Note: Rationing of health care services is a contentious topic and in our view, current discussions of rationing often are unproductive or harmful. We suggest two reasons for that result. First the discussants are imprecise in their use of the term rationing – often for politically motivated reasons. Second, the discussants write about rationing as an activity that “we” will undertake, with significant consequences for “you.” It would be more helpful to discuss the prudent use of health care services in terms of the health plan that we would want for ourselves and our families.

In part 1 of this post below, we address the first problem by offering a more precise definition of rationing. In part 2 of this post [which will appear on Health Affairs Blog tomorrow], we address the second problem by describing the type of health plan we would like for ourselves and our families.

Editor’s note: You can also read a response to these posts by Uwe Reinhardt.

“Rationing,” and particularly “government rationing” is politically charged rhetoric. Merely mentioning the possibility of government rationing of health care services has a chilling effect on health policy debates. For example, some of the early criticisms of comparative effectiveness research (CER) were based on the possibility that the research would be used as a basis for coverage decisions in public insurance plans, especially Medicare. During the health care reform debate in September, 2009, Senator Orrin Hatch remarked:

We don’t want to turn health care over to a bunch of bureaucrats in Washington, who then will determine what kind of health care we have … (a)nd you know that rationing is going to happen.

Concerns over a possible link between CER and rationing were reflected in language included in the American Recovery and Reinvestment Act of 2009 CER funding legislation (H.R.1,pp. 62-64, 73-74), which prohibits the Federal Coordinating Council for Comparative Effectiveness from construing any of the reports or recommendations made by the Council as “mandates or clinical guidelines for payment, coverage, or treatment.”  The Patient Protection and Affordable Care Act of 2010 (PPACA, p. 621) contained similar provisions.

Be that as it may, Martin Feldstein’s Wall Street Journal editorial in 2009 was entitled “ObamaCare Is All About Rationing: Overspending is far preferable to artificially limiting the availability of new procedures and technologies” and Scott Gottleib of the American Enterprise Institute expressed similar sentiments in the Health Care News. These skeptical views of CER have support in the general population. Carman, et al., found that:

To the extent that consumers perceive that the application of comparative effectiveness research to decision making could limit their choice of providers, inappropriately interfere with physicians’ recommendations for treatment, or appear to “ration” care based on cost, the efforts will encounter consumer resistance and could lead to a broad consumer backlash.

Despite public distaste for rationing, the term remains remarkably resilient. Steven Rattner, former Obama Administration Treasury Secretary counselor, recently opined that, “unless we start allocating health care resources more prudently — rationing, by its proper name — the exploding cost of Medicare will swamp the federal budget.” Rattner criticizes both the PPACA (with its Independent Payment Advisory Board) and Paul Ryan’s proposals as having insufficient teeth for curtailing cost compared to Britain’s quality adjusted life year (QALY) benchmark of approximately $48,000 per year. He suggests that the “last year of life,” which accounts for a quarter of Medicare spending, would be a good starting point.

Yet what exactly is rationing, and should “rationing,” properly defined, be the focus of any current health policy discussion? We will address these questions by offering a precise definition and discussion of rationing in the first installment of this Health Affairs Blog contribution. The second installment will provide a proposal for improving public discussion regarding the prudent consumption of health care services.

What Rationing Is

Many authors discuss the effects of rationing, but fewer provide a definition. Dictionaries define rationing in different ways, but virtually all include references to the field rations of the U.S. (or other) Army. Merriam-Webster adds the helpful supplement, “a share especially as determined by supply” (our emphasis). These definitions make it clear that rationing involves limits on consumption that are (1) set by someone other than the consumer (such as the supplier) with (2) the intent of limiting choices that some consumers otherwise would be willing and able to make.

In other words, rationing is a binding restriction (without alternative) on either the maximum or minimum quantity of a commodity that a consumer is permitted to consume or willingly would forgo. The rationing agent does not need to be an individual. Obviously, the government is able to engage in true rationing. Furthermore, the choice that the agent limits must be a choice that otherwise would be available and one that some individuals would exercise.

Importantly rationing can be either a ceiling or a floor constraint on consumption. Telling a person that they must buy a Honda Civic rather than a Lexus SUV is a ceiling constraint. Telling them they must buy a Lexus SUV whether they want one or not is a floor constraint. Denying medical treatment on the basis of the patient’s age or disability status would be a ceiling constraint. “Essential” benefit legislation or requiring individuals to purchase health insurance (or, if refusing, to pay a fine or tax, or give up their U.S. citizenship) is a floor constraint. A required minimal level of consumption may be the opposite of the way most people think about rationing, but it is rationing nonetheless.

Some constraints may not be effective rationing for some or all consumers. Mandating consumption of Honda Civics for people who would have bought them anyway may not be effective rationing for them, but it satisfies the requirement for rationing and potentially is binding for all consumers if they might change their minds.

What Rationing Isn’t

There are two ways to abuse the meaning of rationing. The first is to use the term so broadly that every consumer choice becomes a form of rationing. The website mentions physicians and health insurers as possible rationers of health care services, but also mentions “self-rationing” by patients. True rationing, as the Army example makes clear, is quite the opposite of self-rationing.

Peter Singer of Princeton University takes the position that “(h)ealth care is a scarce resource, and all scarce resources are rationed in one way or another.” This expansive use of “rationing” equates the consumer’s choice of one breakfast cereal over another with the denial of medically effective care to a disabled person simply because they are disabled. (For other expansive definitions of rationing, see Henry J. Aaron & William B. Schwartz with Melissa Cox, Can We Say No? The Challenge of Rationing Health Care; and Peter Ubel, Pricing Life.)

Expansive definitions of rationing are a form of linguistic larceny, robbing the term of its ability to distinguish importantly different contexts. Referring to all consumption choices as rationing simply because they involve a budget or supply constraint is equivalent to referring to all conflicts as war. There are passionate disagreements, duels, skirmishes and battles, and although it may be politically expedient in some contexts to refer to all conflicts as war, doing so robs war of its distinctive and horrific meaning.

The second way to abuse the meaning of rationing is to define it too narrowly. Rationing rarely is used to describe health care providers who acquire market pricing power and subsequently reduce the quantity of services supplied in order to command a higher market price. Nor is rationing often used to describe efforts to prevent lower cost health care personnel such as nurse anesthetists and physician assistants from assuming expanded roles in the delivery of health care services. Yet both of those examples involve limits on consumption that are set by someone other than the consumer with the intent of limiting choices that some consumers otherwise would be willing and able to make.

Expansive definitions of rationing are politically useful to those who favor true rationing, because if every choice is a form of rationing, then having the government deny certain medically effective care to the elderly, for example, is simply another form of an activity that already is pervasive. Similarly, failure to define rationing expansively enough is politically useful to those who would prefer to conduct their choice-limiting activities in the absence of politically volatile rhetoric. Neither misuse advances public discussion of the prudent use of health care services or the barriers that prevent it.

Rationing discussions sometimes focus on health plan coverage decisions. For example, an individual might prefer a health plan that provides free transportation to doctors’ offices, but that individual cannot find a health plan in her market area that offers that coverage. Another individual might prefer a health plan that covers an experimental therapy or an untested medical device, but no plan offers that coverage. Are those health plans rationing care? The answer is “No” for two reasons. First, lack of insurance coverage does not necessarily prevent the individual from purchasing the treatment on her own. (We take up the topic of insurance as a financing mechanism in a later section.) Her inability to afford the service in the absence of insurance may represent economic injustice, as discussed in the next section, but it is not rationing, per se.

Second, insurance is a group-purchased product, not the result of solitary preferences. The terms of coverage are determined by the preferences of the group, not single individuals. If an individual can find enough like-minded consumers who are willing to pay for the same type of coverage that she wants, and that coverage is not offered by insurers currently in the market, then the group can form its own insurance pool. If she cannot find such like-minded consumers, then there is no more rationing involved than if she is unable to find enough financial partners to purchase a piece of real estate that she would like to co-own.

Distinguishing Rationing From Injustice

Aside from political expediency, one reason rationing often is used to refer to normal consumption decisions is the observation that some consumers do not buy commodities that other people believe they obviously need. In those cases, we might say that the person cannot afford a basic necessity such as a certain type of health care service, and thus we have an example of “price rationing.”

The problem is that the term “price rationing,” in the absence of context, is too vague to be helpful. Certainly, there are cases involving prices that represent true rationing. We already have cited the primary example: monopoly pricing power. By restricting the supply of a service, the monopolist is able to generate a higher price per unit, and since there is no alternative supplier, some consumers will not be able to buy the service for which they would be willing and able to pay a competitive market price, whether directly or in the form of the actuarially fair cost of covering that service in an insurance premium. However, even a perfectly competitive market can produce prices of commodities that some consumers are unable to pay while maintaining subsistence levels of other basic necessities. For example, after purchasing minimum subsistence levels of food, shelter and clothing, a person might have nothing left to pay for their prescription medications at prices determined in a perfectly competitive market. Is that a case of “price” rationing?

While there might be universal agreement that the destitute consumer’s economic position represents an intolerable example of economic injustice, one demanding a remedy through either a cash or in-kind transfer of resources, a competitive prescription drug price that a consumer finds unaffordable would not be an example of rationing: there is no identifiable agent, external to the consumer, who is acting with the intent of denying the consumer a choice they otherwise would be willing and able to make. However the case of the Oregon Health Plan rejecting coverage of a prescribed lung cancer drug of demonstrable benefit, albeit not meeting a 5-year survival criterion for 5 percent of patients, would be rationing. (Interestingly the denial letter from the state reminded the patient that a lethal prescription would nonetheless be covered.)

The competitive market price of a prescription medication is the result of the interaction of millions of well-informed consumers and competition among multiple producers of the drug. If a monopolistic supplier maintains a price above the competitive level, there might be true rationing, but the rationing would be due to market failure, e.g., in the form of restricted entry or decreasing marginal costs, not “prices” per se. These examples illustrate the difficulty of finding true examples of rationing in perfectly competitive markets and the ease of finding true rationing in cases of market failure.

What about individuals who refuse to provide help to destitute individuals, either privately or through their opposition to public programs? Are they not engaged in true rationing? Again the answer is no, because the needy individuals currently are not otherwise able to make a choice they are being denied. If failure to support income transfers, per se, is rationing, then failure to provide subsidies to owners of Chevy Aveos so that they can drive Ferraris also is rationing. At that point, the term “rationing” has lost any useful meaning. Those who oppose income transfers for destitute people might be hard-hearted, tight-fisted louts, but not all louts are rationers, just as not all rationers are louts (unless simply declared a subset). However, louts can be forced to participate in tax-financed redistributive transfers if there is sufficient public support for the transfers.

Rationing and Market Failure

There are several types of market failure, some of which can lead to true rationing.

1. Pure public goods. Pure public goods exhibit two characteristics. Consumption by one individual does not affect the consumption possibilities of other individuals; and once the pure public good is provided, no one can be prevented from consuming it. The market typically cannot collect the revenue needed to finance the optimal level of a pure public good because consumers need not reveal their willingness to pay for it. National defense against incoming ballistic missiles is the textbook example. Failure to provide the socially optimal level of a pure public good could represent true rationing. However, neither health insurance nor health care services are pure public goods.

2. Declining marginal costs. Industries that exhibit declining marginal cost (increasing returns to scale) deteriorate into monopolies, which typically require regulation as a public utility. The production and distribution of electrical power is the (somewhat dated) textbook example. Failure to regulate a declining marginal cost industry, or inefficient regulation, could lead to true rationing. Neither health insurance nor health care services are declining marginal cost industries.

3. Externalities. Externalities or “spillovers” occur when one person’s or firm’s activity produces costs or benefits that affect another person or firm. Externalities can occur in consumption or production and they can be positive (benefits not paid for by the beneficiaries) or negative (uncompensated costs borne by the sufferers). Inoculation against infectious disease is a textbook example of a positive consumption externality, because one consumer’s inoculation reduces the likelihood that other consumers will contract the disease. Pollution is an example of negative production externality. Although many externalities can be resolved privately through the assignment of property rights (e.g. compensation), in other cases, taxes and subsidies are used to incentivize the socially optimal level of consumption or production. In severe cases, mandatory consumption requirements, such as mandatory inoculation, may be deemed necessary and can represent true rationing.

4. Poor information. Information is expensive, and the failure to provide perfect information or to become perfectly informed may be a rational economic decision. However, an attempt to restrict the public’s access to information on provider quality, for example, could represent true rationing if the information otherwise would be available at a price that consumers were willing to pay.

5. Restricted market entry. Market entry that is restricted either publicly, e.g., through licensure restrictions or certificate of need laws, or privately, e.g., early efforts by physicians to keep health maintenance organizations out of the market, could represent true rationing. For example, a consumer might prefer to receive a certain type of care from a nurse practitioner rather than a physician, but find that licensing laws have eliminated that choice. The licensing laws might be justified or unjustified, but the restriction on choice remains. Failure to enforce the antitrust laws can allow providers or insurers to amass significant market pricing power, reduce output, and thereby engage in true rationing.

6. Distorted prices. Insurance itself alters the point-of-purchase price of health care services, and massive government subsidies (so-called ‘tax expenditures’) — including the tax deductibility of health insurance premiums, out-of-pocket spending, and Medicare subsidies — distort the price of both health insurance and health care services. One could argue that these subsidies result in demand for insurance and health care services that establishes a standard of care, much of which appears to be unnecessary. As a result, some individuals, particularly those who lack access to those subsidies, e.g., the self-employed, are unable to purchase policies that offer coverage of a more conservative standard of care.

This is an interesting example, because it clearly is not the intent of the government (in providing subsidies for health insurance), or private insurers (in offering insurance to consumers), or subsidized consumers (in responding to the government subsidies) to restrict the insurance choices of unsubsidized consumers. So in that sense, this example of distorted prices would not represent true rationing. Still, should the effect of the subsidies be firmly established empirically, it would be incumbent upon the government to explain why the benefits of the subsidies, if any, outweigh the deleterious effects on both subsidized and unsubsidized consumers.

In sum, the right language and right motivation for rationing are important. Artificially imposed limits on consumption should be motivated by identification of clear cases of market failure or fairness problems, because rationing is itself a form of market failure in the form of restricted market entry. If analysts and politicians are unable to identify a clear source of market failure that justifies imposition of yet another form of market failure, we recommend that they abandon the term rationing and speak instead of a situation that is economically unfair or unjust and propose legal or economic remedies for that specific problem.