A Question about the FDA

by Don Boudreaux on January 13, 2005

in FDA, Risk and Safety

In response to Russ Roberts’s commentary that aired Tuesday on NPR’s Morning Edition — and to my posting it here on Cafe Hayek — I’ve gotten a surprising amount of feedback.  Most of it positive, none of it hostile.  But some of it skeptical.

"Can people really be trusted to make such life-and-death decisions?" is the best one-line summary of this skepticism.

Much can said in favor of answering ‘yes’ to this question — and perhaps I’ll add more in a later post.  (Hint: any such answer begins with its own question: As compared to what?)  But recognizing that ‘irrational’ behavior does exist — pretty foolish to deny it — I here ask you to ponder the following:

Suppose that Mr. Jones is so eager for relief of his arthritis pain that he downs Vioxx before seeking the advice of an MD or searching out other sources of information on it.  Further suppose that had Mr. Jones known of the likely side-effects of Vioxx, he would voluntarily refrain from taking it.  That is, he takes it ‘foolishly.’  Mr. Jones is precisely the sort of person that FDA-proponents want to protect.

Suppose also that Ms. Smith has her own ‘irrationality’ — her own mistaken knowledge or distorted judgment.  This distorted judgment is such a deep suspicion of profit-seeking pharmaceutical companies that she refuses to take a drug that will save her life even though, in fact, this drug has almost no serious side-effects.  She will die soon simply because she refuses to take this drug.  And, were she ‘rationally’ better informed about the true state of the world, she would take the drug; she doesn’t want to die.  Her irrationality, her mis-information, her distorted judgment will kill her.

If we correct for Mr. Jones’s irrationality by preventing him from taking Vioxx, should we also correct for Ms. Smith’s irrationality by forcing her to take the drug that will save her life?  If not, why not?


1 comment    Share Share    Print    Email

Previous post:

Next post: