct interaction with the people that worked with these offenders. 

“We feel more at arm’s length from probation.” - Magistrate

The range/quality of locally available rehabilitation provision 

The volume of work that CRCs are paid for, as measured by the payment mechanism, has been much lower than expected and promised in the original contracts from the MoJ. This reduced funding, combined with rising caseloads (following the introduction of statutory supervision, which is unpaid) has impacted upon the ability of CRCs to commission innovative/new provision i.e. anything which is not on an agreed ‘rate card’. It is also clear, as indicated in Table 3, that CRCs’ ability to deliver the core requirements of community sentences varies greatly across the country. As far as existing data suggests, in the worst performing CRCs, three in ten programme requirements are not successfully delivered, and one in five unpaid work requirements is not fully completed.

The impact of TR on local delivery 

To understand how TR has impacted on the confidence and perceptions of sentencers and service users, qualitative research was undertaken in Greater Manchester and Norfolk. The aim was to test whether TR has exacerbated existing factors that have contributed to the decline in community sentences, and ascertain whether further decline can be expected. The findings are explored below. 

Inflexibility of the TR model to tailor provision to local needs 

The one-size-fits-all TR model commissioned at the national level is too rigid to be adapted for all offenders and all contexts at the local level: 

● The model doesn’t necessarily fit with Greater Manchester’s vision for managing women offenders, who, according to TR, must be assessed by a probation officer in custody, rather than at a women’s centre 

● Rigid scrutiny from central government, with three separate MoJ teams regularly demanding input/engagement from CRC staff, generates a significant amount of work. The sentiment from CRC staff is that they feel like they are ‘feeding the machine’ rather than actually delivering for local communities. 

Contract specifications and payment schedules are based around processes and volumes, rather than outcomes 

Whilst the MoJ operating model talks about TR as an outcomes-focused programme, with maximum flexibility for providers, the reality, post-award, is that CRCs are being micromanaged and innovation is being stifled: 

● CRCs are measured according to 16 separate service levels (including the number of starts on unpaid work; how many plans are completed in 10 days; timeliness for recalling people to prison etc). There are financial penalties attached for not meeting these service levels. 

● The completion rates for behavioural programmes have generally been much lower than anticipated. Even though the pipeline of participants is not in the CRC’s gift (it is the NPS that makes advice to the court on those programmes), it is the CRC that ends up being financially penalised by the MoJ. This creates a vicious circle whereby the overall fee for service falls, discretionary funds are reduced and the CRC becomes more risk averse about innovating/commissioning new things. 

CRCs are supposed to be rewarded for reducing reoffending through a system of ‘payment by results’ (PbR). However, PbR doesn’t really feature in the CRCs’ day-to-day work as it is simply too far away to drive behaviours. This has led to CRCs chasing the short term fee for service. 

The rate card does not incentivise new services to be commissioned 

There is little evidence of new commissioning, as a result of: 

● The economics – unless the CRC can guarantee a certain volume of referrals, it will not go to the effort of commissioning new services 

● Lack of responsiveness – all CRCs have a standard rate card with a fixed price; there is no mechanism to vary the price and make it more attractive to the purchaser (e.g. a drink and drivers’ programme should cost a tenth of what it costs to deliver a violence reduction programme, yet it costs the same price) 

● Central government control – even when the CRC and NPS have managed to work together in co-commissioning a new discretionary service/product for ex-offenders, they have been unable to progress it to implementation because the MoJ insists on signing off on any new services added to the rate card – a process which has proved to be lengthy and bureaucratic.

Contracts do not incentivise breach enforcement

In addition to the problems identified above, there is some anecdotal evidence to suggest that breaches are less likely to be enforced as a result of TR. It has been suggested to us that the way in which the contracts for CRCs have been designed has created perverse incentives to disengage with an offender if the breach process is triggered. Once an offender breaches their conditions, the MoJ issues a financial penalty on the CRC, as an offender is only eligible for PbR the first time they commence an eligible sentence, meaning that those who breach their sentence are no longer eligible for payment, regardless of the support the CRC may have already provided. 

HM Inspectorate of Probation has found variability in enforcement processes across the CRCs, with some experiencing high rejection rates from the NPS for their breach recommendations. HMIP found that many recommendations were returned because of minor spelling and grammatical errors, and the NAO also identified an “inherent risk that offender managers may avoid ‘breaching’ offenders where this would affect CRC performance against targets for successful completion of orders if the court order is subsequently revoked”. A recent Justice Inspectorate report followed these earlier warnings, and stated “we were disappointed to find, in a third of cases where the individual was breached, insufficient effort was made by the CRC responsible officer to re-engage them and encourage their commitment to continued engagement”.