Research and analysis

STEP Ukraine programme: process, impact and economic evaluations – executive summary

Published 10 July 2026

This executive summary sets out key findings, lessons, and recommendations from an evaluation of the Specialist Training and Employment Programme (STEP) Ukraine Programme. The full technical report is available to download as a PDF.

Programme overview

The STEP Ukraine programme was developed by the Ministry of Housing, Communities and Local Government (MHCLG) and delivered by World Jewish Relief in partnership with the British Council. The pilot was developed in direct response to the displacement of Ukrainian nationals following the outbreak of the Russia-Ukraine war. The programme aimed to accelerate the integration of Ukrainian visa holders in the UK by enhancing English language proficiency and improving employment prospects. The intervention consisted of 12 weeks of tailored employment support combined with 10 weeks of intensive online English language training, all provided free of charge.

The programme initially ran from August 2023 to August 2024. The first stage of the programme engaged 17,282 participants, of whom 9,813 were onboarded. Following a contract extension, delivery was continued from August 2024 to March 2025. In total, the programme engaged 45,512 participants, with 14,850 signing up for the programme. This evaluation focuses on the initial programme from August 2023 to August 2024.

Evaluation objectives

Alma Economics and RAND Europe were commissioned by MHCLG to conduct an independent evaluation of STEP Ukraine to assess its processes, impact on language skills and employment outcomes, and value for money.

The evaluation aimed to:

• assess the design, delivery, and experience of the STEP Ukraine programme
• quantitatively estimate the impact of the programme on English skills
• quantitatively estimate the impact of the programme on employment outcomes
• quantitatively estimate the programme’s value for money

To achieve these objectives, the evaluation employed a mixed-methods approach using both qualitative and quantitative methods.

Qualitative methods aimed to gather insights into the programme’s implementation and participant experiences by conducting in-depth interviews and focus groups with Ukrainian guests, and interviews with programme managers, implementers and stakeholders.

Quantitative methods measured the programme’s impact on language skills by analysing pre- and post-programme English proficiency test scores, compared employment outcomes with a counterfactual group of Ukrainian guests who did not participate in STEP Ukraine, and determined the programmes value for money by comparing programme costs to the estimated value of the benefits that it generated for society and public finances.

Key findings

Process evaluation

The programme aimed to address both immediate integration needs and longer-term career aspirations through a number of routes. The process evaluation found the dual-focus model was ambitious and well-conceived. In practice, however, the language and employment support strands often operated in silos, with a lack of communication between English tutors and employment advisers. In turn, this undermined the intended synergy of the model. Operational agility was demonstrated when the programme was launched, as within 6 weeks of the contract being awarded, the programme was operational. However, the speed of the launch may also have resulted in some of the practical difficulties subsequently experienced, including high caseloads (up to 35 clients per adviser), rigid scheduling and timetabling of English provision, and the concurrent delivery of employment support.

Personalised employment support was seen as one of the strongest aspects of the STEP Ukraine programme. Participants consistently valued the culturally sensitive, one-to-one guidance from Ukrainian-speaking advisers, however, STEP Ukraine staff felt that more training could be provided on mental health concerns. English language training appeared to boost the confidence of participants, however, there were some equity concerns related to the starting English level of participants. Findings indicate that learners who started with a lower level of English were particularly disadvantaged, and that the model likely worked best for those who already had a baseline level of English language on entry to the programme.

Barriers to participation and completion of the STEP Ukraine programme were largely related to the flexibility of the programme. Barriers cited include concurrent timetabling of support, which hindered completion, digital access, which limited the ability to participate, and caring/work responsibilities, which prevented attending sessions during the daytime. This highlights a tension between the programme structure and participants’ real-life constraints to participate.

English language impact evaluation

On average, participants in the STEP programme showed significant improvement in their English test scores between programme entry and exit, even when controlling for time spent in the UK. The increase in average English test scores between pre- and post-participation ranged from 20.8% to 29.2%, depending on the assessment measure used. Some increase in test scores would be expected because English tends to improve with time spent in the UK. However, STEP participants saw an additional increase in English test scores, over and above what would be expected based on time in the UK alone, of between 17.6% and 25.9%. These estimated effects are statistically significant, meaning they are unlikely to have been observed by chance.

Employment impact evaluation

At the UK level, STEP Ukraine had a positive effect on employment outcomes, with participants on average 6.6 percentage points more likely to be in employment than if they had not participated, over the 20 months following programme entry. These impacts emerged shortly after participation and grew steadily up to approximately 7 months after treatment. Positive effects of STEP Ukraine were also observed on the probability of sustained employment. Participation increased the likelihood of being in employment for at least 3 consecutive months by 7.7 percentage points, and for at least 6 consecutive months by 6.5 percentage points. It was also estimated to have a positive effect on the monthly earnings of participants who were already employed prior to the programme, raising their monthly earnings by around 10.7% after 20 months.

Impacts were also examined at the level of devolved administrations in Scotland, Wales, and Northern Ireland. The estimated impacts of STEP Ukraine on employment in Scotland were broadly similar to those observed in the UK as a whole. For example, the likelihood of being employed among STEP Ukraine participants in Scotland increased by 6.0 percentage points over the 20 months following programme entry, compared to if they had not participated. It was unfortunately not possible to reach robust conclusions for Northern Ireland and Wales due to lower sample sizes and more difficulty establishing robust control groups of non-participants (see Section 4.3 and Appendix 4 of the full technical report).

Economic evaluation

The cost-benefit analysis of STEP Ukraine demonstrates that the programme delivered value for money, both from the perspective of the public finances and for society overall. The model compared the benefits of STEP Ukraine’s impact on employment, such as increased tax revenues and reduced Universal Credit payments, against costs, including programme implementation costs and increased travel expenses.

The central scenario suggested that for every £1 invested in the programme, it generated gross savings of £2.04 (£1.04 net) for the public finances, via higher tax receipts and lower benefit payments. The programme generated £3.67 in gross benefits (£2.67 net) for society overall, primarily from the value of increased economic output due to higher employment and earnings. Sensitivity analysis explored a range of plausible values around key assumptions required for the cost-benefit analysis. These result in benefit-cost ratios varying from 0.88 to 3.27 from a public finance perspective, and 1.59 to 5.84 from a social perspective.

Limitations

Process evaluation

Time constraints and limited availability of programme partners restricted several planned data collection activities, including non-participating observations of programme activities and interviews with key English language training stakeholders. Additionally, partner facilitation of the survey questionnaire to delivery partners limited the evaluation team’s oversight over its distribution. Coupled with low response rates, the findings do not reflect a representative picture of all STEP Ukraine implementers.

Language impact evaluation

The counterfactual analysis relies on the assumption that earlier arrivals did not differ from later arrivals systematically on unobserved characteristics that affect the relationship between time spent in the UK and English language ability (for example, unobserved ability or selective migration timing). To estimate the impact of STEP Ukraine on English language ability, expected improvements that would have occurred irrespective of the programme due to increased time spent in the UK were modelled. This modelling assumed a linear relationship between time spent in the UK and English language ability. In addition, around 20% of participants exited the programme before completing an exit test, meaning the results may be subject to attrition bias; for example, if participants were more likely to drop out because the programme was not improving their English skills, the estimated positive impact of the programme would be overstated.

Employment impact evaluation

As the analysis window only spanned 20 months after the programme start, it was not possible to examine the longer-term impacts of STEP Ukraine. Difference-in-differences relies on the assumption of parallel trends in outcomes between participants and non-participants. Statistical testing demonstrated that the parallel trends assumption was, on average, satisfied during the months prior to the intervention. However, a ‘stricter’ version of the test showed that the parallel trends assumption did not hold in every individual month, as would be expected from a true randomised experiment. Unfortunately, due to the much smaller sample sizes, it was not possible to attain robust results for the impact of STEP Ukraine on employment outcomes in Wales and Northern Ireland.

Economic Evaluation

Some programme impacts could not be directly estimated from the impact evaluation, requiring assumptions (for example, impact persistence). Where possible, standard assumptions from the Department for Work and Pensions guidance were applied. Where uncertainty remained, informed assumptions were made and tested through sensitivity analysis. The model used captures the main expected channels of benefits of the programme (employment, wages, and output), but may not account for all potential impact channels, such as improvements in health or wellbeing. Similarly, the modelling reflects impacts on Universal Credit and taxation but excludes devolved benefits and non-Universal Credit payments. While this omits some benefit types, Universal Credit is the primary income-related benefit in the UK and across all devolved administrations and captures a substantial share of employment-related fiscal effects.

Policy recommendations

The process evaluation identified a number of recommendations that could enhance the accessibility, effectiveness, and sustainability of future programmes. While many of these changes may involve additional costs that need to be weighed against their potential benefits, they are worthy of consideration to support improved outcomes for participants.

Design for diversity and flexibility

To strengthen future programmes, thorough eligibility and suitability assessments should be implemented during participant registration. Clear, upfront communication about programme intensity, its dual focus on English language development and employment support, and how it differs from direct job placement services would help set realistic expectations and reduce early dropouts.

Course intensity and support tracks should be tailored to account for varying English proficiency, employment experience, motivation to learn, and personal circumstances. A ‘2-speed’ model could offer both a standard, full-intensity track and a lighter, more flexible option with fewer classes or reduced weekly contact hours. Evening provision could further support participants with caring responsibilities or work commitments, enabling them to engage at a pace that suits their circumstances. Additionally, introducing more entry points based on prior English language ability would allow tutors to deliver more tailored support, particularly for participants with lower baseline proficiency.

Operational efficiencies and staff capacity

The introduction of a centralised and intuitive customer relationship management system would simplify data collection, minimise duplication of effort, and enhance real-time monitoring of participant outcomes. Such a system would support effective programme management and allow for timely adjustments informed by operational insights.

Stricter suitability checks, combined with clearer communication of the intensity and commitment required for both English and employment components would strengthen onboarding and help manage expectations. Priority should be given to recruiting staff with backgrounds in refugee support, employment services, or language education, alongside comprehensive onboarding and role-specific training for all new staff, including translated resources where needed.

To ensure high-quality, personalised support, advisers’ caseloads should be limited to around 25 to 30 clients at any one time with planned breaks between programme cohorts to allow for reflection, strategic planning, and completion of administrative tasks.

Integrating individual guidance with group sessions could enhance efficiency and engagement, particularly for foundational employment topics. Offering sessions at varied times and where possible, in multiple languages would further support accessibility and consistent participation.

English language learning

English language delivery would benefit from stronger integration and communication between English tutors and employment advisers. This could be achieved through improved direct contact, structured feedback loops, streamlined attendance monitoring, and better coordination between language and employment support schedules. Staff should also be supported through trauma‑informed and context‑specific training, alongside access to Ukrainian language support where appropriate, to foster inclusive learning environments and reduce isolation in online provision.

Developing dedicated learning pathways for lower-level English learners (A0-A1 on CEFR scale) would further enhance outcomes. These pathways could include more focused grammar support, structured exercises and discussion, and tailored feedback. Separating beginner groups by level would help ensure teaching methods are well-matched to learners’ needs and support engagement. The expectations and structure of a full-immersion approach should be clearly communicated during the application process to allow participants to make an informed decision.

For peer-led sessions, explicit guidance on participation, collaborative learning, and group dynamics should be provided. Assigning staff or teaching assistants to actively monitor these sessions would allow for real-time feedback and correction of persistent language errors. Offering less frequent classes over a longer period, along with mid‑course breaks, could improve flexibility for participants with competing responsibilities.

A broader mix of instructional techniques could also be adopted, such as situational role-plays and practical language exercises linked to participants’ everyday lives and employment goals. To maximise the effectiveness of breakout groups, expectations around turn-taking and participation should be clearly set, with active facilitation to ensure inclusive discussion.

Employment support

The aims of the employment support offer should be clearly communicated to distinguish it from direct job placement services. Emphasis should be placed on empowering participants with the skills and knowledge needed to independently navigate the UK job market, including developing self-directed job search strategies, tailoring applications, and building effective interview techniques.

Advisers should work collaboratively with participants to set realistic, motivating short-, medium-, and long-term goals, taking into account their previous qualifications, work experience, and local labour market conditions. Support should remain flexible and tailored to participants’ ambitions and the requirements of their chosen sectors.

Opportunities for professional networking could be strengthened through organising sector-focused events, workshops, mentorship programmes, and forming partnerships with local employers and community groups. These initiatives, including internships and apprenticeships could support labour market integration and career advancement.

Leveraging community and partnerships

The use of Ukrainian‑speaking support staff and advisers was a key strength of STEP Ukraine and should remain a central feature of future provision. Expanding partnerships with employers and intermediary organisations could provide access to practical employment opportunities, including entry‑level roles, internships, and apprenticeships, particularly where aligned with participants’ prior skills and experience. Community‑building could be further encouraged through group activities, interactive online sessions, and structured peer support mechanisms. Strategic use of word-of-mouth within the Ukrainian community was particularly effective for recruitment and can continue to drive engagement and participation.

Monitoring and evaluation

Future monitoring frameworks should consider capturing a broader range of non-employment outcomes, such as confidence, motivation, hope, and understanding of UK workplace norms, which are key drivers of long-term success. Regular attendance proved to be a strong indicator of improved language and employment results; therefore, it should be prioritised as a core performance metric, alongside proactive support to sustain engagement.

Simplifying post-programme surveys and improving user-friendliness may increase response rates and provide deeper insights into long-term outcomes. Early targets for STEP Ukraine were often considered overly ambitious, especially regarding anticipated improvements in English proficiency. This sometimes created pressure to meet numerical goals at the expense of genuine progress. Introducing flexibility into KPI frameworks and allowing for adjustments over time would better support effective and responsive programme delivery.

Holistic support

Some participants reported feelings of isolation after completing the programme. Establishing ongoing support mechanisms, such as alumni networks, regular drop-in sessions, or continued access to advisers, could extend programme benefits and support sustained progress in employment and language development. Integrating specialist trauma-informed and culturally sensitive mental health support, co-designed with refugee communities and relevant experts could further improve participant wellbeing. This approach would help recognise the impacts of displacement-related trauma, embed trauma‑informed practice across delivery, and minimise the risk of re‑traumatisation.

Childcare responsibilities, especially among single mothers, presented a significant barrier to access and sustained engagement. Embedding childcare support within programme design, or establishing strong referral pathways to local childcare services, could help alleviate this challenge. While housing support falls outside the direct scope of employment and language provision, partnerships with housing organisations or the provision of basic housing advice may further support stability and participation.

STEP Ukraine relied significantly on participants’ digital skills. To support full engagement, future programmes should provide comprehensive digital literacy training that offers introductory sessions for online platforms like Zoom and learning management systems.  Strengthening digital skills would enable participants to fully access programme activities and maximise learning opportunities.

Conclusion

The STEP Ukraine programme demonstrates that rapid, integrated support combining intensive English language learning with personalised employment guidance can produce meaningful and measurable benefits for participants. The findings highlight the importance of flexible, culturally attuned, and responsive programme design to meet the diverse needs of displaced populations. As refugee and forced migration flows continue, the lessons from STEP Ukraine offer valuable insight for policymakers and practitioners seeking to design impactful and scalable integration programmes.

The evidence indicates that STEP Ukraine succeeded in improving participants English language ability beyond what would have occurred naturally through independent progression. The English language component should therefore be considered a successful model for informing future programmes with similar objectives. Positive impacts were also observed across employment outcomes, including employment probability, employment sustainability, and monthly earnings.

Participation in STEP Ukraine increased the likelihood of being in employment, with impacts emerging shortly after programme completion and growing steadily over the first 7 months before stabilising. This trajectory is consistent with the programme design, as participants received a 12‑month action plan upon completion and retained access to the online portal for 6 months. It is therefore plausible that participants continued to implement their action plans over several months, with the full employment effects materialising gradually. For monthly earnings, the estimated impacts continued to increase over the observation period, suggesting that the evaluation may not capture the full longer-term impact on earnings. This likely reflects the time it takes time for improved language proficiency and employability skills to translate into higher-paid and more stable employment.

Finally, the cost-benefit analysis showed that the net benefits of STEP Ukraine outweighed the costs from both the Exchequer and social perspectives, indicating that the programme represented an effective use of public funds. Taken together, the findings suggest that STEP Ukraine offers a robust and evidence‑based approach to supporting refugee integration through combined language and employment support.