An inspection of Probation Services – Wales region

Foreword (Back to top)

We found much that was positive about the Probation Service in Wales. There was strong leadership both in the probation delivery units (PDUs) and regionally, and there were many structures and systems in place to improve practice and support people on probation. Nevertheless, as we have found in every region we have visited to date under our current inspection model, keeping people safe continued to be the area of work that undermined much of what was positive. We saw this not just in the casework we looked at in PDUs, but in other work managed regionally, including work at courts, unpaid work and resettlement. Leaders were not oblivious to this, and had done much to try to address it, but as yet with only limited effect.

As a consequence, the overall rating for Wales was ‘Requires improvement’.

Senior leaders had worked hard, were well respected and had developed strong relationships with partners, both statutory and in the third sector. They were working well with His Majesty’s Prison and Probation Service (HMPPS) and with the Welsh government. This had led to well-integrated planning and engagement that resulted in a number of positive initatives. However, as seen across all recent regional and PDU inspections, more was needed to address the key issue of keeping people safe, ensuring the right information was being received from partners and that practitioners were using the information suffiently to plan and manage risk effectively. More also was needed to improve communication by practitioners with partner agencies and to ensure the right services were consistently available for people on probation to address desistance and reoffending.

The creation of the probation planning, assurance and implementation team was a promising initiative as a model to support quality assurance, and there were signs that this was already having some impact on the quality of work. We were pleased to see indications that, as a model, its remit was being extended, including the safguarding audit carried out during our inspection.

Other postive initiatives to improve effectiveness included the Centralised, Operational, Resettlement, Referral and Evaluation (CORRE) Hub to support desistance, whose work might also be extended. The adoption of a Human Factors model across the region was also supporting staff and teams to become more engaged and to focus more closely on their practice. The region had also been enthusiastic trailblazers for the piloting of Justice Transcribe, a tool which had significant potential to transform the day to day work of probation.

While the initiatives and innovation undertaken by leaders in Wales are positive, it is important to ensure that they have the right impact on frontline activity and that, in particular, they address the key issues around keeping people safe. I wish leaders every success in these endeavours.

Martin Jones CBE

HM Chief Inspector of Probation

---

Ratings (Back to top)

Fieldwork started July 2025

Score 5/24

Overall rating

Requires improvement

1.Organisational arrangements and activity

R 1.1 Leadership	Requires improvement
P 1.2 Staffing	Requires improvement

2. Service delivery

R 2.1 Public protection	Inadequate
R 2.2 Desistance	Inadequate
R 2.3 Court work	Inadequate
R 2.4 Unpaid work	Inadequate
R 2.5 Resettlement	Requires improvement
R 2.6 Victim work	Good

---

Executive summary (Back to top)

Introduction

The Probation Service in Wales is one of 12 across England and Wales. Combined with the five public and one private sector prison, it is part of the Wales region of HMPPS. The region consists of six PDUs: Swansea Neath Port Talbot, Cwm Taf Morgannwg, Dyfed Powys, North Wales, Gwent, and Cardiff and the Vale of Glamorgan.

At the time of the inspection, the Wales probation region was managing 13,427 people on probation, including 3,038 held in custody. Of these, 5,667 were subject to multi-agency public protection arrangements (MAPPA).

At the time the inspection was announced (May 2025), the region employed a total of 1,418 full-time equivalent staff, with a vacancy rate of seven per cent. Across the six PDUs, the average caseload was 33.28 for a probation officer (PO), 39.46 for a probation services officer (PSO), and 27.81 for a Professional Qualification in Probation (PQiP) trainee. Regional Workload Management Tool data indicated that workloads had, on average, come down for practitioners in the 12 months before our inspection, from 102 to 97 per cent for POs, 93 to 85 per cent for PSOs, and 88 to 74 per cent for PQiPs.

Methodology

We conducted fieldwork in each PDU across Wales between 28 July 2025 and 26 September 2025. We inspected 338 cases, of which 224 were subject to a community sentence and 114 were subject to release on licence. From each of these cases, we collated data for our public protection and desistance ratings. We conducted 206 interviews with probation practitioners. We also inspected 265 court reports and 107 cases subject to resettlement provision. We inspected 58 UPW cases and 28 statutory victim cases from across the region where community sentences and licences had begun between 24 March and 04 April 2025.

---

1. Organisational arrangements and activity

R1.1 Leadership

Many aspects of leadership across the Probation Service in Wales were impressive, with a strong focus on innovation to drive service delivery. Nevertheless, despite well-focused initiatives and clear leadership overall, work to keep people safe remained insufficient, and this shortfall permeated most aspects of work across the region and the six PDUs.

Senior leaders had a clear vision of where they wanted to take the service and, in many respects, how they intended to do so. Much of this was appropriately linked to the wider work of both HMPPS and the Welsh government. Priorities were clearly defined, and these formed the primary focus of activities. They also linked directly to the work of PDUs, and we saw good examples of how PDU leads worked well with regional leaders and heads of function to ensure these priorities were reflected in their local delivery.

Leaders had created a supportive culture, which was widely confirmed by those we spoke to, and this had formed a strong basis on which to build practice. A number of positive initiatives had been introduced to inform the service of where its shortfalls lay and to anticipate future pressures. The change board was developing pilot projects to support service developments. The Justice Transcript artificial intelligence tool was a notable example of this.

Services were commissioned through the collaboration and commissioning forum to meet the needs of people on probation. This demonstrated how senior leaders worked with partner agencies and stakeholders to develop provision for people on probation.

Leaders had taken a positive approach to addressing identified shortfalls in some aspects of service delivery. The development of the probation planning, assurance and implementation team since our last inspection was the culmination of this. The creation of the Wales Assurance Tool (WAT) focused on quality assurance and building effective practice, with the first round of activity (May 2025) largely reporting what we found in our inspections. We hope that these findings will provide a springboard to improvement.

Diversity and inclusion was a strength in Wales. There was a strong emphasis on developing specialised provision for those needing it. The introduction of the disproportionality taskforce also demonstrated a commitment to exploring and addressing significant issues of concern.

More work was necessary to focus attention on keeping people safe, as this was the area of work that consistently undermined the effectiveness of service provision. Leaders needed to ensure that consistent and accurate information was received from partner agencies, in particular the police and children’s services, and to develop staff’s confidence and ability to analyse all information to inform both assessments and planning.

---

R1.2 Staffing

Staffing levels across the Probation Service in Wales were generally reasonable, with a vacancy rate of around seven per cent. Workloads were not excessive, and most staff found them manageable. The region recognised some of the shortfalls in key staffing areas and had recruited a number of additional posts, including extra quality and scrutiny managers in the serious further offence team (SFO) and a positive actions officer. There was a clear strategic workforce plan in place, with workforce leads in each PDU and escalation routes to the regional leadership team.

There was a strong focus on learning and development, and good completion rates of core aspects of training. The embedding of quality development officers (QDOs) and protected learning days added to this approach. Similarly, levels of supervision were generally good and most staff found it helpful in developing their practice. The region had prioritised support for senior probation officers, and had introduced a management coordination hub and delivered a range of specific training events.

The Human Factors model, with an emphasis on creating a supportive learning culture within the service, was also having a positive impact across the region. Most staff spoken to were positive about the way it was being used. They said it supported them by providing the space and time to focus on their practice, as well as helping them in decision-making.

Nevertheless, pressures remained on some staff where there were pockets of significant staff shortage, and sickness levels remained too high overall. Attempts by leaders and PDU heads to mitigate the worst effects of sickness were to be commended but too many staff across PDUs still felt their welfare was not sufficiently attended to and most felt that there were not enough staff in the service.

Managers’ oversight of work was insufficient across most aspects of the service that we reviewed. This undermined much of the good work being done with people on probation.

---

2. Service delivery

2.1 Public protection

Work relating to keeping people safe was the weakest aspect of service delivery across the Probation Service in Wales. In all four key questions relating to safety in our casework review, less than half of cases met our standard for sufficiency. This was the case across five of the six PDUs we visited. While North Wales PDU was better, with the best overall results across our inspection programme so far, more work was still required there.

Senior leaders appeared to have good relationships with both police and children’s services across Wales, with good communication and protocols. Information-sharing agreements with the four police services in Wales and the 22 local authorities were hard won and a positive step. However, these agreements were still not working as intended, and we saw variations as we went to different PDUs.

Although a relatively low number of enquiries to the police and children’s services were not responded to, in far too many the information received was not sufficiently detailed to fully inform an assessment or support a sentence plan. On too many occasions when this happened, practitioners appeared to lack the confidence to escalate the concern or request more information, or did not know how to do so. Engagement with partner agencies to support public protection was also limited and was too often absent in both the planning and implementation of work with people on probation. The police provided daily information on reportable incidents across Wales. This was an excellent initiative, and we saw many cases where it improved the management of individuals. However, in some cases practitioners were using this information as an alternative to making direct enquiries at the relevant time to inform their practice.

Although our case reviews indicated that practitioners did generally see people on probation sufficiently frequently to manage and minimise risk of harm, it was evident that this did not consistently result in risk being managed effectively.

---

2.2 Desistance

Across all six PDUs, the assessment and planning of work to promote desistance and reduce reoffending were generally both consistent and effective in most cases, practitioners drew on all available information to inform this work and focused sufficiently on individuals’ needs.

The role of the CORRE Hub across Wales was central to these positive findings. Practitioners from the Hub worked with the practitioner holding the case, which helped to focus attention on the key issues relating to desistance, set appropriate targets and ensure that referrals were made to relevant services. However, people on probation were not involved in these discussions, and it was therefore not clear how well they were engaged and committed to targets set for them.

While we saw some very good work to address issues relating to reoffending, for instance around housing, this was not consistent enough. Although Commissioned Rehabilitation Services (CRS) were provided in around three-quarters of all cases we reviewed, only a little over half of them were assessed as the services most likely to support desistance.

Our inspections took place at a time when, in all regions across England and Wales, services were transitioning to the new suite of accredited programmes. As a consequence, we did not see many examples of this work in our casework. However, we did see a number of cases where structured interventions were being delivered by PDU practitioners, and in many of these this work was viewed positively by inspectors.

---

2.3 Court work

Despite generally positive satisfaction in the most recent (2025) sentencer survey undertaken by Wales, the information provided to courts was sufficiently analytical and personalised in less than a third of the cases we reviewed across all PDUs.

A number of factors impacted on this work. Staff shortages across some parts of Wales had put pressure on staff, and this was compounded by the lack of an alternative casework management tool, as was available in offender management teams.

Some aspects of court work were positive, for example the focus on issues relating to the likelihood of reoffending and the individual’s motivation to change. However, there were shortfalls in relation to domestic abuse and child safeguarding enquiries. This varied across PDUs and appeared largely to reflect the different approaches taken by police forces and children’s services, despite information-sharing agreements being in place across Wales with both. This was also what we found in the broader picture of public protection.

Practitioners did not consistently use the available escalation routes, and in some cases did not appear to recognise the shortfalls in some of the information they received. In broad terms, there were too few cases where the practitioner had used all available information to inform reports presented to courts. We were nevertheless pleased to note that in a reasonable majority of the cases we reviewed, the advice to courts did include factors related to risk of ham.

---

2.4 Unpaid work

Although it was encouraging that, overall, UPW had achieved well against its performance targets, some elements of delivery were not consistently effective. These shortfalls manifested in the early stages of casework relating to assessment and planning. As we found in many aspects of the work we reviewed during our inspections, in too many cases the practitioner failed to identify all factors relating to risk of harm or to engage sufficiently with the other people (mainly UPW staff) involved in the case.

Leaders had emphasised the need to improve breach procedures and enforcement action, and we were encouraged to see this reflected in the cases we reviewed.

There was also a strong emphasis on the rehabilitative element of UPW, and we saw this reflected in many of the cases we reviewed. We were also encouraged to see a number of work initiatives to support those on UPW with a protected characteristic.

---

2.5 Resettlement

We saw some positive work in relation to resettlement, particularly in identifying and addressing issues around desistance, during our community-based casework across all PDUs. In a reasonable majority of cases, community offender managers were identifying key resettlement and desistance issues and were generally also working to resolve them They were supported in this work by colleagues in the CORRE team.

Again, as we found with our broader review of casework, shortfalls in resettlement provision centred on keeping people safe. The community offender manager identified and addressed issues relating to harm in less than half the cases we looked at where this was relevant.

We did note some variations across the six PDUs. However, the PDUs with the highest proportion of cases assessed as sufficient had maintained the tasking function (allocating specific activity to individuals) of the former short-term sentence teams or had kept the teams in place. This had not happened in the PDUs with lower scores.

---

2.6 Victim work

The structure and organisation of statutory victim work in Wales had been reorganised since our last inspection in 2021 to create the Wales Victim Unit. This incorporated all victim work staff across Wales and was designed to offer greater assurance of the quality of work undertaken. At the time of the inspection, the unit was managing over 3,000 cases.

There were a number of positive features of this work. Initial contact was consistent and supportive, and victim liaison officers (VLOs) were making every effort to support and encourage victims to take advantage of the scheme. In the relevant cases, victims were supported in making representations to the parole board and in suggesting no-contact conditions.

Although in a reasonable majority of cases there was appropriate information exchange and communication between VLOs and probation practitioners, further attention was necessary to ensure this extended to MAPPA meetings and reviews.

---

Recommendations (Back to top)

As a result of our inspection findings, we have made a number of recommendations that we believe, if implemented, will have a positive impact on the quality of probation services.

The Probation Service in Wales should:

1. improve arrangements for information exchange with both the police and children’s services to ensure the information shared is consistent, timely and sufficiently detailed to inform assessments and plans to keep people safe
2. ensure that MAPPA level setting for custody and community cases is timely, taking into consideration the earliest possible date of release and any temporary releases, and be fully informed by information from all relevant agencies in all cases
3. improve practitioners’ information-sharing and liaison with partner agencies to support the effective delivery of rehabilitative services and work to keep people safe
4. improve the quality of court reports to inform sentencing, ensuring that information on domestic abuse and child safeguarding is obtained and used effectively
5. review the delivery of CORRE, and consider including people on probation in casework planning and its routine use in case reviewing
6. evaluate the effectiveness of current resettlement tasking arrangements and consider whether a consistent model across all PDUs would improve delivery.

---

Background (Back to top)

Wales is one of 12 regions of the Probation Service in England and Wales. It is made up of six PDUs: Swansea Neath Port Talbot, Dyfed Powys, North Wales, Cwm Taf Morgannwg, Gwent, and Cardiff and the Vale of Glamorgan. The service employs 1,524 staff. Wales has a total population of 3,186,581.

Across Wales, there are four approved premises (Ty Newydd, Plas Y Wern, Quay House and Mandeville House), five public sector prisons (HMPs Berwyn, Usk, Prescoed, Cardiff, and Swansea), and one private prison (HMP Parc).

Wales currently has four police services (North Wales, South Wales, Gwent, and Dyfed Powys), 22 local authorities and 17 youth justice services. There are 10 magistrates’ courts and three Crown Courts.

While criminal justice services are not devolved in Wales, some services that impact on the probation service are, including health services and housing. Criminal justice work in Wales is largely coordinated by the Criminal Justice Board for Wales, with work managed through area planning boards, public services boards and Community Safety Partnerships.

At the time the inspection was announced, the region was responsible for managing a total caseload of 13,427. Of these, 6,669 were subject to a community order, 3,720 were subject to a post-sentence licence, and 3,038 were in custody.

The total number of offenders subject to MAPPA across Wales is 5,667. Of these, 42 per cent are in South Wales.

Regional management arrangements are in place to manage some service functions, including UPW, accredited programmes, and victim liaison services.

---

1. Organisational arrangements and activity (Back to top)

R 1.1. Leadership	Rating
Regional leadership drives the delivery of a high quality, personalised and responsive service for all people on probation.	Requires improvement

Strengths

• The integration of probation across Wales with HMPPS and also the Welsh government remained a significant strength. It offered a substantial opportunity to improve identified shortfalls in the casework and create a strong service.
• Senior leaders articulated and managed the strategic vision of the service through a range of criminal justice forums and boards. They promoted a clear agenda, based on defined priorities that linked to the priorities of both the Welsh government and HMPPS. Regional strategic boards within the service included regional leaders, functional heads and PDU leads. This ensured that activity driven centrally was consistently reflected at PDU and operational level.
• Across the region as a whole, there was a strong and supportive staff culture. Seventy-two per cent of respondents in our regional survey and an aggregate 61 per cent in our PDU surveys said that the culture promoted openness, constructive challenge and ideas always or most of the time.
• Commissioning of services reflected the engagement of partner agencies and other organisations that worked with the service. The development of the collaboration and commissioning forum was an excellent example of this working in practice. Feedback from people on probation, the courts, third-sector organisations and the disproportionality taskforce was brought together to determine commissioning priorities. A wide range of provision had been commissioned in this way, both at PDU and regional level. This included the DRIVE domestic violence programme in Swansea prison, a brain injury support initiative (BrainKind), and a health pilot in Cardiff.
• The Women’s and Youth Justice Blueprints, an integrated strategy across devolved and non-devolved services across Wales, further demonstrated the importance of developing services collaboratively across the region with support from both HMPPS and the Welsh government.
• There was a strong emphasis on equality, diversity and inclusion across the region. This included work with criminal justice agencies as part of the anti-racism taskforce. The regional disproportionality taskforce had taken a proactive approach to identifying, monitoring and addressing disproportionality across service outcomes. Examples included work relating to employment outcomes for women and the completion of UPW for those with a disability. An appropriate focus on diversity, equality and inclusion extended to PDUs where, in a large majority of cases, people on probation were asked about their diversity and protected characteristics. In two-thirds of all cases, practitioners considered these when assessing the person’s ability to comply with their order.
• The introduction of both the change and targeting boards demonstrated a strong commitment by senior leaders to anticipating change, planning resources effectively and supporting innovation. The latter was a prominent feature across all six PDUs; for example, the Justice Transcribe artificial intelligence tool was being used to record cases quickly and efficiently. By the conclusion of our regional inspection, we were told it had already been adopted by 90 per cent of all practitioners.
• Senior leaders were acutely aware of pressures on staff, given the number of national policy changes in the previous two years. The introduction of the Human Factors model to create a supportive culture across the service had been embraced by the majority of staff.
• The development of the probation planning, assurance and implementation team had also been a positive, and unique, initiative in Wales. Creating a cycle of activity from tier one assurance to PDU-based accountability reviews and with a commitment to improvement activity, the focus was primarily on quality and effectiveness and less on audit. The introduction of the WAT had also been a positive initiative. While this was still a relatively new initiative at the time of the inspection, there were signs that it could support improvement in the quality of casework.
• Although forums for people on probation were not consistently in place across all PDUs, leaders focused on engaging with people on probation and eliciting their views about the service provided by the region. They used this information to support development across PDUs.
• The CORRE Hub, designed to improve the quality of planning in casework, was demonstrably effective. The plans in 261 of the 338 cases we reviewed (77 per cent) focused sufficiently on reducing reoffending and supporting desistance. In some PDUs the level of sufficiency was even higher.
• The four police services across Wales shared information on reportable incidents with probation staff. This was a positive approach to managing the risks posed by people on probation and was indicative of the work done by senior leaders to foster positive relationships with stakeholders and to build confidence and trust in such processes.
• Probation and police services across the region had negotiated an information-sharing agreement to facilitate domestic abuse enquiries by the probation service. This was further evidence of positive relationships at a strategic level. A similar agreement had been struck with the 22 children’s services departments in Wales, so that information could be exchanged to keep children safe. However, leaders acknowledged that, despite these agreements, more work needed to be done to ensure that there was consistency in the detail received and its application by practitioners across the region.

Areas for improvement:

• Although there were some extremely positive and innovative initiatives to manage services effectively, these had not consistently resulted in delivery of a high-quality service for all people on probation.
• While the CORRE Hub consistently supported effective assessments and planning, in all six PDUs the number of cases assessed as sufficient in terms of delivery was much lower relative to assessment and planning, despite sufficient contact by practitioners with people on probation.
• Both desistance and keeping people safe were rated as inadequate across the region. Court work and UPW were also rated as inadequate overall across the region.
• In five of the six PDUs, the number of cases assessed as sufficient to keep people safe across each of the four AsPIRe (Assessment, Planning, Implementation and Reviewing) standards was inadequate. When aggregated across all six PDUs, the number of sufficient cases remained well below half.
• In a reasonable majority of core cases, people on probation were being seen often enough by practitioners to potentially manage risk. However, too often practitioners appeared unclear about when, where or how to follow up information to inform effective assessments and plans to keep people safe. There was sufficient multi-agency working in only a minority of cases.
• Despite the model of reportable incidents, staff did not always know when or how to use the information they had received. In some cases, staff saw this information as an alternative to making specific domestic abuse enquiries.
• Although information-sharing agreements were in place with children’s services across Wales, in far too many cases the information received was not sufficient for practitioners to make informed judgements in their assessments of risk. Work to address some of these shortfalls had not yet had a sufficient impact.
• More work was needed to ensure the effective management of MAPPA. At the time of this inspection, a substantial number of MAPPA-eligible cases required a management level to be set and/or reviewed. As a consequence, there was a potential that the risk posed by some people would not be sufficiently managed.

---

R 1.2 Staffing	Rating
Staff are enabled to deliver a high-quality, personalised and responsive service for all people on probation.	Requires improvement

Strengths

• Workloads for regional staff were not excessive, with 71 per cent of those completing our regional survey saying their workloads were manageable. Although across the six PDUs we inspected only 46 per cent of survey respondents said their caseloads were manageable, this increased to 60 per cent of those we interviewed (206, all frontline practitioners). Of the practitioners who completed our PDU survey, 74 per cent said they had a caseload of between one and 40.
• Workloads did not have a negative impact on the level of contact with people on probation. Of the 316 core cases we reviewed, contact by the probation practitioner was sufficient to reduce reoffending and increase desistance in 200 out of 254 relevant cases. There were also sufficient levels of contact to manage risk.
• There was a strong culture of support and supervision across the region. In our regional survey, 82 per cent of respondents said they received supervision sufficiently and frequently, as did 67 per cent of those completing the survey across PDUs. Also, 73 per cent of staff in our regional survey and 65 per cent of staff across the six PDUs said that they received supervision that improved the quality of their work. QDOs had been introduced as part of the probation planning, assurance and implementation team in all six PDUs, which had further improved supervision. This role usually focused on supporting practitioners and developing their skills.
• Leaders were working hard to embed a learning culture across the service. Completion levels for core training (for example on safeguarding and domestic abuse) were high. Completion rates for the SEEDS2 (Skills for Effective Engagement, Development, and Supervision) training were also high; all leaders and 93 per cent of managers had completed it.
• In our regional survey, 106 of the 128 respondents who answered the question said that the service actively promoted a culture of learning and continuous improvement. Almost three-quarters of all staff across PDUs also expressed this view.
• We also saw practice improvement and development across all PDUs, particularly in the form of protected learning time, daily team briefings and the adoption of the Human Factors model. Leaders had introduced the situation, background, assessment, and recommendation checklist, which was helping staff to learn while maintaining individual responsibility and autonomy.
• Senior leaders had positively embraced a model of rotating PDU heads and functional leads on, broadly, a three-year cycle. This ensured a fresh approach to local leadership and encouraged insight into the various activities of the service.
• Leaders across the service acknowledged that both staff turnover (at 9.4 per cent across the region) and sickness (13.2 average days lost per year) were too high. Work to address this had been identified as a priority, and leaders were analysing data on absence and setting clear actions. This included a monthly absence panel, which reviewed the most complex cases. There were indications that this approach was achieving some, albeit small, improvements.
• There had been a strong focus in the preceding year on developing support for middle managers, and in particular senior probation officers. For example, a regional management coordination hub had been created to provide support for many functional activities. A variety of training events had been rolled out under the banner of ‘people management essentials’, including supporting attendance, conducting investigations and conduct and discipline. These had been generally supportive and empowering. We saw examples of these approaches being used in PDUs. This also reflected the philosophy of ‘unapologetic management’ that was followed by senior leaders.
• In most PDUs, we saw evidence that managers were actively managing pressures on staff created, primarily, by levels of sickness. Across the region 84 per cent of all sickness was classified as long term.

Areas for improvement:

• Sickness levels across Wales were too high. In each of the six PDUs, the average number of days lost to sickness was slightly higher in the 12 months before this inspection was announced than in the previous year. This was particularly the case for PSOs and case administrators. Attrition rates had stayed consistent in this same timeframe.
• Staffing levels were reasonable across most PDUs. However, sickness levels were having a significant impact on staff, with 83 per cent of respondents to our staff survey saying that there were not enough staff in the service. This view was also reflected in our regional survey, although at a slightly lower level of 60 per cent of respondents.
• Furthermore, despite considerable attempts by PDU leaders to respond to the needs of staff across the region, only 53 per cent of staff in our survey felt that sufficient attention was paid to their welfare. This is often a sign of a stressed workforce, and the sickness levels within the service were a further indication of this.
• Although leaders had made a push to improve supervision, created support for middle managers and introduced the Human Factors model, management oversight was generally insufficient across all the casework we reviewed. This was of particular concern given the number of inexperienced staff in the workforce, and the lack of oversight that we identified in serious further offence cases. Leaders needed to give greater attention to this crucial element of case management.

---

Diversity and inclusion (Back to top)

• The region had a strong emphasis on equality, diversity and inclusion. Priorities were clearly defined and managed through the probation equality board. This was underpinned by the work of the regional diversity and inclusion team, which was integrated well within HMPPS.
• An understanding of the diverse needs of people on probation was well established across all six PDUs. Practitioners had asked about diversity and protected characteristics in a large majority (82 per cent) of the cases we saw. In a reasonable majority of core cases (66 per cent), they had assessed the impact of these on the person on probation’s ability to comply and engage with service delivery.
• The Welsh language played a significant role in the PDUs we inspected. We were assured that, when requested, interviews and other work could be carried out in Welsh. The service’s commitment was reinforced through a dedicated strategic action plan reviewed quarterly within the organisation and annually with the Welsh language commissioner. We were encouraged to see many documents available in Welsh.
• It was also encouraging to see that the Wales anti-racism taskforce had been integrated within the service. This is an initiative between the Welsh government, criminal justice partnerships and the probation service. At the time of the inspection, over 250 line managers had completed the leadership module of the inclusive behaviours programme and cultural awareness training, and 450 staff had completed the all-staff module.
• The Wales disproportionality taskforce was also a positive initiative. While it was relatively new (introduced in early 2024), there were already indications that it was identifying and taking action in some key areas, such as with women and those with disabilities. It was also trying to improve the disproportionately low employment opportunities for those aged between 40 and 49. The board had recently commissioned a study to examine why a disproportionate number of women and Black, Asian and minority ethnic people on probation were recalled.
• We saw diversity champions across all six PDUs. It was encouraging that Wales employed a positive action officer, who worked with diverse communities to promote recruitment to the service.
• Work undertaken in diversity and inclusion was resulting in the commissioning of services to meet the needs of some minority groups across PDUs

Areas for improvement:

• Despite the commitment shown to the Welsh language, we were not assured that all engagement with people on probation and the public supported the principle of ‘language courtesy’, such as answering the telephone in Welsh first and English second. There was a need to strike a consistent balance between business needs and a commitment to the principle of using the Welsh language.

---

2. Service delivery (Back to top)

R2.1 Public Protection	Rating
High-quality, personalised, and responsive services are delivered to protect the public.	Inadequate

Our rating[1] for public protection is based on the percentage of cases we inspected being judged satisfactory against four key questions and is driven by the lowest score:

Key question	Percentage ‘Yes’
Does assessment focus sufficiently on keeping other people safe?	35%
Does planning focus sufficiently on keeping other people safe?	43%
Does the implementation and delivery of services effectively support the safety of other people?	34%
Does reviewing focus sufficiently on keeping other people safe?	42%

• Overall, the level and nature of contact practitioners had with people on probation to manage and minimise risk was sufficient in over 70 per cent of cases. We also found that home visits were undertaken where necessary in 191 of 286 relevant cases to manage the risk of harm. However, this did not result in effective management of risk in a sufficient number of cases.
• In our PDU inspections, less than half of the cases we reviewed were assessed as meeting the sufficiency standard for all four key questions relating to keeping people safe. The level was slightly higher than half in North Wales PDU but was still assessed as ‘Requires improvement’.
• Information-sharing agreements were in place both with the four police services across Wales and with the 22 local authorities. However, these were not having a sufficient impact on keeping people safe. We saw variations in how these agreements were being interpreted. For example, we found that data provided by South Wales police was heavily redacted in many cases, while redaction was not a significant factor in North Wales. A protocol for sharing reportable information was also in place across all four police forces. In many respects, this was a very positive initiative, and we saw the information being used in a number of cases. However, in other cases, practitioners were using this an alternative to requesting specific information about domestic abuse.
• In most cases where domestic abuse or child safeguarding enquiries were made at the start of a community order or at the point of release from custody response were received. However, in too many cases, the information received was either insufficient to fully inform assessments or no enquiry was made. In 236 of 338 cases, practitioners received sufficient information about domestic abuse from the police to inform assessments. However, in 40 cases the information received was not sufficiently detailed, and in a further 48 no information was requested at all. Information on child safeguarding was worse. We assessed that this was required in 327 cases. However, in 100 of these, the information received was not sufficiently detailed and in 43 the information was not requested.
• In far too many cases practitioners appeared to either lack the confidence to make enquiries or did not know how to escalate the request or make further enquiries. This was further compounded by too few practitioners drawing on other forms of information, either held on file or with partner agencies. Overall, this lack of professional curiosity meant that in too few cases sufficient information was used to inform assessments or related plans to keep people safe.
• Work by practitioners to keep people safe and work with other agencies was also poor. Again, this spoke of a lack of confidence and/or knowledge in working with other agencies. During planning, practitioners made appropriate links to other agencies involved with the person on probation and multi-agency plans in only 147 of 299 relevant cases. Furthermore, there was effective multi-agency working, including information-sharing, in only 84 of 235 relevant cases regarding safeguarding children, and 101 of 241 relevant cases regarding domestic abuse.
• More work was necessary to ensure information from both the police and children’s safeguarding services was requested and was sufficiently detailed to inform work to keep people safe. More work was also required to ensure all practitioners knew what information was needed to manage people on probation safely and to be professionally curious in their work with people on probation and partner agencies.

---

R 2.2 Desistance	Rating
High-quality, personalised, and responsive services are delivered to promote desistance.	Inadequate

Our rating[2] for desistance is based on the percentage of cases we inspected being judged satisfactory against four key questions and is driven by the lowest score:

Key question	Percentage ‘Yes’
Does assessment focus sufficiently on the factors linked to offending and desistance?	72%
Does planning focus sufficiently on reducing reoffending and supporting desistance?	77%
Does the implementation and delivery of services effectively support desistance?	47%
Does reviewing focus sufficiently on supporting desistance?	59%

• Assessment and planning regarding desistance was generally strong across all six PDUs, although slightly less so in Cardiff and the Vale of Glamorgan. In most cases assessments identified and analysed offending-related factors and in 78 per cent of all cases assessments identified and analysed the strengths and protective factors of the person on probation. In contrast to what we found with keeping people safe, 62 per cent of assessments regarding desistance did draw sufficiently on available sources of information.
• Similarly, in 78 per cent of cases, planning set out the services most likely to reduce reoffending and support desistance. In 243 of 319 relevant cases, planning built on the individual’s strengths and protective factors, drawing on potential sources of support.
• In large part this appeared to be due to the work of the CORRE Hub. Most staff spoken to across the region spoke positively of this model. They saw it as a helpful approach to focusing on the needs of people on probation and facilitating access to the most appropriate services.
• However, meetings held between probation practitioners and CORRE workers did not include people on probation. As a consequence, it was not clear how well people on probation were engaged with the plan created for them or how willing they were to work with the services they were referred to. This was largely reflected in our findings. Although we found that CRS were provided in 203 out of 276 relevant cases, these were not always sufficient. In many cases, while the sentence plan comprehensively identified the services needed by the person on probation, less attention had been given to sequencing them. In only 52 per cent of relevant cases were the delivered services those most likely to reduce reoffending and support desistance, with sufficient attention given to sequencing and available timescales. Too often, this was further compounded by a lack of engagement by practitioners with partner agencies. The involvement of other organisations in delivering services was sufficiently well coordinated in only 140 out of 284 relevant cases, despite many service providers being co-located in probation offices. This was particularly disappointing given that, in 200 out of 254 relevant cases we reviewed, the level and nature of contact were sufficient to reduce reoffending and support desistance.
• Across the PDUs we visited we did see some good work in relation to providing accommodation. This reflected the positive partnership work carried out through the HMPPS Wales accommodation framework and Wales Accommodation Board. Sufficient services had been delivered in 100 of 164 relevant cases that we reviewed where accommodation was identified as a need.
• At the time of our inspections across Wales, accredited programmes were just beginning to adopt the Building Choices programmes (June 2025), following the national transition and training for staff between March and June. Although we saw very few people on probation on programmes in the cases we looked at in our inspections, we did see some who were subject to structured interventions and delivered by probation practitioners. We were told that 176 people on probation were currently (October 2025) on a programme. The programme team had a target to reduce the current waiting list to three months within the next year.

---

R 2.3 Court work	Rating
The pre-sentence information and advice provided to court supports its decision-making.	Inadequate

Our rating[3] for court work is based on the percentage of cases we inspected being judged satisfactory against one key question:

Key question	Percentage ‘Yes’
Is the pre-sentence information and advice provided to court sufficiently analytical and personalised to the individual, supporting the court’s decision-making?	32%

• Engagement by senior leaders with sentences across the region was generally positive. In the most recent survey of sentencers (2025) the region received a 77 per cent satisfaction rate. The region had established a quarterly strategic sentencers forum and produced a regular newsletter.

• There had a been a push from regional leaders and the regional court lead to focus strongly on the principle of ‘start right – end right’, emphasising the importance of court reports and engagement with people on probation. There were some challenges, however. Across the region, only 84 per cent of court staff were in post. Some areas were struggling more than others for staff, although attempts to improve this had had some impact more recently. Compounding this was a lack of an alternative to the workload management tool used for offender management. As a consequence, it was difficult to know how manageable workloads were for court staff.

• There were some strong elements to court work. In 86 per cent of cases we reviewed, the individual was meaningfully involved in preparing the report and their views were included. Advice to the court considered factors related to the likelihood of reoffending in 81 per cent of cases and the individual’s motivation and readiness to change in 72 per cent of cases. In 85 per cent of cases, reports also included a sufficient record of the advice given and the reasons for it.

• However, there were some significant, and consistent, omissions. With police domestic abuse enquiries, we found that sufficient information was received in time to inform reports in only 48 per cent of cases where this was required. However, there were considerable variations across PDUs. Sufficient information was provided in 80 per cent of cases in North Wales, but in only 37 per cent of cases in Cardiff and the Vale of Glamorgan and 28 per cent of cases in Dyfed Powys. We found further variations in the proportion of cases where the information received was sufficiently detailed and where requests for information were not responded to. For instance, in Dyfed Powys, the information received was sufficiently detailed in all cases. However, in 17 cases (59 per cent), the information requested was not received in time to inform the report. Such variations largely reflected what we found in enquiries to police generally. Police forces took different approaches to information-sharing. For example, South Wales police often redacted information, despite a Wales-wide information-sharing protocol. Information was not requested when it should have been in 28 out of 251 relevant cases across all PDUs.

• Enquiries related to children’s safeguarding were sufficient in only 117 out of 252 relevant cases across all PDUs. The variations across PDUs were less stark than for domestic abuse enquiries. However, these variations largely reflected relationships struck with services in each PDU, despite an information-sharing protocol being in place across all 22 services. In 31 of 252 cases reviewed information was not requested when it should have been.

• Overall, in 14 per cent of relevant enquiries to the police and 21 per cent to children’s services, the information received was insufficient. Although escalation routes were in place for such occurrences, these were not being used sufficiently. There was a lack of professional curiosity on the part of report authors, which was a missed opportunity to explore potential concerns in more detail.

• Leaders had attempted to address the delays in information-sharing from safeguarding services and the police, which had resulted in some PDUs piloting safeguarding enquiry officers. However, on the basis of our findings, these were having only a limited impact to date.

• Audits using the court case audit tool did not sufficiently identify underpinning issues such as a lack of analysis of information received from child safeguarding enquiries. There was a move to extend this work and include support from the probation planning, assurance and implementation team and WAT reports which was anticipated to support quality improvement.

• Despite the limitations of the information received at court about safeguarding and domestic abuse, we were encouraged to note that advice to courts included factors related to risk of harm in 70 per cent of cases.

---

R2.4 Unpaid Work	Rating
Unpaid work is delivered safely and effectively, engaging the person on probation in line with the expectations of the court.	Inadequate

Our rating[4] for unpaid work is based on the percentage of cases we inspected being judged satisfactory against four key questions and is driven by the lowest score:

Key question	Percentage ‘Yes’
Is the assessment and planning of unpaid work personalised?	45%
Do arrangements for unpaid work maximise rehabilitative elements and support desistance?	74%
Is unpaid work delivered safely?	64%
Is the sentence of the court implemented appropriately?	76%

• Regional leaders, including the regional lead for UPW had focused on improving both performance and activity. They had introduced an UPW taskforce to focus on these issues, which had had some success. In May 2025 UPW had exceeded five out of its six primary performance targets, including the target for completion of UPW (92 per cent). The standdown rate was 1.25 per cent, against a national target of 2.5 per cent. In the previous 12 months there had been a strong emphasis on reducing the number of dormant cases. At the time of our inspection announcement the percentage of UPW cases with hours outstanding beyond 12 months was five per cent, down from nine per cent 12 months earlier.
• Middle managers and UPW staff met regularly across PDUs to resolve obstacles and to improve levels of enforcement. This was having some impact, as we found that enforcement action had taken place where required in 27 of the 35 cases where it was relevant, and appropriate professional judgements were made in relation to missed appointments in 35 out of 48 relevant cases.
• Across the region there was a strong emphasis on supporting rehabilitative activity and desistance for those subject to UPW. In 28 out of 47 cases reviewed, UPW offered opportunities to develop employment-related opportunities. We saw examples of appropriate action to link those undertaking UPW to placements appropriate to their needs and/or interests. Various initiatives were in place to support diverse needs, including a workshop for disabled participants and those with needs related to neurodiversity. At the time of the regional inspection, the service was about to recruit a further female supervisor, which would enable it to provide women-only projects across all PDUs.
• However, in only half the cases we reviewed did assessment consider the diverse needs of participants and their likely impact on compliance and engagement on UPW. More significantly, and as we found across most aspects of work in Wales, assessment and planning for unpaid work identified factors related to risk of harm in less than half of relevant cases.
• In too many of the cases we inspected, assessments did not consider all factors affecting the individual’s ability to complete their order. In some cases, the electronic assessment tool was considered too basic and was completed perfunctorily. There were some indications that the pre-placement work session was seen as an opportunity to explore risk further. However, this was inappropriate, as the session was provided by UPW staff who were not trained in such assessments. In too many cases there was insufficient communication between probation practitioners undertaking assessments and UPW staff. Partnership working between all staff involved in a case was evaluated as effective in only 34 per cent of the cases we reviewed.
• A new risk sharing process was launched in Wales (and nationally) in July 2025 and leaders hoped that this would improve some of the shortfalls we identified.

---

R2.5 Resettlement	Rating
Resettlement work is timely, personalised, and coordinated, addressing the individual’s resettlement needs and supporting their integration into the community.	Requires improvement

Our rating[5] for resettlement is based on the percentage of cases we inspected being judged satisfactory against one key question:

Key question	Percentage ‘Yes’
Is resettlement timely, personalised, and coordinated, and does it address key resettlement needs and support the individual’s integration into the community?	53%

• There was a strong strategic focus on resettlement work across Wales. The prison group director and regional probation director alternated in chairing the monthly senior leadership forum. This incorporated PDU heads and governors from all the prisons in Wales, as well as the two women’s prions of Styal and Eastwood Park, which served Wales. A number of initiatives had been adopted through this forum, including a guided interview format to replace the basic custody screening tool used in England and a women-specific pathway.
• Much of this was reflected in what we found in our analysis of 107 cases across the six PDUs. In 74 out of 102 applicable cases, resettlement services were delivered in line with resettlement needs, prioritising those that were most critical. In 77 per cent of cases, the community offender manager identified and addressed the key resettlement or desistance needs before release.
• Despite this, the key shortfall in casework we identified was in relation to keeping people safe. In only 52 of 101 relevant cases did the community offender manager identify and address key risk of harm issues before the person was released. In all six PDUs the number of cases sufficiently meeting this standard dropped relative to work focusing on desistance and resettlement needs. This largely reflected what we found in all casework relating to keeping people safe.
• Across the six PDUs, we saw considerable variation in the quality of work to support resettlement. In Cardiff and the Vale of Glamorgan, 11 out of 14 cases reviewed met our standard for resettlement, whereas in Swansea Neath Port Talbot it was only five out of 16.
• The introduction in 2021 of short-term sentence teams in Wales had, at the last inspection of Swansea Neath Port Talbot and Gwent (also in 2021), been seen to be effective in supporting individuals returning to the community. Across Wales, short-term sentence teams had been disbanded in some PDUs, including Swansea Neath Port Talbot and Gwent. However, in Cardiff and the Vale of Glamorgan they had been maintained. In the three best-performing PDUs in relation to resettlement (North Wales, Cardiff and the Vale of Glamorgan, and Gwent), the tasking function that had been part of the short-term sentence teams had been maintained.
• Broadly speaking, the cases we reviewed on short sentences were less likely to meet our overall standard of sufficiency. Swansea Neath Port Talbot had a higher proportion of such cases.
• Overall, although across the region prison offender manager to community offender manager handovers were completed in a reasonably high number of cases (84 per cent as of May 2025), this did not always reflect the quality of some of these reviews. While reports from the probation planning, assurance and implementation team and WAT may identify some issues as part of the overall evaluation of PDUs, it was hoped by leaders that the role of probation planning, assurance and implementation team would be extended to support resettlement work more specifically.

---

R 2.6 Statutory victim work	Rating
Relevant and timely information is provided to victims of a serious offence, and they are given the opportunity to contribute their views at key points in the sentence.	Good

Our rating[6] for victim work is based on the percentage of cases we inspected being judged satisfactory against three key questions and is driven by the lowest score:

Key question	Percentage ‘Yes’
Does initial contact with victims encourage engagement with the victim contact scheme and provide information about sources of support?	92%
Is there effective information and communication exchange to support the safety of victims?	75%
Does pre-release contact with victims allow them to make appropriate contributions to the conditions of release?	81%

• Statutory victim work relates to the region’s activity in relation to the victim contact scheme. This scheme provides victims with the right to support from the Probation Service for the duration of the sentence, where the perpetrator of a violent or sexual offence is sentenced to 12 months or more in prison.
• In 2023 Wales launched the Wales Victim Unit. In January 2024, this incorporated the victim notification scheme, which offers victims of stalking and harassment offences where sentences are under 12 months the ability to make representation for licence conditions and information about the sentence. This was a positive initiative to focus appropriate attention and resources on work with victims.
• Two senior probation officers provided line management to all 21 VLOs across the region. Staff had three weekly check-ins with managers. There were weekly tasking meetings, monthly team meetings and monthly protected learning sessions. Overall, the unit was managing 3,620 cases at the end of September 2025. The regional lead for victim work attended the Wales Criminal Justice Board to provide assurance about the effective functioning of the system.
• In nearly all the cases we inspected, victims had received both a letter and leaflet explaining the nature of the scheme. In all cases, where no response was initially received, the VLO followed this up, usually by email or a telephone call. In many cases this increased the likelihood of engagement. We also saw examples of perseverance on the part of VLOs when no response was initially forthcoming.
• We were encouraged to note that in the three cases where it was appropriate, the victim was supported in making a personal statement to the parole board. In all relevant cases, no-contact conditions were used in licence conditions.
• The regional lead for victim work also met monthly with the heads of PDUs to pick up any issues relating to work with probation practitioners. We saw this in 12 of 16 relevant cases. Nevertheless, this was an area of work that required some further attention and vigilance, especially in how it applied to work with those subject to MAPPA.

---

Learning from Serious Further Offences (Back to top)

• The serious further offence (SFO) team functioned under the overall management of the regional head of public protection. There was a deputy head of community public protection, along with three quality and scrutiny managers (QSMs, known in England as reviewing managers). The team also included a business manager and a single point of contact administrator. Although the target for the team was to have three QSMs, two further QSMs had been recruited and joined the team during our inspection. This reflected the demands of the team. The service planned to have one specialist QSM linked to the offender personality disorder team to undertake SFO reviews through a psychologically informed perspective.
• At the time of our inspection there was a backlog of 15 SFO reviews that had not yet begun, and a further seven that had not yet been submitted. Data further reflected a gradual increase in notifications to the team over the last five years and the increase in staff was a proportionate response to this.
• Of the 11 SFO quality assurance reviews completed during the year 13 May 2024 to 12 May 2025, 73 per cent met the required standard (‘Outstanding’ or ‘Good’), although it was noted that some action plans were lengthy.
• Key deficits identified by SFO reviews were that risk factors had been underestimated, the risk assessment did not consider all known relevant factors, and there was insufficient management oversight, and a lack of professional curiosity. These were similar to our findings across our case reviews on this inspection.
• Key factors underpinning the work of the SFO team were the Human Factors principles (which applied across the whole of the Wales Probation Service) and, for SFO work in particular, fishbone analysis (sometimes known as a cause-and-effect model). This analysis was carried out six weeks after notification. It had been a feature of work in Wales since May 2024, although it was not a feature of full SFO reviews until October 2024. The benefits of fishbone analysis were felt by the whole of the SFO team, particularly the learning in relation to obtaining a more detailed and informative look at the systemic and procedural factors that were evident in the case.
• From an analysis of the 11 SFO quality assurance documents it was determined that the team was mainly performing well in respect of the standard ‘analysis of practice and learning’. Its work demonstrated effective explorative and analytical approaches, clear messaging where practice fell below the expected standard, and action planning relevant to omissions and/or deficits in practice.
• Learning from SFOs was underpinned by the serious case review board, which reviewed all early look cases and action plans. This was attended by the head of the relevant PDU to consider what actions had taken place from a PDU, individual or regional perspective. The team also produced ‘practice matters’ bulletins, which went to all departments and ‘case chronicles’. These were similar to seven-minute briefings, but more interactive.
• The SFO team managers were involved with sharing the findings of SFO reviews with victims/families, alongside PDU heads, and liaising with the relevant VLO. The SFO team’s involvement with sharing reviews appeared to be a unique and positive approach by Wales. However, at the time of inspection no feedback had been gathered from victims/families about it.

---

---

Further information (Back to top)

Full data from this inspection and further information about the methodology used to conduct this inspection is available in the data annexe.

A glossary of terms used in this report is available on our website.

This inspection was led by HM Inspector Keith McInnis supported by a team of inspectors and colleagues from across the Inspectorate. We would like to thank all those who participated in any way in this inspection. Without their help and cooperation, the inspection would not have been possible.

[1] The rating for the standard is driven by the aggregate scores from PDUs for the key questions, which is placed in a rating band. Full data and further information about inspection methodology is available in the data annexe.

[2] The rating for the standard is driven by the aggregate scores from PDUs for the key questions, which is placed in a rating band, indicated in bold in the table. Full data and further information about inspection methodology is available in the data annexe.

[3] The rating for the standard is driven by the aggregate score from PDU and unpaid work case inspections, which is placed in a rating band, indicated in bold in the table. Full data and further information about inspection methodology is available in the data annexe.

[4] The rating for the standard is driven by the lowest score on each of the key questions, from unpaid work cases inspected during regional fieldwork, which is placed in a rating band, indicated in bold in the table. Full data and further information about inspection methodology is available in the data annexe.

[5] The rating for the standard is driven by the aggregate data from resettlement cases in PDU inspections, giving a score for the key question, which is placed in a rating band, indicated in bold in the table. Full data and further information about inspection methodology is available in the data annexe.

[6] The rating for the standard is driven by the lowest score on each of the key questions, from eligible cases inspected as part of regional fieldwork, which is placed in a rating band, indicated in bold in the table. Full data and further information about inspection methodology is available in the data annexe.

---