FINAL EVALUATION OF THE ROUGH SLEEPERS INITIATIVE
1 The RSI was launched in 1997 in response to evidence of increasing levels of rough sleeping across the country. The programme was initially aimed at reducing levels of rough sleeping, following the broad model established by the RSI programme in England in 1990. In 1999, the programme was given a new target to end the need to sleep rough in Scotland by 2003.
2 An evaluation of the Rough Sleepers Initiative (RSI) programme was undertaken by the Centre for Housing Policy, at the University of York, during the Autumn of 2004 and the Spring of 2005. The evaluation involved a critical review and analysis of existing research and statistical evidence, interviews with local authority officers responsible for RSI, national level interviews with individuals responsible for the development and delivery of RSI and interviews with a sample of service providers as well as former, current and potential rough sleepers.
3 The evaluation found statistical and qualitative evidence that the RSI programme had successfully produced tangible reductions in the need to sleep rough across the country.
4 The monitoring of the need to sleep rough by George Street Research is a broadly reliable indicator of RSI success. This monitoring indicated a decline of the number of people sleeping rough presenting to services of more than one third between 2001 and 2003.
5 The data collected by Glasgow Homeless Network from RSI funded projects provide a very detailed demographic picture of rough sleepers across the country. The Glasgow Homelessness Network monitoring also suggests that new rough sleepers continue to present themselves to services in relatively small numbers. At present, the database does not have sufficient coverage of the outcome of service engagement with former, current and potential rough sleepers.
6 HL1 data, collected by local authorities on households presenting as homeless provide a comprehensive picture of the extent of recent rough sleeping among homeless households. Again, these data illustrate that households with experience of sleeping rough, while in a minority, continue to present as homeless.
7 Local authority respondents reported that RSI had enabled the development of a suite of services within the major cities that were able to provide an enhanced range of support to people sleeping rough. In rural areas and smaller urban authorities, local authority respondents reported that RSI had quite often created specific services for the first time.
8 Local authority respondents also reported that RSI had produced important local cultural and political changes in placing rough sleepers and their needs firmly within the policy agenda. RSI was seen by many local authority respondents as the catalyst for more recent policy developments including health and homelessness action plans, homelessness strategies and the integration of homelessness services within Supporting People planning. It was the cumulative effect of these changes that was often seen as most significant in reducing levels of rough sleeping in their localities.
9 RSI had been mainstreamed into the homelessness strategies of most local authorities. In some cases, such as Edinburgh, there was full integration of RSI budgets with Supporting People homelessness budgets and other funding streams, at both strategic and service delivery level. The same pattern existed within most smaller urban authorities and rural authorities. Although a few authorities were less far down this path, all were heading in the same direction.
10 At service delivery level, most RSI funded services were also in receipt of Supporting People funds, often at greater levels than the RSI funds they had access to.
11 Service providers generally shared the positive views of local authority respondents about RSI. They saw it as facilitating the development of effective service responses for people sleeping rough and as acting as something of a catalyst in encouraging wider joint working and strategic planning across homelessness services as a whole.
12 Service providers and local authority respondents saw some limitations in the effectiveness of RSI. Some groups of former, current and potential rough sleepers were difficult to reach, such as a small number of people with multiple needs and challenging behaviour. Poor access to certain care and support services, such as drug rehabilitation, was also seen as diminishing service effectiveness. Most respondents talked about the effects of shortages of suitable and affordable housing in which to re-house former, potential and current rough sleepers in the areas in which they worked.
13 Service users were generally positive about the RSI funded services. Some reported the same difficulties in accessing suitable housing and certain services, particularly drug rehabilitation, as were reported by local authority respondents and service providers.
14 The RSI has been a successful programme that has largely fulfilled its objective to end the need to sleep rough in Scotland. The introduction of a flexible funding programme allowed the development of new services in areas that had previously lacked any specific provision and also enabled the further development of the sector in those areas that had some service provision. RSI was widely seen as having culminated in the adoption of local authority homelessness strategies which are coordinated with both health and homelessness action plans and Supporting People plans. Consequently, services for people sleeping rough are increasingly integral to strategic responses to homelessness. Positive changes in cultural and political attitudes, which raised awareness of the multiple needs among people sleeping rough and placed their needs on local and national agendas were strongly associated with the introduction of RSI. There is statistical and qualitative evidence that significant, tangible reductions in the levels of rough sleeping have occurred since the programme began.
15 There are limits to the effectiveness of RSI. Some groups of former, current and potential rough sleepers are difficult for services to engage with, as much because of their situation and characteristics as because of the finite resources available to those services. In terms of service delivery, beyond the existing provision of services that specifically target the most marginalised and challenging people sleeping rough, it is difficult to see what else might be done. After a certain point, ever increasing levels of expenditure on what is quite a small group of people with high needs, would start to become hard to justify.
16 There are other changes outside direct service delivery that can potentially benefit people sleeping rough. The increased coordination and comprehensiveness of responses to all forms of homelessness has been of general benefit to rough sleepers and it can be anticipated that the ongoing legislative changes will ease their access to accommodation in some respects. At the same time, however, a lack of suitable and affordable accommodation supply is evident across the country and this will continue to limit the effectiveness of responses to homelessness at strategic and service delivery level. There are also issues in respect of access to certain kinds of health and social work services, with the adequacy and accessibility of drug detoxification services for people sleeping rough, being highlighted in the fieldwork for this evaluation.
The future of the programme
1 There are good strategic and logistical arguments for integrating RSI planning, commissioning and service delivery within local authority homelessness strategies and associated Supporting People planning. The process of mainstreaming RSI services at strategic and service delivery level is effectively complete in several areas and should be encouraged where it is not yet completed.
2 Specific targets to ensure services are geared towards the needs of people sleeping rough should be integrated into local authority homelessness strategies and externally monitored, to ensure that the focus brought to rough sleeping by RSI is not lost.
3 If integration of RSI funding with other funding streams were to occur, it would be of central importance to retain the flexibility that has characterised the programme. For example, if RSI funding became integrated into Supporting People, the usual rules with respect to tying funding of services to accommodation would need to be suspended for services for people sleeping rough. Specific modifications to the criteria for funding services for particular client groups are commonplace within the Supporting People programme.
4 There is evidence of a continuing need for rough sleeper services. Any significant reductions in expenditure on homelessness and rough sleeper services are likely to produce corresponding rises in rough sleeping.
5 Further consideration should be given to investigating the effectiveness of preventative services, in the light of evidence of ongoing need.
6 The provision of highly supportive long-term housing settings should be investigated as a possible option for meeting the needs of people sleeping rough with multiple needs and challenging behaviour.
7 Specific initiatives such as RSI are affected by the context set by wider housing and social policy across the country. Issues such as the availability of suitable and affordable housing across different areas will have an impact on the effectiveness of homelessness strategies in relation to rough sleeping. Wider policy debates should take account of homelessness and rough sleeping where applicable.
The monitoring of rough sleeping and rough sleeper services
8 There is a strong case for maintaining a specific national target on rough sleeping to retain appropriate attention on this easily marginalised group. However, when the 2003 legislation is fully implemented, it may be sensible to revise the 'no-one need sleep rough' target to reflect a changed context whereby there will be a duty on local authorities to supply accommodation to all homeless groups. If this revised target relates to reducing the overall numbers of people sleeping rough, it may be possible to assess this with the suggested modified version of the GHN database.
9 The need to continue the monitoring of rough sleeper services is clear, in order to assess effectiveness and to provide data for local and national policy planning and strategy. The existence of the national dataset on rough sleeping provided by GHN gives Scotland a much clearer picture of progress in tackling rough sleeping than is available in England.
10 There is a need to address issues in respect of the data entry systems within the GHN monitoring database, as quality control needs some further development.
11 The GHN monitoring lacks sufficient outcome measures, it collects insufficient information from an insufficient number of organisations. Both the range of data collection and the response rate need to be improved.
12 Although there are problems with the GHN monitoring, this dataset provides a wealth of data within a very small operating budget. There are good arguments for retaining the role of GHN in managing a revised database system, despite some operational problems, because of the degree of success that has been achieved.
13 To maintain a separate 'RSI' database for the foreseeable future is illogical in the context of the mainstreaming of RSI funded services within local and national strategic responses to homelessness. Such a database would represent a increasingly arbitrary set of homelessness projects. Consequently, if the GHN database is to be maintained, it would be logical to roll it out across homelessness services throughout the country.
14 The database developed by Edinburgh City Council, which is outcome led and covers all homelessness services in the city, should be examined in detail and any valuable lessons transferred to a revised GHN database. The capital's database system provides both the outcome measures and the universal coverage of homelessness services that would make a national database of great utility for policy planning at local, regional and national level.
15 Monitoring should enable the logging of whether people sleeping rough are within couples or other forms of household and whether this has acted as an obstacle to services and also take account of whether pets have acted as an obstacle to services.
16 Consideration should be given to one extension to HL1, which would be asking a question about lifetime or sustained experience of rough sleeping. This would provide a greater depth of information and allow analysis of the extent to which local authorities might be housing people with sustained experience of sleeping rough. Again, this modification would be of particular interest following the implementation of the 2003 Act.