SOURCE: KF8961 .B47 AUTHOR: Berger, Margaret A. DOCTITLE: Procedural and Evidentiary Mechanisms for Dealing with Toxic Tort Litigation SECTITLE: Procedural and Evidentiary Mechanisms for Dealing with Toxic Tort Litigation DATE: 1991 SUBJECT: expert evidence United States toxic torts litigation PUBLISHER: Carnegie Commission on Science, Technology, and Government DOCTYPE: Book TITLEID: ISBN_ISSN: Text: PROCEDURAL AND EVIDENTIARY MECHANISMS FOR DEALING WITH EXPERTS IN TOXIC TORT LITIGATION: A CRITIQUE AND PROPOSAL A CONSULTANT REPORT Margaret A. Berger Brooklyn Law School OCTOBER 1991 The goal of the Carnegie Commission on Science, Technology, and Government is a nation better prepared to respond to the opportunities and hazards of scientific and technological advances. The Commission was established by Carnegie Corporation of New York in 1988 to assess, and recommend improvements in, the mechanisms by which the federal government and the states incorporate scientific and technological (S&T) knowledge into policy and decision making. The Commission's special focus is on the organization of government as it affects decision-making processes, rather than on specific policy options. The Commission is considering how government can be better organized so that policy options can be systematically formulated using the best available S&T expertise; what mechanisms for analysis need to be strengthened or created; and what technical competency is needed in government. Since policy-making in a democratic society requires balancing diverse and competing goals and values, the Commission is equally concerned that S&T- based policy options be flamed in ways that are readily intelligible and accessible both to policymakers and the people who elect them. The Commission is an independent bipartisan body with a five-year charter. In addition to eminent scientists and engineers, the Commission and its Advisory Council include former officials who have served at high levels of government, as well as leaders from the private sectors of American society. This consultant report was prepared for the Commission's Task Force on Science and Technology in Judicial and Regulatory Decision Making. The views expressed are those of the author and not necessarily those of the Carnegie Commission on Science, Technology, and Government or the Task Force. CONTENTS 1.0 EXECUTIVE SUMMARY 2.0 INTRODUCTION 2.1 The Formal Framework: In General 2.2 Organization of the Discussion 3.0 EVIDENTIARY REQUIREMENTS FOR EXPERT PROOF 3.1 The Applicable Rules 3.1.1 Rule 702 3.1.2 Rule 703 3.1.3 Rule 403 3.2 Experts Who Present Problems 3.2.1 The expert without expertise 3.2.2 The second-hand expert 3.3.3 The expert whose theory does not fit the facts 3.3.4 The expert without a valid theory 3.3.5 The expert who relies on unreliable data 4.0 THE PRETRIAL STAGE OF A TOXIC TORT ACTION 4.1 Discovery by Parties 4.1.1 The formal rules 4.1.2 The informal mechanisms 4.2 Judicial Management 4.2.1 The formal rules 4.2.2 The informal mechanisms 4.3 Motions by the Parties 4.3.1 The formal rules: summary judgment 4.3.2 The informal mechanism: the motion in limine 5.0 SCREENING EXPERT TESTIMONY 5.1 Need for Additional Information 5.1.1 Ascertaining the expert's qualifications 5.1.2 Ascertaining problems with the theory's fit or basis 5.1.3 Ascertaining problems with the expert's theory 5.2 Need for Additional Assistance 5.3 Special Pretrial Conferences Relating to Expert Testimony 5.3.1 Should the rules require a special pretrial conference relating to expert proof in certain kinds of cases? 5.3.2 Topics for discussion 5.3.3 Particular problems 6.0 CONCLUSION 7.0 ENDNOTES 8.0 ABOUT THE AUTHOR 1.0 EXECUTIVE SUMMARY This paper examines whether there are identifiable, recurring problems with expert proof that could be more capably managed so as to produce opinions and verdicts more consistent with scientific thinking. Because so many of the current complaints about the judicial system's handling of complex scientific information center on toxic tort litigation, the discussion focuses primarily on problems surrounding the proof of causation in this type of litigation. The conclusions in this paper rest on a number of assumptions. In the first place, the paper endorses the pretrial screening of expert testimony in order to make cases more amenable to disposition by summary judgment, to facilitate settlement, and to lead to more focused, speedier trials. In light of the restrictions on expert proof that exist in the evidentiary rules, and the underlying policies that allocate preliminary questions of admissibility to a judge, pretrial handling by a court is not viewed as an unwarranted intrusion on the jury. Second, after surveying the various formal and informal mechanisms currently available to provide a court with information about complex S&T issues prior to trial, the paper concludes that changes in present pretrial procedures would better enable a court to screen expert testimony effectively. It recognizes, however, that a judge so inclined could proceed in the manner suggested under the existing procedural framework. Chapter 3.0 of the paper reviews the evidentiary framework within which courts resolve issues with regard to expert proof, and examines how scientists approach the question of causation. The discussion is focused on a series of recurring problems: the expert without the requisite expertise; the second-hand or summary expert; the expert without a valid theory; the expert who relies on unreliable data. Chapter 3.0 also considers some of the issues currently facing the courts, such as deciding which discipline possesses adequate expertise, resolving conflicting evidence from different disciplines, and choosing between various standards of validity. The paper takes no position on how these difficult questions should be resolved other than suggesting that they raise policy concerns that are more appropriately the province of the judge than the scientist. Chapter 4.0 analyzes the various mechanisms that currently enable parties and the court to obtain the information needed to determine whether a proffered expert will meet the evidentiary requirements discussed in Chapter 4.1. The formal and informal rules governing discovery, case management and motions are surveyed in the context of toxic tort litigation. Chapter 5.0 proposes three modifications of the procedures discussed in Chapter 4.0 that would enable courts to handle more effectively some of the issues discussed in Chapter 3.0: (a) modifying mechanisms for making particular kinds of information more readily available to the court; (b) amalgamating the special master and court-appointed expert rules in order to better provide courts with the expert assistance needed to understand S&T issues; and (c) creating timing constraints and other guidelines for a special pretrial conference relating to expert proof. More specifically, Chapter 5.0 posits the need for a pretrial conference specifically adapted to the issues posed by expert proof about complex S&T issues. A motion for summary judgment would not be permitted prior to such a conference except in an unusual case, and with permission of the court. This conference would ordinarily be held after the completion of all expert discovery. The discovery from experts would be considerably expanded beyond what the present Federal Rules of Civil Procedure mandate. The parties would be required to furnish each other with the information now obtainable through interrogatories pursuant to Rule 26(b)(4). Thereafter, the opposing parties would have to provide their opponents and the court with (a) a written statement of their experts' anticipated testimony, (b) an accompanying written statement by each expert setting forth his or her qualifications and explaining how this particular expertise enables the expert to be of assistance with regard to the opinions expressed in the statement of testimony, and (c) a report by each expert which would have to follow a particular protocol if so ordered by the court. Once a party is in receipt of this information, it may take the expert's deposition as of right. The parties would have to confer before the pretrial conference and to provide the court with a brief written report setting forth the principal issues on which their experts disagree. At the conference, the parties must advise the court whether any challenges will be made to the qualifications of expert or the admissibility of any expert testimony. The court may explore with the parties the need for expert assistance for the court. The paper suggests under what circumstances and for what kind of issues a court is likely to need help from court-appointed to supplement or critique information furnished by the parties' experts. Because of constraints of time or a lack of familiarity with scientific principles, a judge may at times encounter difficulty in assessing the parties' proposed expert testimony to see whether it satisfies evidentiary criteria. Accordingly, in such a case the paper proposes appointing a lawyer, to function much as some special masters do now, to frame questions for the court-appointed expert. After the pretrial conference, the court must issue an order setting dates for motions in limine to exclude expert testimony, to submit nominations for court-appointed assistants, and to set up any further conferences that might be required for such matters as a hearing on who should be appointed, as well as for hearing the motions in limine. If on hearing the motion in limine the court excludes expert testimony that constitutes the plaintiffs sole proof of causation as not satisfying evidentiary requirements, a motion for summary judgment should be made and granted on the record of the in limine hearing, and materials obtained through discovery. The parties should not be permitted to supplement the summary judgment motion with further affidavits from experts who have not been subjected to the discovery process. Furthermore, a party whose expert is rejected at the pretrial hearing should not be afforded another opportunity to obtain an expert except under extraordinary circumstances, as, for instance, when the judge has injected a new legal theory into the case. Even in cases where the court does not exclude any anticipated expert testimony, the pretrial conference affords an opportunity to narrow issues, and to consider techniques for presenting expert testimony in a more focused, understandable manner at trial. 2.0 INTRODUCTION According to some observers, our legal system is seriously flawed because courts do not adequately comprehend complex scientific information. When litigation turns on the resolution of science and technology (S&T) issues, these critics claim that all too often the end product is a judicial opinion wanting in authority and consistency, or a jury verdict at odds with "good" science. Nor is the faulty quality of scientific thinking the sole grievance; complaints also center on the cost and inefficiency with which S&T concepts are handled. Probably the most caustic criticism, however, is reserved for the judicial system's continued dependence on party-selected experts. Although these experts invariably play a crucial role in S&T litigation as the consultants and witnesses who transmit critical information to all the participants -- the parties, the judge and the jurors -- party-selected experts are routinely branded as mercenaries whose opinions in the courtroom would never be tolerated in their own disciplines. Furthermore, the adversary system is accused of forcing opposing experts into extreme positions that exacerbate the differences between them, making it more difficult to achieve settlements or for the fact-finder to ascertain the truth. Over the years, thousands of pages have been written demanding procedures that will bring more neutral and well-grounded science into the courtroom. Numerous suggestions have been made for lessening the court system's reliance on partisan witnesses, improving the quality of expert testimony, and educating judges so that they will be better equipped to deal with scientific data. The chief objective of this paper is to identify those problems with expert proof that are susceptible to more effective and accurate handling through the utilization of various procedural and evidentiary mechanisms. This paper also recognizes, however, that although the judicial system's ability to manage and adjudicate S&T issues can be enhanced, it is wishful thinking to imagine that all problems will vanish because the "truth" will emerge once the correct procedural approach is adopted. As will be discussed below, the assumption that selection of the proper legal mechanism guarantees the "fight" scientific answer, application of which ensures the "correct" legal decision, fails to accord with the reality of either science or the law. Difficult questions will persist even if procedures are improved and judges are provided access to needed expertise. Many of the most troubling S&T issues involve policy choices about how to proceed in a world of imperfect knowledge. A subsidiary goal of this paper, therefore, is to identify some of the areas in which judges purport to make decisions on evidentiary grounds even as they sub silentio engage in substantive analysis. [1] Unwarranted problems do exist, however, because the fragmented and dated formal procedural framework governing expert testimony is not sufficiently responsive to the issues most disquieting to the courts. This does not mean that the hands of an innovative, activist, scientifically sophisticated judge are tied under the present system. As discussed below, informal mechanisms currently exist that can be, and are being, used by trial courts to manage S&T issues effectively. The current rules, however, fail to make these procedures readily available, do not give notice to the trial bar as to how S&T issues should be handled, and are susceptible to charges that ad hoc determinations are too often unreviewable and inherently unfair. 2.1 The Formal Framework: In General Part of the problem is that expert testimony -- the source of so many problems -- is treated in two separate and insufficiently coordinated sets of rules. Rules 702-705 of the Federal Rules of Evidence, enacted in 1975, set forth the requirements for party expert testimony at trial. Rule 26(b)(4) of the Federal Rules of Civil Procedure, adopted in 1938 and last revised in 1970, governs discovery from experts. Nowhere in either set of rules is any explicit consideration given to resolving issues of expert testimony prior to trial. The drafters of the Federal Rules of Evidence assumed that the purpose of expert discovery is to obtain information needed for the effective cross-examination of the expert at trial. [2] Rule 26(b)(4) is not designed to provide a court, or the litigants, with the information needed to determine prior to trial whether anticipated expert testimony will be able to meet evidentiary standards for admissibility. Recent developments suggest that mechanisms for screening expert testimony prior to trial need to be addressed. Effective case management is a potent weapon against the delay and cost encountered in judicial proceedings. [3] Furthermore, the procedural and evidentiary rules relating to expert proof predate the flood of toxic tort and environmental litigation that commenced in the 1970s. The rules need to be more responsive to the complex S&T issues these cases pose. Since these lawsuits have the capacity to swamp a system already sinking under an increasing civil and criminal caseload, identifying situations in which crucial issues can be disposed of short of trial would reap useful dividends.[4] Even if we accept the premise that expert proof requires pretrial handling, however, the problem remains of how to provide a court with the S&T information needed to make the requisite determinations. Although reformers have long advocated procedures enabling judges to obtain input from persons other than the parties' partisan experts in order to improve the quality of adjudication, the evidentiary and procedural rules confuse matters by setting forth two different, inadequately meshed methods for obtaining needed judicial assistance. Rule 53 of the Federal Rules of Civil Procedure regulates the use of Special Masters. Rule 706 of the Federal Rules of Evidence governs court-appointed experts who will testify at trial; it also contains provisions, however, directing the activities of these experts prior to trial. Other mechanisms for giving courts access to needed S&T information also exist that are not addressed in either the evidentiary or procedural rules. The failure of the formal rules to insist that courts examine expert testimony issues prior to trial encourages judges who are uncomfortable with scientific principles to defer decisions about the admissibility of the proffered expert testimony.[5] Under the present system, unless the court is forced to review the record by defendants' motion for summary judgment,[6] a jury verdict for the defendant may well obviate the need to understand the science about which the parties' experts are testifying.[7] This paper concludes that there are issues concerning the admissibility of expert testimony that a court ought to confront rather than to avoid, and that an earlier analysis of proffered expert testimony would lead to timelier, less costly dispositions. This paper also recognizes, however, that some judges firmly believe that issues of uncertainty ought to be left for jury resolution; these judges' refusal to peer more closely at scientific principles reflects a policy concern that may not be amenable to procedural tinkering. 2.2 Organization of the Discussion In order to illuminate the shortcomings of the present system and recommendations for change, the discussion focuses primarily on problems surrounding the proof of causation in toxic tort litigation. The courts' handling of causation, more than any other S&T issue, has spawned charges that the current system produces inconsistent judicial results irreconcilable with good science. Any suggested reform, therefore, must pass the test of handling causation issues more effectively. Chapter 3.1 of the discussion that follows considers the evidentiary requirements for expert testimony and considers some paradigmatic cases with suggestions about how they should be analyzed by the courts. Chapter 4.0 discusses the extent to which the current discovery rules and other mechanisms generate the information required to determine whether a potential expert witness will meet the evidentiary requirements. Chapter 5.0 revisits the problems considered in Chapter 3.0 to determine under what circumstances a court might screen anticipated expert testimony prior to trial, and suggests some procedural innovations that might simplify the trial court's task. Some of these proposals may strike the reader as far too radical and unwarranted. They are put forth in the hope that even if they are rejected, their specificity will at least move the discussion from a description of the problem towards the implementation of solutions. 3.0 EVIDENTIARY REQUIREMENTS FOR EXPERT PROOF In order to enhance the jury's ability to understand expert testimony, the drafters of the Federal Rules of Evidence deleted a number of common law restrictions which they thought interfered unduly with the legal system's ability to use expert testimony rationally and efficiently. The Rules, enacted in 1975, sought to make it easier for experts to testify in a non- legalistic manner uninterrupted by technical objections, and to base their opinions on the kind of data on which experts rely in their professions. Although some critics blame current problems with expert witnesses on the liberalization of the evidentiary rules,[8] recent judicial opinions demonstrate that the language of the rules can be an effective weapon against unfounded expert opinions. 3.1 The Applicable Rules 3.1.1 Rule 702 Rule 702, the basic rule on expert testimony, recognizes that a proffered witness must possess qualifications in order to testify as an expert, and that actual qualifications rather than any particular job description are what count. Rule 702 also states that the testimony must assist the trier of fact. If the proffered witness's expertise will not help to resolve a controverted issue, then the particular physician, physicist or chemist is not an appropriate expert in this litigation even though highly credentialed in his or her particular field of endeavor. Before the court can conclude that the expert's testimony will be of assistance, therefore, the court must be able to (a) identify the precise issue to which the testimony will be directed, and (b) determine that the proffered testimony relates to the disputed issue. A court that fails to undertake this two-pronged analysis risks permitting the possessor of an impressive resume to testify about matters that will not assist the trier of fact. The charisma of the expert, rather than the logic of his explanation, may then become paramount, allowing experts to function like oath-helpers of old in a manner antithetical to notions of rational proof. 3.1.2 Rule 703 Rule 703 is generally viewed as a rule that liberalized the receipt of expert testimony because it permits experts to base their opinions on hearsay and the opinions and analyses of others. Rule 703 has a restrictive side as well. The data relied on must be "of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject." This test ties in with Rule 702's helpfulness test; an expert who disregards the bases sanctioned in his particular field will not be able to assist the trier. The restrictive language in Rules 702 and 703 empowers the courts to deal with a number of recurring problems: the expert without the requisite expertise; the second-hand or summary expert; the expert whose theory does not fit the facts; the expert without a valid theory; the expert who relies on non-peer-reviewed data; the expert who fails to consider certain data. These categories are not mutually exclusive; the proffered expert's testimony may be deficient in a number of respects, and the dividing line between Rule 702 issues and Rule 703 issues is often blurred.[9] Nor do judicial opinions necessarily term the flaws in expert testimony in these precise terms; these categories are being used here for analytical purposes to illustrate the kinds of issues that arise with regard to expert proof of causation at trial. 3.1.3 Rule 403 Some courts have also relied on Rule 403 to limit expert testimony. According to the Ninth Circuit, "if the court finds a witness qualified under Rule 702, it may still exclude testimony under Rule 403."[10] The court weighed the minimum probative value of the proffered testimony against its possible prejudicial effect and cumulative nature and concluded that the district court had properly exercised its discretion. Rule 403 has also been invoked to exclude from evidence the materials that an expert has reasonably relied upon pursuant to Rule 703.[11] The Third Circuit has warned that excluding expert evidence under Rule 403 "at the pretrial stage is an extreme measure."[12] 3.2 Experts Who Present Problems 3.2.1 The expert without expertise "Whether a witness is qualified as an expert can only be determined by comparing the area in which the witness has [expertise] with the subject matter of the witness's testimony."[13] An expert who does not understand the subject matter about which he or she seeks to testify is incapable of assisting the trier of fact. An expert who seeks to testify about causation must therefore have adequate expertise to explain whatever type of scientific information is being relied upon to prove causation.[14] In order to be allowed to conclude that defendant's product caused plaintiff's injury, the expert must base this opinion on something other than a temporal coincidence of exposure and harm. A demonstration of cause and effect requires more than the association of two events in time. An example of the expert without expertise is the physician who claims that plaintiff's birth defects were caused by defendant's product but does not know such matters as the composition of defendant's product, the developmental stages of a fetus or the other substances to which the mother was exposed.[15] Such an expert does not know enough about any type of scientific proof recognized in proving causation. Under these circumstances, the physician is no more qualified to testify about causation than an engineer who has no familiarity with the design of the system about which he wishes to opine.[16] As is discussed below, the expert without expertise should be dealt with prior to trial; in order to do so, the court must know the expected substance of the expert's testimony and his qualifications. 3.2.2 The second-hand expert A more difficult issue arises when the expert, though familiar with general principles relevant to the theory he is prepared to state, has no special expertise with regard to the controverted issue.[17] For instance, unlike the physician in the previous example, the second-hand expert does know about general principles of toxicology and fetal development and is well- acquainted with the mother's medical history.[18] However, he has never done any research or had any experience with the kind of substance implicated in this case. His knowledge about the attributes of defendant's product stems completely from the literature search in which he engaged after being retained as an expert.[19] On the crucial issue of causation, he would be acting solely as a summarize of other experts' work. The problem here is not that the expert is relying on hearsay,[20] but rather that his expertise may not be sufficiently specialized to be helpful, given the issue before the court. Error may also occur when the expert testifies that his opinion is consistent with that of other experts whose identities were never disclosed during discovery.[21] The danger with a second-hand expert is that the trier of fact will be relying on his credibility far more than on his expertise. Tenure at a leading university and ability as a public speaker do not ensure the witness's value in resolving a controverted issue even when the proffered expert is generally knowledgeable about a relevant field. Judges are accorded a great deal of discretion by the appellate courts in deciding when an expert is sufficiently qualified. On a complex issue such as causation, however, the trial court should exercise care to ensure that an appropriate expert has been selected. The trial court should take into account the exact issue about which the expert seeks to testify,[22] the nature of the expert's field of expertise, and the availability of other experts. An epidemiologist, for instance, whose work routinely consists of drawing statistical conclusions from others' studies may be an appropriate witness even though he is recapitulating another expert's data. A professor of toxicology, on the other hand, who has never worked with substances related to those produced by the defendant should perhaps not be qualified to testify about causation though he might be permitted to explain general principles to the jury. Some of the mechanisms discussed in Section 5.0, below, may help courts to become more discriminating about the appropriate degree of expertise required. Obviously, a court needs adequate background information at a meaningful time in order to exercise its discretion appropriately. 3.2.3 The expert whose theory does not fit the facts Reported cases demonstrate a number of instances in which the expert's theory is not consistent with the facts of the case. For instance, in an action against the manufacturer of the Copper 7 intrauterine device, plaintiff's expert testified that there is a three- to five-fold increased rate of pelvic inflammatory disease in women wearing IUDs.[23] Cross- examination revealed that the expert was relying on a study in which only one participant was wearing the defendant's device; that patient did not develop pelvic inflammatory disease.[24] Other reported cases similarly reflect a lack of correspondence between the expert's theory and the facts in evidence.[25] When there is no fit between the facts in issue and the expert's theory or the research on which he is relying the expert testimony is irrelevant and of no assistance to the trier of fact as required by Rule 702. The discussion below will consider mechanisms for making this determination short of trial. 3.2.4 The expert without a valid theory Particularly troublesome for the judicial system are those toxic tort cases in which the plaintiff suffers from a disease or defect which is not attributable solely to a particular substance, but which plaintiff claims was caused by exposure to a substance for which defendant is legally responsible.[26] If the proffered expert witness's theory of causation is untenable, it will not satisfy the helpfulness test of Rule 702. But how can a judge, particularly one untrained in scientific matters, gauge the validity of a theory offered by a seemingly qualified expert who insists that there is a causal link between the plaintiff's condition and exposure to defendant's product?[27] Ascribing causation is fraught with difficulty because of the current uncertainty about how the human body responds to toxic substances. Modern science and the law, moreover, approach the problem in different philosophical terms. Brennan has pointed out that the law is still apt to subscribe to a Newtonian view of causation that demands a mechanical cause.[28] Science treats causation issues as working hypotheses or paradigms that are defended by the scientific establishment until anomalous results lead to adjustment of the paradigm theory.[29] Experts who offer a mechanical explanation may be treated favorably by the judicial system even though the present lack of knowledge about the mechanism causing a particular disease would restrain most scientists from finding a causal connection in the absence of probabilistic evidence that provides a strong plausible association between the alleged exposure and injury. Imperfect knowledge and epistemological differences result in two central problems for the courts in relation to an expert witness's theory of causation: (1) does the witness subscribe to a relevant discipline? (2) what is the appropriate standard of validity? A detailed consideration of the mechanisms available to the courts to handle these issues is deferred until Section 5.0 of this paper. A brief description of the questions that arise in determining the validity of an expert's theory is included at this point, however, as the background for considering what kinds of information judges need to resolve these issues. a. Relevant discipline or methodology. The first issue the judge may have to resolve is whether the expert belongs to a discipline that is capable of ascertaining whether a causal connection exists between the product and the injury.[30] Physicians, clinical ecologists, and toxicologists who rely on animal studies have all been challenged as inappropriate experts. The Frye test[31] is particularly unhelpful in resolving these cases because it posits admissibility on a scientific principle having "gained general acceptance in the particular field in which it belongs." It sheds no light on how a court can identify the relevant "particular field." Courts will also undoubtedly be asked with increasing frequency to determine whether meta- analysis is a reliable scientific technique for determining causation, although at this writing only one reported federal opinion[32] deals with this issue. The testimony of a physician who states that he is convinced to a reasonable degree of medical certainty[33] that defendant's product caused plaintiff's cancer is the kind of opinion that is traditionally admitted in medical malpractice cases when a plaintiff's claim turns on establishing the basis for an injury. Except in the case of certain signature diseases, and certain known carcinogens, however, some would maintain that at this time only probabilistic evidence should be considered in linking any particular case of cancer with exposure to a product.[34] Because no validated theory as yet furnishes an explanation about cancer formation, the physician's testimony may amount to nothing more than describing two events, exposure and disease, that are sequentially, but not causally, connected.[35] Whether a physician's testimony that rests on neither epidemiological or toxicological findings will suffice to prove causation was explored recently. In Christophersen v. Allied-Signal Corp.,[36] a wrongful death action, the plaintiffs claimed that the deceased's death from colon cancer had been caused by his exposure to toxic nickel and/or cadmium fumes at his place of work. The only evidence supporting plaintiffs' claim of causation was supplied in an affidavit by a Dr. Miller.[37] The trial court concluded that Dr. Miller's opinion would be "inadmissible because it was not based on epidemiological, animal or in vitro studies showing a 'statistically significant link between colon cancer and exposure to nickel and/or cadmium.'"[38] Dr. Miller acknowledged that these were the three main types of evidence used to establish causation, and conceded that he knew of no studies linking nickel and/or cadmium exposure to small-cell cancer of the colon. He argued, however, that nickel and cadmium had been linked to other forms of cancer and expressed the opinion that "'[b]ased on what's known about the biochemical nature of small-cell' carcinoma, 'the same sorts of chemicals and exposures that are associated with small-cell carcinoma of the lung are likely to be associated with small-cell carcinoma elsewhere in the body.'"[39] Defendants' experts found no scientific merit to Dr. Miller's theory that the size of the cell in lung cancer and colon cancer indicate a common pathogenesis.[40] Because the trial court found that Dr. Miller's testimony would not be admissible at trial it granted summary judgment for defendant. A panel of the Court of Appeals reversed.[41] It found that the district court had erred in striking Dr. Miller's expert testimony because he relied on "methodologies that physicians traditionally rely on in diagnosing the cause of a particular patient's cancer."[42] According to the panel, the district judge should have permitted the jury to evaluate the weight of the expert testimony. It refused to require causation "to be supported by epidemiological studies establishing a cause-effect relationship."[43] The Christophersen case was reviewed en banc by the thirteen judges then sitting in the Fifth Circuit.[44] The majority first expressed some concern, as had the district court, as to whether Dr. Miller had met the minimum qualifications to testify under Rule 702 since he was neither an expert in oncology or pathology. "The questions . . . do not stop if the expert has an M.D. degree. That alone is not enough to qualify him to give an opinion on every conceivable medical question. This is because the inquiry must be into actual qualification -- sufficient to assist the trier of fact."[45] The en banc court did not, however, base its opinion on Dr. Miller's lack of "specialized experience and knowledge" because the trial court had not excluded his testimony on this ground.[46] The majority refused to adopt either the viewpoint of the plaintiff -- that a treating physician is competent to express an opinion about causation -- or the viewpoint of the defendants -- that "the pathogenesis of a particular kind of cancer requires clearly positive results from one or more of [three] types of testing."[47] Rather, the en banc majority found that Dr. Miller's opinion could not be admitted because the methodology or reasoning that Dr. Miller used to connect the facts to his conclusion was not generally accepted within the relevant scientific community.[48] It found no support for Dr. Miller's presumption that the association between nickel and cadmium with a certain type of cell in lung cancer means that these substances are associated with a similar type of cell in the colon. "All Dr. Miller had was a scientific hunch, which as far as the record shows, no one else shares. This was enough to support further investigation but was inadequate to support a judgment in favor of Christophersen."[49] A case like Christophersen does not lend itself well to a motion for summary judgment determined solely on the basis of affidavits from experts. Section 3.2 below discusses how a pretrial conference followed by a motion in limine would be utilized in such a case. The physician as expert poses additional problems when the physician subscribes to an alternative form of medicine such as clinical ecology, which is frowned upon by the mainstream medical community.[50] A challenge to the expert forces the court to decide whether testimony from such an expert has any value. Clinical ecologists attribute a vast array of symptoms to "chemical hypersensitivity syndrome" caused by agents in the environment and workplace.[51] Testimony by clinical ecologists has been admitted with some frequency,[52] infuriating critics who argue that the clinical ecologists' theories are based on inadequate scientific inquiry[53] and amount to nothing more than "junk science."[54] Although toxicologists produce animal studies that are relied upon in risk assessment, their work has come under attack when used to prove causation in a court of law.[55] Two major problems exist with regard to these studies. In order for them to be relevant it must be assumed, first, that if a substance is found to be toxic in some species of animal it must also be toxic in humans and, second, that one can extrapolate from the dosage level used in the study to the level at which the plaintiff was actually exposed. Although science is willing to make these assumptions in relying on animal studies, provided the animal studies have been conducted in an appropriate manner,[56] a number of courts reject this type of evidence on the ground that the underlying premises cannot be confirmed.[57] Other courts at this time, however, view expert testimony based on animal studies as adequate proof of causation to raise an issue for the jury, even in the face of negative epidemiological evidence. The extensive Bendectin litigation furnishes numerous examples of judges who have refused to grant summary judgment even though plaintiff's proof of causation rests solely on animal studies,[58] and will not set aside verdicts for plaintiffs supported by this kind of evidence.[59] b. Standard of validity. In deciding whether the expert is relying on a valid theory, courts also must make a policy determination as to whether they will accept the scientific standard of proof or a lesser standard. Scientists customarily employ a 95% confidence level in testing a hypothesis -- that is, they do not consider the data in favor of the hypothesis statistically significant unless the observed association has less than a 5% probability of resulting solely from chance.[60] The scientist's use of a 95% confidence level in hypothesis testing is consistent with scientific methodology. Scientists choose to accept a relatively high risk that they will accept a false null hypothesis and therefore decline to adopt the tested theory (a Type II error), in order to minimize the risk that the null hypothesis will be wrongly rejected, and, therefore, that the tested theory will be mistakenly accepted (a Type I error). This approach makes perfect sense in the context of scientific research. The not-proven verdict which is the consequence of a failure to satisfy the 95% confidence level means only that more research is in order -- it is not a statement of an established "truth."[61] False acceptance of the null hypothesis in perhaps 50% of the cases[62] makes little difference because the alternative hypothesis will continue to be tested.[63] A not-proven verdict in court, however, has very different consequences in our present system of civil litigation. We have virtually no mechanisms at the moment for deferring decisions until more proof is available,[64] or correcting decisions erroneously made.[65] The plaintiff or the defendant wins or loses at the moment the case is ripe for decision. The plaintiff, who has the burden of persuasion, bears the risk, and the loss, if the case is not proven.[66] When a court requires epidemiological studies to be statistically significant at the 0.05 level,[67] critics charge that plaintiffs are unfairly being required to meet a burden that is much higher than the "preponderance of the evidence" burden of persuasion in a civil case.[68] Furthermore, if the court insists on a 95% confidence level without looking at the underlying data, the plaintiff is forced to bear the risk of the wide confidence intervals that result when there are few subjects in a study, even though the data suggest some association of the product and plaintiff's disease. Although the plaintiff ordinarily bears the burden of producing evidence, it is perhaps unfair to make plaintiff suffer the consequence of insufficient probabilistic data in toxic tort litigation when the plaintiff, particularly an individual plaintiff, has no control over the amount of data that is available, and no means of compelling anyone, including the defendant, to undertake additional research.[69] Furthermore, even when the sample is large, setting the confidence level at 95% means that cases supporting the alternative hypothesis will be ignored even though there might be quite a few hinting at a positive association between exposure and disease were a lower level, such as 90%, used.[70] A bigger problem is lack of recognition of the risk of type II error and the need for some balance between the errors. 3.2.5 The expert who relies on unreliable data As stated above, Rule 703 provides that an expert may rely on facts or data not admissible in evidence as long as the facts or data are "of a type reasonably relied upon by experts in the particular field in forming opinions or inferences upon the subject." Some courts have relied on this rule, rather than Rule 702, to conclude that expert opinions are inadmissible when they rely on a theory with which the court disagrees.[71] As the Third Circuit has explained, "it can be difficult to determine whether the putative problem with scientific evidence lies in the underlying data itself or the method by which the data is analyzed."[72] For instance, courts that consider extrapolation from animal species to humans meaningless have not admitted the results of an animal study to prove that a disease was caused by exposure to a particular chemical.[73] Rule 703 is also used to reject opinions based on data that the court believes is not sufficiently reliable because of the manner in which the results were obtained or tested.[74] The kinds of issues that might be raised in toxic tort litigation include: Must the study on which the expert relies have been peer-reviewed?"[75] Must an opinion as to causation take into account the claimant's medical history?[76] Was an appropriate animal species selected for the animal study?[77] Did the epidemiological study consider appropriate factors such as dosage?[78] Were the studies replicated?[79] Was the statistical analysis correctly carried out?[80] While a court might not necessarily reject an expert's testimony if his or her data were deficient in any one of these respects, the court might find that cumulative defects are inconsistent with reliable data. Even if the court has no quarrel with the expert's underlying theory, the proffered opinion may be excluded if the court concludes that the data on which the expert seeks to rely have been obtained by methods inconsistent with the demands of good science?[81] Meta-analysis, a statistical method for combining the results from separate published studies on a common scientific issue, is an approach to analyzing statistical data that may be subject to challenge under either Rule 702 or Rule 703. The objective of meta-analysis is to see if all available data looked at collectively produce a different result from that obtained when small studies are looked at individually. Some courts may be concerned with the underlying notion that various studies can be combined in a meaningful manner and cite Rule 702, while other courts may focus more on how the different data are combined, a Rule 703 approach. Although meta-analysis has been used in the scientific world for some times, it is just beginning to surface in the courts as a method, inter alia, for determining whether a particular substance is a carcinogen or teratogen.[82] Combining studies that measure different parameters is controversial. The technique is susceptible to bias in ways that have been summarized as follows: Two different studies rarely measure precisely the same parameter. Differences in study methods and measures, subject populations, time frames, risk factors, analytic techniques, etc., all conspire to make the "common" parameter different in each study. . . . One must then ask what one is testing or estimating when one combines a heterogeneous group of studies. The answer is some ill-specified average of the estimated parameters from the selected studies. Whatever the average is, it depends very much on the sampling scheme by which studies are selected for combination. This is a serious problem for meta-analysis, because so often there is no "scientific" sampling of studies. Obviously, a biased selection of studies is compounded when one limits the selection to all published studies, because of the well-known bias that favors publication of significant over non-significant findings. This is the "file-drawer" problem for meta-analysis: how to access the unknown number of negative findings buried in researchers' file drawers to achieve a balanced cross section of findings. Finally there is an almost definite lack of independence across studies. Successive studies by the same researcher on a subject are often subject to the same systematic biases and error; a recognized authority in a field may determine an entire research program; and in extreme cases, scientific competitors may have hidden agendas that affect what gets studied and reported.[83] We turn now to an examination of the means available to a court to obtain the information needed to resolve the difficult issues sampled in Section 3.1 of this paper. Of course, this does not mean that all of these issues are resolvable prior to trial; Section 3.2 of this paper is not concerned with the extent to which a court should screen expert testimony prior to trial, but rather with the more pragmatic question of how it could be in a position to do so. 4.0 THE PRETRIAL STAGE OF A TOXIC TORT ACTION For purposes of this discussion, I will somewhat artificially separate three components of the pretrial process as instrumentalities for providing a trial judge with the information needed to screen expert testimony prior to trial: (a) discovery by the parties, (b) judicial management, and (c) motion practice by the parties. Actually, of course, these three aspects of the pretrial process are considerably intertwined. Each has a formal side governed by specific Federal Rules of Civil Procedure or Evidence, as well as an informal posture, regulated not by rule but by unwritten practice, local rules, and the dictates of individual judges. In considering the formal rules governing the chronology of a federal lawsuit, it may be helpful to keep in mind two of the premises upon which the Federal Rules of Civil Procedure were drafted,[84] as they explain the choices reflected in the Rules. First, the Rules were designed to be trans- substantive, i.e., a single set of rules was designed for all cases regardless of the substantive issues at stake. Second, the Rules were drafted to provide "flexibility, open-ended texture, simplicity and broad grants of judicial discretion."[85] Consequently, unlike the Codes of Procedure that they replaced, the Rules for the most part lack detail and definition. Amendments proposed in the last few years, however, seem to be getting more complex. Originally, the Rules placed little or no emphasis on judicial management. This has changed. When we look at the current version of the Rules and pending amendments we can see that the need for case management has been recognized explicitly, particularly in Rule 16, which will be discussed below. On their face, though, the Rules still appear to be trans- substantive and remarkably undetailed.[86] After examining some of the informal mechanisms that have blossomed in the interstices of the rules, Section 4.3 will consider whether the formal model suffices for handling expert testimony in toxic tort litigation, or whether more definite, detailed rules might be more helpful in resolving the difficult issues raised in Section 4.1. 4.1 Discovery by Parties 4.1.1 The formal rules The Federal Rules of Civil Procedure governing discovery rest on a number of assumptions. In the first place, although the rules have embraced the concept of judicial management, the court is not directed to play a role in discovery unless the parties so request, except for setting guidelines early in the litigation. The original rules governing discovery assumed that judicial orders would be required for some modes of disclosure, but that discovery would otherwise proceed extrajudicially, except in the aberrant case in which judicial intervention was needed to avert a breakdown in orderly procedures. From 1980 onwards, however, the Advisory Committee began to respond to widespread complaints about discovery abuse. In 1980, Rule 26(f) was added; this authorized a discovery conference at the option of the parties or the court, followed by an order establishing a plan and schedule for discovery.[87] In 1983, Rule 16, the pretrial conference rule,[88] was amended to require courts to enter a scheduling order within 120 days after filing of the complaint, "that limits the time . . . to complete discovery." Since Rule 26(f) gives courts the option of combining the pretrial conference with the discovery conference, in most instances the two conferences will be combined. Neither Rule 16 nor Rule 26 mentions expert witness testimony as a topic to be considered at the scheduling conference. Once the discovery conference is held, the rules do not require the judge assigned to the case to keep informed of the material that is produced or created through the various discovery techniques,[89] although Rule 16(b)(4) does give the court the option of scheduling additional conferences before trial. Since even more discovery is now designed to proceed without judicial orders than originally,[90] the court may have little or no contact with the case until discovery is complete, unless the parties request assistance. This is especially true in those districts in which discovery is routinely handled by magistrates. In those districts, even when there are discovery disputes requiring judicial intervention, rulings will be made by the magistrate rather than the judge. A second basic assumption underlying the present formal governing rules is that discovery is not self-executing.[91] There is no obligation on a party to make information available to an opponent or the court absent a request.[92] The rules govern the circumstances under which a request may be made, and the showing required to obligate the opponent to respond. In addition, they describe the information that may be obtained. Rule 26(b)(4) of the Federal Rules of Civil Procedure treats discovery from experts as a sub-category of discovery of trial preparation materials subject to special limitations not applicable to other witnesses. Subdivision (b)(4) contains provisions dealing with two kinds of experts: (1) the expert who is expected to testify at trial, and (2) the "expert who has been retained or specially employed by another party in anticipation of litigation or preparation for trial" but who is not expected to testify.[93] In order to discover "facts known or opinions held" by the retained expert who will not be a witness, the party seeking the information must make a "showing of exceptional circumstances under which it is impracticable for the party seeking discovery to obtain facts or opinions on the same subject by other means."[94] Discovery from the nontestifying expert is accordingly extremely limited, although a few courts have authorized depositions even when the expert is not the only person capable of expressing an opinion on the subject.[95] With regard to experts expected to testify, the rule currently provides that a party may serve an interrogatory requiring the other parties to identify all such experts and "to state the subject matter on which the expert is expected to testify, and to state the substance of the facts and opinions to which the expert is expected to testify and a summary of the grounds for each opinion."[96] Upon motion, the court may also order "other discovery" from the expert expected to testify.[97] In practice this means that the court may order the expert to be deposed, and/or order experts to exchange reports. Since the enactment of the Federal Rules of Evidence, some courts have recognized that a party will not be able to prepare effectively for expert testimony at trial without a deposition,[98] and a number of local rules provide that parties may depose such experts without having to obtain a court order in the individual case.[99] Were Rule 26 to be amended to reflect a duty by parties to disclose certain information absent a request, automatic disclosure might be required of the information pertaining to potential expert testimony now obtainable through an interrogatory. Furthermore, the rule might authorize testifying experts to be deposed without requiring the examining party to obtain a court order. Although these innovations would reduce paper work, they would affect little change in the present scope of discovery. Even under the present rules, judges are often willing to allow parties to depose their opponents' experts; tactical concerns, however, often prevent parties from exercising this option. They fear that asking questions at a deposition may reveal too much about the inquirer's theory of the case.[100] It is questionable, however, whether the present formula requiring the party "to state the substance of the facts and opinions to which the expert is expected to testify and a summary of the grounds for each opinion" is enough to provide the opponent or the court with sufficient information to determine whether the expert's testimony will satisfy the requirements imposed by the Rules of Evidence discussed in Section 4.1 or the applicable substantive law. 4.1.2 The informal mechanisms In practice, however, parties to complex litigation frequently engage in far more extensive discovery than the formal rules mandate. Parties voluntarily provide their opponents with their experts' reports, stipulate to depose all experts, and agree to maintain joint document depositories.[101] Consequently in some cases, even without taking advantage of special management procedures pursuant to Rule 16, the trial judge may have ready access to considerably more information bearing on anticipated expert testimony than would be available through formal discovery. Whether a court will take the initiative in reviewing the fruits of either informal or formal discovery is currently left to the court's discretion. 4.2 Judicial Management 4.2.1 The formal rules Rule 16 of the Federal Rules of Civil Procedure was considerably amended and expanded in 1983 in order "to encourage pretrial management that meets the needs of modern litigation."[102] The rule mandates an initial conference culminating in a scheduling order discussed above, and subsequent conferences in the discretion of the court at which the judge may exercise control over the progress of the case. In accord with the drafters' preference for trans-substantive rules and flexibility, the rule does not specifically refer to expert witnesses or S&T issues. The Committee Note explained: No particular techniques have been described; the Committee felt that flexibility and experience are the keys to effective management of complex cases. Extensive guidance is offered in such documents as the Manual for Complex Litigation. A number of provisions empower the judge so-inclined to single out cases posing S&T problems for special treatment. Subdivision (c) lists among the topics that the participants at a pretrial conference may consider, and with regard to which they make take action, the following: (5) the identification of witnesses and documents. . . (10) the need for adopting special procedures for managing potentially difficult or protracted actions that may involve complex issues, multiple parties, difficult legal questions, or unusual proof problems; and (11) such other matters as may aid in the disposition of the action. Clauses (10) and (11) provide the court with the authority to experiment with the many informal techniques discussed below that innovative judges have used in handling complex S&T issues. If the case has not settled or been terminated through a dispositive motion, the court may hold a final pretrial conference. Subdivision (d) provides that at a final pretrial conference,[103] the participants "shall formulate a plan for trial, including a program for facilitating the admission of evidence." Certainly, by the time of a final pretrial conference the court will know that it is dealing with a complex case raising difficult S&T issues that may not settle before trial. Again, however, the rule does not explicitly direct attention to the problems posed by highly specialized and controverted expert proof, nor does it suggest that courts consider some of the threshold questions of expert testimony that could be resolved prior to trial. Rule 16(c)(6) also recognizes that a court may need assistance. One of the subjects that may be considered and acted upon at the conference is "the advisability of referring matters to a magistrate or master." Although the rule does not mention the possibility of utilizing a court-appointed expert pursuant to Rule 706 of the Federal Rules of Evidence, the appropriateness of utilizing Rule 706 is certainly a matter covered by the broad language of clauses (10) and (11). a. Rule 53 - special masters. The first sentence of Rule 53(b) provides that a reference to a special master "shall be the exception and not the rule." The exceptionality requirement is buttressed by the second sentence's dictate that "[i]n actions tried by a jury, a reference shall be made only when the issues are complicated." In its order of reference the court may specify and limit the master's powers and direct the contents of the master's report.[104] Subject to these specifications and limitations, the master has the power to hold hearings, to require the production of evidence, to rule on the admissibility of evidence and to question witnesses. Pursuant to Rule 45, witnesses may be subpoenaed to appear before the master. Compensation is determined by the court and is generally charged to the parties in proportions fixed by the court. A number of commentators have expressed misgivings about the increasing reliance on non-judicial private officers who may have a profound effect on the shape of the litigation.[105] Of particular concern is the rule's failure to prohibit ex parte contact between judge and master, especially because the rule is silent about a special master's qualifications. Critics of the present system worry about an appointee's potential conflicts of interest and institutional biases. Obviously, the possible impact of the master on the outcome of the litigation will vary tremendously, depending on the particular task the master is asked to perform. Although traditionally masters were used to handle accountings and to calculate complicated damages, they are nowadays being used more in connection with pre-trial matters, settlement, and to assist in formulating post-trial remedial relief.[106] Masters have at times been used because the appointing court needed help with technical and scientific data.[107] In these cases, the masters performed tasks such as supervising discovery, hearing non-dispositive motions, or making recommendations with respect to dispositive motions.[108] The advantage of a master over a magistrate is that a court can select someone who has far more specialized knowledge than the usual magistrate is likely to have. Rule 53(f) was added in 1983 to provide that a reference to a magistrate may be made subject to the limitations of Rule 53, in which case the proceedings would be subject to the report procedures of Rule 53(e). Some observers believe that masters are being appointed to handle pretrial matters far more frequently than the reported cases indicate. How masters might be used with respect to S&T issues without inappropriately taking on roles that should more properly be played by public judicial functionaries is discussed in Section 5.2. b. Rule 706 - court-appointed experts. Rule 706 of the Federal Rules of Evidence authorizes a judge to appoint expert witnesses. Subdivision (a) of the rule provides that before a court- appointed expert can act there must be (1) a hearing on an order to show cause why the expert should not be appointed, (2) consent by the expert and (3) notification of the expert as to his or her duties either in writing or at a conference. The judge may nominate and select the expert or make a selection from names submitted by the parties. Once appointed, the expert must advise the parties of his or her findings, if any, and is subject to deposition by any party. If the expert is called to testify by the court or by any party, the court has the option of informing the jury of the expert's court-appointed status, and the expert is subject to cross-examination by any party, including a party calling the expert as a witness. Pursuant to subdivision (b), the court has discretion to tax the expert's compensation against the parties in such proportions as it deems proper, taking into account the nature of the case, the reasons why a court-appointed expert is needed, the status of the parties, and the ultimate disposition of the controversy. Court-appointed experts have long been touted as the best available corrective to partisan expert testimony. Nevertheless, studies have consistently found that judges seldom appoint experts to assist them.[109] A number of reasons have been advanced for this reluctance: the judge's fear of appointing a biased witness[110]; judicial lack of knowledge about where to find a competent expert[111]; concern lest the jury be unduly swayed by testimony offered by a witness under the court's aegis[112]; misgivings about imposing the cost of additional experts upon the parties[113]; and the belief that an expert selected by the court is at odds with the traditions of the adversary system. [114] Furthermore, the judge may be apprehensive about not understanding the case sufficiently to be able to select the right expert from the appropriate field and to give him or her suitable marching orders.[115] These problems and possible solutions are discussed further in Chapter 5.0. The extent to which court-appointed experts have been used in connection with scientific causation issues is discussed elsewhere[116] and will not be addressed anew. Certainly, the possibility of providing the court with a nonpartisan basis of expertise is extremely attractive. A number of recent proposals have suggested modifications to Rule 706 that would make the rule more responsive to the problems it is designed to alleviate.[117] 4.2.2 The informal mechanisms In the exercise of their case management function, judges have developed numerous techniques for educating themselves sufficiently to cope with complex S&T issues. These judicial practices seek to achieve a number of different objectives: obtaining more information from the experts than the parties were able to obtain; clarifying questions for the court that persist after it has reviewed the information obtained by the parties and the court; reducing the conflicts between opposing experts; and supplementing the information available from party-selected experts. A court may obtain additional information from the parties' own experts by devices such as requiting the experts to follow certain protocols,[118] to provide glossaries,[119] to reveal the underlying data as well as statistical models and bases on which the experts are relying.[120] A court may utilize a standard pretrial order to inform parties about their experts' obligation to furnish this information. After reviewing the information obtained from the parties' experts, the court may feel the need for further explanation. Some judges ask the parties to have their experts make a presentation to the court on salient issues; others may more informally sit down with the experts for an exchange of views. Judges may also request the expects to answer specific questions or undertake additional calculations to fill in gaps the court considers crucial in proving the case.[121] An oft-repeated complaint about relying on party-selected experts is that the system pushes experts into taking extreme positions. Another aspect of effective case management is to reduce the conflicts between opposing experts by requiring them to identify data and conclusions as to which they are in accord. Some judges order experts to cooperate in creating a joint report setting forth their areas of agreement and disagreement. Other judges require experts' reports to be treated like marked pleadings-- counsel are asked to identify those statements about which there is no dispute, those about which the parties disagree, and those to which the opposing experts have not responded. Courts have also attempted on their own to obtain additional information that they feel is essential to enable them to deal with the S&T issues in the case. Rather than appointing a special master or a court-appointed expert, courts have at times utilized technical advisers, relying on inherent authority as the source enabling them to make such an appointment.[122] In Reilly v. United States,[123] for instance, the district court sought the assistance of an economist in calculating damages for lost earning capacity. After obtaining the approval of the Administrative Office of the United States Courts and the Chief Judge of the First Circuit,[124] the court appointed a professor to assist him.[125] Courts have also at times sought to obtain needed information from administrative agencies or other institutions that have special expertise with regard to the controverted S&T issues. Rule 803(8) of the Federal Rules of Evidence ensures that the reports of public bodies will be admissible.[126] It has been suggested that formal mechanisms are needed to ensure that such information will be forthcoming.[127] All of these informal techniques have enabled many federal judges to cope very effectively with complex litigation. Chapter 5.0 of this paper will discuss to what extent, if any, some of these mechanisms should be put on a more formal basis in order to assist in the resolution of the problems discussed in Section 3.2. 4.3 Motions by the Parties 4.3.1 The formal rules: summary judgment Rule 56 of the Federal Rules of Civil Procedure authorizes a court to grant summary judgment to a moving party when the admissible evidence tendered by the party with the burden of proof on an issue is insufficient to permit a rational fact finder to find for the party on that issue under the appropriate standard.[128] Recent cases in which defendants were awarded summary judgment in toxic tort cases suggest that the granting, and perhaps the incidence, of motions for summary judgment has increased in this type of litigation.[129] The courts may be reacting to the increased emphasis on case management,[130] the Supreme Court's favorable attitude to summary judgment reflected in its 1986 trilogy of cases,[131] a greater willingness to take issues away from a jury,[132] the ever-increasing criminal and civil case load that makes it ever more difficult to find trial time, as well a growing recognition of the costs and consequences of toxic tort litigation.[133] Although the summary judgment motion is usually made after formal discovery is complete, the motion itself operates as a discovery device. This is because the parties present experts' affidavits and other materials in support and opposition of the motion that supplement the information obtained through discovery.[134] Two recent Third Circuit opinions are sufficiently detailed to reveal the abundance and variety of the additional supporting material that may be submitted in connection with a motion for summary judgment in a toxic tort[135] case. In DeLuca v. Merrell Dow Pharmaceuticals, Inc.,[136] one of the last of over one thousand law suits claiming recovery for birth defects allegedly caused by defendant's drug Bendectin, defendant submitted the following: epidemiological studies; published and unpublished judicial opinions in other Bendectin cases; an affidavit of an expert; and copies of various FDA documents relating to Bendectin.[137] Plaintiffs countered with an affidavit by their expert which supplemented his deposition testimony. Attached to the affidavit was an article and chapter by a professor at a medical school explaining the scientific methodology for evaluating epidemiological studies on which plaintiffs' expert relied.[138] Additional materials were also submitted on the summary judgment motion in In re Paoli Railroad Yard PCB Litigation,[139] a toxic tort case in which thirty-eight persons who either worked or lived adjacent to the Paoli railyard claimed that they contracted a variety of illnesses as the result of exposure to PCBs concentrated at the railyard.[140] The Third Circuit reversed the grant of summary judgment to defendants in both DeLuca and In re Paoli. Chapter 5.0 of this paper returns to both of these cases in considering under what circumstances summary judgment is useful to resolve issues with regard to causation. 4.3.2 The informal mechanism: the motion in limine Nowhere in the Federal Rules of Civil Procedure or Evidence is the motion in limine mentioned, and yet the cases reveal considerable reliance on this device as a means of challenging the admissibility of evidence, including the testimony of expert witnesses. A motion in limine is an extremely flexible device that can be tailored by the judge to meet the demands of the particular case. It may be made orally or in writing. In addition to hearing counsel on the motion, the court may be willing to have the challenged expert testify in chambers, thereby enabling the court to consider the admissibility of the testimony in circumstances equivalent to the adversarial process that takes place at trial. Of course, a court will be loath to spend valuable time on previewing expert testimony that will be repeated at trial. The court will conduct an extensive hearing only if it believes that the process will expose flaws that would authorize the exclusion of some or all of the testimony, resulting in a settlement, disposition by summary judgment, or a considerably shorter trial. Information derived from some of the informal case management techniques discussed above may better enable the court to make such a determination. For instance, if the experts' reports have been marked and their expected testimony has been prefiled, the judge will be able to focus on the crucial disputed issues at the hearing on the motion in limine. 5.0 SCREENING EXPERT TESTIMONY Placing the paradigmatic problems posed by expert proof in toxic tort litigation[141] in juxtaposition with the available procedural mechanisms[142] highlights the need for procedural changes that would enable the courts to handle expert witnesses more effectively. This section will propose three innovations as a means of curtailing some of the difficulties that arise in litigating a toxic tort suit: (1) modifying mechanisms for making particular kinds of information more readily available to the court; (2) amalgamating the special master and court- appointed expert rules in order to better provide courts with the expert assistance needed to understand S&T issues; and (3) creating timing constraints and other guidelines for a special pretrial conference relating to expert proof. Although the discussion suggests amending the formal rules in order to accomplish these changes, virtually all of the proposals could be handled instead via local rules or judicial orders. Whether to achieve change by formal rule or otherwise depends in large measure on views about how much judicial discretion and ad hoc adjudication the system should tolerate. 5.1 Need for Additional Information 5.1.1 Ascertaining the expert's qualifications Among the problems discussed in Section I, supra, were a number that turned on the expert's having insufficient expertise to assist the trier of fact. A court should not permit an expert to testify who does not understand the theory on which he or she purports to rely,[143] or who does not have sufficient specialized knowledge to help the trier handle the controverted issues.[144] Excluding testimony on these grounds neither implicates the judge-jury relationship nor raises the troublesome issue of which party should bear the risk in a world of imperfect knowledge. The chief difficulty is informational -- how can the trial judge become adequately informed to make these determinations without inordinate expense to the litigants or a waste of the court's time? The Federal Rules of Civil Procedure should set forth a date -- such as thirty days before the time set for the completion of fact discovery[145] -- by which time the parties must furnish each other with the information that is now obtainable pursuant to Rule 26(b)(4) through interrogatories.[146] Within sixty days after this information is provided, unless the court orders otherwise, the opposing party must provide its opponent and the court with: (1) a written statement of their experts' anticipated testimony and (2) an accompanying written statement setting forth each expert's qualifications (including past stints as an expert retained to testify).[147] These qualifications, whether derived from "experience, training or education,"[148] must be related to the opinions in the statement of testimony; rather than reproducing his c.v., the expert must explain how this particular expertise enables him to be of assistance to the court and jury with regard to the opinions expressed in the statement of testimony.[149] In addition, the party must provide (3) a report by each expert that, if the court so requires, must follow a particular protocol as discussed below. Once a party is in receipt of this information, it may within sixty days, unless the court orders additional time, take the expert's deposition as of right.[150] How a party would then proceed after the close of all expert discovery to exclude an expert's testimony on the ground that the expert does not possess the necessary qualifications is considered in Part C, infra. 5.1.2 Ascertaining problems with the theory's fit or basis As was discussed above in Chapter 3.0, an expert who relies on a theory that does not fit the facts of the case is not providing testimony that is relevant to a resolution of the controverted issues.[151] Expert testimony based on data that are erroneously derived or markedly incomplete or deficient in some manner, that is objectionable to other experts in the relevant field, does not satisfy Rule 703.[152] Under our present system, these deficiencies in the expert's proof are not likely to be revealed by the interrogatories authorized by Rule 26(b)(4). These shortcomings may, of course, be explored either at a deposition or on cross-examination at trial,[153] but in the absence of any prior information about the studies and data on which the expert is relying, the examination of the expert may be costly and inefficient. A statement of anticipated testimony will be of only marginal help because Rule 703 does not require the expert on direct to set forth the bases for his opinion. Requiring the expert additionally to furnish a report will be helpful only if the report contains enough about the theory's fit and basis to provide leads that can effectively be followed up at the oral examination of the expert. In the field of toxic torts, thought should be given to developing specific protocols for each type of expert who regularly testifies in these types of cases, e.g., the physician (with special questions for clinical ecologists), the epidemiologist, the toxicologist who relies on animal studies or other kinds of studies, the statistician or other expert who is testifying about a meta-analysis. [154] Such protocols would have to be developed jointly by scientists in the relevant areas and lawyers knowledgeable about toxic torts. The scientists would first have to educate the lawyers about their particular disciplines[155]; the lawyers would then have to explain what kind of information the courts have required in assessing this type of expertise[156]; and finally the scientists would translate the lawyers' requests into language that would make sense to an expert in the scientific field.[157] The word protocol obviously has very different connotations for scientists and lawyers. Scientists consider protocols as guidelines for future work, while lawyers view them as mechanisms for recapturing the past. Accordingly, scientists may be hesitant about developing what they think of as protocols when there is a divergence of opinion in their discipline about how a particular study should be carried out. Certainly, disagreements in the scientific community may ultimately have important legal ramifications beating on the adequacy of the data to support a theory or the validity of the theory itself.[158] Initially, however, counsel and the court need factual information about how a study was set up and what assumptions were made before they can evaluate its legal significance. For example, epidemiologists may differ about how different dosage levels impact on the proper selection of a population group for study. Before a court can begin to assess the different scientists' contentions, however, it needs to know what assumptions were utilized in setting up the studies relied upon by the expert in the case before it.[159] Protocols would have to be detailed enough to provide information that would assist a court in determining whether the results of the study are legally relevant. If protocols were developed, a court could then order experts to abide by these protocols in their reports. As examples of what generally should be included, the protocol might require the expert to list all studies relevant to the opinion, and the extent to which his opinion is based on any particular study. With regard to any study on which he relies, the expert would have to state (a) the population being studied; (b) the method of data collection; (c) whether the study was peer reviewed; and (d) the results of the study (with confidence intervals). With regard to any study with which the expert disagrees, he would have to explain why he rejects the results of the study or finds the result inapposite in this case. For maximum effectiveness, the questions addressed by the protocol should be tailored as narrowly as possible to the particular discipline of the expert. Although requiring this much information from an expert prior to trial may seem inconsistent with Rule 705 of the Federal Rules of Evidence, which permits the expert to testify without revealing the basis for his opinion,[160] it must be remembered that this information is sought to instruct the court rather than to inform the jury; access to this information may enable the court to dispose of problems of fit and basis prior to trial, or at least help the court to do so in conjunction with other devices discussed below. A similar conflict, as to the appropriate level of detail, presently exists with regard to Rule 705 and the summary judgment rule. Since Rule 56(e), the summary judgment rule, requires that affidavits "set forth such facts as would be admissible in evidence," it has been argued that an expert's affidavit on a summary judgment motion may state an opinion without specifying a factual basis, as such an opinion would satisfy Rule 705. Nevertheless, many courts have reacted to the increasing use of affidavits by experts supporting and opposing motions for summary judgment by requiring a factual basis for the opinion in the affidavits,[161] and by further insisting on an adequate opportunity for additional disclosure before summary judgment may be ordered.[162] These courts recognize that sufficient information prior to trial may obviate the need for trial. Requiring detailed factual bases to be provided in the experts' reports may pare summary judgment motions down to a more manageable size. See Section C, infra. 5.1.3 Ascertaining problems with the expert's theory Protocols will also provide courts with a point of departure for educating themselves about the most difficult issues they face -- determining whether as a matter of law an expert witness subscribes to a valid theory or employs an appropriate standard of validity. In order to answer these questions, a court may need considerable additional help beyond that which is available from the parties through discovery.[163] The next section explores whether a reworking of the court-appointed expert and special master rules might provide a better alternative to party-selected experts than presently exists. 5.2 Need for Additional Assistance As noted above, although court-appointed experts have frequently and enthusiastically been proposed as alternatives to party-selected experts, all studies to date indicate that Rule 706 of the Federal Rules of Evidence is rarely utilized. On the other hand, special masters are being appointed by the courts pursuant to Rule 53 of the Federal Rules of Civil Procedure in cases posing complex S&T issues. This section draws on experience with both rules to suggest a different model of non-partisan assistance for the court. Although Rule 706 explicitly indicates that it will apply at the pretrial stage -- the court-appointed expert may be deposed -- the major concern of the rule is the trial, as is only natural with a rule that is located in the Rules of Evidence. On the other hand, although Rule 53 appears in the "trials" section of the Federal Rules of Civil Procedure, many courts assume that the rule is a source of authority for pretrial references.[164] On its face the rule has nothing to do with helping a court cope with experts. What is needed is one rule that recognizes that effective case management may require providing a court with varied kinds of assistance in dealing with difficult issues of expert proof. The proposed special expert assistance rule should be a subdivision of a rule governing special pretrial conferences relating to expert proof. See Section C, infra. Accordingly, it would be located in the Federal Rules of Civil Procedure in recognition of the court's need for assistance in determining prior to trial whether proffered expert proof is admissible. The rule needs to be sensitive to the concerns that may inhibit courts from using court-appointed experts more readily. Appointing an expert may impose an unwarranted expense on the parties if adequate information is being provided through discovery, especially in view of current settlement rates. The particular field in which the court needs assistance may not be ascertainable until the disputes between the party-selected experts have been narrowed by techniques discussed above. A judge who is having difficulties with scientific or statistical theories may require help in understanding the concepts on which the party-experts' opinions rest rather than additional opinions about the ultimate issues in the case. Furthermore, even when it is quite apparent that the case will turn on disputed S&T issues, the particular nature of that dispute may not be immediately apparent. Until it is, a court may not know to whom to turn. The ultimate issue may be one of policy -- what justice requires -- a question a court is far better equipped to answer than a scientist. For instance, a scientist who assumes that all hypotheses must meet the 95% confidence level test may not be of much assistance to a court trying to decide as a matter of policy what rate of error is consistent with the objectives of burdens of proof in civil litigation.[165] The discussion above suggests that in most instances[166] a court should defer consideration of the need to appoint an expert until party discovery is complete.[167] It also suggests that the expert will be most helpful to the court if asked to report on particular, narrowly focused issues. [168] Science and the law do not necessarily see eye-to-eye with regard to what is relevant; if the court hopes to get a useful answer it will first have to frame an appropriate question based on the governing legal standards. If these assumptions are valid, the court should consider the need for a court-appointed expert at a pretrial conference held after party expert discovery is concluded. (See Section 5.3). The court would thereafter have to conclude within a reasonable specified period (30 days?) that an expert should be appointed, and proceed to select the expert, also within a reasonable specified period (60 days?). Detailed time limits are essential to prevent the case from stalling for an unacceptably long period. If the court is particularly busy, however, or not well-versed in scientific matters, this time frame may not give the court adequate time to familiarize itself with the sophisticated scientific data produced during discovery,[169] to formulate the issues on which it needs help, or to select the appropriate expert either on its own, or from a list of nominations proposed by the parties. Sometimes a magistrate may be able to assist the court in categorizing the controverted S&T issues. In other instances, however, the court may need the help of a knowledgeable legal assistant who would act much as some special masters presently do. Although under current practice, issue formulation by special masters is usually a by-product of expertise obtained by the masters in assisting with discovery, a knowledgeable legal assistant could handle this task without necessarily having supervised discovery.[170] The appointment of the legal assistant would also be governed by rule. The court could ask the parties for nominations,[171] and in any event, would hold a hearing on the nomination. The lawyer could use informal mechanisms to narrow the issues with respect to expert proof, such as conferences, requesting joint reports or briefings from the party experts, etc., subject to time limits set by the court and under the supervision of the court. Unlike the procedure under present Rule 53, all ex parte contacts between the court and the court-appointed lawyer or the court-appointed lawyer and the parties or their experts would be prohibited. Furthermore, the legal assistant would not be permitted to engage in other roles such as mediation.[172] Whether special masters should be used in aid of settlements is a wholly separate issue. This proposal relates solely to using a lawyer as a link between the court and scientific issues by translating legal standards into questions that experts must answer.[173] The lawyer-expert could also make recommendations about whom to appoint as the S&T expert, could comment on nominations submitted by the parties, and would assist the court in framing the particular issues on which the expert will be asked to report.[174] The parties would have an opportunity to comment on all of the legal assistant's recommendations before they are put into effect. The S&T expert would write a report, and be subject to deposition, as under present Rule 706. The court would have the option of requiring the legal assistant to write a report, which would be made available to the parties, detailing in what respects the court-appointed expert's opinion differs from those of the party-selected experts.[175] Both court-appointed experts would attend the special final pretrial conference relating to expert testimony. As under present law, the S&T expert could testify at trial, and the legal assistant's report may be made available to the jury. 5.3 Special Pretrial Conferences Relating to Expert Testimony This section proposes the adoption of a formal rule that would govern pretrial conferences in certain kinds of cases in which S&T issues predominate. Factors would have to be specified that would trigger applicability of the rule. A possible formulation that includes a number of factors that would lead to inclusion of a toxic tort case might be: if expert testimony on a scientific or technical issue determinative of liability will be offered (1) and resolution of the issue may have a substantial impact on numerous persons besides the parties before the court; or (2) and at least one of the parties has identified more than x number of experts expected to testify at trial; or (3) and at least one of the parties has moved (or has notified the court of its intention to move) for summary judgment on the ground that there is no genuine issue as to a particular material fact because the opponent's anticipated expert testimony bearing on that fact must be excluded. More specificity could be added by factors such as a minimum number of cases pending in the federal courts raising the same issue, or requiting a certain amount in controversy. Although the ensuing discussion builds on the proposals put forth in Sections 5.1 and 5.2, above, a special pretrial conference relating to expert proof in certain kinds of cases would be useful even if none of these proposals are adopted. The conference would enable the court to explore with the parties (1) the possibility that issues relating to expert proof could be disposed of prior to trial; (2) methods for sharpening the disputed issues relating to expert evidence, through informal mechanisms or court assistants (or more discovery if the present scope of expert discovery is not expanded), with a view to pretrial disposition or a more focused, less time-consuming trial; and (3) procedures ensuring that the parties will have an adequate opportunity to be heard with regard. to rulings on expert testimony that could have a dispositive effect. 5.3.1 Should the rules require a special pretrial conference relating to expert proof in certain kinds of cases? Amending Rule 16 (or adding a Rule 16.1) to require special treatment for a particular category of issues is compatible with the assumptions underlying the Federal Rules of Civil Procedure, and furthers the recognized objective of promoting effective case management. Singling out certain kinds of expert testimony for more elaborate handling does not transgress the preference for trans-substantive rules; the rule should be drafted broadly so that it will impact on a wide range of cases besides product liability suits. Although this proposed rule would be considerably more detailed than Rule 16(c), a trend to greater specificity is evident in recent amendments to the rules. The crucial question is whether a case has been made for insisting on a more elaborate and detailed procedure than is currently mandated. The preceding discussion has made a number of suggestions for improving the court's ability to screen expert testimony prior to trial, all of which could be implemented more effectively and fairly if they were formally recognized as case management techniques to be explored at a pretrial conference. Recent opinions suggest appellate reluctance to affirm a judgment below that granted summary judgment because the plaintiff's expert testimony will be inadmissible, unless the reviewing court is convinced that the parties had an adequate opportunity to challenge the anticipated expert testimony at an on-the-record adversarial hearing.[176] When a court finds that proffered testimony is incapable of meeting the tests of Rules 702 or 703, it is using judicial notice to conclude that there is no evidential hypothesis justifying receipt of the evidence.[177] Although the terms of Rule 201 of the Federal Rules of Evidence do not apply when a court is ruling on the admissibility of evidence,[178] providing the parties with an opportunity to be heard is consistent with the concept of fair procedure underlying the rule.[179] 5.3.2 Topics for discussion The pretrial conference should be held within a designated number of days (perhaps sixty unless otherwise specified by the court) after the completion of all expert discovery. Furthermore, a pretrial conference should be required before a defendant may file a motion for summary judgment on the ground that plaintiff's expert testimony is inadmissible,[180] and that accordingly no genuine issue exists as to an essential fact on which plaintiff bears the burden of proof. The parties should be under an obligation to have conferred with each other prior to the conference about the following topics. At the conference, they must either provide the court with the requested information, or explain why such a report is premature: 1. The identity of all experts who will be called at trial. 2. Whether additional expert discovery of any kind is needed. 3. Whether any challenges will be made to the admissibility of the testimony of any of the experts designated as witnesses. 4. If a party intends to challenge the qualifications of a particular expert or the admissibility of particular expert testimony, the objecting party must advise the court briefly about the particular defects which justify exclusion. The parties must be prepared to discuss with the court the usefulness of any informal mechanisms to sharpen the disputes about the expert testimony, and the desirability of providing legal and/or scientific assistance to the court. 5. When no request for pretrial screening will be made, the parties must briefly specify the main issues about which their experts disagree. They must be prepared to discuss with the court the most efficacious way to handle these disputes in anticipated expert testimony at trial, in terms of sequencing of witnesses, making expert reports available to jurors, etc. The parties must also consider with the court the extent to which appointing an expert to report about any of the disputed issues would improve the quality of the trial. The court shall issue an order no more than sixty? days after the conference setting dates that limit the time to complete formal or informal discovery, to make motions in limine, to submit nominations for court- appointed assistants, and to set a date for any further conference that might be required for such matters as a hearing on who should be appointed, as well as for a hearing on a motion in limine. 5.3.3 Particular problems a. Qualifications. If the expert's qualifications are challenged, and the court agrees that the specified defects suffice to raise an issue that might be disposed of prior to trial, the court may schedule a motion in limine on the hearing of which the court may voir dire the expert. Motions such as these raise the troublesome issue of whether a party whose expert is rejected at a pretrial hearing should be afforded another opportunity to select an expert. The answer must be in the negative except in the most extraordinary circumstances if pretrial screening is to further efficiency and fairness.[181] Allowing a losing party to look for another expert who would then be subject to discovery would obviously cause delay, expense to the opponent and the additional outlay of judicial time. Since the Federal Rules of Evidence favor the admissibility of expert testimony, and indeed all relevant evidence, a court will exclude an expert on account of insufficient qualifications only in egregious instances that should have been apparent to the party selecting the expert. Extending the pretrial phase of the case to give the party another chance at finding an appropriate expert is not only out of keeping with efficiency concerns but is also inconsistent with current summary judgment practice. Courts have been willing to grant summary judgment after concluding that the plaintiff's anticipated expert testimony is deficient without affording the plaintiff an additional opportunity to rectify the defect. b. Theory. If the challenge is on the ground that the expert is relying on a particular theory that presents a case of first impression, the parties should provide the court with citations to cases in which other courts have considered this issue. The parties should also be prepared to discuss what kind of additional information would assist the court in deciding on the admissibility of this testimony, and be prepared with a list of persons whom the court could recruit as court-appointed experts. For instance, the circuit may not yet have ruled on the admissibility of meta-analysis, and a party may claim that testimony based on meta-analyses should never be admitted in a court of law.[182] In this type of case, the parties will have to brief the legal test that the particular circuit uses in determining the admissibility of a novel scientific technique. Although the court certainly has to have some understanding of the scientific issues in order to make a decision about the appropriateness of a particular theory, its conclusion may well depend on the resolution of underlying policy issues about when compensation is appropriate, which a scientist is not equipped to make.[183] Rather than appointing a scientist as an expert, therefore, the court may prefer to be briefed by the parties' experts, perhaps with the assistance of a lawyer assistant who could help in the framing of appropriate questions. Sometimes the theory in question will be much more case specific. In the Christophersen case,[184] for instance, a physician supported his opinion about causation by claiming that the nature of the cancer cells indicated that the substance in question had caused the cancer because that substance had been implicated in causing similar cancers in other organs. Pursuant to the procedure being suggested, the defendant in that case would have had to advise the court at a pretrial conference that it was going to challenge this testimony before it could bring a motion for summary judgment.[185] The parties would have to discuss with the court the alleged defect in the expert's testimony, and the usefulness of appointing a science expert, as well as utilizing a legal assistant. In deciding whether it wishes the assistance of a scientist, the court would have to take into account not only the likelihood that this procedure might lead to the pretrial exclusion of plaintiff's expert, but also the desirability of a court- appointed expert if the plaintiff's expert is allowed to testify at trial. If the defendant makes a sufficient showing, the court would schedule a hearing in limine before which it might wish to use some informal mechanisms, such as ordering a joint report from the plaintiff's and defendant's expert asking them to specify their areas of agreement and disagreement, asking for a briefing from the party experts, etc. At the pretrial conference, the parties could also request that experts not designated as trial witnesses be permitted to testify at the in limine hearing; if the court agrees, any such experts should be required to submit a written report prior to the hearing. The parties could also request permission to submit affidavits to supplement Brandeis briefs submitted on the motion in limine. At the hearing on the motion, the court could permit court-appointed experts to testify (the parties would of course have been permitted to depose them), and might permit the parties' experts to respond to the additional materials submitted on the motion. For the reasons stated above, however, a court might be wary about appointing a scientific expert, especially as selecting the field from which to appoint the expert might well have a decisive effect on the opinion that will be rendered. If the court concludes that the proffered expert's testimony will not satisfy the requirements of Rule 702, and this testimony constitutes the plaintiff's sole proof of causation, a motion for summary judgment should be made and granted on the record of the in limine hearing. The parties should not be permitted to supplement the summary judgment motion with further affidavits by experts; the hearing on the motion in limine offers the parties a far better opportunity for an effective adversarial confrontation than does the paper record compiled on a summary judgment motion.[186] This would change the current practice which does not explicitly prohibit summary judgment on the ground that the plaintiff will be unable to sustain his burden of expert proof at trial even when the moving party bases its motion on affidavits by experts who have been subjected neither to discovery or cross-examination. [187] c. Reliability. Perhaps the most laborious challenges to expert testimony for a court to handle are those mounted pursuant to Rule 703. In order for the court to determine whether the data relied on by the expert is sufficiently reliable, the court may have to be quite knowledgeable about numerous scientific concepts and statistical models, and be able to understand details about matters such as how a particular kind of study must be set up and analyzed. The appointed legal assistant may be of particular assistance in this kind of case in spelling out to an appointed scientist the questions that are relevant in terms of legal standards that need to be applied. Because the Federal Rules of Evidence favor the admissibility of relevant evidence, opinions challenged on the basis of the underlying unreliability of data will probably pass a threshold test for admissibility unless the unreliability is so great as to make the opinion worthless.[188] In most cases the chief problems for the court will surface at trial -- dealing with battling experts and deciding whether the plaintiff's evidence is sufficient to sustain a verdict if the plaintiff wins. In this type of case, a court-appointed expert who would help the jurors understand what the parties' experts are disputing might be helpful. A court might, for instance, ask an expert to analyze the results of the parties' contentions using different statistical techniques without expressing an opinion about the appropriate choice of model.[189] Case management techniques may narrow the issues so that the jury is able to focus more effectively on the disagreements among the experts. 6.0 CONCLUSION Expert testimony on causation in a toxic tort case poses difficult challenges for a trial judge which many federal judges have handled with great distinction. None of the suggestions incorporated in this paper require new rules; all of these proposals could be implemented through a judge's inherent powers under the Rules of Civil Procedure and Evidence. By making these procedures the norm, however, all judges would be forced to consider the interrelationship of law and science more explicitly and at an earlier stage of a litigation. This paper suggests that some of the issues surrounding expert testimony can and should be disposed of prior to trial with effective case management. Rule 104(a) of the Federal Rules of Evidence recognizes that the admissibility of evidence shall be determined by the court. A court should exclude as irrelevant the opinions of experts whose qualifications do not enable them to aid the jury, or whose theories are invalid or have no hearing on the facts, or whose work is so unreliable as to make their conclusions worthless. Allowing the judge to determine these questions initially rather than leaving them for jury resolution furthers the policies underlying rule 104(a). Among the reasons advanced for allocating preliminary questions to a judge are: (1) simplifying issues for a jury; (2) improving the consistency or predictability of results; (3) sharpening issues on appeal; and (4) increasing the speed of trial.[190] Our experience with toxic tort litigation to date indicates the pertinence of these goals to the problems facing the courts. In order to handle these threshold issues effectively a judge often needs more information than the Federal Rules of Civil Procedure currently provide. Rule 26(b)(4), which governs discovery from experts, was written before the desirability of case management was recognized in Rule 16, before the Federal Rules of Evidence liberalizing expert proof were enacted, and before the explosion of product liability litigation occurred. In light of these developments, the discovery rules should be amended to impose greater obligations on the parties to engage in the kinds of discovery that would be most helpful and economical in providing information that a court could use to screen expert testimony prior to trial. Parties should be required to reveal automatically information that must now be requested, and, in addition, should be obliged to exchange experts' reports and to produce experts for deposition, a practice that is already being followed in many complex cases. Furthermore, courts should order parties to engage in other practices tailored to the particular case that would narrow the contentions of opposing experts, such as following particular protocols, or filing joint expert reports. A special pretrial conference devoted to issues of expert testimony should be held after the discovery stage is over with a view to resolving issues concerning the admissibility of expert proof prior to trial. In some instances the court may need assistance that the parties are not able to furnish. A judge untrained in scientific and statistical principles may need help in determining what it is that he or she must comprehend in order to evaluate the proffered testimony to determine whether it meets the threshold of admissibility required by the Federal Rules of Evidence. Assistance in framing the appropriate questions may be needed before the court can examine the proposed answers. When the parties cannot brief the judge adequately because of insufficient expertise, or mismatched resources, or excessive partisanship, the court may wish to turn to outside assistance. This paper suggests an amalgam of the special master and court- appointed expert rules in recognition of the fact that most causation issues that confront a court are a blend of law and science. The lawyer who understands the legal standards and is knowledgeable about science may be of assistance to a court in framing specific questions for a scientist or statistician. The discussion above also attempts to discern under what circumstances court-appointed assistants are likely to be helpful, either at the pretrial stage or at trial. Although improved procedures may make it easier to resolve some of the problems posed in the discussion above, reforms of this kind can play only a limited role. In making decisions about causation, judges often must decide how the legal system should balance harm, risk and uncertainty, a difficult calculation that they and other governmental officials face in other contexts as well.[191] Scientific knowledge must inform the choice, but abdication to the scientist, though justified by a desire for good science and a lesser burden on the courts, is incompatible with the judge's responsibility to decide the law. Ultimately, in the absence of legislative directives, the courts decide how risk is to be allocated in the guise of rulings on the admissibility and sufficiency of expert testimony. Better information may make it easier for the judge to approach that task, but the difficulty of doing justice remains a constant. 7.0 ENDNOTES [1] See Note, Causation in Toxic Torts: Burdens of Proof, Standards of Persuasion and Statistical Evidence, 96 Yale L.J. 376 (1986) (discusses how evidentiary rulings disguise doctrinal change and proposes approach compatible with tort law goals). For discussions of how causation issues in toxic tort cases should be handled as a matter of substantive law, see Rosenberg, The Causal Connection in Mass Exposure Cases: A "Public Law" Vision of the Tort System, 97 Harv. L. Rev. 851 (1984); Delgado, Beyond Sindell: Relaxation of Cause-in-Fact Rules for Indeterminate Plaintiffs, 70 Calif. L. Rev. 881 (1982). This paper takes no position on how courts ought to resolve substantive issues that in a diversity case often hinge on complex choices of the applicable law. [2] Rule 705 of the Federal Rules of Evidence authorizes an expert to state an opinion without first disclosing the basis for that opinion unless the court requires otherwise. The cross-examiner is given the right to bring out the underlying facts or data on cross-examination. The Advisory Committee explained: "If the objection is made that leaving it to the cross- examiner to bring out the supporting data is essentially unfair, the answer is that he is under no compulsion to bring out any facts or data except those unfavorable to the opinion. The answer assumes that the cross- examiner has the advance knowledge which is essential for effective cross- examination. . . . Rule 26(b)(4) of the Rules of Civil Procedure, as revised, provides for substantial discovery in this area, obviating in large measure the obstacles which have been raised in some instances to discovery of findings, underlying data, and even the identity of the experts." [3] See text at notes 102-103, infra, for discussion of changes in the 1983 amendments to the Federal Rules of Civil Procedure, and see discussion of summary judgment in text infra at notes 123-128. [4] But see Brennan, Helping Courts with Toxic Torts: Some Proposals Regarding Alternative Methods for Presenting and Assessing Scientific Evidence in Common Law Courts, 51 U. of Pitt. L. Rev, 1, 9 and note 40 (1989) [hereinafter Brennan, Helping Courts] (terming as "disturbing" greater activism by federal judiciary in scrutinizing experts' qualifications and deposition testimony more closely). [5] This point was recently raised by Judge Higginbotham in his opinion dissenting from the majority's refusal to rehear en banc an appeal in which a panel of the Fifth Circuit had granted judgment n.o.v. to the defendant in an action claiming that defendant's drug Bendectin had caused birth defects. Brock v. Merrell Dow Pharmaceuticals, Inc., 884 F.2d 167 (5th Cir. 1989). Judge Higginbotham stated that he did "not dissent because I think the panel reached the wrong conclusion -- about that I am not sure." Id. at 169. Rather, Judge Higginbotham dissented because the panel shied away from addressing the crucial issue -- the admissibility of the evidence in the first place rather than its sufficiency after it is admitted: "Yet, while skepticism permeates its opinion, the panel does not seem to engage the question at this juncture. Rather, the panel chooses to accept the admissibility of the testimony and to quarrel with its effect." Id. [6] See text at notes 129-135, infra. [7] In the case of a verdict for plaintiffs, the trial judge ultimately may be forced to confront the viability of the expert testimony by a motion for judgment n.o.v. (see, e.g., cases cited in note 71, infra), but given the minuscule percentage of cases that ever reach trial, the judge at the pretrial stage may believe that an inordinate investment of time in fathoming complex scientific issues is unwarranted. See discussion infra. [8] It should be noted, however, that there were complaints about expert witnesses long before there were Federal Rules of Evidence. See, e.g., Keegan v. Minneapolis & St. Louis R.R., 76 Minn. 90, 95, 78 N.W. 965, 966 (1899) ("there is hardly anything, not palpably absurd on its face, that cannot now be proved by some so-called `expert.'"). [9] See DeLuca v. Merrell Dow Pharmaceuticals, Inc., 911 F.2d 941, 952, 955, note 14 (3d Cir. 1990) (although the court states that Rule 703 "seeks to delimit the acceptable bases for expert testimony" whereas Rule 702 is concerned with the reliability of the expert's methodology, the court agrees that "Rules 702 and 703 intersect. If a study's method of data collection is faulty, it may well be that no expert would rely upon the data generated as a basis for drawing any inference about the studied subject."). See also Imwinkelreid, The "Bases" of Expert Testimony: The Syllogistic Structure of Scientific Testimony, 67 N.C.L. Rev. 1 (1988) (proposal for drawing sharp line between Rules 702 and 703). [10] Rogers v. Raymark, Industries, Inc., 922 F.2d 1426, 1430-31 (9th Cir. 1991) (in suit against asbestos manufacturers for wrongful death, expert's work experience was only tangentially related to that of decedent's and might have confused jury into equating life as an insulator with life as a welder). See also Christophersen v. Allied Signal Corp., 939 F.2d 1106, 1112 (5th Cir. 1991) (en banc) ("Assuming the witness's testimony is relevant and the witness is qualified, and assuming his testimony has the appropriate factual basis and his methodology is well rounded, the only remaining inquiry is the balancing authorized in Rule 403" dictum as majority found that it did not have to address Rule 403). [11] Gong v. Hirsch, 913 F.2d 1269, 1272-73 (7th Cir. 1990) (in malpractice action, district court did not abuse its discretion in refusing to admit into evidence a letter from a former treating physician about cause of plaintiff's disease that plaintiff's medical expert had relied upon; court was concerned with trustworthiness since it did not know whether writer had personal knowledge and letter was written to obtain unemployment benefits). [12] Hines v. Consolidated Rail Corp., 926 F.2d 262, 274 (3d Cir. 1991). [13] Carroll v. Otis Elevator Co., 896 F.2d 210, 212 (7th Cir. 1990) (expert on human behavior and perception properly qualified to testify about the attractiveness of an escalator stop button to children). Compare Edmonds v. Illinois Central Gulf RR Co., 910 F.2d 1284 (5th Cir. 1990) (clinical psychologist not qualified to express opinion on whether stress worsened plaintiff's coronary artery disease); Stull v. Fuqua Industries, Inc., 906 F.2d 1271, 1275 (8th Cir. 1990) (expert in mechanical engineering was not qualified to testify that plaintiff's leg would have broken had lawn mower accident occurred in manner claimed by plaintiff; expert lacked expertise in human anatomy). [14] For a comprehensive discussion of the various kinds of scientific evidence that are relied upon in the scientific community to prove causation see Brennan, Causal Chains and Statistical Links: The Role of Scientific Uncertainty in Hazardous Substance Litigation, 73 Cornell L. Rev. 469 (1988) [hereinafter Brennan, Causal Chains]. I am not at this point considering the more difficult issue of determining how to deal with an expert who is relying on a theory that is in dispute (see discussion of clinical ecology at notes 50-54, infra). Nor am I discussing how the courts should handle the interrelationship between data derived from different theoretical approaches (see discussion of animal studies and epidemiological evidence below). I am concerned here with the far simpler problem of the expert who does not know enough about the particular theory he is trying to put forth. [15] See, e.g., Rubinstein v. Marsh, 1987 WL 3068 (E.D.N.Y. 1987) (not officially reported). (Plaintiffs alleged that their twins' severe birth defects were caused by the ingestion of defendant's birth control pills at a time when the mother did not yet know that she was pregnant. On the issue of causation, the plaintiffs called a pediatrician who had never diagnosed a drug-related birth defect in his own practice, had no experience in obstetrics or gynecology, and had never prescribed oral contraceptives or observed their effect. He did not know when the hands and fingers of an embryo differentiate (one of the twins had a severe hand malformation), did not know what other drugs the mother had taken, and had an extremely tenuous grasp of the relevant medical literature. For instance, he did not know the hormonal composition of the drug that one of the population groups studied had been using. The judge found the testimony of the second expert even less compelling. While the court in this judge- tried case entered judgment for the defendants, noting "the complete failure by the plaintiff to prove any causal connection between the birth defects and the conduct of the defendants," the judge also concluded that the plaintiffs' experts "were qualified by virtue of the fact that each was a doctor." ). Compare Will v. Richardson-Merrell, Inc., 647 F.Supp. 544, 548-49 (S.D.Ga. 1986) (in Bendectin case, court refused to admit testimony on causation by plastic surgeon with "relatively little, if any, scientific knowledge regarding Bendectin, its components, or its effects.") Cf. Apostol v. United States, 838 F.2d 595 (1st Cir. 1988) (air traffic controller unqualified); Wilkinson v. Rosenthal & Co., 712 F.Supp. 474, 478 (E.D.Pa. 1989) (professor of finance unqualified as to certain opinions). [16] See, e.g., Hoban v. Grumman Corp., 717 F.Supp. 1129 (E.D. Va. 1989), aff'd without opinion, 907 F.2d 1138 (4th Cir. 1990). Brown v. Parker-Hannifin Corp., 919 F.2d 308 (5th Cir. 1990) (professor of electrical engineering never saw defendant's product that injured plaintiff; did not know how long the product was in use or under what conditions it was stored; expert also never tested his theory of over- pressurization; expert must present sufficient evidence to enable a jury to find that his proof is the most likely cause of product's failure; not enough to present theory consistent with the facts). [17] A recent case, although it does not deal with S&T issues, is helpful in illustrating the parameters of this problem. In Wilkinson v. Rosenthal & Co., 712 F.Supp. 474, 478 (E.D.Pa. 1989), plaintiff brought suit against defendant for churning his commodities account. Plaintiff's proffered expert, an associate professor of finance at the Wharton School had little if any work experience or academic history in the commodities field. He taught a course in which he touched on some basic principles relating to commodities investing, but did not discuss churning. The court ruled that although the professor could testify about basic principles bearing on commodities, he would not be permitted to testify about what constitutes churning. [18] Cf. Loudermill v. Dow Chemical, 863 F.2d 566, 568-570 (8th Cir. 1988) (plaintiff claimed that defendant's cirrhosis of the liver was a direct result of worker's exposure to DBCP at defendant's plant; plaintiff's expert on causation had extensive academic and practical knowledge in the field of toxicology, but admitted on voir dire that he was not specifically familiar with the relationship between halogenated hydrocarbons, such as DBCP, and liver toxicity; appellate court stated that "Dr. Lowry's credentials are not unassailable in the specific area of the relationship between halogenated hydrocarbons and liver injury" but found no abuse in discretion in permitting opinion on causation based on examination of microscopic specimen slides, pathology and autopsy reports, government records, and publications concerning liver injuries caused by halogenated hydrocarbons). [19] See, e.g., Edmonds v. Illinois Central Gulf RR Co., 910 F.2d 1284 (5th Cir. 1990) (expert relied on studies about relationship of stress to heart disease but not on tests which could support application of studies to case); Thompson v. Merrell Dow Pharmaceuticals, Inc., 229 N.J. Super. 230, 242, 551 A.2d 177, 184 (A.D. 1988) (in Bendectin case, trial court properly barred expert who had acquired knowledge from limited library research conducted in preparation to testify). [20] Rule 703 of the Federal Rules permits experts to base their opinions on hearsay and the opinions and analyses of others provided that the facts and data on which the expert bases an opinion are of a type "reasonably relied upon by experts in the particular field in forming opinions or inferences." [21] Hutchinson v. Groskin, 927 F.2d 722, 724-26 (2d Cir. 1991) (defendant's medical expert testified that he had reviewed three letters from eminent oncologists that had been sent to defense counsel; two of the letter writers had never seen the plaintiff and must have based their opinions on data furnished by defense counsel; court termed the testifying witness "a conduit for hearsay testimony." Id. at 725.). [22] For instance, establishing causation for a so-called signature disease obviously requires different types of expertise than when the disease occurs naturally in the community. See Brennan, Helping Courts, supra, note 4. [23] Marder v. G.D. Searle & Co., 630 F.Supp. 1087 (D.Md. 1986), aff'd sub nom., Wheelan v. G.D. Searle & Co., 814 F.2d 655 (4th Cir. 1987). [24] The court ultimately granted judgment to defendant because of plaintiff's failure to prove causation after the jury was unable to reach a verdict following a three-week trial. [25] See, e.g., Christophersen v. Allied-Signal Corp., 939 F.2d 1106, 1113 (5th Cir. 1991) (en banc) (plaintiff's expert premised his opinion on a twenty-year history of exposure although record indicated that Christophersen had worked in defendant's plant for only fourteen years; majority held that Rule 703 would permit rejection of an opinion "founded on critical facts that are plainly untrustworthy, principally because such an opinion cannot be helpful to the jury." Id. at 1114.); Bernhardt v. Richardson-Merrell, Inc., 892 F.2d 440 (5th Cir. 1990) (affirming grant of summary judgment for defendant when discovery established that drug exposure occurred after that point in pregnancy at which particular birth defect could have occurred); Novak v. United States, 865 F.2d 718, 723-24 (6th Cir. 1989) (even if epidemiological evidence supported expert's contention of an increased incidence of persons contracting DM/PM within fifteen days of a swine flu vaccination, evidence showed that decedent's symptoms commenced considerably after fifteen days). [26] Other kinds of causation problems also arise. When the plaintiff suffers from a signature disease associated with defendant's product, the causation issue is limited to proving attribution in the individual case (i.e., asbestos, DES). In other instances, even in the absence of a signature disease, there may be general agreement by the courts (i.e., Dalkon Shield, benzene cases) or a finding by Congress (i.e., Swine Flu Act, National Childhood Vaccine Injury Act) that exposure to a particular substance may result in certain kinds of harm. Individual attribution issues remain, which, as the asbestos litigation demonstrates may have the potential to inundate the judicial system unless sophisticated and innovative management techniques are developed to lessen the amount of individualized fact-finding required for each plaintiffs claim. These cases have given rise to a good deal of discussion about the desirability of handling such litigation within the traditional tort compensation framework. These cases do not, however, for the most part produce charges that the legal system is reaching decisions on S&T issues that are incompatible with scientific principles. [27] Prior to the enactment of the Federal Rules of Evidence, many courts applied the so-called "general acceptance" test, first postulated in Frye v. United States, 293 F. 1013, 1014 (D.C.Cir. 1923), when confronted with expert testimony based on a scientific hypothesis not yet accorded judicial recognition. Rule 702 does not mention "general acceptance" and there are courts and commentators who do not believe that the Frye test has survived the enactment of the Federal Rules. See, e.g., J. Weinstein & M. Berger, Weinstein's Evidence, par. 702[03]. In addition, as is discussed below, the Frye test is not sufficiently responsive to the kinds of problems raised by issues such as proof of causation in toxic torts. For analyses of the Frye test, see Giannelli, The Admissibility of Novel Scientific Evidence: Frye v. United States, a Half-Century Later, 80 Colum. L.Rev. 1197 (1980); Black, A Unified Theory of Scientific Evidence, 56 Ford L.Rev. 59