Research and analysis

Evaluation of the Refugee Transitions Outcomes Fund

Published 26 September 2025

Executive summary

The Home Office commissioned Ecorys, leading a consortium including Social Finance and a group of academics, to carry out evaluation and learning activities for the Refugee Transitions Outcomes Fund (RTOF). This executive summary details key information about the RTOF, key findings, conclusions and lessons learned.

About the RTOF

The RTOF (from September 2021 to March 2024) was a cross-government initiative that aimed to increase the integration and self-sufficiency of refugees. The RTOF accessed up to £13.99 million (from the Shared Outcomes Fund and the Life Chances Fund) to pilot approaches to improve employment, housing and wider integration outcomes for around 1,800 newly recognised refugees. The Fund also aimed to test innovative delivery partnerships, identify effective practices, improve the evidence base concerning integration, and diversify funding to the refugee support sector.

Four projects in high asylum seeker dispersal areas mobilised in 2022, all adopting Social Impact Bond (SIB) structures (a form of outcomes-based commissioning). Projects comprised social investors, regional migration partnerships, local authorities and service providers (commonly voluntary community and social enterprise organisations (VSCEs)).

Key findings

Employability and employment

The RTOF significantly improved the employment status of refugees supported, with participants 38.5% more likely to find work than those not participating in the RTOF. This reflects a relative increase in the odds of employment for participants compared to non-participants. Translating this into absolute impact, 135 job entries (22.1% of those employed in the treatment group) could be directly attributed to the intervention, after accounting for baseline employment levels in a comparable group without treatment. While the odds ratio highlights the increased likelihood of employment, the additional jobs calculation provides a clearer measure of the programme’s direct contribution to creating employment opportunities.

In terms of sustained employment outcomes, out of the quarter of RTOF participants who found work, 78% were still in employment after 3 months. Participants were more likely to sustain work if they had higher English language proficiency, were in safe and secure accommodation and possessed prior work experience.

Limited English proficiency represented a key barrier to employment for refugees, which often relegated them to low-skilled positions and limited job prospects to roles requiring minimal English interaction. Evaluation evidence thus highlights that enhancing English language support is crucial to enable refugees to secure meaningful work and integrate into the community.

Mental health problems emerged as another key barrier to employment, with many participants experiencing low self-esteem and mental health challenges that hindered job applications and could exacerbate the negative effects of unsuccessful applications. Caseworkers delivering RTOF played a key role in helping to build confidence through providing empathetic support, confirming the importance of broad based and holistic support in helping refugees to move closer to, and into, work.

A third of participants improved their employability while participating in RTOF through engaging in volunteering or work experience, alongside achieving wider employability focused goals. Volunteering or work experience opportunities could be beneficial for participants with greater barriers to finding work but need to be appropriate and meaningful.

Engaging varying types and sizes of employers created valuable opportunities to support the employability of participants, as well as providing direct access to subsequent job opportunities. While requiring significant resourcing, employer engagement activities were seen as highly beneficial in developing employability and hence supporting integration.

Finding safe and secure housing

Over half of the cohort entered safe and secure accommodation while participating in the RTOF. Of those who entered accommodation, 71% were still in safe and secure housing after 6 months. Structural issues, notably the scarcity of social housing and affordable private rentals, posed substantial challenges. Participants frequently failed to meet social housing criteria or encountered lengthy waiting lists, while the private rental market was largely unaffordable.

Evidence indicates that participants in safe and secure accommodation were, all else being equal, twice as likely to enter employment, demonstrating the importance of such accommodation as a precursor to employment entry. RTOF participants often needed ongoing support from staff to access and sustain tenancies, including help with understanding legal and financial documents, indicating the importance of such support in attaining and maintaining secure housing.

The lack of suitable and affordable housing was a critical barrier to long-term stable housing outcomes, however, with access to the private rented sector requiring stable incomes, deposits, credit histories, and references. Accessing social housing, meanwhile, was contingent on prioritisation criteria that often disadvantage single individuals who were common amongst the RTOF cohort. Participants expressed a preference for social housing over private rentals due to its perceived accessibility and stability but were commonly unable to access such accommodation.

Wider aspects of integration

Most participants achieved outcomes related to broader integration areas, with 95% of the cohort creating goals and a plan to achieve them and 71% of the cohort improving on their goals over 6 and 12 months of RTOF support. RTOF participants reported benefits such as increased confidence and access to services and community activities, as well as improved wellbeing. The high levels of achievement in terms of setting integration goals within integration plans can be explained by these activities being a core element of onboarding and delivery.

Social activities were designed to engage participants and foster trust with professionals delivering support, while acquainting refugees with their localities and community members, all of which proved beneficial. These activities strengthened social ties and community awareness and could boost participants’ confidence and sense of belonging.

The evaluation evidence highlighted that integration is not linear for any individual or cohort, being multidimensional and interrelated. Participants felt that employment and wider integration were highly contingent on accessing safe and secure housing, accessing to English language tuition (such as English for speakers of other languages (ESOL)) and, in some cases, mental health support. Aligned with this, most participants valued activities relating to housing the most, with this providing them with much needed stability.

Value for money (VfM)

The RTOF was estimated to reach a positive return on investment shortly after 3 years under the scenarios modelled, restricting the assessment to fiscal benefits only and using at the central assumption for employment benefits (derived from the counterfactual impact evaluation’s estimate of the RTOF’s additionality), and restricting the assessment to fiscal employment and housing benefits. When also including economic and social benefits, a positive return on investment was estimated to have been achieved after one year. As such, the cost-benefit analysis undertaken confirmed the invest-to-save logic of the RTOF. Additional benefits are also likely to exist, related to broader aspects of integration achieved beyond housing and employment. Even though these are unable to be easily monetised, they further highlight the VfM generated by the RTOF.

Effects on the system

The RTOF also achieved some system-level outcomes such as improved relationships between projects and employers (through employer engagement activities), enhanced local provision for newly recognised refugees, and further development of partnerships able to support this cohort.

Effects of the SIB funding mechanism

The SIB structures adopted fostered collaboration and knowledge exchange among providers and investors, enhancing partnership working and understanding of SIB operations. The outcomes-based funding mechanism involved enabled the flexible use of funding, facilitating person-centred approaches and adaptive management to achieve outcomes. The SIBs drove an outcomes-focused, data-driven approach among providers and investors, improving efficiency despite variations across projects due to factors including data system limitations alongside additional management and administration requirements.

However, the annual caps on funding available to finance RTOF outcomes were restrictive for projects, as funds could not be transferred between years. This posed notable challenges for an outcome-based payment approach where achievement of outcomes increased over the programme lifetime, with many coming in years 2 and 3 of delivery; it was also acknowledged that profiling of expected annual outcomes was based on limited evidence.

Gathering documents to evidence outcomes was also a notable challenge for RTOF projects as it could require the input of external stakeholders, such as employers or landlords, and the motivation and understanding of participants themselves.

Programme design and management

Programme design and management generally worked effectively. However, greater clarity on administrative processes, eligibility criteria and evidencing requirements would have been helpful from the outset, as well as clearer and more transparent communications between all stakeholders. More time would also ideally have been allocated to mobilisation phases, particularly where new partnerships were forming within new contracting and financing structures.

Participant engagement and support

Limited language learning opportunities (such as English for speakers of other languages) and long waiting lists were key constraints for RTOF delivery, in the context of language proficiency being key to achieve employment and wider integration outcomes.

Conclusions

The RTOF was effectively implemented, achieving positive outcomes for participants, staff, organisations, local refugee support services and government. However, challenges were faced due to uncertainties over referral and engagement numbers, a perceived lack of flexibility to adapt and change delivery models, and constraints from annual caps on funding. Despite this, the finding that participants were 39% more likely than comparable non-participant refugees to enter work during the programme’s lifetime confirms the effectiveness of the approach in the context of employment being a key dimension of integration. Likewise, the housing outcomes achieved for many within the cohort, in the context of structural constraints around limited accommodation availability, was a similarly positive achievement of the programme and likewise a key pillar of successful integration.

The RTOF also demonstrated effectiveness in terms of supporting refugees with their broader integration goals, beyond employment and housing, with contributions to participants’ social integration and support for wellbeing being key features of positive outcomes. Despite the positive outcomes achieved, the extent to which participants became self-sufficient during the timescales of RTOF delivery appeared to vary. Longer-term programming and evaluation could provide the opportunity to support and further investigate progress towards full self-sufficiency.

Finally, the SIB models adopted had clear benefits, in terms of removing financial risk from providers that would be faced in other outcomes-based funding approaches, while encouraging a focus on achieving key employment, housing and broader integration benefits. However, the use of annual payment caps, limiting the support that could be offered and payments claimed in the final year of delivery, was a key drawback and learning point for future similar models.

Lessons learned

Key lessons from the RTOF pilot and its evaluation of are presented below. The lessons identified are structured according to their likely relevance for particular stakeholder groups in the context of considering future interventions, though may be applicable across them in some cases.

For government

When designing future programmes, as far as possible it is likely to be beneficial to avoid use of annual payment caps, particularly where profiling engagement numbers and outcomes achievement accurately is challenging.

Governance and management of the RTOF benefitted from broad stakeholder engagement, in particular to ensure transparent and effective communications alongside mitigating risks.

The focus on employer engagement as part of delivery proved effective in sourcing work placements and job opportunities

Achieving positive outcomes amongst the refugee cohort can take time. Where possible, longer (funding) timeframes could thus be considered, particularly if activities are intended to facilitate longer term outcomes, such as self-sufficiency.

Allocating sufficient time for project mobilisation was important, particularly where new partnerships were forming within new contracting and financing structures.

Effective systems and mechanisms for data capture and sharing that are joined up across all providers are important to track individual participant situations and implementation progress.

Limited ESOL provision and long waiting lists were key constraints for the RTOF in the context of language proficiency being central to achieving integration outcomes.

For investors

There were clear positives in terms of the RTOF’s approach to partnership working, although regular and clear communication was needed to ensure alignment with delivery requirements.

Specific requirements related to eligibility criteria, and the claiming of outcomes, need to be fully communicated to all providers, including the potential consequences of not achieving progress targets, with sufficient guidance provided to avoid any misinterpretation.

For local authorities (LAs) and VSCE providers

Some projects were successful in engaging landlords in the private rented sector to mitigate the need for credit checks and references, as well as providing up-front payments for rent and deposits, indicating potentially successful approaches if resources and local contexts permit.

Given the centrality of language proficiency to integration, facilitating greater opportunities for English language development within projects, such as buying-in ESOL provision or providing longer term language development activities, could be considered.

Group activities can be beneficial in offering refugees chances to develop English language skills and build social connections, which can have wider benefits for participants’ wellbeing.

Achieving employment outcomes can take substantial time and resource; other needs are likely to require addressing first (particularly housing and improving English language) as key precursors to finding work.

Engaging participants at the earliest opportunity was helpful to minimise the escalation of challenges and barriers to progress, and to ensure participants had a greater length of time to access support services; providing information about life in the UK to people waiting for decisions on applications to remain, for example, may be beneficial.

Ongoing proactive check-ins with RTOF participants helped projects to be aware of issues escalating; while this required substantial resource in the short-term, it prevented issues needing more in-depth support later.

Working flexibly with participants and forming strong, trusting relationships was key to positive engagement; providers needed to be willing to reschedule appointments regularly and undertake significant activity to engage those less inclined to participate in activities.

Volunteering and work experience opportunities can clearly be beneficial for refugees, though ensuring they were matched to suitable volunteering opportunities was important.

RTOF participants could be unclear on available support, particularly in the context of several organisations engaging with them; this highlights the benefits of ensuring there is a clear participant journey established from the outset that can be communicated clearly, including outlining what may be required to evidence intended outcomes where applicable.

1. Introduction

The Home Office commissioned Ecorys in consortium with Social Finance and a group of academics, to provide evaluation and learning support to the Refugee Transitions Outcomes Fund (RTOF). This final report presents the findings from evaluation activity undertaken between 2021 and 2024.

1.1 RTOF background and context

Refugees in the UK can face multiple integration barriers, such as high levels of unemployment, low earnings, inadequate access to secure housing, limited English language skills, few social connections, poor health, and digital exclusion (Atto, Hirst and Hall 2020). Research from 2019 shows that people who came to the UK seeking asylum are less likely to be in employment than people born in the UK and other migrant groups; the employment rate among refugees who had sought asylum was found to be 51%, compared with 73% for those born in the UK (Kone, Ruiz, and Vargas-Silva 2019).

Linked to challenges in entering the labour market, broader integration barriers have negative implications for peoples’ lives, wellbeing, participation in society and prospects. Refugee integration must therefore be addressed as a multi-dimensional issue, with employment, housing and wider elements of integration being interdependent. Integration has many dimensions and is context-specific, involving the shared responsibility of individuals, institutions, organisations, and communities. However, less is known about what works to support successful integration and self-sufficiency, including which funding models are most effective (Home Office 2019).

Responding to this context, the RTOF was launched in 2021 (running until 2024) as a pilot programme to support refugees to find employment and stable housing, while also supporting their broader integration. Alongside supporting refugees, the RTOF sought to facilitate the development of innovative delivery partnerships, as well as identifying and sharing effective practices. The pilot also aimed to improve the evidence base about integration more broadly, diversify funding to the refugee support sector, test solutions, and increase capacity to deliver integration services. The RTOF accessed up to £13.99 million to improve employment, housing, and wider integration outcomes for approximately 1,800 newly recognised refugees. Funding came from 2 sources: up to £10 million from the Shared Outcomes Fund (HM Treasury) and up to £3.99 million from the Life Chances Fund (DCMS).

Four projects in high dispersal areas (areas where greater volumes of asylum seekers and refugees are housed) were selected through competitive tender, mobilising in 2022 after some initial contracting delays. Project partnerships included social investors, regional bodies, LAs, and providers, most commonly VCSEs. The projects were:

• Greater Manchester Combined Authority (GMCA) with Big Issue Invest (BII) operating across 10 LAs
• West Midlands with BII operating across 4 LA areas
• Plymouth Refugee Opportunities (PRO) with Bridges Outcomes Partnership (BOP) operating across one LA area
• North East Refugee Integration Support and Employment with BOP operating across 11 LAs

Projects were able to select their preferred payment mechanism. All successful bids chose SIBs. Traverse (a social purpose consultancy) delivered capacity building support to projects during mobilisation. The Home Office engaged Palladium International Limited to complete the selection process of projects, later taking the administration of the RTOF ‘in-house’ at the Home Office.

1.2 Aims of the RTOF

The aims of the RTOF were to:

• enable innovative and place-based approaches to support newly recognised refugees in selected areas
• improve employment, housing, and wider integration outcomes for newly recognised refugees in the selected areas and increase refugee self-sufficiency
• generate evidence about what works, to evaluate the impact of projects on refugee integration and self-sufficiency
• generate cross-government savings
• provide evidence to inform future policy, programming, and mainstream service delivery

1.3 Evaluation approach

1.3.1 Aims

The key aims of the evaluation were to:

• generate learning about what works to improve employment, housing, and wider integration outcomes for newly recognised refugees
• evaluate whether and how the outcomes funding mechanism and involvement of social investors led to more efficient or effective delivery
• test the economic case for future investment in refugee integration services

1.3.2 Methodology

The evaluation comprised several key pillars:

1. Coping and preparation activity included a document review, supporting with the selection of projects, scoping consultations, Theory of Change development and assisting with development of the Shared Data System (SDS). The SDS was built in consultation with projects and the government to collate data submitted by projects about the demographics and outcomes of individual participants.
2. A process evaluation explored RTOF implementation via interviews with government stakeholders and project stakeholders at 3 points in time, and RTOF participants at up to 2 points in time. Interviews were supplemented by insights from peer consultations conducted by Refugee Steering Group members, alongside a review of participants’ case files and data from the SDS. A steering group comprising refugees with lived experience of seeking asylum and being a refugee was recruited to help inform the evaluation design, as well as conducting consultations with their peers to generate additional evaluation insights.
3. Developing an Impact Evaluation involved a feasibility study to explore potential approaches to assessing the RTOF’s impact; a counterfactual impact evaluation, using a quasi-experimental approach, was implemented to the estimate the effect of RTOF on employment entry. This was supplemented by theory-based evaluation, drawing on qualitative research with a range of stakeholders and participants and analysis of SDS data, to explore the RTOF’s broader outcomes.
4. The ‘Value for money’ assessment methods included cost benefit analysis (CBA) - to monetise the Fund’s outcomes as ‘benefits’ and to compare these benefits with programme costs - and cost effectiveness analysis (CEA) - to compare metrics of value for money across related projects.
5. Learning workshops for projects were facilitated by Social Finance and were held quarterly, informed by analysis of SDS data. ‘Insights’ reports were produced following the workshops and shared with projects and the Home Office, adding additional evidence for the evaluation.
6. Academic partners advised on the methods and materials for the evaluation. A Refugee Steering Group, engaged for the evaluation and comprising people with lived experience of seeking asylum and being a refugee, helped to shape evaluation approaches and carried out consultations with their peers to generate additional insights for the study.

1.3.3 Limitations

As with any large-scale evaluation, there are a few data limitations which should be considered when reading the report, in order of importance:

1. There were some inconsistencies in Home Office date formats in recorded data that had implications for matching the group participating in RTOF and the comparison group to compare employment outcomes.
2. It was not possible to establish a counterfactual (comparison group) for housing outcomes or any wider elements of integration.
3. The SDS was a key data source for the evaluation, capturing outcomes that were claimed by projects which were required to be evidenced. In respect of longer-term (sustained) outcomes, participants may have ended their contact with RTOF projects prior to achieving them, or the outcomes may have occurred after the final SDS data was collected. This means that achievement of these outcomes may be underestimated in the analysis.
4. The SDS had notable levels of missing data across some fields that limited the scope and depth of analysis possible.
5. Some participants only engaged with interviews at one point in time, meaning that understanding their progress through qualitative data was more limited than through the quantitative data.
6. Projects had limited capacity or were reluctant to provide access to information in case files. Only a small number of case files were reviewed, which were then supplemented with wider analysis of SDS data.

More detail on the methodology and limitations can be found in Annex A.

1.4 Report structure

The following sections cover:

• the design and implementation of the RTOF
• the SIB design and effects
• employment, housing and integration outcomes (using a theory-based impact evaluation approach)
• an assessment of the RTOF’s employment impact (based on a counterfactual impact evaluation) and value-for-money
• conclusions and lessons learned

A series of Annexes provide further information and context for the findings, including detailing the composition of the cohort supported by the RTOF.

2. Design and implementation

2.1 Key findings

Governance and management structures supported delivery at the Fund and project level, though this support could have been enhanced by greater collaboration between investor representatives and government stakeholders at the outset.

Participants were engaged through a range of referral routes, with referrals through Home Office and asylum accommodation providers (26%), along with self- or peer- referrals (27%) being most common.

Partnership working was identified as a key benefit of RTOF, with projects bringing together a range of providers and specialisms to enhance support provided to participants.

Benefitting from the SIB approach to funding project delivery, the RTOF enabled the participation of providers who may be precluded from participating in contracts with greater financial risk.

The SDS was most effective for managing performance and progress against targets (such as the achievement of outcomes), but less effective at generating more detail about individual participants. Where projects or individual providers did not have systems able to record and share data easily, this limited the effectiveness of ongoing monitoring at the project and Fund levels, with implications for both project management and evaluation.

Project activities were mostly delivered as intended, with some modifications to the scope, focus and providers selected for delivery, with these aspects facilitating adaptive management to respond to emerging challenges and needs throughout delivery.

Participants generally had positive experiences of engagement, and were grateful for the support provided, particularly around finding accommodation due to the availability and cost of housing and the pressing need to find accommodation following the 28-day move-on period.

Integration activities tended to focus on more practical elements and support, such as opening bank accounts and registering with GPs, which differed slightly from the original conception of integration support that had a greater focus on building social connections, enrichment activities and wellbeing.

2.2 Design of the Fund

The RTOF was designed to primarily focus on improving employment outcomes for participants, alongside supporting with housing and wider integration to offer a holistic service that recognised the connections between various key integration elements. Projects delivered a range of activities to meet the needs of participants, covering employment, housing and wider integration support (including support for language development, social connections and wellbeing). Further detail on findings of the process evaluation relating to RTOD design and implementation, and the delivery models and approaches adopted in each RTOF area can be found in Annex B.

The eligibility criteria for participants included people who:

• received refugee or humanitarian protection status after 24 February 2020
• were economically active (actively seeking work or ready to take active steps towards securing work) and therefore needed to:
  • be of working age (18 or over, and under the State Pension age)
  • not be in full-time education
  • be available for work and actively taking reasonable steps to look for work, and
  • not have an illness or disability that stopped them from working

2.2.1 Governance and management

A quarterly RTOF Programme Board was attended by cross-government partners and, latterly, investors. The fact that the Board did not include investors from the outset was perceived by some stakeholders as a missed opportunity. Including investors was cited as useful to understand performance, challenges and to be able to directly ask questions, while investor and government representatives felt that greater involvement of, and collaboration amongst, stakeholders at the Programme Board would have been beneficial from the outset.

Project-level governance and management varied across projects but generally included investor level governance, monthly whole-partnership meetings, and one-to-one engagement between LAs (with strategic oversight) and VCSE providers. Projects perceived these structures to be effective for bringing providers together and for enabling partnership working. However, some delivery partner staff did cite a view that the additional layers of management and administration associated with outcomes-based financing was ‘cumbersome’ and time consuming and could detract from frontline delivery.

2.3 Implementation of the Fund

2.3.1 Referral and engagement patterns and processes

Overall, referral and engagement processes were generally perceived as effective by project stakeholders because, over time, projects were able to reach their engagement targets. However, inaccuracy of data about prospective participants caused issues through some routes, as discussed further below.

Over a quarter of RTOF participants (26%) were referred through data from the Home Office and providers of accommodation for asylum seekers. In some cases, however, contact details were out of date so contact could not be made to engage these potential participants. Where data was accurate and up to date, these referral sources were seen as effective in enabling projects to contact potential participants at the earliest opportunity. It was evident that participants’ situations could be more acute, for example regarding housing needs, when engaging later through LA or VCSE routes. Around a quarter (27%) of participants approached projects themselves or via peers. This suggests some awareness of the RTOF within the community although participants could also become aware of the programme through other services (for example Jobcentre Plus, or local refugee organisations). Projects also sourced referrals through leveraging connections with local VCSEs in a smaller number of cases (13%).

Project representatives reported that initial recruitment was sometimes rushed to meet targets, as targets would influence later outcomes payments. Ultimately, however, some project staff felt this contributed to slow progress in achieving outcomes or, in some cases, participants disengaging due to what were perceived as inappropriate engagements. Likewise, fewer individuals were granted refugee status or humanitarian protection than estimated by investors, LAs and providers at design phase, with this being linked to the backlog of asylum claims. This led to a decision to work with some people who were ineligible in some projects, including single parents with caring responsibilities and those above retirement age. Some stakeholders also felt that the need for refugees to be ‘economically active’ to meet RTOF eligibility criteria unnecessarily restricted projects’ ability to work with pregnant women, those with young children or with disabilities who may later have been actively looking for work.

Effective approaches that supported engagement included coverage of a range of languages within staff teams (and strong links with communities and refugee community organisations); adopting flexible, person-centred approaches driven by participant interests (including social activities); and having accessible, conveniently located offices, or offering home visits for those with young children or those who may struggle with travel.

2.3.2 Disengagement

There were relatively high levels of disengagement from RTOF support, reportedly higher than projects anticipated, with the SDS recording 958 participants as having disengaged due to a range of factors. Project representatives suggested that disengagement often came when participants needed to provide evidence to validate outcomes, which was sometimes perceived as being “intrusive” or burdensome. Participants were also more likely to disengage after making progress or achieving particular goals. In particular, higher levels of disengagement after successfully achieving housing outcomes were noted by project staff. In some cases, participants moved away from RTOF areas due to perceived greater housing, work and/or education opportunities elsewhere and/or being closer to friends, relatives or familiar communities.

“Once people are more settled, they don’t feel like they need support anymore, so they drop off. Quite a few people have moved out of area, so we can’t support them anymore.”

Caseworker

2.3.3 Partnership working and data systems

RTOF project partnerships generally worked effectively. After initial challenges during early mobilisation, partnerships formed and settled into delivery successfully, with providers working as cohesive teams beyond organisational boundaries. A key finding was that many providers brought together through the RTOF would not have considered working together previously. This led to what was widely recognised as a new and positive partnership approach in some local areas. In more limited contexts, partnerships were less coherent and effective at certain points in delivery, particularly where there was continued staff churn, or limited management and oversight. Where there was underperformance or challenges, projects typically actively managed this by procuring new providers or rescoping activities and remits; this approach was generally seen by project stakeholders as effective in addressing the issues encountered.

Projects’ data systems varied in effectiveness in terms of supporting effective partnership working and delivery. Shared customer relationship management (CRM) systems were evidently effective for measuring progress, informing performance management and, in some cases, supporting case work across providers. However, these systems mainly focused on progress monitoring and recording outcomes achievement, rather than recording qualitative information about participants (such as case notes). This meant that more in-depth information could be recorded inconsistently, either on separate case management systems, retained mentally by staff, or noted in voluntary notes sections. More locally hosted data systems supported measuring progress and performance management but did not support casework across partners, as such information was only able to be accessed by individual providers. As a result of the above data capture and sharing issues, in some contexts delivery and partnership working was seen as less effective than it might have been.

2.3.4 Employment activities

From the provider perspective, offering practical, relationship-focused, holistic, and individualised support was seen as particularly important in supporting participants’ employability. Such approaches included going beyond traditional employability support by, for example, accompanying participants to interviews to reduce anxiety and being on-hand to offer support. This and other similar support was much valued by participants.

Project representatives likewise cited the importance of targeted, employment-focused volunteering or work experience, rather than simply volunteering without an immediate link to employment. The latter was reported to be less attractive or viable for participants, partly as many had already volunteered when they were awaiting a decision on their asylum claim. Identifying opportunities that matched individuals’ career aspirations was seen as key to engagement and as more likely to generate positive outcomes.

There was also widespread consensus across project stakeholders that employer engagement, although resource-intensive, was critical to securing work experience and employment opportunities for participants, while provider staff cited the importance of raising awareness of the benefits of employing refugees. To address this, staff talked to employers about what skills refugees can bring to their organisation, how to make recruitment more accessible and employment more inclusive, as well as answering questions and reassuring employers of refugees’ right to work. Project stakeholders also described substantial targeted outreach work to engage specific employers who could offer opportunities that matched participants’ skills, aspirations, and requirements for employment (for example, offering flexible arrangements for those with caring responsibilities). Such options included hospitality, NHS health and social care (who could offer shift work) and retail employers.

Limited English language proficiency was the main barrier identified by provider staff in terms of supporting clients into work and could restrict participants to low-skilled, entry-level work. The need to support clients with English to enable them to access meaningful employment, work experience and volunteering opportunities was thus seen as key. Participant interviews similarly indicated that English language was the main barrier to employment, having a notable effect on their confidence, leading in some instances to perceived unfair treatment at work, and negatively impacting prospects for employment (often leading them to jobs which required less interaction in English). Mental health issues were also clearly a barrier to finding work. Many participants reported low confidence and mental health problems impacting on their ability to apply for and get jobs, compounded by the employment process (including rejection) which further negatively impacted their mental health further. The RTOF caseworkers were reported to have been particularly helpful in this respect, building participants’ confidence alongside offering understanding and compassion.

“Mental health or trauma has been the reason for multiple people on the project not feeling able to stay in the accommodation that they were in. It’s been a key barrier to people entering employment and becoming isolated so really across all 3 [integration] elements [of the RTOF}. It has played a factor in people not being able to achieve outcomes and not being able to move towards independence. Having said that, what we’re starting to see now is some really nice examples of people who have had, a really difficult time psychologically throughout the programme. But because of the relationship and the trust between the case worker and the client, they’ve been supported where possible through the individuals’ really bad times. And then in those moments, the team have been really good at bringing it back and sort of gently reminding the individual why they’re there, and working on employment in particular.”

Manager

Lack of stable, secure housing close to employment opportunities was also identified as a further key barrier. Project representatives also noted that some participants who had entered employment faced a worse financial situation due to the impact of their working patterns on their universal credit allowance, resulting in them dropping out of employment. The increasing cost of supported housing once in work also deterred some participants from securing employment. Participants who were studying were also reported to be less likely to move into employment in some instances due to their learning commitments. However, delivery staff reported supporting participants to try to access flexible part-time work around their studies where possible, with this seen as being effective in building up experience.

2.3.5 Housing activities

Project representatives suggested that the transition from asylum accommodation into secure housing was the main priority for most participants upon engagement. This resulted in high levels of engagement with housing activities. Provider staff linked effectiveness in generating positive housing outcomes to established partnerships with housing providers. However, structural factors such as a lack of social housing and affordable private rentals were widely acknowledged as key challenges for delivery.

Participants often did not meet the eligibility criteria for social housing (particularly where health needs were not recognised) or were faced with long waiting lists for supported accommodation. In general, the private rented sector was described as unaffordable for most participants, with landlords requiring a secure income that covered the high costs of rent within a context of limited supply and high demand.

“It’s really difficult to find accommodation. Rents are very expensive and supported accommodation is not available for non-priority people in our area. There is some supported accommodation in [the area] that I refer to, but they’ve got quite a long waiting list as well.”

Caseworker

The shortage of housing options meant that some refugees were being placed in, or finding, accommodation that they perceived may be unsafe due to the prevalence of drug and alcohol abuse in proximity. Some participant interviewees expressed concerns at being housed close to drug use.

“Our area’s infested with drug users and drug dealers. And like I have mentioned before, the police station is literally 2 minutes away from us, so it’s not like [the police] are struggling to get out to us and all that…no matter how much we talk to the police, nothing has changed.”

Participant

Participants reported feeling unsafe in their area, having to phone the police, being subject to racism, and the area being inappropriate for children. In addition, many interviewees living in shared accommodation reported that they were reluctant to share a home and would prefer their own space. Many interviewees reported that their accommodation was too small for them and their families and/or was in disrepair (with issues including damp and mould, leaks, broken appliances or problems with utilities).

Where appropriate, projects did support clients to access private rented accommodation. However, this option was only feasible where participants had secure employment. In these cases, staff noted that providing support to participants on an ongoing basis was critical to enable them to manage their tenancies. This support had included advocating for tenancy rights and translating or explaining documents (such as utility bills or contracts).

While the Home Office only paid for outcomes achieved, some projects also sought to address the challenges encountered in securing housing by purchasing furniture, flooring, and other basic items with the RTOF funding. Project stakeholders described this as being important due to the lack of available furnished properties. However, there were mixed views about the use of RTOF for household items; some delivery partners did not use the funding this way, instead encouraging participants to access other provision specialising in offering household items.

Across project areas, interviewees rarely discussed housing activities that had been unsuccessful, instead citing wider contextual or structural challenges accessing housing. However, in one area, stakeholders identified that while their decision to focus on supported accommodation was helpful in the short term, it may act as a disincentive to work as rent would become unaffordable as soon as participants found employment.

2.3.6 Wider integration activities

Overall, wider integration activities were not considered to be as high a priority for participants as housing and employment. Project staff also highlighted that when participants entered employment, they could often disengage with wider integration support. Despite this, aspects of integration support were clearly welcomed and appreciated by participants, with such support being seen by provider staff as an important aspect of more ‘wraparound’ assistance as part of addressing barriers to employment, as well as supporting participants’ integration more broadly.

Social activities with the dual purpose of engaging participants in the RTOF support and building trust with staff, as well as helping participants to get to know their local area and meet other members of the community, were evidently effective approaches for engaging some participants. Provider staff reported that such an approach had been successful in helping participants to build social connections, alongside enhancing understanding of local communities and services. Such activities could have a large influence on participants’ sense of confidence, fulfilment, and belonging.

Projects designed their own templates for integration plans with some suggestions of items or activities to include from the Home Office. Stakeholders generally found the integration plans to be a useful way to record participants’ individual integration goals and keep track of progress. However, they also found them to be limited in the information that could be recorded. Often, details discussed verbally were not included on the plans, although it is unclear if this was due to the limitations of the integration plans themselves, IT systems, or staff capacity or motivation to update them.

As noted elsewhere, English language development was a key, though challenging, aspect of RTOF integration support. Long waiting lists and limited English for Speakers of Other Languages (ESOL) capacity were cited as leading to slower progress than providers hoped, while in some cases participants’ caring responsibilities and distance between their accommodation and college were also barriers to attending English classes. Project areas took different approaches to addressing these issues. Approaches seen as effective included securing additional English language lessons through VCSEs, sometimes alongside other integration activities, engaging ESOL-trained employment support workers to provide employer-specific language training on-site with partner employers, and facilitating 1:1 informal English conversations with service-users to build confidence in conversing. Some caseworkers reported using online translation tools to work with participants and lacked access to interpretation services, which could limit their ability to communicate where there was no shared language.

Accessing NHS services such as GPs, dentists and mental health services were also typically key integration goals for participants. However, project staff reported challenges accessing NHS dentists and mental health services in particular, with long waiting lists up to 4 years reported in one area. Poor mental health emerged as a key need across the RTOF cohort, with project stakeholders identifying this as a key barrier to achieving wider outcomes. Project staff and managers reported the strain that the prevalence of mental health issues could have on people working directly with participants.

2.3.7 Refugee experience of engagement

As indicated above, participants generally felt positive about their experiences of the RTOF projects. In particular, they appreciated the relationships with caseworkers and delivery staff, with trust and compassion commonly referenced in respect of these. It was clear that such positive relationships took time to develop, with some participants reporting being initially dubious about the ‘agenda’ of the RTOF and use of their data. In this context, the perseverance of staff and their perceived genuine interest in participants’ situations and interests was acknowledged as vital for building trust.

“Just once again, really, I want to spread my thanks to the organisation. This whole situation has been really stressful sometimes, it leaves one feeling hopeless, so it’s good they have been so supportive, they help me a lot, they make me feel so good with this organisation. One feels desperate not knowing the future, not having a job or housing.”

Participant

It was also clear that, while the process of integration is not linear for any individual or cohort, participants felt that employment and wider integration were highly contingent on accessing safe and secure housing and, in some cases, mental health support. Aligned with this, most participants valued activities relating to housing the most. Many participants interviewed likewise discussed the difficulties they experienced before the RTOF and the effects these experiences had on their mental health. Securing safe and appropriate housing was, to them, inextricably linked to their mental health, wellbeing, and subsequent employability and integration.

However, reflecting the housing issues cited above, participants spoke of confusing eligibility for social housing, delays with paperwork, issues with repairs and poor communication. Several participants lacked stable access to housing, often awaiting better options or for family reunion to move again. However, analysing participants’ trajectories across interviews showed limited signs of improvement in housing circumstances over time, and signs of increased frustration with their situation. However, participants also generally expressed understanding that such issues were often beyond the control of the RTOF projects.

“[My caseworker] is also kind of limited in how much she can do. I understand their situation. There’s a lot of demand on housing, there’s a lot of people who need social housing. There was another person at the council who was really nice too. On professional levels they are really nice - I think the system is there’s a problem.”

Participant

More generally, participants highlighted the flexibility, convenience, and consistency of the RTOF support in particular. Upon engaging with projects, many participants had low levels of English language and a lack of knowledge about systems and culture in the UK. They appreciated having caseworkers to explain processes in accessible language, alongside providing more general integration support. In particular, enrichment focused activities, such as sports or film clubs, as well as communal meetings with food and conversational English practice, were said to help them to make friends, improve their English, and get to know the local area better.

During follow up interviews, participants often noted that they met with their caseworkers less frequently due to feeling more settled, but were reassured by knowing that they could still call their caseworker whenever they needed.

Despite the positive experiences noted, knowledge about the connectivity and cross-availability of different elements of the RTOF support offer varied. Some participants were not able to distinguish between different services and were unable to attribute certain experiences to the RTOF. In addition, some interviewees were receiving housing support elsewhere (from charities or the council) and were not aware that the RTOF provided housing support.

3. Social Impact Bond design and effects

3.1 Key findings

Working within SIB structures enabled greater collaboration between providers, as well as the combining of expertise across organisational boundaries. These factors had benefits for partnership working and growing knowledge and understanding of working within SIBs for various stakeholders and the sector.

The SIB approach enabled projects to use finances flexibly to achieve outcomes, rather than delivering fixed activities, enabling more person-centred approaches to be developed and more adaptive management to be possible.

There was consensus across stakeholders that the SIBs adopted had driven a focus on outcomes, alongside a strong focus on data and performance that in turn influenced a more performance-driven approach amongst providers. This was seen to generate some positive effects on efficiency, although it was also apparent that such benefits varied across projects due to other factors negatively impacting on efficiency (including limitations in data systems for capturing and recording outcomes).

Within the SIB model adopted, funds financing the RTOF outcomes were subject to annual caps on available funds that could be used to meet outcome claims. While agreed to by investors at the outset, these caps were highlighted as being restrictive, given that funds could not be transferred between years. This was recognised as challenging in the context of an outcomes-based payment approach dependent on estimating numbers of outcomes likely to be achieved in each financial year.

3.2 Design of the Social Impact Bonds

A SIB involves partnership between 3 key parties: an outcomes payer (commissioners), in this case the Home Office, an investor or investors, and providers of services. The RTOF SIBs had similar designs.

As commissioner, the Home Office managed outcomes payments, with funding provided by the Shared Outcomes Fund and Life Chances Fund. The Home Office held contracts with 2 special purpose vehicles (SPVs): one covering the North East and Plymouth projects and one covering the Greater Manchester and West Midlands projects. SPVs are legal entities that are contracted by the commissioner (outcomes payer) and hold contracts directly with providers. SPVs held subcontracts with providers and included performance managers to oversee project delivery. Providers (both LAs and VCSEs) delivered employment, housing, and wider integration activities, and some had management or coordination responsibilities. A rate card defined the amounts paid for each outcome. The Home Office consulted a range of stakeholders and experts to establish the rate card, with the aim of modelling likely payments required against expected volumes of support and outcomes within the funding available.

Key features of the SIB approach included:

• sharing financial risk: providers were shielded from financial risk by investors; SPVs held fee-for-service contracts with providers who were not paid based on achieving outcomes; provider contracts could only be at risk where providers were underperforming without sufficient improvement
• proportion of payment based on outcomes: almost all payment was based on achieving outcomes, alongside a smaller amount for management costs
• involvement of VCSEs: across all projects VCSEs (who may typically be precluded from participating in larger contracts with greater financial risk) were contracted to deliver activities with participants and/or employers alongside local authorities
• rigour of outcome measurement: the Home Office paid all rate card outcomes based on receiving evidence that outcomes had been achieved; the Home Office audited 10% of outcomes and associated evidence at monthly audits
• degree of additional performance management: an entity established by the SPVs, separate from providers, undertook additional performance management
• degree to which the SIB is built on an invest-to-save logic: all projects were built on an ‘invest-to-save’ logic; for example, the assumption was that costs for HM Treasury (such as benefits payments) would be reduced by supporting more people into employment

3.3 Effects of the SIB models

Evaluation evidence shows that the SIB model enabled stakeholders to collaborate and combine their expertise. Providers brought experience of working with participants and investors brought finance and contract management expertise. Blending these skillsets was valuable to projects and for delivery. Investors brought together providers in regular partnership meetings, liaised with the Home Office on behalf of providers, and used data to drive performance. Partnership working increased providers’ knowledge about SIBs, increased investors’ understandings of sector needs and supported cross-government partners to understand more about working with investors and outcomes-based commissioning. Many stakeholders (providers and government stakeholders) were new to working within a SIB. Some providers had initial reservations related to private financing of services for people with vulnerabilities. Over time, these feelings somewhat dissipated, with some providers becoming more amenable to future opportunities to work in a SIB or, more broadly, an outcomes-based contracting context.

In terms of advantages to the SIB model, the approach involved was positively acknowledged across stakeholders as having enabled projects to use finances flexibly to achieve outcomes, rather than delivering fixed activities. This flexibility enabled projects to be person-centred, facilitated adaptive management, and enabled projects to test multiple approaches. For example, some projects accessed flexible funds to provide rent payments up-front to private rented sector landlords, buy white goods or build language development software.

The SIB model was also widely recognised as leading to a greater focus on outcomes. There was strong consensus that the focus on data and performance required drove providers to be more outcomes focused. However, provider representatives did note feeling pressure from investors to achieve payable outcomes in some cases. Likewise, it was noted that balancing the requirement to achieve outcomes with what was best for participants was challenging at points during delivery. This highlights a potential tension between the outcome-based models adopted and the aim of providing holistic support to participants, who tended to progress at different rates in terms of moving closer to employment.

Despite the positive aspects of the SIB approach identified, several elements of the administration of the SIBs by the Home Office were perceived by investors and providers to limit potential flexibility and adaptation. Project representatives felt there was limited flexibility offered to adapt evidence requirements during delivery. Projects had notable difficulties collecting evidence as participants could be reluctant to share this or could disengage after achieving their goals, preventing projects from claiming outcomes. In particular, the need for providers to evidence participants’ active engagement with employability activities was seen as challenging, as this could lead to disengagements and prevent investors being paid for other outcomes. Project representatives suggested the need to provide evidence undermined the ability to be person-centred as it could take time to build trust or start to discuss employment with participants. Conversely, government stakeholders stressed that evidencing outcomes was an important core feature of the model, noting that the requirements should have been clear to investors and subsequently communicated to providers from the outset.

“I don’t think that this has been a well-designed SIB… it is very atypical that the evidence requirements are so stringent. I think that you need to have a bit more trust in your delivery partners.”

Manager

Perhaps most significant in terms of learning from the models adopted was the fact that funds financing RTOF were timebound, meaning that there were caps on available funds that could be used to meet outcome claims within each year of delivery. Contracts outlined the inability to transfer funds between each year of delivery if outcomes were not achieved. However, from the perspective of some investors, while agreed to at the start, these caps were restrictive for projects given that financial modelling and target profiling were underpinned by limited evidence at the outset. Projects surpassed total available funds in October and November 2023, but continued to deliver at their own cost until March 2024.

While acknowledging limitations of government funding in the context of annual cycles, various stakeholders argued that annual funding caps can be a particular challenge for SIBs, as well as other outcomes-based funding models. The need to consider flexibility to transfer funding across financial years was offered as a key learning point for future schemes using similar models.

Delays in contracting and mobilising projects were perceived to shorten the timeframe to achieve longer-term sustainment outcomes, despite contracts being extended. Participants who were engaged closer to the closure of the referral window also had less time to access services and achieve outcomes. Projects felt these factors were particularly challenging when working with a cohort facing multiple barriers. For example, participants often needed to access housing and English language development before finding work. Many stakeholders felt that greater numbers of outcomes (particularly sustainment outcomes) could have been achieved over a longer contract period. While programmes such as the RTOF are typically timebound, this does indicate a need for long(er) delivery cycles when seeking to support individuals with often multiple, complex needs and who can be very distant from the labour market, as many within the RTOF cohort were.

Project representatives likewise reported that many outcomes could not be claimed, despite being achieved. This mainly related to projects misinterpreting outcome requirements or being unable to obtain specific documentation to evidence outcomes. External evidence was required to evidence all housing and employment outcomes (mostly sourced from participants via their employers or housing providers/landlords), whereas only internal evidence was required to evidence wider integration outcomes. Alongside the greater challenges inherent in achieving employment and housing outcomes, this in part explains why greater volumes of wider integration outcomes were achieved (see section 4 below).

While evidence indicates that the SIB models supported efficiency through effective performance management and encouraging a focus on achieving outcomes, this varied across projects. A key challenge for efficiency was the lack of uniform data systems for providers to access and share information about participants. This was a particular challenge as all projects had multiple providers working with the same participants, which implies a need for all to ensure, as far as possible, that adequate systems and data sharing arrangements and mechanisms are in place at the outset of delivery.

4. Employment, housing and integration outcomes

4.1 Key findings

Analysis of the SDS focused on understanding what characteristics or factors were associated with outcomes achieved by RTOF participants. English language proficiency was identified as important across multiple outcomes.

Around a third of participants (33%) achieved outcomes to progress their employability (intermediate employability outcomes. For example, work experience or volunteering), with older age groups, higher English language proficiency, and certain housing situations (for example, being homeless or in temporary accommodation at baseline) being significant factors in the achievement of this outcome.

Over a quarter of participants (26%) achieved employment entry outcomes, with higher likelihood associated with being male, higher English proficiency, prior employment experience (in the UK), and achieving a housing entry outcome. Being from the 6 most represented nationalities and having dependants were linked to lower likelihood of employment entry. This is comparable with other employment support programmes for vulnerable groups, as well as some mainstream DWP employability support programmes.

Barriers to achieving employment outcomes included funding caps for providers, limited timeframes to achieve outcomes with a cohort that is further from the labour market, challenges accessing ESOL provision, and a lack of access to stable and secure housing.

Sustained employment outcomes at 3 months were achieved by 78% of those with an employment entry outcome (or 20% of all participants), with higher English proficiency, being in rented accommodation, prior employment experience, and achieving a housing entry outcome as key factors.

More than half (53%) of participants entered safe and secure accommodation while engaged with RTOF projects. The likelihood of achieving this outcome was higher for those who, at baseline, were in housing provided to asylum seekers or temporary accommodation. Factors including being from older age groups, having higher English language proficiency, and being in education at baseline increased the likelihood of achieving housing entry. Of those who achieved a housing entry outcome, 71% went on to achieve a sustained housing outcome at 6 months, with similar explanatory factors.

Wider integration outcomes were achieved by most participants, with 95% of participants creating a plan and setting key goals related to their integration. Of these, 71% achieved a 6-month improvement in integration outcome, and 72% achieved a 12-month improvement in integration outcome. Factors contributing to these outcomes included older age groups and higher English language proficiency. The high levels of these outcomes reflect that integration plans were a standard element of ongoing activity with participants.

Qualitative data suggests that RTOF also achieved a range of system-level outcomes related to improved relationships with employers, enhanced local provision, greater knowledge and understanding of refugees and their needs, and a greater focus on outcomes as a result of the SIBs.

RTOF participants experienced other benefits through participating in RTOF activities, including reporting greater confidence and increasing their access to statutory services and community activities, as well as improvements in wellbeing.

4.2 Employment outcomes

This section covers outcomes experienced by those supported by the RTOF relating to their employability, along with entering and sustaining employment. As discussed earlier, the RTOF sought to support participants into sustained employment. Reflecting this, the programme paid for achievement of 3 employment-related outcomes, including an ‘intermediate’ outcome to reflect activities designed to help participants progress towards, and into, employment. These outcomes were:

Intermediate employability outcome: undertaking volunteering or work experience alongside improvements on 2 further individualised employability targets.

Employment entry: entering employment or self-employment.

Employment sustained: sustaining employment or self-employment for 3 months.

The number of these outcomes achieved by the RTOF participants are summarised in Figure 1 and discussed in further detail in subsequent sections. Note that intermediate employability and employment entry outcomes are not interdependent. Whilst some participants achieved both, others achieved just one, which reflects their distance from the labour market and support focus.

Figure 1: Employment outcomes achieved

Source: Ecorys analysis of SDS

Base: n=2,340 individuals

To assess the Fund’s impact on employment and the employment-related outcomes achieved, a counterfactual impact evaluation (CIE) approach to estimate the RTOF’s impact on employment entry (see section 5) was combined with analysis of SDS data and insights from the qualitative research.

4.2.1 Increasing employability

Around one-third of participants (33%, n=761) achieved the intermediate employability outcome. Participants from older age groups, with higher English language proficiency, and certain housing situations (for example, being homeless or living in temporary accommodation upon first engaging with RTOF) were more likely to achieve this outcome that aimed to increase their employability via volunteering or work experience.

Participants accessed a range of volunteering or work experience opportunities, some leading to paid employment. Alongside volunteering and work experience, participants gained skills accreditations or qualifications, accessed English language tuition, or received additional employability coaching. The intermediate outcome was designed for participants who needed more support to find employment, for example due to being more distant from the labour market and/or being likely to benefit from being able to evidence work experience or volunteering. Despite this guidance, some projects pursued this outcome for participants before starting to focus on actual employment entry, which reportedly led to some level of disengagement. This approach was adapted during delivery to mitigate further disengagement, moving to a focus, as intended, on using intermediate support only for those requiring it.

4.2.2 Entering employment

Over a quarter of participants (26%, n=613) entered work. Entering employment on zero-hours contracts was not accepted an outcome because of concerns about their potential to hinder the progress of those already vulnerable.

This job-entry figure is comparable to other employability support programmes. For example, DWP’s Work Programme (supporting Jobseekers Allowance claimants likely to be closer to work, as well as ESA claimants who can be presumed to be further away) had a similar performance to RTOF, despite working with a potentially easier cohort. See Work Programme: background and statistics, that shows 31.5% of those completing 24 months on the programme, who were referred on in December 2013, achieved a job entry.

In terms of more comparable cohorts, the ‘early access’ group volunteering for support through DWP’s Work and Health Programme, comprising individuals from groups such as the homeless, ex-armed forces, care leavers, and refugees, also achieve similar job entry rates (just over 20% after 12 months and just over 30% after 24 months)(GOV.UK 2023).

Using logistic regression, analysis of SDS data indicated that participants who were male, had higher English language proficiency, prior employment experience, and lived in safe and secure accommodation were more likely to enter employment. Participants from the 6 most commonly represented nationalities and those with dependants were less likely to enter employment. The qualitative research showed that having dependants implies greater caring responsibilities that can create barriers to work. However, the reasons for more represented nationalities being less likely to enter employment was unclear.

Despite high levels of missing data in relevant fields, analysis of the SDS suggests that some of the roles participants gained in terms of job-entry included warehouse or production operatives, care assistants, hairdressers, delivery drivers, cleaners, and administrators. These findings broadly reflect the qualitative findings, which highlighted that jobs such as warehouse operatives were more accessible for at least some of the cohort due to lower English language requirements. Relatively greater availability of particular roles, including those noted, was also cited as a factor for refugees entering such jobs. However, provider representatives also commented that participants often had to apply for such entry-level roles despite having higher-level skills or professional experience (because of a lack of UK-based experience and existing qualifications not being recognised).

SDS data also showed that around half of participants (n=160) for whom the relevant fields were completed (there were high levels of missing data in respect of this area with useable data only available for 339 individuals) entered employment that aligned with goals they set at the beginning of their engagement with RTOF. However, this analysis is somewhat limited as data was not updated throughout delivery, wherein goals could have changed.

4.2.3 Sustaining employment

Over three-quarters (78%, n=477) of the participants who found work sustained employment for 3 months (representing 20.4% of all participants). Participants with higher English language proficiency, prior employment experience or who lived in safe and secure accommodation were more likely to sustain employment. Participants in age groups 25 to 29, 35 to 39, and 40 to 44 were proportionally less likely to achieve this outcome relative to those aged 18 to 24.For those age 30 to 34 and 45 and over the results were similar (see logistic regression outputs) but not statistically significant (so not reported).

4.2.4 Association of housing and employment

Evidence indicates that entering safe and secure accommodation was a key precursor to finding work. Analysis explored the sequencing of employment and housing outcomes. Logistic regressions indicated that participants in safe and secure accommodation were, all else being equal, twice as likely to enter employment. SDS data shows that entry to accommodation outcomes were claimed for 302 participants before they entered employment. This compares with 94 participants who entered employment before entering safe and secure accommodation. 217 participants entered employment without any accommodation outcomes.

4.2.5 Influencing factors

As noted above, project representatives reported that participants with higher skill levels could face barriers to entering similar job roles in the UK. The main reasons related to limited English language proficiency (a commonly cited and key barrier overall), qualifications not being recognised in the UK, and/or a lack of local opportunities for specialist roles (for example, chemical engineering). People with higher skill levels also pursued retraining opportunities, potentially slowing down their movement into employment. Reportedly, such participants entered lower skilled opportunities to get into work faster.

Project stakeholders noted that entering lower skilled or entry-level opportunities (taxi driving, warehouse operative, cleaning, or food delivery roles) could enable participants to work sooner. Project representatives generally felt that this was a positive step towards more skilled employment, if that might be suitable. However, the nature of some employment contracts meant that these outcomes could not be claimed, for example where zero-hours contracts were involved. Investors and project stakeholders argued that this was problematic and/or inappropriate given the prevalence of such contracts and the fact that some employees may choose them. However, counter views were also expressed by government stakeholders, noting that the RTOF objective of sustainable employment was less likely to be achieved through such routes.

Evidencing requirements reportedly also influenced the achievement or otherwise of the outcome of sustaining work for 3 months. To achieve this outcome, participants were required to share documents to evidence that they were still in work. Project representatives reported that, in many cases, participants could not understand the need to do this, and/or had wider priorities, meaning that gathering and recording evidence was challenging. Providers sought to overcome this by incentivising participants to share evidence by providing vouchers, but this was not always successful. As a result, it is important to recognise that the figures discussed for the proportion sustaining work amongst those entering jobs may well be an underestimate.

Although a minor theme, some participants had reportedly found informal employment. Project staff reported that they had seen some participants working in roles associated with cash-in-hand employment in the informal economy, which in turn could also discourage further, or as much, engagement with RTOF support to find more stable and secure opportunities.

Within the qualitative research, the perceived extent to which RTOF directly led to employment outcomes varied. Some interviewees were entirely reliant on RTOF support for employment (including ESOL, CV writing, job searches and applications), with no knowledge about work in the UK upon referral. For others, RTOF gave help where participants were already on their employment journey (such as helping them to find more suitable jobs, helping with equipment, clothing, or qualifications). A small number of participants interviewed, meanwhile, suggested that they had found employment without RTOF support. These interviews also revealed that employed participants often reported that they were in jobs which were sufficient, but that did not fulfil their needs entirely (both financially and mentally). Many reported having future dreams and aspirations which they were working towards (that is, getting qualifications or attending courses). Largely, participants saw improving their English language as the most vital step on their employment journey, confirming the centrality of this factor discussed elsewhere in this report.

4.3 Housing outcomes

This section covers the extent that participants entered safe and secure accommodation and were living in safe and secure accommodation after 6 months. Safe and secure accommodation was defined by compliance with the Decent Homes Standard, including the absence of category 1 and 2 hazards under the Housing Health and Safety Rating System, and with all relevant gas, electricity, and fire safety checks in place. Within this safe accommodation, the participant user/family must have had appropriate individual space.

4.3.1 Entry to accommodation

More than half of all participants (53%, n=1242) entered safe and secure accommodation while engaged with RTOF projects. However, participants’ housing situation upon engaging with RTOF should be considered. Table 1 details participants accommodation situation at the start of their engagement with RTOF. High proportions of those in asylum accommodation (62%) or temporary accommodation (82%) achieved a housing entry outcome, whereas smaller proportions of those in LA rented (50%), private rented (48%), or other safe and secure (33%) accommodation achieved housing entry outcomes.

Table 1: Accommodation outcomes compared with baseline housing situation

Participants’ accommodation upon engaging with RTOF	Count at baseline	Percentage entering safe and secure accommodation
Asylum accommodation	1167	62%
Homeless and rough sleeping	36	56%
Homeless but not rough sleeping	151	52%
Other safe and secure accommodation	144	33%
Publicly funded temporary accommodation	214	82%
Renting - local authority or housing association	110	50%
Renting - private sector	105	48%

Source: Ecorys analysis of SDS

Base: n=2,340 individuals

The above table shows that 62% (n=1167) of participants initially living in accommodation provided to asylum seekers by the government entered safe and secure accommodation. This compares with 48% (n=105) of participants already living in private rented accommodation, where their need to find alternative accommodation may have been less of a focus. Some of this group may have been in temporary or supported accommodation but looking to find more stable housing.

The above findings were supported by the logistic regression results, where participants already in rented accommodation (both in the private rented sector and social housing) were, all else being equal, around 40% less likely to claim an entry to accommodation outcome. There was a similar association for those who were homeless, which might relate to their further distance from safe and secure housing (where the focus may have focused on finding temporary accommodation in the short-term).

Further factors associated with being more likely to enter accommodation included:

• being from older age groups (specifically, 35 to 39, 40 to 44, and 45 to 49)
• having entry level or above English language proficiency
• being in education at baseline

4.3.2 Sustaining accommodation

Of those entering accommodation, most (71%, n=887) were still in accommodation after 6 months. Factors that increased the likelihood of sustaining accommodation were similar to those for the entry to accommodation outcome (older age groups, higher English language proficiency).

4.3.3 Influencing factors

Projects struggled to support participants to find private rented accommodation or social housing within the 28-day move-on period, with staff suggesting that the time available was insufficient to find more stable housing. This also related to a broader and significant challenge that was widely reported, namely a distinct lack of suitable and affordable housing locally. It was noted that entering private rented accommodation required stable incomes, deposits, credit histories, and references, which took time to build. For social housing, registration on bidding platforms was required based on prioritisation criteria, where single people (many of the cohort) could be lower priority.

Projects therefore often supported participants into temporary accommodation or supported housing in the short-term, while helping participants to find opportunities to improve their employability, English language proficiency and support their integration. However, project staff also commonly reported that waiting lists for supported accommodation could be long. Such accommodation was also perceived to be inappropriate for some participants, particularly families.

Another notable finding from the qualitative research was that entering supported housing could be helpful in the short term but could become a disincentive to work, as rent would become unaffordable when participants began work. The need to move out of supported accommodation to afford to work could thus deter people from finding work. Likewise, interviews with participants identified cases where participants were seeking to improve their English, complete college courses and wait to find other housing before finding a job.

Interviews with project representatives and participants also identified that many participants had a preference to access social housing rather than the private rented sector. Participants reported that this was because they thought social housing was more accessible than private rented housing. They also thought that they were more likely to have to move regularly with private rented housing, that they could be more likely to be evicted (if landlords sold properties), and social housing could give greater stability that they desired. However, as discussed in earlier sections, the assumed availability of social housing when it was very limited represented a key aspect that staff needed to discuss with clients and work on to adjust expectations.

Sustaining housing was supported via discussions around budgeting, interpreting letters for clients (including overdue bills), ensuring participants understood and were adhering to tenancy agreements, liaising with landlords on participants’ behalf, and advice on managing energy costs. All these elements were cited as important by project staff. It was unclear how much support was provided beyond the 6-month sustainment period. However, project staff commonly reported having an ‘open door.’ Indeed, qualitative data indicated that some participants re-engaged with RTOF housing providers when they were facing evictions, because of landlords selling properties (typically Houses in Multiple Occupation).

4.4 Integration outcomes

This section describes the outcomes for participants related to wider elements of integration, beyond employment and housing. Integration goals were set by participants and their caseworkers, documented in an individualised integration plan. Integration outcomes included:

• creating an integration plan
• 6-month improvement on 2 or more progress targets
• 12-month improvement on 2 or more progress targets

Progress targets were goals set by participants and their caseworkers, related to various aspects that would support their integration.

4.4.1 Integration plan creation

Most participants (95%, n=2,206) created an integration plan. Integration plans were a standard element of RTOF provision, demonstrated by the high proportions of this outcome achieved within the cohort. Reasons why this was not achieved for all could include recording errors, (almost immediate) disengagements, moving between RTOF projects or non-engagement with wider integration activities (where participants felt they did not have needs). Recognising most participants achieved this outcome, logistic regression was not conducted, as this analysis requires variability in the outcome to be reliable.

4.4.2 Six-month improvement

For participants who created an integration plan, most (71%, n= 1,560) achieved improvement targets after 6 months. Factors that were associated with an increased likelihood of the achieving 6-month improvements included:

• older age groups (specifically, 35 to 49, 45 to 49, and 50 and over)
• having entry level or above English language

4.4.3 Twelve-month improvement

Twelve-month improvement targets were achieved for a similar proportion (72%, n=1,119) of participants who achieved 6-month improvements. A similar pattern in potential explanatory factors existed here, wherein older age groups were more likely to achieve this outcome. This could potentially be explained by these groups engaging and being in contact with RTOF projects for longer (direct information on support duration was not collected in the SDS).

4.4.4 Influencing factors

Project stakeholders generally found the integration plans to be a useful way to record participants’ individual integration goals and to track progress. However, as outlined in section 2, information was not always recorded consistently. Despite this, the plans and the activities that RTOF was able to flexibly facilitate were seen as beneficial and positive in supporting integration outcomes. Again, English language proficiency emerged as being key to integration, as well as an important precursor and/or supporting factor in achieving employment and housing outcomes.

Reflecting on challenges with accessing ESOL and the slow progress with language development, projects suggested that it would have been helpful to have more focus on English language as a key element of delivery (rather than sitting outside of the support). With English language development being a key priority for participants, projects also reported participants prioritising ESOL above other elements of the support, thus not wanting to engage in volunteering, work experience or paid employment if this could affect their availability for ESOL classes.

4.5 Assessment of overall RTOF outcomes

This section provides an assessment of the overall achievement of outcomes for RTOF participants and illustrates their interconnected nature, further reflecting the need for holistic integration services.

4.5.1 Average time to achieve outcomes

The need to evidence and claim for specific outcomes means that dates were recorded for the achievement of each outcome. Analysis was conducted to understand how long it took for participants to achieve outcomes. Figure 2 below shows the average time between participants engaging with RTOF and outcomes being achieved. Outcomes for creating an integration plan were achieved the soonest (mean 39 days, median 21 days), followed by entering accommodation (mean 120 days, median 80 days), entering employment (mean 217 days, median 183 days), and intermediate employability outcomes (mean 241 days, median 216 days).This suggests that employment outcomes take longer to achieve and claim.

Figure 2: Average days from RTOF start date to achieving outcomes

Source: Ecorys analysis of SDS

4.5.2 Connectedness of outcomes

Overall, there was strong evidence to suggest that the achievement of outcomes as a whole was highly interconnected. Supporting participants into safe and secure accommodation was clearly important for the achievement of other outcomes in particular (as also demonstrated by the logistic regressions discussed above). When participants had safe and secure accommodation, they could focus on their education, developing English language proficiency and finding employment or improving employability through work experience or volunteering. Participants described how destabilising it was not to have somewhere to live, with short timeframes to find new accommodation, and the negative impact this could have on their and their family’s wellbeing.

“I was taking medicine, but now I am ok. I still struggle sometimes; I forget a lot of things. Sometimes it makes me depressed again when I can’t find a place to live. Before I was taking medicine, I was having counselling, she was helping me a lot. It was a really difficult time when I moved [out of asylum accommodation]. It’s not only about moving house, but Home Office when they refuse, they stop everything, it is really horrible, it is hopeless, there is no hope.”

Participant

Being housed also supported participants to feel a sense of belonging to neighbourhoods and local communities. While not all participants felt comfortable in the local area within which they were housed, or their neighbours, being in one place for a sustained time was clearly important in supporting them to regain stability in their lives. Housing situations, where more precarious, also negatively affected participants mental health and wellbeing.

Finding work was an important element of participants’ integration. When participants entered work, they reported feeling more self-sufficient and found meaning in being able to provide for themselves. Alongside generally improving financial situations, employment also supported participants to build social connections and gain experience that they felt could help them into future work and other opportunities. Similarly, many stakeholders noted that participants’ initial jobs could be a stepping stone, helping them to work towards future opportunities that aligned more with their long-term goals.

RTOF’s design recognised that integration was multidimensional and involved factors beyond employment and housing, such as being able to access services, improving health and wellbeing, gaining English language proficiency, achieving education and skills accreditation, and developing social connections, as outlined in the Home Office’s Indicators of Integration. Accessing classes to develop English language proficiency helped participants to communicate with others and to understand key documents and systems. Despite a lack of access to mental health services, accessing GPs and, to some extent, dentists supported participants’ health and wellbeing.

Group activities were found to be important in supporting participants to build social connections, practise English language skills, and helped to generate wellbeing benefits. Opportunities to meet new people and build social connections mostly occurred between RTOF participants, with less evidence of participants being referred to social or enrichment activities in the community (like local football teams). That said, participants could find new social connections through volunteering or work experience opportunities (such as at the local zoo), providers themselves, or when in work opportunities. Participants could also find new social connections through their housing, particularly where living in shared houses or supported accommodation, where some elements of living were more communal.

However, while RTOF clearly generated positive inter-related outcomes for some participants, both project staff and participants recognised that achieving self-sufficiency and all their integration goals was a longer-term endeavour that extended beyond the lifetime of RTOF for many participants.

4.6 Wider system and organisational outcomes

This section presents an assessment of the RTOF’s effect on system-level outcomes. It examines the extent to which providers reported newly developed or improved relationships with employers, as well as exploring any effects on increased organisational and service capacity because of involvement with the programme.

4.6.1 Relationships with employers

Project stakeholders reported that they had much greater engagement with potential employers than they had before RTOF highlighted the need for employer engagement to improve available opportunities for participants and provided dedicated resource to build relationships with local employers. Projects reported portfolios of available opportunities for participants that had been built under RTOF. Certain providers, moreover, reported building positive relationships with local and national employers, which had reportedly resulted in employment opportunities for some participants. However, as noted earlier, engaging employers took time - in particular, to build trust and to understand their wants and capabilities for supporting people with experience of migration.

4.6.2 Organisational and service capacity

The evaluation found evidence of a range of outcomes for staff and organisations involved in RTOF delivery. Providers reported that the funding and partnership working under the RTOF helped them to build networks with other providers, as well as better connections with LAs. VCSEs reported developing greater understanding of how to tailor service offers to meet LA requirements, as well as how to be more adaptive and responsive to changes in the wider ecosystem of refugee support.

Through RTOF delivery, providers reported being able to demonstrate the level of need for refugee support services to LAs. This supported projects to build a case for continued provision, beyond the RTOF lifecycle, drawing on learning from delivery models. However, there was a continued concern reported by project stakeholders and investors about sustaining funding at the required level, especially as there was not a subsequent programme directly following piloting RTOF.

Despite this, providers had hopes for their organisational sustainability being improved because of participating in the RTOF. Provider staff also reported organisational-level improvements in recording monitoring information and data analysis, learning about working with outcomes-funding mechanisms (SIBs rather than grants), and learning from specialists (in housing or employment). Provider representatives hoped these learnings would improve their chances of securing future grants or funding opportunities. However, there was limited evidence identified on how the level of support provided through RTOF could be sustained via other funding streams.

Project staff also reported that RTOF had enabled them to become better networked and established. Developing relationships with the Home Office and cross-government partners was viewed positively. Smaller providers in particular gained understanding of working within a larger-scale government contract. Some providers had also been able to expand their delivery into new locations through the RTOF, extending their reach and growing their understanding of new communities.

For some stakeholders involved in delivery, RTOF was the first time they had worked with refugee cohorts. These stakeholders described gaining much greater understanding about refugees’ needs and the many intersecting elements that affect the lives of people recently granted refugee status or humanitarian protection. Those who had previously worked with resettled cohorts described gaining improved knowledge about the needs of people who had previously traversed the UK’s asylum system.

At the system level RTOF has clearly positively contributed to enhanced partnership working between providers, which has the potential to be sustained in the context of future programmes alongside contributing to the wider support system available for refugees. RTOF also enabled a greater awareness of the capabilities and performance of different providers. Government stakeholders became more aware of working with stakeholders in a SIB context, and the various layers of communication that are required to ensure effective delivery, hence informing future programme design. Similarly, LA representatives reported gaining a better understanding of the capacity of local VCSEs. Collectively, such learning has the potential to aid development of future service models and programmes, as well as positively contributing to the refugee integration support ecosystem.

5. Impact and value for money assessment

5.1 Key findings

The counterfactual impact evaluation (CIE) indicated that participants were 39% more likely to enter employment relative to a comparator group (odds ratio of 1.39). RTOF demonstrated a positive effect on employment status, which was statistically significant at a 95% confidence level. This means that the likelihood of the observed change in employment status being due to chance is less than 5%, providing strong evidence of the intervention’s effectiveness. It is important to note that the odds ratio is a relative measure of increased likelihood, not a direct reflection of the number of additional jobs created.

In absolute terms, the evaluation found that 135 job entries - equivalent to 22.1% of those employed in the treatment group - can be directly attributed to RTOF, after accounting for baseline employment levels in a comparable group without the intervention. This adjustment is necessary because the odds ratio reflects a relative increase in likelihood, which must be scaled to the baseline employment rate to estimate the actual number of additional jobs created.

Based on analysing the costs and benefits of the RTOF, the programme is estimated to achieve a positive return on investment (that is, the monetary benefits outweigh the programme costs) shortly after 3 years. This is based on using the central assumption for employment benefits (derived from calculated additionality) and restricting the assessment to fiscal employment and housing benefits only.

When also including economic benefits (to the individuals concerned) and social benefits (to wider society), a positive return on investment was estimated to have been achieved after one year.

Noting that the cost-benefit analysis does not capture non-monetisable benefits related to broader aspects of integration achieved beyond housing and employment, and hence will not fully account for all benefits observed, the analysis suggests that the programme represents value for money and confirms the invest-to-save logic of the RTOF.

While there are limited direct comparators, in terms of similar interventions also being subject to a robust economic evaluation, available data suggests that the RTOF can be viewed positively in terms of the benefits achieved relative to costs allocated.

5.2 Introduction

This section assesses the RTOF’s impact on employment entry amongst those supported and the value for money (VfM) of the Fund.

A CIE approach was developed to estimate the impact of RTOF on employment entry for those supported (the treatment group) relative to a statistically ‘matched’ comparison group (further explanation of the methodology used for the CIE is included in Annex A). The key difference between this analysis and that presented in section 4 - which focused on the outcomes experienced by those supported by the RTOF - is that it seeks to more precisely attribute the impact of RTOF support on employment entry. In other words, by comparing outcomes between the treatment and comparison groups, the additionality of RTOF support (over business as usual support) can be estimated. The precise approach used propensity score weighted logistic regression models to isolate the effect of support on employment outcomes. Propensity scores were calculated to account for multiple characteristics that influence the likelihood of receiving the treatment. These scores were then used as weights in the logistic regression model, which adjusted for any remaining differences between the groups, isolating the effect of the treatment on employment outcomes.

The main methods used for the VfM assessment were Cost Benefit Analysis (CBA) - to monetise the fund’s outcomes as ‘benefits’ and to compare these benefits with programme costs - and Cost Effectiveness Analysis (CEA) - to compare metrics of value for money across related projects. Although RTOF is unique in its aims and approaches, CEA provides useful context to the assessment. However, inevitable differences in programme design, metrics and monitoring mean that such cross-programme comparisons can only provide indicative evidence.

The approach to the CBA and CEA are summarised below in section 5.4, prior to presenting the results and conclusions of the VfM exercise. Where applicable, additional insights from the qualitative research undertaken are also provided.

5.3 Impact analysis of the RTOF

5.3.1 Impact on employment entry

The primary model used for the CIE indicated that RTOF participants were 39% more likely to enter employment relative to a comparison group of non-RTOF participant refugees (odds ratio of 1.39). The comparison group was constructed from administrative data to mirror the characteristics of RTOF participants (see Annex B for a full description of the methodology used in the CIE).

The primary model in our analysis used the assumptions deemed most reliable and fair to the RTOF programme and data limitations. Sensitivity analysis tested various alternative assumptions to assess the robustness of our results. The sensitivity analyses showed slightly smaller effects relative to the primary model but confirmed the positive impact, with odds ratios of 1.29 and 1.24. The consistent positive results across models validate the effectiveness of RTOF in improving employment outcomes for its participants.

The odds ratios tell us how much more likely people in the treatment group are to find employment compared to those without treatment. The odds ratio shows how much more likely employment is in the treatment group compared to a comparison group, but it doesn’t directly indicate the number of additional jobs created. To calculate the additional number of jobs attributable to RTOF support, the odds ratio - representing a relative increase in likelihood - needs to be combined with the estimated employment rate in the comparison group.

This adjustment removes the effect of the treatment to estimate how many people would have been employed without it. Specifically, we adjust the treatment group’s employment rate by dividing it by the odds ratio, which scales down the rate to reflect the likely outcomes for a similar group without the support. Scaling down refers to adjusting a value to reflect a lower or baseline level, such as removing the effect of an intervention to estimate what would happen naturally or without it. This “scaled-down” rate represents the baseline employment level, enabling us to calculate the difference attributable to RTOF support.

Using this approach, and the odds ratio of 1.39 resulting from the primary model used for the CIE, the central estimate of the additionality generated by RTOF support for employment entry (that is, over and above outcomes that would have happened anyway) was 22.1%. This figure, representing the additionality of the RTOF’s effect on likelihood of employment entry, is used below when calculating applicable benefits as part of the VfM assessment.

5.4 Cost benefit analysis of the RTOF

5.4.1 Costs

The Home Office expected to fund the RTOF programme up to £13.99 million. Funding came from 2 sources: up to £10 million from the His Majesty’s Treasury (HMT) Shared Outcomes Fund (SOF), and up to £3.99 million from the Department for Digital, Culture, Media and Sport (DCMS) Life Chances Fund (LCF).

As well as the projects, the RTOF’s budget funded support contracts, including a:

• Fund Manager (Palladium International Limited), employed in the initial stages of the programme as outlined in the report introduction

• Learning and Evaluation Manager (a consortium of Ecorys, Social Finance and academic experts), to evaluate the programme and share learning

• Capability Building Provider (Traverse) to support bidders and providers

RTOF’s direct costs were £13.2 million, comprising £10.8 million for project funding and £2.4 million in administration costs. Administration costs comprised £1.5 million (including VAT) for the support contracts, plus £0.9 million for the Home Office to administer the programme, including staff costs. The costs of verified outcomes and associated payments are detailed in Table 2 below.

Table 2: RTOF payments by verified outcomes achieved

Outcome	Volume	Unit cost (£)	Total cost (£)
E0 – Intermediate employment	669	2,250	1,505,250
E1 – Entry to employment	420	3,000	1,260,000
E1E – Entry to self-employment	53	3,000	159,000
E2 – Three-month employment sustainment	352	3,550	1,249,600
E2E – Three-month self-employment sustainment	15	3,550	53,250
H1 – Entry to accommodation	1,131	1,600	1,809,600
H2 – Six-month housing sustainment	683	1,600	1,092,800
I1 – Integration plan	2,187	400	874,800
I2 – Six-month improvement (integration)	1,372	500	686,000
I3 – Twelve-month improvement (integration)	588	550	323,400
Total	7,470	-	9,013,700

The support contracts were initially budgeted at £2.3 million (plus VAT), representing a substantial saving due to efficiencies resulting from contract changes, changes on the VAT recovery position, and the termination of the contract with the Fund Manager. However, as a result there were some additional costs for the Home Office for programme administration, while some investors and providers reported experiencing additional costs. This was related to time spent dealing with the early termination of the Fund Manager contract and additional resources said to have been required due to the transfer of its responsibilities back to the Home Office.

In addition to RTOF’s £13.2 million direct costs, we identified £173,881 of indirect costs (for context, this represents just over 1% in addition to the direct costs) incurred by the Home Office outside of the budget envelope outlined above. These were incurred in the development and design of the programme, including contract setup and negotiation (set-up costs). These included:

• supporting the Learning and Evaluation Manager: £20,500
• support from Home Office Commercial on contract management: £10,800
• RTOF development (internal): £27,664
• RTOF development (external): £100,000
• RTOF business case/investment case/bids to the Shared Outcomes Fund and Life Chances Fund: £14,917

There were also indirect costs incurred by investors and/or providers. Again, these were incurred as set-up costs before and after contract award so they could commence delivery of any programme effectively - such as creating the Special Purpose Vehicles (SPVs) - or throughout the programme, as a function of the inherent risk and uncertainty involving in participating in a complex and dynamic programme, which required changes to original plans. As one provider commented:

“We also absorbed a lot of costs because of other things that we’ve put in place to support RTOF but aren’t included in the RTOF contract.”

Provider

Examples of these additional costs to investors and providers included additional staff time (including employing new staff) for project management or employer engagement, or additional activities to support outcomes to be achieved beyond those initially envisaged. For example, the latter were reported to include funding clients to visit third party workshops or potential employer premises by providing bus passes, so as to remove as many barriers to engagement and progress as possible.

Indirect costs to investors and providers have not been quantified, since the competitive bidding model meant that investors and providers budgeted - as much as practicable - to recoup any indirect costs to them though their operational payments. This links to the logic that the SIB model should, in theory, provide the funding required - with any profit or loss being part of investor risk. However, the risks involved do mean that investing in any such programme requires an investor(s) with sufficient capital to be able to cover upfront costs and losses should they arise.

In addition, with parameters specified at the outset of the programme, including the annual financial caps (so losses in the initial years are not able to be recouped in later years), the need for any flexibility as things change over the course of a programme (which could be due to internal or external factors, such as data sharing) fell on investors and providers as discussed elsewhere in this report. A break clause based on annual reviews could de-risk a future programme for investors and providers, potentially attracting more interest, though transferring more risk onto the public purse, which may not be desirable.

Accounting for the direct costs that could be quantified, as outlined above, the total costs of the RTOF programme were thus £13,373,881 - or over £5,700 for each of the 2,331 participants - noting that additional investor or provider costs are ignored as an inherent part of the SIB design.

5.4.2 Benefits

Reflecting the approach developed in the evaluation scoping phase, non-cashable wider integration outcomes, such as developing digital, language and communication skills, were assessed qualitatively and were not included within the cost benefit analysis. This was because there are inherent challenges in monetising such non-cashable outcomes. Equally, non-cashable benefits did not feature in the cost benefit model that fed into the economic case for investment for the RTOF.

Nevertheless, these outcomes will have some value. For example, digital, language and communication skills may be associated with future earnings uplifts, as well as being a precursor to entry to employment (though, in the short-term, these are in part captured through monetising changes in refugee employment as part of estimating programme benefits). The cost benefit analysis undertaken thus offers a conservative estimate of the benefits generated.

To calculate the benefits from employment, the CIE results relating to employment entry were combined with unit costs to monetise the outcomes concerned. Where necessary, reasonable assumptions based on insights from the quantitative and qualitative analysis, or relevant literature (for example from HMT’s Green Book), were used to adjust estimates, as explained further below. In particular, such assumptions are required in ascribing the additional effects of RTOF support on housing outcomes achieved.

Table 3 below summarises relevant outcomes from the SDS that form part of the basis for the calculations made. The cohort was 2,331 participants in total, of whom 613 entered employment, 761 secured an intermediate employment outcome and 1,242 achieved a housing entry outcome. The sustained employment percentage figure is derived from the 613 participants who entered employment. Likewise, the equivalent sustained housing figure is derived from the 1,242 participants who achieved housing at entry.

Table 3: Summary of outcomes from the SDS

Outcome	Number of participants achieving outcome	Number of participants from whom to calculate percentage achievement rate	Percentage of achievements relative to the number of participants concerned
Employment entry	613	2,331	26.3%
Employment intermediate	761	2,331	32.7%
Employment sustained	477	613	77.8%*
Housing entry	1,242	2,331	53.3%
Housing sustained	887	1,242	71.4%**

Source: Ecorys analysis of SDS

Notes:

1. * Percentage figure for ‘employment sustained’ is calculated as a proportion of those entering employment (n=613) rather than of the full cohort (n=2,331).
2. ** Percentage figure for ‘housing sustained’ is calculated as a proportion of those entering housing (n=1,242) rather than of the full cohort (n=2,331).

The CIE found that the central estimate of the additionality generated by RTOF support for employment entry (that is, over and above outcomes that would have happened anyway) was 22.1%. To monetise these benefits, it is thus assumed that 22.1% of the employment outcomes claimed can be attributed to RTOF support. As this figure is an estimate, it was further estimated that there is a 95% chance of the actual figure being between 15.1% and 28.6%. For the purposes of this analysis, these “pessimistic” and “optimistic” estimates sit either side of the 22.1% “base” estimate. This figure compares to an inferred measure of attribution or additionality of 73.7% (attribution is calculated as 100% - 26.3% = 73.7%) from HM Treasury’s Green Book.

This Green Book measure was used to estimate the RTOF’s contribution to the benefits accruing from housing outcomes given that, in contrast to employment entry, it was not possible to use a CIE approach. Primarily, this was due to the lack of housing data through which to construct a comparison group, thereby rendering a CIE approach infeasible. Based on the qualitative research, and the centrality of RTOF support to generating positive housing outcomes, it can be assumed that the programme contributed to entering and/or sustaining housing outcomes in comfortably above half of cases. In this context, the 73.7% figure used reflects the broader evaluation findings.

As per Table 3 above, the SDS data show that 477 participants or 77.8% of those that entered employment were in sustained employment after 3 months. The sustained employment rate, as a proportion of overall programme participation (n=2,331) is thus 20.5%. Three months employment in itself provides a small, fiscal benefit to the public purse. To calculate longer-term benefits, we projected forward to estimate the benefits should those individuals sustain their employment at 12, 24 and 36 months. Sustaining employment at 3 months is assumed to represent a fairly good predictor of sustainment at 12 months, though naturally some RTOF participants will not necessarily sustain employment to 12 months or beyond. As such, we have estimated that employment is sustained in 60%, 50% and 40% of cases respectively at 12, 24 and 36 months, to offer a realistic range of estimates when calculating employment benefits. Additionality was estimated during the observation period for treatment and comparison groups. When projecting benefits it is assumed that this additionality level remains constant over time which may not be the case.

Housing was not adjusted in this way, reflecting its perceived higher level of attribution and - to some degree stability - relative to employment once gained. Therefore, we assume lower levels of drop-off over time. However, using the Green Book deadweight (that is, the inverse of additionality) estimate of 26.3% implies, in itself, that over one-quarter of housing outcomes for RTOF participants should not be considered as benefits in the analysis. This enables some adjustment to gross benefits to account for RTOF support alone not generating housing entries or sustainment. For housing, 71.4% that gained an entry to housing outcome sustained it to 6 months, and benefits were thus estimated based on housing being sustained at 12, 24 and 36 months respectively.

In keeping with principle of when estimating social value to not overclaim, the analysis may risk being an underestimate, as it does not consider employment benefits up to 3 months (that is, the period before sustained employment is recorded) and housing benefits for up to 6 months (Institute for Social Value 2025. Likewise, the 77.8% and 71.4% estimates of employment and housing sustainment themselves may be underestimates due to challenges on the part of providers in evidencing them; hence outcomes being claimed as reflected in the SDS data used for the CBA may be lower than the real sustained employment and housing outcomes achieved.

To calculate the benefits, the outcomes adjusted for attribution were multiplied by a unit cost. The unit costs used for sustained employment and housing were £14,312 and £5,182 per year respectively; these figures were generated by updating the cost benefit model that fed into the economic case for investment for the RTOF. These unit costs only provide fiscal benefits. In addition, to understand wider benefits, including economic and social benefits, other estimates were included: £51,183 for securing sustainable employment per year (comprising £15,912 fiscal, £4,641 economic and £30,630 social benefits)(Mallender et al. 2017), and £11,634 per year for sustained housing (comprising labour market benefits, increase in tax receipts, decrease in Universal Credit claims, savings to health services, reduced homelessness and crime, and fewer disruptions to education)(CEBR 2024). The unit cost used to value the intermediate employment outcome was £4,551 per individual (Royal Voluntary Service 2024); this reflects economic benefits from volunteering associated with an estimated 50% of refugees achieving the outcome. The same additionality estimates as for employment entry were used.

5.4.3 Benefit cost ratios

Benefit cost ratios were calculated to estimate the return on investment for every £1 invested in direct and indirect costs; the central estimates show that the RTOF achieved value for money - that is, has a benefit cost ratio of at least 1.01 and is hence better than ‘break-even’:

• shortly after the end of year one (the benefit cost ratio was 0.93 at the end of year one) based on fiscal, social and economic benefits, using at least the base/central attribution of employment benefits calculated via the CIE outlined in section 5.3 above (22.1%), or when using the HM Treasury Green Book inferred measure of attribution (73.7%)
• by the end of year 2 (the benefit cost ratio was 1.13 at the end of year 2) based on fiscal benefits alone and applying the HM Treasury Green Book inferred measure of attribution (73.7%)
• shortly after the end of year 3 (the benefit cost ratio was 0.96 at the end of year 3) based on fiscal benefits alone, using at least the base attribution of employment benefits calculated via the CIE outlined earlier in this section (22.1%) or when using the HM Treasury Green Book inferred measure of additionality (73.7%)

The above analysis further suggests that the RTOF can be said to represent value for money in terms of achieving a positive return on investment (that is, the monetary benefits outweighing programme costs) shortly after 3 years under the scenarios, using the central assumption for employment benefits (derived from the CIE’s estimate of the RTOF’s additionality) and restricting the assessment to fiscal employment and housing benefits. This assessment is further supported by the fact that the cost benefit analysis does not capture non-monetisable benefits (that is, those unable to be quantified and given a monetary value), and hence will not fully account for all positive benefits achieved.

The benefit cost ratios estimated, implying a positive return on investment, are between 0.34 and 3.86. This range is based on the most pessimistic and optimistic scenarios respectively, on the basis of being calculated between one and 3 years of the cohort having achieved and/or sustained outcomes, and whether only fiscal benefits, or fiscal, social and economic benefits, are taken into account as detailed above.

5.4.4 Cost-effectiveness comparison

In addition to estimating benefit cost ratios, to further contextualise the assessment of value for money a review was undertaken of programmes that a) supported a similar cohort to the RTOF and b) were subject to analysis of their costs and benefits and hence return on investment. This cost-effectiveness review identified 6 broadly comparable programmes; however, all programmes had some notable differences from the RTOF that meant direct ‘like-for-like’ comparisons were not possible. Accepting this limitation, the review did indicate that the benefit cost ratios achieved by the RTOF compare well in this context, being broadly in line with programmes supporting cohorts that were potentially less distant from work. This adds some further weight to the positive impression of the value for money achieved by the programme, albeit subject to the caveats outlined above. Further detail on the cost-effectiveness assessment undertaken is included in Annex A.

5.5 Overall VfM of RTOF

As illustrated above, the RTOF was estimated to reach a positive return on investment shortly after 3 years under the scenarios modelled over and above the counterfactual, and using the central assumption for employment benefits (derived from the CIE’s estimate of the RTOF’s additionality) and restricting the assessment of benefits to solely those accruing to the Treasury (that is, fiscal benefits). As such, the VfM aspect of the evaluation serves to confirm the invest-to-save logic of the RTOF. On the basis of conservative assumptions, cost benefit analysis indicates that the benefits achieved relative to costs (including costs avoided) demonstrate a positive return. There are likely to be additional benefits beyond this also, for instance related to broader aspects of integration; while these are unable to be easily monetised, they serve to further indicate the VfM generated by the RTOF. From a broader cost-effectiveness perspective, while only indicative and subject to caveats concerning comparability, the BCRs achieved by RTOF also appear in line with, or compare positively to, other programmes.

6. Conclusions and lessons learned

The following section details the conclusions of the evaluation alongside lessons learned that may inform future policy and/or practice.

6.1 Conclusions

Overall, the RTOF was implemented effectively, as well as increasing employment outcomes by 39% compared to control group, and achieving a range of further positive integration outcomes for participants, thereby supporting their integration in the UK. Likewise, the evaluation has identified evidence of positive outcomes for staff, organisations, local refugee support services and government. However, delivery was not without its challenges, in particular linked to running an outcomes-based programme in the context of uncertainty over referral and engagement numbers, alongside constraints imposed by annual Departmental funding cycles. In addition, the cohort were further from employment than expected and limited access to ESOL and secure and stable housing were greater barriers than anticipated. Key lessons have also been generated from the successes and challenges of the RTOF pilot, as discussed in the following sub-section.

In terms of implementation, the SIB models adopted by projects had some key benefits, particularly in terms of encouraging a greater focus on outcomes (possible in other outcomes-based contracting (Webster 2016, p.20) and removing financial risk from the delivery organisations engaged (a direct benefit of the SIB model specifically), which were often small and from the VCSE sector. However, administration of the SIBs was perceived to limit potential efficiency and modification throughout delivery, as well as their design involving annual payment caps that limited claims for outcomes achieved. Despite this, governance and management structures supported the delivery of projects, although there was room for improved collaboration at the outset, particularly in terms of fund-level governance.

The partnership approach adopted to enhance the support offer for refugees emerged as a key advantage of the RTOF, with projects bringing together a diverse range of providers and specialisms. In particular, this approach facilitated the inclusion of providers that might otherwise have been unable to participate in outcomes-based contracts, as well as encouraging positive joint working between organisations at local levels. Data systems used were effective for managing performance and progress against targets, though were less effective in terms of delivery partners capturing and/or sharing detailed information (such as case notes) about individual participants. Projects with less developed systems faced challenges in terms of monitoring, recording and submitting outcomes claims and other data to the SDS, necessitating additional collation and analysis activity.

Project activities were generally delivered as planned, with some modifications to adapt to emerging challenges and needs. Participants commonly had positive experiences of activities, particularly appreciating the support with finding accommodation following the urgent need to find accommodation after the 28-day move-on period. Integration activities often focused on practical support, such as opening bank accounts and registering with GPs, with some additional activities around building social connections, enrichment activities and wellbeing. These elements generally aligned with the participant-led integration plans developed and participants similarly broadly welcomed both practical, individually based integration support as well as group activities promoting both the building of connections as well as improving language competencies.

In terms of outcomes achieved, the RTOF had a significant positive impact on the employment status of the refugees engaged, with RTOF participants being 39% more likely to gain employment compared to a similar comparison group (developed from administrative data sources to estimate employment impacts). Around a third of participants also achieved intermediate employment outcomes as part of enhancing employability. Intermediate employment outcomes were only applicable for those participants who were further from the labour market. Over a quarter of participants entered employment, with a higher likelihood of achieving this associated with being male, having higher English proficiency, possessing prior employment experience, and achieving a housing entry outcome beforehand. Sustained employment outcomes at 3 months were achieved by 78% of those who entered employment, with key factors in achieving this being similar to those associated with employment entry.

Entering employment was notably influenced by housing situations, with evidence indicating that participants who had safe and secure housing were twice as likely to enter employment compared to those without. Evidence thus indicates that (stable) housing is a key precursor to achieving employment, rather than employment per se being the key factor in improving housing outcomes. In terms of such outcomes, more than half of the RTOF participants found accommodation, with a higher likelihood for those in housing provided to asylum seekers or temporary accommodation at baseline. Of those who achieved a housing entry outcome, 71% went on to achieve a sustained housing outcome at 6 months. The achievement of housing outcomes was greatly influenced by the availability of local housing stock, with this being the main challenge to achieving stable and secure housing.

Integration outcomes, as defined by the RTOF payment model, were achieved by most participants, with 95% creating an integration plan and over 70% showing improvements at 6 and 12 months. The high achievement rates for integration outcomes reflect the role of developing and monitoring refugees’ integration plans as a standard and integral part of RTOF support. There was also evidence to suggest that positive integration outcomes were interconnected and interdependent, with positive progress in the domains of housing and employment being key alongside broader support in generating positive progress towards integration. This reflected part of the rationale for the RTOF’s design, confirming that integration is multidimensional and comprised of interdependent and reinforcing aspects.

Evidence gathered from qualitative interviews with investors, project and government stakeholders also indicates that the RTOF was successful in achieving a range of organisational and system-level outcomes. These concerned, in many cases, improved relationships with employers on the part of support providers, enhanced local provision and partnership working, greater organisational knowledge and understanding of participants and their needs, and a greater understanding and experience of operating in outcomes-based payment contexts because of the SIB models adopted.

The evaluation also served to confirm the invest-to-save logic of the RTOF. On the basis of conservative assumptions, cost-benefit analysis indicates that the benefits achieved relative to costs (including costs avoided) demonstrate a positive return shortly after 3 years after the achievement of positive, sustained, employment and housing outcomes. There are likely to be additional benefits beyond this also, for instance related to broader aspects of integration; while these cannot be easily monetised, they serve to further indicate the VfM generated by the RTOF.

Despite the positive outcomes achieved, including economic benefits, the extent to which participants became self-sufficient during the timescales of RTOF delivery appeared to be variable, as may be anticipated for a cohort facing multiple disadvantages (precarious housing situations, unemployment, lack of knowledge of UK systems and lower English language proficiency). However, alongside the notable achievement of outcomes related to key integration areas, the support offered will have contributed to self-sufficiency and its more thoroughgoing achievement.

6.2 Key lessons

Key lessons from the RTOF pilot and its evaluation of are presented below. The learning highlighted is intended to be applicable across different audiences. However, the lessons identified are structured according to their likely relevance for particular stakeholder groups; namely, policy makers in government, investors in SIBs, and LAs and VCSE organisations that may implement refugee support programmes.

6.2.1 For government

Annual payment caps were a key constraint linked to RTOF design that prevented the transfer of unclaimed outcome payments into later years of delivery; this can be challenging if it takes time for delivery of interventions to get fully up and running and where employment outcomes may take time to achieve; when designing future programmes, as far as possible it is likely to be beneficial to avoid use of such annual payment caps, particularly where profiling engagement numbers and outcomes achievement accurately is challenging.

Evidencing employment outcomes was challenging for providers, potentially leading to under-recording of job entry and sustainment; this highlights the potential to consider alternative mechanisms for evidencing outcomes, for example using His Majesty’s Revenue and Customs Pay As You Earn (PAYE) Real Time Information (RTI) to evidence employment entry and/or sustainment; this may also enable projects to spend more resource working directly with participants rather than gathering evidence.

Governance and management of the RTOF benefitted from broad stakeholder engagement in the Programme Board, including representation from those overseeing projects on behalf of investors and representation from other government departments; the benefits of such a structure facilitating engagement of key stakeholders can usefully be considered in future interventions, in particular to ensure transparent and effective communications alongside mitigating risks.

The focus on employer engagement as part of delivery proved effective in sourcing work placements and job opportunities; future schemes could usefully consider integrating this into programme design.

The RTOF demonstrated that achieving positive outcomes amongst the refugee cohort can take time. Where possible, longer (funding) timeframes for future interventions could be considered when outcomes take time to achieve and/or where activities are intended to facilitate outcomes that may inevitably be longer term, such as self-sufficiency; longer funding timeframes may also reduce the start-up costs associated with introducing multiple short-term programmes.

As a pilot programme, the RTOF generated and facilitated the sharing of valuable learning, highlighting the benefit of resourcing evaluation and peer learning activities within similar schemes, as well as facilitating evaluation of the programme’s employment impact via access to administrative datasets; this could be considered in the context of future programme design.

The RTOF highlighted the importance of allocating sufficient time for project mobilisation, particularly where new partnerships were forming within new contracting and financing structures; future interventions could consider this when developing implementation timelines.

RTOF delivery demonstrated the importance of designing effective systems and mechanisms for data capture and sharing, with these being key to reliable and efficient reporting alongside monitoring progress and performance; as part of programme design, this also highlights the need to focus on developing effective systems that are joined up across all providers from the outset, and that are able to capture and share detailed information about participants to give all involved in delivery sufficient understanding of individual participant situations and implementation progress.

Limited ESOL provision and long waiting lists for language support were key constraints for the RTOF in the context of language proficiency being central to achieving employment and other integration outcomes; this confirms the centrality of support for language development in the context of similar interventions supporting refugees and could usefully be considered in the context of future intervention design and implementation.

6.2.2 For investors

There were clear positives in terms of the RTOF’s approach to partnership working; such working, supported by capacity building amongst delivery partners, was important from the outset; new partnerships needed regular, clear, in-person communication to ensure alignment on delivering programme requirements; investors and those overseeing or managing projects on their behalf could usefully focus on these aspects to promote effective delivery.

RTOF delivery highlighted the importance of ensuring that clarity on specific requirements related to eligibility criteria, and the claiming of outcomes, are fully communicated to all providers from the outset, with sufficient information and guidance provided to avoid any misinterpretation; likewise, the importance of communicating the potential consequences of not achieving progress targets to providers from the outset of delivery, flagging risks as early as possible, emerged as a related lesson.

The RTOF highlighted the challenges in supporting refugees with housing in particular. In collaboration with providers, investors and those overseeing projects on their behalf are likely to benefit from a significant focus on how projects that rely heavily on supported housing, in the context of limited housing options, can be supported to get participants into other types of accommodation that may better enable them to move in to work (without associated cost barriers).

There were some miscommunications about aspects of RTOF delivery (such as eligibility criteria and specific requirements to achieve outcomes) which affected progress; this further confirmed the critical role of effective communication in such programmes, particularly in SIB structures where there are multiple layers of management and stakeholders; future programmes could usefully focus on ensuring messaging from management through to frontline delivery in such contexts is consistent and information clarified at every opportunity.

6.2.3 For LAs and VSCE providers

Some RTOF projects were successful in their engagement of landlords in the private rented sector to mitigate the need for credit checks and references, as well as providing up-front payments for multiple months of rent and deposits, indicating potentially successful approaches if resources and local contexts permit; engaging private sector organisations and landlords in partnerships using a similar approach could support participants to access the private rented sector more easily.

Given the centrality of language proficiency to integration, where possible consideration might be given to facilitating greater opportunities for English language development within projects, such as buying-in ESOL provision or providing longer term language development activities; alongside this, RTOF delivery highlighted the need for, and benefits of, interpretation and translation services to support engagement of participants with lower English language proficiency, and where there may be no shared language amongst participants and delivery staff.

Linked to the role of language proficiency in integration, the RTOF highlighted that group activities can be beneficial in terms of the opportunities they offer participants for English language development and building social connections, which can have wider benefits for participants’ wellbeing; similarly, projects might consider focusing on looking for opportunities to connect participants with wider communities, including those from non-migrant backgrounds, as a way of supporting positive integration outcomes.

Achieving employment outcomes through the RTOF could take substantial time and resource, and it was evident that other needs must often be addressed first (particularly housing and improving English language) as key precursors to finding work; the design and delivery of similar projects might usefully focus on the type of holistic support offered through the RTOF that this implies.

Engaging employers can create valuable opportunities to increase the employability of refugees (through volunteering and work experience) and lead to meaningful employment; where possible this should be a focus of future delivery of support to refugees.

Engaging participants at the earliest opportunity was helpful to minimise the escalation of challenges and barriers to progress, and to ensure participants had a greater length of time to be able to access support services (such as while in work or once in accommodation); related to this, engaging with participants at the earliest opportunity, providing information about life in the UK to people waiting for decisions on applications to remain for example, may be beneficial.

Similarly, ongoing proactive check-ins with RTOF participants helped projects to be aware of issues escalating; whilst this used up substantial resource in the short-term, projects learned that it prevented issues needing more in-depth support later.

Working flexibly with participants and forming strong, trusting relationships was key to positive engagement with the RTOF; providers needed to be willing to reschedule appointments regularly and undertake significant activity to seek to engage those who were less inclined to participate in activities; this took time, resource and persistence to achieve and is likely to be a useful consideration in similar future initiatives.

Volunteering and work experience opportunities can clearly be beneficial for participants, while participants conversely offered benefits to the organisations involved; however, ensuring refugees were matched to suitable volunteering opportunities was important, while providers acknowledged the time and resources required to provide refugees with a positive experience; careful design and delivery of these elements could usefully be considered in future support.

In some cases, RTOF participants could be unclear on available support, particularly in the context of several organisations engaging with them; future projects could therefore benefit from ensuring that there is a clear participant journey established from the outset that can be communicated clearly to clients; it is also likely to be useful to ensure providers have a clear set of expectations to communicate to participants from their initial engagement, including what will be required of them to evidence intended outcomes if applicable.

RTOF delivery highlighted the importance of staff development, support and organisational capacity building in ensuring effective delivery. Future projects could benefit from a focus on providing wellbeing support for staff to manage the effects of working with participants who can have acute mental health needs, as well as supporting the development of expertise in data collection and entry to ensure that quality data is collected that can support monitoring of, and improvements in, delivery.

Annex A: Technical annex

Process evaluation

Aims

The primary aim of the process evaluation was to generate evidence and learning about what works to improve employment, housing and wider integration outcomes for newly recognised refugees. As a secondary aim, qualitative data was generated to inform the impact and economic assessments.

Methods and approaches

Key methods for the process evaluation included:

• scoping interviews with government stakeholders and representatives of refugee organisations (n=11)
• a series of engagement workshops with each project to test the evaluation approach (n=4)
• longitudinal interviews (at up to 3 time points) with project stakeholders including representatives from investors, local authorities, VCSE service providers and delivery staff (total n=63)
• longitudinal interviews (at up to 2 time points) with RTOF participants (total n=60)
• longitudinal interviews (at up to 3 time points) with government stakeholders (total n=11)
• reviews of participant case files (information collected by projects about participants) (total n=8)
• interviews with representatives from community organisations not directly commissioned to deliver RTOF (total n=3)
• peer consultations carried out by (refugee) members of the Steering Group

The process evaluation also analysed data from the Shared Data System about participants and their outcomes, and incorporated insights from learning workshops run by Social Finance to support projects. Data was sorted and analysed using the software package NVivo, using frameworks to thematically analyse data and examine how experiences or views varied by stakeholder type.

Counterfactual impact evaluation

Aims

The aim of the counterfactual impact evaluation (CIE) was to estimate the impact of RTOF on employment for those supported (the treatment group), relative to a statistically ‘matched’ comparator group. By comparing outcomes between the treatment and comparison groups, the additionality of RTOF support (over business as usual support) can be estimated.

Data accessed

The following datasets were accessed and used for the CIE:

Home Office Refugee dataset: this included National Insurance Numbers (which enabled linking across datasets), key characteristics (age, gender, dependent children, nationality) date of arrival in the UK, date of refugee application approval.

His Majesty’s Revenue and Customs (HMRC) employment data: this included National Insurance Numbers, dates and amounts of employee payments, and data on self-employment. Other variables (for example, sector of employment) were provided but not used owing to high-levels of missing.

Department for Work and Pensions (DWP) Universal Credit assessment data (UC): this included National Insurance Numbers, claimant details, Universal Credit assessment period dates and type.

RTOF Shared Data System (SDS): this included many variables for the treatment group (see separate section on analysis of SDS) but the only ones used for the impact evaluation were RTOF start dates, which were used to simulate start dates for the comparator group (see below) and calculate employment before and during/after RTOF support.

Data linking and cleaning

National Insurance numbers and other identifying variables from the Home Office’s refugee dataset were shared with HMRC and DWP so that any employment and UC records could be matched and shared back to the Home Office. Final datasets, with identifying variables replaced with a unique identification number, were then shared with Ecorys. The unique identification number allowed for matching across the different datasets but was otherwise meaningless (that is, the ID could not be used to identify individuals in other datasets/databases).

During the data cleaning, inconsistent date formats were identified and as far as possible adjusted for to mitigate errors. This included a mix of day/month/year, and month/day/year formats for the dates when refugees arrived in the UK and when their asylum application was approved. Whilst many dates could only be one or the other format (and were retained) others could technically be either (for example, 12/06/2022 could be 12 June or 6 December). Where this was the case, a mid-point was calculated. Whilst this did have implications for simulating RTOF start dates for the comparator groups (see below), sensitivity tests were conducted and are reported.

Simulating start dates

To estimate the difference between the treatment and comparator groups on employment outcomes, it was necessary to calculate employment status/periods before and during/after RTOF support. Whilst those supported by RTOF were newly (or recently) approved refugees and were not anticipated to have substantial (if any) prior employment in the UK, it was important for this to be considered and adjusted for in the CIE. Generally, adjusting for pre-intervention outcomes helps strengthen most CIEs because this is a strong predictor of future outcomes. Analysis of the treatment group also identified some (8%, see Table A1: Treatment and comparison group before propensity score matching) did have some employment prior to RTOF.

RTOF start dates were available for the treatment group in the SDS data. By definition, the comparator group did not have an RTOF start date, so these had to be simulated/estimated. The approach to simulating start dates for the comparator group involved:

Model training: by training the model on the treatment group data, a relationship between refugee application approval dates and start dates was estimated. This included:

• transforming approval dates to ‘days to start’
• trimming implausible start dates and/or ‘days to start’ (for example, where likely owing to recording errors it appear as though some started RTOF before having their application approved)
• adjustment for application approval dates that were before RTOF projects were operational (where individuals had to wait)
• additional covariates were tested but these did not substantially improve the overall explanatory power of the regression model so were omitted

Simulation for comparator group: this trained model was then applied to the comparator group. Using their refugee application approval dates, the model generated simulated RTOF start dates for these individuals.

Prior to calculating propensity scores and impact estimates, the final dataset was restricted to those with start dates when RTOF was operational. This resulting in a final sample size of 7354 in the comparator group and 1447 in the treatment group. There was good balance/overlap between the 2 groups on start dates and other covariates that were subsequently used in the propensity score weighting (see Table A1).

Propensity score weighting

To ensure as robust and reliable as possible impact estimates, it was important to ‘match’ individuals in the treatment and comparator groups based on factors that likely affect outcomes (other than RTOF support). These factors are referred to as covariates. A common approach to matching individuals on many covariates involves estimating the propensity (likelihood) of being treated based on the covariates.

Propensity scores were calculated using logistic regression, where the outcome was treatment status and used the following explanatory variables (that were available for both the treatment and comparator groups):

• age at RTOF start
• nationality (see below)
• gender
• linked dependants
• year of arrival in UK
• RTOF start date (as days since start)
• Pre-RTOF employment (where earning >£1000 = Yes)

For nationality, a binary indicator for common nationality was created to help match the treatment and comparison groups more effectively. With many nationalities present in the dataset, including each one in the regression models would have complicated the analysis and reduced reliability due to small sample sizes in each group. The binary indicator grouped the most frequent nationalities together, simplifying the models and ensuring clearer, more robust statistical comparisons. This approach enhanced the matching process and improved the overall quality of the impact evaluation.

Claiming Universal Credit was also tested to calculate propensity scores but this did not have a substantial impact on the impact evaluation results.

The resulting propensity scores were then used to weight the data. Individuals in the comparator group that were more like those in the treatment group (that is, had a similar propensity score) received a higher weighting than those who were less similar. Propensity score weighting was used rather matching because it allowed for the inclusion of the entire sample, ensuring better statistical power and more generalisable results.

The weighting approach was successful in minimising the differences in covariates between the treatment and comparator groups. An assessment of this covariate balancing is provided below.

Estimating impacts

As described previously, employment status before and during/after RTOF were calculated based on RTOF (or in the case of the comparator group, simulated) start dates. There was a high proportion of missing employment start and end dates in the HMRC data so payment dates and amounts were used to approximate when an individual entered employment. If an individual had a payments totalling £1,000 or more before or after their RTOF start date, this was flagged as in employment. The £1,000 criteria was used to help minimise potential errors or very short term employment periods, where claiming entering employment as an outcome would be potentially misleading.

To estimate the impact on employment, logistic regression models were used. For the primary model, the outcome was in employment after RTOF start date, and the explanatory variable was treated (that is, supported by RTOF or not). The regression models used the propensity scores as a weight.

As noted previously, there were issues with the dates provided for refugee application approvals, which RTOF start dates and pre/post-employment periods were based on / influenced by. To explore whether this potentially biased the CIE, the following sensitivity tests were also conducted:

Sensitivity model 1: Calculating propensity scores using the same variables as the above with the exception of pre-RTOF employment. The rationale for this was that if pre-RTOF employment status was incorrect, this could bias the results.

Sensitivity model 2: As above but without RTOF start date (application approved year was used instead), and the outcome any employment (that is, before or after the RTOF start date) in the regression model. The rationale here was that, theoretically (and apparent in the data), most new/recently approved refugees would not have been in employment prior to RTOF.

Findings

There was a statistically significant and positive impact on employment status for those supported by RTOF, relative to the comparator group, across all models. The treatment effects expressed as odds ratios are presented in Figure A below. In summary:

• the treatment effect for the primary model was 1.385; in other words, those in the treatment group were 38.5% more likely to enter employment than those in the comparator group
• the treatment effects resulting from the sensitivity tests were smaller (odds ratios of 1.28 and 1.24) but the overlapping (95%) confidence intervals between models support the overall finding of a positive impact on employment status

Figure A: Treatment effects on employment status: Odds ratios and confidence intervals

Methodology for calculating comparator group employment rate and attributable employment

Below are the calculations used to estimate the comparator group (baseline) employment rate and determine the number of jobs attributable to the intervention.

s1. Estimating the comparator group employment rate.

The employment rate in the comparator group (P(C)) is calculated using the treatment group employment rate (P(T)) and the odds ratio (OR):

The formula uses 1−P(T) to account for the proportion of the treatment group not employed, ensuring the adjustment considers both employment and non-employment probabilities. The term OR×(1−P(T))+P(T) rebalances these probabilities to reflect what would happen in a comparable group without treatment, maintaining the correct relationship between odds and probabilities.

Inputs:

• P(T): Employment rate in the treatment group (for example, 613/2331 = 26.3%)
• OR: Odds ratio comparing treatment and comparator groups (for example, 1.385)

Worked calculation:

This means the comparator group’s baseline employment rate is approximately 20.5%.

s2. Calculating the number of jobs attributable to treatment.

Once P(C) is determined, the expected number of employments in the treatment group without treatment is calculated:

Worked calculation:

The additional jobs attributable to the intervention are then calculated as the difference between actual and expected employments, which is 613 – 478 = 135. The percentage of attributable employments is 135/613 = 22.1%

Shared Data System (SDS)

RTOF projects collected data on those supported through the SDS. The SDS was developed by Social Finance and Ecorys, with Social Finance then responsible for implementing and quality assuring the data collection. The SDS was used to record outcomes achieved for individuals (allowing projects to claim for outcomes), and for learning and evaluation purposes. Variables included background characteristics, and the details and dates of outcomes achieved/claimed.

Variables in the SDS were largely closed and validated responses and therefore required minimal data cleaning. Data cleaning focused on recoding and grouping responses to facilitate analysis (see logistic regression model specifications below). Whilst completion was generally good across projects, there were some variables with high levels of missing data. Base and missing sizes are reported in all tables, and factored into the interpretation where relevant.

The SDS data provided to Ecorys was spread over multiple Excel Sheets. To allow for a more comprehensive analysis (for example, exploring outcomes achieved by different background characteristics) the Sheets were joined in R (statistical software) based on the (anonymised) unique IDs provided. All individuals with a matching IDs across the ‘person’ and ‘baseline’ Sheets were retained for analysis. The final sample size was 2,340.

All analysis was conducted in R. Descriptive statistics for participant outcomes and characteristics are provided. Logistic regression was chosen over subgroup analysis to better understand the potential factors influencing the outcomes. Logistic regression allows for the examination of multiple variables simultaneously, providing a comprehensive picture of how different factors contribute to the likelihood of a particular outcome. This method enables the control of various confounding factors, ensuring that the true effect of each variable is captured while holding others constant. As a result, the insights gained are more robust and reliable.

To balance sample sizes (cases with missing predictor variables were removed) and insights, 2 logistic regression models were tested on each outcome:

Model 1 included data for 1535 participants that had complete data for the following factors:

• gender: recoded to a binary/dummy variable where male = 1
• age group: grouped by 18 to 24, 25 to 29, 30 to 34, 35 to 39, 40 to 44, 45 to 49, 50 and over
• nationality: recoded as a binary variable (see below)
• language level: grouped by below entry level, entry level (1 to 3), and above entry level
• dependants: recoded to a binary where dependent children = 1
• housing situation at baseline: grouped by asylum accommodation, homeless, other safe or temporary accommodation, and renting (private or local authority)
• RTOF project/area: this was only included to adjust for (unobserved) area-level effects (for example, employment opportunities and wider support services for refugees); results should not be interpreted as differences in the effectiveness of support provided by projects

Model 2 included data for 1,107 participants that had complete data for the above factors and the following additional factors:

• in education at baseline (RTOF start): recoded to binary where in education or training = 1
• prior (to RTOF) employment: recoded to binary where prior employment = 1

An additional model for the employment entry outcome was tested; this was the same as model 2 but with housing entry outcome as an additional factor. The reason for this was to explore if there was an association between housing and employment outcomes

Regarding nationality, a binary indicator for common nationality was created to facilitate the logistic regression analysis. RTOF projects supported many nationalities (80+), which, if included individually, would have made the models overly complex and less reliable due to small sample sizes within each nationality group. By grouping the most common nationalities together into a binary indicator, the analysis became more manageable and focused. This approach ensured clearer insights and more robust statistical comparisons. The definition used for common nationality was 100 or more RTOF participants, and included: Afghanistan, Eritrea, Iraq, Iran, Sudan, Syria. Participants from these nationalities may have had more similar experiences/needs than others that accounted for fewer participants.

Counterfactual impact evaluation additional outputs

This section includes:

• comparison of treatment and comparator groups before propensity score weighting
• comparison of treatment and comparator groups before and after propensity score weighting (that is, covariate balance)
• outputs from the regression models used to estimate impacts

Treatment and comparison groups

Table A1: Treatment and comparator group before propensity score weighting

Variable	Comparator	Treated
n	7354	1447
Average age at start	30.71	32.24
Common nationality	68%	70%
Males	78%	79%
Dependants	13%	14%
Average year of arrival in UK	2019.52	2019.75
Average year of application approval	2021.38	2021.7
Employed before RTOF	13%	6%

Covariate balance

The propensity score weighting minimised the differences in covariates “prop.score” is the average propensity score, which provides an overall measure of balance between the groups

Table A2: Covariate balance for primary model

Covariate	Type	Standardised mean difference before	Standardised mean difference after
prop.score	Distance	0.42	0.06
age_at_start	Contin.	0.18	0.04
nationality_common	Binary	0.02	-0.01
gender_m	Binary	0.01	0
linked_dependants	Binary	0.01	0.01
year_arrived	Contin.	0.22	0.06
days_since_start	Contin.	0.39	0.1
pre_employment	Binary	-0.07	-0.02

Table A3: Covariate balance for sensitivity model 1

Covariate	Type	Standardised mean difference before	Standardised mean difference after
prop.score	Distance	0.4	0.04
age_at_start	Contin.	0.18	0.03
nationality_common	Binary	0.02	-0.01
gender_m	Binary	0.01	0
linked_dependants	Binary	0.01	0.01
year_arrived	Contin.	0.22	0.06
days_since_start	Contin.	0.39	0.08

Table A4: Covariate balance for sensitivity model 2

Covariate	Type	Standardised mean difference before	Standardised mean difference after
prop.score	Distance	0.4	0.07
age_at_start	Contin.	0.18	0.03
nationality_common	Binary	0.02	0
gender_m	Binary	0.01	0
linked_dependants	Binary	0.01	0
year_arrived	Contin.	0.22	0.07
year_approved	Contin.	0.42	0.15

Impact regression output

Table A5: Logistic regression outputs

Model	term	Odds ratio	95% CI (Lower)	95% CI (Upper)	P-value
Primary model	(Intercept)	0.600	0.573	0.628	0
Primary model	treated	1.385	1.242	1.545	0
Sensitivity model 1	(Intercept)	0.609	0.582	0.638	0
Sensitivity model 1	treated	1.280	1.140	1.436	0
Sensitivity model 2	(Intercept)	0.620	0.592	0.650	0
Sensitivity model 2	treated	1.236	1.101	1.387	0

Value for money assessment

Aims

The VfM assessment addressed the third aim identified at the start of the evaluation, namely, to test the economic case for future investment in refugee support. The evaluation objectives related to this aim were to:

• generate data to validate projections of cumulative savings to HM Government after completion of the programme, and non-cashable benefits (where possible)
• assess VfM in respect of the funded projects and make recommendations on how costs can be driven down without compromising outcomes

The VfM assessment also fed into the other 2 evaluation aims - namely, to generate learning about what works in improving self-sufficiency, wellbeing and integration outcomes for newly recognised refugees; and to evaluate whether and how the outcomes funding mechanism and involvement of social investors have led to more efficient delivery.

To inform the approach, evaluation aims and objectives relevant to the VfM assessment were mapped against evaluation questions from the overall evaluation framework, as well as to evidence sources and methods. The approach to the VfM assessment included Cost Benefit Analysis (CBA) - to monetise the fund’s outcomes as ‘benefits’ - and Cost Effectiveness Analysis (CEA) - to compare cost per outcome across comparator projects, where appropriate to do so. Both techniques involved analysis of the costs and outcomes/benefits of the intervention to inform the VfM assessment. In addition, CBA and CEA analyses were supplemented with findings from the qualitative research to clarify and triangulate elements of the analyses.

RTOF VfM approach within the evaluation framework

Three objectives were identified within the evaluation framework that are summarized in Tables B6, B7 and B8 below.

Table A6: Objective 1 evaluation framework

Objective 1: generate learning about what works in improving self-sufficiency, wellbeing and integration outcomes for newly recognised refugees.

Aim	Key evaluation questions	Evidence source	Method
1. Generate recommendations on how to embed learning from the programme in the wider refugee support sector; and how to build the findings into future commissioning of mainstream services and tailored support for refugees.	To what extent and how did the role of social investment contribute to sharing the financial risk of Payment by Results (PbR) in refugee support activities?	Views of government stakeholders / views of social investors / views of project leaders	Qualitative

Table A7: Objective 2 evaluation framework

Objective 2: evaluate whether and how the outcomes funding mechanism and involvement of social investors have led to more efficient delivery

Aim	Key evaluation questions	Evidence source	Method
1. Evaluate total costs of an outcomes funding approach where multiple pay-for-success contracts are launched using a single design process.	To what extent did the programme reduce the costs on the public purse, such as a reduction in benefit claims	Value for money assessment	CBA
2. Evaluate total costs of an outcomes funding approach where multiple pay-for-success contracts are launched using a single design process.	To what extent has the RTOF improved the evidence base on whether the outcomes funding model delivers better outcomes in terms of refugee integration?	Views of HO/stakeholders involved in performance management of RTOF. Value for money assessment	CEA
3. Evaluate whether this type of outcomes fund can deliver greater economic benefits and savings than alternative forms of service provision and, if so, what are the critical success factors for this to hold true.	To what extent did the outcomes funding model incentivise a stronger focus on outcomes? To what extent did the outcomes funding model contribute to increased: - innovation - efficient delivery - adaptive management? How did any observed improvements in performance compare to those observed in non-outcomes funded contracts? What were the key contributing factors that led to any observed differences?	Views of strategic and operational project stakeholders. Value for money assessment Views of HO/stakeholders involved in performance management of RTOF.	CEA / Qualitative
4. Evaluate the ‘value-add’ of an outcomes funding structure, compared to other available contract and grant types, and validate savings projections to HMG.	To what extent and how did the programme increase the capacity of government and other organisations to support refugee integration?	Views of strategic and operational project stakeholders. Value for money assessment Views of HO/stakeholders involved in performance management of RTOF.	CBA / CEA
5. Identify whether an outcomes fund model delivers better value for money than alternatives, and identify the specific service delivery elements to which successful outcomes are most closely linked.	(If relevant data on other programmes is available) how does the cost per outcome in the SIB model compare to cost per outcome of other programmes, for example, AMIF?	Views of strategic and operational project stakeholders. Value for money assessment Views of HO/stakeholders involved in performance management of RTOF.	CBA / CEA / Qualitative

Table A8: Objective 3 evaluation framework

Objective 3: build the economic case for future investment in refugee support

Aim	Key evaluation questions	Evidence source	Method
1. Generate data to validate projections of cumulative savings to HMG after completion of this programme	N/A	Programme costs/evidence of programme impact/estimation of costs/savings	CBA
2. Generate data to validate projections of non-cashable benefits (where possible)	N/A	Programme costs/evidence of programme impact/estimation of costs/negative outcomes avoided (where possible)	CBA (where possible)
3. Assess value for money of the funded projects and make recommendations on how costs can be driven down without compromising outcomes	How could costs be driven down without compromising outcomes?	Views of strategic and operational project stakeholders. Value for money assessment Views of HO/stakeholders involved in performance management of RTOF.	CBA / Qualitative
4. Evaluate the ‘value-add’ of an outcomes funding structure, compared to other available contract and grant types, and validate savings projections to HMG.	To what extent and how did the programme increase the capacity of government and other organisations to support refugee integration?	Views of strategic and operational project stakeholders. Value for money assessment Views of HO/stakeholders involved in performance management of RTOF.	CBA / CEA
5. Identify whether an outcomes fund model delivers better value for money than alternatives, and identify the specific service delivery elements to which successful outcomes are most closely linked.	(If relevant data on other programmes is available) how does the cost per outcome in the SIB model compare to cost per outcome of other programmes, for example, AMIF?	Views of strategic and operational project stakeholders. Value for money assessment Views of HO/stakeholders involved in performance management of RTOF.	CBA / CEA / Qualitative

Cost benefit model review

During the programme design stage, Social Finance, with inputs from the Home Office and Ernst and Young, developed a cost benefit model for the RTOF. The model informed a separate financial model, which fed into the economic case for investment that featured in the fund’s outline business case. This was later the basis for the development of the RTOF rate card and projections for programme expenditure.

To inform the VfM approach, the evaluation team reviewed and updated the cost benefit model. This included reviewing and updating the unit costs and assumptions within it to inform the valuation of benefits resulting from programme outcomes. Where the cost savings arising from the outcomes identified related to unit costs or assumptions from the model, the unit costs and assumptions concerned were updated with more recent data and/or sources.

Derivation of benefits from cost benefit model

The table below presents the extent to which the programme generated fiscal cross-government savings. It draws on the unit costs outlined above as applicable, applying these to the outcomes identified in the main body of the report using the approach and assumptions to calculate benefits detailed therein. Figures are presented per individual participant outcome.

Table A9: Benefits per participant outcome for housing and employment

Benefits	Months 0 to 3	Months 9 to 12	Year 1 (Months 0 to 12)	Year 2 (Months 13 to 24)	Year 3 (Months 25 to 36)
Employment
Health	£159	£478	£637	£1,252	£1,866
Indirect tax	£321	£962	£1,283	£2,521	£3,759
Direct tax	£0	£315	£315	£618	£922
National insurance	£188	£564	£752	£1,478	£2,204
Council tax	£266	£798	£1,064	£2,091	£3,117
Universal Credit standard allowance	£1,035	£3,105	£4,139	£8,134	£12,129
Universal Credit child allowance	£466	£1,398	£1,864	£3,662	£5,461
Local Housing Allowance	£1,352	£3,172	£4,524	£8,890	£13,256
Total	£3,787	£10,791	£14,578	£28,646	£42,713
Housing
Housing stability	-	-	£4,537	£8,916	£13,294
Reduction in rough sleeping	-	-	£645	£1,268	£1,890
Total	-	-	£5,182	£10,183	£15,184

Cost effectiveness analysis of RTOF

Table A10 illustrates benefit cost ratios from a desk review of evaluations of similar programmes, as well as programmes funded by alternative forms of service provision and other contract types, where data existed. Although only 6 programmes/projects provided a benefit-cost ratio, the wider review comprised a total of 38 programmes and/or projects. Of these, 23 included at least one of the same outcomes as RTOF, and 19 of them included some kind of cost evaluation (for example, cost per person, or cost per outcome). Programmes that had similarity to RTOF (compatibility was assessed based on the similarity of the Population, Intervention, and Setting - part of PICOTS) but did not have an evaluation yet or any outcome data were still included for completeness

Table A10: Benefit cost ratios of programmes

Study	SIB?	BCR	Link to RTOF
Local Authority Asylum Support Liaison Officer (LAASLO) project 1 / 2	No	0.83 / 1.42	Housing
Evaluation of the Traineeships Programme: final report 2015 to 2019	No	Between 2.71 and 5.89	Employment
The Work Programme: impact assessment	No but Payment by Results	£1.75, with a range of £1.16 to £2.3 per £1 spent	Employment
Innovation Fund	Yes	1.3 for round 1 and 1.25 for round 2 (not based on statistically significant results)	Employment
Supporting Families (also called the Troubled Families programme)	No	For every £1 spent, £2.28 of economic benefits, and £1.51 of fiscal benefits.	Employment
Evaluation of Jobs Growth Wales 2015 to 2019: final summative evaluation report	No	2.41 to 2.61	Employment

It is difficult to directly compare the ratios in Table B1 to those calculated for the RTOF programme, given that the RTOF programme is unique in many regards and quite different from the programmes listed, though does have some similarities.

Support for resettled refugees or asylum populations cannot easily be used as comparable programmes because these cohorts all have very different rights and experiences, and there has been limited support for newly recognised refugees historically. The RTOF appears to be the first programme that has holistically focused on housing, employment, and wider elements that support integration for newly recognised refugees. Arguably, the lack of opportunity for direct comparison demonstrates that RTOF is innovative, and there appears to be little evidence that such a holistic programme has been implemented before. Similar programmes cover employability and/or housing, but there is little discussion of wider integration support, reinforcing that this evidence base is limited in the UK context, for newly recognised refugees.

Nevertheless, it appears that the BCRs appear broadly in line with other programmes. This is particularly positive given the multi-faceted nature of the programme that focuses on employment, housing and integration - as noted in the right-hand column in Table A10, other programmes tend to focus on one (or from the wider review, up to 2) outcome areas. Additionally, as indicated in preceding sections, the cohort is a complex one to support into work. Moreover, the BCRs calculated for the RTOF relate to only 3 years, whereas other programmes stretch beyond that.

Annex B: Design and Implementation

This annex details the longer form process evaluation findings, including additional detail on characteristics of the RTOF cohort, the Design of the Fund and various activities.

Characteristics of the RTOF cohort

Key features of the characteristics of those engaging with RTOF support are summarised below.

Cohort characteristics

Age group: the age distribution of participants varied with, proportionally, the largest age groups being 25 to 29 (26%, n=600) and 30 to 34 (23%, n=547). Smaller proportions were seen in the youngest (18 to 24) and oldest (50 and over) age groups, accounting for 14% (n=327) and 5% (n=117) respectively.

Gender: the cohort was predominantly male (73%, n=1,713), with females representing 23% (n=538) of participants. There were very few participants who identified as non-binary or other genders.

Dependants: a majority of participants (67%, n=1,563) did not have dependants. Around 20% (n=474) had dependants, while there was missing data for around 13% of participants.

Nationality: the cohort included a broad range of nationalities, with the largest groups being Iranian (24%, n=552) and Eritrean (17%, n=393).

Highest qualification achieved: a third of participants (36%, n=836) had above entry-level qualifications, while 22% (n=517) were entry level or below. There were substantial proportions where qualifications were unknown (12%, n=273) or missing (31%, n=714).

Employment status in home country: 44% of participants (n=1,026) had been employed in their home country, while 26% (n=612) had not been employed. There was a substantial amount of missing data (30%, n=702).

Language level on entry: a notable proportion of participants (45%, n=1,048) were at entry level 1 to 3 in English proficiency, while nearly a fifth (19%, n=455) were pre-entry or below formal levels; a similar proportion were above entry level of English proficiency (19%, n=452). There was a notable proportion of missing data (17%, n=385).

Education and training status: almost half (46%, n=1,074) of participants were not in formal education at the start of RTOF, while 21% (n=495) were in formal education. A notable proportion (33%, n=771) of education status data was missing.

Housing baseline accommodation: half of participants (50%, n=1,167) resided in accommodation provided to asylum seekers at the time they engaged with RTOF projects. Other significant categories included publicly funded temporary accommodation (9%, n=214) and various forms of homelessness (8%, n=187).

Prior employment: almost half of participants (47%, n=1,100) were not employed in the UK before engaging with RTOF projects, and 21% were employed (n=490); there was also a notable proportion (32%, n=739) of data missing.

Design of the Fund

The RTOF was designed to primarily focus on improving employment outcomes for participants, alongside supporting with housing and wider integration to offer a holistic service that recognised the connections between various key integration elements. Projects delivered a range of activities to meet the needs of participants, covering employment, housing and wider integration support (including support for language development, social connections and wellbeing).

Delivery models and approaches

Greater Manchester: 10 LAs provided housing and integration activities; VCSEs provided employability activities. All activities used a caseworker model to provide personalised integration support; later in delivery, VCSEs were contracted to provide activities for further integration activities.

North East: delivery models varied across LA areas, with some LAs providing housing and integration support, while in other areas VCSE organisations delivered those elements. A VCSE provider was embedded in project housing and integration teams to support employment outcomes. A specialist employability provider worked across the region to support participants and source referrals.

Plymouth: 2 VCSE providers worked in partnership with the LA, managed by the SPV. One provider delivered housing activities and the other delivered employment and integration activities.

West Midlands: each LA area had their own local model, which usually comprised some LA oversight and VCSE delivery, with one area having a VCSE provider responsible for management and delivery.

Governance and management

Fund-level governance included the RTOF Programme Board attended by cross-government partners (Home Office, Department for Culture, Media and Sport, and Department for Work and Pensions) and investors. Fund-level management included monthly meetings between Home Office and investors and monthly data audits.

The Programme Board did not include investors from the outset, which was perceived by some stakeholders as missed opportunity. Including investors was cited as useful to understand performance, challenges and directly ask questions. Final interviews with investors indicated that greater involvement and collaboration of all stakeholders at the Programme Board would have been appreciated from the outset.

Project-level governance and management varied across projects but generally included investor level governance, monthly whole-partnership meetings, and one-to-one engagement between local authorities (with strategic oversight) and VCSE providers. Projects perceived these structures to be effective for bringing providers together and to enable partnership working. However, some stakeholders thought that the additional layers of management associated with outcomes-based financing was cumbersome and could detract from frontline delivery.

Implementation of the Fund

The following sub-sections cover referral and engagement processes, including the patterns within the overall RTOF cohort engaged, partnership working and data systems, and the effectiveness of activities related to employment, housing and wider integration. Participants’ experience of engaging with RTOF support is also examined.

Referral and engagement patterns and processes

Overall, referral and engagement processes were somewhat effective, although they were challenged by inconsistencies or inaccuracy of data about prospective participants, as well as notable levels of disengagements, further detailed below.

Over a quarter of RTOF participants (26%, n=606) were referred through data from the Home Office and providers of accommodation for asylum seekers. This was not always the most effective referral route as information could be out of date, for example where people had changed their contact details, which was reportedly common amongst the cohort. Conversely, where data was accurate, using these sources was beneficial in enabling projects to contact potential participants at the earliest opportunity. Projects reported that peoples’ situations could be more acute, for example in terms of housing needs, when engaging at a later point through LA or VCSE routes.

Around a quarter (27%, n=636) of participants approached projects themselves (19%, n=442) or via peers (8%, n=194). This suggests that there was some awareness of RTOF within the community. Participants could also be told about RTOF by other services, later approaching projects themselves. Projects also sourced referrals through leveraging connections with local VCSEs. Around 13% (n=311) of participants were referred through VCSEs.

Project representatives reported that initial recruitment was rushed to meet targets, as targets would affect outcomes payments to investors. Some participants were not ready for employment, with notably lower levels of English language proficiency. Ultimately, projects felt this contributed to slow progress in achieving outcomes or, in some cases, participants disengaging.

There were lower numbers of people being granted refugee status or humanitarian protection than estimated at design phase. This led to a decision to work with some ineligible people in some projects, including those deemed ‘less suitable’ to pursue employment (including single parents with caring responsibilities and those above working age). Relatedly, some stakeholders felt that the need to be ‘economically active’ to meet RTOF eligibility criteria could restrict the ability of projects to work with pregnant women, those with young children or people with disabilities who may later have been looking for work within the delivery timescales.

Delays to Home Office decision making on grant applications also meant that some participants were at the end of their decision appeal period. Project representatives noted this could mean their needs had increased in complexity, most notably in respect of mental health challenges (which was also a barrier to sustained engagement in itself).

Quantitative data only shows referrals that led to successful engagements. In the context of future programmes, gathering data covering all referrals, including people who never engaged and people who disengaged very quickly could provide a greater picture of referral routes.

Table B1: Referral sources

Referral source	Count	Percentage
DWP	14	0.6
HO providers (of asylum accommodation)	143	6.1
Home Office	463	19.8
LA	360	15.4
Local VCS	311	13.3
Migrant Help	7	0.3
Other	162	6.9
Peer	194	8.3
Primary care	1	0
Self-referrals	442	18.9
Missing	242	10.3

Source: Ecorys analysis of SDS

There was limited evidence to indicate that projects took a targeted approach to generating referrals for people with particular demographic characteristics, with projects instead taking a broad approach to engagement.

Factors that supported engagement included coverage of a range of languages within staff teams (and strong links with communities and refugee community organisations); adopting flexible, person-centred approaches driven by participant interests (including social activities); and having accessible, conveniently located offices, or offering home visits for those with young children or those who may struggle with travel.

Disengagement

There were relatively high levels of disengagement from RTOF support, reportedly higher than projects anticipated, with the SDS recording 958 participants as having disengaged due to a range of factors discussed below. However, there are likely to be greater numbers of those who disengaged recorded on local data systems that were not provided in data submissions.

Project representatives suggested that the highest levels of disengagement came when participants needed to provide evidence to validate outcomes, which was sometimes perceived as being “intrusive” or burdensome. Participants were also seen as more likely to disengage after making progress or achieving particular goals, including housing entry, as they could feel they no longer required support. Other factors influencing engagement included mental health and wellbeing, English language proficiency, status of family reunion application, or housing circumstance. In particular, higher levels of disengagement after successfully achieving entry to accommodation outcomes were noted by project staff (often the priority issue for participants, given many were engaged after receiving their 28-day move-on notice from asylum housing).

Another reason for disengagement was participants moving away from RTOF areas. Key reasons for moving (reported in the SDS and by project stakeholders) included perceived greater availability of housing, work and education opportunities elsewhere and/or being closer to friends, relatives or familiar communities. Where participants moved to another area offering RTOF provision, some projects sought to maintain engagement via arranging handover meetings or, otherwise, continuing support remotely. There was limited data to understand the effectiveness of these processes.

Partnership working and data systems

RTOF partnerships, formed of investors, regional bodies (such as migration partnerships in some projects), LAs, and providers (most commonly VCSE organisations), generally worked effectively. After initial challenges during early mobilisation, partnerships formed and settled into delivery successfully, with providers reported to be working as cohesive teams beyond organisational boundaries. It was clear that providers generally worked through any partnership challenges, exchanged knowledge, and shared their respective expertise through regular partnership meetings, facilitated by co-location or close office proximity. In particular, a key finding was that many providers brought together through RTOF would not have considered working together previously. Project representatives thus considered partnerships, and the joint working encouraged, to have introduced a new and positive partnership approach in some local areas.

In more limited contexts, it was also apparent that partnerships were less coherent and effective at particular points in delivery, particularly where there was continued staff churn, or limited management and oversight. Where there was underperformance or challenges, some projects actively managed this by procuring new providers or rescoping activities and remits, which was seen as effective in addressing the issues encountered. For example, one project brought integration and housing teams together under one provider, while another added employer engagement into a well-performing provider’s scope. Collaborating with Jobcentre Plus and the Department for Work and Pensions also reportedly supported effective employment activity in terms of identifying local opportunities and understanding additional local provision.

To aid partnership working and delivery more generally, projects employed a range of data systems, including both unified Customer Relationship Management (CRM) systems across whole delivery areas and more localised or segmented approaches. Data systems varied in effectiveness in terms of supporting partnership working and capturing progress and outcomes. Shared CRM systems were evidently effective for measuring progress, informing performance management and, in some cases, supporting case work across providers. However, shared CRMs were often mainly oriented around progress monitoring and recording outcomes achievement, rather than recording qualitative information on participants. This meant that more in-depth information could be recorded inconsistently, either on separate case management systems, retained mentally by staff, or noted in voluntary notes sections. It was acknowledged in some contexts that this hampered effective delivery and partnership working.

More segmented or locally hosted data systems supported measuring progress and performance management but did not support casework across partners, as such information was stored locally and able to be accessed only by individual providers. This created additional layers of information and data sharing requirements and issues between providers to support casework. Collating data to input into the SDS required additional resource as multiple sources needed to be combined before submitting. In addition, in earlier phases of delivery, the SDS functionality did not include data for outcomes claims. This meant that RTOF projects had to submit 2 monthly returns, which was perceived as duplicative and inefficient (by RTOF projects). This process was unified during delivery to streamline data submissions.

The issues identified above confirm the importance, where possible, of establishing effective systems to capture casework data and facilitate effective partnership working across providers. Additionally, in cases such as RTOF where a shared system is needed for monitoring outcomes and evidencing/processing claims, findings indicate the importance of ensuring functionality and availability from the outset of delivery.

Employment activities

RTOF employability activities included coaching and mentoring around CV writing, goal setting, interview preparation, and job-searching. Projects delivered one-to-one and group activities, which were reportedly (separately or in combination) successful in seeking to meet a range of needs. Stakeholders across 2 of the project areas described connecting participants with external accredited skills courses (such as health and safety, food hygiene and digital skills courses), and funding them via RTOF to remove the cost barrier, as being an effective approach in preparing participants for employment and enhancing opportunities. Projects also provided or connected participants to volunteering fairs and introduced participants to volunteering opportunities, with such signposting and support to access opportunities being likewise seen as important and effective by project staff.

From the provider perspective, offering practical, relationship-focused, holistic, and individualised support was seen as particularly important in supporting participants’ employability. Such approaches included going beyond traditional employability support by, for example, accompanying participants to interviews to reduce anxiety around travelling to the interview and being on-hand to offer support as needed. Caseworkers also described how they were often on hand for ad-hoc employability support as and when participants had questions, and building trust with participants to enable employment support workers from partner providers to manage participants’ expectations about intended activities. This was highly valued by participants.

Project representatives also cited the importance of focusing on targeted, employment-focused volunteering or work experience, rather than simply volunteering without an immediate link to employment. The latter was reported to be less attractive or viable for participants, partly as many had already volunteered when they were awaiting a decision on their asylum claim. In particular, identifying opportunities that matched individuals’ career aspirations was seen as key to engagement and as more likely to generate positive outcomes. One project described the importance, for example, of encouraging volunteering by linking volunteering opportunities with work trials or work experience, which could potentially lead to an interview or job. Relatedly, some projects encouraged participants into apprenticeships; however, it was noted that there was often little interest in this route due to the low-income provided, particularly when participants could instead enter low-skilled work such as warehouse operations for a much better hourly rate.

There was also widespread consensus that employer engagement, although resource-intensive, was critical to securing work experience and employment opportunities for participants. In terms of effective practice, provider staff also cited the importance of raising awareness of the benefits of employing refugees and people with lived experience of displacement amongst employers. To address this, staff talked to employers about what skills refugees can bring to their organisation, how to make recruitment more accessible and employment more inclusive, as well as answering questions and reassuring employers of refugees’ right to work. Project stakeholders also described substantial targeted outreach work to engage specific employers who could offer opportunities that matched participants’ skills, aspirations, and requirements for employment (for example, able to facilitate flexible arrangements for those with caring responsibilities or other commitments such as studying). Such options were noted to include hospitality, NHS health and social care (who could offer shift work) and retail employers.

Project representatives recounted several barriers to participants entering employment, work experience, or volunteering that matched their experience and work aspirations. Limited English language proficiency was the main barrier identified, which provider staff noted could lead participants into low-skilled, entry-level work. The need to support clients with English to be able to help them access meaningful employment, work experience and volunteering opportunities (this included referring clients to ESOL, providing on-the-job, sector-specific English language sessions, using employability sessions to also practice English, or using separate funding to offer additional English Language support to participants) was thus seen as key.

Reflecting this, participant interviews indicated that English language was the main barrier to employment, having a notable effect on their confidence, leading in some instances to perceived unfair treatment at work, and negatively impacting prospects for employment (often leading them to jobs which required less interaction in English). Some participants were highly qualified in their home countries (such as with PhDs or Master’s degrees), but having to do lower skilled work due to language barriers or lack of experience of the UK labour market and its requirements, including the lack of recognition of qualifications (in some cases even where qualifications were converted). Mental health issues were also clearly a barrier to finding work. Many RTOF participants reported low confidence and mental health problems (anxiety, depression, stress, trauma) impacting on their ability to apply for and get jobs, compounded by the employment process (including rejection) which negatively impacted their mental health further. RTOF caseworkers were reported to have been particularly helpful in this respect, building participants’ confidence alongside offering understanding and compassion.

Housing was also identified as a further key barrier to entering employment, as limited housing availability meant that participants could be housed in places distant from employment opportunities, while public transport was not necessarily reliable or accessible to them. Project representatives also noted that some participants who had entered employment faced a worse financial situation due to the impact of their working patterns on their universal credit allowance, resulting in them dropping out of employment. The increasing cost of supported housing once in work also deterred some participants from securing employment. Participants who were studying were also reported to be less likely to move into employment in some instances due to their learning commitments. However, delivery staff reported supporting participants to try to access flexible part-time work around their studies where possible, with this seen as effective in building up experience.

Housing activities

Project representatives suggested that the transition from housing provided for asylum seekers into secure housing was the main priority for most participants upon engagement. This resulted in high levels of engagement with housing activities across RTOF areas and particularly with specialist housing providers. Housing activities often involved discussing available housing options, registering clients on systems to find social housing, identifying suitable accommodation, bidding for social housing, attending viewings with clients, explaining terms of agreement, contracts, or their rights in an accessible way, and liaising with landlords. Effectiveness was linked to established partnerships with local VCSE housing providers, or housing providers forming partnerships, which helped to streamline processes and achieve housing outcomes more quickly compared to liaising with LA housing or homelessness teams.

Where appropriate, projects supported clients to access private rented accommodation. However, this option was only feasible where participants had secure employment. In these cases, staff across project areas reported that it had been essential to provide support to participants on an ongoing basis to help manage their tenancies. This support had included advocating for participants’ tenancy rights and translating or explaining documents (such as utility bills or contracts).

Across 2 project areas, stakeholders accessed wider pots of funding to help participants with the costs of housing (like the Household Support Fund). Some projects also purchased furniture, flooring, and other basic items with RTOF funding. Stakeholders described this as being important due to limited choice for housing, where participants may need to move into unfurnished properties. However, there were mixed views about the use of RTOF funding for household items; some delivery partners did not use the funding this way, instead encouraging participants to access other provision specialising in offering household items.

Additionally, staff reported managing clients’ expectations of their housing options, within a context of constrained choice. Having lived experience within the staff cohort was helpful for projects, as staff could explain their experience of finding accommodation following leaving housing provided for asylum seekers.

Across project areas, interviewees seldom discussed housing activities that had been trialled and found to be less successful, instead citing wider contextual challenges with the housing market being the main barrier to success. However, in one project area, stakeholders reflected on their operating model, with some identifying that their reliance on supported accommodation was helpful in the short term, but may act as a disincentive to work as rent would become unaffordable as soon as participants found employment. This suggests that additional housing support may be needed during refugees’ transitions to work.

More broadly, a lack of suitable, affordable housing was consistently reported across projects and all stakeholder groups during delivery as a key challenge. Stakeholders described how participants often did not meet the eligibility criteria for social housing (particularly where health needs were not recognised) or were faced with long waiting lists for supported accommodation. In general, the private rented sector was described as unaffordable for the majority of participants, with landlords requiring a secure income that covered the high costs of rent within a context of limited supply and high demand.

Some stakeholders felt that such housing challenges were exacerbated by Home Office policy changes shortening the notice period for moving out of asylum accommodation, alongside perceived increased pressure on local housing systems caused by making greater volumes of approvals. However, in the main, the unaffordability of accommodation was attributed to the wider economic context and housing policy changes.

While accessing safe and secure accommodation options proved challenging throughout RTOF delivery, there were concerns that the shortage of housing options meant that refugees were being placed in, or finding, accommodation that may be unsafe due to the prevalence of drug and alcohol abuse in proximity. This was reflected in interviews with participants who expressed worry at being housed close to drug use. Participants reported feeling unsafe in their area (the majority reported drug and alcohol use), having to phone the police, being subject to racism (when sharing accommodation with homeless people), and the area being inappropriate for children. In addition, many interviewees living in shared accommodation reported that they were reluctant to share a home and would prefer their own space. Relatedly, many interviewees reported that their accommodation was too small for them and their families and/or was in disrepair (with issues including damp and mould, leaks, broken appliances or problems with utilities such as gas or water).

Wider integration activities

To create RTOF Integration Plans, caseworkers and clients focussed on categories including (but not limited to) health and wellbeing, socialisation, confidence and self-esteem. Clients chose specific goals and associated activities aligned to these categories. Integration activities ranged from social activities to practical support, including registering for healthcare, attaining driving licences, alongside support with financial issues (like opening a bank account, applying for benefits, or budgeting advice).

Overall, wider integration activities were not considered to be as high a priority for participants as housing and employment. Projects highlighted that when participants entered employment, they could often disengage with wider integration support.

Accepting this general finding, effective activities for integration included: social activities with the dual purpose of engaging participants in the RTOF support and building trust with staff, as well as helping participants to get to know their local area and meet other members of the community. Provider staff reported that this had been successful, with participants being more socially connected, having spent time with others from a variety of backgrounds and members of the local community, understanding local transport systems through going out on trips, and visiting local libraries and other VCSEs in the area. It was reported during casefile reviews that such activities had a large influence on participants’ sense of confidence, fulfilment, and belonging, indicating the importance of this type of ‘wraparound support’.

Project stakeholders also generally found the personal integration plans to be a useful way to record participants’ individual integration goals and keep track of progress. However, they also found them to be limited in the information that could be recorded. Often, details that were discussed verbally were not included on the plans, although it is unclear if this is due to the limitations of the integration plans themselves, IT systems, or staff capacity or motivation to update them.

Project stakeholders and participants likewise appreciated the flexibility to create bespoke, person-led integration goals. Tailored integration plans commonly included goals around registering participants with dentists and GPs, setting up bank accounts and facilitating access to Universal Credit, accessing travel documents and provisional driving licences, supporting access to local schools, and activities to support wellbeing such as joining local sports clubs or getting a gym membership.

Project areas took different approaches to addressing the gap or delays in ESOL provision for service users. This included securing additional English language lessons for service-users through VCSEs, sometimes alongside other integration activities, funded through a separate small grants programme available in their LA. Another project area used ESOL-trained employment support workers to provide employer-specific language training on-site with partner employers, and 1:1 informal English conversations with service-users to build confidence in conversing. One project tested providing online English classes for clients on the ESOL waiting list or currently attending ESOL, but this was not deemed effective for many clients. Consequently, the provider concerned began offering face-to-face English conversation classes, though this was only delivered for a short period due to a lack of capacity.

Despite efforts to support English language development, long waiting lists and limited ESOL capacity could reportedly result in slow progress with English language development. Provision was sometimes one or 2 classes per week, which could make progress limited. Achieving positive outcomes in English, particularly to the standard required to secure employment, was therefore seen as a challenging within the RTOF timeframe, particularly given the low levels of English language for some of the cohort upon engagement.

Accessing NHS services such as GPs, dentists and mental health services were key integration goals for many participants. However, project stakeholders reported systemic barriers to accessing dentists, and mental health services in particular, with long waiting lists up to 4 years reported in one area. As noted elsewhere, poor mental health emerged a substantial need across the RTOF cohort, with project stakeholders identifying this as a key barrier to achieving wider outcomes. It was also noted that the prevalence of mental health issues amongst the cohort had placed strain on the delivery staff when seeking to support participants.

Despite coproduction featuring in some projects’ bids, there was limited evidence to understand whether this improved understanding of effective integration for different groups. Where evidence of coproduction was found, this was reported as part of wider organisational delivery (rather than being specific to RTOF delivery), and/or was apparent in some projects’ use of using informal feedback mechanisms to adapt activities.

Refugee experience of engagement

Participants generally felt positive about their experiences of RTOF projects. In particular, they appreciated the relationships with caseworkers and delivery staff, with trust and compassion commonly referenced in respect of these. It was clear that such positive relationships took time to develop, with some participants reporting being initially dubious about the ‘agenda’ of RTOF and use of their data. In this context, the perseverance of staff and their perceived genuine interest in participants’ situations and interests was acknowledged as vital for building trust.

Echoing findings cited above, it was also clear that, while the process of integration is not linear for any individual or cohort, participants themselves felt that employment and wider integration were highly contingent on accessing safe and secure housing and, in some cases, mental health support. Aligned with this, most participants valued activities relating to housing the most. Many participants interviewed likewise discussed the difficulties they experienced before RTOF and the effects these experiences had on their mental health. Securing safe and appropriate housing was, to them, inextricably linked to their mental health, wellbeing, and subsequent employability and integration.

However, finding safe and secure accommodation was clearly not easy for participants. They spoke of confusing eligibility for social housing, delays with paperwork, issues with repairs and poor communication (including one participant going for months without hot water). A number of participants engaged in the research were in a transient state regarding their housing situation, often awaiting better options or for family reunion to move again. However, analysing participants’ trajectories across interviews showed limited signs of improvement of housing circumstances over time, and signs of increased frustration with their situation. However, participants also generally expressed understanding that issues with the housing system were beyond the control of RTOF projects.

More generally, participants highlighted the flexibility, convenience, and consistency of RTOF support in particular. Many appreciated being introduced to activities and opportunities aligned with their interests. Caseworkers often offered support through regular meetings and ad-hoc contact (including phone calls or text messages when the client has a question/ odd check-ins). During follow up interviews, participants said they met with their caseworkers less often because they felt more settled (for example, bi-weekly instead of weekly), but were reassured by knowing that they could still call their caseworker whenever they needed.

Upon engaging with projects, many participants had low levels of English language and a lack of knowledge about systems and culture in the UK. Therefore, they appreciated having caseworkers to explain processes in accessible language, alongside providing more general integration support. Such activities could include accompanying participants on public transport for the first time, or explaining the ‘jargon’ related to housing, employment, or finances. Participants also typically enjoyed and engaged well with activities that facilitated a sense of normality and socialisation in particular. This could include enrichment focused activities, such as sports or film clubs, as well as communal meetings with food and conversational English practice. These activities were cited by participants as supporting them to make friends, improve their English, and get to know the local area better.

However, knowledge about the connectivity and cross-availability of different elements of the RTOF support offer varied. Some participants were not able to distinguish between different services (such as integration support or housing support) and were not able to attribute certain experiences to RTOF. In addition, it was apparent that some interviewees were receiving housing support elsewhere (from charities or the council) and were not aware that RTOF provided housing support. Others secured employment without the help of RTOF. Participants who had interacted with the council services, outside of the RTOF, sometimes felt that the council was pressuring them into accommodation which they did not like, not offering adequate support during the 28-day eviction process, and rushing them out of temporary accommodation. For this reason, projects noted that participants could be initially reluctant to accept RTOF support because they felt that professionals did not have their interests at heart. However, after RTOF caseworkers persevered offering support, clients were pleased with generally very pleased with the help offered as noted.

Participants also shared that they had experienced VCSE provision being at full capacity and felt like they might need more resources/staff to meet the high demand. Some had done volunteering, enjoyed it and got further opportunities from it, whereas others found it difficult to volunteer (unpaid) due to financial difficulties and having the time alongside studies, family responsibilities and applying for jobs. Several participants chose to focus on education as a path towards employment instead. Alongside long waiting lists, some participants’ caring responsibilities and distance between their accommodation and college were the key barriers preventing them from attending English classes.

Publication is licensed under the terms of the Open Government Licence v3.0 except where otherwise stated. View this licence on the National Archives

Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned.

Acknowledgements

Evaluation and learning activities would not be possible without the support and input of many people.

We are thankful to RTOF projects in Plymouth, Greater Manchester, North East England and the West Midlands for their ongoing support and contribution to the evaluation and learning activities. We are especially grateful to RTOF participants who took part and shared their experiences with us.

We are grateful to the evaluation and learning team: Ian Atkinson, James Ronicle, Helen Palmer, Matthew Cutmore, James Whitley, George Horton, Rachel Wooldridge, Maya Hill-Newell, Lilly Monk, Sarah Hamilton, Valeria Miglio, Martin Manuch, Kay Robinson, Angus Elsby, Sophia Stone, Samuel Greet, Johnny Lillis, Monica Andriescu, Mariya Yanishevskaya, David Crane and Jo Llewellyn, of Ecorys, and Jane Newman, Chris Clements, Jessica Davies, Tanyah Hameed, Pavel Rahman, Manon Desert, Sarah Baillie, Tom Rintoul, James Hearn, and Killian Troy-Donovan, of Social Finance.

We express gratitude to the members of the Refugee Steering Group (Ramatoulie Saidykhan, Omran Al Habbal, Nicholas Gawa, Moud Gouba and several others) for their advice and expertise and Helen Baillot, Oonagh O’Brien and Gianluca Palumbo for their facilitation. We thank Jenny Phillimore, Alison Strang and Sin Yi Cheung for their guidance and expertise throughout.

Finally, we thank colleagues at the Home Office, in particular Caroline Peate, Craig Barnes and Chiara Manzoni for their ongoing support and management, as well as cross-government partners in the Department for Work and Pensions and the Department for Culture, Media and Sport.

References

Atto, N., Hirst, C. and Hall, V. (2020) RESPOND Working Papers: Global Migration: Consequences and Responses Paper 2020/56 Integration Policies, Practices and Experiences.

CEBR, 2024. The economic impact of building social housing (p41). London: Centre for Economic and Business Research/Shelter/National Housing Federation. Available here. Comprising fiscal and economic benefits.

GOV.UK, 2023. Work and Health Programme statistics to February 2023. GOV.UK. Accessed 31 July 2024.

Institute for Social Value, 2025. Social Value UK. Accessed: 2024. Available here.

Mallender, J.; Tierney, R.; Baah, B.; Stavridou, M., 2017. Movement Into Employment: Return on Investment Tool, Estimation of benefits from moving an individual from unemployment into sustainable employment. London: Public Health England. Available here.

Kone, Z., Ruiz, I. and Vargas-Silva, C., 2019. Refugees and the UK labour market. COMPAS Report.

Ndofor-Tah, C., Strang, A., Phillimore, J., Morrice, L., Michael, L., Wood, P. and Simmons, J., 2019. Home Office indicators of integration framework 2019.

Royal Voluntary Services, 2025. Royal Voluntary Services. Accessed: 2024. Available here.