Content area
Full Text
Joseph G. Eisenhauer*
Abstract:
This paper critically evaluates a recently introduced theory of the demand for health insurance. The new model makes a valuable contribution to the theoretical literature on moral hazard and the debate over national health insurance, but as a general explanation of the demand for health insurance it fails to provide a robust alternative to conventional theory.
[Key words: Moral hazard, medical care, risk aversion]
INTRODUCTION
In a justly famous article published more than four decades ago, Kenneth J. Arrow (1963) argued that where private markets for insurance-particularly health insurance-were absent, a strong case could be made for governmental provision of insurance. In an almost equally well known comment five years later, Mark V. Pauly (1968) observed that health insurance often induces moral hazard, resulting in an inefficient reallocation of resources, and that institutionalizing such inefficiency through government regulation could potentially be welfare-reducing. Thus, moral hazard weakened the case for national health insurance.1
John A. Nyman's (2003) book, The Theory of Demand for Health Insurance, reconsiders moral hazard and offers a new perspective on the reason why consumers buy medical insurance in the first place. His explanation is a departure from conventional views, and Professor Nyman acknowledges at the outset that his position is controversial in several respects. Though it is framed largely as a refutation of Pauly's analysis, the book is by no means an endorsement of Arrow's work; indeed, it rejects the fundamental notion that risk aversion motivates insurance purchases. A treatise that seeks to overturn established theory so thoroughly demands scrutiny, and to that end, the present paper briefly reviews the literature leading up to Nyman's book, evaluates the author's critique of standard theory, and then examines his proposed alternative.
THE MORAL HAZARD DEBATE
Pauly's (1968) essay assumed a fixed individual demand curve for health care and a constant marginal cost of production. Together, these determined an efficient optimum for an uninsured patient: the marginal willingness to pay for care (as represented by the demand curve) was equal to the marginal cost of care. If the same individual were insured, however, she would perceive a lower out-of-pocket price for care (zero, if there was no coinsurance), and move down the demand curve; unless demand had no price-elasticity, the...