How Foreign-Trained Health Workers Saved the NHS £14 Billion

One in three doctors and nurses working in the UK were trained overseas. If the NHS had paid to train them it would have cost the government £100,000–£120,000 per doctor and over £20,000 per nurse. Multiplying through these numbers adds up to a cumulative £14 billion saving to the NHS in training costs based on workers currently in post, even with conservative assumptions.

This question, how much the UK saves by recruiting foreign-trained health workers, is at the heart of an inquiry by the UK's All Party Parliamentary Group (APPG) on Global Health, supported by Global Health Partnerships (formerly THET) and the Center for Global Development (CGD). This piece sets out an attempt to answer it.

To estimate this, I multiplied the number of foreign national doctors and nurses by an estimate of what it would have cost to train a British doctor or nurse. We follow a similar approach to the Royal Society of Canada and Canadian Academy of Health Sciences (Box 7).

There are plenty of assumptions involved in this calculation. For one, in a counterfactual world without foreign doctors and nurses, we might have just had fewer health workers and worse health outcomes, rather than have paid to have trained more domestically. That would hardly have been a better outcome. Regardless, it is clear the UK health service has in fact benefitted from investments in training made by much poorer countries.

Who are we counting

We start with the total numbers for doctors working in hospitals, from the NHS Workforce Statistics for June 2025. These statistics report staff by nationality (rather than where they trained) but it’s reasonable to assume that the majority of non-British staff trained overseas. The sum of all non-British hospital and community health service doctors, consultants, associate specialists, speciality doctors, and staff grade doctors currently working in the NHS is 81,000. This is 34 percent of the total. The General Medical Council (GMC) reports that doctors who qualified outside of the UK currently make up around 42 percent of the workforce, so using non-British staff as a proxy may be a conservative underestimate of those trained overseas (though the GMC figures include private as well as NHS doctors). The sum of non-British nurses, health visitors, and midwives comes to 129,000 (29 percent of the total). In addition, the General Practice Workforce data for September 2025 shows a total of 9,000 non-British GPs (26 percent of the total).

What would it have cost to train them here?

Estimating the cost of training is a little more difficult. The most commonly cited figure comes from the University of Kent Personal Social Services Research Unit (PSSRU). The average doctor is aged around 50 and will have completed their training around 20 years ago. We therefore use historical rather than current cost data. The oldest version of the annual “Unit Costs of Health and Social Care” readily available online is the 2007 report. The report adds up tuition, placement, and training salary costs, and comes to roughly £350,000 for a consultant, £300,000 for a GP, and £65,000 for a nurse. These are the total economic costs, including both a government subsidy but also a large share of costs borne by individual trainees. That includes portions of fees that are repaid out of loans, and the trainees own living costs. PSSRU doesn't provide a breakdown of government and individual costs. Full Fact estimated that, given loan repayment rates, the cost to the taxpayer is around 70 percent of the headline training cost, with the remainder coming from student loan repayments and private funding. That 70 percent figure includes salaries for trainees. But that isn’t a pure training cost. Trainees are also delivering a service whilst working, even if it’s not at the same level of service as a fully qualified doctor. Putting a number on the share of a trainee salary that should count as training is not obvious. For simplicity, we make the conservative assumption that half of government spending is buying a service, with half being a pure investment spent on training. So 35 percent of total costs are spent purely on training by the government. This is a big assumption, but we think it is conservative, and if anything the costs could be higher.

We can now multiply through these numbers, and arrive at that headline figure, of £13.8 billion in savings. This is a conservative calculation which ultimately likely underestimates savings. It excludes other occupations besides doctors and nurses, for instance. But it provides a meaningful order of magnitude estimate.

£14 billion: the bill the UK never paid

Notes: Data on doctors in hospitals are from the NHS Workforce Statistics for June 2025, and are the sum of all non-British nationalities for hospital and community health service doctors, consultants, associate specialists, speciality doctors, and staff grades. Data for nurses are from the same source and are the sum of non-British nurses, health visitors, and midwives. Data for GPs are from the General Practice Workforce data for September 2025. Costs estimates for doctors are the University of Kent Personal Social Services Research Unit, assuming that 35% of total costs are borne by the government.

What are the annual savings today?

The stock £14 billion figure tells us the accumulated value of decades of overseas recruitment. To understand what this means going forward, it's worth looking at what the UK is saving right now. Over the most recent year for which we have data (June 2024 - 2025 for consultants and nurses, September 2024 - 2025 for GPs), there has been an increase in the stock of foreign nationals working in the NHS of around 4,000 doctors, 4,000 nurses, and a slight decrease in the number of GPs (by 425). If we assume they would all have otherwise been trained domestically at current costs, that adds up to an annual cost of £1.1 billion.

The UK saved over £1 billion in the last year alone from not paying to train foreign health workers

Can the NHS afford to go it alone?

The government’s 10 year plan for the NHS, launched in July 2025, includes a target for only 10 percent of recruitment to be from overseas. In the most recent year, international recruitment was around 30 percent. For the government to have paid to have trained 20 percent of actual recruitment in the last year would have cost an additional £700 million. Closing that gap domestically would require not just funding but a significant expansion in university places, clinical placements, and training infrastructure—none of which is currently costed or planned in the 10-year document.

The bottom line

It’s clear the UK health system is heavily reliant on internationally trained health workers. With savings of around £14 billion in avoided training costs, and over £1 billion annually at recent recruitment levels, the scale of overseas recruitment is not a marginal supplement to domestic training, but a core component of sustaining the NHS workforce.

Against this backdrop, the government’s ambition to reduce international recruitment to 10 percent over the next decade appears highly ambitious. Achieving this would require far more explicit and credible planning to expand domestic training capacity and funding than is currently set out.

The £14 billion figure is an accounting exercise. The policy question it raises is rather more serious.

CGD and the World Bank will be hosting a discussion on this topic at the World Health Assembly in Geneva on May 20, 2026; and releasing two research papers on modalities to manage health worker mobility and channel HRH financing. To learn more and contribute, please contact Anastassia Demeshko ([email protected]).