Double Jeopardy Disaster

Getting a new drug or medical device approved by the FDA is a long and expensive process. The FDA is risk-averse and pays much more attention to the risks of approving a bad drug than to the risks of failing to approve a good drug. As a result, every economist who has ever written a serious analysis of the FDA has come to the conclusion that less regulation would mean more new drugs and more saved lives. (See FDAReview.org for more information. Gary Becker offers a recent statement.).

Approval, however, does not end a firm's problems because even then it faces the risk of a debilitating lawsuit. Consider how bizarre this is: A team of statisticians, physicians and medical researchers pores over years of clinical data to pronounce a product safe (always noting that this means safe relative to the product's expected benefits) and then a jury of 12 randomly selected Joes and Janes second guesses them, awards plaintiffs billions of dollars and drives the firm into bankruptcy. This has happened more than once.

FDA approval ought to be a "safe harbor." Many states already have laws along these lines but they have been weakly enforced. The Bush administration's efforts to limit lawsuits against firms that have passed FDA approval is a therefore a necessary and welcome piece of common sense. This doesn't mean that you can't sue a drug manufacturer. If the manufacturer lies to the FDA or to your physician or if they don't produce the drug according to specification then by all means sue away. Every drug, however, has side-effects and every drug works differently in different people. That means that there has to be some sort of cost-benefit test to decide if a drug should be marketed. There is an argument for using tort law instead of the FDA to do this test - an argument that gets weaker the more out out-of-control the courts become - and there is an argument for using the FDA instead of tort law but there is no argument for adding tort law on top of FDA regulation, that is a double jeopardy disaster.