[3] REAVLEY, Circuit Judge, with whom POLITZ, KING, JOHNSON, WILLIAMS and HIGGINBOTHAM, Circuit Judges, join dissenting from the denial of rehearing en banc:

If the court, by refusing to rehear en banc, approves the panel's application of Boeing Co. v. Shipman, 411 F.2d 365 (5th Cir. 1969), I wonder what is left in this circuit of the Boeing rule or, for that matter, the Seventh Amendment right to trial by jury.

Six highly qualified and experienced experts testified that bendectin is a human teratogen, i.e. capable of causing human birth defects. Three of them testified to the opinion that bendectin was a cause of Rachel Brock's deformation. Their opinions of causation were not the product of faulty syllogisms but were predicated upon medical study and research and upon their explanation of the process by which the doxylamine element in bendectin can interfere with the development of nerve cells in an embryo. The panel picks at details in the testimony lacking expert consensus, but its characterization of this voluminous expert proof as "speculation" could just as well doom virtually all expert testimony. The panel reaches its climax with the novel declaration that only epidemiological studies can prove causal relation between bendectin and birth defects, and it enters into the debate with Dr. Glasser on the statistical significance of the Heinonen study. In the absence of expert consensus must we now always await population studies before a jury verdict may be based upon medical opinion? So says the panel, at least for bendectin cases. This, despite the testimony here that case reports and laboratory research reveal teratogens and that no epidemiological study has ever discovered a teratogen.

The panel tries to draw support from Richardson by Richardson v. Richardson-Merrell, Inc., 857 F.2d 823 (D.C. Cir. 1988), but that court ruled that the plaintiff's medical expert lacked an adequate basis, of the type reasonably relied upon by experts in his field, for his opinion, which left no support for the verdict. That record was a far cry from the one we are called upon to review here. Though it tinkers with bits of the extensive expert testimony in this record, our panel does not purport to hold that the foundation for the opinions of these experts fails to satisfy Federal Rule of Evidence 703.

We dissent.

[8] PATRICK E. HIGGINBOTHAM, Circuit Judge, concurring in Circuit Judge REAVLEY's dissent:

I join Judge Reavley in dissenting from the refusal of the court to consider this case en banc. I do not dissent because I think that the panel reached the wrong conclusion — about that I am not sure. I prefer to examine this case en banc because the panel's opinion shies from a direct confrontation with one of the more vexing problems currently facing the federal courts — the role of experts. This case raises important questions about the role of experts in the federal courts, including whether we should accept opinions of experts not based upon a generally accepted scientific principle and the more broadly stated concern that substantive principles such as tort law are not handling science issues in a rational manner. We should confront them, and state the guiding principle, if we can.

A significant body of judicial and academic comment has expressed fear concerning these dangers. Several examples of such comments were collected by Barry Epstein and Marc Klein in an article for Seton Hall Law Review. The authors found such concerns expressed as early as 1853 in the following language from a federal court opinion:

[T]he experience of all concerned in the administration of justice tends to the conclusion that this species of evidence is less satisfactory than any other.... [W]here there is any room for a difference of opinion, experts, in about equal numbers, will generally be found testifying on each side.

Much more recently, the Attorney General's Tort Policy Working Group reported an "increasingly serious problem" of

reliance by judges and juries on noncredible scientific or medical testimony, studies or opinions. It has become all too "common" for `experts' or `studies' on the fringes of or even well beyond the outer parameters of mainstream scientific or medical views to be presented to juries as valid evidence from which conclusions can be drawn. The use of such invalid scientific evidence (commonly referred to as `junk science') has resulted in findings of causation which simply cannot be justified or understood from the standpoint of the current state of credible scientific and medical knowledge. Most importantly, this development has led to a deep and growing cynicism about the ability of tort law to deal with difficult scientific and medical concepts in a principled and rational way.

It has been eighty-eight years since Judge Learned Hand identified two practical defects with a system depending on a conflict between expert witnesses in front of a lay jury. First, he wrote, "the expert becomes a hired champion of one side":

Enough has been said elsewhere as to the natural bias of one called in such matters to represent a single side and liberally paid to defend it. Human nature is too weak for that....

Second, Hand asserted, the conflict between opposing experts leaves the jury no rational basis for choosing between them:

The trouble with all this is that it is setting the jury to decide, where doctors disagree. The whole object of the expert is to tell the jury, not facts, as we have seen, but general truths derived from his specialized experience. But how can the jury judge between two statements each founded upon an experience confessedly foreign in kind to their own? It is just because they are incompetent for such a task that the expert is necessary at all.

It strikes me that the issue in this case revolves around the admissibility of the expert testimony. Indeed, we ought to be wary of opinions of scientists expressed in court free of their traditional restraints of peer review and scientific consensus. See In re Air Crash Disaster at New Orleans, Louisiana, 795 F.2d 1230 (5th Cir. 1986). 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. I share much of the panel's concern and might join them if I were confident of their focus. Yet, the translation of these concerns into administrable rules is absent. The concerns only manage to bubble to the surface and then not in a coherent way that fully discloses.

Regardless, the issues reach further than questions of correct application of the rules of evidence in a particular case. Surely, these questions are worthy of en banc consideration, particularly when we cannot today close our eyes to the rapidly increasing upward pressure of an expanding base of federal cases. This upward force, capped at the top by the full capacity of the Supreme Court, is generating serious proposals for changing the structure of the Courts of Appeals. We are obligated to use our resources as best we can to mitigate this pressure. This means that we must be more willing to utilize en banc procedures to tackle difficult problems by reducing intra-circuit and inter-circuit conflicts. It also means that we ought not leave this lion in the busy street of our trial courts and inevitably multiply the appellate workload.

Describing the issues in the language of efficiency and impact on the courts presents only a part, and not the most important part, of larger questions of fairness and judicial roles and our commitment to the jury system. Ultimately, it is no more than that we are obligated to state the law, certainly if we truly expect people to order their affairs by its measure. We are not doing that here. When we fail to explain, we must ask if it is because we cannot. I join Judge Reavley's wise caution.