Concept
From Frye to Daubert
Reliability and relevancy
Expert Reliability and Relevancy
From Frye to Daubert
Every expert opinion asks jurors to accept a conclusion they cannot check for themselves, so someone must decide, before the jury hears it, whether the science behind the opinion is good enough. Under the older Frye v. United States standard, the scientific community made that decision.1 In Daubert v. Merrell Dow Pharmaceuticals, Inc., the Supreme Court discarded Frye for federal courts and gave the decision to the trial judge, who must determine whether the testimony is reliable and fits the issues in the case.2
A judge faced with opposing experts has three options. The judge can admit the evidence because the experts are qualified and let the jury resolve the dispute. The judge can decide personally whether the plaintiff's proof is valid science. Or the judge can defer to the scientific community, asking the proponent to show not only what the proof is and what it means, but that scientists generally agree with it.3 Frye most clearly resembles the third option. It lets the judge say, in effect, that any exclusion rests on the verdict of other scientists and not on the judge's own opinion.4
The third option appeals because it spares the judge the hardest question. A judge applying Frye never has to decide whether the expert's methodology is actually reliable; polling the scientific community is enough. Daubert ended that shortcut and imposed a heavy responsibility: the judge must evaluate the reliability and relevance of evidence from a scientific field in which the judge may have no expertise.5
The standard drew objections from more than one direction. One was that Frye excluded too much evidence. By the time Daubert was decided, a different concern had arisen: courts were being flooded with “junk science” and taking on more responsibility for risk management than they could handle. Peter Huber's Galileo's Revenge: Junk Science in the Courtroom (1991) popularized the term, and the Ninth Circuit cited the book in Daubert.6
The Bendectin litigation put the question squarely. Jason Daubert and Eric Schuller, minor children born with serious birth defects, and their parents sued in California state court, alleging that the mothers' ingestion of Bendectin, a prescription anti-nausea drug, had caused the defects. The manufacturer, Merrell Dow, removed the suits to federal court on diversity grounds.7 The plaintiffs answered Merrell Dow's motion with eight experts, each with impressive credentials, who concluded that Bendectin can cause birth defects. Their conclusions rested on three kinds of work: test-tube and live-animal studies linking Bendectin to malformations; pharmacological studies comparing Bendectin's chemical structure with that of other substances known to cause birth defects; and a “reanalysis” of previously published epidemiological studies.8
Credentials did not carry the day under Frye. The Ninth Circuit concluded that the reanalyses could not be shown to be generally accepted as a reliable technique, and it stressed that other courts of appeals considering Bendectin had refused to admit reanalyses that were neither published nor peer reviewed.9 The Supreme Court reversed, holding that Frye was superseded by Federal Rule of Evidence 702.10
The Court reached that holding by reading the Rules, not by judging whether general acceptance is a sound test; it did not address the merits of Frye.11 Rule 402 supplied the central point: all relevant evidence is admissible unless the Constitution, a statute, or another rule excludes it. Frye was a creature of decisional law, not any of those, so it could not stand as an additional barrier. The only criteria the Court found applicable to scientific expert testimony lay in Rule 702 and a few other provisions of the Rules.12
Displacing Frye did not mean that credentials alone open the courtroom door. The Rules give experts wide latitude in what they may say, so some standard of reliability is needed; it would not do to say that a credentialed expert may testify to anything and let the adversary process serve as the only safeguard.13 The Court found that standard in the Rules themselves and adopted a more flexible approach, still designed to insure the reliability of scientific evidence.14 The inquiry survived and its keeper changed. The judge must now determine whether the expert's methodology is actually reliable, and general acceptance counts as only one factor in that determination.15
Winning the Frye question did not win the case. On remand the Ninth Circuit again held that the plaintiffs' proof was not valid science. Without peer review, it reasoned, the proponent must show validity some other way, for example by having the experts explain how they reached their conclusions and show that they followed the scientific method as practiced by at least a recognized minority. The proof also had to “fit” the claim: it had to show that children whose mothers took Bendectin were more than twice as likely to suffer defects as children of mothers who did not. The plaintiffs made neither showing.16
Two misreadings recur. The first treats the end of Frye as the end of screening, as though any credentialed expert may reach the jury; the judge's duty to determine reliability forecloses that reading. The second carries Frye forward by excluding anything unpublished or not yet widely accepted, which turns one factor into the whole test.17 A sound answer states that the federal question is reliability and relevance, decided by the judge. Whether state courts adopted Daubert or kept Frye is a separate matter, addressed later in this chapter.
Daubert concerned scientific evidence. In 1999, in Kumho Tire, the Court extended the Daubert standard to all expert testimony presenting technical or specialized material.20 The case began with a minivan accident: the right rear tire blew out, a passenger died, others were severely injured, and the Carmichaels sued the tire's maker and distributor in a diversity suit. They rested their case in significant part on the deposition testimony of Dennis Carlson, Jr., an expert in tire failure analysis.21 The gatekeeping duty described in the next section reaches his testimony as it reached the Bendectin experts.