Migration, vacancies and labour market skills

by Ben Brindle & Madeleine Sumption

The current government aims to link immigration and skills policy more closely. This sounds straightforward, but faces several challenges. This briefing examines the link between skills and migration policy—the first in a series examining immigration in middle-skilled and technician roles. Among the key points:

• Government policy to make employers recruit fewer people on work visas includes restricting work visas so that middle-skilled jobs are only eligible if they are crucial to the industrial strategy sectors or critical infrastructure, and requiring employers in those sectors to create workforce plans to develop domestic skills.
• Work visa policy only affects a small share of labour supply, however. Most migrants have work authorisation but are not main applicants on work visas, instead arriving on family or dependant visas, or as students or refugees. Work visas post-Brexit drove much of the increase in the migrant workforce in some industries, such as health and care, but had a much smaller impact in others, such as construction.
• Increasing the domestically trained workforce does not automatically lead to lower migration. Skill shortages are just one factor that contributes to employers’ demand for migrant workers—for example, local workers can be difficult to attract into certain roles because of poor pay and conditions. Recruitment of migrants is also driven by migrants’ choices to come to the UK, regardless of any shortage.
• Migration restrictions might push some employers to train local workers, but the limited evidence on this point is mixed.
• It is difficult to create effective incentives in the immigration system that explicitly encourage employers to invest in domestic training. A sectoral approach could encourage ‘free riders’, while an individual approach is difficult to monitor.
• Significant data gaps prevent a comprehensive overview of how migration is affecting the UK’s skills base.

What is government policy on the link between immigration and skills?

The government announced in May 2025 that it wanted to make the UK labour market less reliant on the immigration system and encourage employers to train UK workers instead. It promised “much stronger links” between UK training and the immigration system. The specific changes proposed were:

• Restricting work visas in low- and middle-skilled jobs. Graduate jobs continue to have access to long-term work visas. However, the government ended overseas recruitment of care workers in 2025 and ended visa access for middle-skilled jobs not related to the Industrial Strategy, such as butchers and chefs.
• Middle-skilled jobs linked to the Industrial Strategy or the building of critical infrastructure will be eligible for visas if they are on the Temporary Shortage List (TSL). At the time of writing, these were primarily technical roles, such as engineering technicians and skilled construction trades, plus some creative industry positions, such as dancers and fashion designers. It is currently unclear whether ‘temporary’ refers to a role’s inclusion on the TSL, the visa granted to the work migrant, or both.
• From mid-2026, representatives of employers in industrial strategy sectors need to have an approved workforce plan in place if they want their occupations to remain eligible. Workforce plans will be assessed by the new Labour Market Evidence Group, which brings together various government departments and advisory bodies.
• Employers will pay more to sponsor skilled workers: the Immigration Skills Charge increased by 32%, to £1,320 per worker per year for large employers and £480 for smaller employers. This is in line with inflation since the charge was introduced in 2017.

The next sections discuss the data and evidence currently available to evaluate these plans and their potential impacts.

How much of UK migration is affected by work visa policy?

Policy debates often focus on who is eligible for work visas, but main applicants on work visas make up only a small share of overall migration. Of the 3.45 million non-EU citizens who received visas since Brexit and still had a valid immigration status at the end of 2024, only 17% were main applicants on work visas, according to Home Office data (Figure 1). This low share partly reflects large numbers of international students, many of whom will eventually leave. However, many work visa holders leave too, while family members and refugees tend to remain long-term. As a result, the picture does not greatly change even if we examine people who receive settlement (33% of settlement grants in 2024 were to people in the work category when they applied, and this includes partners and children of main applicants).

Most of the non-EU migrants with other statuses have work authorisation (see ‘Understanding the Policy, above), and many appear to have taken up work. HMRC payroll data shows that the number of jobs held by non-EU employees rose by 1.9 million, from 2.1 million in January 2021 to 4 million in December 2024, far more than the number of work visas granted to non-EU citizens over the same period.

The high share of migrants who are not main applicants on work visas is common internationally. Across high-income OECD countries, for example, family migration is much larger than work migration, and work routes make up a minority of all arrivals.

Figure 1

These figures cannot tell us precisely what proportion of working-age post-Brexit migrants were employed, because some of the net increase in non-EU employment will result from people who arrived in previous periods, and some valid visa holders will no longer be in the country. However, as a back-of-the-envelope calculation, if we assume that 82% of valid visa holders were adults (in line with 2024 Home Office data), the increase in non-EU payrolled employment was equivalent to around 67% of adult non-EU citizens granted visas since 2021. This was below the UK-wide employment rate of around 75%, which may be because the visa holder group contains large numbers of international students, who can work part-time but often do not. Higher employment rates for work visa holders will have been offset by lower rates for international students, refugees and family members of British citizens.

Which industries and occupations are most affected by long-term work visas?

Because most migrants are not main applicants on work visas, migration affects the supply of labour and skills even in lower-paying roles that tend not to be eligible for work visas.

Labour Force Survey data on migrants who arrived in the UK since 2000 show that people who say they originally arrived for work or study were more likely to be in graduate jobs than UK-born workers, particularly among non-EU arrivals. People who said they arrived for other reasons were significantly more likely to be in jobs classified as low-skilled (Figure 2). Across all groups, migrants were less likely to be in middle-skilled roles than the UK-born.

Within middle-skill occupations, migrant workers were slightly more strongly represented in occupations that the government deemed crucial to the Industrial Strategy or building critical infrastructure (TSL jobs as of 2025), where they made up 14% of the workforce, more than their share in middle-skill jobs not on the TSL (10%).

Figure 2

Long-term work visas post-Brexit played a small role in job recruitment for the now-prioritised TSL roles. An estimated 29,100 visas were granted in these roles between the 2022/23 and 2024/25 financial years, equivalent to 1% of the number of new hires in these occupations between 2022 and 2024 (3.1 million). By way of comparison, 440,000 Skilled Worker visas were granted between 2022 and 2024, equivalent to 4.8% of new hires in eligible jobs over the same period (9.1 million). See Appendix 1 for more information.

Figure 3 shows the change in the number of jobs held by adult migrant employees in each industry between 2021 and 2024, compared to Skilled Worker visa grants over the same period (represented by the purple dots). Note that the figures refer to all jobs, not only those eligible for work visas.

Across all industries, just over 500,000 Skilled Worker visas were granted, equivalent to 26% of the net growth in the migrant workforce (1.9 million). This varies significantly across industries, however. Skilled Worker visa grants were highest relative to the net growth of the migrant workforce in finance (64%), IT (60%) and health and care (53%). In other sectors, such as retail (7%) and construction (28%), long-term skilled visa grants were much lower, and most of the growth in the migrant employee workforce will have come from people on other visa types.

Figure 3

See Appendix 2 for more information about the jobs held by work migrants’ partners and people on family visas. These groups often had lower median earnings than would be required for long-term work visas.

Will increasing UK training reduce migration?

Surveys asking employers why they hire migrant workers often highlight a lack of domestic applicants with the necessary technical skills as a key reason. A lack of trained applicants can result from several factors, such as underinvestment in training, poor pay and conditions, long training pipelines that make it hard for industry to respond quickly, and temporary spikes in demand (e.g. in construction or NHS recruitment).

In some cases, employers respond to perceived shortages by actively recruiting abroad: creating job adverts in languages other than English, using overseas recruiters, or asking existing migrant workers to share job adverts with their networks. A qualitative Migration Advisory Committee (MAC) review in 2022 found that this was particularly likely when employers wanted to hire many staff at once (e.g. for a new construction project).

Skill shortages and shifts in employer demand for workers are not the only drivers of skilled migration, however—most employers who responded to the MAC said they hired migrants through traditional recruitment processes and simply found that migrants made up a high share of job applicants. In other words, if there are more migrants in the candidate pool, more are likely to be recruited regardless of how many domestically trained workers are available. This may be one reason why the MAC also found no correlation between the intensity of skill shortages in an occupation and the number of Skilled Worker visas granted in the same occupation between 2021 and 2023 (Figure 4).

Figure 4

More broadly, there is no fixed number of skilled workers UK employers need, especially in the private sector. If the number of people with technical skills increases due to migration or successful training initiatives, UK employers are likely to increase the amount of goods and services they produce and the number of people they wish to employ. As a result, training more UK workers will not automatically lead to lower migration.

Will reducing migration increase training?

Employers’ investment in training has decreased, according to the Employer Skills Survey. In 2024, UK employers reported investing an estimated £50 billion, compared to £65 billion in 2011 (adjusted for inflation). As a result, investment in training per employee fell from an estimated £2,410 to £1,700 over this period—driven primarily by employers spending less on each person trained. Many factors affect how much employers invest in training, such as subsidy regimes, cost pressures, and the extent to which jobs require generic skills (e.g., numeracy and literacy). In theory, high immigration may also affect employers’ incentive to train workers, if it means that they can recruit ‘job-ready’ candidates more easily.

In practice, evidence on how much immigration affects training is limited and has produced mixed results. One study looked at vocational training, such as NVQs and apprenticeships, and found that migration reduced UK workers’ participation in less costly and less economically beneficial types of training. Another found that more immigration was associated with higher training rates of UK-born adults in some sectors and lower in others. In particular, training fell in industries where output was not exported, e.g., construction or hospitality, but increased elsewhere. A third study found that skilled migration might increase workplace training, although the results were inconsistent across different methods.

Employers in some jobs have alternatives to recruiting and training local workers when labour is in short supply. In particular, numerous studies indicate that automation increases when lower migration makes workers scarcer, although automation is not feasible in all occupations.

In other words, it is possible that a decline in migration—whether work-related or not—could push some employers to increase domestic training, but evidence does not provide a clear guide on whether and to what extent this benefit might materialise.

Can immigration policy incentivise employers to increase training?

Regardless of how strong the link between migration trends and training patterns is, the government might try to strengthen it by requiring employers who use the immigration system to do more to train domestic workers. There are different ways of doing this and they face different challenges.

Current government policy aims to create a direct link between sponsoring overseas workers and sector-level training for middle-skilled Industrial Strategy and infrastructure jobs (but not for graduate jobs). From 2026, employers sponsoring workers in these technician roles must be in an industry that has an approved workforce plan. The workforce plan must involve credible strategies to boost the domestic workforce. Failure to carry out the workforce plan could mean that occupations become ineligible for visas in future.

The challenge with this approach is that just because a sector has a workforce plan, it does not mean that all—or even most—employers are following it. Few employer representative bodies can genuinely claim to represent the majority of employers in their sector. At the same time, employers who do a lot to train local workers may still see their occupations lose access to visas if other employers fall short. This creates a risk of ‘free riding’.

An alternative approach is to assess employers’ efforts individually. Until 2021, the UK required employers to show that they had attempted to recruit domestically through the “Resident Labour Market Test” (RLMT). This did not scrutinise training plans, although it is in theory possible to do so (Canada’s labour market test does). This approach may create a greater incentive at the employer level, but it puts greater demands on the government to understand what employers are doing. Home Office caseworkers do not have a detailed understanding of the skills required by each firm, and may find it hard to assess how thoroughly an employer has searched the pool of resident workers. As a result, RLMTs may become a box-ticking exercise for sponsors. Research from Sweden and Finland found that RLMTs there had slightly reduced immigration and increased wages in some occupations, but there is little evidence on whether RLMTs increase employer training.

Previous governments have introduced different measures aiming to make domestic training more financially attractive to employers than overseas recruitment. For example, the Apprenticeship Levy encourages large employers to invest in training by requiring them to pay a compulsory charge for apprenticeship training, with funds lost if not used. At the same time, the cost of recruiting migrants on work visas has increased due to the Immigration Skills Charge (ISC), which came into effect in 2017. In theory, ISC revenues were supposed to boost skills and training budgets, though it is unclear whether all the funds reach this destination.

There is remarkably little evidence on the impacts of these programmes and fees on employers’ decision-making. It is possible that they have only had a limited impact on employers’ demand for migrant workers, because these additional costs only represent a small share of the overall cost of hiring high-skilled workers (which includes salaries and recruitment costs)—although they are higher if employers pay other fees on migrants’ behalf, such as application and NHS fees. Studies which examine how employers’ demand for workers changes when real wages rise suggest that demand for high-skill workers is less responsive than demand for low-skill workers.

Does the UK require migrant workers to meet its housebuilding goals?

The government aims to boost residential construction and increase the net number of homes in England by 1.5 million by 2029—equivalent to 300,000 per year on average. This implies a substantial increase in construction activity, since only 208,000 net new dwellings were added in the year ending March 2025. Industry bodies have said the target can only be achieved if the construction workforce is increased (although there is disagreement on the precise increase required).

Migration contributes to housing demand and an increase in the cost of housing, as the Migration Observatory has outlined elsewhere. Migrants also contribute to the construction workforce, although they are not overrepresented. In 2024, an estimated 15% (325,000) of the 2.2 million people working in construction were non-UK born, according to the Annual Population Survey (APS), which is lower than the migrants’ share in the overall workforce (20%). This share is increasing, however: HMRC payroll data show that non-UK nationals accounted for 52% of the increase in the construction employee workforce between January 2021 and December 2024 (31,000 of 60,000 new employees).

Nonetheless, because 85% of construction workers were born in the UK, training and retention are likely to be more important than migration in the medium- to long-term. Analysts have identified various problems in the skills pipeline, such as the loss of skills during downturns and limited investment by major housebuilders.

Overseas recruitment might have a faster impact than domestic training in the short run, although the post-Brexit immigration system is not well-suited to the construction sector. This is because over one-third of construction workers are self-employed according to the 2024 APS, and skilled worker visa holders must be employees (unless the government changes the programme to allow umbrella organisations to handle sponsorship). Construction is also a high-risk industry for exploitation, and employer-sponsored visas can exacerbate this risk further.

Is there enough evidence to understand the links between migration and skills?

Significant data gaps make it harder to examine how migration affects skills, and vice versa. Data on main applicants on work visas is rich and detailed, but there is limited reliable data on the occupations or skills of the majority of migrants newly entering the labour market, i.e., those on other visa types. Newly available data from HMRC tax records linked to Home Office visa records provide an excellent overview of workers’ industries, but not their occupations. These data do not cover those who are not employees, which leaves a large gap in industries like construction.

In the past, it might have been possible to use rich data from the Labour Force Survey to fill this gap, but the response rate to the survey fell dramatically during the pandemic. This makes the figures less reliable, particularly if we want to assess changes over time. The LFS is still a good way to get a sense of how large groups compare (such as work and non-work migrants), but fine-grained analysis of specific job and skill types is not possible. There is also limited data on the characteristics of employers who hire migrant workers, and how they differ from employers who do not (such as how much they invest in their workforce).

There has also been a particular problem with the coverage of recently arrived migrants in the LFS, where response rates appear to be particularly low. This has left a large gap in the understanding of what skills new migrants have brought under the post-Brexit immigration system, and how and whether this has changed since the previous system. HMRC data suggest surprisingly modest changes in the initial earnings of newly arriving cohorts of migrants, but these data are only a crude proxy for skill (especially since we do not know how many hours people worked, and in what occupation). We know that after the end of free movement from EU countries, non-EU migration replaced EU migration as the main source of growth in the migrant workforce, but it is harder to ascertain how comparable these two groups of workers’ skills are. Given the enormous change to the scale and composition of migration after Brexit, this leaves a significant gap in the understanding.

Finally, data on workforce training are also limited. Analysis primarily relies on the Employer Skills Survey (ESS), a survey of employers conducted every two years. While the ESS contains information on the proportion of employees trained and types of training provided, it does not disaggregate training by employees’ seniority, occupation, or skills. As a result, training patterns cannot be linked to specific employees, making it difficult to understand how employers respond to changes in the workforce, whether through migration, education policy, or other factors.

Appendix 1

The Migration Advisory Committee (MAC) estimates that 29,119 work visas were granted to the 82 occupations eligible for the Temporary Shortage List (TSL) following their stage 1 report. These estimates are uncertain because the TSL was created in July 2025, after the Home Office began using an updated occupation classification, SOC 2020, in April 2024. To estimate visa grants to jobs eligible for the TSL before April 2024, the MAC had to convert SOC 2020 job codes to older SOC 2010 codes, which do not always perfectly align. In addition, visa grants by occupation are not available from April to September 2024; therefore, the MAC uses data for April to June 2025.

To calculate the share of new hires recruited on work visas, we divide this figure by the number of people who started a job eligible for the TSL between 2022 and 2024 (3.05 million), using figures from the 2024 Annual Population Survey.

Appendix 2

Almost half of health and care workers’ partners who were employed in the year ending March 2024 worked in healthcare, while a quarter of people on family visas worked in administrative services (Figure A1). Note that the Home Office was unable to match a small number of visa holders to HMRC payroll data, meaning the actual number working in each industry may be higher.

Median annualised earnings were £22,100 for partners or health and care workers, for example, and £20,200 for family members of British citizens—lower than the £28,300 for the overall employee workforce. The exception was the partners of skilled work visa holders (excluding health and care), whose earnings were slightly higher than the median (£30,200). Select ‘by visa’ in Figure A1 to view earnings for partner and dependant visa holders by industry.

Figure A1

Acknowledgements

This report was produced with the support of the Gatsby Education.

Gatsby Education’s mission is to champion and support the delivery of a world-class education and skills system – one that drives opportunity, boosts productivity, and supports a stronger, future-facing economy. We work in partnership with government, education providers, employers, and others, to support and deliver innovative projects, build lasting collaborations, and shape national education and skills policy. Our focus is on long-term, systemic change – ensuring impact that endures well beyond our direct involvement. www.gatsby.org.uk.

Thanks to Daniel Sandford Smith, Jonathan Wadsworth and Chris Wright for comments on an earlier draft. Any errors are the responsibility of the authors.

This publication uses data from the Office for National Statistics, Social Survey Division. (2025). Annual Population Survey, 2004-2024: Secure Access. [data collection]. 33rd Edition. UK Data Service. SN: 6721, DOI: http://doi.org/10.5255/UKDA-SN-6721-32.