Not mission impossible: An evidenced-driven approach to ending homelessness is within our grasp
The Centre for Homelessness Impact (CHI) is engaged in an epic quest. Through its End It With Evidence campaign it seeks to make evidence-credible programmes, practices and policies in association with robust data intrinsic and essential in pursuit of the challenging, but achievable goal of ending homelessness. Where specific programmes, services or interventions can be evidenced as helping people escape homelessness, it believes these should be widely disseminated and appropriately replicated in accordance with the identified needs and priorities of local communities. Those that cannot demonstrate they are transforming lives and bringing homelessness to an end should be remodelled, replaced or ended. Indeed, how can we expect to end homelessness without this pragmatic, clear-eyed rigour?
So why does it need a campaign at all if it is so obviously the right approach? And what can be done to speed up progress at a moment when, as a result of the Everyone In initiative that has helped so many people experiencing street homelessness and the positive impact of the Homelessness Reduction Act, we have an exceptional, though perhaps transitory opportunity to take a great leap forward towards ending homelessness for good? In my view there are three key challenges:
Scaling up evidence-based research and robust data
The CHI is relentless in uncovering and assessing research. Its website provides access to a vast range of research, convincingly interrogated to produce a strong evidence base of what works and what doesn’t. This should be gold-dust for organisations seeking to develop and deliver high quality, transformational services that can end homelessness.
Yet the most recent update of the CHI’s Evidence and Gap Map (2021) provides a salutary reminder of how far we need to travel before we can confidently claim to understand what works, why and for whom. Of the 394 pieces of research on interventions to improve the situation of people experiencing homelessness, only 28 are from the UK, and 70% score a rating of ‘low confidence’ as a measure of being able to convincingly provide a strong evidence-base, 25% were of medium quality and just 10 studies (5%) were of high quality. Many other pieces of research were simply not strong enough to even make the cut.
There were enough high and medium quality evaluations of some areas of delivery such Housing First, effective housing responses to people with mental health challenges and Critical Time Intervention (CTI) to confidently be able to understand how such programmes and interventions help people escape homelessness effectively and sustainably. In the UK, Housing First services are now commonly provided, though availability is far from comprehensive. CTI, which has become a familiar delivery model in parts of North America and Scandinavia, is still comparatively rare in this country.
In contrast, the CHI review did not uncover a single piece of research on hostels that reached the necessary standard to be deemed convincingly evidence-based, even though hostels are a major part of the delivery response to homelessness nationally.
At Thames Reach where I was formerly Chief Executive, we managed a number of hostels which were well regarded and popular with commissioners and local communities. We produced plenty of outcome information and continually sought to improve the hostels physically, respond to new needs and, where they were evidence-based, adopt new ways of delivering high quality services. But crucially, what impact the hostels were making on the lives of people experiencing street homelessness as a contributor to the wider local and sub-regional system of service delivery was much more difficult to assess.
This is the challenge. Because of severe limitations on the availability of shared data we struggle to provide convincing evidence of how hostels within a place-based network of services can contribute to ending homelessness, even though in London alone there are, according to the Atlas of London Homelessness Services, 245 accommodation-based services comprising 10,048 bed spaces for homeless people, the great majority of which are either hostels or some form of shared housing.
At a strategic level, how can we ascertain whether a local area or region should be increasing or reducing its hostel provision?
The conflict between an evidence-based approach and organisational priorities
Which begs the question, how can it be that programmes and services for people who are experiencing homelessness have evolved with apparently little regard to whether they are making a positive difference? Clearly this is not a case of wilful intent; homelessness organisations and funders of homelessness services seek to deliver and support projects they are convinced are helping make people’s lives better. Most organisations actively engage with people using their services to understand how services can adapt to meet changing needs.
The homelessness sector delivers a complex hotchpotch of services. Most organisations specialise in a particular area of delivery and in supporting specific cohorts within the homelessness population. There is competition between organisations to ensure that the needs of their people using their services are highlighted and the form of service provision they specialise in is recognised and appreciated. An organisation neglecting to do this would, understandably, feel they were letting down their people using their services and possibly putting at risk their sustainability as profile and funding, especially via public donation, are closely linked.
These conditions often blunt the incentive to reflect collaboratively with other homelessness organisations on how projects and programmes across a locality, region, even nationally, could be remodelled to provide more interconnected, symbiotic and transformational services.
As in the example of the Thames Reach hostels, individual services are often, as stand-alone projects, effective and perform well as measured against the outcomes they have been set. Yet if services are not considered in the context of a wider operating environment at local, sub-regional and regional levels, we will not get close to ending homelessness. To do this, a level of data sharing is required which is rare in the UK. London has the immensely valuable CHAIN client recording system for people experiencing street homelessness, but beyond the capital it is hard to find a city that has any real data-led understanding of their homeless population. Even CHAIN has limitations in being able to track the journey of people through services.
One encouraging recent example that indicates that there may be increased interest in more comprehensive, system-wide analyses is the Heriot-Watt evaluation of Greater Manchester’s ‘A Bed Every Night’ programme which explicitly considered the broad range of services linked to the programme and measured the effectiveness of the response to homelessness in Greater Manchester against international evidence of what works. The result is an evaluation that, while not assessing causal impact, acknowledges good progress and makes convincing recommendations for change.
In the United States, supported by Community Solutions, a not-for-profit organisation, some communities are now able to effectively track people through services to produce quality, real-time data appropriate for a homelessness population which is naturally going to be dynamic rather than static. Collecting data in this way requires trust between organisations and there is undeniably a necessary financial cost and time commitment. However, if we have a serious intent to end rough sleeping and other forms of homelessness, improved data collection and assessment is not optional but essential.
Making the collection and use of robust data the cultural norm
Do we have the appetite to take the data-leap required? We often seem curiously disinclined to turn to data to help understand a problem. For example, in meetings with London local authority leads on street homelessness, in my experience an impressive group of people determinedly committed to reducing rough sleeping, there are frequent conversations about the movement (flow) of people sleeping out around London. It is a very important issue because investigating and understanding the shifting patterns of this population could lead to better prevention, improved collaborative working and effective reciprocal arrangements to house people in the borough best placed to meet their needs. Given the robustness of the CHAIN data base, it is possible to track movement around boroughs. In my experience, on the few occasions when a data exploration exercise has taken place, the information produced has led to some unexpected conclusions. Typically, borough A had not, as feared, experienced a big increase in rough sleeping as a consequence of people being ‘displaced’ into it from borough B. Things were not as they seemed.
Yet despite this, anecdote and supposition usually prevail. The discipline of seeking data where it is available has to be strengthened and the End It With Evidence campaign can play an important part by relentlessly making the case that credible data, where it exists, must always trump anecdote, even when the latter is from credible sources. And alongside this, we should champion evidence-led communication based on an honest acknowledgement of the limits of our understanding and an aversion to what Ligia Teixeira, CHI’s Chief Executive, refers to as ‘the dangerous half-truth that can pass as wisdom’.
Brave New World
Something has happened during the period of the Covid-19 pandemic that previously would have seemed preposterously optimistic. Faced by the real possibility of a very significant number of people sleeping on the streets and in precarious housing situations losing their lives through infection, a coalition formed with a mission to help as many people as possible to rapidly come in off the street and into a place of safety and then to find a way for them move into settled, sustainable accommodation. The coalition included homelessness charities, local and central government, health services, faith groups, housing providers and the public. Lives were saved - at least 266, according to an authoritative analysis published in The Lancet.
For the coalition to be successful, organisations and bodies had to operate to a different model which required suppressing narrow organisational needs for the common good, responding to developing events with a level of urgency that previously would have been considered impossible and collecting, collating and studying data obsessively in order to fully comprehend what was happening on the ground and the effectiveness, or otherwise, of interventions being put in place to help people avoid infection and maintain accommodation.
As a member of the Covid-19 Rough Sleeping Task Force during this time, it was personally an enormously instructive and ultimately inspirational experience as I know it was for others. Health colleagues, for example, were able to gather crucial data on the vast, disparate health needs of the street homeless population that greatly enhanced their knowledge and understanding of this deeply disadvantaged group of people.
It will be difficult to sustain this level of engagement and commitment as the vaccination programme continues, infection rates fall, and fear recedes. Yet, if there was ever a moment to commit to an evidence-based approach to ending homelessness, this is it. Street homelessness numbers are low and the government’s commitment to tackling homelessness endures with a further £203 million of funding recently announced to support the next stage of the Rough Sleeping Initiative.
It has pledged to end rough sleeping before the end of this Parliament. Encouragingly, we have witnessed the strengthening of relationships across sectors and disciplines. As the pandemic raged, health services, homelessness organisations and local authority housing departments found new ways of working together to support people experiencing street homelessness being placed in hotels. Central government worked closely with faith groups to help people move out of crowded shelters and to find new ways in which homeless people could be supported in response to the threat of infection. We must hope that these are not temporary changes but permanent realignments.
But we are hamstrung by our limited knowledge of what works. Through a CHI impact report we know accommodation-based programmes that offer high levels of tailored support with an unconditional housing offer are proven to be successful. We also have evidence that basic projects providing a bed and food will often produce worse health and housing stability outcomes than no intervention at all. They remain a prominent part of the current patchwork of services in this country.
As a tangible next step, I propose that a group of homelessness organisations with significant hostel provision commit to rigorously scrutinising the role of hostels in the delivery of services to people experiencing homelessness in the UK and that a partnership of funders agrees to collaboratively support this research which will be scrupulously evidence-focused and commit to the type of shared data approach necessary to deliver comprehensive, sustainable, evidence-based systems change. What hostel-based interventions are most effective in helping people take the step to settled housing? How do the alternatives to a hostel as effective and enduring routes off the street measure up? Is there evidence that hostels operating in particular ways do more harm than good? How well do hostels address health issues and for which groups do they work best?
So much to learn, so much to change, so much to gain.
Jeremy Swain is an adviser on homelessness and former CEO of Thames Reach
The philosophy of using evidence to end homelessness
Our CEO, Ligia Teixeira, spoke at a graduation ceremony at the Glasgow University and explained her commitment to empirical evidence to end homelessness.
Comparing data on homelessness deaths
How do data on homelessness deaths compare across the UK? This blogpost summarises the different data sources used.
Five questions we'll explore at our upcoming Impact Forum
In our upcoming Impact Forum, on 15 + 16 November, we’ll be bringing together global evidence of interventions to prevent and relieve homelessness, looking at what works to prevent worsening homelessness in a cost of living emergency. Find out more about the topics we'll be exploring.