Bogus Science in the Courtroom: A Consistency Check

Here in Pennsylvania, where Scopes II is pending in federal court, the respective parties have lined up their expert witnesses, according to a Philadelphia Inquirer story.

If you didn't know, Scopes II (a.k.a. Kitzmiller v. Dover Area School Dist.) involves a decision by the school board in Dover, PA, to teach the theory of "intelligent design" in biology classes alongside the Darwinian theory of evolution. A group of parents has filed suit, alleging that the school board's plan would violate the Establishment Clause under the Supreme Court's decision in Edwards v. Aguillard, 482 U.S. 578 (1987).

We're not constitutional scholars, but surely the school board's position would be buttressed if "intelligent design" were shown to be a theory supported by reliable scientific methodology. That is presumably why the Dover school board has retained a biochemistry professor and "intelligent design" proponent, Michael J. Behe, to testify on its behalf.

Let us assume, then, that Prof. Behe will indeed opine that the theory of "intelligent design" draws reliable support from legitimate application of the scientific method. Will that opinion face a Daubert challenge? Such a challenge, if mounted, might place the trial judge in the awkward position of effectively adjudicating the merits of a hotly contested controversy in the context of an evidentiary ruling. But so much is true, of course, in many cases where the admissibility of expert testimony is contested.

If mounted, should a Daubert objection to such testimony be sustained? On its face, Prof. Behe's theory would seem to face an uphill Daubert battle.

Not wishing to be embroiled unduly in a debate over the scientific details -- a debate, in any case, in which we possess only the most dubious qualifications to participate -- we will abstain from any analysis of whether the theory of "intelligent design" has been subjected to empirical testing, except to wonder what evidence would be counted, by the theory's adherents, as falsifying it.

The theory does suffer, however, from a want of support in the published peer-reviewed literature. Prof. Behe has himself defended the theory in such periodicals as the New York Times, the American Spectator, and the Weekly Standard. He has also found a sympathetic forum for his views at the Discovery Institute -- a conservative redoubt. But refereed scientific publications have proved much less receptive, as Prof. Behe admits:

Much of the material shown posted as "responses to critics" on this website was originally submitted to several science journals for consideration for publication. In every case it was turned down.

Prof. Behe chalks this up to a bias in favor of "orthodoxy" at peer-reviewed scientific journals. (Where have we heard that before?) In fairness, he also says "my experience with philosophy journals has been quite different." But of course the Dover school board is not proposing that the theory of "intelligent design" be taught as part of a unit in a philosophy course.

Do the methods employed by Prof. Behe have a known error rate? Are they subject to recognized standards governing their application? Maybe. But not that we can tell.

Does the theory of "intelligent design" enjoy general acceptance in the scientific community? Not by a long shot, as Prof. Behe would apparently concede. (He's fighting the "orthodoxy," remember?) The theory might fare better with the general American population. But it is the theory's popularity with parts of the general public that arguably creates the problem in the first place, insofar as the basis for that popularity is not scientific, but religious.

This brings us to a last consideration. In Kumho Tire, the Supreme Court urged that experts of all stripes should follow at least that degree of intellectual rigor prevalent in the routine practice of their specialties, outside the litigation context. Does the theory of "intelligent design" enjoy any normal scientific application -- say in facilitating empirical analysis of the detailed evolutionary process? Or is its sole natural habitat in the political arena, as an ideological adjunct to the religious right in the culture wars?

We intend that question seriously, not just rhetorically. Empirically serious proponents of "intelligent design" might well be quite curious, for example, about the frequency of divine intervention in the evolutionary saga. (One can imagine theological debates about whether an omnipotent deity would have the power to structure the physical laws of the universe in such a way that evolution might occur without further divine intermeddling -- and, if so, what evidence existed that the deity had chosen to proceed otherwise.) If divine assistance is necessary, as Prof. Behe contends, for "irreducibly complex" biological systems to develop, then it might be of interest to catalogue the instances in which new elements of "irreducible complexity" were introduced at various evolutionary stages. It is possible that somewhere, candidates for the Ph.D in "intelligent design" are pursuing just such investigations. But we doubt it. We suspect that the theory enjoys little currency, except in the context of essentially political efforts to repel the perceived cultural onslaught of nontheistic perspectives.

So should Prof. Behe's testimony come in? Our own general outlook on expert testimony is to favor a permissive approach, and to repose substantial trust in the trier of fact (in this case, the same federal judge who would rule on admissibility) to sort things out. But some species of expert evidence put our tolerant attitude to the test. Polygraph testimony is one example. And this, we must say, is another.

Perhaps there are readers, meanwhile, somewhere off our starboard ideological bow, who feel that "intelligent design" testimony should plainly come in. To them, we pose a question. Should the same result obtain, if an expert for a toxic tort plaintiff opines, based on no firmer methodological foundations than Prof. Behe's, that complex carcinomas cannot plausibly arise in humans except through industrial intervention?

Professional Discipline for Expert Physician Witnesses: A Consistency Check

Should physicians be vulnerable to professional discipline if they offer spurious testimony for the plaintiff (or the defendant, for that matter) in a medical malpractice case?

It is a recurring question in the "tort reform" debate. Advocates of malpractice "reform" generally defend the legitimacy of disciplinary proceedings against expert witnesses whom they see as having pressed "junk medicine" in frivolous cases. But others, including us, worry that such proceedings may have an unacceptable chilling effect on legitimate expert testimony.

Now, then. Should Sir Roy Meadow be hauled before a professional disciplinary tribunal and made to answer for his frequent appearances as a prosecution witness, in which he has routinely advanced the theory, now largely discredited, that under "Meadow's law," multiple crib deaths in the same family are generally the result of homicide?

Well, he's about to be, and for some reason our reaction is different. But should it be? Are we simply responding to media coverage that has often painted Sir Meadow as an angel of darkness? Or can a legitimate distinction be drawn? Is it worse, when innocent parents are jailed, or lose custody, based on dubious medical testimony, than when innocent physicians are held liable in malpractice cases? Does it matter, that evidence on the standard of care is mandatory and directly relevant in medical malpractice cases, whereas medical testimony about the statistical incidence of crib death is a very poor substitute for the evidence we would really want before sending someone to prison for murdering a child -- i.e., directly probative evidence that the defendant committed a crime? Are worries about the chilling effects of professional discipline less acute, in the context of evidence offered by prosecutors in criminal cases?

We're not sure. But we're thinking it over.

Washington Judge Hears Expert Testimony but Defers Ruling on Admissibility in Lawsuit over Gubernatorial Election

If any idea is more nightmarish, in a democracy, than having electoral contests decided by the judiciary, it may be the notion that electoral outcomes might depend on discretionary rulings on the admissibility of expert evidence. A state court judge in Washington (a Frye state) now finds himself in precisely that uncomfortable thicket, according to a story in the Seattle Times. In the pending litigation over who really won the vote in Washington's last election for governor, the trial judge has decided to hear evidence from the Republicans' expert witnesses for now, but to defer ruling on its admissibility until later.

"I can imagine how frustrated counsel has been with the court because you don't know me and you come to this court and I start making rulings which I'm sure some of you think are just not supported by any rules of evidence you've ever read," the judge told the lawyers. But the approach he has chosen is guided, he says, by the "desire to make sure the Supreme Court has as large a record as possible, for which they may not thank me, actually."

The disputed testimony is over how to estimate the impact of illegal votes by convicted felons. The Republican expert advocates a statistical method of "proportional reduction" at the precinct level. A Democratic expert says a better way to find out how the felons voted would be to ask them.

New Study Suggests Link Between Pesticides & Parkinson's

New Scientist is reporting on a newly released study affording new evidence of a link between pesticides and Parkinson's Disease. The study canvassed a population of 2756 subjects in five European countries, comprising 767 Parkinson's sufferers and 1989 healthy controls. Light pesticide users such as amateur gardeners were 9% likelier to develop the disease; high exposure users, like farmers, had a 43% higher risk. The study's authors acknowledge that they cannot say, from their data, just which pesticides may be responsible.

The same study also identified other, stronger risk factors. A family history of Parkinson's increases risk by 350%; getting knocked unconscious on one occasion, by 32%; getting knocked unconscious several times, by 174%.

Costs of the Tort System -- The Debate Continues

We learn from a post by Walter Olson at Point of Law of a new round in the debate over the costs of the U.S. tort system. In one corner, insurance consulting firm Tillinghast-Towers Perrin defends its figures and approach. In the other, the Economic Policy Institute responds, standing by its criticisms. For background, see our post of 5/18/05.

Doubting for Dollars

Worth reading: this post from Effect Measure on the industrial manufacture of scientific doubt.

11th Circuit Revisits Polygraphy

One result of the Supreme Court's 1993 ruling in Daubert was to create new ambiguity on what had previously been a largely settled point: the admissibility vel non of polygraph testimony in federal court. Yesterday's decision in United States v. Henderson, No. 04-11545 (11th Cir. May 23, 2005) (Barkett, Hill, & Farris, JJ.), shows that not all uncertainty on this issue has yet been dispelled.

First, some background. In 1989, before Daubert was decided, the Eleventh Circuit, sitting en banc, had announced two limited exceptions to the general rule then prevailing, under Frye, that polygraph testimony was inadmissible per se. See United States v. Piccinonna, 885 F.2d 1529 (11th Cir. 1989) (holding that "the science of polygraphy has progressed to a level of acceptance sufficient to allow the use of polygraph evidence in limited circumstances where the danger of unfair prejudice is minimized"). The first exception allowed for polygraph evidence to be admitted by stipulation. The second permitted its use for impeachment or corroboration of a witness at trial, provided the proponent gave advance notice to the opposition and afforded the opposing party an opportunity to conduct its own polygraph of the subject. The Piccinonna court explained that it was not adopting a rule of per se admissibility in cases where the exceptions applied, and went on to enumerate three specific grounds on which polygraph evidence might still be excluded notwithstanding the exceptions: "1) the polygraph examiner's qualifications are unacceptable; 2) the test procedure was unfairly prejudicial or the test was poorly administered; or 3) the questions were irrelevant or improper." Id. at 1537. The court also noted that its holding was not intended to "preempt or limit in any way the trial court's discretion to exclude polygraph expert testimony on other grounds under the Federal Rules of Evidence." Id. at 1536 (emphasis added). The precise scope of these "other grounds" was not defined, although the opinion did give exclusion under Fed. R. Evid. 403 as one example.

The Supreme Court's 1993 ruling in Daubert, which overruled Frye and substituted new standards for the admissibility of expert evidence, placed the continuing validity of Piccinonna in some doubt. The Eleventh Circuit was confronted with the issue of Piccinonna's continuing viability in United States v. Gilliard, 133 F.3d 809 (11th Cir. 1998), and responded by upholding a trial court's discretionary exclusion of corroborative polygraph evidence. The holding in Gilliard, however, was ambiguous, because the trial court decision affirmed in Gilliard rested not only on Fed. R. Evid. 702, but also on the alternative ground that the probative value of the evidence was outweighed by the potential for prejudice and confusion under Fed. R. Evid. 403. Thus the outcome in Gilliard might have been reckoned consistent with the view that the list of three permissible grounds enumerated in Piccinonna for exclusion of polygraph testimony (in cases subject to the two exceptions) still governed, so far as Rule 702 was concerned, and indeed that it represented an exhaustive list of permissible bases for exclusion under Rule 702 in such cases.

The majority opinion in yesterday's Henderson decision rejects the latter view. Judge Barkett's opinion, joined by Judge Farris of the Ninth Circuit (sitting by designation), upholds the trial court's adoption of a magistrate's ruling that a polygraph should be excluded for want of reliability under Daubert. The majority opinion cites the following language from the Piccinonna decision itself: "Our holding states merely that in the limited circumstances delineated above, the Frye general acceptance test does not act as a bar to admission of polygraph evidence as a matter of law." Piccinonna, 885 F.2d at 1536. Piccinonna, in other words, does not pronounce on whether the Daubert test may bar the testimony in such circumstances, in the panel majority's view.

Judge Hill, however, filed a dissent. In his view, Piccinonna held that "polygraph evidence is sufficiently reliable to be generally admissible at trial in this circuit," and that holding cannot be overruled, he says, except by another en banc panel of the Eleventh Circuit. To the argument that the Supreme Court's intervening decision Daubert changed the landscape, Judge Hill responds that the Daubert Court announced its ruling "without any mention of Piccinonna." That response may sound somewhat weak, but Piccinonna, after all, is an en banc decision in which the Eleventh Circuit devoted considerable effort to gauging the scientific reliability of modern polygraphy (as Judge Hill apparently remembers, having dissented, back in 1989, from so much of the Piccinonna opinion as held polygraph testimony sufficiently reliable to be admissible in any context, absent stipulation).

Can Daubert be assumed to have overthrown whatever jurisprudence on expert evidence antedated the Supreme Court's decision? That view seems bold. Is Daubert enough to undo the binding force of en banc decisions rendered by the Courts of Appeals before it was decided? Two Eleventh Circuit judges from the Henderson panel apparently hold opposite views on that subject, with a Ninth Circuit judge, sitting by designation, breaking the tie. What should an Eleventh Circuit panel now do, in the next case, if it wants its holding to be faithful to precedent? We, at least, are unsure. We know only that a decision excluding polygraph testimony would be hard to condemn on the theory that it reached the wrong result.

Daubert and the "Bush Seven"

Those seven candidates renominated for appellate judgeships by Pres. Bush after their confirmation was blocked by Senate Democrats the first time around? A reader inquires whether they have any track record on Daubert. Not the foremost question on people's minds about these nominees, perhaps. But an interesting issue, in its way, and one that may warrant some level of consideration in the judicial confirmation process.

So we've taken a quick look. All in all, we conclude that only one -- Justice Priscilla Owen of the Texas Supreme Court -- is on record taking explicit positions on the admissibility of expert evidence that could fairly be counted as controversial. Of course, two of the nominees have never been judges, and so they have no track record at all. If we've missed something of note, we invite readers to let us know.

The Hon. Janice Rogers Brown, a nominee for the Ninth Circuit Court of Appeals, has been a Justice of the California Supreme Court since 1996. She has authored only a few opinions touching on the admissibility of expert evidence under California's "Kelly/Frye" test, mostly in criminal cases. Her opinions do not suggest any unusual or exotic perspective on the subject. In People v. Brown, 16 Cal. Rptr. 3d 447, 94 P.3d 574 (2004), she dissented from the majority's ruling that expert evidence on "battered women's syndrome" -- in particular, on the tendency of domestic violence victims to issue false recantations -- is admissible under California law even where only one incident of abuse has occurred.

The Hon. Richard Allen Griffin, a nominee for the Sixth Circuit, has been a judge on the Michigan Court of Appeals since 1988. In Anton v. State Farm Mut. Auto. Ins. Co., 238 Mich. App. 673, 607 N.W.2d 123 (1999), Judge Griffin wrote the majority opinion upholding an endocrinologist's testimony that stressful events can precipitate Graves' disease (an autoimmune thyroid disorder), rejecting arguments that the causal link did not enjoy sufficiently general acceptance under Michigan's "Davis/Frye" test. In People v. Hubbard, 209 Mich. App. 234, 530 N.W.2d 130 (1995), appeal denied, 450 Mich. 965, 548 N.W.2d 634 (1996), he authored the court's decision holding that "drug profile" evidence should be inadmissible as proof of guilt. His opinion called the reliability of such evidence "suspect" and noted that "profile evidence often changes to meet the facts of any given case."

The Hon. David W. McKeague, another Sixth Circuit nominee, was appointed in 1992 as a U.S. District Judge for the Western District of Michigan. In Shanks v. Home Depot, Inc., No. 1:00-CV-383 (W.D. Mich. Dec. 7, 2001), he excluded an engineer's opinion that a design defect caused a stepladder to fail -- a garden-variety outcome, where, as in Shanks, the engineer performs no case-specific calculations or tests, and relies in vague and conclusory fashion on his experience and expertise. In Javetz v. Board of Control, Grand Valley State Univ., 903 F. Supp. 1181 (W.D. Mich. 1995), an employment discrimination suit by a university professor, Judge McKeague refused plaintiff's request to reconsider summary judgment in light of plaintiff's late-tendered expert evidence. The reasons were primarily procedural, but the judge did also comment that the expert's mere conclusory opinion on the ultimate question of discrimination would be of little assistance to the trier of fact.

William G. Myers, III, a Ninth Circuit nominee, is former Solicitor for the Department of the Interior. He has never been a judge, and reportedly he has never participated in a jury trial, so any indication of his stance on Daubert would be indirect at best. He has been criticized for taking a strongly anti-environmental, pro-industry line at Interior, but we could locate no specific evidence of his positions on scientific inquiry or evidence.

The Hon. Priscilla Owen, nominated for the Fifth Circuit, has served on the Texas Supreme Court since 1995. She authored the opinion in Merrell Dow Pharms. v. Havner, 953 S.W.2d 706 (Tex. 1997), which held that plaintiffs' evidence was insufficient to show that Bendectin caused their baby's birth defect. That is not an unusual outcome for a Bendectin case; most courts have reached similar conclusions. The Havner opinion, however, goes on to set standards for scientific evidence in toxic tort litigation that many would consider unusually stringent. Among other things, the opinion endorses a test to measure causation by whether epidemiological data show the plaintiff's toxic exposure to have doubled his or her statistical risk of disease -- a standard long popular with the defense bar, but one broadly rejected by courts in most other jurisdictions.

William H. Pryor, Jr., a nominee to the Eleventh Circuit, has served as Alabama's Attorney General since 1997. We have found nothing from Pryor explicitly addressing the admissibility of expert scientific evidence. He is on record, however, opposing litigation against the tobacco, gun, lead paint, and managed care industries, on the ground that such cases "leave only the personal injury lawyers better off." Some might think that a clue to his sentiments.

The Hon. Henry W. Saad, another Sixth Circuit nominee, has served on Michigan's Court of Appeals since 1994. In Forest City Enters. v. Leemon Oil Co., 228 Mich. App. 57, 577 N.W.2d 150 (1998), appeal denied, 459 Mich. 948, 616 N.W.2d 170 (1999), Judge Saad authored the court's opinion upholding the admissibility of testimony from defendant's hydrogeologist in a groundwater contamination case.

5th Circuit Affirms Exclusion of Causation & Damages Testimony in Antitrust Suit

The Fifth Circuit has issued an unpublished opinion affirming the trial court's exclusion of plaintiffs' evidence on causation and damages in an antitrust action involving the tortilla market. Neither expert sufficiently grounded his opinion in the facts of the case. See El Aguila Food Prods., Inc. v. Gruma Corp., No. 04-20125 (5th Cir. May 17, 2005) (Davis, Smith, & DeMoss, JJ.).

LexisNexis Assimilates Jurispro Database

LexisNexis has struck a deal with Jurispro to make the Jurispro database of expert witnesses available to Lexis subscribers. Details here.

Costs of the Tort System -- Another Take

We have posted previously in this space about the periodic reports of Tillinghast-Towers Perrin, an insurance consulting firm, on the "costs of the U.S. tort system." (See our posts of 12/28/03 and 1/19/05.) Now a new report from the Economic Policy Institute reflects a decidedly different take. The EPI study, entitled "The frivolous case for tort law change: Opponents of the legal system exaggerate its costs, ignore its benefits," offers this summary of its findings in the opening paragraph:

Tort litigation has been blamed for driving liability insurance premiums to excessive levels, reducing real wages and overall employment, undermining corporate profits, dampening productivity growth, discouraging research and development, and generally reducing the willingness of corporations and individuals to take reasonable risks (such as introducing new products) that may benefit themselves and society. There is scant evidence for any of these claims. To the contrary, macroeconomic trends since the early 1990s are especially inconsistent with the argument that supposedly high and rapidly rising tort costs have inflicted serious harm on the economy. Yet the legal system's critics continue to argue that there is a tort liability "crisis" that warrants changing the tort system, and that change is a key element in bolstering economic growth in the future. Yet these critics provide no credible evidence to support their assertions. In fact, what little effect changing the tort system will have on the economy might hurt job creation rather than help it.

The Kansas City Star has more -- including a quote from Russ Sutter, primary researcher for the Tillinghast-Towers Perrin report, to the effect that "tort reform" advocates have used the TTP data "in a way that's probably misleading."

Project on Scientific Knowledge and Public Policy -- 2005 Awards Program

The Project on Scientific Knowledge and Public Policy at George Washington University has announced a program to fund empirical pilot studies to measure the use and effects of the Supreme Court’s Daubert decision and subsequent rulings. The project will fund up to four pilot studies at an average of $25,000 each. If pilot studies are successful, project staff and advisors will support investigators in seeking additional funding for more comprehensive research. The program invites initial letters of intent by Friday, July 15, 2005. To avoid futility in proposal-writing, the program intends to invite full but relatively short proposals only from investigators whose proposed work is thought to have strong prospects for an award. Awards will be made no later than Friday, December 2, 2005.

To obtain a copy of the Request for Proposals, go to: http://www.defendingscience.org/.

For questions, contact Molly Jacobs at daubert@bu.edu.

Swedish Study Links Cell Phones with Brain Tumors

Multiple press sources are reporting on a new study linking rural cell phone usage with brain tumors in a Swedish population. The sample comprises 1400 adults residing in central Sweden who were diagnosed with malignant brain tumors between 1997 and 2000. Subjects having used cell phones in rural areas for three years or more were three times as likely to suffer brain tumors as their urban counterparts. Five-year users in rural areas suffered a risk quadruple that of urban dwellers. It is hypothesized that the stronger signal strength used in rural areas may be the reason. The study has been published in Occupational and Environmental Medicine and is available online.

Update 5/18/05: Our original post inadvertently linked to a news story instead of the study itself. We've fixed the mistake. The text of the study can be found here. The URL is:

http://press.psprings.co.uk/oem/june/390_om17434.pdf

1st Circuit Says Law Enforcement Officer's Testimony on Modus Operandi of Drug Dealers Is Lay Opinion

Last week, the First Circuit held that testimony from an arresting officer on the modus operandi of drug dealers should be treated as lay opinion, not expert testimony -- at least where the testimony is grounded in the officer's own law enforcement experience. The ruling, which departs from the general pattern prevailing in other circuits, would imply that such modus operandi testimony is not subject, within the First Circuit, to the pretrial discovery requirements applicable to expert evidence. See United States v. Ayala-Pizarro, No. 04-1038 (1st Cir. May 12, 2005) (Torruella, Lynch, & Lipez, JJ.).

8th Circuit Upholds Exclusion of Testimony on Future Damages in Manufacturer's Suit Against Supplier

The Eighth Circuit has published an opinion upholding the trial court's exclusion of expert damage testimony offered by a manufacturer in its suit against one of its suppliers. The supplier's pigment turned out to be unsuitable for use in the manufacturer's outdoor PVC fencing product. To honor its product warranties, the manufacturer was obligated to replace the fencing when consumers complained that the color had faded. The manufacturer's expert estimated future damages based on the assumption that every last bit of defectively pigmented fencing already sold would be subject to warranty claims. The testimony was excluded for lack of fit, because it did not account for record evidence suggesting that warranty claims would be made for only a fraction of the fencing. See Nebraska Plastics, Inc. v. Holland Colors Americas, Inc., No. 04-2035 (8th Cir. May 13, 2005) (Arnold, Bowman, & Gruender, JJ.).

We Now Resume Our Regularly Scheduled Programming

We are back from three weeks of trial, and so we can once again devote our lives to keeping you updated on developments in the law of expert evidence. Stay tuned for news of what the courts have been doing during our hiatus.