Changes to our inspections of probation services in 2019-2020

In 2018, HM Inspectorate of Probation introduced new inspection methodology for youth offending and probation services. This was coupled with new standards and ratings, which were designed to drive improvement.

Ahead of the 2019-2020 inspection cycle, we have reviewed and made changes to parts of our methodology and standards for inspections of probation services. We considered a number of factors including: learning from inspections that took place in 2018-2019, responses to our four-week consultation, and the evidence base for high-quality probation services.

We are not making any changes to our ratings. We will continue to rate individual aspects of work as well as the overall quality of services using our four-point scale: ‘Outstanding’, ‘Good’, ‘Requires improvement’ and ‘Inadequate’.

Key changes to our methodology for the 2019-2020 cycle include:

  • a clearer link between our judgments on leadership and organisational delivery (domain one) and the data from our detailed review of individual case supervision in domains two and three
  • a stronger voice for those being supervised by probation services in our methodology. As a first step we will be interviewing the service users involved in the cases we select for domain two to hear from them direct about their experience under supervision. And we will meet with service user fora or their equivalent in every area we inspect
  • improving the evidence we collect on unpaid work delivery, including site visits and interviews with supervisors
  • additional questions on diversity and inclusion in our domain one methodology and a greater focus on how we report this data.

Changes to our standards in detail

Domain one – Leadership

We have found a disconnect in some inspections between our ratings in domain one (organisational delivery) and domains two and three (which look at the quality of casework). This was particularly pronounced in CRC inspections – some providers achieved ‘Good’ or ‘Outstanding’ ratings in domain one, despite ‘Inadequate’ ratings in domains two and three.

When we developed our methodology for the 2018-2019 inspection cycle, we made the decision to treat each domain separately as this would help us to identify and describe where problems lay. In practice, this has limited our ability to evaluate the effectiveness of management on the quality of work with people under probation supervision.

To address this issue, we have revised our domain one standards to create a clearer link between management inputs and delivery outputs. We have not changed the standards themselves, but we have changed some of the questions and prompts to guide our inspectors. Data from case inspections in domains two and three may be used to influence judgements in domain one.

We have added three new prompts to domain one covering: diversity of the service’s workforce; staff access to promotions and reward and recognition schemes; the way in which disproportionality is addressed in the delivery of services. When we assess the learning culture of services, we will now specifically ask about learning from Serious Further Offence Reviews.

In 2018-2019, we found some of our standards did not differentiate sufficiently between actions and decisions taken at a local level and those taken at a higher level by CRC owners or national NPS directives. We found that CRC providers or NPS divisions were not always able to influence operating models or staffing levels – both of which can have a major impact on the quality of the service’s work. We are currently undertaking a national inspection of the NPS and a review of CRC ownership influences to enable us to better understand and comment on these dynamics.

Domain two – case supervision

Our standards are based on the well-established and recognised ASPIRE model for case supervision (Assessment, Sentence Planning, Implementation, Review and Evaluation). We are not making any changes to this domain.

Domain three – NPS and CRC-specific work

For inspections of NPS divisions, we look at the court work and case allocation, and statutory victim work.

Court work and case allocation – we have made one change at prompt level. When we ask if appropriate checks are being made, we will now specifically ask about safeguarding and domestic abuse checks.

Statutory victim work – we have not changed the standard; the focus is still on capturing victims’ views and the provision of timely information to victims. However, we have made changes to some of the questions and prompts to guide our inspectors. These changes draw on the Victims’ Commissioner’s rapid evidence assessment (2016) and include looking at: information sharing and ongoing communication with victims, procedural justice, multi-agency working and the professionalisation of victims’ services.

For inspections of CRCs, we look at the delivery of unpaid work and Through the Gate services.

Unpaid work – we have not changed the standard but we have changed some of the questions and prompts to guide our inspectors. We have created a better link between the assessment of the individual and the delivery of unpaid work.

Research shows that it is important for unpaid work to have a rehabilitative element, to focus on desistance from further offending, and to join up with probation supervision. Pro-social modelling by supervisors and the types of opportunities available are also important. We will be taking these factors into account in 2019-2020, to bring our inspections closer to the evidence base.

Through the Gate – we have not made any changes to this standard. However, there are two developments in 2019-2020 which could have an impact on future inspections: the embedding of the revised Through the Gate specification and the full rollout of the Offender Management in Custody (OMiC) model. We will keep this standard under review.

Ratings for domain two and three work will be based solely on data from inspected cases (and for unpaid work, qualitative information from observations and outcome information).

Changes to our methodology in detail

Inspection order

As detailed in the section on changes to domain one standards, there was disconnect in some inspections between our findings in domain one (organisational delivery) and domains two and three (which look at the quality of case work).

In 2018-2019, we inspected domains two and three during the first two fieldwork weeks, and domain one throughout the three fieldwork weeks. Evidence from each domain was treated separately.

In 2019-2020, we are changing our methodology to create a clearer link between all three domains. We will still inspect domains two and three over the first two fieldwork weeks, alongside initial inspection of domain one. That will be followed by a week off-site for data analysis and planning, then a further week of fieldwork focusing on finalising the gathering of evidence for domain one. These changes will give inspectors an opportunity to look at a sample of casework first, and to analyse and reflect on findings before finalising inspection of organisational delivery.

Case sampling

We have made some changes to the way we sample cases. There are minor changes to domain two and unpaid work cases, and major changes to Through the Gate cases. Full details are in the inspection manual for 2019-2020.

Service users

In 2018-2019, we asked staff at inspected services to explain how they engaged service users in the design and review of their services and we will continue to do this. In 2019-2020, we will be reintroducing interviews with service users in inspected domain two case so we can hear about their experiences directly. We will also meet with service user fora or the equivalent where one exists.

Domain one

Prior to fieldwork, the inspected service will be asked to submit three to five pieces of evidence for each key question in domain one. We will be asking for the organisation’s best evidence, rather than all evidence, along with the option to provide 300 words of narrative to explain their current position.

Domain two

We are not making any changes to the way we inspect against domain two standards.

Domain three

We have revised our methodology to reflect the changes to the inspection of unpaid work. As well as inspecting cases, we will be introducing a more qualitative approach including: interviews with supervisors, site visits, observation of inductions, interviews with workers, assessments of additional evidence in advance, project assessments, and analysis of training and employment opportunities.

Changes to the way we report on our recommendations

In 2018-2019, we made recommendations in each inspection report. In 2019-2020, we intend to assess whether inspected services have followed up on our recommendations and whether there has been sufficient progress, insufficient progress or no progress at all.

When we assess progress, we will be measuring the result of the activity taken to address the recommendation, as opposed to simply the activity itself; we are recognising achievement rather than effort.

Our recommendations are made at prompt, question or standard level – so even if sufficient progress has been made against a specific recommendation, this will not always lead to an impact on the rating at standard level.

We will continue to assess how well an organisation responds to audit, inspection and monitoring in the broad sense as part of domain one (standard 1.4.4).

Changes to the inspection timetable

In 2018-2019, we introduced an annual inspection for all 21 CRCs and seven NPS divisions. The next two years – 2019-2020 and 2020-2021 – will see significant changes to probation services, with the transition to 12 probation areas across England and Wales.

Our inspection cycle for 2019-2020 will take more than 12 months. Our first inspection was announced in July 2019 and we plan to finish in December 2020.

Visit our standards and ratings page for further information and to download the documents.

Our case assessment rules and guidance and inspection manuals are available on our probation inspection documentation page.