- Shopping Bag ( 0 items )
Coming from diverse backgrounds—economics, law, political science, and the health care industry itself—the contributors use Arrow’s article to address a range of present-day health-policy questions. They examine everything from health insurance and technological innovation to the roles of charity, nonprofit institutions, and self-regulation in addressing medical needs. The collection concludes with a new essay by Arrow, in which he reflects on the health care markets of the new millennium. At a time when medical costs continue to rise, the ranks of the uninsured grow, and uncertainty reigns even among those with health insurance, this volume looks back at a seminal work of scholarship to provide critical guidance for the years ahead.
Linda H. Aiken
Kenneth J. Arrow
Gloria J. Bazzoli
M. Gregg Bloche
Richard A. Cooper
Victor R. Fuchs
Annetine C. Gelijns
Sherry A. Glied
Mark A. Hall
Peter J. Hammer
Clark C. Havighurst
Peter D. Jacobson
Michael L. Millenson
Richard R. Nelson
Mark V. Pauly
Mark A. Peterson
Uwe E. Reinhardt
James C. Robinson
William M. Sage
J. B. Silvers
Frank A. Sloan
Joshua Graff Zivin
* General Equilibrium and Marketability in the Health Care Industry
Kenneth Arrow's 1963 article, "Uncertainty and the Welfare Economics of Medical Care," has become a seminal essay in the field of health economics. Its fundamental contribution is a detailed and thoughtful comparison of the deviations between the workings of markets for medical care and the competitive ideal. As Arrow demonstrates, a variety of factors prevents the medical care market from yielding an optimal allocation of resources. Prime among those factors is the "lack of marketability" for many products. Essentially, certain products that would improve the allocation of resources if they existed are not available for purchase (nonmarketable).
Arrow provides a complementary analysis of how "nonmarket social institutions" may arise to fill the gaps left by the lack of markets for certain products and thereby improve resource allocation. By nonmarket social institutions, he largely means norms of behavior that deviate from those typically observed in a competitive model.
This essay examines nonmarketability in the health care sector. The first section outlines Arrow's notion of general equilibrium in the health care sector and the problem of nonmarketability. The second sectionexamines the markets (and market failures) in the early 1960s and how those market failures can be traced to a lack of markets for several types of products. It concludes with a discussion of how nonmarket institutions could be viewed as filling the gaps for those missing markets. The final two sections discuss how, since 1963, there has been an expansion in markets and an associated change in the role of nonmarket institutions. The central thesis of this essay is that market and nonmarket institutions have a symbiotic relationship, with nonmarket institutions serving to improve resource allocation in areas where markets fail or do not exist. As the role of the market has expanded, the role of social institutions has changed to fill new gaps that have arisen in the increasingly market-oriented environment.
ARROW'S GENERAL EQUILIBRIUM ORIENTATION AND THE PROBLEM OF NONMARKETABILITY
Much of Arrow's acclaim reflects his exposition of the theory of general equilibrium. In economics, general equilibrium refers to the situation in which all markets (consumer and producer markets as well as markets for inputs such as labor and capital) are in equilibrium (namely, supply meets demand). It is largely a theory in which prices adjust to achieve this balance, and the theory recognizes the interconnection between markets.
The theory of general equilibrium is founded on decentralized action by consumers and firms, with consumers maximizing their well-being (utility) and firms maximizing profits. In standard models, individuals are assumed to be perfectly informed. Perfect information does not mean that everyone knows what the future will hold, only that they know the probabilities with which different events may occur. Outcomes are uncertain, but individuals are not uninformed (or misinformed). General equilibrium models also assume that, when faced with a set of prices, individuals (and firms) are cognitively capable of maximizing their well-being through their behavior. A variety of other assumptions, such as those that relate to market power (or lack thereof ), complete the general equilibrium model but are less salient for this discussion.
As Arrow notes in his essay, much of the appeal of the general equilibrium model relates to its implications for economic efficiency. Specifically, in general equilibrium settings two theorems of welfare economics link optimal resource allocation and competition. First, if a competitive equilibrium exists, it will be optimal; second, if we do not like the particular competitive equilibrium that arises from the market, we could reallocate incomes to achieve, through competition, any other optimal allocation we desired.
Such an analysis relies heavily on one's definition of optimal resource allocation, and Arrow is careful to introduce early in his essay the standards he uses and their precise meaning. Specifically, he adopts the economic concept of Pareto optimality, which defines an optimal allocation of resources as one in which no one person can be made better off without making at least one person worse off. He is careful to note that this is a weak definition of optimality. Many such allocations may exist, and though some value judgments would be required, society may not view each of them as equally desirable.
Arrow recognized several prerequisites for competitive equilibrium. Among the key requirements is that markets exist for all of the relevant goods and services (all relevant goods and services are marketable). Arrow views nonmarketability essentially as synonymous with a lack of markets. Marketability is necessary if individuals are to be able to match their purchases to their preferences, a fundamental feature of optimality. Without markets, there are commodities that would enhance welfare if produced, but they are not produced.
Arrow recognized that in the competitive ideal, general equilibrium would be characterized by a very rich set of markets. In the absence of this rich set of markets, Arrow contended that nonmarket institutions would develop so that resource allocation would come closer to the competitive ideal than would otherwise occur if only the incomplete set of markets were relied upon.
MARKETS, MARKETABILITY, AND THE ROLE OF NONMARKET INSTITUTIONS IN THE EARLY 1960S
The medical care market, as outlined by Arrow, really comprises two distinct, but interrelated, markets-the market for health care services and the market for health insurance. Arrow recognized that risk-averse individuals desire insurance against the financial and nonfinancial consequences of illness. Individuals purchasing health insurance products in the early 1960s typically purchased policies that reimbursed them for some portion of their expenditures. In this dominant insurance model of the time, the insurer did not interfere with the patients' choice of physician or the recommended treatment. Payment from insurer to physician was on a fee-for-service (FFS) basis.
This system could mitigate the financial risks associated with illness, at least for those with some insurance coverage. Yet several market failures were apparent. First, many individuals lacked coverage, a suboptimal outcome in a general equilibrium model if one assumes individuals are risk averse. Second, even insured individuals were not insured against the nonfinancial consequences of illness or treatment. Third, the system of insurance encouraged medical care prices and utilization to rise above their optimal levels (i.e., prices rising above marginal costs and utilization rising above that where the marginal benefit equals marginal cost). The phenomenon of consumption rising above optimal levels is commonly referred to as moral hazard (Pauly 1968; Manning et al. 1987; Newhouse 1992).
Each of these market failures could be traced to a gap in markets (a lack of marketability). Before discussing missing markets, it is important to recognize that the number of potential markets is enormous. The ramifications associated with the lack of any particular market depend on both the number of consumers who would benefit from the existence of that market and the ability of other markets to approximate outcomes that would arise had markets been complete. For example, the lack of a market for an insurance product with an 18 percent coinsurance rate for office visits might not be a big deal if not many people would have wanted such a policy or if policies with, say, 20 percent coinsurance rate existed.
In his analysis of the health care sector, Arrow identified several important market gaps. First, because the health care environment is particularly complex, characterized by considerable uncertainty, complete markets would imply that markets must exist for contingent contracts. These contracts would allow the consumer to purchase prespecified products, at prespecified prices, from prespecified vendors in different states of the world.
One type of contingent contract would involve the commitment to purchase health care goods and services only in certain states of the world. For example, a contract to provide open-heart surgery (or the financial equivalent), if, at some later date, one has a heart attack, would be one such contingent contract. If there were meaningful differences in open-heart surgery facilities, the contract would specify which facility would provide the service. In theory, widespread use of contingent contracts would reduce the price of services because individuals could shop for providers at a point in time when they were price sensitive as opposed to once they were ill and aware that insurance would share the cost. The contract would also limit excessive utilization because individuals could commit to the amount of care they desired when purchasing insurance, thereby avoiding insurance-induced excess consumption.
Another type of contingent contract would involve purchasing a warrant to compensate the patient for nonfinancial consequences of illness. This includes health status changes that cannot be remedied by medical care as well as failures of treatment to achieve intended outcomes. Warrants to insure against bad treatment outcomes are analogous to contingent contracts to mitigate the financial cost of illness. Ideally, a portion of the cost of treatment failure would be transferred to physicians and thereby improve incentives. Of course, given the importance of health, it is difficult to see how markets for such warrants could exist to transfer all of the risk from consumers to other parties. In the absence of markets for warrants, risk-averse individuals are forced to bear the risk associated with poor health and treatment outcomes.
In general, markets for contingent contracts do not exist, at least not at the level of detail envisioned by Arrow in a general equilibrium setting. The lack of marketability reflects the vast number of possible contingent contracts that would be required and the tremendous informational burden that writing, verifying, and enforcing those contracts would entail. Moreover, the transactions costs associated with individuals purchasing such insurance contracts, prior to illness, for care in the eventuality of any possible disease would be prohibitive.
A second type of market gap, which, along with excessive prices and moral hazard, deters insurance purchase, involves the absence from the market of certain insurance policies that would ideally be available. Considerable research, largely in the mid-1970s, examined the consequences of the asymmetric information regarding consumer risk and built on the basic outline contained in Arrow's article (see Rothschild and Stiglitz 1976; Wilson 1977; Cave 1985). The work in this area emphasizes that free entry into a market with asymmetric information limits the set of policies available. Certain standard products will not be sold, at least not at attractive prices.
This gap in markets arises because individuals can sort themselves in insurance markets so that relatively low-risk individuals separate from high-risk individuals in the insurance pool through their actions or through those of insurers. The process of sorting is referred to as adverse selection and arises because of the inability of insurers to assess enrollee risk. Specifically, individuals may have a better sense than the insurer if they are relatively high risk or relatively low risk. Low-risk individuals will try to avoid plans that attract high-risk individuals. In markets where the process of adverse selection is important, under certain conditions, equilibrium may not exist. Under other conditions it may exist, but certain groups of individuals may receive no coverage or incomplete coverage. In either case, Arrow notes the outcome is suboptimal. A recent summary of the empirical literature suggests that the quantitative impact of adverse selection is large (Cutler and Zeckhauser 2000).
A third important market gap arises because many types of information are not marketable. Lack of information is central to many of Arrow's arguments. In ideal conditions, information would be treated as a commodity in general equilibrium models. Imperfect information exists at all levels of the market, but Arrow emphasizes imperfect information among consumers. Prior to becoming ill, individuals may be unaware of the probabilities that they will contract various ailments and their associated consequences. After falling ill, patients may be unaware of the various treatment options and associated probabilities of various outcomes. They may be unaware of the natural course of the disease (which may be stochastic), and they may be unaware of the quality of their physician. In an ideal competitive environment, information would be marketable so that individuals could purchase the desired information.
Yet Arrow recognizes, for a variety of reasons, markets for information violate standard assumptions regarding products. For example, information has some of the properties of public goods. Even conceptualizing these markets (and associated demand curves) is often challenging, because of, as Arrow notes, "the elusive character of information as a commodity" (946). This lack of markets for information is a marketability problem. It contributes to the inability of other markets, such as those for contingent contracts, to exist and contributes to the adverse selection problem.
Given the marketability problems in the health care sector, Arrow contends that nonmarket norms and institutions arise to fill the gaps. Perhaps the most important of the nonmarket responses is the physician code of behavior. Because warrants are not marketable, patients must place extraordinary trust in their physician, both because the physician may not have the ideal incentives and because the patient is not insured against treatment failure. Arrow sees the code of physician ethics as an outgrowth of the need for trust in the physician-patient relationship, though others, such as Paul Starr (1982) and Charles D. Weller (1984), have questioned Arrow's interpretation of the reasons for the development of these ethics. This code of ethics results in unique expectations of physician behavior. These nonmarket standards of behavior and their extraordinary importance in this sector represent the most fundamental deviation of health care markets from other markets and, in part, fill gaps arising due to the absence of a market for the insurance against poor treatment outcomes.
The code of physician ethics, in theory, could mitigate problems associated with a lack of markets for contingent contracts for health care services, specifically over utilization and excessive prices. If physicians acted as perfect agents for their patients, utilization would be optimal, with physicians prescribing care based on what the patients would have desired at the time of insurance purchase, and pricing would be based on marginal cost.
Interestingly, the principles of physician conduct shaped the norms of behavior in the insurance industry, which tended to prevent marketing of products that could minimize the market gaps associated with the lack of contingent contract markets for health care services. Specifically, insurers did not interfere with the physician practice patterns and did not steer patients to particular providers. Any other set of norms for the insurance industry would have conflicted with the preeminent role played by physicians and perhaps threatened the central role of trust in the physician-patient relationship. In essence, the nonmarket norms that addressed, in part, the lack of markets for warrants resulted in norms of behavior in the insurance market that limited the ability of markets to address the nonmarketability of contingent contracts for medical care.
Arrow notes that other nonmarket norms existed as well, which could be interpreted as filling market gaps. For example, he indicates that for the purposes of price setting, market forces tend to be replaced by direct institutional control, but he does not elaborate. Additionally, nonphysician participants in health care markets, most importantly hospitals and insurers, tended to be organized as nonprofit institutions. Nonprofit status in these areas of the health care industry addressed many of the same issues that relate to the absence of markets for contingent contracts, such as trust, excessive use, and excessive pricing. For example, in theory nonprofit status reduced incentives for hospitals to take advantage of consumers for a profit.
Excerpted from Uncertain times by Peter Joseph Hammer Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.