Journal of Health Politics, Policy and Law 26.5 (2001) 1197-1203
[Access article in PDF]
Reflections on the Reflections
Kenneth J. Arrow
An author cannot help being flattered by the attention paid in this issue. Disagreement and criticism are almost as welcome as praise and extension; the important thing to any author is that his or her ideas matter and are important enough to fight about. Further, these essays offered me a very informative history of the last forty years in health care in the United States, from which I have learned much.
A detailed analysis of the writings in this volume would not be possible in any brief compass. Enough new ideas and challenges have been raised here to lead to a short volume in terms of specific agreements, disagreements, and extensions. My response takes the form of five somewhat linked sets of remarks on the economics of medical care and on more general aspects of economics in which similar themes emerge.
As the authors well know, I have not engaged in research on medical care for a long time and so am certainly not up to date on many specific empirical aspects of the subject. I have some impressions, and I will not let my lack of knowledge preclude my raising points, secure in the understanding that I am privileged to ask questions without knowing the answers.
I am stimulated to respond to critical comments about the relevance of my analysis to the present and recent states of the medical care industry. Some of my remarks are hypotheses about the evolution in the medical care system; some, clarifications or extensions of the positions in my first article on medical care. [End Page 1197]
I will restate my thesis in a few words: The role of moral hazard in medical insurance arises from inequalities of information between the insurer on one hand and the physician and patient on the other. By itself, this phenomenon was well known in other kinds of insurance (where the term moral hazard arose) and was met by various devices, such as deductibles and ceilings. Direct controls came later, as I conjectured. I then came to the view that the same degree of asymmetric information also occurred in the relation between patients and providers. I proposed the hypothesis that many of the special features of the medical care market arose in an attempt to meet the inefficiencies created by this generalized moral hazard.
The Increased Demand for Medical Services
One point made in several of the articles is the greatly increased scale of expenditures for medical services. There were plenty of concerns about the great expense of medical treatment in 1963, when the costs were about 4 percent of U.S. national income. Today the costs reach 14 percent or so of national income, and the fear is that still further increases will occur. Prices of medical personal services did rise, at least for a while. The consensus, though, seems to be that the increased availability of expensive technologies is a bigger factor now than increased wages.
One obvious factor, of course, is the great extension of government insurance in the form of Medicare and Medicaid. Clearly, the demand for medical care rises sharply and becomes markedly less elastic under these conditions; that is one of the basic implications of moral hazard in medical insurance. The demands for the services of both labor and capital will clearly rise under these conditions.
But the effect on technology is multiplied by innovation. There has been a large development of new technologies, especially in imaging, and these apparently are very costly. These become medically valuable and even expected, especially under conditions of inelastic demand. Why this sudden spate of innovations? Of course, it could be chance. Undoubtedly, also, the very large expenditures on basic research in medicine, through the National Institutes of Health, certainly must have laid the groundwork for the development of new projects.
There is another systematic reason, which may be relevant here. Ever since Adam Smith, it has been repeatedly asserted that increases in scale lead...