All Commentary
Thursday, August 1, 2002

Free-Market Medicine

Why Has the Health-Care Market Failed to Produce High-Quality, Low-Cost Medicine?

Larry Van Heerden is the author of website Free Market Medicine (

The health-care system in the United States is beset by problems. After years of feeling shortchanged by managed care, doctors and hospitals are demanding and getting greater compensation; the elderly (under Medicare) have no prescription coverage; and many people find health insurance of any kind unaffordable. Managed care, which was hailed as the answer to spiraling costs, is under legislative and legal assault, while health-care costs are rising at double-digit rates. Proposed solutions range from a Canadian-style single-payer system to medical savings accounts to staying the course with managed care.

First, before discussing these issues, some characteristics of the health-care system that get little attention:

Inefficiency. In 1999, health-care costs in the United States totaled about $1.2 trillion (13 percent of GDP). One measure of the medical industry’s inefficiency can be gleaned from examining the experience of a 2,000-patient clinic in Renton, Washington. Using a concept known as SimpleCare, the clinic has been able to reduce the price of outpatient care by requiring patients to pay in full at the time they are seen. The clinic charges 30-50 percent less than what a regular clinic would charge for such things as office visits and X-rays. The savings come from bypassing the bureaucracy: No staff are needed to get managed-care approval, file endless claims, or coax payment from insurance companies. (Patients are encouraged to get low-cost insurance to cover catastrophic illness.)1

Preventable Disease. Tobacco use, diet, obesity, and lack of exercise can be linked to nearly two-thirds of cancer deaths in the United States.2 A proper diet can substantially lower high blood pressure.3 Sedentary habits account for a substantial portion of deaths due to coronary artery disease, type 2 diabetes, as well as loss of functional ability in older adults.4 Osteoporosis can be prevented by not smoking, consuming calcium and vitamin D, engaging in weight-bearing exercise, limiting alcohol and caffeine consumption, and (for some women) taking supplemental estrogen.5 According to Dr. Christopher Fanta, head of the asthma treatment program at Brigham and Women’s Hospital, “Asthma is considered the most common preventable cause of hospitalization.”6 Fewer than a third of women of childbearing age take the vitamin folic acid, which has been shown to prevent disabling birth defects.7

Although making lifestyle changes that prevent disease is not easy, the fact remains that consumers have neither the financial motivation nor the access to information that would allow them to select doctors who have a proven track record in educating and motivating their patients.

Unnecessary Treatment. Studies have shown that various medical procedures and regimens are overused. Examples include treatment for Lyme disease,8 back surgery,9 myringotomies (insertion of tiny tubes to prevent ear infection),10 hysterectomies,11 and laparoscopic gallbladder and anti-reflux surgery.12 A common assumption is that treatment is driven by medical necessity. However, in a series of studies, Dr. John E. Wennberg found that the amount of health care consumed by individuals is highly dependent on where they live, on the capacity of the local health-care system, and on the practice styles of local physicians.13

Poor Performance. A substantial number of physicians are falling short in the practice of medicine. Examples include poor control of hypertension,14 potential errors made by sleep-deprived interns,15 lack of cardiac stethoscope proficiency,16 under-use of beta blockers for preventing subsequent heart attacks,17 inadequate control of cholesterol,18 under-use of blood-clot dissolving therapy,19 inadequate pain management in cancer patients,20 and poor communication with patients.21

Imperfect Knowledge

Medical Uncertainty. Standardized algorithms and guidelines based on scientific evidence have shown themselves to be successful in treating disease and saving lives. However, in a 1997 article Dr. David Mirvis made these points: Uncertainty remains a fact of life in medical practice. Physicians don’t know the full extent of a patient’s disease or the best single approach to diagnosis and treatment. Biological variability is a major challenge. The same condition may result in different outcomes in different patients due to genetic predispositions, co-morbidity, or other factors. Knowledge of the characteristics of disease and the effects of therapies is far from complete. Diagnostic tests are imperfect and many treatments are based not on scientific principles but on learned practices, assumptions, and other nonscientific formulations. Much uncertainty results from lack of knowledge of the outcomes of what physicians do. A consequence of the high level of medical uncertainty is the existence of a range, rather than a single point, of acceptable practice. This range is broad enough to allow substantial variations in practice patterns.22

State Mandates. A 1989 study found that states have passed more than 700 statutes requiring health insurers to cover specific providers, diseases, or high-risk individuals. Mandated benefits run the gamut from hair-transplant and acupuncture coverage to mental health and maternity coverage.23 A follow-up study, published in 1999, found that 20 to 25 percent of uninsured Americans lack coverage because of benefit mandates. The proliferation of mandated benefits has increased the cost of health insurance, disproportionately hurting employees who work for small businesses, which can’t afford to self-insure and thereby avoid the mandates.24

Bone marrow transplantation, for example, is a complex, expensive ($60,000), and painful experimental treatment for late-stage breast cancer, which places the patient at high risk for infection. Beginning in the early 1990s, after preliminary indications of promising results, lawsuits and political pressure led to legislation in ten states, forcing insurers to cover the procedure. Such laws made it difficult to recruit women into clinical trials because women did not want to take the chance of being “randomized” into a control group. By 1999, 5,000 women a year were undergoing transplants nationwide. By the spring of 2000 it became clear that such transplants were no more effective than chemotherapy alone at standard doses.25

Income Tax Distortion. Many writers have pointed out that for the 90 percent of people who obtain their health insurance through their employers, the tax-exempt status of employer-provided health insurance encourages a huge overconsumption of health insurance and hence overconsumption of health-care services, compared to how people would spend their money in the absence of tax exemption.26 For someone in a 33 percent tax bracket, for example, $2 of taxable income turns into $3 of tax-free income, which can be spent on health insurance without any change in disposable income.

Stiffing the Piper. Health-care costs have risen dramatically in the last few decades, while patients’ out-of-pocket share of those expenses has declined: In 1960 patients paid, on average, 20.8 percent of hospital costs and 61.6 percent of physician costs. By 1999 those figures had dropped to 3.2 percent and 11.4 percent, respectively.27

Failed Prescriptions

Managed care, which came into prominence in the 1990s, was initially successful at holding down health-care costs. However, doctors, hospitals, and patients were soon fighting back, and the inherent weaknesses of third-party control were revealed: By requiring patients to pay no more than a token copayment, managed care removes the incentive to economize and undermines patient control of health-care encounters. The central tenet of managed care is that consumers are ill-equipped to deal directly with health-care providers; managed-care organizations must act as intermediaries, handling the complexities of medical payment and quality assessment, leaving consumers to make their wishes known by choosing from a list of rival health plans provided by their employers. This is an anemic form of competition, which is as effective in securing cost-effective health care as a passenger would be in arriving at his destination by telling a blindfolded driver when to step on the accelerator, hit the brake, or turn the steering wheel. To achieve the goal of cost-effective care, consumers need to choose at the level of the individual provider and medical procedure, and face both the costs and benefits of their choices.

A more fundamental problem with managed care is that many medical decisions fall into a gray area where definitive scientific judgment cannot be rendered for individual cases. This gray area is the subject of a tug of war between patients and managed-care administrators. Patients, many of whom are being treated for diseases partly of their own making, want no expense to be spared in their treatment, since someone else is footing the bill. Managed-care organizations, on the other hand, make money (or stay solvent) by limiting the amount of care rendered to subscribers. This gray area is large enough to make the difference between financial success and failure for the organizations and large enough to give patients who are denied care plenty of ammunition when seeking legislative and legal action against those organizations.

A Canadian-style single-payer health care system is nothing more than a massive managed-care arrangement with government bureaucrats in control and without a meaningful appeals process for care denied or delayed. The same problems inherent in private managed care arise in a government-run system. Moreover, as a rule, government programs cannot satisfy consumer demand. Since all goods and services are finite and require human effort to produce, rationing is unavoidable. Only the method of rationing is subject to choice. The free market rations on the basis of income; the method of rationing is the familiar pricing system. When this system is circumvented by the government to provide a “free” good or service, all constraints on demand are removed, making inevitable the explicit rationing of supply by some government authority or the disappearance of the good or service altogether. Regarding universal access to health care, it should be noted that before government intervention in the health-care system, a variety of private organizations provided free medical care to the poor.28

Medical Savings Accounts

Medical savings account (MSA) health plans were introduced at the federal level as a demonstration project in 1996. The central feature of these employer-provided plans is a savings account controlled by the insured individual and used to pay for routine health care. An accompanying low-cost catastrophic insurance policy covers health- care expenses that exceed the high yearly deductible. MSA plans enjoy the same tax advantage as other employer-provided health insurance. Although unspent MSA funds roll over from year to year, they can only be spent on health care.29

The high deductible associated with MSA health plans leads to substantial savings in administrative costs because many low-dollar claims for routine medical care are never filed. In addition, having patients spending their own money on health care makes them more prudent consumers, which means less spending on unnecessary health care services.

However, MSA health plans have drawbacks. They perpetuate the income tax distortion of health-care spending (discussed above) and are subject to legislative manipulation: Under current law, MSA plans are hamstrung by limited availability and growth, unnecessary complexity, and design features that put them at a disadvantage in the marketplace.30 Finally, MSA critics argue that the very idea of government direction or control of consumer spending is inimical to a free market.

The Solution

The solution to the problems discussed above is to treat health care more like other products and services. This means repealing all tax exemptions for health insurance and health-care spending, enacting a compensating tax cut unrelated to individual health-care consumption, eliminating all health-insurance mandates and other regulation, and letting the market sort things out.

The market would probably respond to such deregulation the same way it did before government intervened in health care: As early as the 1940s commercial insurers included deductibles and copayments in their sickness insurance offerings and excluded many elective treatments from coverage, all in an effort to restrain demand for unnecessary and costly medical services. Commercial insurers also used actuarial risks to calculate premium payments and paid individual subscribers, instead of hospitals.31 Giving patients a substantial financial stake in the cost of their care will make them interested in the cost-effectiveness of that care.

A second part of a market-driven solution would likely be giving patients access to information comparing the performance of competing physicians, just as consumer magazines provide information on competing products. To do this, independent organizations might determine what physicians accomplish in a clinical setting by measuring the health status of patients before and after treatment. To be cost-effective, such measurement would probably make use of electronic medical records.

Employees on expense accounts spend much more freely than when making purchases with their own hard-earned money. Similarly, patients consume medical services with little regard for cost when someone else is paying for them. In a climate of unnecessary medical care, preventable disease, and medical uncertainty, insulating consumers from the cost of choices they or their doctors make guarantees inefficiency and runaway costs.

Cost-sharing refers to the requirement that patients bear a significant share of the cost of all medical care rendered in their behalf. It does not refer to paying insurance premiums (which do nothing to constrain health-care consumption). The RAND health-insurance experiment showed that patients who have to pay for part of their care cut back substantially on the use of medical services.32 While the market will figure out the right mix of deductibles and copayments, it seems likely that as an individual’s yearly health-care expenses rise, his out-of-pocket share of new health-care expenditures will decline. However, from an economic point of view, it would be optimal if no one’s out-of-pocket share of medical expenses ever dropped to zero, giving every consumer a stake in the cost of every medical visit, test, procedure, hospitalization, or prescription drug he consumes. An immediate effect of such cost-sharing would be to give physicians a newly found interest in cost control for the benefit of their patients and as a means to attract business.

Measuring Physician Performance

In many cases, poor physician performance (discussed above) is due to lack of feedback; in the absence of evidence to the contrary, it is natural to assume that one’s performance is adequate or even superior to that of one’s peers.

In an attempt to control costs, managed-care organizations have been measuring the process of health-care delivery, rather than identifying physicians who keep their patients healthy. In a newly deregulated market, one can imagine managed-care organizations dropping their review and oversight functions in favor of collecting and disseminating (for a fee) information on the performance of physicians (and eventually hospitals). If such a service were to periodically measure a patient’s health status during the course of treatment, the change in these measurements, collected for a sufficient number of patients (and adjusted for severity of illness, co-morbidity, and patient demographics), could be used to measure doctors based on the results they achieve in their patients.

Assessing outcomes is appealing because of its narrow focus: As long as a patient’s health status can be objectively measured, none of the intervening steps that are part of medical treatment need be evaluated. Concern about the number and type of tests performed, improper use of high-tech equipment, medications prescribed, or the appropriateness of the treatment chosen would be superfluous. Poor choices by a physician in such matters would either be reflected in higher costs or worse outcomes than those of other physicians.

For preventive medicine and most chronic diseases the performance of physicians is inextricably linked to patient compliance and cooperation. As a result, the performance of physician and patient would probably have to be measured jointly. Nonetheless, in the context of a system controlled by any third party, measuring a physician based on the behavior of his patients would likely be unacceptable to the marketplace.

However, in a health-care system that includes patient cost-sharing, measuring the performance of physician and patient jointly makes sense. No third-party coercion would be needed; a patient’s financial stake in the cost of care would serve as a necessary and sufficient constraint on his behavior.

Doctors with performance scores showing a low-key approach to smoking cessation, for example, would attract patients with no intention of quitting. Such doctors would suffer no financial penalty other than that brought on by the indifference of smokers looking for a more aggressive approach to their problem. Moreover, doctors with low cessation scores who have high scores in other areas would attract nonsmokers who have no reason to care about quitting.

Meanwhile, patients with a desire to quit smoking, lose weight, improve their diet, or begin an exercise program could choose physicians with performance scores showing a successful track record in preventive medicine. And doctors who felt certain patients weren’t living up to their end of the bargain (regarding compliance with treatment plans) would be free to refer them elsewhere. Thus the goals of physician and patient would be in alignment.

Indicators to be measured would probably be those known to be closely related to good health and closely related in time to physician intervention. Examples of possible indicators are cholesterol levels, blood pressure, blood-glucose levels, and patient satisfaction.

The government would play an important role in establishing and enforcing a patient’s right to control his medical information. Beyond that, patient privacy would be protected because the measurement system would not need identifying information.

The health-care market has failed to produce high-quality, low-cost medicine for two reasons: Consumers are insulated from the cost of medical care by third-party payers, and information on the performance of competing physicians is not available. Fixing the incentives and providing consumers with physician performance data will cause unnecessary surgery to decline, physician performance to improve, disease prevention to increase, and health-care efficiency to rise.


  1. See
  2. “Harvard Report on Cancer Prevention. Volume 1: Causes of Human Cancer,” Cancer Causes and Control, November 1996, Suppl. 1, p. S55.
  3. Lawrence J. Appel et al., “A Clinical Trial of the Effects of Dietary Patterns on Blood Pressure,” New England Journal of Medicine, April 17, 1997, p. 1123.
  4. Andrea L. Dunn et al., “Comparison of Lifestyle and Structured Interventions to Increase Physical Activity and Cardiorespiratory Fitness,” JAMA, January 27, 1999, p. 327.
  5. “Osteoporosis: It’s Never Too Late to Protect Your Bones,” Mayo Clinic Health Letter (supplemental medical essay), October 1997, pp. 1-8.
  6. Larry Tye, “Asthma Figures Double Since 1980,” Boston Globe, February 25, 1998, p. B1.
  7. “Women Are Told Anew of Need for Folates to Stem Birth Defects,” Boston Globe (Reuters), August 3, 1999, p. A10.
  8. M. Carrington Ried et al., “The Consequences of Overdiagnosis and Overtreatment of Lyme Disease: An Observational Study,” Annals of Internal Medicine, March 1, 1998, pp. 354-62.
  9. “Needless Surgery, Needless Drugs,” Consumer Reports on Health, March 1998, pp. 1-4.
  10. L.C. Kleinman et al., “The Medical Appropriateness of Tympanostomy Tubes Proposed for Children Younger than 16 Years in the United States,” JAMA, April 27, 1994, p. 1250.
  11. M.S. Broder et al., “The Appropriateness of Recommendations for Hysterectomy,” Obstetrics and Gynecology, February 2000, p. 199.
  12. “Keyhole Surgery: Promise and Pitfalls,” Consumer Reports on Health, February 2002, pp. 8-10.
  13. See
  14. Dan R. Berlowitz et al., “Inadequate Management of Blood Pressure in a Hypertensive Population,” New England Journal of Medicine, December 31, 1998, pp. 1957-63.
  15. N.J. Taffinder et al., “Effect of Sleep Deprivation on Surgeons’ Dexterity on Laparoscopy Simulator” (Research letter), Lancet, October 10, 1998, p. 1191; Steven R. Daugherty et al., “Learning, Satisfaction, and Mistreatment During Medical Internship,” JAMA, April 15, 1998, pp. 1194-99.
  16. Salvatore Mangione and Linda Z. Nieman, “Cardiac Auscultatory Skills of Internal Medicine and Family Practice Trainees,” JAMA, September 3, 1997, pp. 717-22.
  17. Harlan M. Krumholz, et al., “National Use and Effectiveness of ß-Blockers for the Treatment of Elderly Patients After Acute Myocardial Infarction,” JAMA, August 19, 1998, pp. 623, 627.
  18. Richard A. Knox, “Heart Study at 50: New Goal,” Boston Globe, September 27, 1998, p. A1.
  19. Harlan M. Krumholz et al., “Thrombolytic Therapy for Eligible Elderly Patients with Acute Myocardial Infarction,” JAMA, June 4, 1997, pp. 1683-88.
  20. Charles S. Cleeland et al., “Pain and Its Treatment in Outpatients with Metastatic Cancer,” New England Journal of Medicine, March 3, 1994, pp. 592-96.
  21. Clarence H. Braddock III et al., “Informed Decision Making in Outpatient Practice,” JAMA, December 22/29, 1999, pp. 2313-20.
  22. David M. Mirvis, “Managed Care, Managing Uncertainty,” Archives of Internal Medicine, February 24, 1997, pp. 385-88.
  23. Jon R. Gabel and Gail A. Jensen, “The Price of State Mandated Benefits,” Inquiry, Winter 1989, pp. 419-31.
  24. Gail A. Jensen and Michael A. Morrisey, “Mandated Benefits Laws and Employer-Sponsored Health Insurance,” Health Insurance Association of America, January 1999,
  25. Richard Saltus, “Last-Hope Cancer Treatment ‘Ineffective’” Boston Globe, March 4, 2000, p. A1; E.A. Stadtmauer et al., “Conventional-Dose Chemotherapy Compared with High-Dose Chemotherapy plus Autologous Hematopoietic Stem-Cell Transplantation for Metastatic Breast Cancer. Philadelphia Bone Marrow Transplant Group,” New England Journal of Medicine, April 13, 2000, pp. 1069-76.
  26. “Subsidizing Employer Provided Health Plans Unfair to Individuals Who Don’t Have One,” National Center for Policy Analysis,
  27. M.S. Eberhardt, D.D. Ingram, D.M. Makuc et al., Urban and Rural Health Chartbook: Health, United States, 2001 (Hyattsville, Md.: National Center for Health Statistics, 2001), Table 118, p. 333.
  28. Marvin Olasky, The Tragedy of American Compassion (Washington D.C.: Regnery Gateway, 1992), pp. 82-83.
  29. Victoria Craig Bunce, “Medical Savings Accounts: Progress and Problems under HIPAA,” Cato Institute Policy Analysis No. 411, August 8, 2001.
  30. Ibid.
  31. Susan Feigenbaum, “‘Body Shop’ Economics,” Regulation, Fall 1992, pp. 26-27; John C. Goodman and Gerald L. Musgrave, Patient Power (Washington D.C.: Cato Institute, 1992), pp. 187-88; Terree P. Wasley, What Has Government Done To Our Health Care? (Washington D.C.: Cato Institute, 1992), p. 49.
  32. Emmett B. Keeler and John E. Rolph, “How Cost Sharing Reduced Medical Spending of Participants in the Health Insurance Experiment,” JAMA, April 22, 1983, p. 2220.