Week of October 16, 2000
TRUTH IN TEST TUBES
Standard for Screening Scientific Evidence Is Still Muddles Years After Daubert
Paul R. Rice
By what burden of persuasion does a judge screen scientific and technological evidence after the famous decisions of the Supreme Court in Daubert v. Merrell Dow Pharmaceuticals, Inc.(1993) and Kumho Tire Co. Ltd. v. Carmichael (1999)? In Daubert we were told that for scientific evidence to be admissible, the court must determine that it reflect "scientific know ledge." Although the decision offered an abundance of platitudes, it provided virtually no practical guidance.
Both before and after these two decisions, the Federal Rules of Evidence Advisory Committee has failed to perform its stewardship responsibilities in maintaining the evidence code. Instead, has permitted the rules to float — like foreign currency before intervention by the International Monetary Fund to stabilize it — waiting to see how the courts resolve the problem. The Advisory Committee has not encouraged meaningful public debate.
With the proliferation of new scientific principles and forms of technology, new methodologies for using them, and novel applications of both, we are expecting judges, who generally have no special training or experience in the disciplines brought before them, to do what the experts cannot agree upon — assess whether evidence represents good or bad science.
The Supreme Court and the Advisory Committee have failed us in two ways. First they have not explained how judges should accomplish this task without Frye-type reliance on practitioners of the relevant science. Under Frye, if a scientific principle had not receive general acceptance in the relevant science community, it was not admissible. Second, they have never addressed how the high standard by which reliability is determined for a science (in the hard sciences having to pass a reliability standard of 98% accuracy for use in scientific research) should be employed to help determine a social dispute by a mere preponderance (over 50%) of the evidence.
How Did We Get Here?
With an implicit understanding that judges generally are not capable of making these determinations, courts initially followed the test promulgated in Frye v. United States (1923) — general acceptance in the relevant science — for screening novel scientific evidence. While this test admittedly was a nebulous standard, in that it didn't define what "general acceptance" meant, or how the relevant scientific community was to be defined, it became the universally accepted test because it placed the determination of reliability (and therefore the ultimate question of admissibility), on those who knew most about the evidence — the people in the relevant science.
The use of this "general acceptance" standard incrementally grew. Courts began to apply it to novel methodologies that employed accepted scientific principles. Ultimately, courts extended the test to require general acceptance of novel applications of accepted principles and methodologies (usually in toxic tort cases when causation was being assessed). At some point, of course, the use of the test became unfair because the issue before the court was so specialized that it was of no interest to the scientific community as a whole. Consequently, general acceptance could never be proven. Therefore, some courts began to employ a balancing test, whereby the presiding judge made her own independent assessment of reliability.
With the adoption of the Federal Rules of Evidence in 1973, revolutionary changes were enacted with regard to expert testimony. The two most significant were in Rules 702 and 703. Rule 702 eliminated the "necessity" standard. It made expert testimony admissible if it merely assists the finder of facts. Rule 703 expanded the basis upon which the experts can form admissible opinions. It permits reliance on inadmissible evidence if it is shown to be "of a type reasonably relied upon by experts in the particular field" in forming opinions on the subject matter. Nowhere in this new evidence code, and the accompanying Advisory Committee Notes, was there even a suggestion that the standard by which scientific evidence is to be screened was being reassessed, much less changed.
Nevertheless, in Daubert v. Merrell Dow Pharmaceuticals, Inc.(1993) the Supreme Court held that the Evidence Code had staged a silent revolution — an interesting conclusion since, at that time, virtually every state and federal jurisdiction followed the Frye test. Because the standard for the admissibility of expert testimony in Rule 702 was stated simply as "scientific . . . knowledge that will assist the trier of fact," the Court held that scientific validity was no longer to be determined exclusively by the scientific community's acceptance. Validity was now to be determined by the presiding judge. The balancing advocates appears to have prevailed.
Regardless of the accuracy of this assessment, the Court in Daubert theoretically shifted the focus of the screening, or gate-keeping, function from the scientists to the judge. The judge is now charged with determining the reliability of science that the scientists cannot agree upon. Because of the way in which this usually will be accomplished, Daubert's shift in focus, and therefore change in the standard of admissibility, may be more apparent than real.
You Can't Get There From Here
The Daubert decision offered four nonexclusive factors that should be considered by the presiding judge in screening scientific evidence: (1) testing; (2) peer review; (3) potential rate of error; and (4) general acceptance. How the delineation of these factors will help judges move away from reliance on people in the relevant science (the Frye approach) is unclear when judges generally do not have the expertise to evaluate and apply them.
For example, most judges employing these factors will have no independent basis for evaluating the tests that have been applied to the scientific principles and methodologies to assess their reliability. They will not have the expertise to determine an acceptable error rate, or measure the adequacy of the controls that were in place to insure that the principles and methodologies that were employed produced accurate results. If people in relevant scientific communities had conducted peer reviews and reported negatively, would the judge be privileged to ignore those reviews based on her own assessment of the science?
As a consequence, judges are compelled to return to the relevant scientific communities whose general acceptance of the principles and methodologies would have been pivotal under the Frye test. Therefore, in practical application, Daubert is really little more than Frye in drag. With neither the time nor the ability to perform this task, judges will simply "retool" Frye by anointing a single expert and substituting that expert's opinions for the relevant scientific or technological community's. The only difference is that judges now lay out the technical criticisms and concerns of the expert as if they were their own — in much the same way that some do with the opinions drafted by their law clerks.
In a recent article by Justice Breyer, Science in the Courtroom, in Issues in Science and Technology, he acknowledged the degree to which judges rely on guidance from the scientific community. He discussed the many ways that judges have managed this gatekeeping function. Some have appointed specially trained law clerks and special masters. Others have appointed independent experts, in addition to those presented by the parties. Still others have appointed a neutral panel of experts. All of the case-management techniques emphasize the continued importance of the opinions of those in the relevant sciences.
Problems of Consistent Standards in Screening Continue
The degree to which experts continue to be crucial to the screening function accentuates the urgency of the unaddressed second problem — the unstated burden of persuasion, or the level of reliability that must be demonstrated. In the hard sciences, where scientific truths are being sought through research, the standard of reliability is often above 98%. Why would courts adopt that standard for determining whether the science can be employed in a judicial proceeding where liability is being determined by a preponderance (over 50%) of the evidence? As long as we continue to resort to people in the sciences for advice and direction, the elevated scientific standards will continue to find their way into admissibility decisions.
The Federal Rules of Evidence Advisory Committee has refused to even consider a proposal by the Washington College of Law Evidence Project to incorporate an explicit level of reliability in the expert opinion rules. Advisory Committee members expressed the belief that Rule 104 establishes a preponderance of the evidence standard for admissibility that is applicable to all rules.
The Committee's decision is unfortunate for several reasons. First, there is no guarantee that the Supreme Court will concur in the conclusion that Rule 104 establishes the standard of reliability for scientific evidence since the sciences employ a much higher level of reliability — something that is often more akin to "beyond a reasonable doubt." Then we must confront the problem of how to coordinate a preponderance standard with a "beyond a reasonable doubt" standard?
A decade ago a similar problem arose with regard to the relationship of Rules 403 and 609. Rule 403 generally regulates the standard for the exclusion of relevant evidence because of potential unfair prejudice. Rule 609 controls the admissibility of prior convictions for impeachment purposes. The Advisory Committee withdrew proposed revisions to Rule 609 that explicitly referred to Rule 403, on the belief that the obvious relationship would be confirmed by the Supreme Court in Green v. Bock Laundry Machine Co. To their surprise, however, the Court did not find what was so obvious to the Committee. Consequently, a specific reference to Rules 403 subsequently had to be incorporated in Rule 609.
Second, the standard by which scientific testimony is screened is influenced by the disproportionate impact that such evidence will likely have on lay jurors. As a consequence, it might be expected that many judges believe that a higher standard than preponderance must be employed when determining the reliability of such evidence. This, of course, is not known because no surveys have been conducted by the Advisory Committee. Such information is generally not volunteered in the written opinions of judges because they understand that the less they say about subjects about which there is uncertainty, the less likely it is that they will be reversed on appeal — in silence, appellate courts assume that trial courts applied appropriate standards.
Finally, even if preponderance were generally acknowledged as the standard of reliability, constant reliance upon experts, who employ a much higher standard of reliability, inevitably will influence admissibility decisions. As a consequence, relevant testimony that should have been heard will be excluded. This cannot be avoided if the standard by which reliability is being measured is not explicitly delineated and the judges instructed on how to integrate the lesser standard with the higher standards being employed by those advising them.
The Advisory Committee's Inadequate Response
In predictable fashion, the Federal Judicial Conference through its Advisory Committee on the Federal Rules of Evidence has done nothing to alleviate the confusion created by Daubert. Recently, its response has been simply to rubber-stamp the Supreme Court's interpretation of Rule 702. Rather than promulgating a different rule that adequately addresses the problems that are now being confronted (which the Advisory Committee has the power to do since the Court was only interpreting a rule), the Committee simply codified the inadequate directives of Daubert to ensure that the testimony is "the product of reliable facts." "[T]he product of reliable principles and methods," that are "reliably" applied to the facts. (Emphasis added)
Contrary to the claims by the Chemical Manufacturers Association and the Defense Research Institute, in a recent article in the ABA Journal by William C. Smith entitled No Escape From Science (August, pp. 60-66), these revisions neither clarify the gate-keeping function of the courts nor set "meaningful standards for determining the reliability and the admissibility of testimony." They leave the test as nebulous as it was under the common law. But unlike the common law, the responsibility is now reassigned to the judge who has no external working standard to bring to the decision.
Where Do We Go From Here?
Under Frye the question of reliability, and therefore admissibility, was delegated to the sciences. Each scientific discipline had a different standard of reliability that was accepted by he courts on its face. If judges, rather than the scientists, are now expected to make this determination, they must be given a standard by which reliability is to be measured. Until we clarify the gatekeeper's screening mechanism, the secret password will remain "general acceptance in the relevant science," and too many admissibility decisions will be as unpredictable and unfair as they were under Frye.
The Manual on Science in the Courtroom that Justice Breyer wrote about is a worthwhile endeavor. It cannot meaningfully accomplish its goal, however, until other problems are acknowledged and addressed. This requires a meaningful public debate that can happen only if a leadership role is taken by those with the power to change the rules. If Chief Justice Rehnquist and the Federal Judicial Conference's Advisory Committee on the Federal Rules of Evidence are not willing to assume this responsibility, Congress should reassert its authority over the rules and refine the evidence code that it had the wisdom to promulgate more than a quarter of a century ago.
Return to the evidence project home page