Externality

Externality in Health Economics

In economics, an externality is a side effect of producing or consuming something that lands on people who took no part in the decision and are not compensated for it. A negative externality imposes an external cost, as when heavy antibiotic use speeds the spread of antimicrobial resistance; a positive externality confers an external benefit, as when vaccination protects unvaccinated people by reducing transmission. Because the person deciding does not weigh these third-party effects, an unregulated market tends to supply too much of activities with external costs and too little of activities with external benefits, which is why externalities are a classic source of market failure. In an economic evaluation, whether such effects are counted depends on the perspective taken and, for infectious disease, on whether the model captures transmission.

How private decisions create social effects

Individuals and organisations usually make choices using the costs and benefits they experience directly. An externality exists when part of the effect falls on third parties and is therefore omitted from that private calculation. This creates a difference between the private incentive facing the decision-maker and the activity's overall effect on society.

• A negative externality imposes an uncompensated cost on third parties.
• A positive externality provides an uncompensated benefit to third parties.
• An externality can arise from either the production or consumption of a good or service.
• An externality is not simply any indirect effect; it must affect parties outside the relevant decision and lie outside the costs and benefits that the decision-maker weighs.

The difference between private and social value

Health economists represent an externality by separating effects experienced by the decision-maker from effects experienced by others. Social cost or benefit includes both components. This distinction shows why a market price or private choice may not reflect the full consequences of an activity.

For costs, the relationship is as follows.

Marginal social cost = Marginal private cost + Marginal external cost

MSC = MPC + MEC

For benefits, the corresponding relationship is as follows.

Marginal social benefit = Marginal private benefit + Marginal external benefit

MSB = MPB + MEB

The terms are defined as follows.

• MPC is the additional cost borne directly by the producer or consumer.
• MEC is the additional cost imposed on third parties.
• MSC is the combined additional cost to the decision-maker and third parties, that is, to society as a whole.
• MPB is the additional benefit received directly by the producer or consumer.
• MEB is the additional benefit received by third parties.
• MSB is the combined additional benefit to the decision-maker and third parties, that is, to society as a whole.

An efficient quantity is reached where marginal social benefit equals marginal social cost, as follows.

MSB = MSC

This condition differs from a private market decision when an external cost or benefit is omitted from the incentives facing buyers or sellers. By convention, production externalities are shown on the cost side and consumption externalities on the benefit side. An external benefit from production therefore lowers marginal social cost below marginal private cost, and an external cost from consumption lowers marginal social benefit below marginal private benefit.

Why externalities can cause inefficient outcomes

A negative externality can lead to too much of an activity because the decision-maker does not bear its full social cost. A positive externality can lead to too little because the decision-maker does not receive its full social benefit. In either case, the privately chosen quantity can differ from the quantity that maximises net social benefit.

The direction and size of an externality must be established from evidence. A socially beneficial activity is not automatically a positive externality, and a harmful activity is not automatically a negative externality.

Why externalities matter in health economics

Health-related decisions frequently affect people beyond the patient, purchaser or provider directly involved. Infectious-disease transmission, antimicrobial resistance, environmental exposure, unpaid caregiving and the spread of medical knowledge can all create effects outside the immediate transaction. Whether these effects count as externalities depends on how the decision problem, perspective and included costs and outcomes are defined.

Culyer's analysis of health care as a commodity separated two kinds of externality. Physical externalities arise with communicable disease, where one person's failure to be immunised raises the probability that others become infected. Caring externalities arise because people care about the health care that others receive, so one person's consumption of care enters another person's wellbeing. Culyer used this caring externality to analyse the case for collective financing of health care, while noting that the existence of a marginal externality does not on its own justify further subsidy.

Excluding externalities can change several parts of an analysis.

• The estimated costs and benefits of an intervention.
• The quantity of a health service that appears desirable.
• The comparison between private and publicly funded provision.
• The economic justification for taxes, subsidies, regulation or collective action.
• The distribution of costs and benefits across affected groups.

An effect should not be counted twice. If an economic evaluation already includes a consequence within its stated perspective, that consequence remains socially relevant but is no longer omitted from that particular analysis.

Externalities in economic evaluation, perspective and transmission

In an economic evaluation, the chosen perspective decides which third-party effects are counted, so an effect that is external to one perspective may be internal to another. The reference case in the NICE manual (PMG36, last updated 13 October 2026) counts health effects for patients and, when relevant, other people (mainly carers), while costs are those of the NHS and personal social services. Benefits to other government bodies, such as reduced crime from treatments for drug misuse, are considered only in non-reference-case analyses agreed with the Department of Health and Social Care.

Transmission is the clearest health example of a spillover that the model structure itself can omit. A static model treats each person's risk of infection as fixed, so it counts only the direct protection of those vaccinated. A dynamic transmission model lets the force of infection fall as coverage rises, so it also captures the indirect protection of unvaccinated people through herd immunity. Brisson and Edmunds compared the two approaches to illustrate why these herd-immunity externalities matter when assessing the cost-effectiveness of a vaccination programme.

Antimicrobial resistance shows the opposite problem. Coast, Smith and Millar argued that resistance is an important social externality of antimicrobial use that economic evaluations comparing strategies for infectious disease had generally left out as a cost. They found no methodological justification for the omission and concluded that the practical difficulty of measuring the cost was the main reason it was likely to go on being ignored.

Illustrative worked example of vaccination and transmission

Vaccination can provide a direct benefit to the vaccinated person and an external benefit to other people through reduced transmission. The external benefit strengthens the social case for vaccination, but its size depends on disease epidemiology, vaccine performance, coverage and population contact patterns.

Suppose, using illustrative values, that the marginal private benefit of vaccination is valued at 55 units and the estimated marginal external benefit from reduced transmission is 20 units.

MSB = MPB + MEB

MSB = 55 + 20 = 75 units

Assume that vaccination imposes no external cost, so its marginal social cost equals its marginal private cost of 60 units, and that the individual bears this cost in full. The marginal net social benefit is then calculated as follows.

Marginal net social benefit = MSB − MSC

Marginal net social benefit = 75 − 60 = 15 units

The individual, however, compares only the private benefit with the private cost.

Marginal net private benefit = MPB − MPC

Marginal net private benefit = 55 − 60 = −5 units

In this illustrative case, the individual declines vaccination because the private return is negative, even though society gains 15 units from the extra vaccination once the transmission benefit is included. A subsidy of more than 5 units at this margin would reverse the private decision. The calculation does not establish the optimal vaccination programme on its own; a real analysis would require evidence on population effects, uncertainty, implementation costs and how the external benefit changes as coverage increases.

How externalities are identified and measured

External effects are often more difficult to observe than direct costs and outcomes because they occur outside the original transaction. Analysts must specify who is affected, how the effect occurs, which comparison is relevant and whether the consequence has already been captured elsewhere in the analysis.

A practical assessment can proceed in seven steps.

1. Define the activity and comparator. State which production, consumption or policy decision is being assessed.
2. Identify the decision-makers and directly affected parties. Establish which costs and benefits enter their private calculation.
3. Identify affected third parties. Specify who experiences consequences without being fully represented in the original decision.
4. Describe the causal pathway. Explain how the activity produces the external cost or benefit.
5. Measure the additional effect. Estimate the difference in third-party outcomes attributable to the activity compared with the relevant alternative.
6. Value the effect where appropriate. Apply monetary values, health outcomes or other measures consistent with the analytical framework.
7. Test uncertainty and potential double counting. Examine alternative assumptions and confirm that the same consequence has not been included more than once.

Evidence may come from epidemiological models, transmission models, environmental studies, administrative data, observational research, trials or economic valuation methods. The appropriate evidence depends on the pathway through which the externality occurs.

Tobacco and alcohol show how much the answer depends on what is counted as external. Manning and colleagues estimated the lifetime, discounted costs that smokers and drinkers in the United States impose on others through collectively financed health insurance, pensions, disability and group life insurance, fires, motor-vehicle accidents and the criminal justice system. Non-smokers subsidised smokers' medical care and group life insurance while smokers subsidised non-smokers' pensions and nursing home payments, and on balance the authors judged that smokers probably paid their way at the excise taxes then in force, whereas alcohol taxes covered only about half of the costs that drinkers imposed on others. Harm that smokers and drinkers bear themselves is a private cost in this framework, so it does not by itself justify a corrective tax.

How policy can internalise an externality

A policy internalises an externality when it changes incentives or decision-making so that external effects are taken into account. The appropriate response depends on the cause of the externality, the available evidence, administrative feasibility, distributional consequences and the risk that intervention creates additional costs or distortions.

Pigou described externalities as divergences between marginal social net product and marginal private net product, and argued that the State could remove such a divergence through bounties and taxes. A Pigouvian tax set equal to the marginal external cost at the efficient quantity makes the decision-maker face the full social cost, and a Pigouvian subsidy set equal to the marginal external benefit does the same for the social benefit.

Coase challenged the Pigouvian approach. He treated harm as reciprocal, since restraining the party who causes it harms that party, and showed that if market transactions were costless, bargaining between the parties would reach the same value-maximising outcome whatever the initial assignment of legal rights. The limit is the assumption itself. Once bargaining is costly, the assignment of rights affects the outcome, and many health externalities involve very large numbers of people, which makes private bargaining impractical. Culyer made a related point for caring externalities, where each person has an incentive to leave charitable provision to others, so private action can leave the externality uninternalised.

The existence of an externality does not prove that a particular intervention will improve welfare. The expected reduction in external harm or increase in external benefit must be weighed against policy costs, unintended consequences, implementation constraints and equity effects.

Externality, spillover and market failure are related but distinct

These terms overlap, but they should not be treated as interchangeable. Keeping them separate helps analysts describe the mechanism precisely and avoid labelling every effect outside the patient as a market failure.

• An externality is an uncompensated third-party cost or benefit omitted from the relevant private decision.
• A spillover effect is a broader term for an effect extending beyond the person or organisation directly targeted; a spillover is not necessarily an externality.
• A market failure is a broader condition in which market allocation is inefficient; externalities are one possible cause.
• A distributional effect concerns who gains or loses and may matter even when resource allocation is efficient.
• A within-family effect, such as a caregiver outcome, may be included directly in the analysis rather than treated as an omitted externality; the NICE reference case, for example, counts health effects on carers where relevant.

Common analytical errors

Externalities can be overstated when analysts classify every wider consequence as external or assume that a desirable activity must create an external benefit. They can also be understated when the analytical perspective excludes real third-party consequences without making that limitation visible.

Several errors recur in practice.

• Calling a consequence an externality without identifying the affected third party.
• Treating an effect already included in the analysis as an additional external benefit or cost.
• Assuming that a private benefit automatically produces an equal social benefit.
• Using average external effects when the decision requires marginal effects.
• Ignoring how the external effect changes as activity or coverage increases.
• Treating evidence of an externality as automatic proof that government intervention will produce a net benefit.
• Combining efficiency and equity arguments without explaining which concern supports the proposed policy.

Sources

• Brisson M, Edmunds WJ. Economic evaluation of vaccination programs: the impact of herd-immunity. Medical Decision Making. 2003;23(1):76-82. doi:10.1177/0272989X02239651
• Coase RH. The problem of social cost. Journal of Law and Economics. 1960;3:1-44.
• Coast J, Smith RD, Millar MR. Superbugs: should antimicrobial resistance be included as a cost in economic evaluation? Health Economics. 1996;5(3):217-226.
• Culyer AJ. The nature of the commodity 'health care' and its efficient allocation. Oxford Economic Papers. 1971;23(2):189-211.
• Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th edition. Oxford: Oxford University Press; 2015.
• Manning WG, Keeler EB, Newhouse JP, Sloss EM, Wasserman J. The taxes of sin: do smokers and drinkers pay their way? JAMA. 1989;261(11):1604-1609.
• National Institute for Health and Care Excellence. NICE technology appraisal and highly specialised technologies guidance: the manual (PMG36). Published 31 January 2022, last updated 13 October 2026.
• Pigou AC. The Economics of Welfare. London: Macmillan; 1920 (4th edition, 1932, consulted for Part II, Chapter IX).