Judgement on National Statistics Status
Statistics on Statutory Homelessness in England are published by the Department for Levelling Up, Housing and Communities (DLUHC). DLUHC has published statistics on statutory homelessness since 1998, but they were substantially redeveloped following the introduction of the Homelessness Reduction Act (HRA) in 2017. The legislation has been reflected in an enhanced Homelessness Case Level Information Collection (H-CLIC) data specification for local authorities (LAs) to follow since April 2018, resulting in a detailed case-level data set, on which the statistics are based.
The introduction of the H-CLIC data requirement has considerably enhanced the potential of the statistics to answer key policy questions, through new insights on household flows through the homelessness system. DLUHC statisticians have developed new data dashboards to help communicate key insights, which are directly informing policy initiatives designed to reduce homelessness and its impacts. Plans to link H-CLIC case-level data with administrative and other sources, promise to enhance the value of the statistics further still.
However, the transition to the H-CLIC system has not gone smoothly for many LAs, some of which are still not able to return data against the new specification. DLUHC recently removed the old data submission method (P1E), to encourage all outstanding LAs to make the transition to H-CLIC, though this has so far led to an increase in the number of LAs not submitting data for key measures such as the number of households in temporary accommodation.
Despite these challenges, the statisticians have engaged positively with LAs to support them during the transition. They have developed methods to impute for LA data that are missing, or deemed to be of insufficient quality to publish, and a new dashboard to help communicate the quality of published LA data. However, ongoing LA data incompleteness and quality issues are limiting the overall public value of the statistics, and it is not clear how DLUHC plans to enable data linkage for those LAs not yet able to submit data to the H-CLIC specification.
The overall public value of the statistics could also be enhanced by the statisticians improving the overall accessibility of the statistics package, and the source data for further analysis, and providing clarity around the statistics’ key insights and messages. DLUHC taking a more transparent and strategic approach to user engagement and its communication around future developments, as an ongoing way of working, would help realise the potential public value of the case-level data.
We have identified six actions for DLUHC to address to enhance the public value and quality of the Statutory Homelessness in England statistics and to achieve National Statistics status. These are described in chapters one and two of this report.
Once the statistics team demonstrates that these steps have been undertaken OSR will recommend that the UK Statistics Authority designate the statistics as National Statistics.
Key Findings - Public Value
The introduction of the H-CLIC data requirement has considerably enhanced the potential of DLUHC’s statutory homelessness statistics to answer key policy questions, through new insights on households’ progression through the homelessness system. These insights are directly informing policy initiatives designed to reduce homelessness and its impacts. For example, insights around homeless prison leavers being at increased risk of sleeping rough before making a statutory homelessness application are leading to policy efforts designed to help prison leavers find accommodation before their release.
Users highly value the statistics and welcome where value has been added through the additional commentary provided to aid interpretation during the COVID-19 pandemic, and the annual flows analysis, which showcases the potential of the case-level data. In terms of presentation, the overall statistics package is presented as a report on households at various stages of the HRA system and is not very accessible to non-expert users. Insufficient narrative is provided to explain what the statistics show. This means that the statistics are not easy to navigate, and it is difficult to determine their value in terms of the insights they offer, or the questions they help answer. The public value of the statistics would therefore be strengthened by improving accessibility to the statistics’ key insights and messages. It would also help if the narrative placed the statutory homelessness statistics more clearly within the broader homelessness landscape.
DLUHC carried out a range of engagements as part of the H-CLIC development process, providing particular support to LAs to help them during the transition. More recently, the statisticians ran an online user event in May 2021 where they presented some of their planned developments. However, some users told us that they were not aware of the event or the development plans, which are not very prominent on the DLUHC website. Users told us that they would like a range of additional analyses and access to H-CLIC data for analysis. DLUHC should take a more transparent and strategic approach to user engagement and the communication of future developments, as an ongoing way of working. This would allow a broader range of users to feed in views on planned developments before they are introduced, and to help evaluate or make suggestions for further improvements once they are.
LAs also told us that H-CLIC is a much more burdensome data requirement than the previous P1E collection. However, the introduction of H-CLIC has increased the value of the statutory homelessness statistics and data and their potential to answer key questions. Plans to link H-CLIC case-level data with administrative health and benefits data and other sources, such as DLUHC’s rough sleeper survey, promise to enhance the value of the statistics for policy making. To enable data linkage, DLUHC is setting up separate data sharing agreements with participating LAs to allow them to share personal data associated with individual cases submitted on H-CLIC. However, it is not clear how DLUHC plans to enable data linkage for LAs not yet submitting data to the H-CLIC specification. DLUHC told us that it is developing a new API (Application Programming Interface) collection method that has the potential to reduce burden on LAs once introduced.Back to top
Key Findings - Quality
Users and data suppliers that we spoke to told us of the positive relationships they have with the statistics team at DLUHC, with the team also described as being helpful with any queries or issues and that the statisticians were timely in their responses. Users report that the quality of the statistics appears to have improved over the past few years since the introduction of the H-CLIC data collection. However, the switch from the previous P1E data collection to H-CLIC has not gone smoothly for some LAs, with some still unable to return H-CLIC data.
Some data suppliers and users we spoke to raised concerns around aspects of H-CLIC data quality, such as LAs that may not be recording their data on a consistent basis, or not asking certain sensitive questions, such as on gender identity and sexuality. Equally, applicants may not be comfortable supplying sensitive information. We also heard that housing options officers may prioritise collecting data essential to supporting a homeless application, with some fields, such as characteristics not related to priority need, being less likely to be completed accurately. This was particularly for cases where officers may have had only one interaction with an applicant, compared those where follow up interactions enabled the collection of further data. The statistics team told us that it is working to enable all LAs to return data through H-CLIC, and improve overall LA data quality.
Since our previous assessment in 2015, we have seen improvements in the supporting information made available to accompany the statistical release, for example the Technical Note, which is published each quarter. We welcome some of the changes made to the latest technical notes. However, some of the value has been lost through the removal of other information from the most recent technical note, such as information on the comparability H-CLIC and P1E data. We recommend that the team uses the guidance set out in our Quality Assurance of Administrative Data (QAAD) framework to help determine its assurances around the extent of data quality issues, and to communicate these assurances to users.Back to top
Key Findings - Trustworthiness
The statistics are released in an orderly way and are presented impartially and objectively. The statisticians maintain an up-to-date list of individuals granted pre-release access to the statistics in their final form ahead of publication and have a published revisions policy. The statistics team also has good relationships with the homelessness policy team within DLUHC, with briefing lines agreed between the teams to avoid potential misuse of the statistics.
The team told us that strong support is provided by the Head of Profession for Statistics at DLUHC, including in terms of training and resources. New starters attend Code of Practice Core Curriculum sessions and training on handling sensitive information appropriately. This assessment has not highlighted any concerns around the team’s data governance practices.Back to top
DLUHC should publish an action plan alongside the statistics on its website which sets out its proposals for addressing the assessment requirements. DLUHC is working towards having implemented the requirements in time for the next annual publication in Autumn 2022. We expect the DLUHC statistics team to report back to us after the publication of each quarterly bulletin, outlining the steps that it has taken to address the requirements. The UK Statistics Authority will take advice from OSR based on the evidence received and decide whether to award the National Statistics designation.Back to top