Evidence-based policy

What works in violence reduction? Learning from the London Vanguards evaluation

/ jlowther2015 / Leave a comment

Dr. Jason Lowther

This is the last (for now) of our series of blogs on recent evaluations of local public services – if you have published an evaluation you think may be of interest to Inlogov readers, please send me a link to it ([email protected]).

Much of the evaluation literature on violence reduction has focused on system-wide approaches such as Violence Reduction Units. The London Vanguards independent evaluation offers a close look at one defined service model designed to support children and young people affected by violence through integrated, community-based provision.

The London Vanguards programme, delivered across 12 boroughs between 2022 and 2025, brought together health, community and psychological support into a single, coordinated offer for young people at risk of, or affected by, violence. The independent evaluation, led by the Anna Freud Centre, used a mixed-methods realist design, combining service data with interviews and focus groups involving young people, families and practitioners. The aim was not only to assess outcomes, but to understand how and why the model worked in practice.

A holistic, multi-systems approach

At the core of the Vanguards model is a holistic understanding of violence risk. Rather than focusing narrowly on offending behaviour, the programme recognises that many young people experience multiple, overlapping adversities – including mental health needs, family instability, educational exclusion and unmet developmental needs.

The evaluation shows that this approach enabled services to identify needs that had often gone unrecognised elsewhere, particularly around neurodevelopmental conditions, speech and language difficulties and mental health. These underlying issues frequently shape both vulnerability to violence and the ability to engage with support.

The model therefore combined psychological support, practical help and wider system navigation, rather than relying on a single intervention. In doing so, it reflects a shift from treating violence as an isolated problem to addressing it as part of a wider set of life circumstances.

Outcomes: improvements across multiple domains

The evaluation presents a broadly positive picture of outcomes, although with important caveats around data completeness.

Around 1,500 children and young people were supported over the life of the programme. Among those with available data, there were reported improvements across several domains:

  • mental health improved for around 47 per cent of participants
  • education and accommodation outcomes improved for around one third
  • reductions in offending and high-risk behaviours were reported for a significant minority

Qualitative evidence reinforces these findings. Young people, parents and practitioners consistently described increases in confidence, wellbeing and sense of agency, as well as improvements in family relationships and engagement with education or employment.

Importantly, the evaluation suggests that these changes were not always short-lived. Many participants maintained positive outcomes over time, which was attributed to the longer-term, relationship-based nature of the support.

What drives change: relationships, persistence and coordination

Among the most valuable insights from the evaluation are those examining how change was achieved.

First, trusted relationships between practitioners and young people emerge as a central mechanism. The model’s flexible and persistent approach allowed practitioners to engage individuals who might otherwise disengage from services. This sustained engagement appears to be critical in supporting progress.  The importance of this relational approach to public services is increasingly recognised.

Second, multi-agency coordination proved essential. Although establishing partnerships was initially challenging, over time services became more aligned, enabling more coherent and responsive support. This coordination allowed practitioners to address needs across different domains without requiring young people to navigate multiple disconnected systems.

Third, the programme’s holistic design enabled it to respond to changing needs. Rather than fixed pathways, support could adapt as circumstances evolved, which is particularly important given the instability many participants experienced.

The data challenge: promising evidence, incomplete picture

As with many complex service evaluations, the report highlights limitations in the evidence base. Missing data rates are high across several outcome measures, and there is no control group for comparison. This makes it difficult to draw definitive conclusions about impact.

What does this mean for local authorities?

The London Vanguards evaluation suggests several lessons for local government and partners.

First, violence reduction is inseparable from wider needs. Effective responses must address mental health, development, family context and structural disadvantage alongside risk behaviours. Single-issue interventions are unlikely to be sufficient.

Second, relationship-based practice is central. The strongest evidence of change comes from sustained, trust-based engagement between practitioners and young people. This has implications for workforce stability, caseloads and commissioning approaches.

Third, integration works, but takes time. The evaluation shows that multi-agency coordination improved over time, but required sustained effort to establish.

Fourth, flexibility is a strength. The ability to adapt support to individual needs was a key feature of the model. Highly standardised approaches may struggle to achieve similar outcomes with complex populations.

Finally, the evaluation highlights the importance of taking a long-term view. Improvements were gradual, interconnected and not always immediately visible. This may challenge some performance frameworks but better reflects the reality of supporting young people affected by violence.

Taken together, the evidence from the London Vanguards suggests that what works in this area is not a single intervention, but a way of working: holistic, relationship-based and coordinated across services. The challenge for government and local authorities is creating the conditions to sustain this approach over time.

From top-down diagnosis to co-design: what youth‑led evidence adds to the Government’s review of Young people and Work

/ jlowther2015 / Leave a comment

Sonia Bussu

The government’s Young People and Work interim report presents a stark diagnosis. Nearly one million young people in the UK are not in education, employment or training (NEET), and the problem is structural, longstanding and getting worse. The report describes a system failure spanning education, labour markets, health and welfare, and it concludes that the UK lacks a coherent “participation system” to support transitions into work. It is a powerful diagnosis of a system that no longer supports young people. But it is also striking that much of what the report “discovers” has already been articulated vividly by young people themselves, many times over!

The INSPIRE programme in the West Midlands offers exactly this missing perspective: lived experience. Through participatory research and legislative theatre, a diverse group of young people aged 14–17 from across the region identified specific barriers, staged a play to reflect on these barriers with the community and policymakers, and co‑designed policy proposals to address them. Set alongside the government’s diagnosis, this youth‑led evidence helps move from description of problems to concrete and youth-led solutions.

Structural barriers are visible in lived experience

The government review highlights three major structural issues: reduced entry‑level jobs, fragmented systems and unequal access to support. The INSPIRE young co-creators showed us how they encounter these issues day to day. Their play, You’re Fried! The realities of youth employment, depicts a system where career guidance is rushed and superficial, where broken weblinks and empty words replace meaningful support, where opportunities are inaccessible, in a cycle of “entry‑level” roles that always require prior experience.

You’re Fried! The Realities of Youth Employment
Scene One: “What Should I Do With My Life?” revealed how love, when bound by fear, could become a cage. A young apprentice’s dream was crushed beneath the weight of parental expectations, where university was seen as the only safe route to independence.
Scene Two: “It’s All on You” exposed the hollow machinery of career guidance – rushed advisors offering broken links and empty words, leaving hope behind in the rush to tick boxes.
Scene Three: “Link Not Found” brought a computer to life, its customer-service smile masking the cold indifference of digital systems that fail those who need them most.
Scene Four: “This Isn’t What I Signed Up For” pulled back the curtain on the myth of ‘real-world experience’ – a McDonald’s shift where understaffing and blame replaced training and support.
Scene Five: “It’s Your Fault” returned to the family living room, where systemic failure was rewritten as personal shame, completing a devastating circle.

The government review links NEET risk to socioeconomic background, race, disability and geography. The INSPIRE young people described the same dynamics of discrimination and constrained expectations. Systemic failures are often reframed as individual shortcomings, reflecting a system that assesses young people but does not adequately support them, placing the burden of navigating complexity on individuals.

While employers face uncertainty and capacity constraints, young people also reported the limited support on accessing valuable work placements or apprenticeships, as they suffer from limited training, little supervision and weak protections. In these situations, responsibility again falls on the young person to adapt. The issue is not a lack of motivation from young people, but a lack of fit between systems and the realities young people face. Young people want to work but are navigating fragmented and often inaccessible pathways.

Youth‑led proposals provide concrete solutions

One of the most innovative aspects of the INSPIRE project is its methodology. Legislative theatre allowed young people not just to describe problems, but to stage them, rehearse alternatives, and co-create policy responses. We co‑designed  proposals to improve access to work experience and employment through creative and structured dialogues between young people, policymakers, employers and communities.

Several of these proposals directly address the issues identified in the government review.

1. Earlier and family‑centred career support
The review highlights the importance of early intervention, noting that disadvantage accumulates over time. The INSPIRE young people proposed starting career conversations at primary school level and involving families, recognising that aspirations and knowledge about career pathways are shaped early and collectively.

2. Accessible and community‑based careers guidance
To address fragmentation and inequality in provision, participants proposed delivering career support also through community organisations and the voluntary sector.

3. Reform of work experience and employer engagement
In response to declining entry points into the labour market, young people proposed structured work experience, mentoring and exposure to workplaces through site visits and flexible placement formats. They also highlighted the need to prepare employers to support young people effectively. Working with local schools, we are now co-creating standards for youth‑friendly employers.

4. Cross‑sector coordination
The government review identifies fragmentation across institutions as a core problem.
INSPIRE proposals include the creation of a cross‑sector alliance on youth employment in Birmingham bringing together young people, education, employers, public bodies and community organisations to coordinate action.

5. Youth‑led evaluation and accountability
Young people know best what they need. They proposed mechanisms for evaluating careers services and employer practices, embedding youth perspectives into ongoing policy development. A youth-led evaluation of career and employment support is now being implemented, co-led by Birmingham City Council and the University of Birmingham.

Rethinking policy: from programmes to relationships

A recurring criticism in participation is the lack of follow-through from policymakers. Young participants expressed frustration with commitments that are vague or delayed, or carelessly forgotten, which deeply undermines their trust in institutions.

INSPIRE suggests that the answer is not simply better engagement programmes, but different relationships between institutions and young people, where trust depends on feedback loops, transparent commitments, and long-term engagement rather than one-off consultations. It’s not simply about redesigning systems but redistributing power.

Young people do not need to be fixed or made “work-ready”. They need systems that recognise their knowledge, respond to their realities, and involve them as partners in shaping policy that affects them.

Sonia researches and teaches public policy. Her main research interests are participatory governance and democratic innovations, and creative and arts-based methods for research and public engagement.

What works in community ownership and local assets?

/ jlowther2015 / 1 Comment

Dr Jason Lowther

Alongside major programmes on local growth and regeneration, there has been increasing policy interest in the role of community-owned assets in strengthening local places. The Community Ownership Fund (COF) was a £150 million programme designed to support communities to take control of valued local buildings and spaces at risk of loss, from pubs and community centres to sports facilities.

The interim evaluation provides an early but valuable insight into how this model works in practice. It focuses on both delivery processes and emerging outcomes, based on a mixed-methods approach combining quantitative survey analysis, qualitative case studies and value-for-money assessment across a sample of projects. Rather than offering definitive conclusions, it highlights the conditions under which community ownership can succeed, and where it struggles.

A distinct model: enabling communities to “save” assets

The COF was designed to address a specific gap in the system: the loss of locally valued social infrastructure. By providing capital funding to support acquisition and refurbishment, the programme enabled community groups to intervene where assets would otherwise have closed or been sold.

The evaluation finds that, in this respect, the fund has been effective in enabling communities to secure assets that were at genuine risk. In many case study areas, projects would not have gone ahead without this funding, particularly given wider economic pressures. The scale of capital support is identified as a key feature, allowing communities to take on projects that would otherwise be financially out of reach.

Early outcomes: social value more visible than economic impact

At this interim stage, the evaluation points to positive early social outcomes, although evidence on longer-term impact is still developing.

Qualitative findings from case studies suggest that community-owned assets are already delivering benefits in terms of:

  • strengthened local identity and pride
  • increased community participation and volunteering
  • improved access to services and social spaces

These outcomes reflect the broader theory of change behind the fund: that ownership can generate not just service provision, but engagement, empowerment and social capital.

However, the evaluation is more cautious on economic outcomes such as financial sustainability and local economic growth. These are necessarily longer-term and more uncertain, particularly given the time required to refurbish and reopen assets.

What enables success in practice?

The process evaluation provides important insight into what makes community ownership projects work.

First, existing capacity within community organisations matters. The fund was most accessible to groups that already had the skills, governance structures and experience required to develop complex capital projects. Where this capacity was lacking, projects found it more difficult to apply for and deliver funding.

Second, support and advisory provision are critical. Changes made to the programme over time, including increased support for disadvantaged areas and adjustments to match funding requirements, improved accessibility. This suggests that financial investment alone is not sufficient; communities also need technical and developmental support.

Third, relationships with funders and partners shape delivery experience. Many projects reported the need for clearer communication, more consistent engagement and greater flexibility from central government during the delivery phase. This reflects a wider lesson across local programmes: delivery quality depends as much on how funding is administered as on the funding itself.

The challenge of accessibility and equity

A recurring theme in the evaluation is the tension between opportunity and inequality.

While the COF provides a mechanism for community empowerment, it does not operate on a level playing field. Communities with stronger organisational capacity, access to professional skills and existing funding networks are better placed to take advantage of the opportunity. Less advantaged areas may face greater barriers, even where need is higher.

Programme adjustments, such as reducing match funding requirements and targeting support, have helped to address this. But the evaluation suggests that structural inequalities remain a significant factor in shaping who benefits from community ownership.

Sustainability and risk

The evaluation also highlights the question of long-term sustainability. Taking ownership of an asset creates ongoing responsibilities, including maintenance, staffing and financial management.  While many projects have strong business plans, the wider economic environment poses risks, particularly for organisations reliant on trading income or voluntary effort.

What does this mean for local authorities?

The COF evaluation suggests several implications for local government.

First, community ownership can deliver real social value, particularly in maintaining assets that would otherwise be lost. For councils facing financial pressure, this provides an important mechanism for sustaining local infrastructure.

Second, capacity and capability are the critical enablers. Local authorities have an important role in supporting community organisations through advice, development support and partnership working, rather than simply acting as funders or commissioners.  Central government should invest in this support.

Third, early investment in support pays off. The evaluation shows that changes to improve accessibility made a tangible difference. This suggests that programmes should prioritise capability-building alongside capital funding.

Fourth, equity requires active intervention. Without targeted support, community ownership risks reinforcing existing inequalities between places and groups. Local authorities are well placed to identify and support communities that might otherwise be excluded.

Finally, sustainability must be considered from the outset. Transferring assets to communities is not a cost-free solution; it shifts responsibility rather than removing it. Ongoing support, realistic business planning and long-term partnership will be essential to ensure success.

Taken together, the evaluation suggests that what works in community ownership is not simply funding assets, but supporting communities to own, manage and sustain them effectively.

What works in mental health and employment support? Learning from recent evaluation

/ jlowther2015 / 1 Comment

Dr Jason Lowther

Alongside the growing evidence base in areas such as homelessness, local growth and skills, there is increasing attention on how local systems respond to the intersection of mental health and employment. This is a critical issue for local authorities and their partners, given the strong relationship between work, wellbeing and wider social outcomes.

The recent evaluation of Individual Placement Support (IPS) integrated within Improving Access to Psychological Therapies (IAPT), delivered through the Mental Health Trailblazers programme, provides a valuable contribution to this evidence. The programme was designed to address two linked challenges (poor mental health and unemployment) through a combined intervention offering both therapeutic and employment support. It was delivered across three areas (Blackpool, the North East and West London) as locally commissioned “growth deal” projects aimed at improving both economic and health outcomes.

An integrated approach to mental health and work

The overarching premise of the programme is simple but important: mental health and employment are interdependent. Traditional services often treat them separately, with employment support and mental health treatment delivered through different systems. The Trailblazers programme sought to integrate these approaches by embedding employment specialists alongside psychological therapy services.

The evaluation employed a mixed‑methods design, combining an impact evaluation – using a trial‑style comparison between IAPT alone and IAPT plus IPS – with a detailed process evaluation examining implementation, service design and user experience. This dual approach reflects the complexity of the intervention: it is not only about whether outcomes improve, but about how services work together in practice.

Findings on outcomes: modest but promising

On outcomes, the evaluation presents a cautious but broadly positive picture. While the evidence on impact is not definitive, there are indications that combining IPS with psychological therapies can support improvements in both employment and mental health outcomes compared to standard provision.

This aligns with a wider evidence base for IPS, which consistently shows strong performance in helping people with mental health conditions move into and sustain work. Employment itself is recognised as beneficial for recovery and wellbeing, reinforcing the rationale for integrated approaches.

However, the evaluation also highlights the difficulty of demonstrating impact in complex, real‑world settings. Data limitations, variation in local models and challenges in maintaining experimental control all affected the strength of conclusions. This is a recurring issue across many local public service evaluations: outcomes are shaped by multiple interacting factors, making attribution difficult.

What makes the model work in practice?

The process evaluation provides rich insight into the mechanisms behind the model. Several features emerge as particularly important.

First, the relationship between service users and employment specialists is central. IPS is explicitly client‑led, focusing on individual preferences, strengths and readiness rather than predefined pathways. This personalised, relational approach appears to be a key driver of engagement.

Second, integration between services matters. Embedding employment specialists within IAPT teams helped create a more holistic offer, reducing fragmentation and enabling better coordination of support. Where integration was stronger, services were better able to respond to the complex and fluctuating needs of clients.

Third, the model benefits from being less target‑driven than traditional employment programmes. The evaluation notes that a focus on client needs, rather than rigid job outcome targets, enabled more sustained engagement – particularly for individuals with more severe or complex mental health challenges.

At the same time, implementation was not without difficulties. Referral processes, administrative requirements and clinical wait times all created friction in the system. Experiences and outcomes  were inconsistent.

The system challenge: integration is difficult

Perhaps the most important learning from the evaluation is about the difficulty of integrating services across organisational boundaries. Bringing together health and employment support requires alignment between different funding streams, professional cultures and accountability frameworks.

The evaluation highlights challenges such as eligibility criteria, information sharing and differences in service priorities. These issues are not unique to this programme; they reflect broader structural barriers within public services. Even where the case for integration is clear, delivering it in practice requires sustained effort and coordination.

There were also challenges in engaging employers and navigating local labour markets. Employment outcomes depend not only on individual readiness, but on the availability and quality of jobs. This again points to the importance of seeing mental health services within a wider economic context.

What does this mean for local authorities?

The evaluation offers several implications for local government and system partners.

First, integration across services is both necessary and challenging. The evidence supports the case for bringing together health and employment support, but also shows that this requires deliberate design, strong relationships and ongoing coordination.

Second, personalised, relationship‑based support is critical. Models like IPS work because they focus on individuals, not categories. This has implications for commissioning and performance management, which often rely on standardised models and metrics.

Third, employment should be seen as a health outcome. The evaluation reinforces the idea that good work is not just an economic goal, but a key component of wellbeing and recovery. This has implications for how local systems define success.

Fourth, local variation matters. The programme was delivered differently across areas, reflecting local labour markets, service configurations and partnerships. This flexibility is a strength, but also makes evaluation and scaling more complex.

Finally, the evaluation highlights the importance of longer‑term thinking. Supporting people with mental health conditions into work is not a short‑term intervention. Outcomes take time to emerge, and services need stability to build the relationships and capability required.

As with other areas of local government, the evidence increasingly shows what works in principle. The challenge is less about identifying effective models, and more about creating the conditions – organisational, financial and cultural – that allow them to be implemented flexibly and at scale.

What works in local growth and skills? Learning from recent evaluations

/ jlowther2015 / 1 Comment

Jason Lowther

Following the previous blog on homelessness and rough sleeping, this piece turns to another major area of local government activity: local growth and skills programmes. Here too, evaluation activity has expanded rapidly, with a mix of national frameworks, programme‑level syntheses and place‑based studies. Taken together, these evaluations offer a valuable, and still evolving, picture of what is working, what is proving harder, and what local systems actually need to deliver economic outcomes.

Four strands of evidence stand out.

MHCLG local growth evaluation

The MHCLG local growth evaluation programme is significant not just for its findings, but for its approach to evaluation itself. Rather than focusing on single programmes, it introduces a portfolio‑level strategy covering multiple funds aimed at improving sub‑national economic performance.

Recent work, including the process evaluation of the Local Growth Fund and Getting Building Fund, highlight both strengths and tensions in the model. Decentralised decision‑making and the “single pot” approach enabled locally tailored investment and stronger alignment with local strategies. Private sector involvement and local prioritisation were widely valued.  However, delivery was shaped by pressures to deliver “shovel‑ready” projects quickly, particularly in the Getting Building Fund, which sometimes limited strategic coherence and innovation. Governance arrangements, while locally responsive, were often complex, and approaches to monitoring and evaluation were variable. More broadly, the evaluation underlines the difficulty of measuring long‑term economic impact, particularly where interventions are diverse and outcomes unfold over many years.

Multiply deep dives (Scotland, Wales, Northern Ireland)

The Multiply deep dives bring a skills and employability perspective, focusing on adult numeracy provision across the devolved nations. Multiply was a £559 million UK‑wide programme designed to improve functional numeracy, with flexible, locally designed delivery models.

The deep dives use qualitative case studies, interviews with delivery partners and analysis of monitoring data, focusing on one area in each nation and drawing on wider place‑level evidence. A central finding is that local flexibility enabled innovation, particularly in embedding numeracy in real‑world contexts such as employment, parenting or financial capability.

At the same time, the evaluations highlight familiar delivery challenges. Short delivery timescales, in some cases just a year, created pressure to scale quickly, often leading to adaptation of existing provision rather than genuinely new approaches. Partnership working across councils, colleges and the voluntary sector was essential but time‑consuming to establish. Engagement with target groups remained difficult, particularly where low confidence rather than low skill was the primary barrier.

Overall, the evidence suggests that contextualised, learner‑centred approaches are promising, but require time, trust and sustained funding to embed.

UK Shared Prosperity Fund (UKSPF) interim evaluation synthesis

The UKSPF interim synthesis report provides perhaps the most comprehensive current view, drawing together 34 place‑based evaluations across the UK. It focuses on process learning rather than impact, reflecting the relatively early stage of delivery.

A clear headline is the importance of local autonomy. Across almost all areas, the ability for Lead Local Authorities to design interventions around local needs was strongly valued, particularly compared to the perceived rigidity of previous EU funds. This flexibility supported alignment with local strategies, more responsive delivery, and better integration across policy areas.

Other success factors included strong local programme management teams, continuity of provision (using UKSPF to sustain previously funded services), and the ability to combine funding streams to create coherent local offers. However, challenges were equally consistent. Tight central government timelines constrained planning and procurement, limited consultation, and created recruitment difficulties. As with other programmes, evaluation and outcome measurement remained underdeveloped.

The synthesis highlights a key tension: local freedom within central constraints. While devolution of decision‑making was real, the operating environment still imposed significant limits on what places could achieve.

UKSPF place‑based evaluations

The place‑based evaluations add depth to this picture by examining how UKSPF worked in specific localities. Using mixed‑methods approaches – including contribution analysis, surveys, interviews and case studies – across 34 areas, they explore how combinations of interventions interact within local systems.

These studies show that outcomes are highly context‑dependent. In some areas, UKSPF supported visible improvements in community facilities, local business support, and employability outcomes. In others, impacts were harder to detect, reflecting both the early stage of delivery and the complexity of local economies. What emerges clearly is that programme success depends less on individual projects than on how they are aligned and sequenced locally.

The evaluations also reinforce the importance of existing capacity and partnerships. Areas with mature governance arrangements, strong voluntary sector links, and prior experience of managing regeneration funding were better able to mobilise quickly and deliver coherent programmes.

What does this mean for local authorities?

Across these evaluations, several consistent lessons emerge.

First, local flexibility works, particularly when supported by capacity and stability. Both UKSPF and Multiply demonstrate the value of devolved decision‑making. However, the benefits are uneven, depending on local capability, existing partnerships, and the time available to plan and deliver.

Second, time is the missing ingredient in local growth policy. Tight delivery timescales appear across all programmes, driving a focus on “shovel‑ready” activity, limiting innovation, and constraining partnership development. Economic change, skills development and behaviour change all take longer than funding cycles typically allow.

Third, integration matters more than individual interventions. The strongest evidence, particularly from the place‑based evaluations, is that impact depends on how interventions fit together. Skills, business support and community investment are interdependent, yet funding streams and evaluation frameworks often treat them separately.

Fourth, measurement remains a weak spot. Across the local growth portfolio, there are persistent challenges in demonstrating impact and value for money. This is partly methodological, but also reflects the reality that many outcomes (productivity, employment, resilience) are long‑term and influenced by wider factors.

Finally, these evaluations underline a familiar but important point: local systems deliver national priorities. Where programmes align with local strategies, build on existing partnerships and allow room for adaptation, they show promise. Where they are constrained by short timescales, fragmented funding or complex governance, delivery becomes more transactional.

The conclusions from the local growth and skills evaluations strongly align with, and are reinforced by last month’s excellent report from the Institute for Government, Designing and delivering employment support.  The IfG goes further in diagnosing why these issues persist and what structural reform is needed. Both emphasise the value of local flexibility, integration and tailoring to place, with the IfG explicitly arguing that strategic authorities are best placed to design joined‑up employment support aligned to local labour markets and services. Likewise, both bodies of evidence highlight fragmentation and poor coordination across programmes as major barriers, with the IfG noting longstanding failures to “shift the dial” despite multiple national schemes, echoing local growth evaluations on disjointed funding and siloed interventions. The IfG report places significant emphasis on the limits of centralised systems and the need for multi‑year funding, capability and accountability frameworks.

In short, the local growth evaluations provide grounded evidence of what works in practice, while the IfG report offers a more explicit systems diagnosis: that without sustained devolution, integration and long‑term investment, the conditions needed for those “what works” approaches to succeed will remain constrained.

What works in homelessness and rough sleeping

/ jlowther2015 / 1 Comment

Jason Lowther

We’ve just started a new series of Inlogov blogs summarising the growing body of evaluation evidence in local government, and what it tells us about how councils are learning what works. Nowhere is that learning more urgent or more complex than in homelessness and rough sleeping. A series of recent national evaluations provide a rich, if sometimes uncomfortable, picture of how programmes are operating on the ground, what difference they are making, and where the system continues to struggle. 

Systems‑wide evaluation of homelessness and rough sleeping

The systems‑wide evaluation represents a deliberate attempt to step back from individual programmes and understand homelessness as a complex, interconnected system rather than a set of discrete services. Led by the Centre for Homelessness Impact with research partners, the early reports combine systems‑mapping, policy analysis and qualitative fieldwork in five local authority areas.

The core finding is stark: the system is not working as intended. Public spending and organisational effort are disproportionately focused on crisis response rather than prevention, even though this approach places increasing pressure on local authorities and delivers poorer outcomes. Fragmented funding streams, short‑term grants and inconsistent incentives across departments actively undermine joined‑up working. The evaluation does find examples of strong local partnership practice,but these are often working around the system rather than being supported by it.

The central conclusion is that meaningful progress requires sustained, cross‑government commitment to prevention, better alignment between housing, health, justice and welfare systems, and a clearer understanding of how national policy choices shape local outcomes.

Rough sleeping and complex needs process evaluation

The rough sleeping and complex needs evaluation zooms in on services supporting people facing the most entrenched disadvantage. This process evaluation examined interventions funded through the Rough Sleeping Grant and Rough Sleeping Social Impact Bonds, focusing on people with co‑occurring mental health and substance misuse needs.

Using case studies across 12 areas, interviews with service users and staff, and cost analysis, the evaluation explored how different models worked in practice. It found that progress, including improved housing stability and engagement with services, was most likely where support was flexible, persistent and relationship‑based. Small caseloads, psychologically informed approaches and multi‑disciplinary working were all important.

However, delivery was often hampered by structural barriers beyond local control: gaps in mental health provision, restrictive criteria in mainstream services, workforce instability and the limitations of short‑term funding.

Rough Sleeping Initiative process evaluation

The Rough Sleeping Initiative (RSI) process evaluation complements earlier impact analysis by explaining how and why the initiative achieved results. While the impact evaluation estimated a significant reduction in rough sleeping in RSI areas, the process evaluation explored local delivery through surveys of all funded authorities and in‑depth qualitative case studies.

Local authorities consistently reported that RSI funding enabled them to expand outreach, create specialist roles, and strengthen partnerships with health and voluntary sector providers. Rapid mobilisation, local flexibility and visible political commitment were key strengths. Many areas highlighted the value of multi‑disciplinary teams and assertive outreach in engaging people who had been sleeping rough for long periods.

At the same time, the evaluation identified familiar challenges: recruitment difficulties, reliance on short funding cycles, and the risk of losing skilled staff when funding ends.

Next Steps Accommodation Programme evaluation (briefing paper)

The Next Steps Accommodation Programme (NSAP) evaluation focuses on what happened after the incredible emergency response of “Everyone In” during the Covid pandemic lockdowns. It draws on two waves of interviews with service users in 34 local authorities, around 12 and 18 months after they were placed in longer‑term accommodation.

The findings are cautiously positive. Many people reported improved stability, safety and wellbeing, and a reduced use of emergency services. Sustained accommodation outcomes were more likely where individuals received ongoing, tailored support alongside housing. However, the evaluation is clear about its limits: the absence of baseline data, attrition over time, and the likelihood that those still in contact with services are the “success cases”.

Crucially, it highlights risks to sustainability, particularly affordability pressures, isolation, and unmet support needs.

Capital Letters process evaluation

The Capital Letters process evaluation adds an important organisational and commercial dimension to the evidence base. Capital Letters was established in 2019 as a borough‑owned, non‑profit company to reduce homelessness and temporary accommodation use in London by collectively procuring private rented sector homes, reducing competition between boroughs and driving better value for money.

The evaluation, based on interviews with boroughs, board members, landlords and MHCLG officials, alongside document and performance data review, focuses on how the initiative was set up, governed and sustained. It finds that Capital Letters had early success in demonstrating the potential of scale, shared negotiation and coordinated landlord engagement. Boroughs valued the ambition to change market dynamics rather than simply manage them.

However, the evaluation also highlights significant challenges. Achieving financial self‑sufficiency while meeting social objectives proved extremely difficult, particularly in a highly pressured London housing market. Tensions emerged between commercial risk, borough expectations and the constraints of homelessness legislation. Governance and accountability arrangements were complex, and uneven borough engagement limited the company’s ability to operate at the scale originally envisaged. Ultimately, despite grant support, Capital Letters was unable to become financially sustainable and subsequently wound down operations in 2025.

What does this mean for local authorities?

Read together, these evaluations paint a coherent but challenging picture of what local government can do in homelessness and rough sleeping.

First, prevention and system change demand stability. The biggest barriers to progress lie beyond individual projects. Short funding cycles, fragmented policy levers and misaligned national incentives consistently undermine local efforts, even where practice is strong. Councils can innovate, but without longer‑term certainty the system pulls them back towards crisis response.

Second, relationships and capability are critical. Outreach teams, assertive support for people with complex needs, and sustained tenancy support all rely on skilled staff, trust and persistence. These are precisely the elements most at risk from time‑limited programmes and competitive commissioning.

Third, housing supply and affordability are constraints. The Rough Sleeping Initiative and Next Steps evaluations both underline that service innovation only works when there are viable move‑on options. Capital Letters reinforces this at a system level: even ambitious collective approaches struggle when the underlying market is stacked against local authorities.

Fourth, collaboration is necessary but tricky. Whether through multi‑disciplinary teams or borough‑owned companies, partnership working requires time, governance capacity and shared risk. The evidence suggests collaboration works best when it is supported by clear national frameworks, realistic financial models and space to mature, rather than when it is required to develop and achieve results rapidly.

Finally, these evaluations show the growing value of learning‑focused evaluation in local government. They do not offer simple answers or “magic bullet” models. Instead, they help councils articulate what they are already experiencing on the ground and provide credible evidence to challenge policies and funding arrangements that make homelessness harder, not easier, to resolve.

The learning is no longer about whether local authorities know what works. It is about whether the wider system will allow them to do it.