"Risk Analysis in the Courts: A Roadmap for Risk Analysts"
Menu of 1998 Full-text Online Casebook
Poster 21a
SUMMARY AND SYNTHESIS
Frequent Misunderstandings, Misuses, And Other Mishaps
CONFUSION CONCERNING RISK CONCEPTS
Probability versus Uncertainty
Probability versus Harm
Herskovits v. Group Health Cooperative, 664 P.2d 474 (Wash. 1983).
The Washington Supreme Court held that the plaintiff, whose chance of survival was reduced from 39% to 25% by the negligent failure to diagnose his cancer, was entitled to recover for "damages caused by his premature death" as a result of the lesser chance of survival. The court, apparently reluctant to discount the value of a whole life by the reduced statistical chance of survival, converted loss of a chance to loss of longevity, without evidence of the relationship between the two concepts.
ABUSES OF RISK DATA AND METHODS
Two Wrongs Don't Make a Right
*Wagon Mound Duo.
There are many problems with an adversarial method of determining risk or other intangible facts. One of the most obvious problems is that in some cases it may be in both parties' interests to base the case on a patently "false" factual statement of risk. This possibility is starkly illustrated in two classic Wagon Mound cases, which were brought as a consequence of the same accident, but where the court's factual finding regarding the magnitude of the risk at issue differed radically in the separate cases only because arriving at the "false" probability of risk was legally beneficial to both Plaintiff and Defendant in the first case, but not in the second. The Wagon Mound incident involved a terrible fire in a harbor that damaged, among other things, Plaintiff‑1's repair equipment and Plaintiff‑2's boat. Defendant (the charterer of the boat "Wagon Mound") was responsible for releasing the furnace oil into the harbor which ultimately caught fire. Liability turned on the question of whether the risk of fire was foreseeable, since furnace oil has such a high boiling point and is unlikely to catch fire under normal circumstances. In these two cases, the two sequential plaintiffs argued the risks of fire in opposite ways. Each of these diametrically different presentations of the risk of fire were accepted by the very same court as equally true and valid facts. In the first case, involving Plaintiff‑1, both Plaintiff‑1 and Defendant were eager to say that there was no foreseeability of furnace oil catching fire in a harbor. Overseas Tankship, (U.K.) Ltd. v. Dock & Engineering Co., Ltd., 1961 A.C. 338 (Privy Council). The Defendant was eager for a zero‑risk‑of‑fire finding for the obvious reason that it would establish that a fire was unforeseeable and the Defendants should therefore not be held responsible for such unforeseeable events. Plaintiff‑1 was also eager to show that the risk of fire was unforeseeable because otherwise their cause of action could be barred for contributory negligence (it was a spark from their welding operations that set the fire going). The court ultimately found that the risk of fire was zero and denied recovery to Plaintiff‑1 because it was unfair to hold Defendant responsible for such unforeseeable harm. In the second case, the Plaintiff‑2 owner of the charred ship played no part in creating the fire and thus proved up the "true" fact that there is a positive, ex ante risk that furnace oil might catch fire when spilled in a harbor. Overseas Tankship (U.K.), Ltd. v. Miller Steamship Co.1 A.C. 617 (Privy Council 1967). In this second case, Plaintiff‑2 recovered based exclusively on the finding that the risk of fire (for the same incident) was greater than zero and Defendant should be held responsible for foreseeable risks. [Wagner]
CREDIBILITY: CONFUSING THE ANALYST FOR THE ANALYSIS
Mannerisms of the Witness
Sheila Jasanoff reports, on Science at the Bar: Law, Science and Technology in America 53‑55 (1995), that courts often deploy wholly inappropriate tests for the "credibility" of expert witnesses, confusing issues of deception with issues of foundation and reliability. The latter are at issue in expert testimony, but credibility is often evaluated according to the former. She quotes the decision of Judge Marvin Shoob in Wells v. Ortho Pharmaceutical Corp., 615 F. Supp. 262, 273 (D. Ga. 1985), in which he credited or discounted experts based on their demeanor on the stand and ability to translate technical concepts into lay terms. Jonathan Harr reports a similar instance in A Civil Action (1995), his account of the Woburn groundwater contamination litigation. There, the defendant's lawyer successfully discredited the testimony of plaintiff's groundwater expert, even though (a) the individual was by all accounts the preeminent authority on hydrogeology in the country, and (b) was ultimately proven (after the litigation was over) to be correct.
95% Confidence Level
Hodges v. Secretary of the Department of Health and Human Services, 9 F.3d 958 (Fed.Cir. 1993).
Must a plaintiff resorting to scientific proof of causation under a legal standard that it is more probably true than not true that a toxic exposure caused a disease meet a stringent scientific standard such as 95‑percent certainty? Judge Newman, dissenting in Hodges v. Secretary of DHSS, focused on the distinction between legal probability and scientific certainty. The judge examined the legal consequences of probabilistic analysis and concluded that a causation relationship could be ascertained even from data possessing low confidence levels that may not attain statistical significance. Data bearing lower than 95% confidence might not support a conclusion about causation to a medical certainty, but could satisfy the more‑likely‑than‑not legal standard used in civil litigation. In Ferebee v. Chevron Chemical Co., 736 F.2d 1529 (D.C. Cir. 1984), the court recognized the "irrelevance" of scientific certainty to "more likely than not" causation. In Rubanick v. Witco Chemical Corp., 593 A.2d 733, 747 (N.J. 1991), the New Jersey Supreme Court rejected requiring a level of scientific proof that was unavailable, and stated that: [T]oxic tort litigation does not frequently encounter well established and widely accepted scientific theories of causation that can, at the level demanded by the scientific method, precisely delineate the causal path between the toxin and pathology. The Childhood Vaccine Injury Act, the federal statute under which the plaintiffs brought their claim, requires that causation be proved by a preponderance of the evidence. This case was decided against the petitioner by a divided panel of the Federal Circuit. Judge Newman stated unequivocally that the burden of proof required by the Vaccine Injury Act, as well as general tort liability law, is not the 95% confidence level expected by scientists; instead, it is the greater‑than‑50% confidence level set by Congress. The special master who decided the case had discounted the petitioner's medical evidenceC14 epidemiological studies that support general causation, or that the vaccine in question is capable of causing death in a child. Judge Newman's dissenting opinion relied upon Daubertv. Merrell Dow Pharmaceuticals, Inc., 113 S.Ct. 2786 (1993), as a basis to admit additional types of evidence. The U.S. Supreme Court stressed the obligation of the court to review the scientific reasoning and methodology of the epidemiological evidence, to evaluate the reliability of scientific expert opinions, and to weigh the totality of the evidence as applied to a particular case. Yet the proffered epidemiological evidence was not evaluated by the special master although it was the foundation for a plaintiff's expert's opinion. Expert opinion on specific, individual causation was inferred from the evidence on general causation together with clinical observations and a strong temporal association; the child's death ensued less than four hours after vaccination. The dissenting opinion in this case is preferred over majority rulings in other recent cases that distinguish between legal and scientific standards of certainty, such as Ferebee and Rubanick. Judge Newman's arguments are more comprehensive and they place toxic risk assessment used in civil cases under a substantially reduced level of scientific certainty as compared to that expected in the scientific community or used in criminal prosecutions. [Roth‑Nelson]
Poster 21b
JUDICIAL BIAS
*In a recent article, Richard Revesz argues with empirical data that in reviewing challenges to agency environmental rule makings in the D.C. Court of Appeals, a judge's political affiliation before taking the bench appears to affect that judge's resultant decision‑making, at least in considering challenges to the agency's adherence to process requirements. Revesz also forwards specific recommendations for curbing the extent to which ideological influences can impact appellate court judge's decision‑making. Richard L. Revesz Environmental Regulation, Ideology, and the D.C. Circuit Virginia L. Rev. 1717, 1717 (1997). [Wagner]
Do you have any additions or
comments?
E-mail us at wagner9@attglobal.net
Copyright © 2000.
John Applegate and Wendy Wagner.
Last modified September 28, 2000