Containing health costs in a consumer-based model.

The assumption that consumer choice cannot be used to achieve cost control in health care is invalid. It does not do so today because the tax treatment of health care leads to perverse consumer incentives that encourage cost escalation. By reforming the tax treatment of insurance and out-of-pocket medical costs, it is possible to design an efficient and universal system in which consumer choice is a powerful restraint on cost.

Introduction

There is a spirited debate taking place over fundamental reform of the U.S. health care system because of a widespread belief that it does not achieve three basic goals. It does not assure access to affordable care for many individuals. It is not efficient, in that resources are not being used in ways that bring maximum medical benefits for money spent. Further, there is the perception that America spends “too much” compared with other industrial countries. Almost 9 out of 10 Americans now believe the system is so flawed that it needs rebuilding (Blendon et al., 1990).

To be sure, these concerns are imprecise and subjective. What, for instance, constitutes “too much” spending on health? Americans do not tend to worry about too much being spent on housing or pizzas. Even if the concern about total cost is traced back to a rate of annual price increase that is well above the average level for goods and services, that itself is not a reason for alarm. Logically there are always going to be items that rise in price more quickly than the average, and there are technological and other reasons that health services might be one such item.

Still, many Americans feel that the health care system does not work in a satisfactory manner, compared with the rest of the economy. Indeed, the health care system does operate quite differently from other systems. Goods and services are demanded, generated, and distributed in most sectors through consumer-driven markets. In these markets, decisions made by the consumer in response to prices offered by competing providers create a dynamic favoring the efficient use of resources, a range of products to suit different incomes and preferences, and a total level of spending that reflects the value society places on one product over another. In general, this consumer-led economic system produces a higher level of satisfaction than other economic systems.

Except for small sections of the system, however, health care does not function in this way. Consumers do not encounter accurate market prices when they seek a medical service or insurance plan. For the most part, it is a third party, normally an employer or government agency, that is confronted by realistic prices. Actually, even these prices are only loosely related to the cost of serving individual patients. In the case of employer-provided plans, for instance, the price for covering a particular employee usually includes a large degree of cross-subsidization. Moreover, in the publicly supported programs and in particular Medicare, there is an increasing degree of price setting by the third-party buyer. In the case of the relative value scale price structure, now being introduced, the basis of this price setting for treatments is an objective labor theory of value. Strangely, this notion is more in line with the Marxist value theory than with the principle of subjective value underpinning market economics.

How the current system works

Why does the U.S. health care system operate on such a different foundation from that of other parts of the economy? The reason is that the policies shaping the system rest on two key assumptions about consumers of health care and the nature of health care as a commodity.

The first assumption is that if market prices were to determine the supply and demand for medical services, certain consumers would not be able to afford treatments necessary for their well-being or even survival. Expensive health services also are often needed urgently, making it difficult or impossible for even relatively affluent consumers to arrange finances as they would for other large expenditures, such as buying a car or a house. In a society in which medical care is considered a necessity rather than a privilege, a system that rations simply by market price is unacceptable.

Insurance, of course, is a mechanism to deal with the high cost of a medical service, spreading it over time as well as among individuals. This only reduces the ability to pay problem, however, it does not solve it. All other things being equal, a consumer will be quoted an insurance price in line with the actuarial risk he poses to the insurer. That will still result in many cases where the price is prohibitive.

The second assumption about the market for health is that consumers are not able to make well-informed decisions about medical services. This is not unique to health care, of course, but it is of greater concern in the health sector than in other sectors. For one thing, the emergency nature of certain services, and thus the impracticality of the consumer to “shop around,” can put the consumer at a severe disadvantage compared with the provider. This feature should not be exaggerated, however. Figures for 1977 indicate that only 2 percent of physician visits result in a hospital admission, and only between 15 and 30 percent of hospital emergency room visits are for urgent care (Ricardo-Campbell, 1982).

In addition, even though the purchase of an insurance package may reduce the knowledge problems facing consumers by including only providers selected by the insurer on the basis of quality and price, and by not covering procedures of doubtful value, the purchase of the package itself is a confusing process for many consumers.

These concerns, among others, triggered a number of incremental policy decisions by Government that resulted in the employment-based group coverage that serves most Americans today. In this system, the prices of most services are subsumed within an insurance package, the price of which is set according to the experience rating of a company-based group. The specifics of each package are determined by the employer, or by negotiation between the employer and organized employees, and between the employer and providers directly or through an insurer as intermediary. In addition to the cross-subsidy from low-risk to high-risk employees implicit in the group rating approach, there is cross-subsidy from general taxpayers to covered employees by virtue of the tax exclusion available for company-provided plans.

It should be noted that this system came about more or less by accident. The Canadian and British systems, by contrast, were designed and created after an intense national debate. When cash wages in the United States were controlled during World War II, expanding fringe benefits became a way for employers to attract scarce labor. Rulings by the Internal Revenue Service after the war firmly established the tax-free status of such benefits.

This system does avoid many of the drawbacks of an unfettered market for health care. Although prices do function within the system, they influence choices made by providers, insurers, and third-party payers rather than those made by the consumers of health services. Thus price does not pose an unacceptable barrier to reasonable access for most covered consumers (although limits on coverage, copayments, and deductibles eventually can amount to a financial barrier). Consumer confusion regarding the choice of plan, and in many instances even the selection of a provider, is avoided by the employer taking on this responsibility.

Yet this system is driven, in large measure, by consumers reacting to price at the point of consumption. It is just that the prices encountered by the consumer bear little or no relation to cost. This absence of real prices for the consumer, either for specific services or for a health package, leads to an economic dynamic that makes cost control and the goal of universal access such perplexing problems in today's health care system—hence, the calls for radical reform.

In today's health care system, consumers make what are for them quite rational economic decisions, yet they do not face directly the full economic consequences of their decisions. In particular, they have the incentive to overdemand services that are underpriced or “free” to them. Even though the heavy use of services means that the cost of their insurance premiums rise, they have little incentive to consider that. One reason is that with premiums set according to the group's aggregate demand, one individual sees little connection between his use of services and the premium. Another reason is that the employer is seen as paying the premium—even though labor economists would point out that cash wages and fringe benefits combine to form a total market-determined compensation package, and thus are “paid for” by the employee.

The tax treatment of benefits, moreover, leads to an understandable resistance among employees to attempts by employers to encourage workers to control their health costs by introducing copayments and similar financial devices to raise cost consciousness. Company-provided plans are tax exempt, but out-of-pocket payments with rare exceptions are not. Thus the employee has the incentive to prefer a package with no out-of-pocket element over a medically identical package including a copayment with a lower total cost.

Among the other consequences of this distorted market is the incentive for providers of services to compete for patients on the basis of available services, rather than on benefit compared with cost. Thus an expensive “arms race” occurs in the medical sector, in which hospitals and physicians entice customers with increasingly sophisticated but costly equipment and procedures. Third-party payers respond by questioning in detail the treatments agreed to by physician and patient, to the irritation of both.

In addition, the tax treatment of health care is a significant contributor to the huge gaps of coverage in the U.S. system. Families with company-provided plans receive financial assistance through the tax exclusion, and most individuals without such coverage receive no such tax help. Moreover, the most tax assistance goes to employees in the highest tax bracket with the most generous plans. Meanwhile workers without company plans must pay in after-tax dollars for insurance and face prices for services that are driven up by the rational but inefficient decisions of consumers who are insulated by company-paid plans. Many such workers and their families respond to the effects of this inequitable tax treatment by not buying insurance. In fact, approximately three-quarters of the uninsured are employed individuals or the dependents of workers (Short, Cornelius, and Goldstone, 1990).

The response of many policymakers to this unsatisfactory and increasingly unpopular system is to propose alternatives that try to move even further away from a traditional consumer-driven market. One proposed alternative is a Government-financed single-payer system, based on the Canadian or British models. Another is to try to deal with the gaps in coverage by requiring employers to furnish employees and their families with a minimum level of benefits or contribute to a public program to deliver similar benefits to uninsured workers and their families.

It is not within the scope of this article to provide a critique of these approaches, but it is important to understand their underlying dynamics. The single-payer approach would eliminate consumer choice entirely as a device to achieve economic efficiency and cost control. Instead it would substitute a system of allocation based on fixed budgets and Government-administered fees to providers. In such a system, officials must use objective methods to determine value and benefit, in contrast to the subjective consumer determination in a normal market. Whatever the virtues might be of such an approach, the world's experience with price and budget controls—in health care or the entire economy—is that controls must become ever more pervasive and elaborate, shortages and explicit rationing become a permanent feature of the system, and measurements of value become more arbitrary and contentious.

Mandates on employers do not establish a total budget for health care, nor in their basic form do they change the incentives facing the consumer. Thus mandates pose problems as a discipline for demand-induced treatment cost escalation and as a method for controlling total health costs. Most mandate proposals try to address this by including very complicated mechanisms intended to guide consumer choices and avoid cost shifting (Enthoven and Kronick, 1989a and 1989b). Some versions would introduce into the private sector the fee-setting features of a Government-sponsored single-payer system.

The assumption behind each of these alternatives is, of course, that consumer choice is of little use as a tool for cost control or resource allocation in health care. It is possible, however, to imagine a very different system of health care in the United States that would be based on consumer choice, rather than to see it as irrelevant or as a problem. Under this arrangement, cost control at the macrolevel is achieved by changing the environment in which microeconomic decisions are made by consumers of health care.

Creating a consumer-based system

A consumer-based system would have to contain certain features to be acceptable to society and to lead to efficient cost control. These features include the following:

Consumers would have to face accurate market prices. The key to cost control with efficiency in a market system is to allow consumers to choose among competing producers on the basis of undistorted prices. There are ways to approximate this condition in health, however, without the problems commonly attributed to an unfettered market. In particular, the cost-control features of a consumer-driven market can still operate when insurance is part of the equation. The key is to allow the consumer a complete choice of health plans without that choice being distorted by tax considerations. Ending the discrimination in the tax code in favor of insurance at the place of work would, for the most part, achieve this. In addition, providing the same tax relief for out-of-pocket expenses as for insurance premiums would induce many consumers to accept copayments and deductibles in order to reduce premium prices, thereby encouraging patients to question the prices of providers.

If the tax code was more neutral in this way, the costs and benefits of health packages would be compared without the perverse incentive to choose an expensive one merely because it was provided through an employer. Choosing between health plans is not, of course, the same thing as choosing among individual physicians or hospital services on the basis of prices. However, it would lead indirectly to a similar result. In a tax-neutral market for health plans, consumers would be invited to consider tradeoffs. One plan—such as those still common today in older, unionized companies—might fully reimburse virtually all services and place few limits on choices of physician or hospital. Such a plan would be expensive compared with another that also placed few curbs on physician choices but imposed a significant copayment. It would likely be even more expensive than a managed-care plan with tight limits on consumers' choices once they enter the health care system.

Consider such a situation: If the consumer has a free choice of plan, without any tax bias in favor of one or the other, the consumer's sensitivity to the price of each package will tend to lead to efficient cost controls within each package. In a managed-care plan, consumers in effect choose to accept limits on their choices once they receive care in return for a less costly but satisfactory package of services. To remain competitive, however, the plan must provide overall satisfaction compared with alternatives, and plan managers must, therefore, strive for the best combination of plan price and the quality of service by their chosen providers. Similarly, a plan allowing greater choice of physician and course of treatment must introduce reasonable and efficient price incentives to encourage the patient to economize if its premium price is to be competitive. Thus even though consumers may not directly question the quality and price offered by each provider, their ability to choose among plans induces plan managers to be driven by the consumers' reaction to the overall price they are offered to assemble a package of providers, services, and incentives for the patient that curbs costs and promotes efficiency.

As an analogy to this process in which individual physician and hospital charges and quality are indirectly determined by the consumer, consider the purchase of a car. When a typical American buys a car, he does not and would not want to negotiate with a carburetor manufacturer, a tire company, an upholsterer and so forth to assemble his car. He buys a package, and rival companies compete in offering him a package. Each company knows that what matters to the consumer is the overall performance and price of the package, and perhaps some specific features. Constantly influenced by the consumer's likely reaction to combinations of quality and price in the assembled car, knowledgeable companies engage in a detailed examination of the price and quality of components offered by competing suppliers. Thus although the buyer of the car does not directly question the cost and quality of each element in the package, he does so indirectly through the car company competing for his business. A patient, in effect, would have exactly the same influence on providers in a market with competition among health plans and on consumers with the incentive to compare cost and overall quality.

Changing the tax system to encourage consumers to make such price-sensitive decisions does not in itself satisfy the objections to a consumer model of cost control, however. This simple model still leaves unanswered such issues as a consumer's ability to pay and his ability to choose among complex combinations of services and prices. These issues must be dealt with through refinements of the simple model.

Consumers would need to be subsidized. An open market for medical care and services clearly would lead to risk selection, where consumers who required a heavy volume of medical services would face prohibitively high medical bills or insurance premiums. Critics of a consumer-based market for health care tend to assume that such a result is sufficient reason to reject the market approach. However cross-subsidization can still occur in a market in which prices reflect actual cost and risk.

Today's employment-group system relies on equal premiums for different individual risks as a principal method of cross-subsidy. As insurers and corporate benefits managers know all too well, this system is very vulnerable to healthier members of the group “escaping” to other health plans priced more in line with their lower risk. Another method of cross-subsidization would be to allow consumers the freedom to choose among alternative plans priced according to plan and the buyer's health risk, and then to subsidize the buyer directly so that he could reasonably afford the premium. An example of such a method of cross-subsidization would be a sliding-scale tax credit to offset premium and out-of-pocket health costs, with a voucher or refundable credit for those paying little or no taxes. In this arrangement, the percentage credit would be determined by the individual's total anticipated outlays on premiums and direct health costs compared with his income; the higher that ratio is, the higher is the percentage credit.

Subsidizing through the tax code rather than through equal premiums would avoid most of the problems associated with adverse selection. In fact, adverse selection—considered a problem in health care—is equivalent to consumer choice in other sectors— normally considered a virtue of the market system and the key to efficiency and cost control. Changing the method of cross-subsidization would make it a virtue in health care.

Any form of subsidization does, of course, alter the effective prices encountered by the consumer and influences decisions at the margin. A credit against health costs would encourage the consumer to buy more health care than he would otherwise. To some degree, this is an objective of public policy. Moreover, a credit still encourages consumers to take relative prices into account when choosing among alternative plans.

Admittedly, the incentive to compare prices diminishes as the percentage credit increases. Thus a family facing medical or insurance bills that are very heavy when compared with its income conceivably could require a credit of close to 100 percent if its after-tax, out-of-pocket cost is to be affordable. In this case, there would be little sensitivity to price. If this is in practice a problem, there are two possible solutions. One would be, as Steuerle suggests, to provide a 100-percent credit against part of the cost of a minimum package of benefits, and a sliding scale thereafter (Steuerle, 1991). For the family in question, this would mean that the same dollar amount of subsidy would be achieved with a lower tax credit against the last dollar spent, thereby maintaining price sensitivity at the margin. An alternative approach, favored by this author, would be to limit the maximum tax credit to, say, 80 percent and allow those still facing hardship to enroll in a subsidized risk pool based on managed care. This uses a method of resource allocation and cost control that is less efficient when applied to consumers in general, but which may be more efficient when the objective of limiting the total cost to the consumer conflicts with the cost consciousness.

Consumer ignorance would have to be addressed. An efficient market requires consumers to make well-informed decisions; and on the face of it, this might seem difficult to achieve in a consumer-driven market for health care. For one thing, “user-friendly” information is in short supply for patients in today's U.S. system (as it is, of course, in the government-run systems in Canada and Britain); and even with such information, it would be difficult for many, if not most, Americans to make confident choices.

The lack of usable consumer information today, however, is the result of the weak incentives for rational consumer choice rather than of some inherent imperfection in health care markets. When a system provides little incentive for consumers to compare quality and price—and, even more important, a strong tax disincentive against choosing any alternative to a company-sponsored plan—it is hardly surprising that there is a dearth of information on which a consumer can determine good value for money. Significantly, in the Federal Employee Health Benefits Program, a controlled voucher program in which approximately 10 million Federal employees and dependents can choose among a wide variety of plans, there is abundant information in various forms. Besides official Government handbooks, a detailed comparative guide to competing plans, giving such information as average likely expenditures under each plan for various categories of family, is published by the Center for the Study of Services, a Washington, D.C., area consumers' organization (Francis et al., 1990). In addition, Washington's talk shows and newspapers are full of discussion about rival plans during the period in which plans can be chosen. With consumer choice a feature of the entire U.S. system, there is every reason to expect similar information be demanded and supplied throughout the country.

Similarly, corporate buyers facing mounting health benefits costs increasingly are demanding information on quality and outcome from providers and are using this information to choose providers for their plans. With consumers an active ingredient in the system, rather than a passive observer as they are today, the managers of health plans would have a strong incentive to make such information available to consumers or to consumer organizations, rather than only to corporate benefits managers.

Consumers unable to digest even “user-friendly” insurance or medical information would do what consumers do in similar situations when they are unfamiliar with a product or service—rely on a trusted professional or agent. Primary care physicians and insurance brokers are obvious examples of experts who are routinely turned to for assistance with complex medical or insurance decisions.

In addition, a likely development under the consumer model outlined would be the emergence of new groups acting as organizers or brokers of plans. The current tax treatment of consumer health purchases encourages only employer-sponsored plans. Under the tax treatment outlined, however, other organizations would be natural candidates to organize health plans. Some of these, as today, would be provider-based, such as health maintenance organizations. Others would be groups with the characteristics of a “friendly society.” Unions would be obvious sponsoring groups, because members of the union could reasonably expect the union to act wisely on their behalf. In fact, union-organized plans are a feature of the Federal employee system, attracting non-members as well as members. Other possible groups would include farm bureaus (some of which already organize plans), alumni associations, churches, and groups representing those suffering from particular chronic illnesses, such as diabetes. In the latter case, the plans no doubt would feature a particular set of specialized services needed by group members. Besides acting as reliable agents for consumers, such groups would provide many of the benefits normally associated with insurance groups today. For instance, as bulk purchasers, they would be able to negotiate with providers and insurers and offer lower marketing and administrative costs in exchange for discount.

The consumer could, in addition, be assured of a certain range of basic services within each plan by government regulation. As a legal condition of sale, a comprehensive health plan could be required to contain certain features, such as catastrophic stop-loss protection. In addition, consumers could be required by law to purchase a plan with at least this basic package of features, to assure all families were adequately covered.

Several proposals have been put forward to create a universal health care system based on such a consumer-based universal health care system. In these proposals, the consumer's power and incentive to choose is the key to cost containment, as it is in the rest of the economy. Among the more recent proposals, one by Pauly et al. (1991) would require all Americans to obtain basic insurance coverage. The current tax exclusion for company-based health plans would be replaced by a system of refundable tax credits based on family income and risk category. A fallback subsidized risk pool would be available for those unable to obtain insurance in the private market. Medicaid would be replaced and Medicare could be phased out, say these authors, as individuals with a plan under the terms of the proposal reach the age of Medicare eligibility.

A very similar proposal has been developed by the author and his colleague Edmund Haislmaier (Butler, 1991; Butler and Haislmaier, 1989). This differs from the Pauly plan only in its details. It would exchange the current exclusion in a revenue-neutral manner for a system of credits, and it would require all non-retiree families to purchase a basic health plan, with credits available against premiums and the cost of supplementary medical services. In certain situations it would allow families to enroll in a subsidized risk pool as an alternative. It would retain a limited and modified Medicaid system; and although the proposal recommends reform rather than the replacement of Medicare, the basic proposal could, in principle, be extended gradually to those now eligible for Medicare.

Proposals of this kind, based on the consumer model of cost control and resource allocation that applies in virtually every other sector of the economy, offer a method of achieving the goal of universal access to health care at reasonable cost. Instead of exacerbating the inefficient results of consumer choices in the current system, as the mandated benefits strategy would do, or replacing consumer choice with government allocation and pervasive price controls, as advocates of a Canadian system propose, these proposals would reform the environment in which the consumer makes health care decisions. In this way, costs would be contained by consumers making informed and efficient decisions—a system that routinely outperforms all other economic mechanisms intended to achieve cost control with efficiency.