Marginal Analysis

Marginal analysis helps decision-makers examine what would happen if the level of a healthcare activity increased or decreased by a defined amount. This page explains how marginal changes are measured, how they guide choices within a limited budget, and why marginal results can differ from averages or totals.

Why healthcare decisions should be made at the margin

Most resource-allocation decisions are not choices between providing an entire service and providing nothing. They concern whether to add, remove or reallocate a specific amount of activity, such as funding another clinic session, expanding vaccination coverage or reducing a low-value service. Marginal analysis focuses on the additional costs and benefits created by that change.

A programme may appear valuable when judged by its total or average results but offer little benefit from further expansion. Conversely, a smaller programme may produce substantial additional health from its next increment of funding. Marginal analysis therefore directs attention to the consequences of doing a little more or a little less.

How marginal changes are measured

A marginal value describes the change in a cost, benefit or outcome associated with a defined change in activity. When the change is small but discrete, the marginal value can be estimated using the difference between the current and proposed levels.

Marginal cost per additional unit of activity is:

Marginal cost = ΔC / ΔQ

where:

• ΔC is the change in total cost.
• ΔQ is the change in the quantity of activity or output.

Marginal benefit per additional unit of activity is:

Marginal benefit = ΔB / ΔQ

where:

• ΔB is the change in benefit.
• ΔQ is the change in the quantity of activity or output.

The units must always be stated. Marginal cost might be measured in pounds per additional patient treated, while marginal health benefit might be measured in QALYs gained per additional patient. Values expressed in different units cannot be compared directly without an additional valuation rule or decision framework.

When expanding or reducing a service improves allocation

In a simple setting where both benefits and costs are expressed in money, expansion may continue while marginal benefit exceeds marginal cost. The familiar condition MB = MC describes a theoretical optimum when changes are divisible, all relevant consequences are included and no additional constraints alter the decision.

Health benefits are often measured in natural units or QALYs rather than money. In those settings, decision-makers may use a cost-effectiveness threshold, opportunity-cost estimate or another explicit decision rule to compare the additional health produced with the additional resources required.

Incremental net monetary benefit can be calculated as:

INMB = (λ × ΔE) − ΔC

where:

• λ is the monetary value or decision threshold applied to one unit of health effect.
• ΔE is the difference in health outcomes between the proposed change and its comparator.
• ΔC is the difference in cost between the proposed change and its comparator.

A positive incremental net monetary benefit indicates that the additional health is valued more highly than the additional cost under the selected threshold. This result supports a decision but does not remove the need to consider uncertainty, affordability, implementation constraints, equity and the care displaced elsewhere.

How marginal analysis works within a fixed healthcare budget

When the available budget is fixed, funding one expansion means that another use of resources may have to be reduced or forgone. Marginal analysis compares the health gained from the proposed expansion with the health that could have been produced by the best displaced alternative. This makes opportunity cost central to the decision.

For divisible activities with comparable outcomes, an efficient allocation tends to direct the next unit of funding towards the available activity producing the greatest marginal benefit per unit of resource. Resources can then be reconsidered as programmes expand and their marginal returns change. In practice, capacity limits, minimum service levels, indivisible investments and other constraints may prevent exact equalisation across services.

Worked example: reallocating a limited budget

A local health system has £80,000 available for service expansion. Three programmes each offer one additional £40,000 expansion step, but the expected health gain differs across the programmes.

Expansion step	Additional cost	Additional QALYs	QALYs per £40,000
Diabetes outreach	£40,000	8	8
Vaccination programme	£40,000	6	6
Mental health follow-up	£40,000	5	5

Because the budget can fund only two expansion steps, the marginal comparison favours diabetes outreach and the vaccination programme in this simplified example. Together, those expansions are expected to produce 14 additional QALYs, compared with 13 QALYs from choosing diabetes outreach and mental health follow-up.

If £40,000 had already been assigned to the mental health expansion, reallocating that amount to vaccination would exchange 5 expected QALYs for 6 expected QALYs. The marginal reallocation would therefore produce an expected net gain of 1 QALY without increasing the total budget.

This example is illustrative rather than a complete healthcare decision. A real analysis would also examine uncertainty, feasibility, distributional effects, service dependencies and whether the estimated gains are comparable across programmes.

How marginal analysis differs from related measures

Marginal, average and incremental measures answer related but different questions. Confusing them can lead to an incorrect conclusion about whether a service should expand, contract or remain unchanged.

Measure	Question answered	Typical calculation
Average cost	What is the cost per unit across the activity as a whole?	Total cost ÷ total quantity
Marginal cost	What additional cost results from a defined increase in activity?	Change in total cost ÷ change in quantity
Marginal benefit	What additional benefit results from a defined increase in activity?	Change in total benefit ÷ change in quantity
Incremental cost-effectiveness ratio	What additional cost is associated with one additional unit of health effect when two alternatives are compared?	Difference in cost ÷ difference in effect

An incremental cost-effectiveness ratio is not automatically a marginal cost. It compares the cost and effect differences between specified alternatives, while marginal analysis focuses on the consequences of changing the level of an activity. The concepts overlap when the alternatives represent successive changes in scale, but they should not be treated as interchangeable.

Practical limitations and safeguards

Reliable marginal analysis requires information about how costs and outcomes change at the relevant level of activity. Healthcare data are often reported as totals or averages, making the consequences of a small expansion or contraction difficult to estimate. Results may also change when a service reaches a capacity limit or requires a large, indivisible investment.

Decision-makers should therefore:

• Define the current activity level and the proposed marginal change clearly.
• State whether the change is an expansion, contraction or reallocation.
• Use costs and outcomes that correspond to the same change and time period.
• Report the units used for every marginal value.
• Avoid comparing monetary costs directly with non-monetary benefits without an explicit decision rule.
• Identify the service or activity likely to be displaced when the budget is constrained.
• Test uncertainty around costs, outcomes, capacity and implementation.
• Consider equity, feasibility and service dependencies alongside the efficiency result.

Marginal analysis is most useful when realistic alternatives and consequences can be specified. It informs resource allocation by revealing the value of the next change, but it does not by itself determine which objectives, populations or outcomes should receive priority.