Dear Sandra 

Management information on support for people rough sleeping in England

I am writing to endorse the publication of new management information on support for people rough sleeping in England, in February this year. These data were collected during the UK Government’s efforts to protect those sleeping rough, or at risk of sleeping rough during the COVID-19 pandemic, following the ‘Everyone In Initiative.’ They provide monthly estimates at local authority level for: people sleeping rough on a single night; people in emergency and short-term accommodation; and people that have moved on to long-term accommodation. My team has carried out a review of the information against the pillars of the Code of Practice for Statistics. We have shared our findings with DLUHC analysts and I appreciate their positive engagement with my team during this process.  

OSR supports the stated aims set out in the published information to: ‘provide a fuller picture of the dynamic and seasonal nature of rough sleeping’… ‘driving better outcomes through increased data and transparency’. This approach is very much in line with OSR’s published guidance on intelligent transparency. The need for data to better inform planning to end rough sleeping in England was also highlighted in the 2021 Public Accounts Committee report: COVID-19: housing people sleeping rough, which stated that plans were being made for the monthly publication of the data at that time.  

The published information is a significant step forward in terms of the level of insight provided about the UK Government’s support for rough sleepers, or people at risk of rough sleeping. It brings together new monthly data on the numbers of people supported into emergency, and longer-term settled, accommodation and a new monthly rough sleeping snapshot series. To help realise the significant public value of these data, our view is that they should be published as official statistics. Their value would also be significantly enhanced by releasing them at more timely and regular intervals throughout the year. This approach would also support the Public Accounts Committee’s interest in better data to inform future planning to end rough sleeping in England. 

Below are some more detailed recommendations to support DLUHC analysts in their further development of this data, in line with the three pillars of the Code of Practice for Statistics – Trustworthiness, Quality and Value. 

Trustworthiness 

• DLUHC analysts told us that they have recently refined the rough sleeping data requirement for local authorities, and that there is the potential to publish this information at more timely and regular intervals in future. OSR supports such an approach. It would be good practice for DLUHC analysts to set out their short and medium-term plans for the development and more regular publication of the data.  
• As they look at options for publishing these data in a regular and timely manner, DLUHC analysts should provide advance notice of future publications, as done with other established homelessness and rough sleeping statistics. Doing so will support users in finding out when and where the information will be made available.
• In light of the potential publication of more detailed and timely data on rough sleepers, we recommend that DLUHC analysts carefully consider the need for data suppression of small numbers to prevent individual rough sleepers being identified.
• We strongly support the publication of this data official statistics as soon as is feasible. Until this occurs, you may wish to consider the voluntary application of the Code pillars – as DLUHC has done for other published management information – as a framework for informing further development work and to signal the standards that DLUHC analysts are applying. 

Quality 

• DLUHC analysts published some quality information alongside the February publication as part of the annual snapshot technical report. This includes helpful quality insights for new monthly rough sleeping snapshot estimates, such as correlations between the monthly and annual snapshot data. DLUHC analysts should further develop their understanding and commentary on quality and uncertainty in the data, especially for the new emergency accommodation, and move on accommodation series. The team might find it useful to refer to OSR’s Quality Assurance of Management Information guidance, as it undertakes this work. 
• DLUHC analysts should consider what new quality documentation might be helpful to support the publication of future data made available on a more regular basis. This should communicate analysts’ judgements about how the information should or should not be used, based on their understanding of known strengths and limitations, or uncertainty in data quality.
• It would be helpful if more detail of the methods used to collect the data could be published – for example, whether the new monthly rough sleeping estimates are based on a count, estimate or spotlight method, or what approaches are used to help ensure local authorities record emergency and move on accommodation data, in a consistent manner.

Value 

• The new monthly data are a significant step forward in terms of offering an enhanced understanding of the numbers of people that have been supported into emergency, and longer-term accommodation since the start of the pandemic. The new rough sleeping snapshot series also fills a longstanding gap for year-round estimates, and complements the annual snapshot measure. The figures are presented clearly and effectively to show the seasonal variation in rough sleeping over the year and trends in emergency, and move on accommodation. 
• The bulletin states that the main use of the information is to measure local authority performance on ending rough sleeping, inform the ongoing response to the pandemic and support future policy development. Our view is that these uses would be much better supported by the more timely and regular publication of the data throughout the year – rather than on an annual only or ad hoc basis – and with its production carried to the standards required of full official statistics.  
• The overall presentation and the detail provided is fairly limited compared to DLUHC’s other homelessness and rough sleeping publications which include, for example, expanded analysis, commentary and interactive data tools. There is a lot of potential to add further insight and value – for example, though interactive data tools, enhanced demographic information across all three series, or the further disaggregation of the accommodation series to enhance insights on outcomes. To inform potential further developments to the published information, DLUHC should seek feedback from a range of users to maximise the value added from future improvements.
• DLUHC analysts should also consider how the further development of these data could support a more holistic view of homelessness and rough sleeping as a theme – for example, how they might be used to complement both DLUHC’s quarterly statutory homelessness statistics, and the annual snapshot estimates, and to explain the relationships and overlaps in coverage.
• DLUHC analysts should look to share this work with other UK analysts throughout the Government Statistical Service, as a good example of how advances in the understanding important societal topics can be gained by developing management data sources and making them public.  

In summary, we very much welcome the publication of this important data and look forward to seeing you publish them as official statistics that meet a range of user needs. Our Housing, Planning and Local services team will continue to liaise with DLUHC analysts over the coming months as they look to take this forward.  

I am copying this letter to Jon White and Sally Frost, the Responsible Analysts.

Yours sincerely

Ed Humpherson
Director General for Regulation