Dynamic Inspection of Public Protection in Wales
Chief Inspector’s judgement (Back to top)
Our inspection of public protection work in Wales found a region with clear ambition, strong leadership, and a coherent vision for improvement. In the relatively short period since our last inspection, leaders had acted with purpose and pace, introducing a wide-ranging programme of change designed to strengthen the quality of work to keep people safe. There were early signs that this was beginning to have a positive impact.
We saw notable improvement in planning, which was a clear strength. Practitioners were producing more focused and structured risk management plans, supported by targeted initiatives such as ‘Start Right, End Right’ and improvements to systems and guidance. Access to information, particularly from police partners, had improved, although there was still work to be done.
However, these improvements were not yet translating consistently into effective practice. Assessment remained too variable and, in too many cases, practitioners were not analysing information with sufficient depth or using it to inform sound risk management decisions. This weakness undermined the quality of work across the sentence. Too often, opportunities to identify patterns of behaviour, and dispersed victims, were missed.
Despite a significant volume of improvement activity, most changes were still in their early stages. Leaders had been proactive in introducing new frameworks, governance arrangements, and learning opportunities, but these were not yet embedded and needed to translate more consistently into frontline delivery.
Leadership was visible, collaborative, and focused on improvement, with a strong culture of reflection, feedback, and learning. Staff understood the direction of travel and spoke positively about being engaged in change, rather than subject to it. However, many practitioners described feeling overwhelmed, and this was affecting their ability to manage complex work consistently. While staffing levels were relatively stable compared with some regions, they were not yet sufficient to mitigate these pressures fully.
Strategic relationships, particularly with the police and children’s services, were well developed and improving. However, operational delivery was less consistent, with ongoing variability in the quality and timeliness of information sharing, and limited confidence among some practitioners in challenging partners constructively. Further work was required to ensure that strong strategic intent translated into effective information sharing and joint practice.
Overall, while Wales was moving in the right direction, the quality of public protection work was not yet consistently good enough. Sufficient time and emphasis were now needed to embed changes, strengthen analytical practice, and improve information sharing. A sustained focus on these areas would be essential if the region was to realise its ambition of delivering consistently high-quality work to keep people safe.
Context (Back to top)
The inspection looked at the public protection work delivered by the Wales probation region, with a focus on the work to keep people safe. We inspected 55 cases, covering assessment, planning, implementation and delivery, and reviewing, focusing on public protection activity. At the time of the inspection announcement, Wales was supervising 10,686 people, either in the community or custody.
At the time of the inspection, the Wales probation region was complex, comprising six probation delivery units (PDUs): Cwm Taf Morgannwg, Dyfed Powys, Swansea Neath Port Talbot, Cardiff and the Vale of Glamorgan, Gwent, and North Wales. It included four approved premises (APs) and six prisons, operating across 22 local authorities and four police forces. The geography was diverse, with significant rural communities alongside more densely populated urban areas. Complexity was further shaped by the devolved context of Wales, where responsibility was shared between HM Prison and Probation Service (HMPPS) and the Welsh Government. This created both opportunities and challenges for partnership working, particularly as key services such as health, housing, and social care were devolved.
Wales also had distinct socioeconomic and cultural characteristics, including areas affected by post-industrial decline, and variable infrastructure, especially in more rural regions, which influenced service delivery and outcomes. Wales also benefited from strong community identity and some well-established partnerships, which supported collaborative approaches to service delivery.
Organisational data highlighted relative stability in staffing, although teams were not working at full capacity, with 89 per cent of probation officer grades and 84 per cent of senior probation officers (SPOs) in post. Wales had experienced further pressure in implementing the national policy changes designed to address workloads across probation and prison services. These included Probation Reset, prison early release schemes, Impact, and ongoing change due to the Independent Sentencing Review.
The Wales region was last inspected in 2025, with the report published on 21 January 2026. At that time, inspectors found that probation services in Wales had strong leadership, some positive partnership working, and a clear commitment to improvement, with well-developed relationships across statutory and third-sector partners. However, the quality of work to keep people safe was not consistently good enough, with ongoing weaknesses in the sharing and use of information to assess and manage risk. Inspectors found that information-sharing arrangements with the police and local authorities were also complex and inefficient. Four regional recommendations were made.
Inspection commentary (Back to top)
This inspection in Wales found evidence of improvement activity across key areas, with some early signs of impact, but longstanding issues in assessment and analysis continued to limit the consistency and effectiveness of practice.
Leaders had responded to their own assurance activity by introducing initiatives such as information-gathering workshops and the ‘Start Right, End Right’1 approach, which were beginning to strengthen the quality and availability of information. However, assessment remained variable and leaders acknowledged that practitioners should prioritise using a methodical approach to obtaining high-quality information and then move the focus to analysis. While most cases demonstrated an appropriate overall risk classification, there were persistent weaknesses in analytical depth. Domestic abuse histories were not always explored fully, particularly where there were aggravating complexities such as coercion and control, and practitioners did not consistently link behaviour to the impact on children or recognise patterns across cases. This was underpinned by challenges in developing core analytical skills. Inspectors saw the same issues with practitioners not always acting on real-time ‘reportable incidents’ from the police, along with some complacency about how proactively they needed to seek information as a result. Encouragingly, this was on a smaller scale than we found at the previous inspection and there were signs of improvement in access to information overall. However, information received was not sufficiently embedded to inform the safeguarding of victims.
Planning had improved significantly and was a clear strength of practice, reflecting sustained and targeted activity which had started before the previous inspection. Changes to the CORRE2 function, and revised offender assessment system (OASys) briefings had supported practitioners to produce more focused and strategic risk management plans. However, this improvement in planning was not fully matched by implementation and review. There remained gaps in coordinating activity with other agencies, integrating plans, and the use of review processes to reassess risk dynamically. While developments such as the evolving CORRE model and improvements in multi-agency public protection arrangements (MAPPA) Level 1 processes provided a strong foundation, further work was needed to ensure that effective plans translated into consistently robust delivery and multi-agency risk management.
A significant volume of improvement activity had been undertaken since the last inspection; however, given the short timeframe, this had understandably resulted in only early and incremental improvements in practice to date.
Since the previous inspection, regional leaders had implemented a structured programme of improvement focused on strengthening public protection. This included the establishment of the Start Right, End Right approach, with dedicated guidance and implementation boards, alongside targeted court, custody, and community interventions to ensure that risks were addressed early. Work had also been completed in the tightening of domestic abuse, safeguarding, and address-checking policies, supported by enhanced police and local authority information-sharing processes, new workshops, and a stronger emphasis on using intelligence effectively within risk assessments and management plans. Regionally, governance and assurance were strengthened through a more coordinated community public protection framework. These developments were consolidated within the Wales public protection strategy 2026–2029, which embedded consistent standards for risk assessment, MAPPA practice, safeguarding, and data-led oversight, with a clear focus on defensible decision-making, professional curiosity, and continuous learning.
Leadership across the region was energising, with clear intent and early evidence of impact. The leadership style we observed was distinctive, collaborative, and clearly embedded, with evidence of it being adopted at multiple levels of the organisation. Leaders did not defer to systemic pressures such as understaffing or the pace of change, taking clear responsibility for delivery and improvement.
Cohesion was a hallmark found throughout our fieldwork, underpinned by a commitment to operating as a single, integrated service. This was evident across courts, prisons, APs, and wider partners, as well as the well-established joint prison–probation model supported by shared executive forums. The same strength was seen in the approach to women’s services, where provision was coordinated from arrest through to post-sentence, reflecting a genuine end-to-end model.
The organisational culture was intentional and enabling. In managing ongoing change in probation services, the region had reflected and sharpened its approach, introducing structured mechanisms such as operational readiness boards and change checklists, which supported staff to anticipate and engage with change. However, sustained change led to fatigue, with findings from our staff survey and interviews indicating that many staff felt overwhelmed, affecting their ability to deliver high-quality work consistently.
A strong culture of reflection and feedback was evident, reinforced by the region’s embedded use of a human factors approach, including daily reflective discussions on decision-making, consideration of cognitive bias, and learning from Serious Further Offence (SFO) reviews to strengthen psychological safety and defensible decision-making in public protection work.
Direct approaches were being adopted to strengthen the consolidation of training – an issue we have identified as a barrier across many regions. Protected learning days were held monthly across all PDUs. In Gwent, the PDU was trialling a three-month cycle on one topic, with the final session focusing on consolidation through ‘Dysgu Cymru’ action learning sets. Alongside this, a strong culture of informal learning was evident and appeared to enhance the overall effectiveness of the formal training offer through consolidation. For example, the use of quality development officers (QDOs) aligned to PDUs had involved taking a bold step away from a time-intensive, case-by-case audit approach. QDO work focused on embedding learning and quality practice and enabling staff through interactive sessions and action learning sets.
This work was informed by the breadth and ambition of audit and assurance activity in the region, along with learning from SFO reviews. While the level of scrutiny was impressive, leaders were keen that it extended beyond oversight. The probation planning, assurance, and implementation team was driving improvement through using insights from internal and external sources of feedback and audit. The implementation of systems, messages, briefings, and effective practice priorities supported operations to improve. This had created a clear and coherent link between learning, assurance, and action. Staff were made aware of the benefits of this work – ‘the why’, which we heard articulated consistently.
Although staffing levels were relatively stable overall, workloads continued to be affected adversely by vacancies. This position was further compounded by ongoing challenges with recruitment being managed centrally and the well-documented delays in vetting processes, which meant that some newly recruited staff could not wait for the vetting to complete, and sought other employment opportunities.
Sickness levels were impacting teams, with regional ‘hot spots’ already well understood by leaders, who identified long-term sickness as underpinning overall sickness data. Staff in the region complied with the national directive of spending a minimum of 60 per cent of their working time in the office. However, regional leaders had recognised that operational demands required sufficient presence at PDU sites, necessitating increased probation practitioner attendance. Taking into account bespoke reasonable adjustments, PDUs were currently operating at 80 per cent staff presence, with practitioners needing line management agreement to work away from the office for specific tasks, including report writing and assessment activity.
This approach was intended to promote more cohesion in teams, increase opportunities for informal learning, and mitigate burnout linked to intensive engagement with people on probation within condensed timeframes. Notably, in Cwm Taf Morgannwg, leaders reported some early improvement in sickness levels, which leaders felt might have been linked to this decision. Although the initial staff response was mixed, regional leaders maintained a consistent direction, bringing staff with them through emphasising ‘the why’, which included opportunities for informal, naturally occurring learning from peers in everyday practice. This message was replicated in induction processes and reflective practice sessions, to ensure that staff understood the importance of office-based practice. Over time, staff began to acknowledge the benefits of the model. Our case inspections confirmed that the vast majority of cases we inspected involved the person on probation being seen at the appropriate frequency and that the nature of the appointment was also appropriate. This was evident in more face-to-face reporting and fewer appointments being covered by colleagues.
The quality of probation services officer (PSO) work had been identified by the region as an area for focus through its internal audit activity. Our case sample reflected similar themes, including gaps in work involving cases with multiple or dispersed potential victims and in the assessment of unconvicted behaviours. We were encouraged by emerging innovative practice, including the piloting of PSO progression panels in Dyfed Powys, which combined a focus on accountability with professional development. Cwm Taf Morgannwg had piloted specific training and peer-to-peer reflective learning sessions for PSOs. At a strategic level, there was a clear and consistent emphasis on PSO practice within accountability meetings.
Access to information, particularly from police partners, had improved since the last inspection. In a substantial majority of cases inspected, sufficient information had been obtained from police to support an assessment of the risks associated with domestic abuse. There was evidence of more strategic-level engagement with the police through joint meetings and training, along with a shared recognition of the challenges. However, variability remained across forces, often influenced by differences in professional background and interpretation of data protection requirements. Police partners were positive about working relationships and described the increase in joint events as transformative. While there were encouraging signs of progress, including police partners’ openness to reducing the amount of redaction in the information they share, in some areas further work was needed to ensure that services were equipped sufficiently to manage the volume and sensitivity of the information being shared.
Staff survey findings indicated that a small but notable proportion of staff did not feel confident in following up or constructively challenging partners. Leaders had taken appropriate steps to address this. Tools such as SBAR (situation, background, assessment, recommendation) had become embedded in routine practice, and the use of case administrator to SPO support to schedule reflective supervision demonstrated a clear organisational commitment to strengthening professional curiosity and confidence. Despite this, the ability to challenge partners appropriately remained one of the key factors limiting the consistency of quality in practice. The actions put in place were credible and well considered; however, given the relatively short period since the last inspection, it was too early to assess fully the impact of these significant changes.
Since the last inspection, regional leaders had identified key priorities within their 2026–2029 public protection strategy. This included standardisation of enquiry and response templates, and the completion of a mapping exercise. This sought to understand the variance in quality and timeliness of local authority responses across the region. Each PDU formed a working group with a focus on identifying and escalating barriers more swiftly, with embedded assurance.
Relationships with children’s services were less developed than those with police, although there was clear commitment on all sides. Strategic partnerships were well regarded, with probation practitioners seen as engaged and collaborative in audits, case reviews, and governance boards. However, barriers were more evident at an operational level, including gaps in the mutual understanding of roles, responsibilities, and assessment processes. While both sides recognised these challenges and took shared responsibility, further work was needed to translate strong strategic relationships into consistent and effective operational practice. Positively, arrangements for probation practitioner involvement in strategy meetings were well established throughout Wales, which meant that information exchange took place at the earliest opportunity.
MAPPA arrangements were largely working effectively. Coordinators, along with the community head of public protection, were all aligned in their approach to thresholds and level setting. Reputation and credibility achieved through longstanding relationships meant that partners were engaged without issue, and this area of work had also been strengthened by the learning culture and commitment to continuous improvement. Most cases managed at Level 1 had sufficient reviewing, which was aided by coordinators’ involvement in reviews. The key area for development was highlighted through case inspection data, which suggested that MAPPA was not always enhancing multi-agency work. This related to cases managed at Level 1 and the same issues with redacted police intelligence that were seen in the core caseload.
There were robust arrangements to manage serious and organised crime. Effective use was made of regional organised crime unit intelligence and joint intelligence programme (JIP) processes, with leaders demonstrating a clear understanding of case volumes and identified hotspots. Our case inspections reflected this, with JIP involvement adding clear value to risk management and oversight in the cases reviewed.
Community sentence treatment requirements and associated contract management arrangements were a strength. The use of dual requirements to address co-occurring mental health and substance misuse needs was particularly positive. In cases where both needs were present, their intersection compounded complexity and risk. The dual‑order approach provided a flexible response that addressed limitations in single‑pathway provision. 101 people on probation were sentenced to these dual orders over a 12-month period, demonstrating the need and impact of this work. There was also evidence of services adapting to changing patterns of need, including shifts in drug use and the commissioning of culturally specific provision to address under-engagement from Black, Asian, and minority ethnic groups.
Case inspections involving people on probation who had been released from custody were mostly managed well. Many of the identified themes with community cases, such as poor analysis and inter-agency working, were less a feature in custody cases. ‘Start Right End Right’ for custody cases had driven earlier engagement, joint working between custody and community, and clearer accountability – supported by an increased MAPPA focus and higher levels of oversight for cases prior to release. Strong regional relationships, particularly with APs and accommodation teams, enabled effective communication and allocation of bed spaces. The use of AP training and pre-release keyworkers improved preparation and stability, while integrated accommodation pathways (including tier 3 community accommodation services (CAS3) and local authority partnerships) ensured a holistic response. Amid wider concern about disproportionate recall and ‘stuck’ indeterminate sentence for public protection (IPP) prisoners, the region had reduced recalls by around a third, helping to mitigate unnecessary returns to custody and support progression. This was achieved by staff briefings, increased scrutiny and enhanced recall discussion.
Regional recommendations (Back to top)
1. Embed and evaluate existing improvement activity to achieve consistent practice.
2. Develop practitioners’ confidence and skills in the use of professional curiosity and challenging conversations to identify, analyse, assess, plan, and respond to indicators of risk effectively.
3. Improve the consistency of multi-agency work at an operational level to manage domestic abuse and child safeguarding, such as with the police and children’s social care services, to ensure that actual and potential victims are sufficiently protected.
4. Implement arrangements for monitoring and improving the quality of sentence management work delivered by practitioners to keep people safe.
HMPPS recommendations (Back to top)
5. Develop a national strategic approach to information sharing with the police and children’s services, to support regions in achieving consistency and compliance with legislation to obtain and use information to protect the public.
6. Reduce vetting delays and address workforce instability by implementing streamlined recruitment processes.
Scoring (Back to top)
| Key question | Percentage ‘Yes’ |
| Does assessment focus sufficiently on keeping other people safe? | 40% |
| Does planning focus sufficiently on keeping other people safe? | 64% |
| Does the implementation and delivery of services effectively support the safety of other people? | 44% |
| Does reviewing focus sufficiently on keeping other people safe? | 47% |
Follow-up activity (Back to top)
In line with the recommendations identified, a range of follow-up activity will take place. HM Inspectorate of Probation will work with the region to identify what can be done to guide and support their work, increase knowledge and confidence, and provide a solid foundation for further improvement. The Inspectorate will also seek to share what effective practice looks like by drawing upon inspection findings, identify blockers to progress, and highlight opportunities to improve accountability.
Key contextual facts (Back to top)
| Number of people supervised (on 31 December 2025)3 | 15,144 |
| MAPPA-eligible offenders (on 31 March 2025)4 | 6,110 |
| Victim satisfaction performance SL021 (April 2024 – March 2025)5 | 79.4% |
| Staffing level (staff in post full-time-equivalent (FTE))6 | ||
| Senior probation officer (SPO) | Probation officer (PO) | Probation services officer (incl. Professional Qualification in Probation (PQiP)) |
| 84% | 89% | 110% |
| Average caseload at the point of inspection (FTE) | ||
| PO | SPO (exc. PQiP) | PQiP |
| 31.61 | 41.14 | 26.67 |
| Recall rates (in the 12 months prior to inspection) | 25.71% |
| Average rehabilitation activity requirement (RAR) wait time (in the 12 months prior to inspection) | Not available |
| Percentage of RAR days completed (in the 12 months prior to inspection) | 68% |
| Percentage of accredited programme requirements completed for individuals convicted of a sexual offence (in the 12 months prior to inspection) | 61.19% |
| Percentage of accredited programme requirements completed for individuals not convicted of a sexual offence (in the 12 months prior to inspection) | 35.54% |
| Risk of Serious Harm classification of inspected cases | ||
| Low | Medium | High/very high |
| 2% | 55% | 44% |
Further information (Back to top)
This inspection was led by HM Inspector Donna Waters, supported by a team of inspectors and colleagues from across the Inspectorate. We would like to thank all those who helped plan and took part in the inspection; without their help and cooperation, the inspection would not have been possible.
- In response to Probation Reset, the Wales Probation Service had embarked on a journey to ensure that people on probation received the right interventions earlier. Start Right, End Right was created and launched in September 2024. To ensure that they were getting things right first time, the initiative saw the creation of guidance manuals for court, custody, community, and unpaid work teams, and the creation of ‘final third’ teams across PDUs where Probation Reset and Impact cases were managed. ↩︎
- Dedicated CORRE (Centralised Operational, Resettlement, Referral and Evaluation) teams are responsible for completing referrals and coordinating key elements of sentence plans. The team ensures that referrals to services and interventions are appropriate, timely, and aligned to the individual’s assessed circumstances, risks, and rehabilitative needs. ↩︎
- https://www.gov.uk/government/collections/offender-management-statistics-quarterly. ↩︎
- https://www.gov.uk/government/collections/multi-agency-public-protection-arrangements-mappa-annual-reports. ↩︎
- https://www.gov.uk/government/statistics/community-performance-annual-update-to-march-2025. ↩︎
- Workforce data included in this report comes from internal management information and some of this data has been derived from a different data source to the published HMPPS Workforce Statistics bulletin and accompanying Probation Officer Recruitment Annex. The Inspectorate needs access to the latest data available, and internal management information is deemed the best source to allow this. As such, there could be discrepancies between the data in this report and the data contained in the publication. ↩︎