National Audit Office — Written evidence (JCS0039)    

We are pleased to provide evidence to the Committee’s inquiry into community sentences. Our evidence draws on insights from our published work where relevant and on publicly available information and data. We have not focussed in depth on community sentences in our recent value for money studies, so our insights are limited in scope. This brief covers:

1.     Trends in the use of community sentences

1. Historical trends
2. Potential causes of the decrease in use of community sentences
3. The future:  possible effects of prison capacity levels on sentencing options and choices

2.     Obstacles to the delivery of community sentences

3.     The cost effectiveness of short custodial sentences vs community sentences

1. Cost information on short custodial sentences and community sentences
2. Potential for cost savings by reducing short custodial sentences and increasing community sentences
3. Effectiveness of community sentences vs short custodial sentences in reducing reoffending
4. Conclusion on what this means when comparing the value for money of both options

1. Trends in the use of community sentences

1a. Historical trends

1. The number of community sentences issued by courts in England and Wales has significantly declined between 2012 and 2022, decreasing by more than half (54%) in that time period (Figure 1). Part of this decline is because the total number of people sentenced over the period has declined by 14%. However, the proportion of all offenders given community sentences has decreased significantly. In 2012, 12.3% (151,183) of all offenders sentenced in England and Wales were given community sentences, but by 2022 this had fallen to 6.6% (68,994). In contrast, the proportion of people sentenced to immediate custody has remained relatively stable at around 6 to 8% of those sentenced (Figure 2).

2. Between 2012 and 2022, there have been some changes to the types of offences for which people are sentenced. For example, the proportion of offenders sentenced for summary motor offences as a proportion of all offenders sentenced has increased by almost 20 percentage points. However, data do not indicate that changes in the frequency of different offences explains the decline in community sentences. For each type of offence, the proportion of sentences issued that were community sentences has decreased, with the exception of violence against the person, for which there was a small increase of one percentage point (Figure 3). 

3. The largest decreases in the proportion of community sentences issued by offence type were for fraud offences (14 percentage points), possession of weapons (13 percentage points) and drug offences (10 percentage points).[1]

Number of offenders supervised in the community

4. The number of offenders starting supervision with the Probation Service under a community sentence fell by 44% (from 106,073 to 59,800) between 2012 and 2022, reflecting the decline in community sentences over the same period. The volume of offenders supervised under a community sentence by the probation service at the end of each year (on 31 December) fell by a smaller proportion of 9.7% (from 78,263 to 70,682) between 2012 and 2022.[2] The decline in those supervised at the end of each year may be smaller due to differences in the rate that offenders start and complete probation, for example those who do not complete their probation period successfully may incur further sanctions and remain on the probation caseload for longer.

1b. Potential causes of the decrease in community sentences

5. There are various factors that may contribute to the decline in the proportion of community sentences issued by sentencers. As we have not reviewed this topic in depth, we do not seek to present a full list of factors. Rather, we highlight some relevant findings from our published work. These include:

• Quality and availability of information about offenders available to sentencers;
• Sentencers’ confidence in community sentences due to probation arrangements; and
• Awareness and availability of options for managing offenders in the community

Quality and availability of information about offenders available to sentencers

6. Where requested by the judiciary, probation staff can prepare pre-sentence reports (PSRs) to provide the court with a greater understanding of the background and the context of the offending behaviour to aid sentencing decisions. In 2022 we reported that the Ministry of Justice’s (MoJ’s) 2018 female offender strategy identified the need to improve the quality and timeliness of PSRs to support its aim of seeing greater proportions of women being effectively managed in the community.[3] This followed a significant decline in the number of PSRs prepared for the court since 2012, particularly in Magistrates’ courts which issue the majority of community sentences (Figure 4).

7. In May 2021, MoJ announced that it had launched a PSR pilot in 15 magistrates’ courts to improve the quality of information presented, following a decrease in the number of PSRs being requested by the judiciary between 2010 and 2018.[4]

Sentencers’ confidence in community sentences due to probation arrangements

8. In 2019, our report Transforming Rehabilitation: Progress review, found that the declining use of community sentences was in part influenced by the low confidence sentencers had in Community Rehabilitation Companies (CRCs), which MoJ contracted at that time to manage low or medium risk offenders in the community. The Transforming Rehabilitation reforms introduced by MoJ in 2014 had split the role of probation services and the National Probation Service (NPS) provided all advice to courts via pre-sentence reports.[5] However, HM Inspectorate of Probation had reported that these arrangements did not support sentencers’ confidence as CRCs were detached from the court process, while the NPS lacked knowledge of, and confidence in, CRC services.[6] Following its Transforming Rehabilitation reforms, MoJ decided to terminate its CRC contracts early and in June 2021 MoJ brought all probation services back into the public sector as part of its probation reform programme. We do not know what effect this unification of probation has had on sentencers’ confidence in community sentences, but the proportion of sentenced offenders given a community sentence remained unchanged between 2021 and 2022 at 7% (see Figure 2).

Awareness and availability of options for managing offenders in the community.

9. Our 2022 report, Improving outcomes for women in the criminal justice system, highlighted that MoJ’s 2018 female offender strategy identified the need to develop and maintain more options for managing women in the community, and to increase awareness of different sentencing options. This was to support its aim to see a greater proportion of women managed in the community. Providers of women’s community services can help support women before or during their contact with the criminal justice system, and some provide services specifically linked to probation. As part of its 2021 probation reforms the Ministry let contracts worth £46 million for services “tailored to female offenders”. It is not clear how the contracts’ value compared with previous years because the preceding contracts, with Community Rehabilitation Companies, did not separate out payments for gender-specific services.

10. Prior to awarding the probation contracts, MoJ initially focused its female offenders funding on grants for providers of women’s services in the community. However, despite this, we highlighted that the MoJ only spent £9.5 million on grants for community services over four years (2018–19 to 2021–22) and the short-term nature of grants made it difficult to sustain services. At the time of our report, providers of women’s community services told us that they were concerned about the financial sustainability of the services they provide. In January 2023, MoJ published its new female offender strategy delivery plan, which stated that it had launched multi-year grant competitions to invest up to £24 million in community support services for women.[7] On 26 May 2023 MoJ announced that it had awarded grants worth £15 million to women’s centres and other providers that work with women.[8] We have not recently looked at community sentences for men.

1c. The future: possible effects of prison capacity levels on sentencing options and choices

11. In recent months the prison estate has been operating close to full capacity. As at the end of May 2023, the prison estate was at 99% of its total capacity, with 85,415 people in prison.[9] Our May 2023 report, Improving resettlement support for prison leavers to reduce reoffending, highlighted that MoJ forecasts that the prison population could reach between 93,100 and 106,300 by March 2027. This represents increases of 9% and 24% compared with May 2023 figures. This is primarily due to the impact of police officer recruitment and longer prison sentences for serious offenders. MoJ is currently building additional prison places as part of its prison expansion programme, but it has also considered a range of measures to help manage prison capacity pressures, such as building rapid deployment cells and making provision for temporary housing of up to 400 places for prisoners in police cells through Operation Safeguard[10]. The increased prison population will have a follow-on effect for the probation service, which will experience more demand for resettlement services.[11]

12. The context of capacity pressures may influence the sentencing options considered and determined by the judiciary. A Court of Appeal judgment in March 2023, R v Ali, clarified that when courts are considering sentencing for an offence that may warrant a short custodial sentence, the court can take into account the impact of a high prison population. This is because high capacity levels may impact the conditions prisoners are held in and could result in reduced access to rehabilitative programmes.[12] Additionally, it is possible that policy makers may also seek to explore making greater use of non-custodial sentencing options such as community sentences for particular categories of offence. For example, the Ministry has recently changed its eligibility criteria for Home Detention Curfew (HDC) electronic monitoring to increase its use. Since 6 June 2023, prisoners can now be released on HDC up to 6 months early rather than the previous four-and-a-half months.[13]

2. Obstacles to the delivery of community sentences

13. Community sentences can include a range of conditions, including unpaid work in the community, treatment programmes to help tackle addictions and restrictions on where offenders can go and when.
14. The probation service is responsible for supervising offenders in the community, including people issued with community or suspended sentences. Since 2015, it is also responsible for supervising all offenders who are released from prison.[14] In our routine work, we and others have identified some challenges for probation staff, which are likely to impact across all probation activity, including supervising community sentences.

15. In a 2021 report, HM Inspectorate of Probation highlighted that the Probation Service had been under significant pressure dealing with both the COVID-19 pandemic and the unification of the probation service in June 2021. The Chief inspector cautioned that transferring cases from private providers to the civil service would not deal with underlying issues that inspections had uncovered in recent years. This included:

• a critical lack of frontline staff;
• excessive caseloads; and
• a lack of investment in the wider system of mental health, drug treatment and multi-agency partnerships on which the probation service relies.[15]

16. Our recent report, Improving resettlement support for prison leavers to reduce reoffending, identified pressures on the probation service which affected delivery of services. We found that:

• In December 2022, 1,762 out of 6,158 probation officer roles were unfilled, a vacancy rate of 29%.
• Across prisons and probation regions, there were high levels of sickness absence in 2020-2021 and high turnover in 2021-2022.
• In 2022, HMPPS’s audit of 98 cases and feedback from 15 service managers found that many probation staff were managing more than 70 cases, against a suggested case load of 30 to 60.
• For the probation officer grade in March 2023, 104 out of 113 of HMPPS’s probation sub-regions with available data (92%) were operating at or above 100% of their operational capacity.[16]

17. MoJ does not publish data on the time probation officers spend supervising offenders serving prison sentences compared with those serving a community sentence, but high workloads for probation staff will make it harder for staff to supervise community sentences effectively. In a March 2023 debate, the Minister of State said that the MoJ is investing more than £155 million in additional annual funding to recruit more staff and reduce probation caseloads.[17]

3. Cost effectiveness of short custodial sentences vs community sentence orders

18. Comparing the value for money of community sentences with short custodial sentences requires information on both the cost of each option and the corresponding outcomes. The following paragraphs summarise the results of our review of publicly available information, covering:
1. Available information on the costs of short prison sentences and community sentences
2. The potential for cost savings by increasing types of by reducing short custodial sentences and increasing community sentences
3. Available information about the effectiveness of sentencing types on reducing reoffending
4. Conclusion on what this means when comparing the value for money of both options.

3a. Cost information on short custodial sentences vs community sentences

19. In this section, we set out cost information available on both short prison sentences of less than one year, and community sentences.

Cost of short prison sentences

20. MoJ publishes annual data on the unit costs of prisons. In 2021-2022, MoJ spent £3.75 billion on overall prison running costs, which equates to an average cost of £47,434 to hold one prisoner for the full year or around £4,000 per month[18]. 

21. This figure does not include additional costs such as post-release supervision and in some cases the overall cost to the taxpayer may be much higher. For example, our report Improving the outcomes for women in the criminal justice system noted that in 2016-17, a cross-government project estimated that the total cost to government of a typical custodial sentence of six weeks in prison followed by 12 months on licence for a low/medium risk female offender in 2015-16 was £20,000. [19]

22. Based on MoJ’s unit cost data we can give a broad estimate of the cost to MoJ of holding people sentenced to short sentences of less than a year in 2022 (37,284) to be around £550 million (see Figure 5). This estimate is based on the average cost of holding a prisoner for a month in 2022 (£4,000), multiplied by the sentence length. There are a range of limitations with this estimate:

• people sentenced during 2022 will not necessarily complete their sentence in the same year, so the unit cost may change;
• As sentence lengths are grouped in the data (e.g. up to and including one month), we have also assumed that people complete the average sentence length in each range when estimating the total cost;
• The estimate does not include any wider costs such as post-release supervision. Since 2015 all prison sentences require a period of at least 12 months’ probation supervision.

Cost of community sentences

23. There are less data available on the costs of community sentences. HMPPS’s predecessor had previously published some unit cost data which included the cost of a community sentence. However HMPPS does not publish this unit cost data in its annual reports.

24. We reviewed available data within the annual accounts but found it difficult to give a current estimate on the cost of community sentences. There are data available on the number of people given a community sentence in 2022 (68,994) and HMPPS’s accounts state that its total resource expenditure on probation services in 2021-2022 was around £976 million. [20][21] However, we do not know how much of this cost will have been spent on providing supervision for community sentences as opposed to supervision for offenders released from prison or those serving suspended sentences.

25. Although we could not find current data on cost, previous published information suggests that community sentences were significantly cheaper than short prison sentences: 

• A 2014-15 National Offender Management Service Business Plan set out that the average overall cost of holding a prisoner in 2012-13 equated to £34,766 about eight times the average cost per community order or suspended sentence order, at £4,305[22][23]
• The 2016-2017 cross-government study referenced previously found that a typical community sentence for women (7-month community sentence) in 2015-16 cost around £12,000, just over half the cost of a typical prison sentence.

We did not find any more recent information on the costs of community sentences, nor on different types of community sentences.

3b. Potential for cost savings by reducing short custodial sentences and increasing community sentences

26. Despite the information set out about the lower cost of community sentences, increasing the proportion of people given community sentences instead of short custodial sentences may not result in immediate savings for MoJ. This is because there are many costs associated with running the prison estate such as staff salaries and the cost of its estate, which would require substantial changes in the prison population before they can be adjusted. For example issuing one offender with a community sentence instead of a prison sentence would not mean that fewer prison staff were required.[24]  Additionally, most people in prison at any given time are those serving sentences of 12 months or more. As at March 2023, just 4% of the prison population were serving sentences of less than 12 months, though this was higher in the women’s estate at 9%. (Figure 6).[25]

27. There may however be wider costs to society. For example, in some cases short prison sentences could involve children needing to go into care. Our report, Improving the outcomes for women in the criminal justice system, noted MoJ’s female offender strategy acknowledged that women’s imprisonment had a greater impact on children as women are more likely to be their main carers.[26] Custodial sentences may also impact an offender’s housing or employment status. Our May 2023 report, Improving support for prison leavers to reduce reoffending, highlighted that prison leavers are more likely to reoffend if they are not resettled into the community, for example if they have nowhere to live or no job.[27]

3c. Effectiveness of community sentences vs short custodial sentences in reducing reoffending

Overall cost of reoffending

28. In a 2019 report, MoJ estimated that the total economic and social cost of reoffending by adults in 2016 was £16.7 billion (stated in 2017-18 prices). It found that the cost of reoffences committed by adults who had previously received a custodial sentence of less than 12 months was £5.0 billion.[28]

Reoffending rates by sentence type

29. MoJ publishes data on reoffending outcomes for offenders given different types of sentences. Latest available figures, which cover a cohort of offenders from April to June 2021, showed that a significantly higher proportion of adults (54.4%) released from custodial sentences of less than 12 months reoffended than those starting a community or suspended sentence (29.5%).[29][30] 

30. However, to understand whether community sentences are more effective than short custodial sentences it is important to control for differences in the types of offenders in each group, which may impact how likely people are to reoffend. For example, those sentenced to custody tend to have greater criminal histories than those given court orders, so the two populations do not have exactly the same characteristics. This means it is not easy to determine whether the reoffending rate is lower for community sentences because the sentence is more effective, or because the offender’s characteristics make them less likely to offend.

31. MoJ has attempted to look at this by using statistical analysis to match people given particular sentences with other similar ones given different sentences. It includes various different variables in its analysis such as information on offenders’ demographics, offending history, employment history and elements of the Offender Assessment System (OASys), which assesses the risks and needs of offenders. [31]  In both its 2015 and 2019 reports, MoJ found that short-term custodial sentences of  less than 12 months in prison were consistently associated with higher rates of proven re-offending than community orders and suspended sentences.[32] [33] MoJ notes that care should be taken in generalising results from the studies as cases had to be discarded where data was incomplete. Findings therefore likely represent offenders with more entrenched problems as they are more likely to have had a comprehensive offender assessment.

32. Its most recent report in 2019 report specified that;

• the one year re-offending rate following short-term custodial sentences of less than 12 months was 4 percentage points higher (around 76%) than if a community or suspended sentence had been given instead (around 72%).
• The one year average number of reoffences per sentencing occasion was also higher for short term custodial sentences compared to a community sentence; and
• additional analysis showed that the one-year reoffending impact of short-term custodial sentences compared to community sentences was of a similar scale (around 4 percentage points) whether the short-term custodial sentence length was less than 3, 6 or 12 months.[34]

33. In its 2015 study, MoJ looked at the effectiveness of different requirements that can be included with community sentences or suspended sentences. It found that those that included supervision by the probation service were generally associated with reduced proven reoffending. But it acknowledged that further work was needed to understand which combinations of requirements were most effective.  It noted that this can be difficult to determine due to difficulties in differentiating between different programmes and activities.[35]

3d. Conclusion on what this means when comparing the value for money of both options

34. Comparing the value for money of community sentences with short custodial sentences requires information on both the costs of each option and the corresponding outcomes. As detailed in the sections above, we found that there is evidence to suggest that community sentences may typically cost less and achieve better outcomes than short prison sentences. However, we identified some barriers to making a robust assessment. This includes:

• a lack of current data on the cost of community sentences, at both a total level and for different sentence types;
• where cost data exists, it does not always include wider costs to government; and
• a lack of evidence on the effectiveness of different types of community sentences.

June 2023