Why the UK Must Put Labour Supply At the Heart of its Visa Access Policy

The UK’s Migration Advisory Committee (MAC) has published its recommendations on which occupations should be eligible for Skilled Worker visas. They’ve done a good job under challenging conditions: ensuring strategically important labour supply while also cutting immigration to incentivise domestic training. But much more needs to be done to ensure coherence between industrial policy, skills policy, and labour migration policy. In this blog we assess the MAC’s recommendations and suggest ways to improve their process.

Selecting the chosen occupations

In May 2025, the UK unveiled a new approach to labour migration policy, tying it tightly to its new Industrial Strategy. Occupations below level 6 of the Regulated Qualifications Framework (RQF)—in short, below degree level—would no longer be eligible to recruit using Skilled Worker visas.

In July 2025, the Home Office asked the MAC to draw up a Temporary Shortage List (TSL), giving time-limited visa access for strategically important occupations facing pressing shortages. Occupations could be included only if the relevant government department could demonstrate there was a genuine shortage, and submitted a Jobs Plan setting out targeted actions to increase the domestic supply of workers.

Occupations were considered for the TSL based on feedback from government sector experts and rates of employment in key sectors. In total, Jobs Plans were submitted for just 46 of the 82 occupations under review. Eight of those without Jobs Plans were occupations that government stakeholders had previously advocated should be on the TSL.

The MAC interpreted non-submission of a Jobs Plan as “a conscious decision by the relevant government department(s)” to deprioritise an occupation due to a recognition of low need. But it could just be down to a lack of administrative capacity. In Table 1 we show the proportion of proposed occupations within a sector for which a Jobs Plan was ultimately submitted, and the departments responsible.

Table 1

Few Jobs Plans compared projected future demand with likely supply of suitably skilled workers. Where this was done, they primarily considered new apprenticeship graduates; very few considered alternative adjustment margins (such as efforts to attract or retain more existing workers), and none discussed whether training programmes could actually scale to meet needs.

Importantly, the MAC “was not asked to consider historic visa volumes as a decision-making criterion”. This was a mistake. As prior work by CGD has found, for some of the considered occupations, Skilled Worker visa holders accounted for the majority of new entrants to the workforce. For these occupations, training would need to be massively—perhaps implausibly—scaled up to make up for the loss of visas.

For example, the “Construction and building trades (not elsewhere classified (n.e.c.))” occupation currently receives 56 percent of new entrants from visas; replacing them will require an increase in starts equal to more than 100 percent of current training levels, with workers taking a year and a half to complete training. At the same time, demand is expected to rise by 17,000 workers by 2030, thanks in part to a government housebuilding effort.

The MAC noted that departments’ demand estimates also varied massively in terms of quality and methodology. Several appear to have used Skills England’s Skills Needs Assessments, which only project total employment demand and ignore replacement demand, an oversight given that demand for additional workers may be as influenced by cohort retirement rates as they are by industrial policy priorities.

The MAC is therefore trying to walk a tightrope between “the government’s objectives of reducing employers’ reliance on migration and reducing risks to the delivery of the Industrial Strategy.” This has previously led to inter-ministerial tensions. The Industrial Strategy Advisory Council (ISAC) and Skills England have argued that it would not be “worth risking adverse impacts on business for relatively small reductions in net migration in some occupations.” The MAC, in response, noted its obligation to focus on Jobs Plans.

Assessing the MAC’s recommendations

The MAC has recommended that 28 of 82 considered occupations have access to the TSL. Those 28 occupations account for around 3,800 of the 9,700 annual visas issued in recent years. They will only be allowed to use the TSL for 18 months; none of the 28 were considered to have produced a strong enough Jobs Plan to justify the full three years of access. (An improved Jobs Plan could see an occupation allowed onto the list for three years in the future.)

Our previous research found that Skilled Worker visa holders made up an average of 27 percent of new workforce entrants to the 82 considered occupations. We then asked whether domestic training pipelines could plausibly scale to replace the workers lost to visa changes and, if so, how quickly. Based on these quantitative indicators, we created an index showing which occupations should be on the TSL.

The MAC’s recommendations align relatively well with ours: of the 28 recommended occupations, 68 percent are in the top 30 of our index, and 79 percent within the top 40. Only six recommended occupations were outside the top 40. Most are justifiable on the basis that they are sufficiently strategically important that it is worth ensuring labour availability. This suggests that the imperfect narrative evidence process got enough information through. But formalising a consideration of visas within the balance of labour supply would be a useful improvement to the process.

Figure 1 shows visa usage for the occupations recommended for inclusion and exclusion. Many of the exclusions, including the largest ones, are highly justified. "Sales accounts and business development managers” and “Marketing associate professionals” averaged more than 1,000 visas per year, but were ranked very low on our index. As the MAC observes, they overlap at the task level with similar roles, allowing for domestic substitution.

Figure 1

But there are also a few surprises.

“Chemical and related process operatives” are included, despite the fact only nine visas were granted in 2024/25. We couldn’t even rank them, given insufficient data. The MAC gives them the benefit of the doubt because of future and area-specific pressures, credible actions, low exploitation risk, and compatibility with sponsorship.

“Air-conditioning and refrigeration installers” are excluded. This is somewhat surprising: they ranked #27 on our index. The MAC accepts that future demand could be substantial, but felt the Jobs Plan was too vague, and that small firms and self-employment may limit sponsorship. Yet the domestic pipeline is extremely weak: replacing visa inflows would require apprenticeship starts to rise by around 19 percent versus 2024/25 levels, and training takes almost three years. Tightly targeted access for strategically relevant heat-pump or commercial-refrigeration roles should have been considered, rather than closing the route altogether. Other occupations—plasterers, floorers, and wall tilers— were included based on strategic importance, despite weak Jobs Plans.

“Stonemasons” were also excluded. They ranked #24 on our index: visas account for around 62 percent of observed new entrants, and replacing them would require apprenticeship starts to rise by around 122 percent. Training new apprentices takes roughly two years, and the apprenticeship pipeline is very small. No Jobs Plan was submitted, so the case for inclusion was never substantively tested, but its pipeline looks fragile.

Overall, however, the MAC’s recommendations are sensible. In Figure 2 we show the percentage increase to 2024/25 apprenticeship starts needed to replace average annual visas per occupation. With the exception of those occupations discussed above, we agree that the occupations most reliant on visas for new labour supply—and without easy domestic substitution options— have been selected.

Figure 2

One recommendation we particularly welcome is the choice to include “Overhead line workers”, contained within the broader “Electrical and electronic trades not elsewhere classified” occupation. These workers are essential to grid expansion; our analysis found that their reliance on Skilled Worker visas has risen sharply, suggesting rapid growth in demand. Ensuring this specific role is defined will support industrial policy goals, while reducing the risk that a catch-all occupation code is misused.

Recommendations for improvements

Some of the MAC’s rationales for including or excluding occupations appear inconsistent, reflecting uneven narrative evidence from government departments. The next iteration of the TSL process should therefore do the following:

1. Proactively assess the importance of visas.

Add a formal “withdrawal-impact test”, in which estimates of the impacts of the loss of labour on industrial policy goals are calculated. This should include consideration of the capacity of the domestic training system to scale up to meet increased demand, and of retirement rates. (This may require better data on training-to-occupation pipelines, e.g., by making use of education outcomes data held by the Department for Education.) It should also consider the impact of the loss of labour inputs and outputs in achieving strategic targets, which may require better data on how employers respond to changes in labour supply.

2. Measure progress towards scaling training over a reasonable timeframe.

For the included occupations, the MAC recommends only 18-month access to international recruitment while Jobs Plans are improved. As a timeframe for improving plans this is understandable, but as a timeframe in which to increase the throughput of trained workers it will be very challenging. Several included occupations have training lengths of 40 months or more, and high apprenticeship attrition rates. For some—such as electricians—longstanding efforts to scale training haven’t worked.

The government should therefore measure progress in domestic training; the impact of TSL access on the scale of domestic training; and give industry appropriate time to meet these benchmarks. Further assessments should consider apprenticeship starts, completions, and retention rates; whether training capacity is being proactively expanded; the extent to which employers are using relevant skills funding; and evidence that attractiveness problems are being addressed.

3. Ensure small occupations don’t fall through the cracks.

It seems very plausible that some occupations’ Jobs Plans were overlooked by under-resourced departmental teams. Requiring departments to explain in writing why they are not submitting a Jobs Plan would provide reassurance, and may indicate where greater capacity is needed.

We welcome the MAC’s efforts to try and align labour supply and industrial policy goals, but these recommendations show that much more needs to be done to properly take supply (of both domestic and international workers) into account.