Where Is the Market Failure Argument for the Interventions We Have?

According to the Food and Drug Administration (FDA), some 20 cents of every consumer dollar purchases products that come under the purview of the FDA. It regulates the safety of America’s food supply and cosmetics, the safety and effectiveness of pharmaceuticals and medical devices, and the claims that can be made about these and other products. Specific FDA interventions include the permitting of new drugs, control of manufacturer speech, and the imposition of prescription requirements. We argue that there is no market-failure rationale for these three interventions.

The quotations section of this Web site demonstrates that economists who express judgments on the FDA in print are overwhelmingly supportive of liberalization. Some economists deny that liberalization is strongly supported by “economics,” but at their heart they seem to agree that there is no respectable market-failure rationale.

A rationale for the observed restrictions would need to assert, at least somewhat plausibly, that what amounts to FDA veto power somehow corrects systematic erring in the face of such grave uncertainty. But no grounds are ever offered for any such systematic erring, much less for the corrective effect of FDA veto power.

To say there is a market-failure rationale for layering premarket approval on top of prescription requirements is to say that doctors systematically fail in prescribing new drugs and that premarket approval somehow corrects such systematic erring by doctors. But is there any evidence of such systematic erring by doctors, or of the corrective nature of premarket approval? No. Nor is there any sensible theoretical basis for claiming that doctors systematically err in prescribing drugs to their patients—particularly in light of the incentives exerted by possible malpractice suits.

Litigation weighs heavily on product safety and medical treatment. More importantly, people have demands for ex ante assurance of quality and safety, and those demands give rise to supplies. Reputation is but one form of assurance, and it suffers by product recalls (Jarrell and Peltzman 1985). Assurance and litigation do not work perfectly. But there is no theory contending that they err systematically and that FDA policies constitute any plausible correction to such erring. A defense of government intervention must open with a rationale and proceed to a reasonably well-rounded case—a case that recognizes important costs and the imperfections of the alternative arrangement. As regards the longstanding policies, we do not find even a coherent opening rationale, much less a well-rounded case.

A market-failure rationale requires a plausible story about how government, with its special abilities, might improve matters. It is not legitimate to look at the situation, notice risk, ignorance, and uncertainty, and declare “failure.” If we go down a checklist of market-failure rationales—adverse selection, externalities, natural monopoly, equity, etc.—we do not find one that can be invoked for FDA policies. Danzon and Keuffel (2007) argue that safety and efficacy information has public goods properties. Surely there is something to this, just as the knowledge conveyed by Underwriters’ Laboratories certification may be said to have some public goods properties. But, at best, public good aspects of information create an argument for information provision, not for the premarket banning of all pharmaceuticals.