The first appellate shoe has dropped in the litigation involving the herbicide Roundup, Johnson v. Monsanto Co., decided July 20, 2020, by California’s 1st District Court of Appeal, Division One. We discussed the verdict and the trial court’s post-trial rulings here, and we now follow through with an update.

Initially, the price tag for allowing questionable science into the courtroom, as measured by this verdict, has been reduced. The court of appeal lowered the compensatory damages award from $39 million to about $10.25 million, concluding the jury had improperly awarded noneconomic damages that plaintiff would likely never suffer. Because plaintiff’s counsel had argued to the jury that plaintiff’s Non-Hodgkins Lymphoma had reduced his future life expectancy to two years, the jury could not award pain and suffering damages beyond that two-year span. And, agreeing with the trial court that constitutional limits required a 1:1 ratio between compensatory and punitive damages, the court slashed the $78 million punitive award to about $10.5 million.

But as we observed previously, the amount of the judgment does not begin to measure the cost. Allowing (1) liability based on a risk most or much of the scientific community doubts exists and (2) punitive damages based on a company’s failure to include a warning inconsistent with most of the scientific studies and the conclusions of most regulators inflicts incalculable costs on the civil justice system.

After a great deal of study over many years, there is no scientific consensus that glyphosate, the active ingredient in Roundup, increases the risk of NHL. IARC (International Agency for Research on Cancer) determined in 2015 that glyphosate is a “probable human carcinogen” based on “sufficient evidence” that it caused cancer in experimental animals and “limited evidence” that it does so in humans.

But hazard classification bodies, including the U.S. Environmental Protection Agency, the World Health Organization and the European Chemicals Agency have concluded that the scientific evidence does not support labeling glyphosate a human carcinogen. The EPA reaffirmed that conclusion last year. And just last month the Eastern District of California barred California from requiring a warning label that glyphosate is carcinogenic under its Proposition 65 because the scientific and regulatory evidence proves IARC an outlier on this issue. Consequently, the statutory warning that the ingredient is ”known to cause cancer” would be “misleading” and therefore not advance the state’s interest in consumer protection, violating the First Amendment. National Ass’n of Wheat Growers v. Becerra, 2020 WL 3412732 (E.D. Cal. 6/22/20).

Back to the Johnson appeal. Just a month after California was barred from requiring a cancer warning on Roundup, the court held that a California jury was justified in finding that Monsanto could be liable for failing to warn of the cancer risk because that risk was “known or knowable in light of the scientific and medical knowledge generally accepted in the scientific community” in June 2012 and thereafter.

How did the court deal with the prevailing scientific uncertainty and debate? It adopted an overly broad definition of the qualitative evidence of risk needed to require a warning. The duty under strict liability is to warn of “potential risks,” which the court defined as risks “existing in possibility” or “capable of development into actuality,” based on language in a case dealing with a different issue. Valentine v. Baxter Healthcare Corp., 68 Cal.App.4th 1467 (1999) (mentioning the dictionary definition of “potential” in holding that a verdict of no strict liability for failure to warn of a risk precluded a claim for negligent failure to warn).

The court’s interpretation of the duty to warn is ill-advised and should be rejected by California courts (the liability issues portion of the decision is unpublished, and therefore non-precedential). Any theoretical, albeit unproven, risk can be considered a “possibility” under a broad (bordering on the metaphysical) interpretation of the term. The better reading would align with the “capable of development into actuality” definition, which is more congruent with the cognate toxic tort standard for “general causation”: generally accepted in the scientific community as capable of causing the condition in the human population. General causation leads into evaluation of specific causation, i.e., actual cause of the condition in the individual. So defined, toxic tort and products liability law would be more symmetrical and provide greater protection against liability burdens based on scientific certainty, while still allowing recovery for risks recognizable by the prevailing scientific consensus.

The court also sustained the jury’s finding that Roundup was defective in design, based on the consumer expectations test, a design defect standard discussed here and here. The issue was the applicability of this “standardless” test rather than the “risk-benefit” test of design defect. The former applies only where the risk of the design that produced the injury is within the common everyday experience of jurors, so that they can assess whether the design is defective without the aid of expert testimony educating them about the design’s risks and benefits.

It is difficult to see how the jury can be expected to bring to the courtroom a solid foundational understanding from which to assess risks of a design that has been and remains the subject of substantial ongoing debate in the scientific and regulatory communities. Determination of whether a particular substance poses a risk of causing cancer ordinarily entails evaluating epidemiological studies, animal studies and other toxicological evidence such as laboratory assays of genotoxicity and mutagenicity, meta-analyses, dose-response relationships and biologically plausible mechanisms of action, among other principles and evidence. That knowledge and process seems quite foreign to the everyday experience of jurors, and therefore outside the purview of the consumer expectations test. The court’s effort to distinguish other product liability cases – opinions holding that products requiring an understanding of complex biological processes to evaluate their risks must be evaluated based on expert testimony and the risk-benefit test – is unpersuasive. It threatens to muddy further the already murky consumer expectations test issue if adopted, though this aspect of the decision also is unpublished.

The presently unpublished nature of the court’s analysis of the failure-to-warn and design defect issues offers hope that these approaches will remain aberrations that do not contaminate citable California law. The published portion of the opinion, however, upholding punitive damages liability based on the same uncertain science, may prove more troublesome if it avoids depublication and further review.