TCS Daily

Doctoring the Market

By Arnold Kling - March 14, 2005 12:00 AM

"A fully developed [consumer-driven health care] market would be chaotic, to say the least, and in such a system continuity of care would be virtually non-existent. It is hard to imagine such a system improving the quality or the efficiency of care. The problems with the quality of our current delivery system would probably worsen." -- Arnold S. Relman writing in The New Republic

Writing in The New Republic, Arnold S. Relman outlines an argument against market-based health care and in favor of a state-controlled, technocratic alternative. He plans to spell out the argument in a book. The purpose of this essay is to raise some issues that I believe his book should address in order to make a persuasive case.


It appears that Relman's argument boils down to this:


1. Health care is not like other goods and services. To provide quality health care, doctors must adhere to a high level of professional and ethical standards.


2. Market competition tends to undermine professional and ethical conduct.


3. Therefore, a nationwide, centralized, technocratic approach is required, in order to insulate doctors from market pressures.


The Need for Ethics


I think it is a mistake to believe that ethics are only needed in certain fields. My recent experience buying a computer showed how important it is to avoid unethical dealers. When I take my car in for service, I prefer a local garage to JiffyLube, because I want to be confident that if the mechanic tells me that my radiator needs flushing it is not because he is trying to fill some sort of corporate quota.


In today's complex economy, many of our major purchases are made under conditions of uncertainty. How many people can evaluate the engine design in different cars? How many people can evaluate the quality of all of the construction materials in the home that they buy, or follow the chain of title to make sure that their ownership will be secure?


Health care providers, unlike auto mechanics or computer salesmen, deal with life-or-death issues. However, that does not mean that the ethical dilemmas of the health care profession are more challenging than those in other fields. Most of the ethical dilemmas for doctors do not involve life-or-death decisions. An ethical issue might be whether to fudge a diagnosis for insurance purposes, whether to order an unnecessary procedure or follow-up visit in order to pad the bill, etc.


Certainly there are instances in which treatment decisions affect the quality and length of life and also raise unusual ethical issues. But the nature and extent of such conflicts needs to be spelled out and documented in order to make a convincing case that the ethical challenges regarding health care and market incentives are truly a major issue in practice.


Economists have come to expect every economic actor to complain that competition is unethical and unfair when it applies to his or her occupation or business. However, it is rare for consumers to share this perception. In fact, when it comes to providing an incentive for sellers to behave ethically, my experience is that market competition is one of the best forces around.


Another argument Relman makes is that health care requires co-ordination. I agree. A couple of years ago, I had a skin irritation, and so I went to a dermatologist. He sent me to an allergy specialist to get patch-tested. The allergy specialist thought that this was unnecessary, and he prescribed an ointment instead. The dermatologist told me to stop using the ointment, and insisted that I needed to get patch-tested. To break the stalemate, I went to a different dermatologist.


This is not the first time I have experienced such behavior. My observation is that when two egotistical doctors disagree, they try to over-ride each other. To me as a patient, this certainly feels like a lack of co-ordination. However, I fail to see what the market has to do with it.


I would speculate that perhaps the same diagnosis applies to the problem of technological inefficiency in medical administration. Many observers complain that the medical care system is backward in its use of information technology, with data records that are not in formats that can be exchanged across users. I do not believe that anyone knows for certain why computer technology is used so poorly in medicine. But it would not surprise me if it turns out that what differentiates medicine from other industries is not the nature of the market but the egos of the practitioners.


If it were impossible for a consumer to co-ordinate a complex service, then we could not hold weddings, build custom houses, or take overseas vacations. In fact, consumers are able to do all of those things. For those who desire a single point of co-ordination, the market offers wedding planners, general contractors, and travel agents. There is no reason to doubt that similar intermediaries could emerge in health care. Some general practitioners are well suited to playing a coordinating role. Others may be too egocentric. Consumers can sort that out.


Old vs. New Economics


Relman mentions an article by Kenneth Arrow, who won the Nobel Prize in economics in 1972. Arrow listed a number of factors that create information uncertainty in health care. However, as I noted above, these factors also exist to some extent in other important markets.


Arrow, a supremely gifted intellect, nonetheless epitomizes the type of economist that I criticized when I wrote about The Emerging Paradigm in economics. Arrow's mathematical models were suited to allocating a given set of resources relative to a given set of consumer preferences. They are not suited to analyzing the process by which new production methods are discovered.


(As an aside, one claim of the mathematical proponents is that their predecessors wrote in imprecise language that was subject to conflicting interpretations. I invite anyone who believes in the clarifying power of formal models to look up Arrow's "Impossibility Theorem," one of the discoveries for which he earned the Nobel Prize. What you will find is that to this day, half a century later, people still cannot agree on what exactly it was that this theorem proved.)


According to the approach that I call Learning Economics, what is important about the market is its ability to operate as a problem-solving mechanism. New ideas are tried and unproductive methods are discarded.


The learning model and the older paradigm suggest different approaches for dealing with markets that are imperfect. The older paradigm suggests a technocratic approach, in which government acts as a benevolent, omniscient dictator, devising a system that overcomes the flaws of the market. That is what Professor Relman proposes -- omniscient, benevolent doctors working for an omniscient, benevolent bureaucracy.


The learning model suggests that the market, which to a technocrat appears to be merely "chaotic," in fact through trial and error gradually generates better approaches for addressing problems. In my view, even if a government technocrat happens to come up with an outstanding solution at any one point in time, the process of further experimentation and adaptation tends to be severely impaired.


Gaming the System


In any situation, participants try to take advantage of features that create opportunities for easy gain. Economists say that people learn to "game the system." For example, in a market where health insurance reimburses health care providers for procedures, a provider could "game the system" by piling on procedures rather than focusing on the health of the patient.


One of the reasons that there is not a perfect market in health care is that quality is difficult to measure. In his article, Relman is guilty of doing something that economists are particularly known for: he assumes away the problem. In the summary of his plan, he writes,


"not-for-profit, prepaid multi-specialty groups of physicians should provide all necessary medical care on the approved list of insured services. The physicians in the groups should be paid salaries from a pool of money that would be a defined percentage of the total patient income received by the group from the central payer. The groups should be privately managed but publicly accountable for the quality of their services..."


For me, this begs the question of how quality will be measured. Clearly, the government agency that Relman envisions will need to perform such measurement, in order to know whether or not physicians are entitled to their salaries.


In my view, if the problem of measuring health care quality had a solution, then this would take care of many of the economic challenges of any health care system, whether it is market-oriented or government run. Improving the measurement of quality is at the crux of any approach for making health care work better. When Professor Relman writes his book, I believe that it would be disingenuous for him not to spell out how quality measurement will be performed.


Depending on how accurately quality is measured, there will be opportunities to game the system. For example, if low infant mortality is considered a measure of quality, then medical practices will tend to steer away from high-risk pregnancies and premature infants. If following conventional procedures is a measure of quality, then doctors will tend to steer away from cases that are complex, unusual, or difficult to diagnose.


The market does not depend on a government agency to measure quality correctly. Certainly, market participants would be happy to use any data and measures that government statisticians are able to develop. In addition, however, the market would encourage private consumer-oriented rating agencies to emerge (some already have). Also, insurance companies have begun gathering data that would allow them to correlate treatments, costs, and outcomes.


Whatever measures of quality emerge as popular in the market, they too will be imperfect and subject to gaming the system. Once sellers learn how to game the system, they will have a stake in keeping the system stable so that they can continue to exploit its flaws.


In a free market, consumer sovereignty and competition tend to create instability when sellers learn to game the system too well. For example, in the automobile market, we have seen the erosion of dealers' ability to game the system and charge high markups. New competitors and information services have empowered consumers.


In a technocratic system, it is more difficult for consumers to exercise countervailing power. Innovative competitors are often precluded by regulation. Suppliers tend to apply concentrated lobbying power to protect their interests, while the diffuse interests of the consumer are poorly represented in the political process.


Centralized, regulated systems look good on paper, and they may be effective as they start. However, market systems learn faster, because competitive innovation prevents a market from getting captured by the incumbents who have learned how to game the system.


In fact, one sign that incumbents are threatened by market innovation is when they assert a need for regulation. When doctors claim that we need a government-run health care system, perhaps this represents their response to Internet medical sites, continuing innovation in pharmaceuticals and medical technology, and other looming threats to their place in the system.


I hope that Professor Relman addresses some of the issues raised here when he completes his book. Otherwise, if all he does is attack what he calls "free-market ideology," then I fear that we will simply talk past one another.



TCS Daily Archives