Recently in Scientific Evidence Category

Tuesday, junk science led a Philadelphia jury to award three women (ages 66 to 68) verdicts ranging from $20M to $28.75M in suits alleging that Pfizer's Prempro caused their breast cancer. [Bloomberg; Phil. Inquirer]

Pfizer has won over 3000 of these cases, including 44 that were set for trial, but when juries can award $20M/plaintiff (plus potential punitive damages: Bloomberg discusses a "second phase" of the trial without being clear what that means), it's still profitable to bring long-shot cases like these were, especially when the state courts are less skeptical than federal courts about allowing experts to quack-ily apply one study's results to disparate facts. Which is perhaps why plaintiffs' lawyers have twisted themselves in knots to avoid federal court. Coincidentally, Monday's WSJ discusses the problem of forum shopping for the Philadelphia judicial hellhole.

As David Oliver points out, we correctly don't hold scientists liable for making mistakes in cutting-edge science; when an Italy even considers the issue, as they did in an unsuccessful criminal prosecution of scientists that failed to predict a deadly earthquake, we justifiably ridicule them. Science is a process of trial and error, and we want scientists to explore hypotheses that might be wrong without fear of legal consequence for failure. So why do we hold manufacturers liable for failing to anticipate the future judgments of lay juries evaluating the predictions of scientists presented at trial testimony, most of which are arguable at best?

Of course, this sort of double-standard is common. We don't hold lawyers liable for the larger effects on society when they use abusive litigation to make us less safe by driving safe and effective products from the marketplace. We let juries second-guess doctors when experts disagree over the correct course of medicine, but give that discretion to attorneys free from liability.

On Friday, the Center for Class Action Fairness LLC filed its opening brief in the Third Circuit case of Dewey v. Volkswagen (10-3618). The case presents some interesting jurisdictional issues as a side effect of Devlin, plus run-of-the-mill economic quackery and an inexplicable decision to arbitrarily include in the class a million vehicles without providing the same pecuniary reimbursement benefit available to the rest of the class. The settlement as a whole is inexplicable: the class counsel's economist calculated that defendant Volkswagen was going to spend $55 million on a service action that would prevent $24 million in future damage. This, the economist concluded, was worth $103 million to the class: the $55 million spent on repairs, plus the $24 million in future damage avoided, plus the $24 million in increased resale value from not being damaged!

Most entertaining, however, was the economist's insistence that the injunctive relief—a letter to the class informing them of a revised maintenance schedule—was worth millions of dollars to the class because now the class would get the benefits of purchasing maintenance. As we argued:

Imagine three hypothetical class-action settlements with Apple. In Hypothetical Apple Settlement #1, every class member receives a free $700 iPad. It is easy to conclude that the value of this settlement would be $700/per class member; even if a class member did not want the iPad, they could sell it on the market.

In Hypothetical Apple Settlement #2, every class member receives a brochure describing the iPad that contains a coupon for $50 off an iPad. Here, the Class Action Fairness Act dictates the valuation of the settlement: it is worth $50 times the number of coupons redeemed. 28 U.S.C. § 1712.

Now imagine Hypothetical Apple Settlement #3, where every class member simply receives the iPad brochure (or say "educational iPad information") without the coupon. How should that settlement be valued? According to Eads's methodology, the educational information tells a class member that an iPad exists (whether or not they already knew that: the Eads Report assumes 100% ignorance replaced by 100% knowledge), and could lead a class member to purchase an iPad worth $700 to the class member. Therefore, according to the Eads Report, Hypothetical Settlement #3 is worth $700 per class member, as much as Hypothetical Settlement #1, and more than a settlement that included a coupon.

The Center for Class Action Fairness LLC is not affiliated with the Manhattan Institute.

When family court reaches levels of self-parody: man's first wife kills his children. She pleads temporary insanity, implausibly blames Prozac for deaths, escapes criminal punishment. Man remarries, has more kids, divorces, takes up again with killer of his first children. Second ex-wife has no luck in family court expressing reluctance leaving kids with child-killer. [Alkon; AP/NYT]

Meanwhile, in the same jurisdiction, a horrifically crooked expert in the same family court jurisdiction got disciplinary records sealed, and had a reign of terror harassing women as an evaluator in child-custody cases. If he hadn't been caught videotaping his employees using the restroom, he might still be running amuck in family court. [Seattle Times via OL]

Father's Day special: In "facilitated communication," an aide helps a profoundly autistic child type answers to questions. This is pure quackery: in reality, it's the communicator who's answering the question, rather than the child, as demonstrated by experiments where the communicator is unable to hear the questions. But it provides enough false hope that some believe in it, and it can lead to false sex abuse allegations as happened in the case of the Wendover family. The Detroit Free Press runs an extraordinary six-part series (via @walterolson) on the trauma inflicted on this one family by junk science and gullible prosecutors. The Wendover family civil suit against government officials is pending, but they seem relatively fortunate that they were only separated from their children for 106 days before prosecutors dismissed the case (while covering their butts by claiming it was because the ostensible victim was too scared to testify); we've certainly seen other prosecutors (cough, cough, Martha Coakley) continue witch-hunts until they lead to prosecutions. Worth noting: Brian Dickerson's criticism of the "cowardice and cronyism" of judges who delayed matters for four weeks past when it was obvious prosecutors had no case.

The Abnormal Use blog points us to a quack expert report hypothesizing that a defective (rather than unattended) stove caused a fire that destroyed a home, defeating summary judgment, and likely leading to a settlement between the stove manufacturer and the home-owner's fire insurance. The subrogee likely won the battle, but insurance companies (and consumers as a whole) lose the war when they encourages nonsensical litigation unrelated to the facts for short-term profit. A collective action problem, to be sure.

One problem with the idea of court screening of scientific experts under Daubert is that some judges simply don't have the scientific training not to be snowed by the same junk science that has the risk of unfairly snowing jurors. (I saw this in a class action settlement where a junk-economics expert waved his hands and gave an ipse dixit that requiring defendants to sell larger gallons of gas would mean that class members would get millions of dollars of free gasoline in the future. Oy.)

Attorney Nathan Schachtman (via Childs) finds a very good example in the Avandia MDL, which independently raises the question why judges who don't understand science or questions of statistics are assigned to MDLs where science and questions of statistics are going to be decisive factors. [more from Schactman]

Schactman also has an interesting post about the misuse of historians as expert witnesses.

A major tort reform package could pass the Legislature and be on its way to Wisconsin Gov. Scott Walker for his signature by the end of the week. While other new Republican governors -- notably Tom Corbett in Pennsylvania and Rick Scott in Florida -- have also declared civil justice reform to be a priority, Walker is being most aggressive, aided by heavy Republican majorities in both houses of the Legislature. His first act as governor was to call the Legislature into special session to enact tax, government-reform and economy-related legislation, including tort reform.

Walker's Executive Order #1 set specific goals for the liability reform bill:

Limiting noneconomic damages awarded in actions against long-term care providers; actions against manufacturers, distributors, sellers, and promoters of certain products; confidentiality of health care services reviews; use as evidence of information regarding health care providers; reporting of quality indicators identifying individual hospitals; homicide or injury by negligent handling of a dangerous weapon, explosives, or fire; criminal abuse of individuals at risk; criminal abuse and neglect of patients and residents; evidence of lay and expert witnesses; damages for frivolous claims; and punitive damage awards.

Both the Senate and Assembly Judiciary Committees held hearings on the legislation last week, with the Senate Judiciary passing the bill out for a vote on the Senate floor on Tuesday. A House vote would follow Thursday.

The bill is SB1JR1. After being amended by the Senate committee, the bill would impose a cap of punitive damages of twice compensatory damages or $200,000, whichever is greater. State standards of proof would now reflect the Daubert standard, and the law's provisions would be applied prospectively only.

In the British Medical Journal, Brian Deer provides extraordinary detail of the extent of fraud in Andrew Wakefield's paper, which, at the behest of trial lawyers, created a gigantic anti-vaccine scare that persists to this day, with untold numbers of measles cases resulting. Can you imagine the fuss if a doctor paid by a pharmaceutical company falsified study results for the profit of that corporation with adverse health effects? No signs of any class actions against trial lawyers or Robert F. Kennedy, Jr., though.

The Lancet paper was a case series of 12 child patients; it reported a proposed "new syndrome" of enterocolitis and regressive autism and associated this with MMR as an "apparent precipitating event." But in fact:

• Three of nine children reported with regressive autism did not have autism diagnosed at all. Only one child clearly had regressive autism

• Despite the paper claiming that all 12 children were "previously normal," five had documented pre-existing developmental concerns

• Some children were reported to have experienced first behavioural symptoms within days of MMR, but the records documented these as starting some months after vaccination

• In nine cases, unremarkable colonic histopathology results--noting no or minimal fluctuations in inflammatory cell populations--were changed after a medical school "research review" to "non-specific colitis"

• The parents of eight children were reported as blaming MMR, but 11 families made this allegation at the hospital. The exclusion of three allegations--all giving times to onset of problems in months--helped to create the appearance of a 14 day temporal link

• Patients were recruited through anti-MMR campaigners, and the study was commissioned and funded for planned litigation.

Update: Daniel Fisher independently makes the same point at Forbes; and Jonathan Adler @ Volokh has more links and quotes.

Was a Texas gas company fraudulently shortchanging customers, or was it a function of a statistical calibration error? A $30M jury verdict rides on the answer. And, quite frankly, it's a simple matter of math: did the errors systematically disfavor the consumer such that it could not have occurred by chance, or were they randomly distributed with the realm of chance and tolerances and the lawyers cherry-picked? I have no idea who's right, but only one of the two competing experts would have been testifying consistent with the laws of probability, so it's disturbing that the Houston Chronicle (via ABAJ) simply treats it as a "he said, she said" issue, since it suggests that the judge let both testify. (Ceteris paribus, I'm inclined to believe the Texas attorney general in this case if, as they claim, there was an attempt at last-minute calibration once a raid began. But given the availability of statistical evidence, there's no reason for the ceteris to be paribus.)