Market Failure

Market Failure in Health Care and Health Insurance

Market failures in health care are the ways in which markets for medical care and health insurance, left to themselves, miss the efficient allocation of resources. The main sources are uncertainty about illness and treatment, unequal information between clinicians and patients, externalities from communicable disease, public goods such as disease surveillance and research, adverse selection and moral hazard in insurance, and market power among hospitals and professions. Health economists use the diagnosis, which builds on Arrow's 1963 account of medical care, to decide which policy instrument, if any, fits the problem, because each failure calls for a different response. Government failure is the counterweight, since an intervention can create inefficiency of its own.

The efficiency benchmark behind the term

A market fails in the economic sense when its outcome differs from an efficient one, even if it trades actively and stays profitable. Arrow's 1963 paper on the welfare economics of medical care takes as its benchmark the first optimality theorem, under which a competitive equilibrium in which every commodity affecting costs or utilities has a market price is Pareto optimal. Market failure is the case where these conditions break and the market fails to reach a Pareto efficient state. The theorems and their conditions are set out on the welfare economics page.

At the margin, the efficient quantity of a service is where marginal social benefit equals marginal social cost. A market responds to private benefits and costs, so a failure appears as a wedge between the private and social values, set out in the two identities below.

$$MSB = MPB + MEB \qquad MSC = MPC + MEC$$

where $MSB$ and $MSC$ are marginal social benefit and cost, $MPB$ and $MPC$ are the marginal private benefit and cost that buyers and sellers act on, and $MEB$ and $MEC$ are marginal external benefit and cost falling on people outside the transaction. Externalities create the wedge directly; information problems and market power instead distort the private values or the quantity supplied.

Arrow 1963 on uncertainty and unequal information in medical care

Kenneth Arrow traced the special economic features of medical care to uncertainty, both about who falls ill and about whether treatment works. Demand for care is irregular and unpredictable, since medical services apart from preventive services give satisfaction only in illness, and uncertainty about the quality of the product is perhaps more intense than for any other important commodity.

The second strand is information asymmetry. Arrow wrote that the physician's information about the consequences and possibilities of treatment is necessarily much greater than the patient's, and that both parties know it. The patient cannot judge the product as the competitive model expects a buyer to, so the doctor becomes the patient's agent, a relationship analysed as a principal-agent problem.

Arrow's interpretation was that when the market fails to reach an optimal state, society recognises the gap and non-market institutions arise to bridge it. Trust in the physician, an ethical code under which advice is supposed to be governed by concern for the patient's welfare, and licensing all fit this reading. Licensing has a cost. Arrow observed that it restricts supply and so raises the cost of care, and is defended as guaranteeing a minimum of quality.

Supplier-induced demand as an agency concern

Because the same clinician often advises on and supplies treatment, the agency relationship can be abused. Johnson's 2014 review attributes the concept of supplier-induced demand to Evans in 1974 and adopts McGuire's definition of it as a physician influencing a patient's demand for care against the physician's own interpretation of the patient's best interests. It reported convincingly identified studies showing that physicians are not perfect agents, for example obstetricians performing more caesarean sections as fertility declined, but concluded that more work is needed before the economic importance of induced demand can be stated. The idea is often used to argue for changes to provider payment systems.

Externalities and public goods that prices do not capture

An externality arises when the market does not make a person pay for costs imposed on others or reward benefits conferred on them. Arrow's medical example was communicable disease. A person who declines immunisation weighs the risk to their own health against the benefit of avoiding the procedure, but not the risk to others. He argued that a price system paying or charging each affected person could not be practical, so approaching an optimal state needs collective intervention through subsidy, tax or compulsion. How indirect protection is valued in evaluations of immunisation programmes is covered on the vaccination externalities page; the instruments are covered on the Pigouvian subsidy and Pigouvian tax pages.

Public goods are the limiting case. Smith and MacKellar define pure public goods as non-rival, not diminished by use, and non-excludable, freely available to all once produced, and note that the free market provides less than the collectively optimal level of them. The state can produce the good itself or arrange for a private firm to produce it. They also stress that health itself is a private good, as are most goods and services used to produce it. Their example separates the two cases. HIV treatment is rival and excludable, while prevention through media campaigns and testing is non-rival and non-excludable. Following Smith and colleagues, they apply the global public good concept to two areas of health, research and development on the one hand and communicable disease control, including surveillance and immunisation, on the other.

Missing insurance markets, adverse selection and moral hazard

Uncertainty creates a demand for health insurance, and Arrow called the welfare case for insurance overwhelming, concluding that government should provide insurance where the market has failed to emerge. Insurance markets then face two failures of their own.

Adverse selection arises when buyers know more about their risk than the insurer. Arrow saw that, in a fully competitive market, plans charging lower premiums to preferred risks would draw them off and leave a plan that pooled all risks with an adverse selection of them. Akerlof showed, with medical insurance for people over 65, that adverse selection can leave no sales at any price, and treated this as an argument for Medicare; the adverse selection page covers the mechanism and the responses.

Moral hazard arises after cover is bought. Arrow noted that spending depends on the doctor's choices as well as on the illness, and that widespread insurance raises the demand for care. Because insurance removes the incentive to shop for better prices, Arrow argued that market forces tend to be replaced by direct institutional control. Insurance is therefore both a response to one failure and the source of others; the design trade-offs are on the cost-sharing page.

Market power among hospitals and professions

Market power lets a seller raise price or lower quality without losing enough patients to make it unprofitable. The modern evidence concerns hospital consolidation. In 2019 testimony to a US House antitrust subcommittee, Martin Gaynor summarised research showing that consolidation between close competitors leads to substantial price increases for hospitals, insurers and physicians without offsetting gains in quality or efficiency, with increases of 20 or 30 per cent common after hospital mergers in concentrated markets and some as high as 65 per cent. He reported that not-for-profit hospitals exploit market power as much as for-profits. Market power matters under regulated prices too. After the English NHS reforms of 2006, which were intended to increase patient choice and hospital competition and introduced administered prices, risk-adjusted heart attack mortality fell more at hospitals in less concentrated markets. Among his proposals were stronger antitrust enforcement and greater transparency on costs and quality.

Illustrative worked example of a hospital merger and the price of a procedure

The figures are illustrative and not taken from any study. Two hospitals in one area each charge an insurer GBP 6,000 for a hip replacement, equal to its marginal cost. After they merge, the insurer has no close alternative and agrees a price of GBP 7,500. The Lerner index measures the mark-up as a share of price, as follows.

$$L = \frac{P - MC}{P} = \frac{7500 - 6000}{7500} = 0.20$$

where $L$ is the Lerner index, $P$ is the price per procedure and $MC$ is its marginal cost. The mark-up is GBP 1,500, 25 per cent above cost. Suppose the insurer funded 1,000 procedures before the merger and funds 900 after it. On those 900 it pays GBP 6.75 million instead of GBP 5.4 million; the extra GBP 1.35 million is a transfer from premium payers to the hospital, not extra resource use. If the 100 patients dropped valued surgery evenly between GBP 6,000 and GBP 7,500, the deadweight loss is 0.5 × 100 × GBP 1,500 = GBP 75,000. An evaluation of a falls-prevention programme that avoids 40 such operations would count GBP 300,000 of savings at the merged price, while the resources released are worth GBP 240,000, so from a societal perspective the savings are overstated by GBP 60,000.

Matching the response to the diagnosed failure

The diagnosis matters because a remedy for one failure can be useless or harmful against another. The table pairs each failure with the response the sources above support and a cost of that response.

In economic evaluation and health technology assessment the same reasoning decides what to count. An evaluation of a vaccine that ignores transmission misses the externality, and a price inflated by market power can overstate the opportunity cost of the resources a technology uses.

Government failure as the counterweight

A market failure shows that the market outcome can be improved. It does not show that a given intervention will improve it. Arrow said as much, noting that the institutions that arise to bridge the gap are not uniformly successful in moving closer to optimality once all consequences are counted, and licensing, one of the institutions he described, raises costs.

Le Grand's theory of government failure runs parallel to market failure. Building on public choice work that assumes all relevant agents pursue their self-interest, it examines provision, subsidy and regulation and concludes that each may create inefficiency and inequity, in forms that differ by activity. Le Grand emphasised that whether government failure is larger or smaller than the market failure it addresses is an empirical question. Coase made the same point about externalities, arguing that the proper comparison is the total social product of the alternative arrangements, starting from the situation that actually exists and counting the costs of operating a market or a government department. The Coase theorem page covers his bargaining argument.

Common misreadings of market failure in health care

The term is often stretched to cover any outcome someone dislikes. The following errors recur in policy debate and in teaching.

• Inequity is not market failure. An efficient market can produce a distribution society rejects. Distributional aims can justify action on their own terms and are better stated openly; the equity-efficiency trade-off page sets out how they are weighed.
• Health is not a public good. Health itself, and most of the goods and services that produce it, is a private good. The public good argument applies to specific activities such as surveillance and research.
• A high price or unmet need is not proof. The analysis must name the mechanism that makes price or quantity inefficient.
• Unequal information alone is not proof. Information gaps matter when they change behaviour or outcomes, and the size of that effect has to be measured.
• Market failure does not settle the policy. It opens the comparison with feasible alternatives, including their own failures.

Sources

• Arrow KJ. Uncertainty and the welfare economics of medical care. American Economic Review. 1963;53(5):941-973.
• Akerlof GA. The market for "lemons": quality uncertainty and the market mechanism. Quarterly Journal of Economics. 1970;84(3):488-500. https://doi.org/10.2307/1879431
• Coase RH. The problem of social cost. Journal of Law and Economics. 1960;3:1-44. https://doi.org/10.1086/466560
• Gaynor M. Diagnosing the problem: exploring the effects of consolidation and anticompetitive conduct in health care markets. Statement before the Subcommittee on Antitrust, Commercial, and Administrative Law, Committee on the Judiciary, US House of Representatives. Washington, DC; 7 March 2019.
• Johnson EM. Physician-induced demand. In: Culyer AJ, ed. Encyclopedia of Health Economics. Vol 3. San Diego: Elsevier; 2014:77-82. https://doi.org/10.1016/B978-0-12-375678-7.00805-1
• Le Grand J. The theory of government failure. British Journal of Political Science. 1991;21(4):423-442. https://doi.org/10.1017/S0007123400006244
• Smith RD, MacKellar L. Global public goods and the global health agenda: problems, priorities and potential. Globalization and Health. 2007;3:9. https://doi.org/10.1186/1744-8603-3-9