Edward J. Imwinkelried is a professor of law at the University of California at Davis. He is the former chair of the Evidence Section of the American Association of Law Schools. Professor Imwinkelried consulted with the the plaintiffs' attorneys in preparing briefs before the United States Supreme Court in Daubert. He is an original member of The Champion Advisory Board. This article is reprinted with the kind permission of Criminal Law Bulletin.
This article was originally published as "Forensic Science: The Second Prong of The Daubert Test: Disturbing Implications of Two Recent Civil Cases," in Criminal Law Bulletin, a West Group Publication, November-December 1997.
In 1993, the United States Supreme Court handed down its decision in Daubert v. Merrell Dow Pharmaceuticals, Inc.1 Prior to Daubert, the Frye general acceptance test had governed the admissibility of scientific testimony in most federal courts.2 Under that standard, an expert had to base any purportedly scientific testimony on a theory or technique which was generally accepted in the pertinent scientific community. In Daubert, the Court ruled that the general acceptance test had not survived the enactment of the Federal Rules of Evidence in 1975. The Court acknowledged that the general acceptance test was the almost universal common-law standard,3 but the Court reasoned that the Federal Rules had superseded that test. The Rules contained no statutory language codifying the general acceptance test; and the Court concurred with the late Professor Edward Cleary, the Reporter for the Rules, that "in principle, under the Federal Rules no common law of evidence remains."4
The Court substituted an empirical validation standard derived from the language of Federal Rule 702. That statute refers to "scientific . . . knowledge."5 Writing for the Daubert majority, Justice Blackmun declared that to qualify as reliable "scientific knowledge," an expert's testimony must be validated by scientific methodology. Justice Blackmun described the methodology in classical, Newtonian terms, that is, the forms of hypotheses of engaging in experimentation and observation to verify or falsify the hypotheses.
The Daubert decision immediately became a cause célèbre. The Supreme Court's decision was front-page news in most metropolitan areas. Law reviews rushed to conduct symposia devoted entirely to dissecting Daubert.6 Most of the commentary focused on the implications of the Daubert Court's definition of "scientific knowledge."7
Although that definition is an important component of the Daubert opinion, it is sometimes forgotten that that definition was only one of the two prongs of the test enunciated by Justice Blackmun. After setting out his understanding of the scientific method,8 the Justice quoted another part of Rule 702 and wrote: "Rule 702 further requires that the evidence or testimony 'assist the trier of fact to understand the evidence or to determine a fact in issue.'"9 Later in the opinion, Justice Blackmun stated the majority's holding:
Faced with a proffer of expert scientific testimony, then, the trial judge must determine at the outset, pursuant to Rule 104(a), whether the expert is proposing to testify to (1) scientific knowledge that (2) will assist the trier of fact to understand or determine a fact in issue.10
While the Daubert majority lists the judicial inquiry about the helpfulness of the testimony as a separate prong of the new test, the Daubert opinion itself describes that prong in relatively innocuous terms. In the only extended discussion of the prong in the opinion, Justice Blackmun asserts:
This condition goes primarily to relevance. "Expert testimony which does not relate to any issue in the case is not relevant and, ergo, non-helpful." 3 Weinstein & Berger para. 702[02]. . . . See also United States v. Downing, 753 F.2d 1224, 1242 (CA3 1985)("An additional consideration under Rule 702 -- and another aspect of relevancy--is whether expert testimony proffered in the case is sufficiently tied to the facts of the case that it will aid the jury in resolving a factual dispute"). The consideration has been aptly described by Judge Becker as one of "fit." Ibid. The study of the phases of the moon . . . may provide valid scientific "knowledge" about whether a certain night was dark, and if darkness is a fact in issue, the knowledge will assist the trier of fact. However . . ., evidence that the moon was full on a certain night will not assist the trier of fact in determining whether an individual was unusually likely to have behaved irrationally on that night. Rule 702's "helpfulness" standard requires a valid scientific connection to the pertinent inquiry as a precondition to admissibility.11
The Justice evidently conceived of the prong primarily as a corollary of the relevance requirement. In federal practice, the relevance threshold is quite lax.12 In light of the relaxed relevance test in federal court, there was good reason to assume that the second prong of the Daubert test would have few teeth and prove to be non-controversial.
However, that assumption has proven false. In two recent cases authored by distinguished federal jurists, Daubert II13--the Ninth Circuit's opinion on remand -- and Hall v. Baxter Healthcare Corp.,14 the courts have breathed new life into the second prong of the Supreme Court's new admissibility test. The first part of this article describes the court's treatment of the second prong in both cases. The second part of the article argues that the courts' treatment of the helpfulness requirement is dubious as a matter of admissibility analysis. The court's treatment tends to confuse admissibility and sufficiency -- a confusion that could have a dramatic impact on the admissibility of prosecution scientific testimony. The third part contends that the courts' treatment of the helpfulness prong is suspect even as a matter of sufficiency analysis.
Courts' Treatment of 'Helpfulness Prong'
In 1995 on remand to the Ninth Circuit, Judge Kozinski wrote the opinion for the court of appeals.15 As in the original Ninth Circuit opinion, Judge Kozinski concluded that the plaintiffs' purported scientific testimony was inadmissible and that without that testimony, Merrell Dow was entitled to summary judgment.In the course of developing his argument that the plaintiffs' testimony was inadmissible, Judge Kozinski resorted to both prongs of the new Daubert test.
The judge invoked the first prong to justify his ruling that the testimony by Dr. Palmer was inadmissible.16 As Judge Kozinski stressed, Dr. Palmer was "the only [plaintiffs'] expert willing to testify 'that Bendectin did cause the limb defects in each of the [plaintiffs'] children.'"17 The judge found that the information Palmer relied on was too "limited" to pass muster under the validation test.18 Quoting a Sixth Circuit opinion,19 Judge Kozinski concluded that "[p]ersonal opinion, not science, is testifying here."20 The judge specifically stated that he was holding Dr. Palmer's testimony "inadmissible under the first prong of the Rule 702 analysis."21
Unlike Dr. Palmer, the other plaintiffs' experts stopped short of opining that Bendectin had either definitely or even probably caused the limb defects suffered by the plaintiffs' children. In Judge Kozinski's words, "they were willing to testify only that Bendectin is 'capable of causing' birth defects."22 To exclude their testimony, Judge Kozinski explicitly turned to the second prong of the Daubert test.23 Initially, Judge Kozinski merely quoted Justice Blackmun's reference to "fit" and "a valid scientific connection to the pertinent inquiry."24 However, immediately after quoting Justice Blackmun, Judge Kozinski added a significant dimension to the second prong:
Here, the pertinent inquiry is causation. In assessing whether the proffered expert testimony 'will assist the trier of fact' in resolving the issue, we must look to the governing substantive standard, which in this case is supplied by California tort law. California tort law requires plaintiff to show not merely that Bendectin increased the likelihood of injury, but that it more likely than not caused their injuries. In terms of statistical [epidemiological] proof, this means that plaintiffs must establish not just that their mothers' ingestion of Bendectin increased somewhat the likelihood of birth defects, but that it more than doubled it--only then can it be said that Bendectin is more likely than not the source of their injury.25
Reviewing the record below, Judge Kozinski noted that none of the plaintiffs' epidemiological experts claimed that a mother's ingestion of Bendectin during the first trimester more than doubled the risk of birth defects.26 In epidemiological analysis, the expert compares the rate of a disease in an exposed population to the incidence in an unexposed population.27 The comparison yields a relative risk or risk ratio. Although some of the plaintiffs' experts claimed that Bendectin use increased the risk, none was prepared to testify to a study finding that the relative risk exceeded two28 -- the epidemiological equivalent of proof by a preponderance of the evidence.29 They could testify only that "Bendectin could possibly have caused plaintiffs' injuries."30 Consequently, in Judge Kozinski's mind, the "studies [relied upon by the plaintiffs' experts other than Dr. Palmer] would not be helpful, and indeed would only serve to confuse the jury. . . ."31 The judge reiterated that "the substantive legal standard has always required proof of causation by a preponderance of the evidence."32 Those opinions were therefore "inadmissible under the second prong of Fed. R. Evid. 702."33
In 1996, District Judge Robert Jones filed his opinion in Hall v. Baxter Healthcare Corp.34 Unlike Daubert, Hall involved breast implants. The plaintiffs contended that their breast implants were the cause of a host of illnesses. As in Daubert, the procedural setting was a defense in limine motion to exclude plaintiffs' expert testimony -- not a summary judgment motion. The parallel continues; as in Daubert II, Judge Jones decided to grant the defense motion and bar much of the plaintiffs' proposed testimony. In framing his decision, Judge Jones drew heavily on Judge Kozinski's opinion in Daubert II.
After describing the "scientific knowledge" prong of the Daubert test,35 Judge Jones took up the topic of the helpfulness or "fit" prong.36 After quoting Justice Blackmun's relatively unassuming language prescribing the prong,37 Judge Jones quickly turned to Judge Kozinski's opinion in Daubert II. Judge Jones commented that elucidating "the second prong," the Ninth Circuit had "explained that in assessing whether proffered expert testimony 'will assist the trier of fact' in resolving the causation issue, the court must look to the substantive standard . . . ."38 Just as California tort law necessitated proof of causation by a preponderance of the evidence,39 the governing Oregon law demanded that the plaintiffs establish causation by "[t]he substantive standard" of "more likely than not."40 Judge Jones then flatly asserted: "Under this substantive standard, if an expert cannot state the causal connection in terms of probability or certainty, the expert's testimony must be excluded under the second prong of Rule 702."41
Admissibility Analysis
The importance of the holdings in Daubert II and Hall emerges once we appreciate two points. The first is that the Federal Rules of Evidence themselves do not insist that the expert vouch for his or her opinion to any particular degree of scientific certainty or probability. Rule 704 governs the phrasing of otherwise admissible expert opinions,42 and it is silent on the question of whether the expert must be prepared to couch his or her scientific opinion as a reasonable scientific certainty or probability. There were common-law decisions imposing that requirement,43 but the Federal Rules nowhere codify that requirement. As previously stated, in Daubert the Court echoed Professor Cleary's position that "in principle, under the Federal Rules no common law of evidence remains."44 The common-law decisions are therefore no longer controlling.The second point is that in epidemiology, standing alone a showing of a relative risk exceeding 2.0 is usually deemed sufficient to support the inference that "an agent was more likely the cause of a disease than not . . . ."45 In short, Daubert II and Hall are arguably authority that to be of enough assistance to be admissible under Federal Rule 702, standing alone purported scientific testimony must possess sufficient probative value to prove the fact in issue. Shortly after the rendition of the Supreme Court's Daubert decision, one perceptive commentator voiced the fear that in applying Daubert, the lower courts would improperly conflate the standards for admissibility and legal sufficiency.46 That fear seems to have been realized. This is a matter of concern because the conflation of admissibility and sufficiency standards is dubious as a matter of logic, precedent, and policy.
The incorporation of sufficiency standards into the second prong of Daubert's admissibility test is certainly debatable as a matter of logic. Both at common law and under the Federal Rules, to be admissible, an individual item of evidence must be relevant.47 However, the test for logical relevance is minimal.
In Dean McCormick's view, the proper test is whether the item of evidence has any impact at all on the balance of the probability of the existence of any fact in issue.48 The Dean drew a sharp distinction between relevancy and sufficiency.49 Even if an item of evidence fell far short of being sufficient to establish so much as a bare probability of the existence of a fact in issue, the item could be relevant.50 In his classic statement, "a brick is not a wall."51 Suppose, as was evidently the case in Daubert II, that the plaintiffs' epidemiologists were prepared to testify at most that their studies documented an increased risk of birth defects among children born to women who had used Bendectin.52 To be sure, that testimony does not satisfy the 2.0 relative risk threshold that must be exceeded to make out a submissible, sufficient case to sustain the plaintiff's initial burden of production on causation.
The issue here is relevance, not sufficiency. Testimony about an elevated risk is relevant as Dean McCormick understood the concept, and the Advisory Committee Note to Rule 401 endorses McCormick's understanding.53 Quoting a "brick is not a wall," the Committee emphasizes the need to "avoid[] confusion between questions of admissibility and questions of the sufficiency of the evidence."54
The evidence that is being excluded under Daubert II and Hall is not only relevant; ex hypothesi, it is also reliable. Both cases rely on the second prong of the Daubert test. There is no need to reach that prong unless, at the first prong of the analysis, the court has already determined that the proposed testimony qualifies as reliable "scientific knowledge" under Rule 702. At this stage in the analysis, the court is at least assuming arguendo that the proffered testimony is not "junk science"55 in the sense that it does not rest on sound scientific methodology. Daubert II and Hall both declare that even relevant, reliable scientific testimony can be excluded by virtue of their refinement of the second, helpfulness prong.
Further, it is difficult to reconcile Daubert II and Hall with the prior precedents, including Justice Blackmun's opinion in Daubert. To begin with, although the Justice mentions the second prong of the test at several junctures in his opinion, he never so much as faintly suggests that the lower court judge must find that the opinion is legally sufficient to support a finding of fact before concluding that the proponent has met the second prong. Moreover, the Justice goes to some length to differentiate questions of admissibility from issues of legal sufficiency. He cites Civil Rules 50 and 56, respectively, as permitting trial directed verdict and pretrial summary judgment when the judge concludes "that the scintilla of evidence presented supporting a position is insufficient to allow a reasonable juror to conclude that the position more likely than not is true."56 The Justice noted that the trial judge may apply those "devices" to "scientific testimony . . . meet[ing] the standards of Rule 702."57 However, there is no intimation that the judge should incorporate the tests for those "devices" into "the standards of Rule 702."58
In the passage in Daubert in which he explains the second prong, Justice Blackmun specifically cites59 the Third Circuit's 1985 decision in United States v. Downing.60 In that decision, Judge Becker coined the expression "fit."61 On remand in that case, Judge Becker directed that the trial judge inquire "whether the expert testimony proffered in the case is sufficiently tied to the facts of the case that it will aid the jury in resolving a factual dispute."62 In Downing, the defense attempted to introduce expert testimony about the unreliability of eyewitness identification. Judge Becker wrote that the defense "offer of proof should establish the presence of factors (e.g., stress, or differences in race . . . between the eyewitness and the defendant) which have been found by the researchers to impair the accuracy of eyewitness identification."63 Judge Becker mandated that on remand the judge address the second prong because "[t]he encounters between the identification witnesses and appellant were apparently not under conditions of stress, nor as far as we know, was there a potential cross-racial identification problem."64 The Third Circuit's opinion strongly suggested that the trial judge should exclude the proffered testimony if the testimony related to factors which were not present in the instant case.
However, the opinion does not imply that if those factors were present, the trial judge could or must exclude the testimony simply because the expert could not vouch that those factors would definitely render the eyewitness identification inaccurate. The Third Circuit treats the "fit" issue as a relevancy issue, not a sufficiency problem.
In Downing,, Judge Becker said only that the trial judge should, for example, bar testimony about the unreliability of cross-racial eyewitness identification if there was no cross-racial identification in the case. That testimony might satisfy the lax relevance test of Rule 401; the testimony establishes the fallibility of eyewitness identification, and defense counsel might contend that proof of such fallibility is relevant whenever the prosecution relies on an eyewitness identification. However, the Downing court is surely correct in indicating that if there is no cross-racial identification in the case, a trial judge would be justified in excluding the testimony under Rule 702 as being of minimal assistance. Downing was not a case like Daubert II or Hall where the expert testimony related to a lively issue in the case such as causation but, standing alone, the expert's opinion was legally insufficient to prove the issue.
Before there was Daubert, there was Downing; and before there was Downing, there was the Advisory Committee Note on Rule 702. The Note mentions the textual requirement that the expert testimony "assist the trier."65 However, the Note describes that requirement in a limited fashion:
There is no more certain test for determining when experts may be used than the common sense inquiry whether the untrained layman would be qualified to determine intelligently and to the best possible degree the particular issue without enlightenment from those having a specialized understanding of the subject involved in the dispute. Ladd, Expert Testimony, 5 Vand.L.Rev. 414, 418 (1952). When opinions are excluded, it is because they are helpful and therefore superfluous and a waste of time. 7 Wigmore
{ 1918. The Note seems to require the judge to inquire whether the topic in question warrants, in the words of Rule 702, "specialized knowledge" or rather whether it is a matter which the jury can competently resolve without the benefit of expert testimony. As Dean Ladd explained in the article cited by the Note, the question is whether "the triers of fact are confronted with issues which cannot be determined intelligently on the basis of ordinary judgment and practical experience gained through the usual affairs of life."66 Restating his view, Dean Ladd added that "expert testimony is not admissible to prove or disprove matters within common knowledge as to which facts may be so described that the triers of fact may form a reasonable opinion themselves."67 Dean Ladd argued against the admission of opinions on "matter[s] of common knowledge."68 On such matters, expert testimony is "unnecessary."69 However, it could hardly be said that the expert opinions proffered in Daubert II and Hall were excludable on the ground that they addressed matters of common knowledge.
More importantly, like Downing and Justice Blackmun's opinion in Daubert, the Note contains no hint that the trial judge must directly incorporate the legal sufficiency standard into the decision on whether the expert testimony satisfies the helpfulness prong. The Note nowhere indicates that under Rule 702, a judge may exclude an opinion on the sole ground that, without more, the expert opinion would be legally insufficient to sustain the proponent's burden on the issue.
Finally, if Daubert II and Hall do mean that the judge should exclude scientific testimony as unhelpful unless standing alone the testimony is sufficient to meet the proponent's ultimate burden of proof, the cases will have a devastating impact on the government's ability to introduce scientific evidence in criminal cases.
Suppose that, seizing upon the language of Daubert II and Hall, a criminal defense attorney presents the following argument. As with scientific testimony proffered in those cases, prosecution forensic evidence must satisfy the second prong of Daubert. Daubert II teaches that to "assess[] whether the proffered expert testimony 'will assist the trier of fact,'" the judge "must look to the governing substantive standard."70 The rub for the prosecution is that here the "the governing substantive standard"71 is far more rigorous, namely, the constitutionally mandated standard of proof beyond a reasonable doubt.72 According to Hall,73 in a civil action, "under [the] substantive standard [of a preponderance of the evidence], if an expert cannot state the causal connection in terms of probability or certainty, the expert's testimony must be excluded under the second prong of Rule 702." Criminal defense counsel would argue that a fortiori, "under [the] substantive standard [of proof beyond a reasonable doubt], if a[ prosecution] expert cannot state [his or her conclusion] in terms of [that degree of probability], the expert's testimony must be excluded under the second prong of Rule 702."74
If the court were to embrace that argument, there is a good likelihood that the courts will begin excluding many types of prosecution scientific evidence which the government has routinely employed in the past to link the accused to crimes. Trace evidence, including hair,75 fiber,76 glass,77 and soil,78 will be particularly vulnerable to this attack. In the typical case involving trace evidence, an honest, competent analyst will have to concede that he or she cannot truthfully vouch that standing alone, his or her analysis establishes any fact in issue beyond a reasonable doubt. If the court adopts a broad reading of Daubert II and Hall, those cases will have a devastating impact on the prosecution's ability to marshall scientific evidence against an accused.
To minimize that impact, later courts may attempt to distinguish Daubert II and Hall. For example, a court bent on limiting the precedential value of Daubert II might point to the passage in which Judge Kozinski opines that the introduction of the plaintiffs' epidemiological evidence "would only serve to confuse the jury."79 A court could conceivably argue that rather than announcing any sweeping rule, the Ninth Circuit was merely making a case-specific, ad hoc ruling. As attractive as that option might at first appear, that option seems flawed. To begin with, if the Ninth Circuit had intended to make a limited, case-specific ruling, it would have more properly predicated its decision on Rule 403. Rule 403 is the source of the court's authority to make ad hoc rulings that in a given case, probative dangers such as the risk of confusion outweigh the probative worth of the evidence.80 Rule 403 expressly mentions "confusion of the issues" as a probative danger cognizable under that statute. In Daubert II, the court did not appeal to Rule 403; rather, it purported to rest its decision squarely on the second prong of Daubert construing Rule 702.81 Moreover, Hall certainly reads Daubert II as announcing a hard-and-fast requirement under Rule 702.82
At another level, later courts might endeavor to limit the scope of Daubert II and Hall by construing those opinions as prescribing a categorical rule confined to a particular type of case, evidence, or issue. It might be argued that the cases are restricted to civil actions. However, Daubert itself did not limit its holding to civil cases, and numerous subsequent decisions have applied Daubert in the criminal context.83
Or it might be contended that Daubert II and Hall deal only with the problem of causation. However, Daubert itself does not indicate that its validation standard governs only scientific evidence offered on the question of causation, and again subsequent courts have applied the standard to forensic testimony introduced on a wide variety of issues.84
In civil negligence and product liability cases, causation is an essential element of the plaintiff's cause of action; but the accused's identity as the perpetrator is just as essential an element of the charged offense in the typical prosecution. Likewise, in a drug prosecution, the identity of the substance seized from the accused as a contraband drug is a required element of the charged crime; and that identity is arguably as essentially a scientific question as the issue of general medical causation.
Alternatively, a court intent on distinguishing away Daubert II and Hall could argue that the cases deal only with epidemiological evidence. However, that attempted distinction will also fail. At the end of his opinion in Daubert II, Judge Kozinski stated that his refinement of the second prong of Daubert applied to the plaintiffs' "animal studies and chemical structure analyses" as well as their epidemiological reanalysis.85 For his part, after stating a single set of admissibility standards,86 Judge Jones proceeded to apply that set of standards to four different types of scientific evidence to: "epidemiology; rheumatology; immunology/toxicology; and polymer chemistry."87
The most fundamental defect in any of these attempts to restrict Daubert II and Hall is that in neither case did the court present any textual or policy argument to limit its refinement of the second prong to a particular type of case, issue, or scientific evidence. The fairest reading of both opinions seems to be that the courts were announcing a categorical rule that in determining the admissibility of scientific evidence under the second prong, the judge must always assess the helpfulness of the evidence in light of the pertinent ultimate burden of proof. More specifically, they seem to direct the judge to conclude that the evidence fails the helpfulness standard unless the opinion is certain enough to be legally sufficient proof of a fact in issue. The enforcement of that rule in criminal cases would deal a tremendous blow to prosecutors. The conclusion seems inescapable that Daubert II and Hall have read too much into the second prong of the Daubert test for admissibility.
Legal Sufficiency Analysis
Perhaps Daubert II and Hall would be less troublesome if we recast them as legal sufficiency decisions. Suppose that in those cases, rather than filing in limine motions challenging the admissibility of the plaintiffs' expert testimony, the defense had merely moved for summary judgment. Assume further that all the plaintiffs' epidemiological experts proposed to testify only that "Bendectin could possibly have caused plaintiff's injuries."88 Could the outcome in Daubert II then be rationalized on the theory that as a matter of law, the plaintiffs' expert testimony was legally insufficient to meet their initial burden of production?The difficulty with that theory is that in Daubert, the plaintiffs were not relying exclusively on epidemiological testimony. As the Ninth Circuit noted, the plaintiffs presented three different types of expert testimony to establish causation: epidemiology, animal studies, and chemical structure analyses.89 In two footnotes in Hall, Judge Jones recognized the possibility that a proponent might "combine[]" several types of expert testimony90 and argue that "the collective weight"91 of the testimony sufficed to meet the burden.
If three different types of scientific experts testified to "a good possibility" of a causal nexus, cumulatively the mosaic of their testimony might be adequate to sustain the burden. The question is not whether a scientist would regard the hypothesis as proven. In a civil action, the issue is "reasonable and fair-minded" laypersons would be willing to draw a permissive inference from the entire body of evidence presented by the burdened party.92 When the various experts vouch for several different types of evidence pointing to a connection, it seems arbitrary to rule that it would be irrational for the jurors to find a connection.
If that type of sufficiency reasoning is unacceptable, prosecutors will be at a tremendous disadvantage. The Supreme Court has decreed that there is a higher standard for satisfying the initial burden of production in criminal cases.93 In that setting, the question is whether the hypothetical juror could find the existence of every element of the charged offense beyond a reasonable, lingering doubt.94 Suppose that a prosecutor presents a wide variety of types of trace evidence to link the accused to the crime scene. The prosecutor proffers expert testimony about hair, fiber, glass, soil, and paint.95 Each expert attests that his or her findings are consistent with the hypothesis that the accused was at the crime scene; but each witness candidly concedes that standing alone, his or her finding does not prove the accused's presence at the scene. If the combination of epidemiology, animal studies, and chemical structure analysis is insufficient on the facts of Daubert II, a criminal defense attorney could argue that by parity of reasoning, this combination of prosecution experts does not amount to a submissible case.
In Hall, Judge Jones did add a dictum that reliance on a combination of types of expert testimony would be permissible if the burdened party can "demonstrate that [the various types of expertise] differ in some significant way. . . ."96 That dictum holds out the hope that plaintiffs and prosecutors may piece together several types of expert testimony to satisfy their burden of production when they present additional scientific testimony that each type of evidence is truly distinct. Conceivably, a court might even restrict the requirement for such testimony to cases such as Daubert II and Hall. In a criminal case in which the prosecutor relied upon hair and soil evidence, it would be obvious even to lay jurors that the two types of trace evidence are different and independent. It might be less obvious to the average juror how "rheumatology [and] immunology/toxicology" differ.97 It could be argued that to decide intelligently whether and how to combine evidence furnished by those disciplines, the jury needs expert guidance as to the relationship between the disciplines.
The most extreme step for the courts to take would be to mandate that the burdened party present expert statistical testimony quantifying the probability created by each type of scientific evidence and explaining how the probabilities cumulate. In the past, the principal battles over statistical testimony have been waged over the questions of whether such testimony is admissible at all98 and whether a "naked statistical evidence" presentation is ever legally sufficient to reach the jury.99 In short, the issues have been whether the proponent may either present statistical testimony or rely exclusively on that type of testimony to meet his or her burden of production. It would be unprecedented for a court to leap to the polar extreme and adopt the view that the proponent must present statistical testimony to have a submissible case when none of the proponent's individual scientific experts can present testimony which, without more, satisfies the burden.
If all the relevant probabilities could be quantified, there might be techniques for cumulating them.100 In many cases, though, the real difficulty would be quantifying the probabilities in the first place. What probability of guilt arises from the fact that a hair strand found on the accused's clothing is microscopically similar to a sample of the deceased victim's head hair? How does the expert estimate the probability of guilt created by a showing that the elemental composition of the soil found on the accused's shoe matches that of the soil found at the crime scene? To generate such probabilities, the expert would need a database; and in the case of most varieties of trace evidence, such databases do not exist.101 It would be arbitrary to assign a numerical probability to the significance of a match.102 If the proponent of scientific testimony were required to reliably quantify the probability created by each type of scientific evidence, it would be far more difficult to introduce such evidence at trial.103
Castastrophic Effect
If one agrees with Huber104 and Angell105 that "junk science" is rampant in American courts, one might be inclined to applaud the results in Daubert II and Hall. However, whether or not one agrees with the results in those two cases, one thing is clear: The reasoning of those courts about the second prong in the Daubert test is debatable and disturbing. In their exegesis of the helpfulness prong, the courts seemed to conflate admissibility and sufficiency analysis.106 As a result, the cases lend themselves to the interpretation that even relevant, reliable scientific testimony is inadmissible when, standing alone, the testimony would be insufficient to satisfy the proponent's ultimate burden of proof. Epidemiological evidence establishing a relative risk ratio exceeding 2.0 would suffice to satisfy the burden in a civil case; and Daubert II and Hall strongly suggest that any epidemiological finding under 2.0 is inadmissible because it would not assist the jury -- the second prong of Justice Blackmun's test.In his opinion, Justice Blackmun was careful to distinguish between the issues of admissibility and legal sufficiency. It consequently seems wrong-minded to construe Daubert as mandating the inadmissibility of scientific testimony simply because, without more, the testimony would be inadequate to satisfy the legal sufficiency test. Under Rule 403107 and the second prong of Daubert, it is legitimate for the trial judge to weigh the factor of the quantum of probative value of the proffered testimony.
It is quite another matter, however, to hold that without more, the testimony must suffice to support a finding of the existence of a consequential fact in the case. Neither Downing nor the Advisory Committee Note to Rule 702 go that far. Worse still, the reasoning in Daubert II and Hall could potentially have a catastrophic effect on the prosecution's ability to marshall scientific evidence at criminal trials. Daubert II and Hall state that in assessing the helpfulness of the proponent's expert testimony, the judge must use the standard of reference of the ultimate burden of proof. There are notable exceptions, namely, fingerprint analysis108 and DNA evidence;109 but, considered in isolation, most varieties of prosecution scientific evidence do not satisfy the standard of proving guilt beyond a reasonable doubt. It is inconceivable that jurists as insightful as Judges Kozinski and Hall meant to damn virtually all prosecution scientific testimony to inadmissibility. However, to prevent the Daubert II and Hall opinions from triggering that untoward result, the courts must either develop a creative, limited reading of those opinions -- or eschew the broad language in those opinions.110
Notes
1. 509 U.S. 579 (1993).2. P. Giannelli & E. Imwinkelried, Scientific Evidence 1-5 (2D Ed. 1993).
3. Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579, 585 (1993).
4. Id. at 588.
5. Fed. R. Evid. 702, 28 USCA.
6. See generally Symposium, Scientific Evidence After the Death of Frye, 15 Card. L .Rev. 1745 (1994).
7. E.g., Bert Black, Francisco J. Ayala & Carol Saffran-Brinks, Science and the Law in the Wake of Daubert: A New Search for Scientific Knowledge, 72 Tex. L. Rev. 715 (1994).
8. Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589-90 (1993).
9. Id. at 591.
10. Id. at 592-93.
11. Id. at 591-92.
12. United States v. Nason, 9 F.3d 155, 162 (1st Cir. 1993)("The threshold for relevance is very low under Federal Rule of Evidence 401"), cert. denied, 114 S. Ct. 1331 (1994); United States v. Casares-Cardenas, 14 F.3d 1283, 1287 (8th Cir.)("Relevance is established by any showing, however slight, that makes it more or less likely . . . ."), cert. denied, 115 S. Ct. 147 (1994); United States v. Foster, 986 F.2d 541, 545 (D.C. Cir. 1993)("Under this test, there is no such thing as 'highly relevant' evidence or . . . 'marginally relevant' evidence. Evidence is either relevant or it is not").
13. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311 (9th Cir. 1995).
14. 947 F. Supp. 1387 (D.Or. 1996).
15. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311 (9th Cir. 1995).
16. Id. at 1319.
17. Id.
18. Id.
19. Turpin v. Merrell Dow Pharamceuticals, Inc., 959 F.2d 1349, 1360 (6th Cir. 1992).
20. Daubert v. Merrell Dow Pharamceuticals, Inc., 43 F.3d 1311, 1319 (9th Cir. 1995).
21. Id. at 1321.
22. Id.
23. Id. at 1320.
24. Id.
25. Id.
26. Id. at 1321.
27. Michael Green, Expert Witnesses and Sufficiency of Evidence in Toxic Substances Litigation: The Legacy of Agent Orange and Bendectin Litigation, 86 NW. U. L. Rev. 643, 647 (1992).
28. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1321 (9th Cir. 1995).
29. Id.
30. Id. at 1322.
31. Id. at 1321.
32. Id. at 1322.
33. Id.
34. 947 F.Supp. 1387 (D.Or. 1996).
35. Id. at 1395-97.
36. Id. at 1397.
37. Id.
38. Id.
39. Id. at 1397-98.
40. Id. at 1398.
41. Id.
42. Fed. R. Evid. 704, 28 USCA.
43.Joseph, Less Than Certain Medical Testimony, 14 Trial 51 (Jan. 1978).
44. Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 588 (1993).
45. Linda A. Bailey, Leon Gordis & Michael Green, Reference Guide on Epidemiology, in Reference Manual On Scientific Evidence 121, 168-69 (Fed. Jud. Ctr. 1994).
46. Thomas J. Mack, Scientific Testimony After Daubert: Some Early Returns from Lower Courts, 30 Trial 23, 28-30 (Aug. 1994).
47. Fed. R. Evid. 401-02, 28 USC A.
48. C. McCormick, Handbook of the Law of Evidence 152 (1954).
49. Id. at 154, at 317.
50. Id.
51. Id.
52. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1321 (9th Cir. 1995).
53. Adv. Comm. Note, Fed. R Evid. 401, 28 USCA.
54. Id.
55. Peter W. Huber, Galileo's Revenge: Junk Science In The Courtroom (1991).
56. Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).
57. Id.
58. Id.
59. Id. at 591-92.
60. 753 F.2d 1224 (3d Cir. 1985).
61. Id. at 1242.
62. Id.
63. Id.
64. Id.
65. Adv. Comm. Note, Fed. R. Evid. 702, 28 USCA.
66. Mason Ladd, Expert Testimony, 5 Vand. L. Rev. 414, 418 (1952).
67. Id. at 419.
68. Id.
69. Id.
70. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1320 (9th Cir. 1995).
71. Id.
72. In re Winship, 397 U.S. 358 (1970).
73. Hall v. Baxter Healthcare Corp., 947 F. Supp. 1387, 1398 (D. Or. 1996).
74. Id.
75. P. Giannelli & E. Imwinkelried, Scientific Evidence 24-2, 24-3 (2d ed. 1993).
76. Id. at 24-4, 24-5.
77. Id. at 24-6, 24-7.
78. Id. at 24-10, 24-11.
79. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1321 (9th Cir. 1995).
80. Edward J. Imwinkelried, Judge Versus Jury: Who Should Decide Questions of Preliminary Facts Conditioning the Admissibility of Scientific Evidence?, 25 Wm. & Mary l. Rev. 577, 614-15 (1984).
81. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1322 (9th Cir. 1995).
82. Hall v. Baxter Healthcare Corp., 947 F. Supp. 1387, 1397-98 (D.Or. 1996).
83. P. Giannelli & E. Imwinkelried, Scientific Evidence 1-10 (1996 Cum.Supp.)(collecting cases).
84. Id.; G. Michael Fenner, The Daubert Handbook: The Case, Its Essential Dilemma, and Its Progeny," 29 Creighton L. Rev. 939 (1996).
85. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1322 (9th Cir. 1995).
86. Hall v. Baxter Healthcare Corp., 947 F. Supp. 1387, 1397-98 (D.Or. 1996).
87. Id. at 1402.
88. Daubert v. Merrell Dow Pharmaceuticals, Inc., 43 F.3d 1311, 1322 (9th Cir. 1995).
89. Id. at 1314.
90. Hall v. Baxter Healthcare Corp., 947 F.Supp. 1387, 1398 n.26 (D.Or. 1996).
91. Id. at 1398 n.28.
92. 2 McCormick, Evidence 338, at 433 (4th ed. 1992).
93. Id. at 433-35, citing Jackson v. Virginia, 443 U.S. 307 (1979).
94. Id.
95. See generally, 2 P. Giannelli & E. Imwinkelried, Scientific Evidence Ch. 24 (2d ed. 1993)(trace evidence).
96. Hall v. Baxter Healthcare Corp., 947 F.Supp. 1387, 1398 n. 26 (D.Or. 1996).
97. Id. at 1402.
98.1 P. Giannelli & E. Imwinkelried, Scientific Evidence 15-7(B), at 470-74 (2d ed. 1993).
99. Id. at 15-7(B), at 474-76.
100.1 P. Giannelli & E. Imwinkelried, Scientific Evidence 15-7(A), 17-9(A) (3d ed. 1991)(describing the multiplication or product rules and Bayes' theorem).
101. See generally , 2 P. Giannelli & E. Imwinkelried, Scientific Evidence Ch. 24 (2d ed. 1993).
102. In People v. Collins, 68 Cal. 2d 319, 438 P.2d 33, 66 Cal. Rptr. 497 (1968), the court faulted the prosecutor for essentially pulling the numbers assigned to various probabilities out of the air.
103. As in the case of the admissibility analysis in Daubert II and Hall, courts concerned about the implications of those cases for sufficiency analysis might endeavor to distinguish and limit those precedents. They might attempt to distinguish them on the ground that in those cases the experts proffered direct testimony on the issue of causation while in the hypothetical prosecutions the prosecutor is offering circumstantial evidence. It is true that the legal sufficiency cases recognize the distinction between direct and circumstantial evidence. McLaughlin v. LIU, 849 F.2d 1205, 1208 (9th Cir. 1988); Randle v. LaSalle Telecommunications, Inc., 876 F.2d 563 (7th Cir. 1989). In cases such as Daubert II, the expert contemplates testifying directly on the issue of causation while the prosecution's trace evidence experts will not opine expressly on the issue of identity. It could then be argued that the courts have greater freedom to scrutinize whether the plaintiff's direct testimony suffices to meet the plaintiff's initial burden of production.
There are several difficulties with this attempted distinction, though. To begin with, while the courts recognize the distinction between direct and circumstantial evidence in the sufficiency cases, the distinction usually cuts in the opposite direction; the courts ordinarily state that they have less power to secondguess the expert when the opinion is direct testimony. For example, while there is authority that the court may direct a verdict when the connection between the circumstantial evidence and the fact is issue is "implausible," the cases do not extend the "implausibility" standard to direct testimony.
Moreover, in this context, the purported distinction is illusory. It is true that in criminal cases, judges rarely permit the prosecution expert to testify directly to a fact in question. However, in both cases the expert's distinctively scientific contribution to the factfinding process is an inference just short of the ultimate fact. Perhaps civil cases more frequently allow the expert to take the final step and opine directly on causation, but the epidemiologist's scientific contribution is the finding of relative risk. While the criminal judge might preclude the trace evidence expert from testifying directly that the accused was at the crime scene, the expert can testify that the soil at the scene has the same elemental composition as the soil found on the soles of the accused's shoes.
104. Peter Huber, Galileo's Revenge: Junk Science in the Courtroom (1991).
105. M. Angell, Science On Trial: The Clash of Medical Evidence and the Law in the Breast Implant Case (1996); Marcia Angell, Shattuck Lecture -- Evaluating the Health Risks of Breast Implants: The Interplay of Medical Science, The Law, and Public Opinion, 334 New Engl .J. Med. 1513 (1996).
106. In Scientific Testimony After Daubert: Some Early Returns from Lower Courts, 30 Trial 23 (Aug. 1994), Thomas J. Mack warned that the courts might be tempted to merge the two standards.
107. Edward J. Imwinkelried & Robert G. Scofield, The Recognition of an Accused's Constitutional Right to Introduce Expert Testimony Attacking the Weight of Prosecution Scientific Evidence: The Antidote for the Supreme Court's Mistaken Assumption in California v. Trombetta, 33 Ariz. L. Rev. 59, 69 (1991).
108. 1 P. Giannelli & E. Imwinkelried, Scientific Evidence Ch. 15 (2d. ed. 1993).
109. 2 P. Giannelli & E. Imwinkelried, Scientific Evidence Ch. 18 (2d ed. 1993)
110. One court has already expressly rejected that interpretation of the second prong of Daubert. In Ambrosini v. Labarraque, 101 F.3d 129, 135 (D.C. Cir. 1996), the court wrote: "The dispositive question is whether the testimony will 'assist the trier of fact to understand the evidence or to determine a fact in issue,' . . . not whether the testimony satisfies the plaintiff's burden on the ultimate issue at trial."