11th Circuit Upholds Exclusion of Chemical Engineer's Testimony

In 1997, Mediterranean fruit flies were found in Florida. Government officials wanted to kill them, or at least to encourage them to fly back to the Mediterranean. So they decided to spray Florida with a mixture of protein bait and a pesticide known as malathion. The malathion was initially shipped from a Denmark plant to warehouses in Florida, Georgia, and Texas. From there, it was transported to Tampa, where authorities mixed it with the protein bait and then sprayed the city and surrounding areas. In short order, various Tampa domiciliaries became ill. They blamed the malathion and brought a putative class action. When malathion is stored above certain temperatures, you see, a process of chemical decomposition produces isomalathion, which is toxic to humans. And it can get hot in those southern warehouses in the summer.

But did it get hot enough? To show it did, plaintiffs retained Dr. Jack Matson, who boasted not only a Ph.D. in chemical engineering but also experience with pesticides (though not with malathion). Matson took weather data for the various storage locations and plugged them into a formula. Because of some indirect evidence that the temperature inside the Texas storage facility was actually 18 degrees higher than the ambient outdoor temperatures recorded by the weather service, he bumped up his temperature estimates for the warehouses in all three states, based on their structural similarities and basic latitudinal equivalence.

The district court excluded Matson's testimony, largely because Matson didn't actually investigate conditions at the non-Texas warehouses. Plaintiffs' causation case thereupon collapsed like a house of cards, and the claims were dismissed on summary judgment.

The Eleventh Circuit has released its opinion affirming the trial court's ruling. See Link v. Cheminova, Inc., No. 04-10160 (11th Cir. Feb. 24, 2005) (Birch, Barkett & Cox, JJ.). In its opinion, the panel rejects as unfounded the plaintiffs' argument that the lower court impermissibly evaluated Matson's credibility. According to the appellate opinion, the district court properly focused not on credibility, but on the reliability of Matson's methods. But the opinion does not stop there. It goes on to gloss the credibility/reliability distinction in a footnote, which we reproduce below:

We note that the argument made by the putative class representatives misconstrues the difference between a district court's evaluation of an expert's reliability, which is required by [Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993)], and an expert's believability or persuasiveness, which is reserved for the trier of fact, see [Quiet Tech. DC-8, Inc. v. Hurel Dubois UK, Ltd., 326 F.3d 1333, 1341 (11th Cir. 2003)]. As we explained in Quiet Technology, a district court's exercise of its gatekeeper function under Daubert "is not intended to supplant the adversary system or the role of the jury." See id. (citation omitted). Accordingly, a district court may not exclude an expert because it believes one expert is more persuasive than another expert. Additionally, a district court cannot exclude an expert because it believes the expert lacks personal credibility because of prior bad acts or other prior instances of untruthfulness. Vigorous cross-examination ensures that these issues of general credibility are properly presented for consideration by the trier of fact. See [Daubert, 509 U.S. at 596]. In evaluating the reliability of an expert's method, however, a district court may properly consider whether the expert's methodology has been contrived to reach a particular result. See General Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997)] (affirming exclusion of testimony where the methodology was called into question because an "analytical gap" existed "between the data and the opinion proffered"); see also Elcock v. Kmart Corp., 233 F.3d 734, 748 (3d Cir. 2000) (questioning the methodology of an expert because his "novel synthesis" of two accepted methodologies allowed the expert to "offer a subjective judgment . . . in the guise of a reliable expert opinion").

Leaving Dr. Matson to one side, how far is it really legitimate for a district judge conducting a Daubert inquiry to "consider" whether "the expert's methodology has been contrived to reach a particular result"? We will not finally resolve this vexing problem in a weblog post. But we would like to discuss a hypothetical.

Suppose then that we have videotape of Dr. Al Sayit, notorious hired gun, who will opine for anybody if the money is right, meeting in a seedy bar with counsel for plaintiffs. Suppose that the tape captures him saying: "Don't worry. There are several different methods and techniques by which this problem can be assessed, and I'll contrive to find one that yields the result you want." Suppose further that defense counsel somehow secures this tape, plays it for Dr. Sayit at his deposition, and asks Dr. Sayit how he can sleep at night. Suppose, finally, that the expert answers as follows: "Yes, it's true. I did exactly what I said I would do. I intentionally chose the one available method that would yield the result favorable to plaintiff, and I chose the method for that very reason. That's how I get paid. But the method has been tested and peer-reviewed, has a low error rate, is governed by standards controlling its application, is generally accepted in the field, and was faithfully applied to the data, all of which is transparently shown in my report. Whatever my motivations, then -- and I'll freely confess they were mainly pecuniary -- my work is scientifically valid."

We incline to the following view. Dr. Sayit's mercenary attitude may be a reason to scrutinize the reliability of his work with special care. But if his methods do indeed satisfy Daubert, then his motivations should be irrelevant to the admissibility of his testimony. There is a role for credibility findings by district courts in determining whether the factual predicates for admissibility are satisfied. If an expert on the standard of care in a medical malpractice case suddenly starts to stutter and perspire when he says he has performed this type of surgery on many occasions, a judge may properly conclude that the witness's testimony on that point is dubious. But the court's legitimate reliance on credibility determinations ends there. That's the way things have to be, we think, under a regime that purports to gauge the admissibility of scientific evidence by reference to the degree of objective epistemic support that the evidence can claim.

Another way to put this point may be to say that the word credibility isn't very precise at capturing what's at stake when an expert chooses among valid alternative methods with an eye to their outcomes. As suggested above, the term motivation is more exact. And when that more exact term is used, the case for excluding "contrived" evidence grows much weaker. All testimony is "contrived" in the sense that various motivations may plausibly be ascribed to the witness. The remedy for "contrived" but scientifically valid testimony should not be exclusion. It should be more testimony (including testimony from the contriving witness on cross).