Thank you for your letter dated 20 May in response to our concerns about the use of the Ministry for Housing, Communities and Local Government’s (MHCLG) most recent rough sleeping figures in public debate. I am writing to set out our thinking on the apparent reductions in the number of rough sleepers in areas receiving Rough Sleeping Initiative (RSI) funding.
I appreciate the inherent difficulties in the accurate counting of rough sleepers in England and the limitations of the snap-shot methodology. These were of central concern to the users we spoke with during our 2015 assessment. Users also highlighted that the statistics were insufficient for assessing policies designed to reduce rough sleeping at the time.
I welcome your update that MHCLG plans to improve its broader rough sleeping evidence base and local authorities year-round recording of rough sleepers. I hope that this leads to a broader statistical view of rough sleeping in England beyond the snap-shot count alone, in the same way that MHCLG statisticians are currently working to provide a fuller statistical picture of homelessness more generally.
My team recently met with you to discuss some of our concerns about the recording of rough sleepers in RSI areas – the 83 areas with the largest numbers of rough sleepers in England. While we recognise that the methodology allows local authorities to decide whether a ‘count’ or an ‘estimate’ is the best collection approach in their areas, between 2017 and 2018, 19 RSI areas changed their data collection approach from an estimates approach to a count. One local authority changed their approach from a count to an estimate. We are concerned that these changes may be material for the comparability of the statistical series.
I therefore welcome your plans to publish an evaluation of the Rough Sleeping Initiative this summer. In order to establish the comparability and overall accuracy of the statistics, it seems particularly important that the evaluation can determine the relative differences between the levels and trends recorded by the counts and estimates approaches. Clarity here is essential to ensure that these statistics can be considered a reliable source for monitoring rough sleeping policy. The inclusion of many new additional RSI areas in this autumn’s rough sleeping count only increases the urgency with which this clarity is needed.
Until the evaluation confirms otherwise, I am not confident that changes in the number of rough sleepers are reliable. They may in part be a result of changes in measurement approach. In the light of this concern, I understand that you have improved the ministerial guidance to avoid too much weight being placed on these statistics before the evaluation is published, and I welcome this guidance. I also appreciate your continued professional scepticism, as well as your openness to review the snap-shot methodology should the evaluation reveal significant limitations for the counts/estimates approach for effectively evaluating the Rough Sleeping Initiative.
Director General for Regulation