Dear Alastair 

National Statistics designation for the Scottish prison population statistics

We have reviewed the actions that your team has taken to address the requirements in our Assessment of compliance with the Code of Practice for Statistics: Scottish prison population statistics. I would like to thank you and your team’s engagement with us during the assessment, demonstrating Scottish Government’s commitment to developing and improving the statistics for the public good. 

On behalf of the Board of the UK Statistics Authority, I am pleased to confirm the designation of National Statistics for the Scottish prison population statistics. These statistics provide granular, accessible, and insightful information about the Scottish prison population. 

Your team responded positively and quickly to implement the requirements and recommendations of our assessment. We particularly welcome the joint working group your team has set up with the Scottish Prison Service, which should lead to be more coherent statistics on the Scottish prison population, and the further quality assurance work that is currently underway. More detail about our judgement is included in the Annex.  

We regard National Statistics as meeting the highest standards of trustworthiness, quality and value. 

I am copying this letter to Amy Wilson, Head of Justice Analytical Services at Scottish Government, and Rose Munenura, Head of Justice Systems Analysis Unit at Scottish Government.   

Yours sincerely,  

Ed Humpherson
Director General for Regulation 

Annex: Review of actions taken in response to Assessment of compliance with the Code of Practice for Statistics: Scottish prison population statistics, produced by Scottish Government 

Requirement 1 

To enhance coherence and insight, and minimise the risk of undermining the official statistics, Scottish Government should: 

a) work collaboratively with the Scottish Prison Service (SPS) to explore if they can produce more joined-up statistics about the prison population in Scotland. 

b) add links to other related statistics, such as homelessness statistics and equivalent prison population statistics for England and Wales, and Northern Ireland. 

Actions taken and commitments made by Scottish Government to meet the requirement 

1a) Scottish Government (SG) and the Scottish Prison Service (SPS) have agreed to work collaboratively to review outputs relating to the prison population to ensure more joined-up statistics are published. A working group has been established and will convene in spring/summer 2023 to commence the review. This work will be supported through a new Memorandum of Understanding (MoU) which has been agreed between SG and SPS on the production of the official statistics. 

In the interim, both SG and SPS have agreed to review and amend the narrative accompanying their main statistical/information releases. These include the SG-produced annual official statistics (next release scheduled for autumn 2023) and the SPS quarterly public information page (next update scheduled for May 2023). The enhanced narrative will clearly outline the purpose and scope of the respective outputs and place them in context to each other and to other data sources (see 1b) below). This is expected to improve clarity for users on the purpose and scope of the respective outputs, as well as helping them identify with ease any differences in methodologies adopted, timelines applied etc to aid interpretation. 

1b) Section 10 of the 2021-22 Scottish Prison Population Statistics (SPPS, released November 2022) provided a breakdown of key related statistical and other publications. This includes descriptions of and links to for example: Recorded Crime in Scotland; Criminal Proceedings in Scotland; the Justice Analytical Services (JAS) Monthly data report; and the Council of Europe Annual Penal Statistics (SPACE-I) collection. SG has committed that future releases will continue to provide this. 

The sources highlighted contextualise the SPPS in the wider justice system and allow comparisons of prison population levels and composition between Scotland and the rest of Europe (including England & Wales (E&W) and Northern Ireland (NI)). Links to current Scottish Prison Service outputs are also provided. 

For the 2022-23 SPPS (release expected autumn 2023), SG plans to signpost users to a broader range of statistical outputs to further contextualise the data, including the Homelessness in Scotland statistics, the Scottish Census and related population estimates, and the Scottish Index of Multiple Deprivation. SG also intends to provide links to the equivalent prison population statistics for E&W and NI in the 2022-23 release and explain their key differences to aid interpretation in relation to the SPPS. 

SG has committed to add further information on the developing outputs from both SG and SPS to future SPPS releases, as the work plan detailed in 1a) progresses. 

OSR’s evaluation of evidence 

We welcome the establishment of the SG/SPS working group and the refreshed MoU. The working group sounds like an excellent forum for closer and more collaborative working with SPS and should lead to more joined-up statistics on the Scottish prison population. 

The enhanced narrative and additional information about other data sources will further enhance the coherence of the statistics. 

Requirement 2

To reassure users about data quality and demonstrate transparency about its QA approach, Scottish Government should: 

a) explain the strengths of the PR2 system and why it is confident in the quality of the data. 

b) review its QA process and publish more-detailed information about data collection, checking and validation. Our Quality Assurance of Administrative Data (QAAD) framework will be helpful for this. 

Actions taken and commitments made by Scottish Government to meet the requirement 

2a) Additional text on the role and importance of the prisoner records system (PR2) within SPS has been added to the Technical Manual for the SPPS. Information on the internal quality controls in place in SPS has also been added. 

2b) Additional text has been added to the Technical Manual to outline the quality assurance work undertaken by the SG during data processing 

Further quality assurance work, in line with the QAAD framework, is currently underway with SPS. This work commenced in April 2023 and SG expects it to be completed by the next statistical release (autumn 2023). SG intends to publish details of the outcomes of this work and a breakdown of the levels of assurance that can be provided in the Technical Manual. 

OSR’s evaluation of evidence 

The additional information about PR2 and the QA process that has been added to the Technical Manual is clear and helpful. It emphasises the operational importance of collecting accurate information on prisoners and briefly explains how SPS ensures data accuracy.  

It is good that Scottish Government is now undertaking a fuller review of the QA process and that this is being doing in collaboration with SPS. This will lead to a better shared understanding of PR2 data quality. Publishing the findings of the review will give users a more detailed account of the QA process and assure them that it is robust and generating high-quality statistics.  

Requirement 3

To help users interpret the statistics, the information about uncertainty should be expanded, by explaining the nature of the prison population estimates and the confidence intervals around the general population estimates. 

Actions taken and commitments made by Scottish Government to meet the requirement 

Text was added to the Introduction section of the 2021-22 SPPS report to provide explicit guidance about the figures being estimates, alongside the explanation of the dataset on which the figures are based: ‘The measurements throughout this publication therefore represent estimates based on the data as constructed’. 

Estimates are compared against other data sources in Section 9 of the 2021-22 SPPS and specific limitations are presented. SG has committed to retaining this additional text in future statistical releases. 

Text has been added to the Comparator Population Rates section of the Technical Manual to provide an explanation of the confidence intervals around the general population estimates and guidance on how to interpret the rates calculated using the estimates. SG plans to add similar text to future statistical releases. To further support users, a link has been provided to the methodology section of the Scottish Surveys Core Questions where the method used to calculate survey confidence intervals is explained. 

OSR’s evaluation of evidence 

The additional guidance about the prison population estimates and the expanded information about confidence intervals around the prison population rates by ethnicity are supporting appropriate use of the statistics. 

Requirement 4

To improve the team’s resilience and ensure the process can be understood and used by multiple team members, the team should prioritise the development of documentation and coding skills. 

Actions taken and commitments made by Scottish Government to meet the requirement 

SG acknowledges that the team delivering the SPPS is small and there are not (currently) multiple team members with capacity to deliver the releases. To aid the team’s resilience it therefore recruited for an additional statistics post to support the delivery of the SPPS in spring 2022. Unfortunately, the successful applicant later rejected the offer shortly before their expected start date. Since then, SG has not been able to recruit, as wider organisational restrictions on recruitment were in place. It is anticipated that these will be eased in the coming year and that recruitment boards for analysts will be re-instated, allowing us to recruit for the vacant post. 

To aid business continuity and swift handovers, the existing team is developing enhanced documentation on the data processing and production of key outputs for the statistical releases. At present, the team works from annotated code and tabulated instructions for data processing. Full desk instructions are being prepared to support the on-boarding of a new team member or temporary resource from other statistics teams within JAS or the wider SG statistics profession in the event of unexpected absences of current team members. SG has committed to having these desk instructions fully drafted ahead of the next statistical release. 

All team members involved in the production of the SPPS have been trained in R and are members of the supportive R Users Community within SG (a forum for sharing training, expertise and troubleshooting). Continuous skills development will be offered to ensure the team keeps up to date with new developments in R. 

OSR’s evaluation of evidence 

We recognise the challenges that the team, and Scottish Government as a whole, has faced in recruiting new analysts in 2022/23. We are therefore pleased to hear that there may be an opportunity to recruit a new team member in the coming year.  

It is good that the team is preparing enhanced documentation and desk instructions for future releases of the statistics, and that there is a contingency solution in place for producing the statistics. This enhanced resilience will ensure that the statistics can continue to be produced.  


Related Links:

Assessment of compliance with the Code of Practice for Statistics: Scottish prison population statistics