Twelve Jurors, One Molecule: How Science Was Weighed in the Roundup Litigation
Photo: Gardiner P. Bissel, Oregon Journal, Public domain, via Wikimedia Commons
In 2018, a California jury awarded $289 million to Dewayne Johnson, a former school groundskeeper with non-Hodgkin lymphoma. The verdict opened a decade of litigation in which tens of thousands of plaintiffs alleged that Monsanto's Roundup herbicide caused their cancers. What makes the Roundup litigation a landmark is not merely its scale but its structure: it forced two different systems—federal regulatory science and the adversarial courtroom—to answer the same question, and they answered it differently.
The Classification That Started It
In 2015, the International Agency for Research on Cancer (IARC) classified glyphosate as 'probably carcinogenic to humans' (Group 2A). The designation rested on 'limited evidence' of carcinogenicity in humans and 'sufficient evidence' in experimental animals. IARC's mandate is hazard identification: does a substance have the capacity to cause cancer under some circumstances? It does not weigh dose, exposure route, or real-world risk.
Regulatory agencies approached the same literature from a risk perspective. The EPA's 2017-2020 evaluations concluded that glyphosate was 'not likely to be carcinogenic to humans' at relevant exposures. The European Food Safety Authority reached a similar conclusion. The discrepancy was not primarily about the underlying studies—it was about which studies were weighted, how tumor findings were interpreted, and whether statistical significance was the threshold for action.
The Evidence in the Courtroom
Plaintiffs' expert witnesses built their case on three pillars. First, epidemiology: pooled analyses of the Agricultural Health Study cohort and case-control studies, including some showing elevated odds ratios for lymphohematopoietic cancers among the most highly exposed. Second, mechanistic toxicology: laboratory evidence that glyphosate formulations can induce oxidative stress and genotoxicity in mammalian cells, and that commercial formulations—with their surfactant adjuvants—can be more biologically active than the isolated active ingredient. Third, regulatory context: internal Monsanto correspondence, disclosed in discovery, showing company scientists engaging closely with external researchers whose work downplayed carcinogenic concern.
Monsanto's defense emphasized the weight of evidence: the largest agricultural epidemiological study available found no statistically significant association at registered use rates, decades of regulatory review, and the argument that lymphoma subtypes in plaintiffs did not match the biological signature glyphosate would be expected to produce.
Why Juries and Agencies Diverged
The courts applied a different standard than the agencies. Under the federal rules of evidence, expert testimony must rest on reliable methodology—but the jury weighs plausibility, not proof to a scientific certainty. Plaintiffs needed to show that exposure was a 'substantial factor' in causing cancer, a standard more permissive than the 'more likely than not' causation analysis used in epidemiology.
The adversarial format also mattered. In a regulatory review, industry-commissioned studies sit alongside independent ones in a synthesis. In a courtroom, the discovery process can expose the provenance of every study, and a jury's perception of corporate conduct colors its assessment of the science itself. Several verdicts—including the Johnson decision—reflected both toxicological evidence and institutional distrust.
Judges, meanwhile, split. Federal multidistrict litigation in the Northern District of California excluded some plaintiffs' experts under Daubert while allowing others, and the Ninth Circuit's 2022 ruling in Hardeman v. Monsanto largely upheld the admissibility of causation testimony, finding the evidence sufficient for a reasonable jury.
What the Verdicts Reveal
The Roundup litigation demonstrates that scientific uncertainty is not resolved by more data alone—it is allocated by institutions with different thresholds, different procedural rules, and different conceptions of proof. Agencies ask what is known with confidence at population scale. Juries ask whether it was reasonable for this plaintiff to believe this product caused this disease. Both questions are legitimate. Only one produces a verdict.
For pesticide science, the legacy is twofold: heightened scrutiny of formulation ingredients and adjuvant toxicity, and a durable lesson that when regulatory conclusions diverge from public perception, the courtroom becomes the venue where the gap gets litigated—literally.
