We need to talk about opportunity cost

The trade-offs are already happening. The question is whether we can see them.

As a medical student at Queen’s in the 1990s, on a renal attachment at Belfast City Hospital, I heard a consultant describe the constraints around who could be offered dialysis. Age was a factor. It was the first time I can remember hearing rationing named in clinical practice.

That unit had been set up by Dr Mollie McGeown. In 1990 she published Prevalence of advanced renal failure in Northern Ireland in the BMJ, as part of a Renal Association programme to estimate unmet need for dialysis and transplantation. She had built a new service and was arguing to expand it. But she also named what that expansion would displace: “More resources for dialysis mean fewer for geriatrics, mental diseases, etc. Rational planning demands accurate definition of the size of the problem.”

What McGeown was describing was opportunity cost: resources used for one purpose cannot be used for another. Applying it means looking beyond the service, treatment or pathway in front of us and asking what becomes harder to provide when that service expands. When a service shrinks, we should ask what value is being lost, and whether that loss is justified by the alternative use of the resource.

It would be easy to treat the renal example as belonging to an older, harsher world. But the problem has only grown. Health now takes a larger share of public spending, and the Department of Health and Social Care’s day-to-day budget rose from 26% of all departmental day-to-day spending in 1998–99 to 43% in 2022–23. In a constrained public budget, more for health means less for other public services, including some that also produce health but sit outside the NHS.

Some of that growth is demographic. Some of it reflects medicine’s own expansion: more conditions are treatable, more tests and drugs are available, more monitoring is expected, and more people are eligible for intervention. Each expansion draws on money, workforce, beds, diagnostic capacity, organisational attention and public tolerance. Opportunity cost is the discipline of asking what those claims might replace.

In Wales in 2011, I took part in a national Programme Budgeting and Marginal Analysis exercise on health improvement spending. PBMA is designed to make opportunity cost practical. It starts with existing activity and expenditure, then asks what could be improved, reduced or stopped, and where released resources could do more good. The Welsh exercise reviewed a defined budget and recommended disinvestment in some initiatives, with child health, mental health and wellbeing, and tobacco control emerging as priorities for investment.

Requirements for reallocation in the NHS, 2015. Photo: Anne Marie Cunningham. Slide: Gwyn Bevan / LSE, Hellish Decisions in Healthcare conference, 2015.

A few years later, I took a photograph of a slide by Gwyn Bevan at Muir Gray’s Hellish Decisions in Healthcare conference at Oxford. The slide was titled “Requirements for re-allocation to improve value in NHS”. The point was that reallocation is not only analytical. It also needs political support, stakeholder engagement, and managerial effort. In other words, opportunity cost can be described on paper, but acting on it is a political and organisational task.

That is where the difficulty really begins. Some forms of healthcare are easier to describe in programme terms than others. McGeown’s renal service had a defined population, a clear intervention and visible unmet need. Drugs, procedures and condition-based pathways usually come with prices, codes, activity data and waiting times. Those features make their claims easier to compare, defend and expand.

Care organised around people rather than conditions is harder to see. Primary care, frailty care, district nursing, social care, rehabilitation, continuity and coordination do not sit neatly inside one disease programme. Palliative care is a good example. Its value may lie in maintaining dignity, symptom control, support for families, fewer crisis admissions and enabling someone to die at home rather than in hospital. Often that work is most noticeable when it is absent.

The same is true of other forms of relational and preventive care. Their value is cumulative and often counterfactual: fewer tests, safer care, supported carers, feeling known, more care delivered at home. But the evidence needed to turn that value into a funding decision is often weak. There may be no single budget, no comparable activity data, no agreed outcomes, and no institution responsible for counting the work as a whole.

This matters most under constraint. When there is little or no new money, and probably less money, commissioning is no longer mainly about expansion. It is about deciding what is redesigned or what stops. If a service is reduced, the question is not only how much money is saved, but what capacity, relationship, skill or outcome is being lost.

The form of funding affects whether opportunity cost is tested at all. Block contracts and core provider budgets can hide programme-level opportunity cost because costs are bundled together and marginal savings are hard to release. By contrast, contracts, grants, pilots and enhanced schemes are administratively easy to remove, so they are often the first place a system goes when it needs to find savings - a contract is not renewed, an uplift falls behind costs, a specification changes or a pilot simply runs out.

Some of these decisions may be defensible under scrutiny. A service may be ineffective, outdated or no longer the best use of scarce resource. But cutting a service because it is easy to cut is different from cutting it because the alternatives have been considered. Making trade-offs explicit does not make them easy, and it does not remove disagreement about what counts as value. But without explicitness, disagreement is displaced into subtler forms of rationing: delay, attrition, unfunded workload and services allowed to fail.

This is why the current language of reform needs a clearer account of resources and trade-offs. We talk about shifting care from hospital to community, from treatment to prevention, from reactive care to proactive care. The practical question is how that shift is translated into resource decisions.

We need to talk about opportunity cost not as an abstract economic principle, and not as a euphemism for cuts, but as a way of making choices explicit. Losing care that creates more value than the intervention, or pathway, that continues to be funded is a bad outcome. And that becomes more likely, and harder to correct, when the reasoning is opaque. A system that cannot explain what it is losing, what it is protecting, and why, is not making trustworthy decisions.