The American tort system has generated many litigation campaigns and completed verdicts we now recognize as scientific embarrassments. Among them are lawsuits claiming that silicone breast implants caused auto-immune disease, common childhood vaccines caused autism, the morning sickness drug Bendectin caused birth defects, one or another make of car suddenly accelerated without any input from the driver or gas pedal, and so forth.

A San Francisco jury’s August 10 verdict ordering Bayer/​Monsanto to pay $289 million to a school groundskeeper who contracted non-Hodgkin’s lymphoma may add to this list.

Most bodies of expert opinion both private and public around the world, including the U.S. Environmental Protection Agency and National Institutes of Health, do not consider glyphosate, the active ingredient in Roundup, a probable cause of cancer in humans. Here’s one discussion of how a large study (>50,000) looking directly at the health of agricultural workers went far to dispel earlier concerns based on weaker and more indirect studies. Similarly: Steven Novella at Science-Based Medicine after the verdict, and Health Canada, Guy-André Pelouze at Slate and Angela Logomasini at Science20 before.

The outlier exception was a pronouncement by a French-based advisory panel to the World Health Organization, the International Agency for Research on Cancer, classifying glyphosate as a “probable” carcinogen based on animal studies. This was subsequently (and quite controversially) picked up and endorsed by the state of California as an addition to a list of suspected carcinogens. The California courts have upheld that listing, although in June a federal judge told the state that it could not require a warning label on Roundup because the evidence against the compound was too flimsy.

Following the San Francisco verdict, some of the more credulous press accounts cited IARC’s pronouncement without mentioning the many assessments to the contrary. But the problems with that pronouncement go well beyond “unconvincing” or “tenuous” and on into territory some, like Geoffrey Kabat at Forbes, have termed “scandal.”

“A scientist who advised a United Nations agency to classify the world’s most widely used weed-killer as carcinogenic received $160,000 (£121,500) from law firms bringing claims by cancer victims against the manufacturer,” reported London’s Times last fall. “He did not declare his links to the law firms in a letter to the European Commission urging it to accept the IARC classification.” One of the law firms that engaged his services now advertises for claimants to sue over Roundup in U.S. courts.

Meanwhile, the Environmental Working Group, a private outfit known for promoting scare stories, has seized the moment to promote the notion that the exceedingly tiny residues of glyphosate found in oats mean that parents should consider avoiding them as a breakfast cereal. That claim is rebutted by Susan Matthews at Slate (again) and by the knowledgeable food science writer Tamar Haspel.

Eventually, our liability system does often get around to rejecting baseless scientific claims of causation, especially since the improvement in the handling of expert evidence embraced by the U.S. Supreme Court in Daubert v. Merrell Dow(1993). Before it gets there, however, it sometimes redistributes large sums—often to claimants, even more reliably to lawyers—and often destroys large amounts of value. In the days after the San Francisco verdict, the value of Bayer stock dropped by more than 10 billion euros. It’s expensive when error prevails.