By John A. Nyman
Journal of Health Policy, Politics and Law
October 2007
Abstract
Much American health policy over the past thirty-five years has focused on reducing the additional health care that is consumed when a person becomes insured, that is, reducing moral hazard. According to conventional theory, all of moral hazard represents a welfare loss to society because its cost exceeds its value. Empirical support for this theory has been provided by the RAND Health Insurance Experiment, which found that moral hazard — even moral hazard in the form of effective and appropriate hospital procedures — could be reduced substantially using cost-sharing policies with little or no measurable effect on health.
This article critically analyzes these two cornerstones of American health policy. It holds that a large portion of moral hazard actually represents health care that ill consumers would not otherwise have access to without the income that is transferred to them through insurance. This portion of moral hazard is efficient and generates a welfare gain. Further, it holds that the RAND experiment’s finding (that health care could be reduced substantially with little or no effect on health) may actually be caused by the large number of participants who voluntarily dropped out of the cost-sharing arms of the experiment. Indeed, almost all of the reduction in hospital use in the cost-sharing plans could be attributed to this voluntary attrition. If so, the RAND finding that cost sharing could reduce health care utilization, especially utilization in the form of effective and appropriate hospital procedures, with no appreciable effect on health is spurious.
The article concludes by observing that the preoccupation with moral hazard is misplaced and has worked to obscure policies that would better reduce health care expenditures. It has also led us away from policies that would extend insurance coverage to the uninsured.
Introduction
The direction of health policy in the United States over the past thirty-five years is well known: increased cost sharing in the form of higher deductibles and coinsurance rates; the “management” of care by utilization reviews, exclusive contracting with certain providers, capitation, and bundling of services; and most recently health savings accounts (HSAs) and “consumer-driven” health care. Less well understood, however, are the foundations of these policies, especially among those who see such policies as mainly restricting access to care.
Two of these foundational studies are (1) the theoretical analysis of the welfare implications of moral hazard (Pauly 1968; Feldstein 1973) and (2) the RAND Health Insurance Experiment (HIE) findings concerning the health consequences of reducing moral hazard (Manning et al. 1987; Newhouse and the Insurance Experiment Group 1993).
This article presents a critical view of these two foundational studies. First, it suggests that much of the additional health care that consumers purchase when they are insured — that is, much of moral hazard — is actually efficient and welfare increasing: its value to patients exceeds (and often far exceeds, in the case of expensive, life-saving, hospital procedures that patients would not be able to afford without insurance) the cost of providing that care. Second, it suggests that the RAND HIE did not really capture consumers’ willingness to substitute other goods and services for medical care in response to increases in the price of medical care, but instead the study largely captured participants who became ill, dropped out of the experiment, and received the needed medical care under their original insurance policies outside the experiment. Because these participants received treatment outside the RAND experiment, their health care utilization was not recorded. Thus, the finding that moral hazard, especially the portion of moral hazard represented by additional hospital procedures, could be reduced dramatically by cost sharing with only a negligible health effect is spurious.
New Theory
The fundamental error in conventional theory is that it did not recognize that, while the price of medical care might drop to zero for all those who are insured, for most medical care — especially the expensive, hospital-based procedures and associated care that comprise the bulk of medical care expenditures in the United States — it is really only those who are ill who respond to that price reduction. For example, what healthy person would purchase a coronary bypass procedure, a leg amputation, or a liver transplant just because the price has fallen to zero? This means that only those who become ill are responsive to the insurance price reduction. If so, the price reduction becomes the vehicle by which income is transferred from those who purchase insurance and remain healthy to those who purchase insurance and become ill. This is an important distinction because it changes the welfare implications of moral hazard dramatically and implies that much, if not most, of moral hazard is efficient and generates a welfare gain (Nyman 1999, 2003).
New Interpretation
Of the various responses to cost sharing that were observed in the participants of the RAND HIE, by far the strongest and most dramatic was in the relative number of RAND participants who voluntarily dropped out of the study over the course of the experiment. Of the 1,294 adult participants who were randomly assigned to the free plan, 5 participants (0.4 percent) left the experiment voluntarily during the observation period, while of the 2,664 who were assigned to any of the cost-sharing plans, 179 participants (6.7 percent) voluntarily left the experiment. This represented a greater than sixteenfold increase in the percentage of dropouts, a difference that was highly significant and a magnitude of response that was nowhere else duplicated in the experiment.
The explanation that makes the most sense is that the dropouts were participants who had just been diagnosed with an illness that would require a costly hospital procedure. If they dropped out, their coverage would automatically revert to their original insurance policies, which were likely to cover major medical expenses (such as hospitalizations) with no co-payments. This is because HIE participants had agreed to relinquish their existing policies to RAND as a condition of participation, but they could revert back to their existing policies whenever they wanted if they dropped out of the experiment. Thus, when faced with a large hospital expense, the participants could have simply dropped out of the experiment but received the inpatient care outside the experiment.
As a result of dropping out, these participants’ inpatient stays (and associated health care spending) did not register in the experiment, and it appeared as if participants in the cost-sharing group had a lower rate of inpatient use. In reality, however, this finding would have been because participants who remained in the cost-sharing group simply had fewer diagnoses that required hospitalization.
Conclusion
The conventional theory of the moral-hazard welfare loss and the conventional interpretation of the RAND HIE results have created a policy environment in which health insurance is viewed as more of a problem than a solution (Gladwell 2005). We must recognize, however, that much of moral hazard generates a welfare gain that society should encourage rather than discourage. Because of this reevaluation of moral hazard, we must also realize that health insurance is much more valuable than has been recognized. Indeed, there is probably no other investment that we can make as a society that would generate as great a net return on welfare as finding a politically acceptable mechanism for insuring the large portion of U.S. citizens who are currently uninsured.
Abstract:
http://jhppl.dukejournals.org/cgi/content/abstract/32/5/759
Full article (25 pages):
http://jhppl.dukejournals.org/cgi/reprint/32/5/759
Comment:
By Don McCanne, MD
This landmark report turns upside down the policy basis of moral hazard. The entire article should be downloaded so that it can be used to refute those who insist that patients who decline health care when they have to pay for it out-of-pocket will not adversely affect their own health because they will decline only that care that has no measurable benefit on their health outcomes.
These two cornerstones of American health policy, the moral hazard theory and the RAND Health Insurance Experiment, have crumbled, taking down with them the rationale of the cost-sharing basis of the consumer-directed health care movement.
Professor Nyman has shown us that the new interpretation of the moral hazard of insurance insulation against health care costs does not produce a welfare loss, but, in fact, produces a welfare gain. That is why we need a national health program that covers all of us and that eliminates financial barriers to beneficial services.
We really don’t have to worry, after all, about unleashing a stampede for free leg amputations or free liver transplants.