JUDICIAL GATEKEEPING IN TEXAS

Introduction
     Expert witnesses have become fixtures in today's courts. From fiber comparisons to economic projections and psychiatric evaluations, the range of offered expertise covers the span of human knowledge. Hardly a case of any consequence goes to trial without expert witnesses of some kind.

     What is the trial judge's role in overseeing the testimony of expert witnesses? Unlike lay witnesses, whose testimony a jury can evaluate with common sense and experience, expert witnesses offer conclusions based on practices and knowledge beyond the ken of the average juror. As a consequence, testimony by unpoliced expert witnesses can have a potentially prejudicial effect on jurors, who may be more inclined to believe the testimony based solely on the witness' expert status. How is the trial judge to know whether the expert is merely speculating, or whether the evidence on which the expert bases his or her testimony is sufficient to support the conclusion? Judges certainly prevent lay witnesses from speculating and can direct verdicts when evidence is insufficient to support a finding. How then does the judge's screening of expert witnesses differ from this practice? 

     This question poses profound issues about the nature of the judge's role. To make an informed judgment about whether to admit an expert, the judge would seem to need at least a modicum of expertise in that witness' field. How else can a general trial judge be expected to say with assurance that the expert's opinion is speculative? Such a demand suggests the possible need for greater specialization within the judiciary, even the creation of scientific courts. More realistically, judges can meet this demand by appointing neutral experts to advise them on scientific issues. While these ideas may be at odds with conventional models of the judicial role, they speak forcefully to the fact that if judges do not police the outer boundaries of expert testimony, they open the gates to verdicts based on speculation rather than sound evidence.

     Should judges be gatekeepers with respect to expert testimony? And if so, how should judges perform this function? The United States Supreme Court brought this question sharply into focus with its 1993 decision in Daubert v. Merrell Dow Pharmaceuticals, Inc., which dealt with the standard in the Federal Rules of Evidence for admitting the testimony of scientific experts.⁽¹⁾ The Supreme Court articulated "a gatekeeping role for the judge,"⁽²⁾ ruling that "trial judges must ensure that any and all scientific testimony is not only relevant, but reliable."⁽³⁾ No longer were federal trial judges bound to apply the Frye standard, according to which they determined only whether the expert's methodology was "generally accepted" by other scientists in the expert's field.⁽⁴⁾ Now federal trial judges must decide for themselves whether an expert's methodology--the often highly complicated basis for his or her conclusion--is reliable.

     This new gatekeeping role for federal judges might not, at first, seem terribly important to state trial judges. After all, the Daubert decision concerned the interpretation of the Federal Rules of Evidence, which do not govern state courts. In addition, the case involved scientific testimony on the issue of causation in a toxic tort claim, which might seem like the type of case unlikely ever to make its way into a state trial court. But in fact, the issues of gatekeeping put in play by the Supreme Court are tremendously important to all trial judges. Like many other states, Texas models its rules of evidence on the Federal Rules, and in the case of DuPont v. Robinson,⁽⁵⁾ has "adopted" Daubert. Moreover, nothing in Daubert or Robinson necessarily limits the trial judge's new gatekeeping responsibilities to scientific experts in Agent Orange-like toxic tort cases. The gatekeeping role extends to expert scientific testimony in any kind of case, and perhaps to any kind of expert testimony, scientific or otherwise.

     Like much of the evidence offered under the Daubert standards, the judicial mechanisms for dealing with this evidence are still at the frontier stage. Lawyers will learn new means of both offering and challenging expert testimony, and judges must be ready. Challenges to the methodologies of experts from all fields have already appeared. For one prescient example, consider the dispute over the methodology of handwriting experts in the Oklahoma City bombing case. There, as in the trial courts of this state, judges face an unprecedented new demand, exacting and still loosely defined, one which stretches their capabilities and resources up to and perhaps beyond their conventional limits. 

     It is the intent of this paper and the papers that follow to explore the many questions surrounding the judge's gatekeeping role. Acting as an aerial map to the gatekeeping terrain, this paper examines the development of Texas law regarding scientific expert testimony, focusing primarily on DuPont v. Robinson and Merrell Dow Pharmaceuticals, Inc. v. Havner,⁽⁶⁾ and introduces the specific gatekeeping issues discussed in more detail by later papers.

DuPont v. Robinson: "Relevant and Reliable"
     In Robinson (involving expert testimony about whether a DuPont fertilizer injured the Robinsons' pecan trees), the Texas Supreme Court held that the Texas Rules of Civil Evidence require an expert witness not only to be qualified, but also to provide testimony that is relevant and based upon a reliable foundation. In doing so, Robinson effectively adopted the Daubert standard. The major difference in Robinson was the fact that Texas courts had never used the Frye "general acceptance" standard specifically rejected in Daubert, but rather had applied an analysis of whether the expert testimony "assisted the trier of fact" under Rule 702. The Texas Supreme Court viewed the Daubert relevancy and reliability standard as a logical progression of this Rule 702 inquiry and explicitly adopted it as the state's standard for admitting expert witnesses.⁽⁷⁾

     How should a trial court decide if an expert's testimony is relevant and reliable? The Texas Supreme Court offered several non-exclusive guidelines:

1. the extent to which the theory has been or can be tested;
2. the extent to which the technique relies upon the subjective interpretation of the expert;
3. whether the theory has been subjected to peer review and/or publication;
4. the technique's potential rate of error;
5. whether the underlying theory or technique has been generally accepted as valid by the relevant scientific community; and
6. the non-judicial uses which have been made of the theory or technique.⁽⁸⁾

     The burden of proving that the expert testimony is relevant and reliable rests upon the party seeking to enter it. Regardless of which party bears the ultimate burden of persuasion at trial, "once the party opposing the evidence objects, the proponent bears the burden of demonstrating its admissibility."⁽¹⁰⁾ (This issue will reappear in Havner, below.)

     The standard of review of the trial court's decision whether to admit expert testimony is abuse of discretion. An abuse of discretion occurs when "the trial court acted without reference to any guiding rules or principles." This standard of review is a highly deferential one, dictating that "a reviewing court cannot conclude that a trial court abused its discretion if, in the same circumstances, it would have ruled differently or if the trial court committed a mere error in judgment."⁽¹¹⁾

Early Developments
     Before Robinson, Texas courts had already begun to implement a Daubert-type approach to determining the admissibility of expert scientific testimony. The first such decision, 1992's Kelly v. State, pre-dates Daubert itself, yet reaches the virtually identical conclusion that relevance and reliability are the threshold elements for this type of testimony.⁽¹²⁾ In affirming both the trial court and Court of Appeals, the Court of Criminal Appeals held that DNA "fingerprint" evidence linking the defendant to the murder victim had been properly admitted in accordance with the framework of the Rules of Criminal Evidence. Pursuant to Rule 104(a), the trial court made a preliminary determination of admissibility, considering primarily whether the testimony would "help the trier of fact understand the evidence or determine a fact in issue" as required by Rule 702. The court also properly weighed the probative value of the evidence against its possible prejudicial effects (Rule 403).⁽¹³⁾ The bulk of the court's gatekeeping role in Kelly centered around the Rule 702 requirement of reliability, determined by a set of non-exclusive factors similar to those enumerated in Robinson.⁽¹⁴⁾ Indeed, for the gatekeeping judge, the winnowing of reliable expert testimony from speculative assertions is a much more demanding exercise than determining relevancy, since even unreliable science can "relate" in some way, no matter how tangentially, to the facts of a case. The Kelly court noted that the Frye "general acceptance" test should be merely one of the factors in a court's consideration.⁽¹⁵⁾

     Gatekeeping in Texas civil courts does not differ significantly from that in criminal courts, and it is in the context of civil litigation that the gatekeeping role has expanded most significantly. Before the Robinson decision, the Dallas Court of Appeals followed both Kelly and Daubert by adopting gatekeeping standards in North Dallas Diagnostic Center v. Newberry.⁽¹⁶⁾ Like Robinson, Newberry asserted relevance and reliability as the predicate gatekeeping evidentiary factors, and upon review found that the trial court had indeed abused its discretion by admitting evidence that was not "grounded in valid scientific knowledge."⁽¹⁷⁾ In evaluating the expert testimony, the court emphasized that the purpose of gatekeeping was to "focus solely on the validity of principles and methodology underlying the testimony, not the conclusions generated."⁽¹⁸⁾ The interface, and indeed confusion, between the concepts of methodology and conclusion presents a uniquely difficult puzzle for the gatekeeping judiciary, leading some to assert that a Daubert-type analysis requires judges to become nothing less than amateur scientists.⁽¹⁹⁾

Merrell Dow v. Havner: Burden Shifting and Uncertainty
     The concern for maintaining a clear distinction between the methodology and conclusions of an expert's testimony appeared in the first major appellate decision to follow the Robinson model, Merrell Dow v. Havner.⁽²⁰⁾ With a decision currently pending from the Texas Supreme Court, Havner is a products liability action brought against Merrell Dow for birth defects allegedly caused by ingestion of the anti-nausea drug Bendectin (the same drug litigated in Daubert) during pregnancy. After an initial trial verdict for the Havners, the appellate court found that the plaintiff's expert evidence was legally insufficient to prove causation. While the court stated it would not intrude into the jury's province of judging a witness' credibility, it asserted its duty to keep from the jury evidence "based merely upon speculation and conjecture."⁽²¹⁾ After examining the evidence, the court blasted the plaintiffs' expert testimony as simply self-serving, conclusory statements: "Reasonable probability cannot be created by the mere utterance of magic words ['reasonable medical probability'] by someone designated as an expert."⁽²²⁾ This opinion demonstrates how elusive the line between methodology and conclusion can be: how can a trial court separate the "magic words" conclusion from the reasoning behind it? And once an expert has uttered these magic words, how intrusive can the judge become before he or she encroaches on the duties of the jury?

     A rehearing en banc was held after the publication of Robinson,⁽²³⁾ and this time the court found that the Havners had met their burden of establishing the predicate relevancy and reliability of their expert testimony. The rehearing opinion admitted the plaintiffs' expert testimony because it provided at least a plausible explanation of the causation issue. The Havner opinion is inconsistent with Robinson in this respect, however, because it places the burden of proving the excludability of the evidence on the party opposing it, the defendant, instead of placing the burden of proving the admissibility on the plaintiff.⁽²⁴⁾ This inconsistency seems to have originated in Merrell Dow's motions for summary judgment and later in limine suppression of the expert testimony: within the context of these motions, Merrell Dow bore the burden of proving its motions should be granted. However, the shifting burdens of persuasion inherent in these motions do not explain the departure from the Robinson standard that the party proffering the evidence has the ultimate burden of establishing its admissibility once it has been challenged.⁽²⁵⁾

     Havner demonstrates the powerful impact procedure can have on the substantive outcome of a case in which expert testimony plays a pivotal role. It also reminds observers that judicial gatekeeping has its historical roots in procedures such as summary judgment, those predicated on the judge's prerogative to take a case from the jury under certain circumstances. How the state supreme court resolves the issue of burdens in this case--that is, how strictly it adheres to the Robinson analysis--will undoubtedly bear on the future character of Texas judicial gatekeeping. 

Gatekeeping Issues
     The following series of questions offers an introduction to the specific issues confronting gatekeeping judges in cases like Robinson and Havner. These questions correspond to the later papers in this volume which explore the complexities and contradictions involved in each area.

1. How has the role of judicial gatekeeping evolved in the United States? How has the balance of power between judge and jury shifted in reaction to trends in American history?

2. What is the influence of summary judgment on the development of judicial gatekeeping? Gatekeeping is clearly rooted in the traditional division of responsibilities between judge and jury, itself predicated on the distinction between fact and law. One way to view Daubert gatekeeping is as another means by which the judge may assert his role as arbiter of questions of law. In this light, judicial examination of expert testimony echoes the decisions a judge makes when ruling on a motion for summary judgment. 

3. What is the precise issue to be determined in a Daubert hearing? The methodology/ conclusion distinction is certainly elusive, and the mechanism for reaching it is not much clearer. Different courts interpret the meaning and scope of "methodology" in different ways: does methodology describe an expert's general approach, or does it include all the steps underlying that expert's conclusion? After evaluating testimony for reliability, a judge then turns to the testimony's relevancy; this question of "fit" has also met with varying approaches by different courts.

4. What is the scope of a gatekeeping judge's admissibility determination? How do judges meet the Daubert command to exclude unreliable evidence without attempting to decide what constitutes "good" science? 

5. Does Daubert apply to areas other than new science? Is science a special case? Or does Daubert-type gatekeeping also apply to experts in such fields as economics and psychiatry? If so, might the judge's evaluation of social science methodology differ from the judge's evaluation of the methodology of a "hard" science like biology or chemistry? 

6. How should testimony relying on epidemiological evidence and differential diagnosis be treated under Daubert? In toxic tort cases such as Daubert, Robinson and Havner, the plaintiffs all attempted to establish causation through epidemiological studies, which endeavor to measure a statistical relationship between exposure to an agent and the incidence of a condition. Are such studies necessary to prove causation? Are such studies helpful? Similarly, how should a judge evaluate a physician's testimony based on differential diagnosis? 

7. What are the procedural issues surrounding judicial gatekeeping? Practically speaking, how is the judge supposed to accomplish this task?

8. Can Daubert hearings be judically noticed or otherwise be given precedential value? What effect has Daubert had on pre-existing precedents? Should one judge's assessment of certain scientific evidence command that other judges follow the same path? Or can the issue of adequacy of expert testimony be repeatedly relitigated? How much should judges worry about conflicting court decisions on the validity of certain scientific methodologies?

9. What is the proper standard of review for a Daubert hearing? In Robinson, the Texas Supreme Court held that the standard of review for a judge's gatekeeping decision should be only for an abuse of discretion. Is this the proper standard, or should another standard, such as de novo, be applied? How thoroughly should appeals courts scrutinize the trial court's evaluation of the legitimacy of an expert's methodology? 

10. Should differing standards of stringency apply in determining admissibility depending on the type of case, evidence or witness before the court? For example, should different standards apply to civil and criminal cases?

11. What is the relationship between legal and scientific standards of proof? How much weight should judges lend to statistical measures of scientific proof? More specifically, how important are significance levels and confidence intervals in determining the reliability of a study? Nowhere is the uneasy interface between science and evidence more apparent than here.

1. 509 U.S. 579 (1993).
2. Id. at 597.
3. Id. at 589.
4. Frye v. U.S., 293 F.1013 (D.C. Cir. 1923).
5. E.I. Du Pont de Nemours and Co., Inc. v. Robinson, 923 S.W.2d 549 (1995).
6. Merrell Dow Pharmaceuticals, Inc. v. Havner, 907 S.W.2d 535 (Tex.App.--Corpus Christi 1995, writ granted).
7. 923 S.W.2d 549, 556.
8. Id. at 557.
9. Id.
10. Id.
11. Id. at 558.
12. 824 S.W.2d 568 (Tex.Crim.App. 1992).
13. Id. at 572.
14. Id. at 573. The Kelly factors are not identical to those outlined in Daubert and Robinson. They include 1) acceptance by scientific community, 2) the qualifications of the expert, 3) the existence of any literature substantiating the evidence, 4) potential rate for error, 5) availability of other experts to evaluate testimony, 6) clarity with which the testimony can be explained in court, and 7) the experience and skill of those applying the technique in question.
15. For a post-Robinson criminal case implementing similar gatekeeping procedures, see Jordan v. State, 928 S.W.2d 550 (1996), affirming the trial court's gatekeeping duty to scrutinize novel scientific theories offered as proof. Issue #3 addresses this case in more detail.
16. 900 S.W.2d 90 (Tex.App.- Dallas 1995).
17. Id. at 93. As later stated in Robinson, the standard of review for admitting expert evidence is "abuse of discretion." The Newberry court found that by admitting legally insufficient evidence (here evidence that lacked enough testing to "prove" causation of injury), the trial court abused its discretion.
18. Id. at 95.
19. 509 US at 601 (Rhenquist, S.J., dissenting). 
20. 907 S.W.2d 535. See also Suzanne B. Baker, "Gatekeeping" in Texas: The Practical Impact of Full Implementation of the Texas Rules of Civil Evidence Regarding Experts, 27 St. Mary's L.J. 237, 256 (1996).
21. Id. 907 S.W.2d at 541.
22. Id. at 542.
23. Id. at 535 (opinion on en banc rehearing August 10, 1995).
24. Id. at 550-551. "...the burden was on Merrell. This differs from the burden established in DuPont which is on the proponent to demonstrate the admissibility of its proffered evidence."
25. 923 S.W.2d at 557. "Once the party opposing the evidence objects, the proponent bears the burden of demonstrating its admissibility."