Proposals to stop the annual national rough sleeper count: summary of responses

Welsh Government sought users’ views on our proposal to cease requesting local authorities to undertake the annual rough sleeper exercise.  

The evidence for the proposed change was that there are limitations to the annual rough sleeper count. It is inherently difficult to identify all people who sleep rough and to therefore include them in data collections when compared with people with more secure forms of accommodation.

The national rough sleeper count provides a snapshot estimate of rough sleeping levels on a single night. The methodology provides robust, verified estimates but despite best efforts some people rough sleeping may be missed. Snapshots can be difficult to conduct and findings can be affected by external factors such as the weather, making comparisons over time difficult. Also, the individuals and number of people sleeping rough on one night will not be the same as on other nights. Some information on rough sleepers is now captured as part of the monthly Homelessness accommodation provision and rough sleeping data collection which was established during the coronavirus (COVID-19) pandemic.

Many thanks to all those who took the time to provide feedback. 

We received 19 responses, of which 16 agreed with the proposal to cease the annual national rough sleeper count. Of those that disagreed, their concerns were around potential loss of coherence/comparability on rough sleeper statistics across the UK and concerns around the quality of alternative information.

We have considered the feedback and have concluded that now is an appropriate time to cease the national annual rough sleeper count, with immediate effect. Further detail is provided in the Conclusion section. 

Following this decision, the Welsh Government will monitor and review the impact of these changes to ensure that users’ needs continue to be met and that the quality of information published is not adversely affected.

From 2015 to 2019, the annual national rough sleeper exercise was carried out by local authorities, in partnership with other local agencies, to gauge the extent of rough sleeping across Wales. 

Data was collected in two ways: a two-week information gathering exercise (usually undertaken during a period in October), followed by a one-night snapshot count in November. The results were published by Welsh Government (National rough sleeper count).

The national rough sleeper count was initially suspended in 2020 due to the coronavirus (COVID-19) pandemic, but was subsequently also suspended in 2021, 2022 and 2023. 

Since August 2020, in addition to the collection and publication of annual data on statutory homelessness, local authorities in Wales have submitted monthly information which has been published by Welsh Government in Homelessness accommodation provision and rough sleeping. This includes the number of people placed in temporary accommodation and the number of people sleeping rough. 

Further information on the alternative source of rough sleeper information and future plans for the collection are available on the Proposals to stop the national annual rough sleeper count page.

Respondents were asked a number of questions on the proposal to cease the annual rough sleeper count. 

Potential users of the annual rough sleeper count data and known users of housing statistics were informed of our proposal and invited to provide feedback. These included national and local government, academics and charities. The proposal was also published on the Welsh Government website to allow members of the public to respond with their views.

Of the 19 respondents, 11 were from local authorities, with the remaining 8 being from charities, academics and the private sector.

Question 1

Please indicate whether the proposal to cease the annual rough sleeper count will impact your work? (and if so, how)

15 respondents stated that ceasing the annual rough sleeper count would have no adverse impact on their work, including all 11 local authorities. 2 of these local authorities also stated that the proposal would have a positive effect, as ceasing the count would reduce costs and burden on the local authority. 

Multiple respondents stated that they actively work with outreach services to gather more timely and accurate data on rough sleepers, making the annual rough sleeper count redundant. A number of responses also highlighted issues with the one-night count methodology and stated that they did not believe it was an accurate reflection of rough sleeping figures across local authorities and Wales. 

2 respondents stated that ceasing the annual rough sleeper count would have an adverse impact on their work. 1 response stated that moving away from the point-in-time snapshot entirely would limit Wales’ ability to be measured against other UK nations and its global peers. 

Furthermore, multiple respondents stated that they had some concern about the robustness and quality of the monthly data collection and believe some improvement work would need to be carried out if Welsh Government was to cease the annual count. 

Question 2

Please explain which statistics you use from the annual National Rough Sleeper Count for Wales; how do you use these statistics? 

The majority of responses stated that they do not use the annual rough sleeper count data, with the introduction of the monthly collection being a major factor in this. Although the data was used previously by some, the monthly collection allows a number of respondents to monitor trends, inform policy and provide the most appropriate support provision. 

2 respondents stated that they use the annual rough sleeper count data for both policy development and to examine trends across the UK. They believe discontinuing the count would significantly diminish the available data for tracking progress and again reiterate the concern in the current monthly collection’s robustness. It would also limit the ability to monitor historic trends in Wales. 

Question 3

Please indicate the impact of the suspension of the annual rough sleeper count between 2020 and 2024, and what alternative data sources you’ve used.

The majority of respondents stated that the suspension has had little to no impact on their services. A recurrent response was the use of the monthly data collection, as well as data captured via their own unique systems. They also stated the importance of using outreach services to provide more accurate and comprehensive data. 

1 respondent stated that the suspension of the annual rough sleeper count has detracted from their ability to show trends over time in Wales and make useful comparisons between Wales and other nations across the globe. They also mentioned that there is a lack of reassurance on the quality of data in the intervening period. Another respondent stated that the data feeds into a large publication and that the current rough sleeper statistics are not a suitable alternative due to differences in data collection, data definition and quality assurance. 

Further comments

Other comments included

• Support the plans described regarding the development of a new, individual-level homelessness data collection system. 
• Despite other sources of data being available, there is a need for increased data on homelessness to better reflect the complexities of homelessness across Wales and help to develop an efficient (as well as preventative) response to cases of homelessness or an individual's risk of homelessness.
• Suggest that any new data parameters should highlight the individual’s reason/cause for rough sleeping, previous accommodation/residence and key support needs. 
• Unclear what “local intelligence” means in practice as this phrasing is open to interpretation. The lack of detail on methodology within this guidance is likely contributing to a failure to capture the true number of people sleeping rough in the community. 
• Suggest that the two-week exercise is conducted by all local authorities, and audited by the Welsh Government. The results of this exercise should also be compared to the monthly statistics collected within the same month to determine the level of variance between the two datasets and assess whether the monthly statistics are fit for purpose.
• Important to retain clear and accurate data in the interim to inform the Welsh Government’s new Ending Homelessness Outcomes Framework. 
• Believe that there are significant data gaps at the intersections of housing, mental health, and criminal justice.

Most respondents agreed with the proposal to stop the annual national rough sleeper count. Those that disagreed raised concerns around potential loss of coherence on rough sleeper statistics across the UK and concerns around the quality of alternative information. 

Whilst England and Northern Ireland currently conduct snapshot surveys each autumn to identify and record the number of people sleeping rough at a point in time, there is no rough sleeping snapshot conducted in Scotland. 

In their report, Rough sleeping in the UK (Office for National Statistics) (ONS)), the ONS highlighted that current statistics between countries are not directly comparable because of small variations in methods, and that the snapshots are best used to highlight the general level of rough sleeping at a point in time or to indicate possible trends in changes in rough sleeping over time 

Welsh Government believe that data from its monthly Homelessness accommodation provision and rough sleeping collection can be used to monitor broad trends and, therefore, ceasing the national rough sleeper count won’t lead to a loss in coherence across the UK. 

Another concern raised was data quality of the monthly data collection. Following the introduction of the monthly publication Welsh Government and local authorities have worked together to improve the data quality by strengthening the data collection guidance. This has led to improvements in the consistency and robustness of the data and we have recently moved to publishing as Official Statistics, rather than as Management Information. Welsh Government will continue to seek user feedback on these statistics and work with local authorities to improve guidance where appropriate, to ensure data is of the highest quality  and continue to meet the standards set out in the Code of Practice for Statistics. 

Local authorities actively work with outreach services and other local intelligence to gather more timely and accurate data on rough sleepers when compared to the previous one-night count. It is believed this approach is much more robust and shows a true reflection of rough sleeping across Wales. 

For these reasons, we have decided to cease the annual national rough sleeper count.

Statistician: Sam Voyce
Email:  stats.housing@gov.wales

Media: 0300 025 8099