Dear Glyn


We have recently conducted our review of the compliance of Welsh Government’s (WG) Rough Sleeper Count official statistics against the Code of Practice for Statistics.

While these statistics are not National Statistics, they are important official statistics valued by users and so we have made a number of recommendations to support your continued development of these statistics. We considered the Trustworthiness, Quality and Value of these statistics in relation to the Code and have appreciated the positive and constructive way that the team has engaged with us during our review, especially at this particularly challenging time.

We welcome the news that new management information is being collected on the numbers of homeless and rough sleepers in Wales being assisted into emergency accommodation since the start of the COVID-19 pandemic. We would like to recognise the positive steps being taken by your statisticians, working with other Welsh Government officials, to determine how this management information can be best used to complement the existing statistics. The plans for how this may develop over the coming months demonstrates real innovation within this area. We commend your team on the timeliness and transparency shown through the recent release of this management information which provides further insight into the homeless landscape, and we feel that this sets a good example of the processes needed to release data that is used publicly in ministerial statements accessibly and promptly.

Within the National Rough Sleeper count we found a range of positive features that demonstrate the trustworthiness, quality and value of the statistics:

  • Upfront guidance about the limitations of the single-night count and providing users with a range of factors that can influence the accuracy of the information and the steps taken to try to reduce the impact of the issues;
  • Having strong internal quality assurance processes within the team, some of which are documented within the release, gaining insights from policy colleagues during the quality assurance process, as well as having built-in validation checks within the data collection form itself;
  • Providing clarity and insight through the use of some good examples in the release of maps, tables and charts to present the data as well as providing some context around the data;
  • The team’s active engagement with topic experts to review and further develop the methods for measuring the rough sleeping population in Wales.

We identified some areas for improvement that would enhance the quality and value of the statistics:

  • Having greater oversight of the data collection methods used and information collected by Local Authorities (LAs), as well as seeking further assurances around how LAs ensure the accuracy of their data returns, will help enhance the understanding of the comparability of the data between the different areas and over time;
  • It is unclear what level of assurance is given on the quality of the data sources. To assure users of the level of quality of the data, information should be provided on this and how they were assessed. This will also help to further demonstrate your application of our required standards for the quality assurance of administrative data;
  • As the statistics are developed further, we welcome your ambitions to explore collecting case-level data and recommend exploring the collection of demographic characteristics of the rough sleeper population to add to the value for users and those working in WG policy areas. We are aware that the Welsh Local Government Association (WLGA) publishes its own analysis of the information collected during the two-week national rough sleeper count and so we would suggest considering building links with organisations such as WLGA with an interest in this area, to help deliver better insights into the rough sleeper landscape;
  • We recommend engaging with the Government Statistical Service (GSS) Good Practice Team to explore the potential for further improvements to these statistics, particularly in light of recent improvements that the team supported Ministry of Housing, Communities and Local Government’s (MHCLG) analysts make to their Rough Sleeping Snapshot in England statistics, including innovative new ways of releasing these data.

Thank you for engaging effectively with us during this review. We welcome the updates on your continued collaborative work as part of the Cross-Government Homelessness Statistics Working Group. With management information now being collected across the UK on the numbers of rough sleepers and homeless helped into emergency accommodation since the start of the pandemic, alongside other administrative sources, we look forward to seeing GSS statisticians work together to more fully illustrate the complexity of the overall UK rough sleeping and homelessness picture.

Our Housing, Planning and Local Services Domain Lead will continue to engage with your team on progress in the coming months and we would welcome a progress update from you upon the next publication of these statistics.

Please do not hesitate to get in touch if you would like to discuss any aspects of this letter further.

I am copying this letter to Sue Leake (Head of Education and Public Services Statistics); Luned Jones from the Housing Statistics team; and Lee Thomas from the Data Collection team.


Yours sincerely

Mark Pont

Assessment Programme Lead