5. Health and Social Care
Risk level: High - June 2026
The NHS Has No Headroom That is the sentence that should be at the top of every health policy document the government produces in 2026. Not the waiting list. Not the staffing crisis. Not the social care funding gap. The headroom. The ability to absorb an unexpected shock, manage a winter surge, or respond to a compounding pressure without the whole system tipping into crisis.
It does not have that ability. And the food price nutrition effect is about to make that worse, on a timeline that standard NHS planning does not account for.
The Compounding Problem
Let us be clear about what is hitting simultaneously.
The NHS waiting list stands at over 7 million people. That is the visible backlog. The invisible one - people who have learned not to bother presenting - is larger. Elective capacity is constrained. Diagnostic services are stretched. The system is running at utilisation rates that leave no slack.
Winter surge is a known problem and a worsening one. The combination of an ageing population, more people living with multiple chronic conditions, and a social care system that cannot absorb demand means that every winter since 2018 has produced NHS performance that would have been considered a crisis a decade earlier. The baseline is already unacceptable. The question is how much worse it gets.
Social care collapse is not a metaphor. It is a specific mechanism by which the NHS loses bed capacity. When an elderly patient is medically ready for discharge but cannot be moved to a care home or domiciliary care package because none is available, they occupy an acute hospital bed. That bed is not available for the next emergency admission. The corridor in A&E becomes the result. This is not a failing of NHS staff. It is a system design failure with a specific causal mechanism.
And now: the food price nutrition effect. When food becomes more expensive, households shift toward calories that are cheaper per unit of energy - processed foods, refined carbohydrates, vegetable oils, sugar - and away from fresh vegetables, protein, whole grains, and fruit. The health consequences of that sustained shift arrive 18 months to three years later. Cardiovascular disease, Type 2 diabetes, deterioration in elderly nutritional status. The harvest being reduced in June and July 2026 arrives as reduced supply in autumn. The food price peak hits in spring 2027. The health consequences peak in 2027 and 2028. By the time they appear in the data, the prevention window has closed.
The Workforce Crisis: The Numbers
Before any solution can be designed, the scale of the workforce problem needs to be stated honestly.
The NMC register holds approximately 788,000 registered nurses (UK-wide, excluding midwives and nursing associates). Every year, some of those nurses retire, emigrate, leave the profession due to burnout or ill health, or simply stop renewing their registration. In the year to March 2025, approximately 28,789 nurses left the register - roughly 3.5% of the total register.
That figure is the first number that matters.
To maintain current NHS staffing levels - not expand, not improve, just hold the current position - the NHS needs to recruit and retain approximately 29,000 new nurses per year, every year, simply to replace those who leave. That is the break-even point. Any expansion requires net additions above that line.
Against that baseline: in the same year to March 2025, approximately 52,833 nurses joined the NMC register - giving a net annual addition to the register of roughly 24,000. That net figure sounds encouraging until you break it down. Roughly 39% of new joiners are internationally educated nurses, despite a 30% year-on-year decline in international recruitment. The domestic pipeline is not yet growing fast enough to compensate. And the net figure is deteriorating.
The second number that matters is the vacancy rate. The NHS currently has approximately 110,000 vacancies across all staff groups. The largest single vacancy gap is in nursing.
The third number that matters is the 1-in-5 figure: approximately one in five newly qualified nurses leaves the NHS within their first two years. The single biggest driver of that early departure is not pay alone - it is a combination of pay, workload, and the band 5-to-6 career progression bottleneck. An experienced nurse who cannot progress and who is working under sustained pressure will, eventually, leave. Every departure of an experienced nurse represents a training investment wasted.
Year 1–5 Net Addition Model
A workforce programme needs a year-by-year pipeline. Here is what the numbers look like, stated honestly.
Starting point: approximately 788,000 nurses on the NMC register. Annual exit rate: approximately 3.5% (roughly 28,800 leavers per year). Annual net addition under current trends: approximately 24,000 per year.
The programme's ambition is to grow the register by enough to close the vacancy gap and provide headroom. That requires net additions materially above the 24,000 baseline - sustained over years.
The table below shows the year-by-year NMC register net addition under the programme's workforce strategy, with each component stated separately. Numbers are rounded estimates based on current trends and announced programme ambitions; they are not precise forecasts.
| Year 1 | Year 2 | Year 3 | Year 4 | Year 5 | |
|---|---|---|---|---|---|
| Domestic graduates (new to register) | ~21,000 | ~22,000 | ~23,000 | ~24,000 | ~25,000 |
| Accelerated 2-year graduate entry | ~1,000 | ~2,000 | ~3,000 | ~4,000 | ~5,000 |
| International recruits | ~5,000 | ~6,000 | ~6,000 | ~5,000 | ~5,000 |
| Return-to-practice | ~1,000 | ~1,500 | ~2,000 | ~2,000 | ~2,000 |
| Total inflows | ~28,000 | ~31,500 | ~34,000 | ~35,000 | ~37,000 |
| Less: projected register exits | ~28,800 | ~29,000 | ~29,500 | ~30,000 | ~30,500 |
| Net annual addition | ~−800 | ~+2,500 | ~+4,500 | ~+5,000 | ~+6,500 |
What this table shows: The programme does not produce significant net register growth until year 2, and meaningful growth does not arrive until years 3–5. Year 1 is approximately break-even or slightly negative - the international recruitment uplift and return-to-practice programmes are ramping up while the accelerated degree expansion has barely begun.
Three assumptions drive this model, and all three carry honest uncertainties:
Domestic graduates (~21,000–25,000 per year): This is the largest single inflow, and also the slowest to expand. The binding constraint is not university places - it is clinical placement capacity in NHS trusts. Each student nurse requires a supervised placement in an NHS setting for typically 2,300 hours of clinical training. NHS trusts have a finite number of placement slots. Expanding them requires additional practice educators (senior nurses who supervise students), trust capacity to take on the supervision burden, and funding. NHS England's Long Term Workforce Plan 2023 targets a significant expansion in nursing degree places, but the placement capacity to deliver on those targets has not been costed or resourced at the level required. The Year 3–5 domestic graduate figures above assume that placement capacity expansion is funded and delivered - which is not guaranteed and is the single largest risk to the domestic pipeline.
International recruitment (~5,000–6,000 per year): The 5,000 figure in the current post is too small to be a headline, but it is in the right order of magnitude for a sustainable annual rate under a reduced-dependence strategy. The government's own NHS Long Term Workforce Plan targets international recruits constituting less than 10% of the NHS workforce by 2035 - down from historically much higher levels. The NMC data shows internationally educated nurse joiners have fallen approximately 30% year-on-year and are declining from all major source countries (India, Philippines, Nigeria). The ethical question - whether the UK should recruit from countries that need those nurses more than the UK does - is legitimate. The operational answer is that a modest, ethically-sourced international pipeline (5,000–6,000 per year) is a bridge to when domestic supply scales, not a permanent strategy. It is also the only route to net additional nurses in years 1 and 2.
Return-to-practice (~1,000–2,000 per year): There is a pool of nurses with lapsed NMC registration who have the training to return. The realistic annual return rate under a well-resourced programme - with CPD costs covered, a structured re-entry pathway, and attractive terms - is estimated at 1,000–2,000 per year. The current annual returner rate is substantially lower (hundreds, not thousands). Achieving the higher end of this range requires that the barriers to return - primarily the cost of revalidation CPD, the loss of income during re-training, and the bureaucratic complexity of restoration - are explicitly funded and simplified. The figure above assumes that a properly funded return-to-practice programme is in place.
The programme's five-year target of approximately 20,000–25,000 net additional nurses on the register is plausible but not certain. It depends on the placement capacity constraint being resolved (the biggest risk), the return-to-practice programme being properly funded, and international recruitment being managed ethically and sustainably. If all three deliver, the register grows by approximately 20,000–25,000 over five years - closing part but not all of the current 110,000 vacancy gap.
Retention: Why Nurses Leave and What Fixes It
The pipeline numbers above only matter if the nurses recruited also stay. Retention is not a secondary consideration to the workforce programme. It is half the programme.
The Nuffield Trust's April 2026 analysis of the NHS 50,000 nurses programme (covering 2015–2023) is the most authoritative account of why NHS nurses leave. The findings are specific enough to be useful:
Mental health is the strongest single driver. Nurses with more than 84 days of mental-health-related sickness absence in a three-month period were more than six times more likely to leave than those with no mental health sickness absence. The NHS staff survey consistently shows elevated rates of anxiety, depression, and burnout among nurses compared to other professions. This is not a soft concern. It is a workforce retention metric.
Pay has eroded significantly. Real-terms nurse pay is down approximately 10.7% since 2010, compared to a 6.9% average public sector decline (IFS). The 2026 NHS pay deal was described by the Nuffield Trust as "a step in the right direction" - an honest assessment of a partial improvement.
The band 5-to-6 career bottleneck is a structural problem. Nurses who want to progress cannot do so because band 6 positions are limited. Staff at the top of band 7 - who have been in post longest and accumulated the most experience - leave at higher rates than those at the bottom of band 7. The career structure that the NHS offers does not retain experienced staff who cannot progress further.
The international nurse retention gap is narrowing. Internationally educated nurses were previously significantly less likely to leave the NHS than UK-educated nurses. That gap has narrowed substantially: in 2020/21 the leaving risk for internationally educated nurses was 36% lower than for UK nurses; by 2023/24 it was only 18% lower. International recruits are now leaving at rates closer to domestic staff, meaning international recruitment is less effective as a long-term retention strategy than it used to be.
The retention interventions that follow from this evidence are specific:
A meaningful pay award - not just the 2026 deal, but a sustained commitment to pay restoration. The 10.7% real-terms erosion cannot be reversed in one year. The direction of travel matters as much as the quantum.
Mental health support as a workplace intervention, not a benefit scheme. The evidence shows that nurses with high mental health sickness absence leave. The intervention is not a confidential counselling line - it is a reduction in the workload and conditions that produce the sickness absence in the first place. Mental health support wrapped around an unchanged working environment is a retention intervention that will fail.
Band 5-to-6 progression reform. Expanding band 6 and 7 positions, creating new advanced clinical practitioner roles, and fund postgraduate training for nurses who want to develop. This costs money. It also reduces the training investment lost to early departure.
The honest cost of a comprehensive retention programme - pay restoration pathway, mental health and working conditions reform, career progression expansion - is not yet precisely modelled. The 2026 pay deal has an estimated annual cost of approximately £3–4bn across all staff groups; extending and deepening it for nurses specifically would be a significant additional commitment. The expected return: a reduction in the annual exit rate from approximately 3.5% toward 3.0% - saving approximately 4,000 nurses per year from leaving. At replacement training cost of approximately £50,000–£80,000 per nurse (university fees plus supervision cost over three years), avoiding 4,000 departures is worth £200m–£320m per year in training cost avoided, before productivity and quality effects are counted.
Role Substitution: Nursing Associates and Physician Associates
"Train more nurses" is necessary but not sufficient. The NHS also needs to use its existing registered nurses more efficiently, and it needs to develop new roles that extend what the current workforce can do.
Nursing associates are the fastest-growing professional group on the NMC register. There were approximately 12,800 on the register as of March 2025, growing at 17.5% year-on-year. They are an England-only, two-year training programme that sits between healthcare assistants and registered nurses - a bridging role that allows registered nurses to focus on higher-scope work while nursing associates manage more routine care. They are not registered nurses and they do not replace registered nurses. They are a workforce multiplier: they allow each registered nurse to practise at the top of their scope rather than spending time on tasks that a nursing associate could manage.
The programme's target should be to expand nursing associate numbers to approximately 40,000–50,000 by 2030, with the training capacity and supervision infrastructure to support that growth.
Physician associates are a different kind of addition - a role that extends the medical workforce rather than the nursing workforce. There were approximately 3,528 physician associates registered with the GMC as of September 2025, with GMC regulation now mandatory (deadline: December 2026). They are not yet on the NMC register (they are GMC-regulated, not NMC-regulated) and they are not a substitute for nurses. They are a supplement to the medical workforce, working under consultant supervision in secondary care.
The role substitution argument is not that nurses are being replaced by lesser-trained staff. It is that the NHS has been under-using both the potential of new non-nurse roles and the potential of registered nurses working at the top of their scope. Expanding nursing associates and developing physician associates together gives a workforce model that grows the total clinical capacity without requiring every additional clinician to be a fully trained registered nurse or consultant.
The Social Care Workforce: The Gap Nobody Talks About
The NHS does not employ the largest part of the social care workforce. Local authorities and independent providers do. And that workforce is in a more fragile state than the NHS workforce.
There are approximately 740,000 people working in domiciliary care (care at home) in England, making it the largest single part of the social care workforce. The majority are paid at or close to the minimum wage. The majority of international recruits in social care are in this sector - the care worker visa route has been a significant source of labour for providers who cannot recruit domestically at the wages they are permitted to pay.
This matters for the NHS workforce programme in a specific, mechanical way: the NHS and social care compete for the same low-wage labour pool. A healthcare assistant in the NHS is paid on band 2 of the NHS pay scale. A domiciliary care worker in the independent sector is typically paid at or close to the minimum wage. The King's Fund's Social Care 360 (April 2026) estimates that the average care worker was paid 31 pence per hour less than a newly employed NHS health care assistant in 2024/25. That is not a large pay gap in percentage terms. It is large enough to cause an exodus - care workers who can move to an NHS band 2 position for better pay, a pension, and more stable working conditions will do so, and they are doing so.
The consequence: NHS workforce expansion without a simultaneous social care workforce plan does not just fail to solve the social care problem. It actively makes it worse, by drawing staff out of the sector that is already most understaffed.
The social care workforce programme needs three things simultaneously:
A real-terms pay improvement for care workers. Not the national minimum wage - the real-terms value of the minimum wage after inflation. The Low Pay Commission's analysis suggests that the adult social care sector has high turnover driven by pay and working conditions. A pay improvement that closes some or all of the 31p/hour gap with NHS band 2 reduces the outflow incentive and improves retention.
A supervised career structure. Domiciliary care work is not currently designed as a career. It is designed as a job - one with high physical and emotional demands, zero-hours contracts in many providers, and little progression. A career structure for care workers - with progression to senior care roles, community nursing associate pathways, and training time funded - changes the retention picture.
An honest acknowledgement that the current international recruitment reliance is not sustainable. The social care sector's use of international recruits at the current scale is a symptom of a domestic recruitment failure, not a solution to it. The ethical question for social care is more acute than for the NHS: the care workers recruited internationally from lower-income countries are often in the roles with the lowest pay, the hardest conditions, and the least support. A sustainable social care workforce model needs to recruit domestically as its primary route.
The Dilnot Problem: Why This Time Might Be Different
The £86,000 lifetime cap on social care costs was Dilnot's central recommendation in 2011. It was accepted by the coalition government. It was delayed from October 2023 and then delayed again. As of mid-2026, the care cap and the means-test reform have been referred to the Casey Commission (reporting phase 1 in 2026, phase 2 in 2028), with no confirmed implementation date and no legislative vehicle.
This is the Dilnot problem. Every government since 2011 has acknowledged that something must be done. None has done it.
The honest reasons for the failure are: (1) Treasury fiscal concern - the cap requires significant up-front government expenditure before savings materialise; (2) political priority - it is never at the top of any government's agenda when it comes to choosing between spending commitments; (3) implementation complexity - local authority capacity to handle the means-testing changes is limited; (4) parliamentary arithmetic - Conservative governments with small majorities have been unable to absorb the political cost without backbench rebellion.
This programme should implement the care cap. The honest statement of what that requires: primary legislation, a three-to-five year implementation timeline from the point a Bill is introduced, and a fiscal commitment that is front-loaded (implementation costs in years 1–3, savings from reduced emergency hospital admissions and improved workforce stability in social care arriving in years 4–8). The Health Foundation's estimate of the cap element's ongoing cost is in the range of £1bn–£3bn annually depending on the cap level set.
The honest answer to "why would this government succeed where previous governments failed" is: it depends on parliamentary arithmetic. If the government has a working majority and is willing to treat social care reform as a first-order priority, it can be done. If parliamentary arithmetic is difficult, it requires the political courage to make the case that the cost of inaction - measured in NHS bed days, cancelled electives, and a social care workforce that cannot be recruited or retained - is higher than the cost of the reform. That case has not been made successfully yet. Whether it can be made depends on political conditions this programme cannot control.
What This Costs vs What Inaction Costs
The health programme totals approximately £5.3–8.5bn per year across its components. Here is how that number breaks down into specific workforce expenditure lines:
NHS workforce expansion (international recruitment infrastructure, domestic training capacity, accelerated programme funding, placement educator expansion): approximately £2–2.5bn annually. Against this: the current cost of vacancy-driven agency staffing across the NHS is running at approximately £3bn annually and rising. A properly implemented workforce strategy costs roughly the same as the failure to implement one.
Retention programme (pay restoration pathway, mental health and working conditions reform, career progression expansion): approximately £1–1.5bn annually above the baseline pay award. The 2026 NHS pay deal is the first step. Completing the pay restoration pathway and extending it to the specific retention-sensitive groups (newly qualified, band 5–6 transition, experienced band 7) is the second.
Social care workforce (domiciliary care pay uplift, career structure, domestic recruitment campaign): approximately £1–2bn annually. This is the most underfunded component of the current programme and the one whose absence would undermine the NHS workforce strategy by starving social care of the staff it needs to function.
Free school meals expansion and Healthy Start uprate: approximately £1bn annually. This is the nutrition intervention described above - a public health intervention that costs approximately £1bn and prevents health consequences measured in hospital admissions and GP appointments that cost an order of magnitude more to treat.
Winter surge and emergency discharge fund: approximately £0.5–1bn in years 1 and 2, tapering as the structural workforce and social care programmes take effect.
Social care Dilnot implementation: implementation costs of approximately £1.5–2bn in transition, with ongoing costs of £1–3bn annually depending on cap level - partially offset by savings from reduced emergency hospital admissions.
Against all of this: the NHS winter crisis of 2023–24 produced approximately 500,000 cancelled electives, an estimated 20,000–30,000 excess deaths attributable to delays in emergency care, and excess emergency admission costs of approximately £2–3bn across every trust in the country. A winter 2026 that is worse than 2023, with less headroom and the nutrition lag effect beginning to bite, is the central case, not the pessimistic scenario.
The return on investment in health and social care is not theoretical. The cost of preventing an emergency admission through a social care package is roughly £1,000–£2,000 per week. The cost of the emergency admission it would have prevented is roughly £4,000–£6,000 per day. The arithmetic is not complicated.
What This Post Depends On
Health policy cannot be written in isolation from the other policy areas in this series.
Food security. The nutrition lag effect described in this post is the downstream health consequence of the food price shock described in the food security post. The free school meals expansion and the Healthy Start uprate are direct health interventions that depend on the food supply being available at any price. If the food security programme fails and supply becomes genuinely constrained, the health interventions in this post are working on top of a much harder problem.
Social security. The work versus welfare disincentive discussion in the social security post connects directly to the health workforce question. If social security policy creates high effective marginal tax rates for low-income workers, the NHS will struggle to recruit and retain the support staff and lower-paid clinical roles where the vacancy gap is largest.
Fiscal framework. Every cost figure in this post is a real commitment that has to be funded. The fiscal framework post needs to treat health and social care as a first charge on any additional borrowing capacity, not as a competing claim on a constrained budget. The cost of inaction in health and social care is higher than the cost of action, and that comparison needs to be made explicitly.
Energy. The hospital sector is a significant energy user. The social care sector is an even larger energy user. Fuel poverty, addressed in the energy post, is also a health issue: cold homes produce respiratory infections, cardiovascular events in elderly people, and mental health deterioration.
The health system in June 2026 is a system under compounding stress with no headroom. The nutrition lag effect is about to add a new pressure that no one is planning for. The social care system is one bad winter away from a total collapse that makes the NHS impossible to run. And the workforce that is supposed to hold all of this together is shrinking faster than it can be replenished.
The government that faces this honestly has hard choices and a narrow window. The government that pretends these are separate problems that can be managed independently will fail all of them simultaneously.
Who Owns the Care? The Case for Co-operative and Non-Profit Models
The current social care market is fragmented across private equity-backed chains, large corporate providers, local authority in-house services, and a large residual of small independent providers. The ownership structure of a care provider affects the quality of care delivered through the employment model it can sustain - and the evidence consistently shows that for-profit providers, at the wages the current market permits, cannot consistently fund the training, supervision, and staffing levels that quality requires. This is not a criticism of individual managers. It is a structural consequence of a ownership model that requires a margin.
The policy mechanism is a deliberate shift in commissioning and market structure toward co-operative and non-profit ownership. The transition pathway: first, create a right for care workers to request an Employee Ownership Trust take-over when private providers fail or exit a market - similar to the right introduced for worker buyouts in other sectors; second, establish a national development fund for social care mutualisations, capitalised from the savings generated by reducing emergency hospital admissions; third, make co-operative and non-profit providers the preferred commissioning partner in all new local authority social care contracts, reversing the presumption in favour of competitive tendering that has pushed the market toward large for-profit chains. The sector will not transform overnight. A ten-year horizon for a managed transition is honest.
The precedent that makes this realistic rather than utopian: Italy's social care co-operative sector employs approximately 300,000 people in health and social care through worker-owned and social co-operative models, with documented better retention rates and stronger voice for care workers in service design. The Japanese welfare company model - which gives care workers an equity stake in the organisations they work for - has reduced turnover in a sector with similar demographic pressures to the UK's. Closer to home, the John Lewis Partnership demonstrates that large-scale employee ownership is compatible with complex service delivery and sustained quality. A government that wants a social care workforce that stays and that has a genuine stake in the quality of what it delivers has a century of evidence to draw on.
NHS Data Infrastructure and Sovereignty
There is a genuine data sovereignty question in the NHS that is not currently being asked in the right way. US technology companies contracted to build or operate NHS data infrastructure - Palantir being the most prominent current example - are subject to US surveillance law in ways that UK patient data is not. Data held on US-owned or US-hosted infrastructure may be accessible to US intelligence agencies under FISA Section 702 or the CLOUD Act, without UK court jurisdiction and without UK government knowledge. That is not a theoretical risk. It is the current legal position for any NHS operational data processed through US-owned cloud infrastructure.
The policy mechanism is straightforward: NHS data infrastructure should be built and owned by NHS bodies or by a statutory public-benefit data trust governed under UK law, with all data processing subject exclusively to UK jurisdiction. No NHS operational data - patient records, capacity data, workforce data - should be held on infrastructure that is subject to foreign surveillance law by default. This is not a ban on US technology suppliers. It is a requirement that any supplier operating NHS data infrastructure does so on NHS-owned or UK-statutory infrastructure, under UK legal governance, with UK court jurisdiction over any access request. The alternative - continued outsourcing of NHS data to US tech platforms on US legal terms - is a choice to treat British patients' health data as a US legal exposure, not a UK sovereign asset.
Part 5: Health and Social Care - part of The Country That Works For You series
Previous: 4. Industrial Strategy · Series Index · Next: 6. Housing and Planning