Issues in Focus: Science and the Law
Expert Testimony: The Supreme Court’s Rules
Three decisions in the 1990s have clarified the judge’s role as the gatekeeper who controls what testimony a jury will hear.
At the beginning of the 21st century, it is not surprising that the question of how to handle scientific and technological information in judicial proceedings has moved into the limelight. The explosive growth and importance of scientific and technological knowledge in our society has run a parallel course in the courtroom, where an ever-increasing number of legal disputes cannot be resolved without the assistance of scientific and technological expertise. But although remarkable new scientific findings are reported every day, there is still much we do not know. Consequently, the courts have been struggling with the difficult problem of determining when expertise will actually help the trier of fact (usually the jury but sometimes the judge) in making a determination. An expert witness who claims to have specialized knowledge will be permitted to testify only when that specialized knowledge can really be of assistance. It is in the context of disputes about the admissibility of expert testimony that courts decide what kind of science and technology (S&T) information the legal system will take into account.
One particularly troublesome area for the courts has been the proof of causation in so-called “toxic tort” cases, a subspecies of product liability litigation. These are cases in which the plaintiffs bringing the action allege that their injuries or disease were caused by exposure to the defendant’s product. In the past 20 years there has been an enormous increase in toxic tort litigation, which even when it does not result in huge awards (and we all know about asbestos and tobacco) may bankrupt or seriously damage a defendant’s financial standing because these suits are so costly to litigate. Except in the case of signature diseases, such as those associated with exposure to asbestos or DES, the injuries of which the plaintiff complains are also found in people who were never exposed to the defendant’s product. Consequently, scientific proof that the product in question is capable of causing injuries such as the plaintiff’s, and more likely than not did, is crucial.
The use of expert testimony to prove causation has recently captured the attention of the United States Supreme Court, perhaps because of the huge amounts of money at stake or because of allegations that experts in these cases have often relied on “junk science.” In any event, since 1993 the Supreme Court has issued a trilogy of opinions dealing with the admissibility of expert proof. Taken together, the trilogy establishes the ground rules for introducing expert testimony in all cases brought in the federal system, criminal as well as civil. Furthermore, although these opinions do not bind state courts, approximately three-quarters of the states have already opted to adopt the Supreme Court’s new test, and more will undoubtedly do so in the future. Consequently, anyone who acts in an expert witness capacity in judicial proceedings in the United States is likely to be affected by the trilogy and its progeny.
Not only relevant, but reliable
The first case in the trilogy, Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993), was one of many cases in which plaintiffs claimed that their birth defects were caused by Bendectin, an anti-morning sickness pill that had been taken by their mothers and more than 20 million other women. As a result of the litigation, the defendant manufacturer took the drug off the market even though it never lost its Food and Drug Administration approval. In determining that the epidemiological and toxicological evidence offered by plaintiffs’ experts was inadmissible, the lower court in Daubert had applied the so-called “general acceptance” test first enunciated by a federal appeals court in Frye v. United States (1923). The general acceptance test, which was used by some federal courts primarily in criminal cases and is still used by some state courts, conditioned expert testimony about a novel scientific principle on there being a consensus about the theory in the relevant field. In Daubert, the Supreme Court first held that Frye was a dead letter in the federal courts and then spelled out a new two-pronged test for the admissibility of scientific evidence, geared to ensuring that testimony “is not only relevant, but reliable.”
In order to satisfy the reliability prong, the expert’s proffered opinion must be the product of scientific reasoning and methodology. That is, the judge must determine whether the expert reached his or her conclusions by a scientific method. The Court suggested a number of factors that bear on this analysis. First and foremost, the Daubert Court viewed science as an empirical endeavor: “Whether a theory or technique can be (and has been) tested” is the “methodology” that “distinguishes science from other fields of human inquiry.” Also mentioned by the Court as indicators of good science were peer review or publication and the existence of known or potential error rates and of standards controlling the technique’s operation. General acceptance of the methodology within the scientific community, although no longer dispositive, still remained a factor to be considered. Second, the Court explained that by relevancy it meant that the expert’s theory must “fit” the facts of the case. The expert may not testify about a hypothesis that cannot properly be applied to the facts of the case, such as, for instance, that substance X can cause plaintiff’s nonsignature disease, when there is no evidence that plaintiff was ever exposed to substance X.
Perhaps the most significant part of Daubert is the Court’s anointment of the trial judge as the “gatekeeper” who must screen proffered expertise to determine whether the relevancy and reliability prongs are met. Although there was nothing particularly novel about a trial judge having the power to exclude inappropriate expert testimony, Daubert stressed that the trial court has an obligation to act as gatekeeper even though some courts would rather have left this task to the jury, especially when the screening entailed complex scientific issues. The Supreme Court did not apply its new test to the scientists who were seeking to testify that Bendectin caused birth defects. Instead, the Court sent the case back to the lower court, which again excluded the testimony of the plaintiffs’ experts and granted summary judgment for the defendants.
In the second case in the trilogy, General Electric v. Joiner (1977), the 37-year-old plaintiff, a long-time smoker with a family history of lung cancer, claimed that exposure to polychlorinated biphenyls (PCBs) and their derivatives had promoted the development of his small-cell lung cancer. On the authority of Daubert, the trial court excluded the plaintiff’s expert testimony and granted summary judgment. The intermediate appellate court reversed. The Supreme Court held that in reviewing a trial judge’s evidentiary ruling, an appellate court must use an abuse of discretion standard. This standard requires the reviewing court to defer to the rulings of the trial court unless they are manifestly erroneous. The Court also took the opportunity to spell out how the abuse of discretion standard operates in a case such as Joiner.
After examining the record in detail, the Supreme Court concluded that the trial judge had not abused his discretion when he concluded that plaintiff’s experts had not explained “how and why” they could extrapolate proof of causation from animal studies conducted under circumstances far different from those surrounding the plaintiff’s exposure. The studies involved infant mice that, after having massive doses of PCBs injected directly into their bodies, developed a different type of cancer than the plaintiff did. The plaintiff’s exposure was through physical contact with fluids containing far lower concentrations of PCBs.
The Court further found that the trial court had not erred in also rejecting the proffered epidemiological evidence. The authors of one study had refused to conclude that PCBs had caused a somewhat higher rate of lung cancer at an Italian plant than might have been expected; the results of another study were not statistically significant; a third study did not mention PCBs; and the workers in the fourth study cited by the trial judge had been exposed to numerous other potential carcinogens. Consequently, the Court found that the trial judge could conclude that the statements of the plaintiff’s experts with regard to causation were nothing more than speculation. “[I]t was within the District Court’s discretion to conclude that the studies upon which the experts relied were not sufficient, whether individually or in combination, to support their conclusions that Joiner‘s exposure to PCBs contributed to his cancer.”
The third opinion in the trilogy was issued in 1999 in Kumho Tire Co. v. Carmichael (1999). It dealt with the admissibility of engineering testimony to prove causation in a product liability action. In Kumho, the plaintiff claimed that the tire on their minivan blew out, causing a death and serious injuries, because of a defect in a tire. The tire was on the minivan when it was bought second-hand. To substantiate this allegation, the plaintiff relied primarily on testimony by an expert in tire-failure analysis, who concluded on the basis of a visual inspection that the tire had not been abused and that therefore it must have been defective. When the defendant tire manufacturer moved to exclude the plaintiff’s tire expert, the district court initially concluded that his proposed testimony had to be examined in light of the four factors mentioned in Daubert: the theory’s testability, whether it was the subject of peer review or publication, the known or potential error rate and standards, and general acceptance within the relevant field. The district court concluded that none of the Daubert factors was satisfied and excluded the plaintiff’s testimony and granted the defendant’s motion for summary judgment.
The plaintiff asked for reconsideration, arguing that the court’s application of Daubert was too inflexible. The district court agreed to reconsider and agreed that it had erred in treating the four factors as mandatory rather than illustrative. But that concession did not help the plaintiff. The district court stated that it could find “no countervailing factors operating in favor of admissibility which could outweigh those identified in Daubert,” and consequently it reaffirmed its earlier order.
The intermediate appellate court reversed on the ground that Daubert applies only in the scientific context, a conclusion about which federal courts were split. The court drew a distinction between expert testimony that relies on the application of scientific theories or principles, which would be subject to a Daubert analysis, and testimony that is based on the expert’s “skill- or experience-based observation.” It was the disagreement in the courts about Daubert’s applicability to nonscientific evidence that was the Supreme Court’s stated reason for reviewing the Kumho case, but some commentators thought the Court might also use the opportunity to clarify the role that an expert’s experience plays in determining admissibility. For although the applicable federal rule of evidence specifies that an expert may be qualified through experience, the Court’s emphasis in Daubert on science as an empirical endeavor suggested to some that an expert was no longer entitled to base a conclusion solely on experience if the expert’s opinion could somehow be tested.
All the justices, in an opinion by Justice Breyer, agreed that the trial court’s gatekeeping obligation extends to all expert testimony. The Court noted that the governing rule of evidence “makes no relevant distinction between ‘scientific’ knowledge and ‘technical’ or ‘other specialized’ knowledge” and “applies its reliability standard to all . . . matters within its scope.” And the Court emphasized that the objective of the gatekeeping requirement is to ensure that the expert testimony satisfies the reliability and relevancy prongs set out in Daubert. The Court explained that this requires the trial court to “make certain that an expert, whether basing testimony upon professional studies or personal experience, employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.”
On the other hand, the Court declined to find that the gatekeeping obligation means that the four factors mentioned in Daubert must be applied. In Kumho, the defendant had stated at oral argument that all the Daubert factors were “always relevant.” The Kumho opinion rejects this notion: “The conclusion, in our view, is that we can neither rule out, nor rule in, for all cases and for all time the applicability of the factors mentioned in Daubert, nor can we now do so for subsets of cases categorized by category of expert or by kind of evidence. Too much depends upon the particular circumstances of the particular case at issue.” Quoting from the Brief for the United States as Amicus Curiae, the Court explained that admissibility will depend “on the nature of the issue, the expert’s particular expertise, and the subject of his testimony.”
The question of which factors are indicative of reliability in a particular case cannot be resolved solely by slotting the expertise in question into a particular category such as engineering. The opinion explained that sometimes, “(e)ngineering testimony rests upon scientific foundations, the reliability of which will be at issue in some cases . . . In other cases, the relevant reliability concerns may focus upon personal knowledge or experience.”
The Court refused to find that the methodology that the plaintiff’s expert had used could never be used by an expert testifying about tire failures: “[C]ontrary to respondents’ suggestion, the specific issue before the court was not the reasonableness in general of a tire expert’s use of a visual and tactile inspection to determine whether overdeflection had caused the tire’s tread to separate from its steel-belted carcass. Rather, it was the reasonableness of using such an approach, along with Carlson’s particular method of analyzing the data thereby obtained, to draw a conclusion regarding the particular matter to which the expert testimony was directly relevant. That matter concerned the likelihood that a defect in the tire at issue caused its tread to separate from its carcass.”
In Part III of its opinion, the Court then engaged in a remarkably detailed analysis of the numerous case-specific facts that made it reasonable for the district court to conclude that in this case the expert testimony was not reliable. The Court appears to be illustrating what it meant when it wrote that everything depends on “the particular circumstances of the particular case at issue,” and its comment in Joiner that experts must account for “how and why” they arrived at their opinions. The Court noted that the tire was old and repaired, that some of its treads “had been worn bald,” and plaintiff’s expert had conceded that it should have been replaced. Furthermore, although the expert claimed that he could determine by a visual and tactile inspection when a tire had not been abused, thereby leading him to conclude that it was defective, the tire in question showed some of the very marks that he had identified as pointing to abuse through overdeflection. Perhaps even more troublesome to the Court was the fact that although the expert claimed that he could tell from a photograph, before ever having inspected the tire, that the tire had not been abused, he had no idea whether the tire had less than 10,000 or more than 50,000 miles of wear. Finally, the Court remarked that there is no indication in the record that other experts, papers, or articles support the expert’s theory and that “no one has argued that Carlson himself, were he still working for Michelin, would have concluded in a report to his employer that a similar tire was similarly defective on grounds identical to those upon which he rested his conclusion here.”
What light does Kumho shed on what the Supreme Court said in Daubert and Joiner? Nothing the Supreme Court says in Kumho is explicitly inconsistent with what the Court said in Daubert. Nevertheless, some of the things Kumho doesn’t say may be significant. After Daubert, some observers undoubtedly expected the Court to continue what seemed the beginning of an attempt to articulate a rigid classification scheme for different fields of expertise. Numerous commentators and publications and some courts had organized their Daubert discussions around a four-factor admissibility test. But the Court now appears less interested in a taxonomy of expertise; it points out that the Daubert factors “do not necessarily apply even in every instance in which the reliability of scientific testimony is challenged.” The Kumho Court contemplates that there will be witnesses “whose expertise is based purely on experience,” and although it suggests that Daubert’s questions may be helpful in evaluating experience-based testimony, it does not, as in Daubert, stress testability as the preeminent factor of concern. It offers the example of the “perfume tester able to distinguish among 140 odors at a sniff” and states that at times it may be useful to ask such a witness “whether his preparation is of a kind that others in the field would recognize as acceptable.” But this is somewhat different than requiring the perfume tester to pass proficiency tests to prove that he can do what he purports to be able to do.
Although the Supreme Court has endorsed an extremely flexible test for determining the admissibility of expert testimony, this certainly does not mean that all experts will be allowed to testify. It is worth noting that in all three cases before the Court, the end result was the exclusion of the plaintiff’s expert proof in accordance with the trial judge’s ruling. Even though the abuse of discretion standard mandates deference to the trial court regardless of whether the ruling below excluded or admitted evidence, it remains to be seen whether the appellate courts will tolerate rulings by a trial judge that allow the plaintiff’s expert to testify, especially in the toxic tort cases that have been at the center of the controversy about junk science. Justices Scalia, O’Connor, and Thomas joined in a brief concurring opinion in Kumho to warn that the abuse of discretion standard “is not discretion to abandon the gatekeeping function” or “to perform the function inadequately.”
Sterling credentials are not enough. The “intellectual rigor” test enunciated in Kumho means that an expert’s outstanding qualifications will not make the expert’s opinion admissible unless the expert has a valid basis for how and why a conclusion was reached. Experts must be prepared to establish that their conclusions were reached by methods that are consistent with how their colleagues in the relevant field or discipline would proceed to establish a proposition if they were presented with the same facts and issues. In Rosen v. Ciba-Geigy Corp. (1996), for instance, a case in which Chief Judge Posner of the Seventh Circuit formulated the “intellectual rigor” test that the Supreme Court subsequently endorsed in Kumho, the court found that the trial judge had properly excluded the testimony of a distinguished cardiologist who was a department head at the University of Chicago. The expert proposed to testify for the plaintiff in a product liability action brought against the manufacturer of a nicotine patch. The plaintiff, a heavy smoker with a history of serious heart disease, continued to smoke, despite having been told to stop, while wearing the patch, which had been prescribed in an effort to break the plaintiff’s cigarette habit. The plaintiff suffered a heart attack on the third day of wearing the patch, and the plaintiff’s expert sought to testify that the patch precipitated the attack. The court found that exclusion of the expert’s opinion was proper because there was no “explanation of how a nicotine overdose . . . can precipitate a heart attack, or a reference to a medical or other scientific literature in which such an effect of nicotine is identified and tested. Since [the expert] is a distinguished cardiologist, his conjecture that nicotine may have this effect and may well have had it on Rosen, is worthy of careful attention, even though he has not himself done research on effects of nicotine. But the courtroom is not the place for scientific guesswork, even of the inspired sort. Law lags science; it does not lead it.”
One consequence of the Supreme Court’s insistence on the screening of expert testimony by the trial court is growing sophistication on the part of the federal judiciary. In the future, courts will undoubtedly be less likely to allow experts to self-validate their fields of expertise by pointing to a consensus in a narrow field they themselves have defined. For instance, before Daubert, a group of scientists from Bell Labs convinced many courts that they could testify that a device they had invented was capable of matching voice samples–for instance, a tape of a bomb threat with a recording of a defendant’s voice. A study by the National Research Council ultimately demonstrated that the results were no more accurate than voice identifications made by lay witnesses familiar with the defendant.
The Daubert line of cases has undoubtedly heightened the courts’ sensitivity to the application of science in the courtroom and has focused their attention on factors that may be significant in determining the reliability of proffered expertise. But as Kumho recognizes, it is impossible to devise a magic formula that will resolve all the complex issues posed by expert testimony. For example, proof of causation in toxic tort cases will continue to present the courts with difficult questions until the causal mechanisms responsible for the conditions for which plaintiffs seek compensation are better understood.
Among the many recurring issues that courts must resolve is the question of whether a plaintiff can make out proof of causation solely on the basis of animal studies. In Joiner, the Supreme Court found that the trial judge did not err in refusing to allow the plaintiff’s experts to testify on the basis of such studies, because they varied so substantially from the facts of Joiner‘s exposure. Obviously the match between the results in the animal studies and Joiner‘s disease would have been closer if the studies had involved adult mice who had developed tumors more similar to his. But the expert is always going to have to extrapolate from the animal species used in the study to humans, and from the high doses given the animals to the plaintiff’s much lower exposure. Indeed, there are those who argue that because of these differences, high-dose animal studies have no scientific value outside the context of risk assessment and should be irrelevant to prove causation. They note that proof of risk and causation differ because risk assessment frequently calls for a cost-benefit analysis. An agency assessing risk may decide to bar a substance or product if the potential benefits are outweighed by the possibility of risks that are largely unquantifiable because of presently unknown contingencies. Consequently, risk assessors may pay heed to any evidence that points to a need for caution. Some would argue that because this pragmatic accommodation does not signify what causation means in a courtroom–that an inference of cause and effect is more likely than not–expert testimony based on animal studies or other evidence used in risk assessment should be excluded.
There is also a debate about whether a physician relying on the methodology of clinical medicine is qualified to testify about causation. Cases in the Fifth Circuit Court have suggested that such testimony is admissible only if sufficient proof exists that the medical establishment knows how and at what exposures the substance in question can cause the injuries or disease in question; if the disease’s etiology is unknown, the physician may not offer an opinion based on a differential diagnosis and a temporal association between the exposure and the plaintiff’s symptoms (Black v. Food Lion, Inc.; Moore v. Ashland Chemical, Inc.). An opinion in the Third Circuit Court is at the opposite end of the spectrum with regard to testimony by medical experts. It states that the physician’s testimony would be based on a sufficiently valid methodology if he or she had conducted a thorough differential diagnosis that had ruled out other possible causes of the plaintiff’s illness and “had relied on a valid and strong temporal relationship between the exposure and the onset of the plaintiff’s problems” (Heller v. Shaw Industries, Inc.).
Other unresolved questions relate to statistical issues. Should courts reject expert testimony based on epidemiological studies if the data do not satisfy the 0.05 level of statistical significance that scientists often use? Is an epidemiological study with a relative risk of less than 2 probative in proving causation?
These questions about what is relevant in proving causation point to a pronounced difference between science and the law. When research is inconclusive from a scientific perspective, the consequence is that more research is in order. Epidemiological studies may get funded if animal studies point to a possible problem. Additional research may be done if the relative risk is elevated but less than 2.0 In a court of law, however, rejection of the plaintiff’s expert proof means that the plaintiff will lose regardless of what future research might show, for the law has virtually no mechanisms for deferring the resolution of disputes until additional scientific information is gathered. Furthermore, defendants, who we know from past experiences are not always forthcoming, are far more likely than plaintiffs to have relevant data on causation; and plaintiffs, particularly the individual plaintiff, cannot compel anyone, including the defendant, to undertake or fund additional research. Whether and to what extent these factors should play a role in judicial determinations on the admissibility of expert testimony raises questions of public policy, not science, on which judges may disagree. Issues such as these are due to the uncertainty and complexity that lie at the heart of many of the disputes that end up in our courtrooms and are not resolved by the Supreme Court’s trilogy.
But the Supreme Court’s trilogy is most useful in emphasizing that expert witnesses may not make claims in court that they would never make in the context of work in their professional fields. It is to be hoped that the S&T community will also take this pronouncement to heart and voice some disapproval of members of their disciplines who are willing to offer conclusions in the context of judicial proceedings that they would never make outside the courtroom. After Kumho, courts are less likely to tolerate such behavior, and the expert who does not take the lessons of the trilogy to heart may find that his services as an expert witness will no longer be required.
Margaret A. Berger is Suzanne J. and Norman Miles Professor of Law at the Brooklyn Law School in Brooklyn, New York.