Research and analysis

Tenancy sustainment and complex needs: A rapid evidence assessment

Published 11 December 2025

Applies to England

Emma Bimpson, Stephen Green, Angela Maye-Banbury,
Lindsey McCarthy, Kesia Reeve, and Beth Speake

Centre for Regional Economic and Social Research, Sheffield Hallam University

Executive summary

Introduction

This is a rapid evidence assessment (REA) about issues and practices associated with tenancy sustainment for people with a lived experience of homelessness and dealing with issues associated with addiction and poor mental and physical health, commonly referred to as complex needs. It was commissioned by the Ministry of Housing Communities and Local Government (MHCLG) and conducted by a team at the Centre for Regional Economic and Social Research (CRESR) at Sheffield Hallam University.

Evidence suggests that there are significant numbers of people with complex needs in England. For example, the Hard Edges report on severe and multiple disadvantage found that over a quarter of a million people in England each year have contact with at least 2 out of 3 of the homelessness, substance misuse and/or criminal justice systems, and at least 58,000 people have contact with all 3 (Bramley et al., 2015). There is also some evidence of significant differences in housing outcomes for homeless people with complex needs, compared to those without complex needs. Harding and Willett (2008), for example, conclude that having complex needs is 1 of 2 factors (the other being an institutional background) that makes resettlement after homelessness more difficult.

Methods

After a literature search and sift to discard sources that were irrelevant, 102 relevant sources were identified and assessed. Very few sources were identified that specifically explored tenancy sustainment for people with complex needs, though relevant information and insights were found in outputs about homelessness more broadly. Housing First was a particularly well evidenced type of housing intervention that describes and, sometimes, measures tenancy sustainment. As such, this body of evidence has been contained in a distinct section of this report.

All studies were assessed against a weight of evidence criteria, outlined in Table 1.1. Quality assessment was not used to exclude studies that were included in the final sample. Rather, it was used to guide how the evidence was used in the review.

Housing First

The rapid evidence assessment found that Housing First and associated models provided key evidence for tenancy sustainment for those with complex needs and a lived experience of rough sleeping. This is partly a function of the focus it has received in policy and practice circles and the amount of research and evaluation that has occurred. Housing First schemes have been established in the UK which have been (or are due to be) subject to independent evaluation. The evidence base in the UK - while still more limited than robust international studies - is growing, and existing evidence clearly demonstrates positive housing retention outcomes (consistent with those in countries where Housing First is more established).

The Housing First literature reviewed provides relatively robust evidence that a model providing rapid rehousing with intensive support produces positive tenancy sustainment outcomes for homeless people with complex needs, at least in the short-medium term. This broad conclusion is useful. However, the evidence does not provide definitive conclusions about why such a model works so well and for whom - about the specifics of those factors supporting or undermining tenancy sustainment.

Tenancy sustainment drivers and barriers

Evidence draws attention to the fact that successfully ending homelessness is a process, and one which involves addressing multiple, overlapping issues (Chamberlain & Johnson, 2018), often relating to the complex reasons for homelessness, which must be understood for resettlement to succeed. Reflecting the complexity of pathways into homelessness and the multi-directional influence of challenging life experiences, the evidence is explicit that promoting tenancy sustainment (or successful resettlement) for people with complex needs requires attention across multiple domains. Factors found to be most important to tenancy sustainment, therefore, should be regarded as just one factor amongst a complex set of overlapping issues.

Evidence focused on ‘resettlement’ suggests that access to housing is a prerequisite for a sustainable tenancy and highlights contextual challenges of housing supply, availability and accessibility that need consideration in efforts to promote tenancy sustainment.

There is very little evidence capable of clearly isolating group characteristics with tenancy sustainment or failure in the UK. The methodological approach commonly taken to studies of resettlement or tenancy sustainment programmes in the UK precludes this kind of quantitative analysis and/or studies are too small for robust sub-group analysis.

Tenancy sustainment in different types of accommodation

Most of the evidence identified about tenancy sustainment amongst people with complex needs (including Housing First) focuses on independent tenancies in the social or private rented sectors, or tenancies in medium-term supported housing projects. In relation to the former, the tenancies in which study participants live is often accompanied by some form of floating support, including from Housing First or tenancy sustainment team workers. Very few sources were identified reporting evidence of tenancy sustainment amongst people with complex needs in independent tenancies without support.

With no evidence identified on tenancy sustainment amongst people with complex needs in other forms of housing - for example hostels or other short-term or emergency supported housing - it is not possible to draw conclusions about whether the factors driving tenancy sustainment vary by accommodation type.

Tenancy sustainment support

The evidence relating to effective tenancy support is typically about people who are homeless, but not always about people with complex needs. There is evidence about the type of support, and the principles of service delivery that are most effective for people with complex needs, but in different policy areas. However, relevant insights can be gleaned from this evidence. It appears that having formal tenancy support arranged prior to a move to individual accommodation is very important, as is having to support to gain key skills to independently manage a tenancy effectively. The immediacy of support, skills and expertise of the workforce, and the availability of long-term support also appear to be important factors.

Conclusions

The overall aim of this rapid evidence assessment was to identify and explore the barriers and drivers of tenancy sustainment for people with complex needs in an effort to identify what works best to support people with complex needs - and sub-groups within this population - to sustain appropriate accommodation. In an effort to meet this aim we have gleaned what insights there are from an extensive and relatively rich body of evidence about homeless people, but one which rarely measures or explicitly explores tenancy sustainment outcomes and experiences for people with complex needs. The exception to this is the body of literature evaluating Housing First schemes. Compared to the international evidence base, which commonly uses randomised control trials, the quality of outcome evidence in the UK is more limited.

There is compelling evidence that a Housing First style approach can offer high housing retention outcomes. There also exists a weight of evidence about the types of tenancy sustainment support, and the way they are delivered, that appear most effective for people with complex needs (with the caveat that many studies do not provide robust quantitative evidence of tenancy sustainment). Some broad principles can be identified - that positive sustainable outcomes for people with complex needs are most likely to be achieved if support is flexible, intensive, not time-limited, and person centred. It was also found that outcomes may be improved by developing a workforce that has the requisite skills to support and stick with people whose complex needs mean that there are steps backwards, as well as steps forward, on the way to their recovery.

The evidence is more limited in terms of the precise circumstances under which people with complex needs are most likely to sustain a tenancy. People resettled into the private rented sector have been found to be at higher risk of tenancy failure than those in the social rented sector, and social isolation amongst those who move away from social and familial connections is suggested to negatively impact on tenancy sustainment. However, the evidence is far from conclusive on these points and pinpointing why this might be the case is much more difficult. We do not, for example, know with any certainty which features of the private rented sector pose a risk to tenancy sustainment, or under what circumstances feelings of isolation do and do not translate into tenancy failure.

The evidence base does not, at the present time, provide clear answers to the question of what factors are most important in driving tenancy sustainment, and for whom. As a result, it remains difficult for policy and practice to develop services, initiatives and programmes for homeless people with complex needs, with confidence that they will effectively support this group of vulnerable people to sustain appropriate accommodation in the long term.

Gaps in the evidence

1. There are few statistically robust UK studies capable of clearly isolating group characteristics and the correlation with tenancy sustainment. Studies tell us a lot about what people value in their housing and support, or about difficulties they face, but there is little robust outcome evaluation capable of isolating specific variables driving tenancy sustainment.

2. There is a need for longitudinal studies which track long-term outcomes of clients. The definition of ‘sustainment’ has a temporal dimension in the sense that something is only said to be sustained if it has been stable over time. Yet many research studies and evaluations are conducted over a year or 2, including evaluations of Housing First schemes.

3. There is very little sub-group analysis in the evidence base. The methodological approach commonly taken to studies of resettlement or tenancy sustainment programmes in the UK precludes this kind of quantitative analysis and/or studies are too small for robust sub-group analysis. We do not, therefore, understand differential experiences and outcomes or the extent to which research findings apply to different groups.

4. In particular, there is a notable gap in evidence relating to sustaining tenancies amongst Black and Minority Ethnic (Sheikh & Teeman, 2018) and LGBT+ people with complex needs.

1. Introduction

1.1 Aims of the REA

This report details the results from a rapid evidence assessment (REA) about issues and practices associated with tenancy sustainment for people with a lived experience of homelessness and dealing with issues associated with addiction and poor mental and physical health, commonly referred to as complex needs. It was commissioned by the Ministry of Housing Communities and Local Government (MHCLG) and conducted by a team at the Centre for Regional Economic and Social Research (CRESR) at Sheffield Hallam University.

The key questions for this REA were as follows.

1. What are the drivers and barriers of tenancy sustainment for people with precarious housing histories and complex needs?
2. What factors are important to tenancy sustainment?
3. What forms of tenancy sustainment support have been successful? Particularly, what evidence is there about the type and duration of support, and who provides it.
4. Are there tenancy sustainment issues different for different accommodation types (e.g. hostels, other supported housing, to more independent housing in the social and private sectors)?
5. When people transition from temporary or emergency accommodation to more stable accommodation, do support needs change?
6. What can be done to help improve tenancy sustainment for those with complex needs?
7. What are the key evidence gaps?

1.2  Background

In 2018, MHCLG commissioned CRESR to conduct a 2-year process evaluation of initiatives for people with complex needs who are sleeping rough supported through 2 funding streams: the Rough Sleeping Grant (RSG) and the Rough Sleeping Social Impact Bonds (SIBs). The evaluation reported on the effectiveness of funded projects to rehouse and support people who were rough sleeping with complex needs. While the evaluation touched on the role that projects played in helping to sustain tenancies, it was clear that more research was required to understand the drivers and issues relating to tenancy sustainment, particularly for people with a lived experience of rough sleeping and dealing with issues associated with addiction and poor health (mental and physical health). Responding to this gap, and as a first step in understanding the issue, MHCLG commissioned this REA to establish what is already known about the factors supporting tenancy sustainment for people with complex needs.

Evidence suggests that there are significant numbers of people with complex needs in England. For example, the Hard Edges report on severe and multiple disadvantage found that over a quarter of a million people in England each year have contact with at least 2 out of 3 of the homelessness, substance misuse and/or criminal justice systems, and at least 58,000 people have contact with all 3 (Bramley et al., 2015). Within the homelessness data analysed for the report, only 34% of homeless people were classed as ‘homeless-only’ with the remaining two-thirds having additional needs.

There is also some evidence of significant differences in housing outcomes for homeless people with complex needs, compared to those without. Harding and Willett (2008), for example, conclude that having complex needs is 1 of 2 factors (the other being an institutional background) that makes ‘resettlement’ more difficult. Research exploring multiple disadvantage in Scotland found that while 69% of homeless households overall achieve a ‘settled housing outcome’, this drops to 36% for those with experiences of homelessness, offending and substance use (Watts et al., 2018, p. 91). Specifically, people with ‘three dimensions’ of severe and multiple disadvantage (homelessness, substance misuse issues and offending history), people with homelessness and substance misuse issues, and people with homelessness and mental health issues were least likely to secure a social housing tenancy (in Scotland) through homelessness services (Bramley et al., 2019).

1.2. Methodology

The REA took the following stages.

1. Inception: review questions were agreed and the inclusion criteria for the literature was defined.
2. Search for studies: this used a range of sources, including academic studies and ‘grey’ literature, with a focus on evidence-based literature.
3. Screen studies: initially to sift out irrelevant studies by title, followed by a second sift based on content.
4. Describe and appraise studies.
5. Synthesise findings and report.

A range of resources was used to identify sources, including: the existing knowledge of the team; the IDOX Service; databases and online libraries (including Google Scholar, JSTOR, Science Direct, SpringerLink, and SSNR eLibrary); and snowballing from reference lists.

Search terms

The literature search was conducted using a combination of different search terms to account for the breadth and limitations of the term ‘sustaining tenancies/tenancy sustainment’ in policy, practice and academic outputs. Tenancy sustainment has evolved around very particular and also very general meanings in policy and practice-based work, which presents challenges for a review of evidence based on keywords in titles or summaries of publications.

An initial scan of literature returned through web searches of ‘sustaining tenancies/tenancy sustainment’ identified a gap in material that focused on people with histories of homelessness or complex needs and highlighted the need to include a variety of different terms which would capture relevant information. For instance, relevant evidence might not have explicitly included ‘sustaining tenancies’ in the title but reported on specific interventions with that aim, such as Housing First.

Homelessness sector reports often use terms like ‘moving on’, ‘resettlement’, or ‘housing outcomes’ to describe the experiences of people leaving street homelessness, homelessness services and other more settled forms of housing. ‘Prevention’ and ‘ending homelessness’ are also commonly applied to outputs with relevant and transferable insights because these activities rely upon successful (re)settling into accommodation.

There are also issues with measurement. Typically, the duration of sustainment being measured ranges from 6 to 18 months, although longer durations of measurement were limited. In many sources, the duration of tenancy sustainment has not been measured at all. The represents a gap in evidence and also highlights the inconsistent ways in which a ‘sustained tenancy’ is measured and defined.

A search on titles and keywords of social science journals was performed in Web of Science and Google Scholar, and an additional hand search was carried out of key journals that include research on this topic. The Idox Knowledge Exchange service was used also. This was followed by an open web search, as well as a search of key research and knowledge centres in the field (e.g. JRF, I-SPHERE, Crisis, Oak Foundation, and Lankelly Chase Foundation).

The search terms used included the following:

• tenancy sustainment; sustaining tenancies; tenancy/ housing outcomes; moving on; resettlement
• social rented sector; social landlords; housing associations; supported housing; temporary accommodation; hostels; private rented sector
• housing-led versus support led interventions
• tenancy support; floating support; support for complex needs; co-occurring health needs; co-morbidity; vulnerabilities
• homelessness; rough sleeping
• mental health; substance misuse; offending histories

Sources used

While very few sources were identified that specifically explored tenancy sustainment for people with complex needs, relevant information and insights were found in outputs about homelessness more broadly, or other client group-specific literature.

Housing First was a particularly well evidenced type of housing intervention that describes, and sometimes more accurately measured tenancy sustainment. As such, this body of evidence has been contained in a distinct section of this report.

Once searches had been carried out, evidence was examined to determine relevance. Firstly, this was done by examining the title and secondly by briefly examining content. These initial searches were not limited by date or location, but these factors were important considerations when assessing the robustness and relevance of each source for inclusion in the evidence assessment.

This initial ‘light touch’ assessment resulted in some evidence being discarded on the grounds of relevance or methodological rigour and left approximately 121 sources for a more rigorous assessment.

It is important to weigh the evidence from different sources. To do so, we adopted the EPPI-Centre’s Weight of Evidence framework. This considers whether a study is well executed, and relevant in answering the review questions (Gough, 2007). It follows 4 dimensions described below in Table 1.1. Studies were scored low, medium or high against each dimension. Studies with an overall low score were not included in the review unless the team felt there was an important, reliable insight worth including. Following this exercise, 19 sources were excluded, resulting in a total of 102 sources being included in the REA.

Quality assessment was not used to exclude studies that were included in the final sample. Rather, it was used to guide how the evidence was used in the REA.

Table 1.1: Weight of evidence framework

A: Coherence and integrity	B: Specific judgement on forms of evidence	C: Specific judgement on relevance	D: Overall judgement
This is a generic and thus non review-specific judgement about the coherence and integrity of the evidence in its own terms. The NICE checklist designed to assess qualitative and quantitative evidence will be used.	This is a review-specific judgement about the appropriateness of that form of evidence for answering the review questions that is the fitness for purpose of that form of evidence.	This is a review-specific judgement about the relevance of the focus of the evidence for the review question. E.g. a research study may not have the type of sample, the type of evidence gathering or analysis that is central to the review question or it may not have been undertaken in an appropriate context from which results can be generalised to answer the review question.	This combines the previous assessments to give an overall weighting.

1.3  Report structure

This report is structured around the following chapter themes:

• Chapter 2: Housing First: insights from the evidence base
• Chapter 3: Tenancy sustainment drivers and barriers
• Chapter 4: Tenancy sustainment in different types of accommodation
• Chapter 5: Tenancy sustainment support
• Chapter 6: Gaps in the evidence base

These chapters address the review questions set out in section 1.1, with the exception of question 5 on tenancy sustainment between different forms of accommodation. No robust evidence was found that provided useful insights in this regard.

Also, the evidence that has been collected concerning Housing First is of key relevance to tenancy sustainment and complex needs, and therefore we begin with a distinct section drawing key insights from that body of literature.

2. Housing First: Key insights from the evidence base

2.1 Introduction

The rapid evidence assessment found that Housing First and associated models provided key evidence for tenancy sustainment for those with complex needs and a lived experience of rough sleeping.  This is partly a function of the focus it has received recently in policy and practice circles and the amount of research and evaluation that has occurred. Housing First evaluations were found to speak most directly to the research questions for this RAE insofar as Housing First programmes are generally concerned with measuring tenancy sustainment as a key ‘outcome’ and are targeted at people who are homeless with complex needs.  This chapter is, therefore, dedicated to what Housing First can tell us about tenancy sustainment for people with complex need. It first provides a background to Housing First and a UK policy context.  The following sections then provide a review of the evidence base on Housing First from the lens of considering its success (and limitations) in terms of tenancy sustainment or housing retention rates^{[footnote 1]}. Given the scope of this evidence review, the main focus was on UK literature; only the most robust and seminal international studies were reviewed. We also acknowledge that an evidence base exists for other outcome variables in addition to tenancy sustainment which is not considered under the scope of this review (Woodhall-Melnik & Dunne, 2016) for a systematic review of evidence outlining other Housing First outcomes such as substance use and mental health, service use and costs, and quality of life).

2.2  Background to Housing First

The Housing First model is understood to be North American in origin, with most sources citing the Pathways Housing First organisation, founded in New York in 1992 by Dr Sam Tsemberis, as pioneering the model as it is known today (Bellis & Wilson, 2018). Differing from other housing models, such as ‘Staircase’, ‘Continuum of Care’ and ‘Treatment First’, Housing First sees housing not as a reward for completion of a programme but as a basic human right to be provided immediately and on an open-ended basis. The Pathway Housing First model is underpinned by 5 core ideas: 1) housing is a fundamental human right; 2) clients should be provided with housing before other services are offered; 3) people undertake recovery at their own speed; 4) there is no requirement for abstinence from drugs and alcohol; and 5) the aim is re-integration into mainstream society (Chamberlain & Johnson, 2017).

Since its origins, Housing First has been adopted by organisations and governments across Europe, the US, Canada and Australia. As Housing First is interpreted and implemented with subtle differences according to local context, some have attempted to devise broad operational definitions of Housing First which both preserve its ethos and take account of local context. Lancione et al. (2018, p. 41) define Housing First as “a housing and support (harm reduction) policy directed at homeless people, in which access to housing and its continuance both take precedence over, and are independent of, any other type of programme”. Other sources refer to Housing First as a set of principles rather than a well-defined system (Bellis & Wilson, 2018). The following principles have been developed by Homeless Link (2017) and are largely based on the core principles in the FEANTSA Housing First Guide Europe (Pleace, 2016) but are adapted for the UK. Homeless Link advise that providers adhere to the following principles as closely as possible in order to provide high quality and successful Housing First services:

1. People have a right to a home - Housing First prioritises access to housing as quickly as possible; eligibility to housing is not contingent on any conditions other than willingness to maintain a tenancy; housing provided is based on suitability (stability, choice, affordability, quality, community integration) rather than type of housing;
2. Flexible support is provided for as long as it is needed - providers commit to long-term offers of support which do not have a fixed end date; recovery takes time and varies by individual needs, characteristics and experiences; the service is designed for flexibility of support with procedures in place for high/low intensity support provision and for cases that are ‘dormant’; support is provided for the individual to transition away from Housing First if this is a positive choice for them; the support links with relevant services across sectors that help meet the full range of an individual’s needs; there are clear pathways into and out of the Housing First service;
3. Housing and support are separated - support is available to help people maintain a tenancy and to address any other needs they identify; an individual’s housing is not conditional on them engaging with support; the choices an individual makes about their support do not affect their housing; the offer of support stays with the individual so if the tenancy fails, the individual is supported to acquire and maintain a new home;
4. Individuals have choice and control - individuals choose the type of housing they have and its location within reason as defined by the context (this should be scattered site, self-contained accommodation, unless an individual expresses a preference for living in shared housing); individuals have the option not to engage with other services as long as there is regular contact with the Housing First team; the individual is supported through person-centred planning and is given the lead to shape the support they receive;
5. An active engagement approach is used - staff are responsible for proactively engaging clients and making the service fit their individual needs; caseloads are small allowing staff to be persistent and proactive in their approach, doing ‘whatever it takes’ and not giving up or closing the case when engagement is low; support is provided for as long as the client requires it; the team continues to engage and support the individual if they lose their home or leave their home temporarily;
6. The service is based on people’s strengths, goals and aspirations - services are underpinned by a philosophy that there is always a possibility for positive change and improved health and wellbeing, relationships and community and/or economic integration; individuals are supported to identify their strengths and goals; individuals are supported to develop the knowledge and skills they need to achieve their goals; individuals are supported to develop increased self-esteem, self-worth and confidence, and to integrate into their local community;
7. A harm reduction approach is used - people are supported holistically; staff support individuals who use substances to reduce immediate and ongoing harm to their health; staff aim to support individuals who self-harm to undertake practices which minimise risk of greater harm; staff aim to support individuals to undertake practices that reduce harm and promote recovery in other areas of physical and mental health and wellbeing.

Although similar in nature, Housing First is distinct from other homelessness services (e.g. floating support) which include an element of tenancy sustainment. As Pleace (2018) - in a major review of Housing First evidence - comments, the intensive, sustained, and choice-led support with an emphasis on recovery offered by Housing First is what makes it distinctive.

2.3  Tenancy sustainment for people with complex needs

International evidence

There is now growing evidence which points to the success of Housing First as a model to end homelessness among people with high and complex needs. The majority of evidence on Housing First is international. In a systematic review of outcomes associated with participation in Housing First programmes, Woodhall-Melnik and Dunne (2016) cite only international (predominantly US) studies under the ‘housing and retention outcomes’. All of the studies they reviewed confirmed that Housing First participants (the homeless population with psychiatric symptoms, addictions, or with concurrent disorders) experienced high levels of housing retention (Palepu et al., 2013); reductions in homelessness (Greenwood et al., 2005); and were housed quicker (Tsemberis et al., 2004), concluding that there is strong, consistent evidence that Housing First is successful in improving housing retention for persons who are typically considered to be hard-to-house. The review does not specify the length of the period of housing retention, however; just that for some groups, levels were higher for those in Housing First than those placed in treatment-dependent housing (Montgomery et al., 2013).  

Similarly, in major reviews of international evidence, Pleace (2018) and Mackie et al. (2019: 87) conclude that evidence of Housing First ending homelessness is strong, “far stronger than is true of any other housing-related intervention targeting rough sleepers”.  Pleace also points to the range of international evidence being a sign that Housing First can be successful when applied to different national contexts with different housing, health, welfare and political landscapes. While there are a range of studies highlighting the success of Housing First in terms of improving the physical and mental health and well-being of clients (Bretherton & Pleace, 2015; Pleace & Quilgars, 2013) as well as social integration (Bretherton & Pleace, 2015; Pleace & Quilgars, 2013), this review focuses on evidence of Housing First as a means of ending homelessness through improving tenancy sustainment.

Based on a review of existing evidence, Pleace and Bretherton (2013a) calculate that Housing First generally ends homelessness for at least 8 out of every 10 people (in the US, Canada and Europe).^{[footnote 2]} As Pleace (2018) also found, the evidence on the success of Housing First in countries where it is more established is inevitably stronger and more extensive. For instance, in 2013, the Housing First Europe project reported that 97% of the high-need homeless people using the ‘Discus’ Housing First service in Amsterdam were still in their housing after 12 months in the service (Pleace, 2016); the French ‘Un Chez-Soi d’abord’ Housing First programme reported that 85% of homeless persons in the programme kept their home after 2 years (Estecahandy et al., 2018); and first results from the ‘Habitat’ Housing First programme in Spain show a large housing retention rate: 100% of users continued in their homes 6 months after its launch (Bernad & Yuncal, 2016). Similar high retention rates have been reported in other European countries based on being housed for one year (Pleace et al. (2019)). Similarly, US and Canadian studies report parallel high success rates. US studies have reported rates of housing sustainment between 80 and 88%; one study in New York reported that 88% of chronically homeless people using Pathways to Housing’s model were stably housed after 5 years (Tsemberis, 2010a), and an evaluation of the Canadian ‘At Home/Chez Soi’ programme reported that Housing First service users spent 73% of their time stably housed over 2 years, compared to 32% of those receiving other homelessness services (Goering et al., 2014). Similarly robust studies of Canadian Housing First schemes showed that Housing First clients experienced better housing outcomes than a comparison group of adults who were homeless with problematic substance use. They moved into housing more quickly, reported a greater proportion of time housed over 2 years, were more likely to have spent the final 6 months housed, and had longer housing tenure with the Housing First clients reporting more days consecutively housed at 24 months (Cherner et al., 2017)^{[footnote 3]}. Evidence showing the effectiveness of Housing First from France and Canada (Goering et al., 2014; Volk et al., 2016; Estecahandy et al., 2018) is based on highly robust randomised control trials and is difficult to refute. Studies comparing the outcomes of different Housing First programmes indicate that those most closely aligned with the Pathways model tend to report the best housing retention rates (Stefancic & Tsemberis, 2007).

Qualitative international studies provide us with more of an in-depth insight into the reasons behind successful retention rates. Drawing on data from an evaluation of 2 trial projects of the Housing First model in Norway, Hansen (2018), after interviewing 16 users of the projects, found that participants’ experience of the tolerant approach by staff, being treated with respect and acceptance, and having their opinions acknowledged was of vital importance regarding their engagement in the programme.

Despite seemingly positive outcomes, Pleace (2018) cautions that Housing First is not entirely effective for all homeless single people with complex needs every time, citing the figure that between 1 and 3 people (out of every ten) using Housing First services do not have a successful outcome. However, he also stresses that no evidence suggests a clear pattern of failures and the weight of evidence so far suggests good results in terms of tenancy sustainment. Attempts to identify certain individual characteristics that might be considered risk factors for early housing instability in Housing First have shown that it is not possible to predict these with confidence, and given the high proportion of people receiving Housing First who do achieve stable housing in the first year (in this case, of the At Home/Chez-Soi group), studies have suggested there is no evidence for prioritising or targeting Housing First to individuals with certain characteristics but to open the programme to all eligible to apply (Volk et al., 2016).

There is additional caution required when applying positive outcomes from international evidence to a UK context. In addition to variations in social welfare and health care provision structures between the US and other nations, there are a range of differences between countries and their Housing First models that demand a need to exercise caution. The Canadian and French programmes which have undergone robust evaluation are heavily resourced compared to levels of spending on Housing First in the UK to date (Pleace, 2018). The model used in Canada and France differs from that adopted in the UK, replicating the original Pathway Housing First approach, arguably with a higher level of support and larger interdisciplinary staff teams. A more appropriate comparison to make, argues Pleace (2018: 28), is the variation in the other homelessness services that Housing First is being compared to: “in terms of the international evidence base for Housing First, the successes are being measured in relation to existing service models that are not widely used in the UK.”  These are models with strict regimes centred on behavioural modification and abstinence, or with basic homelessness services. The UK homelessness sector does not have these characteristics, as use of service-user choice, personalisation, co-production and harm reduction is widespread. Cognisant of these limitations and differences, evidence regarding the effectiveness of Housing First in improving tenancy sustainment for chronically homeless people in the UK needs really to come from the UK. 

UK evidence

Based on work conducted by the Housing First England team since 2016, there were 32 known active Housing First services in England at the time of their fieldwork (Homeless Link, 2018a).^{[footnote 4]} They found that Housing First services in England are still relatively new (30% had been in service 2-3 years; 26% between 0 and 6 months), and many are small and operate with very few staff members. Evidence on the number of Housing First projects in Wales is weaker but sources estimate 10 Housing First projects as of February 2018 (Barker, 2018; Blood et al., 2018). Scotland was the first nation in the UK to pilot Housing First in 2010 (Johnsen, 2013) and is seen to be outpacing England in the development of schemes according to a major piece of comparative research on Housing First in Europe (Pleace et al., 2019). In Scotland, progress has been made through the development of Scotland’s Housing First Pathfinder launched in April 2019, working across 5 cities with a clear commitment to provide Housing First in each local authority and with the wider goal of acting as a catalyst for the adoption of Housing First across all 32 Scottish local authorities, so it becomes the default solution to tackling the systemic issue of homelessness (Housing First Scotland, 2020). Pathfinder aims to support as many people as possible into their own safe, secure home across local communities by 2021, with housing capacity pledged and funding available for more than 800 tenancies. An annual check-up by Housing First Scotland (2020) shows the Pathfinder is moving positively towards this goal after the first year, with 252 new tenancies started as part of the Pathfinder programme, and 232 people remaining in their new home (a 92% tenancy sustainment rate) by the end of March 2020 (Housing First Scotland, 2020). Scotland’s Housing First pathfinder is currently being independently evaluated by Heriot-Watt University who will review and report on the process adopted, the outcomes achieved, and the cost benefit delivered. Northern Ireland’s Housing First pilot (run by DePaul in Belfast) launched in 2013 (Boyle et al., 2016). DePaul has since expanded their scheme to also include Derry (Fitzpatrick et al., 2020). The most recent Homelessness Monitor report for Northern Ireland (Fitzpatrick et al., 2020) noted that while support for Housing First is strong in Northern Ireland, significant concerns were raised about implementation. Suspension of the Northern Ireland Executive in early 2017, and uncertainties and tensions associated with Brexit, had been proving very disruptive to progress on decision-making and the development of homelessness strategy.  

England in particular has been slower to develop Housing First than some other European countries (Pleace et al., 2019). Drawing on findings from a major review of international evidence on what works in ending street homelessness, Mackie et al. (2019) pose the question, ‘if we know what works, why don’t we do it?’, reflecting on why those working to end rough sleeping do not always do what works. Among the potential barriers identified is a lack of affordable and suitable settled accommodation;^{[footnote 5]} funding^{[footnote 6]} (Housing First requires high upfront costs although is likely to make long-term savings); a lack of research on how well Housing First works with sub-groups other than those with complex needs; and political will/underpinning political ideologies^{[footnote 7]} across all levels of government. Nevertheless, recent years have witnessed a gathering momentum and growing cross-party interest in Housing First as a potential policy solution for so-called ‘entrenched’ rough sleepers across the 4 nations. The 2017 Conservative Government pledged in their manifesto to “halve rough sleeping over the course of the parliament and eliminate it altogether by 2027” by setting up a new homelessness reduction taskforce to focus on prevention and affordable housing and by piloting a Housing First approach to tackling rough sleeping (Wilson & Barton, 2020). Their 2019 manifesto further committed to ending rough sleeping through an extension of the Rough Sleeping Initiative, Housing First, in which the government pledged to invest £28 million in 3 Housing First pilots in Greater Manchester, Liverpool City Region and the West Midlands to support those sleeping rough with the most complex needs and using local services to meet the health and housing needs of people living on the streets^{[footnote 8]}. The pilots are testing the effectiveness of Housing First in the English context. They are currently subject to an ongoing impact evaluation (2018 - 2022) commissioned by the Ministry of Housing, Communities and Local Government and conducted by a research consortium including ICF International, Byson Purdon Social Research, Heriot-Watt University and Homeless Link. The rest of this section considers evidence from the UK on Housing First schemes which have been developed and evaluated so far.

Commissioned by Crisis, a study by Blood et al. (2017) which tested the feasibility of implementing Housing First at scale within the Liverpool City Region (LCR)^{[footnote 9]} contains some important learning. The LCR Housing First model would be based on the Housing First principles set out in Homeless Link (2017) and in Section 2.2 of this report. Key to the success of Housing First in the LCR, Blood et al. argue, is its integration into a wider effective and joined-up homelessness strategy. To prevent a Housing First service from becoming overloaded it is recommended that there needs to be investment in prevention activity as well as consistent, clear pathways and good communication between the Housing First service and the rest of the housing/homelessness system. Using Mainstay data, the study identified a cohort of individuals in LCR who met the criteria for the Housing First service and had presented 4 times or more for a Mainstay assessment. Defining a ‘successful outcome’ as someone supported into and sustaining an independent tenancy for 6 months, they were able to compare the likelihood of a person achieving this outcome within current services and within a Housing First model. The analysis suggested that the Housing First model could be around 3 to 5 times more cost effective than current provision for this cohort. A number of factors are deemed as key to the potential success of the LCR Housing First model: its ability to recruit, retain and manage effectively a small and consistent team of workers with excellent engagement skills and who are willing to work to the Housing First principles, which they anticipate will require careful recruitment of people working in other sectors and those with lived experience (the quality of relationships is how they distinguish it from other models); finding the right level of the duration and intensity of support, holding on to clients when required and letting them go when they are ready; being able to provide efficient access to suitable housing (initial engagement with local Registered Providers and other stakeholders suggests there might be an appetite to support the Housing First model); a flexible menu of housing options, which might include individual flat or houses in different areas across LCR, Housing First in mainstream sheltered tenancies for older homeless people with complex needs, shared tenancies and accessible properties; access to health and social care including mental health professionals; and effective referral routes and assessment to Housing First from outreach and Housing Options teams as well as criminal justice and health agencies.

Modelling the use of Housing First helps to understand how services might be implemented, but to fully examine what Housing First as a model might achieve requires evaluation of services already in place alongside comparative analysis. Bretherton and Pleace (2015), in their evaluation of 9 Housing First projects undertaken between July 2014 and January 2015, offer insight into the cost effectiveness of Housing First as an alternative to accommodation-based services and low-intensity floating support in reducing long-term and recurrent homelessness in England. The Housing First service users included many high need people with experience of homelessness, including long-term rough sleepers identified by the CHAIN database system in London. The results are, however, based on short-term research which the authors acknowledge had some limitations. It was not, for instance, possible for them to explore long-term outcomes because even the longest running service had only been fully operational for a few years. Limited resources for the research also meant that each Housing First project could only be visited once. The services evaluated were reported to be adhering to the core philosophy of Housing First, with the exception of one, which was nevertheless very influenced by Housing First in how it operated. While some of the services had not yet been operational for one year, the report shows evidence of housing sustainment. Four people in a council tenancy had been with a Housing First service for an average of 17 months (median 14 months), another 23 in private rented sector housing had been with a Housing First service an average of 9 months (median 8 months) and 13 people in housing association tenancies had been with Housing First for an average of 29 months (median 31 months). When interviewed about what they thought made the Housing First approach effective in terms of tenancy sustainment, service users said they valued the freedom, choice and security of having their own home as well as the open-ended, intensive and flexible support offered. The report calls for wider use of Housing First at a strategic level in England but notes that current commissioning practice does not provide the consistency and duration of funding needed to allow the open-ended support model of Housing First to be sustained.

Adding to the small but growing evidence base that is interrogating Housing First models in an English context, Jones et al.’s (2019) research explores the impact of 2 Housing First services run by St. Mungo’s, in Brighton and Hove, and Westminster. The 2 services were selected from the 12 Housing First projects currently run by St. Mungo’s because one (Brighton and Hove) was well established,^{[footnote 10]} whereas Westminster was a very new service.^{[footnote 11]} The projects sought to target their support at both those people who had long and chaotic housing histories and those who could be reasonably described as multiply excluded homeless. Following longitudinal interviews with service users, consultations with staff and wider stakeholders, and analysis of project monitoring data, they found that both services appeared to be having a positive impact on service users, especially in terms of housing retention. Of the 8 people with multiple and complex needs originally selected for the Brighton and Hove pilot in 2014, 7 of the cohort have remained with the service. In Westminster, by January 2018, 7 people (out of 10) remained in their properties. When interviewed about their perceptions of the Brighton and Hove Housing First service, service users explained that the flexibility associated with their Housing First tenancy, as opposed to a hostel place, was important. All participants reported staying in their flats most nights, if not every night. However, where they had previously experienced eviction due to spending time away from their property (for one participant this was due to a hospital visit), they now felt reassured that this would not happen with the Housing First service. For one participant, the privacy and freedom that was afforded by their own flat was a key benefit to the Housing First model for them. Others cited being able to cook for themselves and being away from people who are drinking and taking drugs. In Westminster, this was reported to be the first time that all participants had had their own tenancy, and this was cited as being the key positive impact of Housing First. All clients were relieved to be away from the hostel system, which they had experienced as unsafe.   

Results of research focused on a small pilot service - Camden Housing First - also appeared to deliver positive outcomes in terms of housing sustainment (Pleace & Bretherton, 2013b). Camden Housing First used scattered, ordinary private rented housing and a mobile team of 2 specialist support workers providing an intensive case management (ICM) service. It was targeted specifically on people whom the Camden Hostels Pathway had assessed as unable to live independently, as not ‘housing ready’, for a period of at least 3 years, and who had never moved into their own housing. Service users had extremely high rates of problematic drug and alcohol use and mental health problems. There was clear evidence that Housing First in Camden was delivering housing sustainment for chronically homeless people who had never lived independently before or had not done so for many years. The scheme had successfully housed 7 of the 13 people using the service and was working to house the others at the point the research concluded in May 2013. One tenancy had failed after 6 months. It successfully used ordinary private rented housing which was secured solely through developing working relationships with estate/letting agents. Service users praised the extent and quality of the support they had received from the Housing First team, and those who had been rehoused at the point of interview reported positive feelings about their homes and valued having their own space. In this study, too, participants made positive comments about being removed from the proximity of drugs and alcohol that was a feature of life in previous hostel accommodation.

Recent research has shown there is potential for Housing First to reduce homelessness among people with high and complex needs across Greater Manchester (Quilgars & Pleace, 2018). The Housing First pilot worked with people with high and complex needs, with a history of long-term and/or recurrent homelessness. Evaluating the Inspiring Change Manchester (ICM) Housing First pilot between April 2016 and March 2018, Quilgars and Pleace (2018) found a 100% tenancy sustainment rate when the pilot ended, with all 18 people who had been found independent housing having maintained their tenancy. The majority of people (11 people) had been in their tenancy for at least 12 months by March 2018. A further 4 people had been in their housing for between 6 and 12 months, another 3 for between 1 to 6 months, and one person for less than 1 month. Service users and stakeholders interviewed attributed this to the significance of finding a stable base. This success did not come without challenges, however, as the evaluation revealed difficulties in finding suitable accommodation for people. The service found getting housing providers signed up and on board with the concept of Housing First difficult. There was also considerable work required in persuading housing providers, particularly private landlords, that people with complex needs would be able to retain a tenancy. However, overall conclusions were positive, with the authors stating that “this evaluation indicates that Housing First can reduce long-term and repeat homelessness associated with high and complex support needs and help people on a recovery journey” (Quilgars & Pleace, 2018, p. 8).

A recent report exploring the cost effectiveness of Housing First in England^{[footnote 12]} estimated that Housing First costs an average of approximately £4,128 in total support costs per-person during the first year of use (including staffing, administration and any personal budgets provided); and the total average costs, including housing, for the first year of use are £9,492 per person if an average social housing rent is being paid and £9,222, if an average rent in the cheapest third of the private rented sector is being paid (Pleace & Bretherton, 2019). It was estimated that 3 years of support from Housing First costs less, on average, than one year in fixed-site services. Evidence - albeit limited - is also beginning to emerge from England which suggests cost reductions for local services as a result of Housing First, due to clients’ better engagement with the health service, and criminal justice systems (Jones et al., 2019). However, it was acknowledged that such savings were difficult to assess over a short timeframe.

Housing First services have also been shown to work in Scotland and Northern Ireland. The evaluation of Turning Point Scotland’s (TPS) Housing First project (Johnsen, 2013)^{[footnote 13]} - the first Housing First project to be developed in the UK (in Glasgow) - shows similar positive outcomes in terms of tenancy sustainment. TPS accommodated service users in independent self-contained housing association flats, on a scatter-site basis, with a rent contract and unlimited lease. The project was staffed by a team of 6 which included 3 peer support workers with histories of homelessness and substance misuse. Support plans were developed on a client-centred basis and assertive outreach and motivational techniques were employed. Staff members assisted service users to access welfare entitlements and other support services. The vast majority of service users retained their tenancies continuously since first allocated the property; half of these individuals had done so for more than 2 years by the end of the pilot period. No evictions were recorded, although a service user ‘lost’ their tenancy due to serving a long prison sentence and another ‘gave up’ theirs after being victimised by other members of the drug-using community. Levels of service user satisfaction with the project were reported as being very high, which was attributed to positive relationships with frontline staff who were seen as non-judgemental, ‘easy to talk to’ and trustworthy. The inclusion of peer support workers was also regarded as a key strength of the project which enhanced service users’ motivation towards recovery. Service users appreciated the flexibility of the support, knowing that they could ask for more or less frequent meetings if their circumstances changed. The evaluation suggests that the Housing First approach can be effective when implemented in the UK.

An evaluation of Northern Ireland’s Housing First pilot (delivered by DePaul in Belfast) found that of the 24 people who were found permanent accommodation in 2014 (the year of the evaluation), 19 (79%) maintained their tenancy for a significant period of time (Boyle et al., 2016). In many cases, their acceptance into the Housing First service was the first time that service users had ever been given an opportunity to sustain a tenancy or been able to sustain a tenancy for any length of time. In some cases, their difficulties went back a number of decades. Service users said in their interviews with the research team that they had been told many times in the past that they would not be capable of obtaining or sustaining a tenancy. They pointed to their background as being an insurmountable hurdle in finding and sustaining accommodation. The type and location of accommodation coupled with the support and brokerage of other services that service users received from the Housing First staff team were cited as the key to tenancy sustainment. The ongoing support to maintain their tenancy was viewed as critical by service users, with Housing First staff as being central to that support. Reference was made to the approach of the Housing First staff, their ethos and their care. A number of service users also pointed to the fact that the harm reduction approach had helped them to reduce their drinking and retain their tenancy.

The results of the above evaluations are cited with the caveat that the majority focus on relatively small pilot Housing First services designed for a specific group of homeless people. The research generally covers short periods of time which means it was not possible to explore long-term outcomes. In some evaluations, a tenancy is sometimes judged to have been sustained, or a positive housing outcome achieved if a tenant has remained in their accommodation for just 6 months. For these reasons, while there is just enough evidence to make a case for the expansion of the individual services, it is still difficult to generalise the results to other groups of people experiencing homelessness (with less severe needs, for instance). It is also worth noting that research rarely compares Housing First outcomes with other forms of tenancy sustainment support (but see Blood et al. (2017) for an exception). High tenancy sustainment rates in Housing First projects are often taken as indication that Housing First is more effective than other models of support, but this has not been conclusively shown (Pleace, 2011). Comparing the results of evaluations of other supported housing models with those of Housing First might be one way of ascertaining the relative merit of different approaches but these are rarely directly comparable. For example, a 3-year evaluation of a ‘homeless to home’ project helping formerly homeless families sustain independent tenancies found similar outcomes to Housing First evaluations, with around 88% still in their tenancy at the end of the programme. However, these were families, while the majority of Housing First beneficiaries are single people, and many will have had lower support needs and so the results are not directly comparable.

Who does Housing First work for in the UK?

Perhaps as expected, a high proportion of Housing First services in England say that their clients experience multiple disadvantage (Homeless Link, 2018a), with 100% of services surveyed^{[footnote 14]} reporting their clients have a mental health issue, either diagnosed or undiagnosed, on entry to the service; 96% of services reporting their clients have a substance misuse issue on entry to the service; 88% of services reporting their clients have been in contact with the criminal justice system; and 75% of services reporting their clients have experienced violence or abuse.

Evidence suggests that Housing First is not always suitable for everyone. DePaul’s experience of running the Housing First service in Belfast (Boyle et al., 2016) demonstrated that, while a majority of people gained significant benefits from the service, it was not appropriate for everyone. A minority of service users said that they would have preferred to live with other people, and loneliness was cited as a very significant factor here. Developing more integrated data linking systems may help here, by tracking people with homelessness histories across a range of services and identifying suitable Housing First candidates based on their support needs (Crisis, 2018).

Users of Housing First are typically middle-aged males (Bretherton & Pleace, 2015; Johnsen, 2013). In the Housing First services evaluated by Bretherton and Pleace (2015), women represented just over a quarter of service users. With growing evidence showing that women’s experience of homelessness often differs from that of men (Reeve et al., 2006), the suitability of Housing First for women, including why women are not more strongly represented among Housing First service users, should be further investigated. The suitability of Housing First for other groups of homeless people, such as young people or those with specific experiences, such as repeated offending and imprisonment, could also be subject of more exploration (Mackie et al., 2019). A similar point was recommended in Woodhall-Melnik and Dunne’s (2016, p. 295) systematic review; that is, a need to better understand the impacts of Housing First for specific sub-populations within the broader homeless population, as some of the international evidence they reviewed cautioned against the use of Housing First with particular sub-groups (veterans with severe alcohol dependency). They caution that, “if this model [Housing First] is to be extended beyond this originally intended population, research should focus on establishing best practices for altering or employing models that meet the needs of specific sub-populations within their local contexts”. There are, however, an increasing number of exceptions as recent studies and projects are beginning to consider the impacts of Housing First for particular sub-groups.

The first significant attempt to develop a specialist form of Housing First was made by Threshold in 2015. Set up as a pilot service for 2 years in 2015, the Threshold Housing First service was designed to support up to 12 women with high and complex needs, with experience of homelessness, and a history of offending. An independent evaluation of Threshold Housing First’s first 2 years^{[footnote 15]} highlighted positive outcomes in terms of tenancy sustainment (Quilgars & Pleace, 2017). By June 2017, 18 of the 33 women using Housing First had been found tenancies by the service. Twelve tenancies were housing association properties; 6 were in the private rented sector. A further 2 women had been referred to the service already living in tenancies (one social; one private rented sector), but where their tenancy was at risk. Of the 20 women in tenancies, 16 women were still in tenancies (including 5 planned moves within this) at June 2017, representing an 80% tenancy sustainment rate. Service users distinguished this service from those they had been involved with in the past, speaking highly of the choice and control it gave them in comparison to other services which were sometimes overly directive; women also respected the team’s empathetic and non-judgemental approach, and were happy with the level of support - which they felt was higher than other services they had received support from in the past. Despite these successes, interviews with staff delivering the service highlighted a number of challenges: delays in identifying housing were a barrier to continuing engagement in a number of cases. However, by mid-2017, the evaluation notes, the service had established relationships with 4 housing providers, a number of private landlords, and every landlord in one area - and finding appropriate accommodation was becoming less problematic. The service has since received a further 3 years funding and is being delivered in 3 local authority areas in Greater Manchester: Tameside, Stockport and Oldham.

Basis runs the only Housing First project specifically for sex working women in the UK. They began their Leeds-based Housing First pilot in November 2016, funded by the Big Lottery and WY-FI’s Innovation Fund to relieve homelessness, alcohol and drug use, reoffending, and mental health issues in West Yorkshire. The pilot funded 6 Housing First tenancies for 12 months, along with a dedicated caseworker from Basis and a housing support worker from Foundation. Working across Leeds, Basis identified a number of gender-specific elements to housing issues in the city which provided significant justifications for the Housing First project (Bimpson, 2018). The only emergency accommodation available, at the time of the report, for women in Leeds was the mixed gender hostel at St George’s Crypt, 8 women-only hostel beds at Francis House and a small amount of Private Rented Sector (PRS) accommodation arranged by the council with floating support. The success of the project is heralded in a recent evaluation, informed by service user questionnaires, interviews and regular discussions with Basis staff (Bimpson, 2018). A combination of assertive engagement, coordination and regular communication between Basis and Foundation ensured that the 6 Housing First tenancies were sustained over the pilot period, with one woman deciding that she did not need such an intensively supported housing programme and another absconding shortly after moving in from prison. The security associated with settled accommodation and continuous and unconditional support was key to tenancy sustainment and it was reported that in many cases mitigated a number of situations which could have resulted in tenancy breakdown.  The project’s flexible approach also added to its success around tenancy sustainment. This was demonstrated when one woman went into rehabilitation shortly after starting her Housing First tenancy. The Housing First place was left open for her and she returned to it after 5 months of rehabilitation. It is also cited as critical that tenancy management issues did not affect the support offered to the women. The report gives the example of one participant who had caused some minor damage in her flat as a result of a difficult period of mental health. Rather than issuing warning letters and reprisals, Foundation’s housing support worker provided reassurance and worked with the client to rectify the damage while support continued.

The previous 2 evaluations clearly demonstrate that the principles of Housing First work well for women. The Threshold Housing First service evaluation found that women valued the relationships of trust formed with workers, the choice and control, and the reliability of having one worker providing long term, high intensity support (Quilgars & Pleace, 2017). The evaluation also highlighted distinct gender-specific support needs, most specifically the prevalence of domestic/gender-based violence and abuse, as well as the important role the service plays in supporting women to manage abusive relationships and stay safe, this being key to ending their homelessness. Developing gender-specific services that sensitively manage gendered issues such as domestic abuse should be recognised as a key operational consideration for Housing First services aiming to end long-term and recurrent homelessness for women.

A promising innovation in the UK is the Housing First for Youth scheme currently being delivered in Scotland by Rock Trust in partnership with Almond Housing Association in West Lothian for young people leaving care.^{[footnote 16]} The Housing First for Youth programme extends the original Housing First model to young people, acknowledging that young people have specific needs related to their age (many young people leaving care will have experienced significant trauma, instability, multiple moves and a lack of security). Their project offers young people accommodation on an immediate, permanent and unconditional basis, with holistic, personalised support from the team in West Lothian. Five fully furnished tenancies have been provided so far. While there is no current evaluation data on the outcomes of the project, the Rock Trust anticipate being able to build the evidence needed to expand this service to more young people over time.^{[footnote 17]}

2.4  Summary

The ultimate test of any homelessness service is whether or not it ends homelessness. Housing First does end homelessness and does so in a relatively cost-effective way. However, it is the human benefits of Housing First, the ways in which it can positively change the lives of people who would otherwise be caught in long-term and repeated homelessness, that are the real measure of its value. (Pleace & Bretherton, 2019, p. 7)

A number of Housing First schemes have been established in the UK - including specialist services targeted at specific sub-populations - which have been (or are due to be) subject to independent evaluation. The evidence base on Housing First in the UK - while still more limited than robust international studies - is growing, and existing evidence clearly demonstrates impressively high housing retention outcomes (consistent with those in countries where Housing First is more established). Studies have also started to emerge which show the considerable cost savings of Housing First when compared to fixed site services (Pleace & Bretherton, 2019).

However, while some still see barriers and hold a degree of hesitancy regarding mass roll-out of Housing First at scale,^{[footnote 18]} there remains a pressing need for further testing in the UK context to add more weight to the evidence base. The majority of UK evaluations either focus on individual projects (Boyle et al., 2016; Johnsen, 2013) and/or are short-term (with the exception of Johnsen, 2013). As existing Housing First projects move forward, then, there is an obvious need for more longitudinal evaluations which track long-term outcomes of clients.

Compared to the international evidence base, which commonly uses randomised control trials, the quality of evidence in the UK is also more limited. This has been identified as a barrier for investment by health and social care commissioners who felt the quality of evidence still did not meet the ‘clinical standard’ which would be required by most health commissioners (Homeless Link, 2015). This calls for more extensive, methodologically robust studies (randomised control trials, for example) to demonstrate the effectiveness of Housing First as a mechanism for ending homelessness. Furthermore, Housing First as a solution to homelessness for specific sub-populations has only recently started to be developed in the UK hence evidence on how Housing First works for these groups is still in its infancy, and owing to the often low numbers of clients in Housing First pilots (and subsequently, studies), it is difficult to draw definitive conclusions about which clients experience the most success from Housing First. A better understanding of programme outcomes for specific sub-populations may prove useful in understanding the outcomes of widespread implementation of Housing First. 

This review has explored both UK and international evidence on the effectiveness of Housing First in relation to tenancy sustainment outcomes, the majority of which predominantly rely on quantitative data on retention outcomes. While the studies reviewed still employed mixed methods and reported on findings from qualitative work with service users, these were usually presented as ‘strengths’ and ‘weaknesses’ of the project, rather than in-depth exploration of why service users were able to retain their tenancies effectively. This calls for the commissioning of more evaluations and research using mixed methods. While not underestimating the value of statistics as a means to demonstrate success, as Pleace and Bretherton (2019) state above, it is also the human benefits of Housing First - the qualitative ways in which it can positively change the lives of those who have been homeless for a long time - that warrant consideration.

The Housing First literature reviewed provides relatively robust evidence that a model providing rapid rehousing with intensive support produces positive tenancy sustainment outcomes for homeless people with complex needs, at least in the short-medium term. This broad conclusion is useful. However, the Housing First evidence does not provide definitive conclusions about why such a model works so well and for whom - about the specifics of those factors supporting or undermining tenancy sustainment. In the remaining chapters we therefore look beyond the Housing First literature in an effort to pinpoint the specific drivers for, and barriers to tenancy sustainment for those with complex needs.

3. Tenancy sustainment drivers and barriers

3.1  Introduction

This chapter reviews literature that provides evidence on the drivers for, and barriers to tenancy sustainment amongst homeless people with complex needs. It addresses Research Questions 1 and 2: What are the drivers and barriers of tenancy sustainment for those with precarious housing histories and complex needs? And what factors are important to tenancy sustainment?

Research evidence draws attention to the fact that successfully ending homelessness is a process, and one which involves addressing multiple, overlapping issues (Chamberlain & Johnson, 2018), often relating to the complex reasons for homelessness, which must be understood for resettlement to success.  Focusing on Scotland, comprehensive research undertaken by Bramley et al. (2019), for example, demonstrated that men and women who have experienced severe multiple disadvantage (SMD) often had traumatic childhoods and lived ‘alongside’ fear of, or actual, violence.  The authors used twelve datasets for SMD statistical profiling purposes.^{[footnote 19]} The statistical data, notably those data sets with a longitudinal element, showed that causation also runs the other way - from poverty to SMD via poverty, debt and educational failure.  Equally, drug or alcohol dependencies were found to contribute to worklessness and exacerbate poverty.  Family breakdown (often involving violence or abuse) had similar effects.

Reflecting the complexity of pathways into homelessness and the multi-directional influence of challenging life experiences, research evidence is explicit that promoting tenancy sustainment (or successful resettlement) for homeless people with complex needs requires attention across multiple domains.  Thus, although this chapter seeks to pinpoint those factors found to be most important to tenancy sustainment, they will always be just one factor amongst a complex set of overlapping issues.

On the basis of interviews with 64 people in 2 Housing First Pilots over a 2 year period, Chamberlain and Johnson (2018), for example, suggested that the process of resettlement into sustained tenancies involved transitioning materially (meeting housing needs), relationally (rebuilding relationships and connections) and psychologically (overcoming the stigma of homelessness).^{[footnote 20]} Although researching in Australia, this is a useful framework because it highlights that support and assistance across multiple dimensions may be necessary to promote tenancy sustainment. Hennessey and Grant (2006), researching different models of housing support in Merseyside, also identified psychological, relational, and material factors as critical to tenancy sustainment (although do not articulate them quite in this way) amongst the 25 previously homeless people they interviewed,^{[footnote 21]} albeit different factors to those suggested by Chamberlain and Johnson (2018). These were: motivation to resettle and engage with support; a positive relationship between client and support worker, based on mutual respect, and flexible support provision. They found that when these factors coincide, ‘successful resettlement’ (not defined) is most likely. This is one of the few sources that proposes an evidence-based model of the most effective housing support to ensure tenancy sustainment (or successful resettlement, in their terms) for homeless people.  However, detail on methodology is very limited, making it difficult to adequately assess the robustness of their conclusions, or assess how context specific their findings might be. A rapid evidence assessment by the Social Care Institute for Excellence also presented similar findings in relation to motivation, flexibility and relational factors when working to support people who are homeless and have complex needs (Sheikh & Teeman, 2018), including, but not exclusively in relation to tenancy sustainment.^{[footnote 22]} Fidelity to service models while allowing for flexibility, client-case management worker relationships and one to one support (including peer support), effective multi-agency working, landlord relationships, and providing the right service at the right time were the most significant factors to emerge from the 35 sources they reviewed. While individual motivation was identified as important, this report highlighted the need for services to adapt to a person’s needs at the right time, rather than relying on the individual’s readiness to change.

Chamberlain and Johnson pointed out that the principles of Housing First broadly attend to dimensions beyond the material, recognising the importance of social and community integration and inclusion to resolving homelessness. However, they cautioned that these important aspects of rehousing/resettlement/exiting homelessness were rarely foregrounded in Housing First evaluations, which focus more on measuring and reporting housing (material) outcomes.  The danger here, they argue, is that important relational and psychological factors, potentially undermining longer-term tenancy sustainment, may be missed. The review of Housing First evidence for this REA (see Chapter 2) would broadly concur with this conclusion.  Some evaluations use mixed methods and report qualitative insights on the strengths and weaknesses of Housing First from clients’ perspectives but generally prioritise, and measure, material rather than psycho-social processes and outcomes.  The evidence base offers insights on 3 different types of drivers and barriers, around which the remainder of this chapter is structured:

1. Broad structural issues
2. Characteristics of people and properties
3. The role of support services and organisations

3.2  Broad structural issues

This REA is primarily concerned with identifying those factors supporting or undermining tenancy sustainment amongst people with complex needs. As such, we are interested in the question of how people with complex needs can be best supported to sustain the accommodation they are living in, or under what conditions are they most likely to do so. However, some evidence - particularly that focused on ‘resettlement’ - makes the point that access to housing is a prerequisite for a sustainable tenancy and highlights contextual challenges of housing supply, availability and accessibility that need consideration in efforts to promote tenancy sustainment. One evaluation of a programme designed to create sustainable tenancies in the private rented sector, for example reported that:

Tenancy creation was proving to be more difficult than tenancy sustainment and reflected the impact of a broad range of changes, including a tighter housing market, competition from local authorities for suitable affordable property, and an altogether less ‘benign’ context for access work. For example, local authorities were less likely to be co-operative. (Rugg, 2014, p. vi)

This does not represent a significant body of evidence capable of generating conclusive lessons about the factors supporting or undermining tenancy sustainment for people with complex needs. Nor is it a discrete body of evidence in the sense of including studies explicitly aiming to identify the structural barriers to tenancy creation (but see Rowe & Wagstaff, 2017).  As such, the insights provided often lack detail or are tangential to the key focus of the study, and the conclusions are rarely drawn from robust measurement or comparison (for example identifying the relative effect of different factors on access to housing). Most of these studies were also conducted by homelessness and other support organisations (but see Fitzpatrick et al, 2019 and Preece 2020). Only those considered sufficiently relevant and robust were included in this REA, but it is worth noting that some of these sources may have also been aligned with particular campaign agendas of those organisations.

Nevertheless, UK based evidence focused on ‘move-on’ (settled) accommodation for people currently living in temporary housing, often identifies access to housing as a key challenge and this is relevant to questions about tenancy sustainment. In particular sources point to limited supply of affordable housing generally for homeless and other low-income households, and problems of access specifically for homeless people with complex needs, as key barriers to resettlement.  They also highlight the way in which the operation of broad ‘systems’ - such as state welfare - can have an impact, albeit unintentionally, on tenancy sustainment.

The following issues are noted in these sources as creating barriers to housing access.

Affordable housing supply

Limited supply of affordable accommodation is frequently cited in the evidence base as a key barrier to ‘resettlement’ (Rowe & Wagstaff, 2017; Homeless Link, 2018b), usually based on reported experiences of organisations providing temporary housing and/or support to homeless people.  This evidence is typically concerned with re-housing options for homeless people, usually ‘single’^{[footnote 23]} homeless people, but as a cohort within the homeless population, it may apply equally to those with complex needs who are homeless. For example, homelessness services participating in a Homeless Link study emphasised that the lack of reasonable and affordable housing had hindered their ability to move people on from their services over the past year (Homeless Link, 2018b). The review notes that, “a large percentage of respondents identified the lack of affordable housing as both a contributory barrier (73%) and the primary barrier (40%) that prevents people from moving on from their services” (Homeless Link, 2018b, p. 37).  Appropriate and affordable accommodation is thus 1 of 4 key aspects of successful move-ons identified in this study (Homeless Link, 2018b).   

More specifically, Holly (2017) noted in relation to women with complex needs that:

…. the lack of social housing stock, which is particularly chronic in London, as well as wider housing affordability issues and benefit reductions, has also been highlighted as a cause of the increased difficulties that women leaving refuge accommodation – many of whom have additional support needs relating to mental distress and substance use – face in accessing secure accommodation. (Holly, 2017, p. 40)

They reference a study by Solace Women’s Aid conducted in 2015/16 that found that only 38% of women who had a secure tenancy on fleeing domestic abuse received an equivalent council or housing association tenure on leaving the refuge and 62% faced long-term housing insecurity by being placed in temporary accommodation, a hostel or staying with family and friends.

Access to housing

In addition to general affordable housing supply issues, a few sources within this body of evidence conclude that social and private landlords are reluctant to house homeless people with complex needs, further restricting the supply of accommodation to this group. Over the last decade, there has been increasing reliance on the private rented sector to meet housing need and for many ‘single’ homeless people, the private rented sector is their only option (Reeve et al., 2016). However, a survey of 949 private landlords in England and Scotland in 2016 found that only 14% were willing to let to homeless people on the open market, with an additional 6% willing to let through a private sector leasing scheme (Reeve et al., 2016). We might speculate that those willing to let to homeless people known to have complex needs would be lower still, given that perception of risk (arrears, property damage and needing to manage the tenancy more intensively) was a key driver of landlord lettings practices in this study. The stigma attached to people with complex needs was also highlighted by MEAM and Lankelly Chase, who argued that local authorities have a bigger role to play in mitigating the exclusion faced by people in these groups (MEAM, 2015).  While there is some research highlighting the perceptions of private landlords about letting to people in receipt of housing benefit, the lettings practices of private landlords in relation to people with complex needs remains a gap in evidence. 

The social rented sector has historically provided housing for those on low incomes and in the most need but there is recent evidence of increasing reluctance of social housing landlords to house tenants with complex needs, due to exclusion on grounds of unacceptable behaviour, such as ASB and histories of offending or substance misuse (Fitzpatrick et al., 2019; Preece et al., 2020; Rowe & Wagstaff, 2017), and as a result of poor availability of floating support services (St Mungo’s, 2018, p. 14). A National Housing Federation survey, for example, asked housing associations about the barriers to providing housing for homeless households, and while lack of stock was most frequently cited, 30% cited a higher risk of arrears and 28% cited higher costs of providing support or services to formerly homeless tenants (St Mungo’s, 2018, p. 15). Recent research with social landlords in England, similarly, found increasing reluctance to accept people with complex support needs, in part due to restricted adult social care provision (Preece et al. 2020). The Prison Reform Trust makes this point in relation to people leaving prison, who are among the least prioritised for social housing. With reoffending rates within a year of release at 48% for adults in the UK, finding appropriate housing is a crucial part of the resettlement picture for this particular group (The Prison Reform Trust, 2019, p. 3).  Thus, regardless of housing supply, people with complex needs may struggle to access the accommodation that is available.

The welfare system

A few studies indicate that issues associated with the benefit system may contribute to, or directly cause, tenancy failure amongst homeless people. This is not a large literature, but it does comprise one of the few longitudinal studies of resettlement amongst single homeless people (Crane et al., 2011: and 2016).

Rent arrears threaten tenancy sustainment because in both the PRS and SRS landlords can evict tenants in arrears. In Crane et al.’s (2011; and 2016) significant longitudinal research of the resettlement of single homeless people over a 6-year period^{[footnote 24]} rent arrears that accrued in the early months of resettlement often stemmed from Housing Benefit administration problems. Subsequently, issues with benefits, including sanctions, changing circumstances such as moving in and out of work, and related suspension/ interruption of HB all contributed to arrears. In a related point about the role of the social security system, one study based on workshops with front-line workers identified Local Housing Allowance^{[footnote 25]} rates as a barrier to tenancy sustainment for people with complex needs. They suggested that the only accommodation affordable to people in receipt of Local Housing Allowance was located in areas where homeless people with additional needs were least likely to sustain a tenancy because of environmental conditions:   

You can’t rent a house for £280 per month, yet they’re saying to homeless people who are addicted with mental health problems go and live somewhere shit in the shittiest possible area because that’s all the money we’re going to be able to give you, so that repeats the cycle again, and you access services forever and ever and ever. (MEAM, 2015, p. 26)

The suitability of accommodation

Inappropriate referrals and move-on accommodation are identified in a few sources as having acted as a barrier to tenancy sustainment. This issue is raised mainly in studies that explore or evaluate the experiences of people in move-on accommodation, and the service providers who work with them.  The small number of sources that provide insights here suggest that appropriate placement into temporary accommodation and then moving people on at the right time and into the right tenancy is important if a tenancy is to be sustained. If these conditions are not met, it is suggested, tenancies are more likely to fail. Sources cite pressures on local authorities and homelessness providers as reasons why people with complex needs are sometimes referred to temporary accommodation without sufficient support or resettled too soon (see issues of limited supply of affordable accommodation above).  A Mental Health Foundation (2016) report, for example, concluded – in relation to a low-level ‘step-down’ accommodation comprising 23 units and floating support for individuals with mental health problems – that financial pressures can lead to inappropriate referrals into the service from the local authority of people with complex needs who need a higher level of support than they can provide. This in turn was reported to contribute to tenancy failure (p. 31). In relation to move-on tenancies, a small scale qualitative study in one LA area in England highlighted how some people with complex needs can require very long-term supported housing, yet are housed in temporary accommodation projects where the service-providers are under pressure to resettle residents within 2 years (Harding and Willett, 2008, p. 446). In this study, moving residents into independent accommodation too quickly was described by stakeholder respondents as a risk – “that they’re not going to manage and they’re going to end up back at the beginning again” or “back were they started” (p. 438).  

3.3  Individual characteristics

There is very little evidence capable of clearly isolating characteristics statistically correlated with tenancy sustainment or failure in the UK. The methodological approach commonly taken to studies of resettlement or tenancy sustainment programmes in the UK precludes this kind of quantitative analysis, and/or studies are too small for robust sub-group analysis. A 2014 evaluation of a programme designed to create sustainable tenancies in the private rented sector, for example, reported that, “The number of funded schemes was too small to be able to draw any robust conclusions on a tendency to succeed or fail in meeting the tenancy or sustainment targets” (Rugg, 2014, p. vi).

We saw in Chapter 2 that some international evaluations of Housing First programmes include comparator samples and randomised control trials (Cherner et al., 2017; Estecahandy et al., 2018; Goering et al., 2014; Volk et al., 2016) but that even using these techniques it has not been possible to predict with confidence the individual characteristics associated with tenancy sustainment or failure. This is partly because of the very high success rates achieved by Housing First projects but it was also noted in Chapter 2 that Housing First clients are usually middle aged and male, with multiple and complex needs (Bretherton & Pleace, 2015; Johnsen, 2013). Until there is greater diversity in the beneficiaries of Housing First in the UK, it will be difficult to compare experiences and outcomes across sub-groups. Qualitative studies provide important insights, and we draw on those here, but they are limited, and cannot determine with confidence who will successfully resettle and under what circumstances.

In a UK context, Crane et al.’s (2011; and 2016) longitudinal resettlement study is a valuable source. Although the sample comprised ‘single homeless people’ rather than ‘homeless people with complex needs’ the sample profile indicates that at least some participants had complex needs.  They reported that: “The respondents’ histories were characterised by high rates of personal problems and disadvantages”.  With a large sample (400 interviewed over 18 month following resettlement and 297 of these re-interviewed 6 years later) they were able to report outcomes by sub-group and found that certain individual characteristics appeared to have greater influence on housing outcomes after initial resettlement. Those in the 2011 study most likely to have become homeless again when their resettlement tenancy ended were:

• the youngest and the oldest (17-19 and 40+). Pawson and Munro’s (2010) statistical analysis of early social housing tenancy terminations in Glasgow also found that young (single) people were at greater risk than other age groups of leaving a tenancy within a year, although these were not specifically formerly homeless or with complex needs. However, the older tenants in their sample were the least likely to leave a tenancy early.
• people with drug dependencies: drug use was linked to issues sustaining a tenancy with 83% remaining housed at 15/18 months compared to 91% of non-users, a variation found to be statistically significant. Use of illegal drugs at the time of resettlement was also identified as a factor associated with tenancy failure in their 2016 follow-up study although not so statistically significantly (Crane et al., 2016).

Crane et al.’s (2016) follow up study identified the following factors as those statistically correlated with tenancy failure amongst the sample:

• being aged 20-24
• having stayed in temporary accommodation such as hostels and supported housing before being resettled
• being rehoused in the PRS

Very few other studies seek to identify specific individual characteristics that predict tenancy sustainment or failure. Pawson and Munroe (2010) is one exception and, like Crane et al., their study of early or ‘chaotic’ tenancy terminations in social housing in Glasgow^{[footnote 26]} identified individual vulnerability as 1 of 3 key causes (along with market failure and managerial practices of social landlords). They referred to some of the same factors as Crane et al., citing history of care, rough sleeping, prison or the armed forces and young people with little experience of independent living. However, their study analysed data on all social housing tenants and so was not focused on homeless people or those with complex needs. Aubrey et al. (cited in Sheikh & Teeman, 2018), in their study of a Housing First programme in Canada, also identify significant vulnerability as a factor undermining tenancy sustainment. In the Chez Soi project there was a significant minority of people who had been chronically homeless and people with severe psychiatric issues who were least likely to achieve longer term stability (Sheikh & Teeman, 2018). However, as an international study, the transferability of these findings is unclear.

As noted above, Crane et al. (2016) found that time spent in temporary accommodation improves outcomes for tenancy sustainment. However, the evidence is mixed on whether time spent in temporary accommodation promotes or undermines successful resettlement into independent tenancies. In Crane et al.’s (2016) study, participants who had lived in hostels or temporary supported housing for longer than a year before they were resettled had a statistically significant higher rate of tenancy sustainment, while stays of 6 months or less in temporary accommodation were particularly likely to have ended in a further spell of homelessness. This was particularly true for those who had long histories of homelessness (10 or more years). These were also those more likely to be people with complex needs. They suggested this was because it allowed people time to:

• address the problems that led to their homelessness
• to access support services and address unmet needs
• to learn or practice independent living skills; and
• to restore confidence and self-belief, becoming engaged in meaningful activities and start planning for the future

On the other hand, by moving away from a ‘treatment’ first model, Housing First has successfully rehoused people sleeping rough directly into independent tenancies and the evidence is compelling of the effectiveness of such an approach (see Chapter 2), without the need to ‘staircase’ through temporary housing. In addition, other evidence suggests problems with living in temporary accommodation that can undermine, rather than support sustainment of subsequent tenancies. Homeless Link, for example suggested that time spent in temporary (presumably referring to supported, shared) provision created ‘a risk of dependency on support which makes successful resettlement harder to achieve…’ (Homeless Link, 2013) and Harding and Willet (2008), in their small qualitative study of homelessness services in one local authority in England found that respondents feared the loneliness of independent living following extended periods in shared hostels. These 2 studies are based on much less robust research than the Crane et al research but Homeless Link does have the benefit of long-standing organisational knowledge and expertise that should not be dismissed.  Fear of loneliness may also impact the success of Housing First, as Johnstone et al. (2016) highlight in their longitudinal study of 119 people in Australia where loneliness and tenancy breakdown were strongly linked. 

Relatedly, in the UK establishing supportive social networks has been found in a small number of studies to be important to effective resettlement (Bowpitt & Harding, 2008), where social isolation can threaten tenancy sustainment (Chamberlain & Johnson, 2018; Pawson & Munroe 2010; Smith et al., 2014). For example, Crisis and Shelter’s Sustain project identified a notable issue with social isolation amongst 128 housed participants in the private rented sector, which was attributed to a risk of a failed tenancy (Smith et al., 2014). Evidence from 2 supported housing projects, one for homeless people with long-term mental health issues, and another for ‘hard to house’ drug and alcohol users also found that the ‘sense of community’ between residents and support workers was important in sustaining the tenancies of residents who moved on to living in independent accommodation (Bowpitt and Harding, 2008).  Other studies, similarly, point to the value of company and community reported by residents in supported housing, and their associated fear of isolation and loneliness should they move on to an independent tenancy (Harding & Willett, 2008; Hennessey & Grant, 2006). However, this is generally raised as an issue in the literature rather than a conclusion based on a robust assessment of the relative impact of isolation on tenancy sustainment outcomes. Some people may also have experienced harms to their social and emotional well-being while living in temporary accommodation such as hostels (McMordie, 2020), making future communal living a difficult prospect.

In their longitudinal, qualitative study of 64 rehoused people who had been ‘chronically’ and long-term homeless in Australia, Chamberlain and Johnson (2018) found that challenges relating to ‘material transition’ (i.e. the practicalities of settling into a new tenancy) were relatively short-lived and more easily overcome, but that relational and psychological transitions proved more difficult. Although the vast majority of respondents remained housed after 12 months, they found that social isolation and other forms of exclusion (such as unemployment) persisted, potentially threatening tenancy sustainment in the longer-term. There may be lessons here for evaluations of projects with tenancy sustainment outcomes. These tend to focus on measuring material housing outcomes (how many respondents remain in their tenancy after a certain period of time) and pay less attention to whether other aspects of respondents’ lives, such as their social connections, might be undermining this apparent stability. 

There are also potential implications for policies that prompt relocation - for example those rehousing homeless people out of one area to another with greater housing supply or moving in order to access housing within the Local Housing Allowance rate. In some cases relocation can be vital for successful resettlement, for example where people are keen to move away from others associated with a problematic lifestyle (drug taking, criminal activity) or to escape violence or abuse (Bimpson, 2018), but in other cases this can also remove people from important sources of support and social connection, (McNaughton & Sanders, 2007; Smith et al., 2014). As a service provider commented in one study:

Our clients must have a local connection to be referred into the hostel but when they are ready to leave they will have to move to a borough where they have no connections due to the LHA rate being too high for Housing Benefit to pay for. We are expecting vulnerable people to move away from what support networks they do have.’ (In-text response).” Service provider quoted in Homeless Link, 2018b, p. 38-39)

3.4  Factors relating to service provision and delivery

There is literature providing evidence on those aspects of housing provision that people with complex needs value. For example, respondents in Burns et al.’s (2020) qualitative Canadian study emphasised the qualities of privacy and safety in their supported housing as created a ‘homelike’ feeling that was highly valued.  Gurstein and Small’s (2005) Canadian study of a supported housing project for people with complex needs (‘hard to house’ in their terms) similarly highlighted the benefits of housing provision which, “is the embodiment of the comfort and security of a home”’. They argued that such qualities are particularly important for people with complex needs for the following reason:

Individuals who are homeless or who have had difficulty integrating into the housing units of conventional housing agencies view creating a home as part of a wider process of personal self-healing. (Gurstein & Small, 2005, p. 732)

Implicitly, the ‘homelike’ security provided by the housing projects of Gurstein and Small (2005) and Burns et al.’s (2020) studies are likely to promote tenancy sustainment, but this small body of evidence does not actively engage with the key question of whether, or how, highly valued qualities of housing and support translate into sustained outcomes.

The evidence is relatively strong in relation to types of support that are most likely to promote tenancy sustainment and we review that evidence in Chapter 4. Beyond this, some insights can be gleaned about factors relating to service delivery that promote or undermine tenancy sustainment, but they are limited and must be taken as indicative only. A further caveat is that some sources are international and so may not translate to a UK context.

Zaretzky and Flatau (2015), reporting on tenancy support programmes for formerly homeless people (although not specifically people with complex needs) in Australia noted a number of service-related issues that constituted barriers to programme success in this international context.  These included:

• long waiting lists for support services, especially mental health and financial/budgeting support
• staff shortages and turnover in the supporting organisations which mitigated against relationship building, and
• a limit on twelve-month support, which was not sufficient for some clients (see Chapter 4 for further discussion about duration of support)

They also identified issues with housing management - including poor property maintenance - as undermining tenancy sustainment in this programme, a conclusion also reached by Pawson and Munroe (2010) in their UK study, where the managerial practices of social landlords were found to influence tenancy sustainment rates amongst social housing tenants. They identified issues with the following as contributing to tenancy termination:

• the allocation process, resulting in tenants accepting unsuitable properties, including those in areas in which they did not want to live, sometimes through perceived coercion. This point is echoed in recent research with social landlords which highlighted the potential for homeless households to be offered housing that is difficult to let (Preece et al. 2020)
• property conditions, including poor decorative conditions, and
• lack of information about the property (how to use the heating and suchlike)

Pawson and Munroe’s study was based on a general sample of social housing tenants and so is not directly applicable to the terms of this REA.  However, they did find that ‘homeless lets’ (i.e. those where applicants were entitled to housing on the grounds of homelessness) were at greater risk of tenancy failure than those ‘waiting list lets’ (i.e. those who were allocated properties through the waiting list and so are less likely to be homeless prior to rehousing), and noted the vulnerability of many of their qualitative respondents (mental health and substance abuse issues, history of care and institutions). As such, their conclusions may be of relevance.

A further issue raised in the literature relating to housing management policy and practice is the way in which previous tenancy failure and rent arrears can restrict access to independent accommodation. Harding and Willet (2008), for example, argued that in the delivery of housing and support to homeless people with complex needs some allowance must be made for tenancy ‘failure’. They suggested that homelessness people may need more than one independent tenancy before they were able to sustain their own accommodation but point to the way in which ‘failed’ tenancies can undermine future attempts to access independent housing, with social landlords excluding applicants to whom this applies. This is a very small, qualitative study, however, and so these conclusions should be treated with some caution.

A recurring theme is that services are often focused on single issues or a target population and so are not effective at meeting the needs of people with multiple, complex needs who either do not meet the single issue ‘needs’ threshold, or present with needs that are ‘too complex’ (Blood et al, 2020; Homeless Link, 2018b; Reeve et al., 2018; Scott, 2011; Seria-Walker, 2018). In terms of the relevance and reliability of this literature, it has the benefit of being focused on homeless people with complex needs but is often concerned with services wider than housing support. Nevertheless, the conclusions drawn are likely to have some transferability to a housing / supported housing context, especially as we do know from housing-related evidence that homeless people with complex needs face barriers accessing generic supported housing because their support needs are thought to be too high. Studies by Reeve et al. (2018)^{[footnote 27]} and Cream et al. (2020)^{[footnote 28]}, of access to healthcare in England by people sleeping rough, for example, highlight how people with co-occurring conditions can fall through the gap between 2 care pathways that seem to be mutually exclusive. They both provide examples of homeless people with mental ill health and substance misuse issues that are turned away from both services because of the other problem.

4. Tenancy sustainment in different types of accommodation

4.1  Introduction

This chapter reviews evidence about whether the rates and drivers of tenancy sustainment amongst people with complex needs vary depending on the type of accommodation in which they live. It addresses Research Question 4: Are tenancy sustainment issues different for different accommodation types e.g. hostels, other supported housing, independent housing in the social and private sectors?

Most of the evidence identified about tenancy sustainment amongst people with complex needs (including in Housing First projects) focuses on independent tenancies in the social or private rented sectors, or tenancies in medium-term supported housing projects.  In relation to the former, the tenancies in which study participants live is often accompanied by some form of floating support, including from Housing First or tenancy sustainment team workers. Very few sources were identified reporting evidence of tenancy sustainment amongst people with complex needs in independent tenancies without support (but see Crane et al. (2011) for a UK study where only half the sample of 400 had a tenancy support worker).  In relation to the latter, where supported housing projects are being evaluated, tenants usually had their own independent unit, sometimes but not always in a ‘cluster’, and a support worker. In many respects, then, the 2 forms of accommodation about which there is the most evidence share key characteristics, prohibiting analysis of differential experiences and outcomes according to type of housing. In both type of accommodation, tenants live independently rather than in shared, staffed accommodation such as hostels, and have support from a worker who is not on site.  The key difference is that supported housing projects are time-limited, in the UK at least, with a ‘maximum’ stay usually determined at the outset (often around 2 years).  In independent social or private rented sector tenancies no such time limit is imposed, although the accompanying support may be time limited. Given the similarities between the 2 forms of tenancy most commonly researched in relation to tenancy sustainment, it is not surprising that, comparing the literature on each, no clear differences emerged in their findings.  It is also worth noting, however, that this is a relatively small and disparate evidence base.  With a larger number of robust, comparable studies, more differences might emerge. 

With no evidence identified on tenancy sustainment amongst people with complex needs in other forms of housing - for example hostels or other short-term or emergency supported housing - it is not possible to draw conclusions about whether the factors driving tenancy sustainment vary by accommodation type.  There is evidence drawing attention to how challenging hostel environments can be for homeless people with complex needs (e.g. McMordie, 2020; Reeve et al., 2018 in relation to homeless people with mental health issues), and a suggestion that long stays in temporary housing increase the risk of eviction or abandonment from that accommodation (Homeless Link, 2013) but no clear evidence on whether these issues do translate into ‘tenancy’^{[footnote 29]} termination or eviction.

We might also assume that some of the factors driving tenancy sustainment identified in Chapter 3 will apply only to certain types of accommodation, but this can be inference only.  For example, we reported in Chapter 3 that social isolation has been raised as an issue potentially undermining tenancy sustainment (although not robustly demonstrated to be the case). This will be specific to independent sole tenancies. In contrast, it is precisely the presence of other residents, and their behaviours that is reported to create problems for some people with complex needs in shared temporary accommodation such as hostels. This might also apply to people living in independent but shared tenancies in the private rented sector (Green and McCarthy, 2016).

The only concrete evidence on this issue identified by the REA related to differential tenancy sustainment rates by tenure, with people with complex needs in the private rented sector less likely to sustain tenancies than in the social rented sector. It is to this that we now turn.

4.2  Tenancy sustainment in different tenures

Evidence from the longitudinal FOR-HOME and Rebuilding Lives studies, which tracked 400 single homeless people resettled into independent tenancies (Crane et al., 2011; 2016), half with and half without tenancy sustainment support, showed that resettlement into the private rented sector was strongly and statistically significantly associated with poorer housing outcomes. In the FOR-HOME study (Crane et al., 2011) less than one half (47%) of the private renters in the sample were still in the resettlement accommodation 15/18 months after resettlement compared with more than four-fifths of social housing tenants. Of these, 27% were without a tenancy at all. In the follow-up study, Rebuilding Lives, which revisited participants 5 years after their last interview, Crane et al. (2016) found a strong relationship between resettlement into the private rented sector and subsequent homelessness, with 58% of those rehoused into the private rented sector becoming homeless again, compared to 22% of local authority tenants and 18% of housing association tenants. They also found that rent arrears - a potential driver of tenancy failure - were more common amongst those housed in the private rented sector. In their 2016 study 43% of private renters had debts of £1,000 or more at 15/18 months, compared to 28% of local authority tenants and 18% of housing association tenants.

The precise reasons why tenancy failure might be more prevalent in the private rented sector than the social rented sector is not clearly determined in the evidence base, but the qualitative experiences of people resettled into the PRS shed some light on why this might be the case. Participants in the Rebuilding Lives evaluation raised issues with the:

… poor condition of their accommodation, conflicts with landlords regarding getting repairs done, difficulties with rent payments, and conflicts with other tenants if in housing with shared facilities. (Crane et al., 2016, p. 56)

A 2018 review of the UK private rented sector (Rugg & Rhodes, 2018) revealed the emergence of a ‘slum tenure’ at the bottom end of the private renting market, which people with the most limited housing options are exposed to.  They found that tenants with the lowest incomes and with characteristics which place them at the greatest risk of harm face the worst property conditions and management.  

The anxiety caused by the poor quality of private rented accommodation, and by managing relationships with their landlord, has been identified as a contributory factor to tenancy breakdown in other studies too.  Research by Crisis and Shelter showed in a longitudinal study of 128 people housed in the private rented sector following homelessness, around two-thirds were unhappy with their tenancy after 19 months (Smith et al., 2014). Where landlords were defensive or hostile about repairs or health and safety issues, tenants reported being more likely to leave their tenancies. This research also highlighted issues with the lack of furnishings, white goods and other essential items in private rented tenancies into which participants were housed, which impacted on their sense of security and comfort that a home should provide. Some participants had taken out pay day loans to fund those items, and by month 19 others had still not been able to afford to buy a bed or other essentials (Smith et al., 2014). Other studies of tenancy sustainment concur that furnishing a tenancy is one practice that makes a house ‘feel like a home’ and that difficulties furnishing tenancies can contribute to tenancy failure (Ambrose et al., 2016). This study of the contribution of furniture provision to tenancy sustainment in Northern Ireland was not related specifically to people with complex needs but they did find that groups at most risk of tenancy failure where absence of furniture had been a contributory factor included formerly homeless people (as well as young people, care leavers and people with physical and/or learning difficulties).  Another study which evaluated the success of cash grants for disadvantaged people also highlighted the importance of carpets and furnishing to avoid tenancy failure early on (Lankelly Chase, 2016).  However, like the other studies cited here, it did not report tenancy sustainment rates. These studies identify problems associated with living in privately rented accommodation and, in some cases, are able to trace some link between these experiences and tenancy failure, but they do not conclusively show a causal relationship or measure the relatively influence on tenancy sustainment rates.

It is also important to note that issues associated with furniture and essential goods are not restricted to the private rented sector.  Indeed, the sample for the Ambrose et al. (2016) study cited above were Northern Ireland Housing Executive tenants, and carpets, furnishings and white goods are generally less likely to be provided in the social rented sector than the private rented sector. In addition, one source found that participants who had been resettled into private rented sector property tended to be relatively happy with the quality of the accommodation including furnishings available (Quilgars et al., 2012). Others suggested that closer management in the private rented sector and use of tenancy support workers can mitigate some of the problems associated with private rented sector tenancies that might undermine tenancy sustainment (Crane et al., 2011).

5. Tenancy sustainment support

5.1  Introduction

This chapter concerns forms of tenancy sustainment that have been successful and, in particular, what evidence there is about the type and duration of support that is most effective. In doing so, it also responds to the review questions 4 and 7: ‘What forms of tenancy sustainment support have been successful? In particular, what evidence is there about the type and duration of support, and who provides it’, and ‘What can be done to help improve tenancy sustainment for those with complex needs?’

5.2  The provision of support

While evidence suggests that people with complex needs require support to help sustain a tenancy, there is a lack of direct focus and evaluation evidence on the actual sustainment achieved.  For example, Harding and Willett (2008, p. 440) reported that, “a failure to provide continuing support could undo the positive work of accommodation projects, meaning that substantial amounts of public funding had been used for no benefit”, a principle of ongoing support that is enshrined in Housing First (see Chapter 2).  Similarly, Homeless Link (2018b, p. 38) found that ongoing support with a direct focus on tenancy sustainment was a key factor in recovery from complex needs.  The report quoted from one service provider:

Client move ons are successful because of the investment made in helping clients address their specific, often complex, issues - issues which may have been a contributory factor in their becoming homeless or contributed to their continued social isolation. Clients are given access to the key personal, financial, work related, and practical skills they will need to be able to lead a successful independent life back in the community.

A recent evidence review on ‘what works’ to end rough sleeping provides some useful insights (Mackie et al., 2017; 2019). While the review was not concerned directly with identifying ‘what works’ in tenancy sustainment, and was not focused explicitly on people with complex needs, implicitly, interventions that provide a long-term solution to rough sleeping include elements of tenancy sustainment. The following measures and interventions were identified as being evidenced to ‘work’ to end rough sleeping:

• housing led solutions (including Housing First)
• person centred support, with particular evidence for the effectiveness of personalised budgets.
• services that specifically focus on addressing wider support needs
• effective collaborations between agencies and across sectors
• swift action, and
• assertive outreach leading to a suitable accommodation offer

The private rented sector (PRS) accommodates people with lived experience of homelessness and complex needs, and local authorities and voluntary services often draw on this (Reeve et al., 2016; Rugg, 2014).  A report from the Making Every Adult Matter Coalition (MEAM, 2015) recommended that local policy should support and encourage landlords in the PRS who are willing to rent to tenants with complex needs, for example through the creation of loan deposit schemes and insurance policies to cover landlords for damage etc (p. 26). The same report highlighted that it was important for properties used to be appropriate and decent.  Other reports highlighted that support was required for more vulnerable people in the PRS (Reeve et al., 2016; Batty et al., 2015), but did not focus specifically on complex needs, nor evidence what kinds of tenancy support were successful.

The evidence base identifies a number of different kinds of provision of support that benefit tenancy sustainment.  While there have been some programmes and projects focused squarely and exclusively on tenancy support (such as intensive tenancy management schemes in the 1990s and 2000s), support for people with a lived experience of homelessness and complex needs tends to address multiple aspects of recovery – and tenancy sustainment plays a part in this.

The Tenancy Sustainment Teams (TST) Programme was founded by the Department of Environment Transport and the Regions (DETR) in London in 2000. It provided resettlement and continuing support to former rough sleepers living in accommodation funded by the Rough Sleepers Unit.  An evaluation (Lomax & Netto, 2007) found that the programme did improve tenancy sustainment. Intensive practical and emotional support was provided to tenants to meet set ‘goals’. The evaluation highlighted good practice, which included:

• shared understanding of the aims and objectives of TSTs with landlords and tenants
• integration of TSTs with parent organisations which enables sharing of good practice in delivering services, staff training, support and management
• flexible needs-centred provision of support, varying as agreed with the tenant, including the regularity of contact or visits and involvement of generic and specialist staff
• the availability of a wide range of support services, including financial, practical and emotional support
• termination of support when not needed combined with the ability to reactivate support at a time of ‘crisis’
• good joint working between TSTs and Contact and Assessment Teams and Pre-Tenancy Teams, facilitated by protocols developed by the Clearing House,^{[footnote 30]} and
• evidence of networking between, for example, health services and the police around the needs of individual tenants

A study of the FOR-HOME Programme (Crane et al., 2011) found that similar percentages of people with and without a tenancy support worker were still housed after 15/18 months.  However, there was a relationship between contact with a tenancy support worker and rent arrears – those who still had a tenancy support worker at 15/18 months were less likely to have rent arrears, and less likely to have been taken to court for arrears (p. 85).  The Clearing House, a London-based service, provides people who have slept rough with a housing association property alongside ongoing floating support provided by 2 Tenancy Sustainment Teams. In its first 25 years of operation, 92% of tenants had never returned to rough sleeping (St Mungo’s, 2018, p. 14).  Crane et al. (2016) support the need to have dedicated teams.  They found that, “tenancy support services for people with complex needs should be provided by designated tenancy support, housing support or floating support workers, who can address the spectrum of problems and needs.” (p. 14).  They also found that, “positive housing outcomes are associated with rent subsidies and access to subsidised housing, enhanced support or case management services, treatment for substance misuse, and involvement in employment and training schemes” (p. 29).

Homeless Link’s Annual Review 2017 (Homeless Link, 2018b) identified holistic and personalised support packages as a key aspect of successful moves into tenancies, though it should be noted that the review did not provide strong outcomes-based evidence.  Robust support packages delivered by dedicated project workers during people’s residency, were identified as key in ensuring that clients could develop the skills, abilities and resources needed to progress towards independent living. Providing assistance that was holistic and tailored to people’s personal needs was also highlighted as particularly important. Other studies (Carver et al., 2020, p. 18; Blood et al., 2020, p. 50) also reported the importance of providing assistance that was tailored to people’s personal needs. The Mental Health Foundation (2016, p. 17) found that person-centred approaches designed to help across multiple areas of need (housing, addiction, mental health, etc.) were essential, as opposed to a compartmentalised approach based on a range of ‘professional distinctions’.

Forms of Floating support were reported as being important for successful moves to a secure tenancy. Homeless Link (2018b, p. 36) found that several homelessness services reported that the lack of ongoing or floating support once a person had moved on from an accommodation project was a limiting factor. As one service stated:

Even when people move on, things can go wrong and so resettlement support does not have time limits … I believe the flexible approach helps enormously and they know they can always come back if things go a bit pear shaped! (Homeless Link, 2018b, p. 39)

Ongoing support is, therefore, broadly regarded as being important, but it was also apparent from evidence that support required flexibility to respond to a person’s specific needs (Harding & Willett, 2008; Hennessey & Grant, 2006). ‘Flexible support’ is not always well defined in the evidence, but it tends to refer to flexibility in duration and intensity (Harding & Willett, 2008) and flexibility of the support worker to respond to the changing and varied needs of the client (Hennessey& Grant, 2006). The latter found that support was most effective where it adapted to individual needs on an ongoing basis. This relied on support workers to be capable of identifying the complexities and intricacies of their client’s needs.

As was identified in Chapter 2 relating to Housing First initiatives, if it is accepted that person-centred approaches are beneficial and that a person’s needs are unique, support should not be time limited. As Boardman (2016) found:

The literature on Housing First suggests that support should be provided either long-term or indefinitely. It may be more helpful to view the support as being provided ‘when people need it’. For people with long-term conditions, needs and wishes for support will change over time. People need to be able to access support easily when required.” (p. 20)

Findings from the Multiple Exclusion Homelessness Research Programme funded by the ESRC also suggested that, “rigid approaches that specify a time period for engagement rather than responding to an individual’s own pace are not appropriate for people with more complex needs” (cited in Scott, 2011, p. 9). Furthermore, the findings also suggested that ‘persistent and ongoing encouragement and support’ appeared to be a factor in improving outcomes for people with complex needs. A 2018 REA of homelessness services identified issues with the short length of support offered to individuals across homelessness services, a lack of sustainability and variable consistency which negatively impacted individual outcomes (Sheikh & Teeman, 2018). For example, evidence from a women’s sanctuary scheme highlighted the wide range of support needs evident among its service users, including legal issues and child contact, as well as more basic tenancy and financial management and support with mental health and other needs (Jones (2010) in Sheikh & Teeman, 2018, p. 32). This suggests a need for varied but co-ordinated efforts between services in order to support individuals living in independent tenancies.

Several studies have reported on the virtues of longer-term support, though it should be noted that outcomes were rarely quantified. Blood et al. (2020) carried out research into the commissioning of homelessness services over the past 10 years and identified several approaches that were seen as effective ways of supporting people with complex needs, and prioritised intensive and long-term support. They highlighted Housing First and ‘wrap-around services for multiple complex needs’ as examples of good practice. Commissioners from one unitary authority reported a rapid growth of complex cases and rough sleeping, that the main housing options service and other welfare services were ill-equipped to cope with. In response they established a team whose remit was to work with people with multiple complex needs who were at risk of sleeping rough. Central to this was wrap-around support, an accommodation offer, and the ability to work with people, “for as long as needed.” Similar teams have been created via the Rough Sleeping Social Impact Bond Programme in England (see Green et al., forthcoming). Carver et al. (2020) echoed this view in their systematic review of what constitutes effective substance use treatment for those experiencing homelessness, that, “lengthy or continuous support was often considered necessary and could be provided in the form of safe housing, such as in Housing First settings” (p. 14). Such interventions are linked to the concept of ‘recovery capital’ (Ibid.) which refers to the resources people can draw on to begin and maintain recovery from problematic substance use. These include social capital (relationships with others), physical capital (income, savings, property), human capital (knowledge, skills, health, education), and cultural capital (values, beliefs and attitudes that promote social norms). The authors argued that people experiencing homelessness often have low levels of recovery capital, making treatment and recovery difficult, and improvements hard to sustain. While in this context it is referring specifically to (self-defined) recovery from alcohol or substance misuse, it may be potentially useful for looking at what is important for people to sustain tenancies after periods of homelessness; an area for further research. 

Crane et al. (2016) also provided insights into the need for long-term support. The identified several areas in which people who were previously homeless were experiencing difficulties 60 months (5 years) after their initial resettlement. This included financial problems, unmet mental health and substance/alcohol use support need, difficulties with household tasks. They concluded that intermittent or regular long-term support is needed to avoid further homelessness (p. 173). Importantly, they highlighted that the people who were still in receipt of various forms of support at 60 months tended to be those with multiple problems and long histories of homelessness. This suggests, therefore that longer-term support is particularly important for those who continue to experience complex needs.

Across the UK, pre-tenancy support has become an established intervention that prepares people to better manage a tenancy. Organisations such as Crisis, Shelter, and other running ‘help to rent’ schemes provide pre-tenancy training that includes aspects such as paying rent and bills, managing the tenant/landlord relationship, understanding a tenancy agreement and residing harmoniously with others in a neighbourhood. There is some evidence that suggests pre-tenancy training is useful in sustaining tenancies, however it is does not necessarily apply to people with complex needs, where engagement with such activities could be more difficult. Batty et al. (2015) found that help to rent schemes used such training effectively with people who had no previous experience of managing a tenancy and were restricted in their housing choices to shared accommodation. Many projects in the Sharing Solutions Programme made pre-tenancy training a prerequisite of a housing offer. Hennessy et al. (2005) evaluated a hostel resettlement service for young people in Merseyside and emphasised how support and assistance provided while homeless people were in temporary accommodation could be the foundation for successful resettlement into independent accommodation. They argued that the resettlement process needed to begin in advance of actual resettlement, assisting people to develop key life skills, access training and employment, and support them with the practical aspects of moving house (furniture and such like). Likewise, Crane et al. (2016) argued that advice and training on money management should happen before resettlement, and that where possible, people should also be involved in education, training and employment (ETE) activities before they were rehoused, as, “people who had been involved in ETE at the time of resettlement were significantly more likely to be engaged in ETE at 60 months” (p. 96). The report also emphasised the importance of being financially prepared for independent living, including having reasonable debt repayment plans in place. A previous study by the authors (Crane et al., 2011) found that in the FOR-HOME programme, less than half of respondents had received advice or training in managing money at their previous hostel or supported housing, and their levels of preparedness and knowledge about bills and finances were low.

Personalised budgets may be effective in supporting the resettlement of people characterised as having experienced ‘entrenched’ homelessness, though the evidence base is not robust. Blackender and Prestidge’s (2014) review of 28 homeless people reported that 20 people were resettled at the end of the personalisation initiative which used multiagency working. When supported by effective inter-agency collaboration, the ability to exercise financial discretion improved outcomes in respect of housing, social and health needs.

There is growing interest and application of trauma-informed approaches being adopted by services caring and providing support for people with complex needs (for example, see McCarthy et al., (forthcoming)). While there is literature about the use and advantage of trauma-informed approaches and psychologically informed environments (PIE), there was little evidence of the impact on tenancy sustainment. For example, Homeless Link (2018b, p. 40) reported that their survey of homelessness services found that, “the introduction of a Psychologically Informed Environment (PIE) and/or trauma-informed approach had reduced the rate of evictions and introduced a better way of dealing with challenging behaviour.” Evidence to support this, is lacking, however.

Beyond direct approaches, the evidence base suggests that there are other services (or pathways) that have a role to play in better tenancy sustainment for people with complex needs. For example, Cream et al. (2020) cite specialist hospital discharge teams as a successful way of preventing rough sleeping. In their study all 4 case study areas had examples of this type of ‘in-reach’ intervention and saw it as an effective way of reducing rough sleeping. The projects worked to advocate for longer hospital stays for people experiencing homelessness, on a clinical basis:

Proactive hospital discharge teams used their clinical expertise to ‘convince a hospital’ that a longer admission period would be time well spent to give a patient a proper assessment and a safe and well-co-ordinated discharge. (p. 27)

5.3  Skills to provide support

Working with people who have lived experiences of homelessness and have a number of complex needs can be challenging, and there is evidence to suggest that staff skills and expertise are a key factor in encouraging people to engage with support and sustain it. Workers who support people with tenancy sustainment, therefore, require some particular skills and training. Green et al. (forthcoming) explored workforce development issues amongst those who worked directly with people with complex needs. In projects funded by the Rough Sleeping Grant initiative and the Rough Sleeping Social Impact Bonds, the report found that projects were best able to meet the needs of this group where staff could be empathetic and patient with their service users and could maintain a positive relationship without sanctioning. This was summarised by one manager as ‘stickability’ – the idea of never giving up on a service user and maintaining an unconditional offer of support. Similarly, other studies (Carver et al., 2020; Holly, 2017) stress that non-judgemental, compassionate staff are crucial to successful relationships and service delivery.

In addition to the support that professionals bring to tenancy support, some projects across the UK have used peer-support, by drawing on the resources and skills of people with lived experiences of complex needs and citing the advantages that this brings for engagement. Crisis, for example, has promoted the use of peer-support in its own services and amongst others that it works with (Batty et al., 2015), and it was seen to provide better engagement and sustainment support that resulted in tenancy sustainment. Similarly, some studies have found that service design and delivery benefited from co-production with people who had lived experience (MEAM, 2015; Cream et al, 2020) as they helped to design and deliver services that are accessible and effective for people with complex needs, and in this case, rough sleepers (Cream et al, 2020, p. 26).

5.4  The role of specialist housing with support

The evidence suggests that many initiatives aimed at tenancy sustainment do not fully integrate support and housing. Housing First is one important exception, as discussed fully in Chapter 2, where it is apparent that for people with complex needs, models built on housing and tailored support can be beneficial to tenancy sustainment. Beyond Housing First, there have been other initiatives directed at people with complex needs and designed to provide sustained accommodation. Mental Health Foundation (2016) report on the Care Support Plus model which was developed through a tripartite agreement between the housing provider, the local NHS Foundation Trust, and the local authority to develop a new type of supported accommodation specifically geared towards people who had often been excluded from supported accommodation due to their complex mental health needs and who might have otherwise been in hospital or residential care. The support drawn together was reported to be pioneering because of high levels of wraparound support, levels of staffing and the presence of NHS staff onsite, which enabled timely responses to problems and more intensive support for residents. In addition, it reported that there was evidence to suggest that people with complex needs considered tailored support and person-centred approaches to be highly valuable, though they cautioned that there was limited evidence about the best model to follow.

5.5  Summary

The evidence relating to effective tenancy support is typically about people who have experienced homelessness, but not always about people with complex needs (although for this review, evidence was only included if it was judged that the sample included people with complex needs, or the findings were likely to be transferable). There is evidence about the type of support, and the principles of service delivery that are most effective for people with complex needs but in different policy areas. However, relevant insights can be gleaned from this evidence. It appears that having formal tenancy support arranged prior to a move to individual accommodation is very important, as is having to support to gain key skills to independently manage a tenancy effectively. The immediacy of support and necessary expertise also appears to be a critical factor.

6. Conclusions and gaps in the evidence base

Broadly, the results of this rapid evidence review concur with the conclusions drawn by Crane et al. (2016), who make the following observation in relation to the evidence on tenancy sustainment (‘resettlement’ in their terms):

There has been very little rigorous research in England of resettlement preparation for homeless people. Little is known about the effectiveness of various housing and support services, and the programmes and interventions that associate with successful and less successful outcomes for homeless people with different problems and backgrounds. The need for more refined understanding of the effectiveness of various interventions and housing and work training programmes for homeless people is well documented by researchers in the US and elsewhere. There is also a need for more understanding about the process of ‘recovery’ from homelessness for people with different problems and needs, including the benefits and/or disadvantages of stays in temporary accommodation before taking on the responsibilities of independent living.” (Crane et al., 2016, p. 178).

The overall aim of this REA was to identify and explore the barriers and drivers of tenancy sustainment for people with complex needs in an effort to identify what works best to support people with complex needs - and sub-groups within this population - to sustain appropriate accommodation. In an effort to meet this aim we have gleaned what insights there are from an extensive and relatively rich body of evidence about homeless people, but one which rarely measures or explicitly explores tenancy sustainment outcomes and experiences for people with complex needs. As such, we are reliant on research and evaluation studies that are certainly relevant, but which never aimed to address the core questions of this REA. It includes, for example: research exploring people with complex needs’ experiences of different types of accommodation and support; evaluations of interventions such as personal budgets or furniture schemes that might support tenancy sustainment; and evaluations of resettlement initiatives for the general homeless population. In addition, information about research samples, methods, and analytical techniques is not always comprehensively presented in publications, making it difficult to derive information specifically about tenancy sustainment, or about people with complex needs.

The exception is a body of literature evaluating Housing First programmes. People who have been rough sleeping and have health and/or addiction needs are a common target group for Housing First interventions, and tenancy sustainment is a key objective and so Housing First research and evaluation provides directly relevant evidence. Compared to the international evidence base, which commonly uses randomised control trials, the quality of evidence in the UK is, however, more limited, usually either focused on individual projects (Boyle et al., 2016; Johnsen, 2013) and/or are short-term (with the exception of Johnsen, 2013).

Despite this limitation, lessons can be derived from the UK evidence base. There is compelling evidence, for example, that a Housing First approach works well for homeless people with complex needs in so far as existing UK evidence demonstrates impressively high housing retention outcomes (consistent with those in countries where Housing First is more established). Studies have also started to emerge which show the considerable cost savings of Housing First when compared to fixed site services (Pleace & Bretherton, 2019). There also exists a weight of evidence about the types of support, and the way they are delivered, that appear most effective for people with complex needs (with the caveat that many studies are not methodologically robust enough to demonstrate this with certainty). This research literature often draws evidence from fields beyond housing, including from health, mental health and addiction services but nevertheless indicates some broad principles that are transferable to a homelessness service context. This body of evidence shows that positive sustainable outcomes for people with complex needs are most likely to be achieved if support is flexible, intensive, not time-limited, and person centred. Many of the housing and wider support services found to be effective in meeting the needs of homeless people with complex needs (with the caveat that measurement is rarely consistent or defined) operate according to at least some these principles. It was also the case that outcomes may be improved by developing a workforce that has the requisite skills to support and stick with people whose complex needs mean that there are steps backwards, as well as steps forward, on the way to their recovery. Peer-support also has a role to play in this regard.

The evidence is much more limited when it comes to pinpointing the precise circumstances under which people with complex needs are most likely to sustain a tenancy, and the sub-groups at greatest risk of tenancy failure. The evidence base provides some insights, but they are limited. People resettled into the private rented sector have been found at high risk of tenancy failure than those in the social rented sector, and social isolation amongst those moved away from social and familial connections is suggested to threaten tenancy sustainment. However, the evidence is far from conclusive on these points and identifying why this might be the case is much more difficult. We do not, for example, know with any certainty which features of the private rented sector pose a risk to tenancy sustainment, or under what circumstances feelings of isolation do and do not translate into tenancy failure.

The evidence base does not, at the present time, provide clear answers to the question of what factors are most important in driving tenancy sustainment, and for whom. As a result, it remains difficult for policy and practice to develop services, initiatives and programmes for homeless people with complex needs, with confidence that they will effectively support this group of vulnerable people to sustain appropriate accommodation in the long term. The evidence is clear that promoting tenancy sustainment for homeless people with complex needs will require attention across multiple facets of their lives - the material, social and psychological - and be capable of holistically addressing multiple needs and in a coordinated effort by multiple services. Any key factor found to be significantly correlated with tenancy sustainment will always be just one factor amongst a complex set of overlapping issues.

Gaps

1. There are few statistically robust UK studies capable of clearly isolating characteristics statistically correlated with tenancy sustainment. Study aims and design tell us a lot about what people value in their housing and support, or about difficulties they face, but there is little robust evaluative evidence capable of isolating specific variables driving tenancy sustainment. However, it is important to note that more robust international randomised control trials of Housing First programmes have not been able to ascertain which groups fare better and worse and under what circumstances.
2. The definition of ‘sustainment’ has a temporal dimension in the sense that something is only said to be sustained if it has been stable over time. Yet many research studies and evaluations are conducted over a year or 2, including evaluations of Housing First Programmes. Some evaluations reviewed for this REA recorded a positive ‘tenancy sustainment’ outcome if a participant remained in their tenancy for just 6 months. With evidence suggesting that less tangible challenges with the resettlement process (building social connections, overcoming stigma, feeling ‘at home’) can persist long after resettlement, there is a need for longitudinal studies which track long-term outcomes of clients. It is no accident that one of the most useful sources for this REA, in the sense that it was capable of providing answers to some of our research questions, was one which tracked homeless people (albeit not specifically with complex needs) over several years (Crane et al., 2011; 2016).
3. There is very little sub-group analysis in the evidence base. The methodological approach commonly taken to studies of resettlement or tenancy sustainment programmes in the UK precludes this kind of quantitative analysis and/or studies are too small for robust sub-group analysis. We do not, therefore, understand differential experiences and outcomes or the extent to which research findings apply to different groups. This applies to the more robust and extensive Housing First literature as much as to other evidence. Owing to the often-low numbers of clients in Housing First pilots and projects in the UK (and in the associated evaluations) it is difficult to draw definitive conclusions about which clients experience the most success from Housing First. A better understanding of programme outcomes for specific sub-populations may prove useful in understanding the outcomes of widespread implementation of Housing First, including people with more significant mental health issues.
4. Picking up on the point above, certain sub-groups are relatively absent from the evidence base. Study samples may include diverse population groups, but full profile information is not always presented and/or their distinct experiences not reported.  In particular, there is a notable gap in evidence relating to sustaining tenancies amongst Black and Minority Ethnic (Sheikh & Teeman, 2018) and LGBT+ people with complex needs. This silence calls for further research to understand the distinct experiences and challenges these groups face accessing and sustaining tenancies, who are also overrepresented amongst the homelessness population. There is also very little evidence about tenancy sustainment amongst older people (but see Burns et al., (2020) from Canada) reflecting a gap in the homelessness literature more generally.

Bibliography

Ambrose, A., Batty, E., Eadson, W., Hickman, P. & Quinn, G. (2016). Assessment of the need for furniture provision for new NIHE tenants. Sheffield Hallam University. 

Barker, N. (2018, February 6). Welsh Government to roll out Housing First across Wales. Inside Housing.

Batty, E., Cole, I., Green, S., McCarthy, L. & Reeve, K. (2015). Evaluation of the Sharing Solutions Programme. Crisis.

Bellis, A. & Wilson, W. (2018). Housing First: Tackling Homelessness for those with Complex Needs. Briefing Paper 08368. House of Commons Library.

Bernad, R. & Yuncal, R. (2016). Introducing the Housing First Model in Spain: First Results of the Habitat Programme. European Journal of Homelessness, 10(1), 53-82.

Bimpson, E. (2018). An Evaluation of Basis Yorkshire’s Housing First Pilot. Leeds Social Sciences Institute, University of Leeds.

Blackender, L. & Prestidge, J. (2014). Pan London Personalised Budgets for Rough Sleepers. Journal of Integrated Care, 22(1), 23-26.

Blood, I., Copeman, I., Goldup, M., Pleace, N., Bretherton, J. & Dulson, S. (2017). Housing First Feasibility Study for the Liverpool City Region. Crisis.

Blood, I., Goldup, M., Peters, L. & Dulson, S. (2018). Implementing Housing First across England, Scotland and Wales. Crisis.

Blood, I., Pleace, N., Alden, S. & Dulson, S. (2020). ‘A Traumatised System’: Research into the Commissioning of Homelessness Services in the Last 10 Years. University of York.

Boardman, J. (2016). More than Shelter: Supported Accommodation and Mental Health. Centre for Mental Health. 

Bowpitt, G. & Harding, R. (2008). Not Going It Alone: Social Integration and Tenancy Sustainability for Formerly Homeless Substance Users. Social Policy & Society, 8(1), 1–11.

Boyle, F., Palmer, J. & Ahmed, S. (2016). The Efficiency and Effectiveness of the Housing First Support Service piloted by DePaul in Belfast, funded by Supporting People: An SROI Evaluation. North Harbour Consulting.

Bramley, G., Fitzpatrick, S., Edwards, J., Ford, D., Johnsen, S., Sosenko, F. & Watkins, D. (2015). Hard Edges: Mapping Severe and Multiple Disadvantage. Lankelly Chase Foundation.

Bramley, G., Fitzpatrick, S., Wood, J., Sosenko, F., Blenkinsopp, J., Littlewood, M., Frew, C., Bashar, T., McIntyre, J. & Johnsen, S. (2019). Hard Edges Scotland. Lankelly Chase Foundation.

Bretherton, J. & Pleace, N. (2015). Housing First in England: An Evaluation of Nine Services. Centre for Housing Policy.

Burns, V., Deslandes- Leduc, J., St-Denis, N. & Walsh, C. (2020). Finding Home after Homelessness: Older Men’s Experiences in Single-site Permanent Supportive Housing. Housing Studies, 35(2), 290-309.

Carver, H., Ring, N., Miler, J. & Parkes, T. (2020). What Constitutes Effective Problematic Substance Use Treatment from the Perspective of People who are Homeless? A Systematic Review and Meta-ethnography. Harm Reduction Journal,

Chamberlain, C. & Johnson, G. (2017). From Long-term Homelessness to Stable Housing: Investigating ‘Liminality’. Housing Studies, 33(8), 1246-1263.

Cherner, R., Aubry, T., Sylvestre, J., Boyd, R. & Pettey, D. (2017). Housing First for Adults with Problematic Substance Use. Journal of Dual Diagnosis, 13(3), 219-229.

Crane, M., Warnes, T. & Coward, S. (2011). Moves to Independent Living: Single Homeless People’s Experiences and Outcomes of Resettlement. University of Sheffield.

Crane, M., Joly, L. & Manthorpe, J. (2016). Rebuilding Lives: Formerly Homeless People’s Experiences of Independent Living and their Longer-term Outcomes. King’s College London.

Cream, J., Fenney, D., Williams, E., Baylis, A., Dahir, S. & Wyatt, H. (2020). Delivering Health and Care for People who sleep Rough: Going Above and Beyond. The King’s Fund.

Crisis (2018). Everybody In: How to End Homelessness in Great Britain. Crisis.

Estecahandy, P., Agha, A. & Roebuck, M. (2018). Fidelity Study of the “Un chez-Soi d’abord” Housing First Programmes in France. European Journal of Homelessness, 12(3), 159-181.

Fitzpatrick, S., Pawson, H., Bramley, G., Wilcox, S., Watts, B. & Wood, J. (2018). Homelessness Monitor: England 2018. Crisis.

Fitzpatrick, S., Pawson, H., Bramley, G., Wood, J., Watts, B., Stephens, M., & Blenkinsopp, J. (2019). The Homeless Monitor: England 2019. Crisis.

Fitzpatrick, S., Pawson, H., Bramley, G., Wood, J., Stephens, M., Frey, J., & McMordie, L. (2020). Homelessness Monitor: Northern Ireland 2020. Crisis.

Goering, P., Veldhuizen, S., Watson, A., Adair, C., Kopp, B., Latimer, E., Nelson, G., MacNaughton, E., Streiner, D. & Aubry, T. (2014). National at Home/Chez Soi: Final Report. Mental Health Commission of Canada.

Gough, D. (2007). Weight of Evidence: A Framework for the Appraisal of the Quality and Relevance of Evidence. Research Papers in Education, 22(2), 213-228.

Green, S. & McCarthy, L. (2015) Is sharing the solution? Exploring the opportunities and challenges of privately rented shared accommodation for single people in housing need. People, Place and Policy, 9(3), pp 159-176.

Green, S., Pearson, S., Sanderson, E., Batty, E., Dayson, C., McCarthy, L., Parr, S. & Reeve, K. (forthcoming). Rough Sleeping and Complex Needs Process Evaluation: Learning from the Rough Sleeping Social Impact Bond Case Studies. MHCLG.

Greenwood, R. M., Schaefer-McDaniel, N. J., Winkel, G., & Tsemberis, S. J. (2005). Decreasing Psychiatric Symptoms by increasing Choice in Services for Adults with Histories of Homelessness. American Journal of Community Psychology, 36(3–4), 223–238.

Gurstein, P. & Small, D. (2005). From Housing to Home: Reflexive Management for those Deemed Hard to House. Housing Studies, 20(5), 717-735.

Hansen, I.L.S (2018). Users’ Choice in Providing Services to the Most Vulnerable Homeless People. Social Inclusion, 6(3), 319-326.

Harding, J. & Willett, A. (2008). Barriers and Contradictions in the Resettlement of Single Homeless People. Social Policy & Society, 7(4), 433-444.

Hennessy, C. & Grant, D. (2006). Developing a Model of Housing Support: The Evidence from Merseyside. International Journal of Consumer Studies, 30(4), 337-346.

Hennessy, C., Grant, D., Cook, I. & Meadows, M. (2005). Young and Homeless: An Evaluation of a Resettlement Service on Merseyside. International Journal of Consumer Studies, 29(2), 137–147.

Holly, J. (2017). Mapping the Maze: Services for Women experiencing Multiple Disadvantage in England and Wales.  Agenda & AVA.

Homeless Link. (2013). Effective Action: Resettlement from Homelessness Services. Homeless Link.

Homeless Link. (2015). ‘Housing First’ or ‘Housing Led’? The current picture of Housing First in England. Homeless Link.

Homeless Link. (2017). Housing First in England: The Principles. Homeless Link.

Homeless Link (2018a). The Picture of Housing First in England. Homeless Link.

Homeless Link (2018b). Support for Single Homeless People in England: Annual Review 2017. Homeless Link.

Housing First Scotland (2020). Housing First Check-Up: Annual check-up on Scotland’s Housing First Pathfinder. Housing First Scotland.

Johnsen, S. (2013). Turning Point Scotland’s Housing First Project Evaluation. Heriot-Watt University.

Johnstone, M., Parsell, C., Jetten, J., Dingle, G. & Walter, Z. (2016). Breaking the cycle of Homelessness: Housing Stability and Social Support as Predictors of Long-term Well-being. Housing Studies, 31(4), 410-426.

Jones, K., Gibbons, A. & Brown, P. (2019). Assessing the Impact of Housing First in Brighton and Westminster. SHUSU, University of Salford.

Lancione, M., Stefanizzi, A. & Gaboardi, M. (2018). Passive Adaptation or Active Engagement? The Challenges of Housing First Internationally and in the Italian Case. Housing Studies, 33(1), 40-57.

Lankelly Chase. (2016). Opening Doors, Changing Lives. Measuring the impact of cash grants on disadvantaged individuals and families. Lankelly Chase & NEF Consulting.

Lomax, D.& Netto, G. (2007). Evaluation of Tenancy Sustainment Teams. Department for Communities and Local Government.

Mackie, P., Johnsen, S., & Wood, J. (2017). Ending Rough Sleeping: What Works? . Crisis.

Mackie, P., Johnsen, S. & Wood, J. (2019). Ending Street Homelessness: What Works and Why We Don’t Do It. European Journal of Homelessness, 13(1), 85-96.

McCarthy, L., Parr, S. & Green, S. (forthcoming). Understanding Models of Support for People facing Multiple Disadvantage: A Literature Review. Sheffield Hallam University.

McMordie, L. (2020). Avoidance Strategies: Stress, Appraisal and Coping in Hostel Accommodation. Housing Studies.

McNaughton, C. & Sanders, T. (2007). Housing and Transitional Phases Out of ‘Disordered’ Lives: The Case of Leaving Homelessness and Street Sex Work. Housing Studies, 22(6), 885-900.

MEAM. (2015). Solutions from the Frontline: Recommendations for Policymakers on Supporting People with Multiple Needs. Making Every Adult Matter Coalition Clinks, Homeless Link and Mind.

Mental Health Foundation. (2016). Mental Health and Housing. Mental Health Foundation.

Montgomery, A. E., Hill, L. L., Kane, V. & Culhane, D. P. (2013). Housing Chronically Homeless Veterans: Evaluating the Efficacy of a Housing First Approach to HUD-VASH. Journal of Community Psychology, 41(4), 505–514.

Padgett, D. K. (2007). There’s no place like (a) home: Ontological Security among Persons with Serious Mental Illness in the United States. Social Science and Medicine, 64(9), 1925-1936.

Palepu, A., Patterson, M. L., Moniruzzaman, A., Frankish, C. J., & Somers, J. (2013). Housing First improves Residential Stability in Homeless Adults with Concurrent Substance Dependence and Mental Disorders. American Journal of Public Health, 103, e30–e36.

Pawson, H. & Munro, M. (2010). Explaining Tenancy Sustainment Rates in British Social Rented Housing: The Roles of Management, Vulnerability and Choice. Urban Studies, 47(1), 145–168. 

Pleace, N. (2011). The Ambiguities, Limits and Risks of Housing First from a European Perspective. European Journal of Homelessness, 5(2), 113-127.

Pleace, N. (2016). Housing First Guide Europe. FEANTSA.

Pleace, N. (2018). Using Housing First in Integrated Homelessness Strategies: A Review of the Evidence. Centre for Housing Policy.

Pleace, N. & Bretherton, J. (2012, June). What do we Mean by Housing First? Categorising and Critically Assessing the Housing First Movement from a European Perspective [Conference presentation]. ENHR 2012 Conference, Lillehammer, Norway.

Pleace, N. & Bretherton, J. (2013a). The Case for Housing First in the European Union: A Critical Evaluation of Concerns about Effectiveness. European Journal of Homelessness, 7(2), 21-41.

Pleace, N. & Bretherton, J. (2013b). Camden Housing First: A Housing First Experiment in London. Centre for Housing Policy.

Pleace, N. & Bretherton, J. (2019). Housing First England: The cost effectiveness of Housing First in England. Homeless Link.

Pleace, N. & Quilgars, D. (2013). Improving Health and Social Integration through Housing First: A Review. DIHAL/FEANTSA.

Pleace, N., Baptista, I. & Knutagård, M. (2019). Housing First in Europe: An Overview of Implementation, Strategy and Fidelity. Housing First Europe Hub.

Pleace, N., Culhane, D.P., Granfelt, R. & Knutagård, M. (2015). The Finnish Homelessness Strategy: An International Review. Ministry of the Environment.

Preece, J., Bimpson, E., Robinson, D., Mckee, K., & Flint, J. (2020). Forms and Mechanisms of Exclusion in Contemporary Housing Systems: A Scoping Study. CaCHE.

Quilgars, D. & Pleace, N. (2017). The Threshold Housing First Pilot for Women with an Offending History: The First Two Years. Centre for Housing Policy.

Quilgars, D. & Pleace, N. (2018). The Inspiring Change Manchester Housing First Pilot: Final Evaluation Report. Centre for Housing Policy.

Quilgars, D., Jones, A., Bevan, M., Bowles, R. & Pleace, N. (2012). Supporting Short-term Prisoners leaving HMP Leeds: Evaluation of the Shelter Prisoners Advocacy Release Team. Shelter.

Reeve, K., Casey, R. & Goudie, R. (2006). Homeless Women: Still being Failed yet Striving to Survive. Crisis.

Reeve, K., Cole, I., Batty, E., Foden, M., Green, S. & Pattison, B. (2016). Home. No less will do: Homeless people’s access to the Private Rented Sector. Crisis.

Reeve, K., McCarthy, L., Pattison, B., Parr, S., Batty, E., Maye-Banbury, A., Bashir, N & Dayson, C. (2018). The mental health needs of Nottingham’s homeless population: an exploratory research study. Sheffield Hallam University.

Rowe, S. & Wagstaff, T. (2017). Moving On: Improving Access to Housing for Single Homeless People in England. Crisis.

Rugg, J. (2014). Crisis’ Private Rented Sector Access Development Programme: Final Evaluation Report. Crisis.

Rugg, J. & Rhodes, D. (2018). The Evolving Private Rented Sector: Its Contribution and Potential. Centre for Housing Policy.

Scott, G. (2011). A Place to Live: Securing Stable Accommodation for Offenders with Mental Health Problems. Centre for Mental Health.

Seria-Walker, E. (2018). Evidence Review: Adults with Complex Needs (with a particular focus on street begging and street sleeping). Public Health England.

Sheikh, S. & Teeman, D. (2018). A Rapid Evidence Assessment of what works in Homelessness Services. Social Care Institute for Excellence.

Smith, M., Albanese, F., & Truder, J. (2014). A Roof Over My Head: The Final Report of the Sustain Project. Shelter & Crisis.

St. Mungo’s. (2018). Home for Good: The Role of Floating Support Services in Ending Rough Sleeping. St. Mungo’s.

Stefancic, A. & Tsemberis, S. (2007). Housing First for Long-term Shelter Dwellers with Psychiatric Disabilities in a Suburban County: a Four-year Study of Housing Access and Retention. Journal of Primary Prevention, 28, 265-279.

The Prison Reform Trust (2019). Prison: The Facts. Bromley Briefings Summer 2019. The Prison Reform Trust. 

Tsemberis, S. (2010a). Housing First: Ending Homelessness, Promoting Recovery and Reducing Costs. In I. Gould Ellen & B. O’Flaherty (Eds.), How to House the Homeless (pp. 37-56). Russell Sage Foundation.

Tsemberis, S. (2010b). Housing First: The Pathways Model to End Homelessness for People with Mental Illness and Addiction. Hazelden.

Tsemberis, S., Gulcur, L., & Nakae, M. (2004). Housing First, Consumer Choice, and Harm Reduction for Homeless Individuals with a Dual Diagnosis. American Journal of Public Health, 94(4), 651–656.

Volk, J., Aubry, T., Goering, P., Adair, C., Distasio, J., Jette, J., Nolin, D., Stergiopoulos, V., Steiner, D., & Tsemberis, S. (2016). Tenants with Additional Needs: When Housing First does not Solve Homelessness. Journal of Mental Health, 25(2), 169-175.

Watts, B., Littlewood, M., Blenkinsopp, J. & Jackson, F. (2018). Temporary Accommodation in Scotland: Final Report. Heriot Watt University.

Wilson, W. & Barton, C. (2020). Rough Sleeping (England). Briefing Paper 02007. House of Commons Library.

Woodhall-Melnik, J. & Dunne, J. (2016). A Systematic Review of Outcomes associated with Participation in Housing First Programs. Housing Studies, 31(3), 287-304.

Zaretzky, K. & Flatau, P. (2015). The Cost Effectiveness of Australian Tenancy Support Programs for Formerly Homeless People. Australian Housing and Urban Research Institute.

---

1. Woodhall-Melnik and Dunne (2016) also observe that the evidence sometimes refers to ‘stability’, ‘housing retention’ or ‘tenancy sustainment’. All mean the amount of time spent in a housing unit.  ↩

2. The figure refers to formerly homeless people housed for at least one year by a Housing First service. ↩

3. Housing First clients (n D 89) and members of a comparison group (n D 89) completed structured interviews at baseline and were followed for 2 years. ↩

4. Also, in its annual review of support for single homeless people in England, Homeless Link (2018a) estimated that 19% of the 121 accommodation-based homeless services in England who responded to their survey were using a Housing First model, with a further 26% considering using Housing First. ↩

5. In Blood et al. (2018), practitioners were in agreement that building trust and relationships with private and social landlords was essential to enabling a supply of suitable local properties. ↩

6. Homeless Link (2018b) identified that funding for Housing First services in England is short-term (1-2 years is the most common length of funding), which continues to pose a challenge when the Housing First approach commits to providing an open-ended offer for clients.  ↩

7. For instance, Pleace (2011) notes that the Pathway Housing First model will not sit very comfortably within a wider national strategy that is intended to deliver cessation of problematic drug use. ↩

8. Housing Secretary James Brokenshire awards funding to reduce rough sleeping. ↩

9. The study involved qualitative research with 95 professionals within LCR, 79 people with lived experience, and 8 commissioners from other parts of the UK; analysis of referral and assessment gateway data for commissioned homelessness services in LCR; review of local policies and evidence; and analysis of potential costs and savings from the model. ↩

10. Since 2014. ↩

11. Launched in April 2017. ↩

12. This work was a small, entirely desk-based exercise. This report explores the cost effectiveness of Housing First, drawing on the results of anonymous surveys of 15 Housing First services in England, local authority service commissioners and a group of 29 Housing First service users. ↩

13. The evaluation employed a longitudinal methodology (over 3 years) and involved repeat interviews with TPS staff and representatives of key stakeholder agencies (total n=30), repeat interviews with service users (total n=43), and analysis of service users’ case files. ↩

14. Based on a survey of 28 out of 32 active Housing First services providers, July to September 2017. ↩

15. Drawing on qualitative interviews with women using Threshold Housing First, with the workers and managers providing the service, and partner agencies working with Threshold. Researchers also analysed data collected by the project to examine referral patterns, support needs and service outcomes. ↩

16. Launched in 2017. ↩

17. Housing First for Youth Evaluation. ↩

18. Based on bespoke analysis, Blood et al. (2018) calculate that if Housing First was implemented at scale (in 2018, at the time the report was written) the Housing First cohort in Britain would be somewhere between 18,400 and 32,250. ↩

19. The analysis involved triangulation between 4 administrative areas and datasets; 15 interviews with national level stakeholders, 25 local key informants were interviewed, 8 focus groups, 47 frontline workers and 42 in-depth interviews ↩

20. This was a qualitative study of 64 rehoused people who had been ‘chronically’ and long-term homeless in Australia ↩

21. This qualitative sample of 25 homeless people included people with additional support needs but was not focused on people with complex needs. The study also included 16 interviews with housing and support providers. ↩

22. This REA reviewed 35 sources on ‘what works in homelessness services’ broadly, rather than specifically in relation to tenancy sustainment, Some sources provided evidence about tenancy sustainment, including those about Housing First, and their general conclusions about what works set out a series of principles that might apply to tenancy sustainment services.  ↩

23. In homelessness legislation and policy ‘single’ refers to people without dependent children. ‘Single’ homeless people are less likely that those with dependent children to be awarded priority need for housing.  ↩

24. Comprising longitudinal interviews with 400 resettled single homeless people and followed up in Crane et al 2016 where 297 of the original cohort were re-interviewed. This study represents one of the very few robust studies in England that tracks resettlement outcomes over a relatively long period.  The 2011 study interviewed respondents over an 18 month period following resettlement. Five years later, the 2016 study re-interviewed  those who had still been housed at their final interview  and who were contactable (297 people). ↩

25. The Local Housing Allowance rate is the amount of Housing Benefit that that can be claimed in an area. The rate is calculated on the basis of private market rents in the broad rental market area in which the claimant would reasonably be expected to live. ↩

26. The study comprised statistical analysis of Glasgow Housing Association lettings and interviews with 50 former tenants ↩

27. Based on a survey and interviews with homeless people with mental ill health in Nottingham. ↩

28. Based on interviews with stakeholders and focus groups with rough sleepers in 4 case study areas. ↩

29. Hostel residents are not provided with tenancies but are usually given a license to occupy. ↩

30. The Clearing House provides supported housing in London for people with a history of rough sleeping and is managed by St Mungo’s.  Properties are self-contained one-bed or studio flats, provided by housing associations as part of the Rough Sleepers Initiative (RSI). Tenants are supported by Tenancy Sustainment Teams, who provide a floating support service. ↩