Peter Nordberg and David Bernstein debate the study on Daubert discussed in this post. If I can take up Mr. Nordberg’s challenge to identify problems with the SKAPP report, I’ll just identify a handful that immediately came to mind as I read it. (This is long, so I’ve moved it off the main page. Click the link below.)

This is, mind you, nowhere near a systematic evaluation.

SKAPP cites Professor Berger complaining that Daubert may preclude a plaintiff from “tell[ing] his or her story” when causation cannot be proved. But the courtroom is not a place for storytelling; if a plaintiff has a tort claim against a defendant, causation is (or at a minimum, should be) an element of the plaintiff’s case. Berger’s complaint appears not to be with Daubert, but with the idea that defendants “exposing the public to risk” is insufficient by itself to involve civil courts and plaintiffs’ lawyers. It is not enough that the defendant has done something that a jury perceives as bad; before plaintiffs can collect, they must show that their injury was caused by the defendant’s wrongful conduct.

SKAPP expresses concern that “after an initial spike in the number of challenges to expert testimony, the incidence began to fall off dramatically, suggesting that plaintiffs increasingly decided not to bring actions that relied heavily on scientific testimony unless that testimony met the Daubert standards.” Leaving aside the question of whether the category of “scientific testimony that does not meet the Daubert standards” is social good, it is hardly surprising that a new rule is more heavily litigated in the initial years of its existence. As uncertainty over the scope of the rule decreases because of precedents filling in interstitial gaps, there is less to litigate: proponents of evidence take more care in meeting evidentiary standards; opponents of evidence are less likely to mount challenges to legitimate testimony as it becomes clearer where the lines are drawn.

SKAPP uses “number of trials” as a metric and laments its decline. But even assuming it has accurately performed the measurement, it is a meaningless statistic. Trials occur where parties have wildly divergent evaluations of the merits of a claim. Settlements occur where parties have similar valuations for the lawsuit. We don’t know whether Daubert is causing plaintiffs to forego legitimate claims, or whether it is instead making settlement more likely. There is a ceteris paribus problem: we also don’t know whether corporations’ experience with the increased willingness of juries to award jackpot damages has made them more willing to settle in the 1990s than they were in the 1970s or 1980s. (The expanded use of the class action mechanism alone would seem to account for a substantial decline in tort actions going to trial.) SKAPP seems to assume that corporations have not become better citizens in the wake of expanded tort liability so that trials would remain a constant–an assumption that directly contradicts both the premise plaintiffs’ lawyers use to justify increased punitive damages awards and the complaint of tort reformers that corporations are forced to engage in inefficient practices to avoid liability risks.

There is a confusingly inconsistent complaint that Daubert is rarely applied in the criminal courts to exclude prosecution scientific evidence, and that junk science might be being used to convict criminals. I’m left uncertain what SKAPP believes is the appropriate standard of review of scientific evidence, as they have just spent several pages complaining of supposed unjust effects from the Daubert standard. It’s not even clear that the allegation is true. Peter Nordberg’s site identifies numerous occasions where criminal evidence was challenged and excluded.

Finally, in an age where plaintiffs’ lawyers have received billions of dollars from their role in the tobacco litigation, it is pure propaganda to complain about “wealthy and powerful” defendants. There are resources to spare for litigation against any defendant that has actually committed wrongdoing with more than enough left over to go after defendants that haven’t.

The complaint that Daubert sometimes results in the inappropriate exclusion of legitimate science is an ironic one. The fact that legal rules sometimes result in mistakes of application has been the argument of many a tort reformer against expansion of tort liability to address public policy issues better resolved in the other two branches of government. Courts make mistakes, and the more issues that come within the judicial system’s ambit, the more mistakes that will be made. Cf. Frank H. Easterbrook, “The Limits of Antitrust,” 63 Tex. L. Rev. 1 (1984).

The question should not be whether there are occasionally false negatives in the application of Daubert (there are), but whether the social cost of the false negatives (and, mind you, the false positives that still occur) when Daubert is used as a screening mechanism outweigh the social costs of a different set of mistaken decisions and different set of compliance and litigation costs under a different regime. The SKAPP report makes no effort to answer that question, and seems to simply assume that any false negatives are too many without any acknowledgement that there is a real burden to society from false positives.

Filed under: Daubert, junk science, procedure