Labs Benched

In late July, two seemingly unconnected events transpired within hours of each other. First, President Bush nominated John Roberts, a conservative whose ideology-to-intelligence ratio remains unknown and perhaps unknowable, for the Supreme Court. Second, the American Journal of Public Health released a special supplement issue on scientific evidence and public policy, largely dedicated to excoriating a little-known but highly consequential 1993 Supreme Court ruling on the admissibility of expert testimony in court, known as Daubert v. Merrell Dow Pharmaceuticals.

After discussing the perils of Daubert, the connection between these two events should become clear.

Daubert was a "toxic tort" case, a type of litigation in which plaintiffs claim that a particular company's actions or products damaged their health. The product at issue in Daubert was Merrell Dow's Benedictin, a drug for morning sickness that, plaintiffs claimed, caused serious birth defects. To prove their case, the plaintiffs had to present evidence on the chemical and pharmacological nature of Benedictin, as well as expert testimony in the field of epidemiology in order to demonstrate scientific causation -- i.e., that when taken by pregnant women, Benedictin can indeed induce birth defects.

Lower courts in the case had found that the plaintiffs failed their evidentiary test, ruling that the expert testimony they presented was inadmissible under the 1923 Frye v. United States standard, which required such testimony to be based on concepts that are "generally accepted" in the scientific community. Into this fray leaped the U.S. Supreme Court, overturning the Frye test and substituting a new one. Writing for the majority, Justice Harry Blackmun (of Roe v. Wade fame) laid out what he viewed as the "scientific method" as a means of determining what sort of expert testimony ought to be deemed reliable by courts. Testimony, according to the Court, should be measured by standards of testability, peer review, the disclosure of an error rate, and (as in Frye) "general acceptance" by the scientific community.

Most significantly, Daubert tasked trial-court judges with the duty of serving as "gatekeepers" over expert testimony, which they were to evaluate according to the Supreme Court's standards. In essence, then, the Court attempted to lay out a test that would allow other judges to determine what kind of evidence should count as "scientific" long before a jury had any opportunity to hear the evidence itself during a trial.

If that sounds like a perilous endeavor, the Daubert opinion more than substantiates that concern. As the American Journal of Public Health supplement exhaustively shows, the Supreme Court blundered miserably on a variety of fronts. It propounded a deeply confused philosophy of science, merging together the incompatible views of Karl Popper and Carl Hempel. It was equally naive about the sociological conditions in which scientific information is produced: Standards of testing and peer review vary across disciplines, which also have different canons for what constitutes "proof." No wonder, then, that when applying Daubert, trial-court judges have required wildly different standards of scientific proof.

After a decade, it has become clear that the Supreme Court's blunder has had serious unintended consequences. Under Daubert, defense lawyers for powerful companies have had a field day raising evidentiary challenges against plaintiffs' expert witnesses and getting their testimony thrown out before the trial takes place. This ritual has fostered a wrongheaded piece-by-piece approach to scientific evidence in which attempts are made to poke holes in each individual study or testimony, rather than allowing a court (and jury) to consider the full weight of the scientific evidence. Meanwhile, delighted by Daubert, industry groups such as the U.S. Chamber of Commerce have called for extending the principles embodied in the case to the judicial assessment of scientific evidence used in federal regulatory decision making, which would be a true nightmare given what Daubert has already done to tort litigation.

So who on the Supreme Court foresaw these problems? Not Blackmun, nor moderate David Souter. Rather, the chief critic of the Daubert decision was the archconservative William Rehnquist, who observed that questions of "scientific validity" and "peer review" were "matters far afield from the expertise of judges," whom he worried about turning into "amateur scientists." "I defer to no one in my confidence in federal judges," Rehnquist wrote, "but I am at a loss to know what is meant when it is said that the scientific status of a theory depends on its 'falsifiability,' and I suspect some of them will be, too."

What does this have to do with John Roberts? As a preface, let me say that the judge's apparent views on matters like church-state separation worry me a great deal, as do his unclear views on Roe v. Wade. However, the Daubert opinion shows that beyond these hot-button culture-war issues, matters also come before the Supreme Court where the liberals and moderates can wind up dead wrong, and where someone like William Rehnquist can put them all to shame via concurrence or dissent.

There are areas, in short, where judges aren't predictable -- and where, accordingly, having a good mind matters a lot more than one's basic political leanings. This is not my brief for allowing Roberts onto the court; I think he deserves considerable scrutiny. But at the very least, Daubert ought to make us think twice about merely examining a judge's politics when deciding whether we want him or her deciding cases of this magnitude.

Chris Mooney is the Washington correspondent for Seed Magazine and a columnist for The American Prospect Online. His first book, The Republican War on Science, will be published in September. His daily blog and other writings can be found at