Short answer
NHS trusts go into deficit when the recurrent cost of delivering safe care exceeds the recurrent income available to fund it, and when the underlying causes of that gap are not resolved. The gap is rarely explained by one thing alone. It usually reflects a combination of demand growth, workforce cost, productivity constraints, inflation, capital underinvestment, service configuration, and financial plans that rely too heavily on short-term fixes rather than structural change.
A deficit is a financial result, not a full diagnosis. It tells leaders that expenditure exceeded income in a given period. It does not, on its own, explain whether the problem sits in demand, capacity, workforce, service design, pricing, operating model, or the wider mismatch between what the system is funded to do and what it is trying to deliver.
Why it matters
It matters because boards can respond to the wrong problem. If a deficit is treated only as a budgeting failure, the response tends to focus on vacancy controls, expenditure freezes, procurement rounds and savings targets. Some of that may be necessary, but it often balances the year without changing the conditions that recreate the gap.
That is one reason financial recovery programmes can look active without making the organisation more sustainable. When staffing is reduced but the work remains, pressure often returns through agency use, waits, deteriorating flow or weaker clinical performance. When hospital costs are cut without redesigning pathways or community alternatives, cost and risk can simply move elsewhere. This is our interpretation, but it is consistent with national evidence on non-recurrent savings, workforce instability and weak productivity recovery.
What the national evidence shows
The national evidence does not support a simple explanation such as weak local grip or poor financial discipline. The National Audit Office found that, after the pandemic, NHS demand, waiting lists and case complexity remained high while real-terms funding growth slowed and productivity recovery lagged. It also reported that non-pay inflation added £1.4 billion above budget in 2023/24, that agency spending reached £3.0 billion, and that just over half of the £5 billion of efficiencies it examined in 2022/23 were non-recurrent.1
Nuffield Trust analysis of provider accounts reached a similar conclusion from another angle. After adjusting for the IFRS 16 accounting change, it reported a £1.2 billion provider-sector overspend in 2023/24, equal to 0.9% of income, and argued that the underlying gap between stable income and recurrent expenditure was at least £4.5 billion. It also found that deficits were no longer confined mainly to acute trusts. Acute providers still reported the deepest overspends at 1.2% of income, but specialist trusts had moved from a 1.8% surplus in 2019/20 to a 0.9% overspend, and mental health trusts from a 1.1% surplus to a 0.1% overspend.2
NHS England's planning record shows the same structural pattern from the inside. In May 2025 it reported that all systems were aiming for a net balanced position in 2025/26, but that without revenue support this amounted to a £2.2 billion aggregate system deficit, closing an initial £4.4 billion planning gap. Planned efficiencies averaged 7.1%.3
The 2025/26 outturn, and what it does and does not prove
The most recent position complicates the picture in a way boards should take seriously. The NHS balanced nationally in 2025/26, ending the year with a £70 million underspend, the first time in a decade the position was delivered without a reserve claim from HM Treasury. Agency spend was almost halved, from £2.1 billion to £1.2 billion. The number of organisations delivering their plan rose from 70% to nearly 80%.4
That is a real achievement and it should not be written off as accounting. Grip moves the number.
Underneath it, systems still ran a £563 million deficit, with 15 systems in deficit against 17 the year before, and five systems accounting for 88% of the total. Planned efficiencies of £11 billion were set, of which £4.8 billion, 44%, was judged high risk at the outset. Delivery came in at £10.2 billion, a shortfall of £842 million, and NHS England attributes the remaining overspend to slippage against efficiency plans, including workforce costs above planned levels, and to the cost of industrial action.4
Balance at the national line and a closed structural gap are not the same thing.
This is the distinction that matters for a board. A system can hit its control total through grip, deficit support and non-recurrent action while the recurrent gap underneath remains open. The question is not only whether the year balanced, but what will have to be found again next year.
How the problem works
The most useful way to understand a deficit is to trace the conversion from population need to resource use, activity and outcome.
Where a deficit is actually made
The financial position is the last link, not the first. Most of what determines it is decided upstream.
Demand creates the need for care. Resources provide staff, estates, equipment, digital capability and management capacity. The operating model determines how those resources are deployed, how pathways work, how decisions are made and where capacity sits. That operating model then generates activity, cost and outcomes.
A deficit emerges when that conversion chain becomes financially unsustainable. The trust may be carrying a level of demand it cannot manage safely at funded cost. It may be using expensive temporary staffing because the service model, job design or local labour market does not support stable workforce supply. It may be maintaining fragmented pathways that generate avoidable admissions, duplicated diagnostics, delayed discharge or repeated outpatient activity. It may also be operating from an estate that is expensive to maintain and poorly suited to modern models of care.
This is why headline benchmarking often misleads. A trust may look expensive relative to a peer but have legitimate reasons, such as case mix, specialist roles, geography or estate constraints. Equally, some of the variation may reflect avoidable cost created by the way services and workforce are configured. The management task is to distinguish the legitimate from the improvable, not to turn every variance into a savings target. That is our interpretation.
What drives deficits
Demand, acuity and waiting-list pressure
Demand is not just about volume. Patients are often sicker, more complex and waiting longer by the time they reach treatment. That raises the cost of care and can reduce throughput, because the same capacity produces fewer completed pathways or takes longer to recover performance. The NAO found the NHS still dealing with large waiting lists and elevated pressure on urgent and emergency care while trying to improve performance.1
Workforce cost and workforce-market pressure
Workforce is both the largest input and the most common point of financial strain. Higher sickness absence, industrial action, recruitment difficulty and heavy use of temporary staffing all increase cost while disrupting continuity and productivity. The halving of agency spend in 2025/26 shows both that agency had become a major cost problem and that targeted control can materially affect the position.4
That does not mean every workforce problem is a workforce-size problem. Some are design problems. Roles may not match the work. Senior clinicians may spend too much time on tasks others could do safely. Capacity may exist in aggregate but not at the right time, in the right place or along the right pathway. That distinction between total input and usable capacity runs through our work on trapped value in healthcare.
Inflation and pay pressure
Inflation can turn an already difficult position into a deficit even when leaders are managing tightly. The NAO reported that non-pay inflation added £1.4 billion above budget in 2023/24.1 Pay settlements have the same effect at larger scale, because every additional percentage point on the pay bill lands on the largest single line in the accounts.
Productivity constraints
Productivity matters, but it is often discussed too loosely. More appointments or procedures do not automatically mean better productivity if quality falls, if case complexity rises, or if the extra activity does not change outcomes. Nor does released time automatically become cash. Productivity is better understood through value than raw output, and gains remain trapped when the rest of the pathway does not change with them.
National evidence supports the broader point. The NAO found NHS inputs rising faster than outputs, with productivity below pre-pandemic levels.1 NHS England has since reported improvement, with 2.7% in 2024/25 and 2.6% in the acute sector in the first half of 2025/26, against a Spending Review requirement of 2% a year.5 The lesson is not that productivity is irrelevant. It is that it cannot be treated as a simple substitute for structural recovery.
Capital, estate and infrastructure constraints
Financial discussions often underplay the effect of capital underinvestment. Ageing estates, poor digital infrastructure and maintenance backlogs increase day-to-day cost, constrain throughput and make service redesign harder. The NAO reported an NHS maintenance backlog of £11.6 billion in 2022/23.1 The Nuffield Trust also pointed to a further £400 million raid on capital budgets in 2023/24 to help fill workforce-cost gaps.2 That improves the in-year position at the cost of future productivity and resilience.
Service configuration and flow failure
Many deficits are created in the spaces between services, not within one budget line. Delayed discharge keeps beds full. Weak community alternatives push patients into acute settings. Fragmented pathways create repeated visits, duplicated work and avoidable handoffs. Capacity released in one part of the system often cannot be used because the next stage of the pathway remains blocked.
This is one reason provider-only savings plans often disappoint. A trust may reduce part of its own cost base only to recreate pressure through poorer flow, higher acuity, outsourced activity or transferred demand elsewhere in the system. It is also why the sustainability of individual services, and what makes a clinical service sustainable, sits underneath the financial question rather than alongside it.
Why savings plans often fail
The most common failure is to confuse identified opportunity with recoverable benefit. Benchmarking, productivity analysis or budget challenge may show that expenditure is higher than expected. That is useful, but it does not prove that the difference is locally controllable, that it can be removed safely, or that the cash effect will materialise inside the planning period.
A second failure is to rely on non-recurrent action. The NAO found that just over half of the £5 billion of efficiencies it examined in 2022/23 were non-recurrent.1 NHS England's own estimate of £3 billion of non-recurrent savings in 2023/24 indicates how much temporarily avoided expenditure would need to be found again the following year.2
| What is reported | What it measures | What it hides |
|---|---|---|
| Reported deficit | The position in the accounts after everything done to close the gap in year. | How much of the closure was one-off, and what has to be repeated next year. |
| Underlying deficit | Recurrent income against recurrent expenditure, stripped of one-off measures. | Little. This is the number that tells a board the size of the real problem. |
| Non-recurrent savings | Expenditure temporarily avoided or deferred. | That the cost base is unchanged, so the same saving must be found again. |
| Deficit support funding | Non-recurrent revenue support contingent on plan delivery. | That the reported position is better than the recurrent position underneath it. |
A third failure is to remove cost without removing work. When posts are cut but demand, clinical dependency and pathway design stay the same, the work returns through another route: agency staffing, waiting-list growth, overtime, operational instability or lower-quality decisions. This is exactly the difference between a finance plan that balances and an organisation that becomes more sustainable.
The Strasys perspective
Most financial recovery work starts in the middle of the problem. It looks at budgets, vacancies, procurement and activity, then asks what can be reduced. That is understandable, but it often misses the reason the organisation is spending at this level in the first place. We start earlier, with population need and with the operating model that converts resources into outcomes.
That changes the diagnosis. A workforce problem may begin in service design. A bed problem may begin in discharge pathways or community capacity. A deficit may be sustained by activity that no longer represents the best use of clinical time. Variation between organisations, services or pathways is therefore a signal to investigate, not a verdict. Reconstructing the whole decision from evidence held across separate functions is what we mean by Decision Intelligence in healthcare, and the Strasys Value Index is built to locate where that conversion is weakest.
This perspective also sharpens financial judgement. If a plan assumes fewer beds, fewer clinics or fewer posts, it should also explain which work will stop, which pathway will change, what assumptions are being made about demand, and where displaced activity will go. Otherwise the saving exists in the spreadsheet but not in the service. That is the core practical test of whether a recovery plan is credible.
What leaders should consider
Boards should ask a harder set of questions than the standard finance paper usually provides.
- What part of the deficit is recurrent, and what part reflects one-off shock or timing?
- Which pressures are driven by demand, workforce, inflation, capital, flow or service design?
- Where is the organisation paying more because the operating model is unstable?
- Which opportunities are only theoretical or identified, and which are genuinely accessible and recoverable?
- What work will stop, change or move if a saving is to be real?
- What dependencies sit outside the trust, in community services, commissioning choices, social care or wider system flow?
- How will the board know that a claimed benefit has been delivered and sustained rather than merely recorded in plan?
These questions matter because financial sustainability is not simply the absence of overspend. It is the ability to keep meeting need safely and credibly without recreating the same financial gap each year.
Evidence, limitations and sources
Established external evidence
The National Audit Office has shown that NHS financial sustainability is being shaped by demand pressure, workforce shortages, inflation, weak productivity recovery, reliance on non-recurrent savings and capital constraints. It has also warned that short-term financial action can crowd out the transformation needed for durable stability.1
Nuffield Trust analysis of provider accounts reported a £1.2 billion adjusted overspend in 2023/24 and an underlying provider-sector gap of at least £4.5 billion.2 NHS England's own planning and outturn reporting shows balanced plans depending on revenue support, very high efficiency requirements, and a national position that balanced in 2025/26 while system deficits persisted.34
Strasys interpretation
The explanation in this article uses a conversion chain from population need to resources, operating model, activity and outcomes. That chain, and the view that many deficits are sustained by misalignment within it, are our interpretation rather than a formal NHS accounting definition.
The related distinction between theoretical, identified, accessible, recoverable and realised opportunity is also our management interpretation. It is useful because it stops attractive opportunity numbers from being confused with what can actually be delivered and sustained.
Limitations
This article does not attribute a share of sector-level deficits to particular drivers, because the published evidence does not support that split. Nor does it claim a realised financial outcome from our own work. Where we describe causes, we are describing a way of reading the evidence, not a measured decomposition.