The CEO's Complaint In April 2026, Bayer CEO Bill Anderson stood before shareholders and made an argument that has become familiar in corporate boardrooms. Bayer had spent decades and billions developing products that undergo serious regulatory review. And yet, Anderson asked investors, why continue that work when it leaves the company “at the mercy of a 600-billion-dollar litigation industry”? The implication was clear. Litigation undermines the regulatory process. It second-guesses the scientists. It makes innovation irrational. Anderson was echoing an argument the defense bar has developed systematically for decades. John H. Beisner of Skadden Arps, in a series of reports for the Institute for Legal Reform , has argued that MDL proceedings pressure defendants to settle without examining the merits of individual claims. The Manhattan Institute’s James R. Copland has framed mass tort litigation as an economic drag on innovation . These positions represent the institutional consensus of the defense bar. And they rest on a factual premise that the evidentiary record contradicts. What the Record Shows Anderson was not speaking in the abstract. He was the CEO of the company that acquired Monsanto and inherited the Roundup litigation. That litigation has cost Bayer more than $11 billion in settlements and verdicts , with a further $7.25 billion proposed class settlement announced in February 2026 and granted preliminary court approval in March. Three days after Anderson’s remarks, Bayer’s attorneys stood before the United States Supreme Court in Monsanto Co. v. Durnell to argue that federal regulatory approval of glyphosate should preempt the state-law failure-to-warn claims that produced that liability. The company was not merely complaining about litigation. It was asking the Court to shut down the legal mechanism that had exposed what was in its own files. What was in the files is telling. In late 2025, the journal Regulatory Toxicology and Pharmacology retracted a twenty-five-year-old paper that had been cited for decades as evidence that glyphosate, the active ingredient in Roundup, was safe. The paper, Williams et al. (2000) , had concluded that glyphosate posed no carcinogenic risk to humans. Regulators relied on it. Defendants cited it in proceedings around the world. It shaped the scientific consensus for a generation. The retraction did not occur because new science emerged. It occurred because multidistrict litigation discovery exposed what peer review could not. The paper was ghostwritten . Internal Monsanto documents produced in the Roundup MDL (In re: Roundup Products Liability Litigation, MDL No. 2741, N.D. Cal.) revealed that company scientists had drafted sections of the paper , managed the editorial process, and selected the nominally independent authors whose names appeared on it. A 2015 internal email from Monsanto scientist William Heydens discussed “how we handled Williams, Kroes and Munro,” referring to the company’s orchestration of the very research that regulators treated as independent science . The “rigorous approval process” Anderson invoked was built on a scientific record his company had manipulated. The regulatory system did not fail because litigation interfered with it. It failed because, without litigation, no one had the tools to discover the interference that was already there. The Pattern Monsanto is not the only company whose internal record told a different story than its public one. The pattern recurs across every successful mass tort of the past three decades. Litigation discovery exposes information that no other institution had the tools or incentive to uncover. Johnson & Johnson’s internal documents, produced through discovery in the talc cancer litigation, revealed that the company had known about asbestos contamination in its Baby Powder since the 1970s . Internal testing detected asbestos fibers . Strategic decisions followed about how to manage the information rather than the contamination. The UCSF Industry Documents Library has catalogued approximately 3,500 of these internal J&J documents. They had been inside J&J’s files for half a century. They emerged only because the litigation process compelled their production. In December 2025, a Baltimore jury returned a $1.5 billion verdict against J&J for a woman who developed mesothelioma after using its talc products, a case built on that same documentary record. It was the largest verdict ever awarded to an individual talc plaintiff. Then, in March 2026, The Lancet retracted a 1977 commentary that J&J had cited for decades to defend the safety of cosmetic talc because the author, Francis J.C. Roe, was an undisclosed paid J&J consultant who had shared drafts with the company and revised the paper based on its feedback. What is striking about these cases is not simply that the defendants knew more than they disclosed, or that regulators failed to detect the problem. It is that the system lacked the capacity to respond. The information gap was structural. Regulatory agencies lacked the subpoena power, the adversarial incentive, and in many cases the resources to obtain what litigation discovery produced. As I discussed in the third article in this series, Mass Torts as a Complement to, and Backstop for, Government Regulation , the systemic case for mass torts rests in part on the proposition that private enforcement supplements public regulation. The evidentiary record that MDL discovery produces is how that supplementation operates in practice. It is the foundation on which accountability depends. From Evidence to Accountability As any trial lawyer will tell you, producing documents is not the same as establishing legal facts. The MDL system includes two processes that convert raw discovery into usable evidence: Daubert proceedings and bellwether trials. Both serve filtering and calibrating functions that determine whether the information discovery produces can be translated into accountability. Daubert serves as a form of adversarial peer review. When general causation is contested, the MDL court evaluates the methodology underlying each side’s expert testimony with a rigor that the scientific peer review process itself often lacks. Daubert is no rubber stamp. The Lipitor MDL is the proof. More than 3,000 women alleged that the statin caused their type 2 diabetes, and the science looked plausible at the headline level. Large observational studies had associated statins with new-onset blood sugar changes, and the FDA had added language to the label. But the plaintiffs’ expert methodologies could not survive scrutiny. The district court excluded them, the Fourth Circuit affirmed in 2018 , and the litigation ended without a dollar changing hands. The pattern repeats. More than 300 Zoloft birth defect claims ended the same way in the Third Circuit . The Mirena MDL ended after a 156-page opinion excluding all seven plaintiff experts, affirmed by the Second Circuit . The Onglyza heart failure MDL ended in the Sixth Circuit in 2024 . Four mass torts, four courts of appeals, zero settlements. The system worked precisely as designed. It filtered claims that could not meet the evidentiary threshold, and the appellate courts confirmed it got the answers right. Anderson’s “600-billion-dollar litigation industry” framing implies an indiscriminate machine. The record demonstrates precisely the opposite. Bellwether trials serve a different function. They calibrate. By trying a representative set of cases to verdict, bellwether trials generate the data that makes rational settlement possible. In the 3M military earplugs MDL , sixteen bellwether trials produced ten plaintiff verdicts and six defense verdicts. That distribution informed the eventual $6 billion settlement. The settlement matrix that allocated recovery across nearly 260,000 claims was built on trial data that differentiated by injury type, severity, and evidentiary strength. The verdicts ran in both directions because the system was measuring, not rubber-stamping. That data could not have been generated any other way. When I evaluate a potential mass tort investment, the first questions are specific. Is general causation supported by methodologies that will survive a Daubert challenge, or does it rely on extrapolations that a well-resourced defense will dismantle? Has the discovery produced internal documents showing the defendant knew, and if so, how specific are they? Is there a bellwether track record, and what does the verdict spread tell you about how juries process the evidence? The quality of the evidentiary record is what separates a case I will fund from a case I will not. That record depends on the discovery apparatus the MDL (and its state-court equivalents) provides. The distinction matters to funders. It should matter to everyone who cares about the quality of outcomes the system produces. What Changed Knowledge Produces Return to Anderson’s complaint. The Bayer CEO framed litigation as a threat to innovation, a system that punishes companies for bringing products to market despite regulatory approval. The framing is powerful because it is partly true. No rational company wants to face billions of dollars in liability after spending billions on development. But the framing collapses when you ask a prior question: what was the quality of the regulatory record that approved the product in the first place? In Roundup’s case, the regulatory record was contaminated by the very company now asking the Supreme Court to immunize it from the consequences. The MDL did not undermine the approval process. It exposed the fact that the approval process had already been undermined, from the inside, by the regulated entity itself. That is not a system run amok. It is a system doing what it was built to do. The Roundup retraction. The EPA PFAS limits. The FDA opioid warnings. The $1.5 billion talc verdict built on documents J&J kept from the public for fifty years . These are the observable consequences of a system that forces information into the open. Information that powerful institutions had every incentive to suppress and that no other mechanism was positioned to extract. I freely admit that the system has real costs and that there are legitimate critiques, which the next article, The Case Against Mass Torts (And What It Gets Right), will address directly. But those costs must be weighed against what the system produces. What it produces is not just verdicts and settlements. It is a changed informational landscape, one in which regulators have better data, markets have better signals, scientific literature is more honest, and the public has access to facts that were previously locked inside corporate filing cabinets. None of this is free. The depositions that produced the Monsanto emails and the J&J memos cost real money. So did the experts, the document review, the years of pretrial proceedings. That investment comes from plaintiffs, their counsel, and increasingly from litigation funders who look at an evidentiary record and make a bet that the truth, once forced into the open, will produce accountability. It is not charity, nor is it pure altruism. But the track record strongly suggests the system is socially beneficial, uncovering corporate wrongdoing that has a concrete effect on people’s lives. And the structure that makes it possible is worth defending. Particularly from those with the most to lose when the record comes to light. That is the proposition that anchors this series. Private enforcement is not an accident of American institutional design. It is how the system was built to work. The MDL’s information function, the adversarial discovery process, and the capital that funds it are the mechanism through which private actors supplement public regulation in practice. Whether that mechanism survives the current moment is the question the remaining articles will take up. Preemption challenges are before the Supreme Court. Tort reform is advancing in state legislatures. Litigation funding is under political attack. ——— W. Tyler Perry is the Director of Mass Tort Strategy at Certum Group, a litigation finance advisory firm. He writes about the institutional architecture of the American civil justice system. The views expressed here are his own.