The United States is in the midst of a massive social experiment that will
determine whether a democratic nation can control the inescapable tendency of
widespread insurance to cause excessive health care spending. The problem has
been well enunciated by health economists for a generation since the pioneering
articles of Kenneth J. Arrow and Mark Pauly. These two showed that insurance, by
lowering price at the time health care is consumed, causes rational economic
agents to demand socially excessive amounts of it.
Many other analysts pointed out the possibility of too much, too little, or
the wrong kinds of care being sought by poorly informed or
less-than-fully-rational patients. And physicians, acting as imperfect agents,
could only compound the problem. In any event, patients would demand the “wrong”
amounts or kinds of care—wrong in the sense that care would differ from what
would be demanded by a fully informed patient represented by a perfect physician
agent.
These two sources of “incorrect” demand for health care imply rather
different remedies. If the problem is that rational people face incorrect price
signals, the solution is to get the signals right, meaning that prices should
equal the social cost of resources used in producing the care. If correct price
signals guide rational patients and providers, the market for health care will
deliver the right care to the right people. Actually, two additional conditions
are necessary to reach this goal: the distribution of income must be socially
acceptable and people must be able to call on future earnings when current
health costs are large relative to current earnings. If people are not rational,
if they have poor information, or if objectives of providers and patients
diverge, any number of other interventions may be necessary to get the right
amounts and kinds of care to the right people.
Some health care analysts emphasize the pernicious effects of distorted
prices. Others dwell on irrationalities, poor information, and misaligned
incentives. The differences among health policy analysts on why health care
markets sometimes perform poorly run right through most disagreements on policy.
Of course, price signals might be distorted and people might behave with less
than perfect rationality and incentives of providers and patients might
sometimes diverge and flawed information might obstruct good decisions. It is
striking, however, that few disputants acknowledge that all of these problems
might be present at the same time.
Against this background, advocates of managed competition come down
foursquare in the faulty price signals camp. Insurance, they point out, causes
demand to be insensitive to price and leads to excessive demand. Individuals who
are shielded from the social costs of the care they seek want too much care when
ill. Employers aggravate the problem by paying more if employees buy costly
insurance than if they buy inexpensive insurance, thereby insulating workers
from the financial consequences of choosing insurance with low deductibles and
negligible cost sharing. To be sure, advocates of managed competition also
stress the imperfect quality of information—of individuals about insurance
products and of insurers about the health status of their potential customers.
Indeed, it is less-than-adequate information that requires competition to be
managed so that insurers do not use nonuniform insurance products to confuse
customers and engage in risk selection. To avoid such practices, some form of
intermediary—the health insurance purchasing cooperatives in the original
Enthoven plan and organizations with other names but similar functions in other
plans—is necessary to ensure that individuals confront insurers offering similar
coverage and insurers do not engage in any of the myriad risk-selection
strategies available to them.
If price incentives are properly aligned and marketing is adequately
controlled, customers can compare prices and evaluate care quality. The premise
of managed competition is that markets will work if only we will let them.
Correct the information failures, get the price signals right, and the pervasive
problems with the U.S. health care system will vanish or greatly diminish. The
United States would cease to be the health cost inflation wonder of the world.
The quality of care would improve as individual consumers, voting with their
feet (or their health insurance checkoffs at work) would drive providers to hold
down costs and improve quality. The relentless discipline of the marketplace
would be the lash driving health care reform.
Not everyone is sold on managed competition, however. Many observers believe
that the elements of managed competition are insufficient to achieve the cost
control and quality improvement that its advocates claim. A smaller number think
these elements are not necessary. And those health policy analysts who believe
that real market forces should not sully the delivery of health care do not even
think that the emphasis of managed care on getting prices right is even
constructive (Marmor 2000: 178).
The striking subtext of the James Maxwell and Peter Temin article is the
tension among these disparate views on health care services among major
corporations. Their findings would merit close attention if for no other reason
than the quite remarkable response rate—85 percent of those surveyed—a rate that surely merits recognition by the Guinness Book of Records. But there are many
other reasons to read their findings with care and for edification.
The authors interpret the responses to their survey as a report on the
struggle between two different approaches to the management of health
insurance—managed competition and industrial purchasing. Orthodox managed
competition consists of three elements: consumer choice, good information, and
correct price incentives on the margin. Industrial purchasing involves the
exercise of the first two of these functions by large companies who, not
incidentally, also have some market power in the purchase of health care. The
study describes these two approaches to corporate management of health care
services as polar alternatives. But there is really only one pole. Managed
competition is distinguished by certain practices; in contrast, there is no
single industrial purchasing model. It is a congeries of different practices
that can be combined in almost infinite variety. In comparing managed
competition with industrial purchasing, therefore, one is not comparing two
clear alternatives but, rather, prototypical managed competition with a
bewildering variety of practices. To paraphrase Tolstoy, all managed competition
plans are alike, but all nonmanaged competition plans differ in their own way.
If one takes the virtue of individual choice in selecting health insurance as
axiomatic, then the companies that practiced industrial purchasing of health
care were apostates from the one true faith, substituting corporate decisions
for choice by individual workers. For those who believe that the choices of
rational individuals facing undistorted prices will be superior to choices made
in any other way, the diverse compromises grouped as industrial purchasing must
lead to inferior outcomes. For those who do not share this faith, the package of
practices grouped as industrial purchasing may be practical compromises based on
the facts and circumstances of each local health care system and workplace.
To judge which of the many approaches subsumed under industrial purchasing or
managed competition, in fact, produces superior results is surpassingly
difficult. And, to its credit, the Maxwell and Temin article does not claim to
do so. The first step in a full evaluation would be determining the measures of
success. Presumably, the objective is to choose an insurance regime that
produces a mix of services that both matches patients’ preferences and provides
cost-effective treatment of illnesses or balances these goals when they
conflict. Achieving this goal may or may not result in reduced rates of medical
cost inflation, involve the provision of “enough” information (but not “as much
information as possible”), lead to fewer complaints from patients, and so on. In
short, it is not clear what the sign should be in an equation relating the
various independent variables in Table 6 and the rate of premium increase.
Maxwell and Temin do not claim to have collected the information necessary
for a full evaluation. There is no way to collect that much information in a
half-hour survey or even in a much lengthier survey of benefits managers.
Instead, the authors report (in Table 6) on correlations between cost increases
and various characteristics of each insurance plan, company characteristics, and
limited regional data. The resulting equation reflects no known theory of
medical costs. The independent variables could influence either the level or
rate of increase of health costs. If their effect was fully reflected in the
level of medical costs at the start of the period, there would be no reason to
expect significant effects on inflation even if they were potent influences on
the level of costs.
I found the case studies to be more interesting than the statistical
analysis. The studies illustrate that even those companies influenced by the
tenets of managed competition did not slavishly follow them. Typically, the
companies were somewhat paternalistic. They foreclosed some dimensions of
individual decision making by limiting access to only a few health care plans.
They used unequal company premium contributions to nudge workers into plans that
the company judged superior. Other companies doubted that individuals would
discipline providers sufficiently to improve quality. As Maxwell and Temin
write: “Many companies… do not trust the market to offer the right choices or
consumers to make wise selections.”
What is one to make of this attitude? One response might be that the
companies are simply wrong, that they foolishly and shortsightedly lack faith in
consumer-driven market pressure, and that they underestimate the intelligence of
or show too little respect for the preferences of their own employees. Another
response might be that by limiting the number of insurance venders, the
companies are using old-fashioned monopsony power to extort price or other
concessions. On this view, companies could wring more from health care providers
by buying in bulk than individual purchasers could achieve buying retail. Still
another response might be that the emphasis of managed competition on the
consumer gains from exercise of choice at the level of insurance plans is
misplaced. On this view, the choice that counts is in the selection of providers
and of care at the time of illness. Choice among insurance plans could even be
perverse if healthy people are poor judges of the extent of coverage they will
want when ill. This possibility is strongly suggested by the literature on
self-control (Schelling) and the psychology of decision making (Rabin) and
limited evidence from controlled experiments (Benartzi and Thaler).
None of these views undercuts the insight of advocates of managed competition
that companies can improve employees’ decision making if employees face the
price consequences at the margin of their health insurance decisions and if
health plans are forced to provide clear information about available plans, and
to avoid confusing variations in their insurance offerings. But in the same
sense, the orthodoxies espoused by advocates of managed competition do not
contradict the beliefs of the large majority of benefits managers that they are
better able than their employees to amass and process the complex information
regarding the price and quality of health care and to wring concessions from
hospitals, physicians, and other health care providers. Someone once quipped
that the mark of an educated person is the capacity to hold two contrasting
ideas in one’s mind at the same time. Adherents of single-factor explanations of
the problems of the U.S. health care system might profitably dwell on this
thought.
References
Arrow, Kenneth J. 1963. Uncertainty and the Welfare Economics of Medical
Care. American Economic Review 53(5):941-973.
Benartzi, Shlomo, and Richard H. Thaler 2001. How Much Is Investor Autonomy
Worth? Working paper. March. Available on-line at http://gsbwww.uchicago.edu/fac/richard.thaler/research/workingpapers.htm.
Marmor, Theodore R. 2000. The Politics of Medicare. 2d ed. New York: Aldine
de Gruyter.
Pauly, Mark. 1968. The Economics of Moral Hazard. American Economic Review
58:531-537.
Rabin, Matthew. 1998. Psychology and Economics. Journal of Economic
Literature 36:11-46.
Schelling, Thomas C. 1984. Choice and Consequence: Perspectives of an Errant
Economist. Cambridge: Harvard University Press.
Posted with permission. All rights reserved. This material may be saved or photocopied for personal use but may not be otherwise reproduced, stored or transmitted by any medium without explicit permission. Any alteration to or republication of this material is expressly forbidden. Please direct permissions inquiries to: Permissions, Box 90660, Durham, NC 27708, USA; or fax 919.688.3524.
Commentary
A Funny Thing Happened on the Way to Managed Competition
February 1, 2002