Hostname: page-component-6bf8c574d5-mggfc Total loading time: 0 Render date: 2025-02-26T15:17:46.891Z Has data issue: false hasContentIssue false
  • English
  • Français

Science, Law, and the Search for Truth in the Courtroom: Lessons from Daubert v. Merrell Dow

Published online by Cambridge University Press:  01 January 2021

Joan E. Bertin  and
Mary S. Henifin
Get access Rights & Permissions [Opens in a new window]

Extract

On June 28, 1993, the United States Supreme Court ruled on the admissibility of expert scientific opinion and evidence in federal court cases. The importance of the case can be measured by the interest it stimulated. The scientific community turned out in particular force to register its views. At the heart of the controversy was a debate over the nature of scientific knowledge and its relation to law. More than any other Supreme Court case in recent memory, the amici seemed to view the Daubert case both as posing a potential challenge to basic scientific premises and as an opportunity to reclaim a lost prerogative to determine how science is applied in courts.

Type
Article
Information
Journal of Law, Medicine & Ethics , Volume 22 , Issue 1 , Spring 1994 , pp. 6 - 20
Copyright
Copyright © American Society of Law, Medicine and Ethics 1994

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Daubert et al. v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. ___, 125 L.Ed.2d 469 (1993). Joan E. Bertin was an author, with Berzon, Marsha S., Meyerhoff, Albert H., of the brief amicus curiae submitted by the American Society of Law, Medicine & Ethics et al., in support of the plaintiffs in Daubert in the Supreme Court.Google Scholar
In addition to the briefs presented by the parties, twenty-two different groups offered their views in amicus curiae briefs. The Pharmaceutical Manufacturers Association, the National Association of Manufacturers, the Chemical Manufacturers Association, the Chamber of Commerce of the United States of America, the Association of Trial Lawyers of America, and others expressed arguments reflecting their institutional interests in tort litigation of this sort. The United States Department of Justice and a group of law professors also filed briefs in the case.Google Scholar
Scientists, doctors, and their representatives and organizations were heavily represented. Approximately seventy doctors, scientists, medical journals, and medical and scientific organizations appeared as amici curiae, in eleven separate briefs. The American Association for the Advancement of Science, the National Academy of Science, The New England Journal of Medicine, The Journal of the American Medical Association, The Annals of Internal Medicine, the American Medical Association, and fifteen national medical specialty societies, the American College of Legal Medicine, and others filed briefs on behalf of the respondent, Merrell Dow. The Carnegie Commission on Science and Technology filed on neither side, but advocated reversal of the decision below. The American Society of Law, Medicine & Ethics and numerous individual scientists filed in favor of plaintiffs. In this article, we address the arguments presented in these briefs.Google Scholar
See Sanders, Joseph, “From Science to Evidence: The Testimony on Causation in the Bendectin Cases,” Stan. L. Rev., 46 (1993): 8.CrossRefGoogle Scholar
In the medical community, Bendectin is widely thought to be safe, based on numerous poulation-based studies that failed to demonstrate risk. The plaintiffs in Daubert dispute that assumption, for reasons discussed infra. In this article, we express no view on the question of Bendectin's safety. Our observations are directed solely at the evidence in the record and how the legal system responds to such evidence.Google Scholar
See, for example, Wagner, Wendy, “Trans-Science in Courts,” Yale L.J., 96 (1986): 431, note 20. The parties' briefs understandably were driven by the effort to prevail in this particular case, not to address systemic problems with the use of science in courts. The amici likewise may have had similarly immediate objectives, which would explain the failure to focus on the larger issues exemplified by the case, although the absence of such a focus nonetheless is regrettable.Google Scholar
See Gold, Steve, “Causation in Toxic Torts: Burdens of Proof, Standards of Persuasion, and Statistical Evidence,” Yale L. J., 96 (1986): 376. Notwithstanding the difficulties inherent in tort litigation as a source of compensation and a way of deterring wrongful conduct, injured individuals will continue to seek legal redress, and the stakes in the outcome of personal injury cases will remain high, unless and until there is a uniform system of compensation or coverage for serious injuries. In the interim, however, there are approaches that would likely diminish the perception, often accurate, that litigation is a high-priced game of chance in which only the lawyers win.CrossRefGoogle Scholar
It may take a considerable period of time to “quantify the hazard,” especially if courts demand positive epidemiology to prove causation. Wagner, supra note 6, at 429 and notes 11–12. Although reliance on epidemiology to prove causation, to the exclusion of other types of proof, has been criticized, it is common. Gold, supra note 7, at 392. The Supreme Court in Daubert was invited to adopt such a restrictive approach, but declined to do so, leaving courts with the discretion to accept valid and relevant experimental evidence.Google Scholar
Gold, , supra note 7, discusses the problems associated with proving causation to satisfy the “more probable than not” standard and the consequential need for a probabilistic approach to proof in the toxic torts area.Google Scholar
“[E]ven a single data set may be interpreted differently by reasonable scientists.” Gold, , supra note 7, at 398 note 112.Google Scholar
The adversary system itself fosters scientific uncertainty, because it requires each side to present evidence most favorable to its position, and thereby discourages discussion of subtleties and uncertainties, as discussed infra.Google Scholar
See Bazelon, David L., “Science and Uncertainty: A Jurist's View,” Harv. Envt'l L. Rev., 5 (1981): 209.Google Scholar
A “no-fault” approach to recovery for personal injuries would also avoid the problem. Such an approach has much to recommend it, if it avoids the expensive litigation that now characterizes many tort actions. Limiting the amount of recovery in exchange for the assurance that all basic needs (income replacement, medical care, and so forth) will be met, would seem an appropriate trade-off. Unfortunately, the experience with the workers' compensation system, which was to effect a similar sort of trade-off, makes some suspicious that such a system would not dispose of victims' claims fairly.Google Scholar
727 F. Supp. 570, 572 (quoting United States v. Kilgus, 571 F.2d 508, 510 (9th Cir. 1978)). Because there was no trial, the factual record in the Supreme Court is extremely limited.Google Scholar
This rule is derived from Frye v. United States, 293 F. 1013 (D.C. Cir. 1923), in which a prosecutor in a criminal case sought to introduce scientific testimony about a rudimentary type of lie detector test. Evidence used to obtain a conviction, of course, must be highly reliable, since potential loss of liberty is involved.Google Scholar
Sanders, Joseph, “The Bendectin Litigation: A Case Study in the Life Cycle of Mass Torts,” Hastings L. J., 43 (1992): 301.Google Scholar
The threat of massive liability did not materialize. Most cases were resolved in the defendant's favor. See note 19, infra, and accompanying text. And, see Sanders, , supra note 4, at 6 (Table 1).Google Scholar
Green, Michael D., “Expert Witnesses and Sufficiency of Evidence in Toxic Substances Litigation: The Legacy of Agent Orange and Bendectin Litigation,” Northwestern L. Rev., 86 (1992): 643.Google Scholar
Compare Lynch v. Merrell-National Laboratories, 830 F.2d 1190 (1st Cir. 1987) and Richardson v. Richardson-Merrell, 857 F.2d 823 (D.C. Cir. 1988), cert. denied, 493 U.S. 882 (1989) with Brock v. Merrell Dow Pharmaceuticals, 874 F.2d 307, modified, 884 F.2d 166 (5th Cir. 1989), cert. denied, 494 U.S. 1046 (1990). And see DeLuca v. Merrell Dow Pharmaceuticals, 911 F.2d 941 (3d Cir. 1990) (expert testimony admissible; sufficiency to be determined on remand); Turpin v. Merrell Dow Pharmaceuticals, 959 F.2d 1349 (6th Cir.), cert. denied, 113 S. Ct. 84 (1992) (evidence insufficient to establish causation); Wilson v. Merrell Dow Pharmaceuticals, 893 F.2d 1149 (10th Cir. 1990) (summary judgment inappropriate); In re Richardson-Merrell, Inc. “Bendectin” Products Liability Litigation, 857 F.2d 290 (6th Cir. 1988), cert. denied, 488 U.S. 1006 (1989) (jury verdict for defendant after full trial); In re Bendectin Products Liability Litigation, 732 F. Supp 744 (E.D. Mich 1990) (denying defendant's motion for summary judgment); and Oxendine v. Merrell Dow Pharmaceuticals, 506 A.2d 1100 (D.C. 1986) and 563 A.2d 330 (D.C. 1989), cert. denied 110 S. Ct. 1121 (1990) (jury verdict in the plaintiff's favor reinstated by appellate court). The Oxendine case, the only verdict for the plaintiff upheld on appeal, has had a tortured history, which is traced in Sanders, supra note 4. Defendants are currently seeking a new trial. Id. at 29.Google Scholar
509 U.S. at ___, 125 L.Ed.2d at 484.Google Scholar
Id. at 485.Google Scholar
While scientists should define the elements of scientific credibility, it is not self-evident that the same standards of proof should apply in the laboratory and in the courtroom. Scientists pursue answers to questions over time, indeed, as long as it takes. Judges and juries, in contrast, must often make do with the best available evidence in arriving at factual determinations at a given point in time, as discussed infra.Google Scholar
But, see, for example, Rennie, Drummond, “More Peering Into Editorial Peer Review,” JAMA, 270, no. 23 (1993): 2856–58: “[W]e do not yet know whether peer review reliably improves the quality of the published product…. [M]isconduct by editors may occur and seems hard to control. Fraud by an author is almost impossible for reviewers and editors to detect….”CrossRefGoogle Scholar
Some data and analyses created for litigation would not be appropriate for publication, nor would there be time to publish them prior to use in the legal proceeding for which they were undertaken. Examples include statistical analyses of employment patterns in employment discrimination cases or reviews of medical histories of all individuals exposed to a chemical spill in a personal injury case. The rule in the lower courts in Daubert was so broad that it would have barred all kinds of expert testimony.Google Scholar
For example, the Brief Amicus Curiae of Bloembergen, Nicolaas et al., In Support of Respondents in Daubert et al. v. Merrell Dow Pharmaceuticals, Inc. (U.S. No. 92–102) at 22, states:Google ScholarGoogle ScholarGoogle Scholar
509 U.S. at ___, 125 L.Ed.2d at 476. The claim that “junk science” has pervaded and corrupted the legal system is reflected in Huber, Peter, Galileo's Revenge: Junk Science in the Courtroom (New York: Basic Books, 1991). But see Chesebro, K., “Galileo's Retort: Peter Huber's Junk Scholarship,” Am. U. L. Rev., 42 (1993): 1637, for a refutation of Huber's thesis.Google Scholar
Brief for the American Association for the Advancement of Science and the National Academy of Sciences, as Amici Curiae In Support of Respondent in Daubert et al. v. Merrell Dow Pharmaceuticals, Inc. (U.S. No. 92–102) at 3.Google Scholar
Brief of the American Medical Association et al., as Amici Curiae In Support of Respondent in Daubert et al. v. Merrell Dow Pharmaceuticals, Inc. (U.S. No. 92–102) at i.Google Scholar
Indeed, the FDA concluded that “available data do not demonstrate an association between birth defects and Bendectin,” but advised caution in its use. Quoted in Sanders, supra note 4, at 7 (emphasis added).Google Scholar
Brief of the American Medical Association et al., supra note 28, at 16.Google Scholar
Id. at 4.Google Scholar
This concern is real but should not be limited to legal proceedings. Many forces threaten scientific objectivity. For example, the source of a scientist's funding may influence the topics of research or the analysis of data; concerns about potential liability may influence both research and treatment; personal and professional investment in the validity of particular theories may blind scientists to flaws in their analyses, and so forth. Many academic scientists also hold positions in corporations or own their own consulting companies.Google Scholar
This consists of interrogatories (written questions answered under oath), document production, and depositions (questions, usually oral, under oath).Google Scholar
The amici stressed the need for scientific evidence to be valid and reliable. This is essentially what lawyers mean when they refer to credibility. Scientific evidence is only credible if it is valid and reliable.Google Scholar
One brief cited a case in which a jury refused to accept valid and conclusive scientific testimony that should have disposed of a particular claim. This example contradicts the premise that juries are unduly swayed by experts, but reinforces the dim view of a system in disarray. Brief Amicus Curiae of the American College of Legal Medicine In Support of Respondent in Daubert et al. v. Merrell Dow Pharmaceuticals, Inc. (U.S. No. 92–102), at 8, note 7.Google Scholar
509 U.S. at ___, 125 L.Ed.2d at 484.Google Scholar
Brief Amici Curiae of Rothman, Professors Kenneth, Weiss, Noel, Robins, James, Neutra, Raymond, Stellman, Steven In Support of Petitioners in Daubert et al. v. Merrell Dow Pharmaceuticals, Inc. (U.S. No. 92–102).Google Scholar
Brief Amicus Curiae of ProfessorFeinstein, Alvan R. In Support of Respondent in Daubert et al. v. Merrell Dow Pharmaceuticals, Inc. (U.S. No. 92-102).Google Scholar
Brief amici curiae of Nicolaas Bloembergen et al., supra note 25, in response to the brief amici curiae of physicians, scientists, and historians of science in support of petitioners (Bayer, Ronald et al.).Google Scholar
See, e.g., Cook, D. J. et al., “Should Unpublished Data Be Included in Meta-analyses? Current Convictions and Controversies,” JAMA, 269, no. 21 (1993): 2749–53; Mills, J.L., “Data Torturing,” N. Engl. J. Med., 329, no. 16 (1993): 1196–99. Most amici conceded that meta-analysis is not per se “unscientific,” but that its validity depends on how it is performed. Because there is no discussion of specific flaws or deficiencies in the methodology employed by the experts in this case, the disagreement often appeared to relate more to the results than to the methods by which they were obtained.CrossRefGoogle Scholar
509 U.S. at ___, 125 L.Ed.2d at 482.Google Scholar
Id. at 484.Google Scholar
Indeed, litigation since the Daubert opinion was issued confirms this. See the discussion of In Re Joint Eastern and Southern District Asbestos Litigation, infra.Google Scholar
Accord Black, B., Singer, J.A., “From Frye to Daubert: A New Test for Scientific Evidence,” Shepard's Expert and Scientific Evidence Quarterly, 1 (1993): 40. The article also observes that the Supreme Court's disposition may “eliminate the extremes of hair trigger exclusion and mindless admission of evidence.”Google Scholar
For example, the judge can appoint experts, make credibility determinations, issue instructions to the jury, grant summary judgment or judgment notwithstanding the verdict, and so forth.Google Scholar
The adversary system creates additional disincentives to speedy resolution: Culpable defendants have nothing to lose and everything to gain from delay, parties with greater financial resources may use delay and expense to exert pressure on their opponents, lawyers profit by expending more hours, and lack of cooperation sometimes is employed in the war of nerves that too often accompanies litigation. These faults must be laid primarily at the feet of lawyers, although clients themselves often encourage or demand dilatory tactics.Google Scholar
Anyone can file an action by paying the filing fees and submitting the appropriate forms. Legal checks protect against frivolous claims and dilatory tactics. Sanctions can be imposed; inadequate complaints are subject to early dismissal; summary judgment is available to expedite cases that involve only legal, not factual, issues; motions in limine limit the triable issues; and so forth. See Fed. R. Civ. Pro., Rules 11, 12(b)(6), 56.Google Scholar
Especially if the calculation includes time expended to compile information for discovery, the savings could be substantial. Attorneys fees and costs of litigation often outstrip amounts paid in settlement.Google Scholar
See, e.g., N.Y. Civ. Prac. Law & R., § 214-c (1993).Google Scholar
Questions raised during the course of litigation and congressional inquiries stimulated research into the effects of exposure to Agent Orange. Five such studies and an editorial were reported in a single issue of Environmental Research in 1988. See, e.g., Stellman, S.D., Stellman, J.M., Sommer, J.F., “Health and Reproductive Outcomes among American Legionnaires in Relation to Combat and Herbicide Exposures in Vietnam,” Envir. Res., 47 (1988): 150–74; and Gochfield, M., Editorial: “New Light on the Health of Vietnam Veterans,” Envir. Res., 47 (1988): 109-11.CrossRefGoogle Scholar
Liability is not generally imposed for injuries from socially useful, but unavoidably unsafe, products such as drugs, if reasonable care has been exercised and warnings have been provided. Restatement (Second) of Torts, § 402A comment k (1965).Google Scholar
The less conclusive the evidence, the more room there is for contrary interpretations by expert witnesses.Google ScholarGoogle ScholarGoogle Scholar
509 U.S. at ___, 125 L.Ed.2d at 485.Google Scholar
Bazelon, , supra note 12, at 213.Google Scholar
For this reason, for example, Judge Bazelon proposed greater resort to regulatory decision making in addressing complex toxic risk issues that “[n]egligence law is simply inadequate to deal with.” Bazelon, , supra note 12, at 211.Google Scholar
Weinberg, Alvin, Letters to the Editor, Science, 174 (1971): 546–47, quoted in Wagner, , supra note 6, at 431 note 20.CrossRefGoogle Scholar
Wagner, , supra note 6, at 437. This problem is exacerbated by the refusal of some courts to admit experimental data, even if methodologically sound. The overreliance on epidemiology not only imposes a distinct disadvantage on the party charged with proving injury, but also is peculiar from a scientific perspective. Epidemiology, the study of diseases in human populations, establishes correlations but “can rarely be used to make a causal inference.” Epidemiology is “inherently capable of detecting only comparatively large increases in the relative risk…. Negative results cannot prove the absence of [risk]….” EPA Guidelines for Carcinogenic Risk Assessment, 51 Fed. Reg. 33992, 33995–96 (1986). Experimental data, in contrast, provide information about the biological processes that cause injury, and are heavily relied on by regulatory agencies. The EPA has concluded that “in almost all cases, experimental animal data are predictive of a developmental effect in humans.” EPA Guidelines for Developmental Toxicity Risk Assessment, 56 Fed. Reg. 63798, 63801 (1991).Google Scholar
827 F. Supp. 1014 (S.D. N.Y. 1993).Google Scholar
Id. at 1038, note 20.Google Scholar
Id. at 1046. As this case indicates, extremely difficult proof problems arise when there is only a weak association between cause and effect, or when the event is uncommon. In such cases, it may be nearly impossible to demonstrate causation in an individual case, even though the effect may exist. For a review of the cases decided after the Supreme Court opinion in Daubert, see Black, B., “The Judicial Reaction to Daubert,” Shepard's Expert and Scientific Evidence Quarterly, 1 (1993): 319–35.Google Scholar
Turpin v. Merrell Dow Pharmaceuticals, Inc., 959 F.2d 1349, 1350 (6th Cir.), cert. denied, 506 U.S.___, 113 S. Ct. 84, 121 L.Ed.2d 47 (1992).Google Scholar
Which is not to say that experts do not have opinions about the ultimate issue. The litigation in Daubert demonstrated graphically that is not the case.Google Scholar
Daubert et al. v. Merrell Dow Pharmaceuticals, Inc., First Amended Complaint for Damages, ¶ 15, Joint Appendix, pp. 20–21.Google Scholar
Sanders, , supra note 4, at 53–54.Google Scholar
Bazelon, , supra note 12, at 212. Juries may perform some of the functions of the electorate, in that they reflect community values.Google Scholar
Id.Google Scholar
See, e.g., Brennan, Troyen A., “Causal Chains and Statistical Links: The Role of Scientific Uncertainty in Hazardous Substance Litigation,” Cornell L. Rev., 73 (1988): 469; and Sanders, , supra note 4.Google Scholar
Gold, , supra note 7, proposes, for example, an approach to fact-finding that reflects the need to make probabilistic judgments about causation.Google Scholar
See Wagner, , supra note 6.Google Scholar
See Hymowitz v. Eli Lilly Co., 73 N.Y.2d 487, cert. denied 493 U.S. 944 (1989). See also Martin v. Abbott Labs., 102 Wash.2d 581, 689 P.2d 368 (1984) and Sindell v. Abbott Labs., 26 Cal.3d 588, 607 P.2d 924, cert. denied 449 U.S. 912 (1980).Google Scholar
In 1971, the FDA banned the use of the drug to prevent miscarriages, based on evidence that DES caused vaginal adenocarcinoma, among other injuries, in prenatally exposed girls. Hymowitz, 73 N.Y.2d at 503.Google Scholar
Id.Google Scholar
Id. at 507.Google Scholar
Id. (citations omitted).Google Scholar
Different terminology is applied by different courts (e.g., “market share,” “enterprise liability,” “alternative liability”) and somewhat different formulations were created, but they shared common themes in that proof of injury from DES shifted the burden to the defendant to prove any exculpatory facts and defendants were generally liable in proportion to their share of the market.Google Scholar
Pub. L. No. 99–660, 100 Stat. 3743, codified at 42 U.S.C.A. §§ 300aa-1, et seq.Google Scholar
H. Rep. No. 908, 99th Cong., 2d Sess., pt. 1 at 6 (1986), reprinted in 1986 U.S. Code Cong. & Admin. News, 6287, 6344, and 6347.Google Scholar
See 42 U.S.C.A. § 300aa-11. However, injured parties, after they complete the act's compensation proceedings, may reject the judgment and pursue a civil action. See, e.g., Graham v. Wyeth Laboratories, 906 F.2d 1399 (10th Cir. 1990).Google Scholar
Such policies limited the employment opportunities of fertile women, with the stated goal to prevent the possibility of fetal exposure to hazardous working conditions. For a full discussion of these policies, see Bertin, Joan E., “Reproductive Hazards in the Workplace: Lessons from UAW v. Johnson Controls,” in Bellinger, D., Needleman, H., eds., Prenatal Exposure to Environmental Toxins (Baltimore: Johns Hopkins Press, in press); Becker, Mary, “From Muller v. Oregon to Fetal Vulnerability Policies,” U. Chi. L. Rev. 53 (1986): 1219; and Bayer, Ronald, “Reproductive Hazards in the Workplace: Bearing the Burden of Fetal Risk,” Milbank Mem. Fund Quart., 60, no. 4 (1986): 633-56.Google Scholar
499 U.S. 187 (1991).Google Scholar
Zuniga v. Kleberg County Hospital, 692 F.2d 986 (5th Cir, 1982); Wright v. Olin Corp., 697 F.2d 1172 (4th Cir. 1982); Hayes v. Shelby Memorial Hospital, 726 F.2d 1543 (11th Cir. 1984); and Grant v. General Motors, 908 F.2d 1303 (6th Cir. 1990).Google Scholar
The California courts, following the federal Occupational Safety and Health Administration, which relied on animal as well as human data, concluded that the policy was not scientifically justifiable, while the Seventh Circuit rejected animal data and upheld the policy. Compare Johnson Controls v. California Fair Employment and Housing Commission, 218 Cal. App. 3d 517 (1990) with International Union, UAW v. Johnson Controls, 886 F.2d 871 (7th Cir. 1989), rev'd 499 U.S. 187 (1991).Google Scholar
Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§ 2000e, et seq., as amended, prohibits discrimination in employment on the basis of sex or pregnancy. 42 U.S.C. § 2000e(k).Google Scholar
The employer retains the ability to avoid foreseeable harm to the fetus of a woman worker by conforming to an appropriate standard of care in the maintenance of safe working conditions and providing warnings about unavoidable risks.Google Scholar
To avoid the necessity to resolve competing scientific claims of a different sort, courts would have to treat an issue as “unknowable” once a credible showing to that effect has been made.Google Scholar
See Wagner, , supra note 6, at 430. By injury consistent with exposure, we mean to suggest the requirement to demonstrate biological plausibility, which should be considerably less onerous than the burden to prove causation, and considerably less subject to dispute.Google Scholar
Beshada v. Johns-Manville Products Corp., 90 N.J. 191 (1982). Interestingly, the New Jersey Supreme Court made this comment while striking a “state of the art” defense in a strict products liability action. The court ruled that in an asbestos strict liability case involving an alleged failure to warn the plaintiff of risk, the defendant's culpability was irrelevant. Of course, the plaintiff would still have to prove causation. We are suggesting that in cases where causation is unknowable given current scientific uncertainty, the emphasis should shift to the culpability of the manufacturer.Google Scholar
Compensation issues remain concerning unavoidable and unforeseeable risks, but these might become a much smaller class of events if greater emphasis were placed on insuring the safety of products in the first place.Google Scholar
Beshada v. Johns-Manville Products Corp., 90 N.J. at 207.Google Scholar