This post is from the non-Dechert and non-RS side of the Blog.
Depending on the time, issue, and players, the supposed epithets of “judicial activism” or “activist judge” can be thrown in just about every juridical direction. If we were to try to parse out the most common reason for the use of these terms, we might identify a judge or panel of judges establishing law that should have resulted from legislation. There is also the variant where the judge or judges establish new law that really should come from another court, such as the highest court of the relevant state. While we recognize that many decisions from courts could be said to “establish new law” whenever a non-rote issue is being decided, there is a difference between interpreting existing law in connection with deciding a live issue in a particular case and establishing new law. Under the principle of stare decisis, controlling precedent is to be respected with logical applications of the precedent to novel sets of facts, which can lead to “extensions” or “corollaries” of existing law. (Obviously, stare decisis does not prevent the Supreme Court or state supreme courts from reversing decades of established law, like Brown v. Board of Ed. undoing Plessy v. Ferguson or Dobbs undoing Roe.) When a federal court sits in diversity and has to utilize Erie restraint in applying state law, these dynamics become more than mere semantics.
In a federal diversity case where the highest court of the state whose law is being applied has not ruled on the specific issue in the case, in which direction does restraint pull? As lawyers defending companies that get sued, our first thought on this is that federal courts sitting in diversity should not establish new state law causes of actions. We concede that they also should not establish new state law defenses to causes of action. In many cases, however, the fight is about whether a broad or vaguely defined state law cause of action encompasses the relief plaintiff seeks for the defendant’s allegedly culpable conduct. For instance, a state’s recognition of a cause of action for negligence does not cover conduct as to which the state imposes no duties. Reporting to a federal agency, such as FDA, would be an example. Also, if the state’s product liability law has recognized distinct theories for strict liability failure to warn, design defect, and manufacturing defect, then making up a hybrid theory to fit the plaintiff’s needs in the particular case would be an expansion.
In City of Huntington v. AmerisourceBergen Drug Corp., 609 F. Supp. 3d 408 (S.D. W. Va. 2022), the district court presented its Findings of Fact and Conclusions of Law after a long bench trial. The trial court found that plaintiffs had failed to carry their burdens on a number of the elements of their public nuisance claims asserted against the three main U.S. prescription drug distributors related to the purported connection between their distribution of prescription opioids and the local impacts of abuse of prescription opioids and various illegal drugs of abuse. This defense win in a marquee opioid litigation trial—the plaintiff jurisdictions were considered some of the strongest plaintiffs because of the volume of prescriptions dispensed in them compared to their populations and the extent of the local problem—garnered quite a bit of attention. Predictably, it also garnered an appeal. Last week, the Fourth Circuit reversed. City of Huntington v. AmerisourceBergen Drug Corp., — F. 4th –, 2025 WL 3009526 (4th Cir. Oct. 28, 2025). Our introduction was not subtle in signaling that we think one of these courts got it wrong in large part because of not following Erie.
We have made no secret of our view that public nuisance should not be extended to the sale or distribution of products, especially prescription medical products. We have been saying this since long before the Restatement (Third) of Torts: Liability for Economic Harm § 8 cmt. g (A.L.I. 2020), stated that “the common law of public nuisance is an inapt vehicle for addressing” alleged harms related to products. We also think the municipal cost recovery rule is a problem for the opioid litigation claims, as they are premised on shifting the increased costs of government functions to purported tortfeasors/lawbreakers. Part of the issue with the proposed expansions of public nuisance is that the tort of nuisance—since its roots in thirteenth century England—has always been focused on the use or enjoyment of real property. As private nuisance expanded to public nuisance, both through common law and statute, the connection to real property has become somewhat more opaque but is still present. The remedy has always been abatement, an equitable remedy, not monetary damages for past injury. Even the term “abatement” derives from the term for cutting down a tree, which definitely relates to property. Before opioid litigation pushed the limits of public nuisance to try to encompass liability for marketing, selling, and distributing products, perhaps the most famous litigation seeking to impose broad liability for a purported public nuisance was the Rhode Island lead paint litigation. While that had a much closer link to real property rights—the allegedly harmful paint was on the walls of residences where children lived and consumed paint chips—it still fell in the Rhode Island Supreme Court because there was an insufficient connection between the defendants’ sale of their products and the creation of a dangerous condition at a specific location that affected an indivisible public resource. As the Rhode Island Supreme Court said there, “Expanding the definition of public right based on the allegations in the complaint would be antithetical to the common law and would lead to a widespread expansion of public nuisance law that never was intended[.]” State v. Lead Indus. Ass’n, 951 A.2d 428, 453 (R.I. 2008). By any measure, using public nuisance as a vehicle for governmental entities to obtain damages related to the marketing, distribution, or sale of a product would be an expansion of public nuisance, not a failure to restrict it.
This brings us back to the issue of Erie restraint. The Huntington district court, at the start of its analysis of West Virginia law on public nuisance, wrote:
This court should “‘respond conservatively when asked to discern governing principles of state law’ and take care to avoid interpreting that law in a manner that ‘has not been approved’” by the West Virginia Supreme Court of Appeals.
609 F. Supp. 3d at 472. This sentence cited Knibbs v. Momphard, 30 F.4th 200, 213 (4th Cir. 2002), and its own quotation of Rhodes v. E.I. du Pont de Nemours & Co., 636 F.3d 88, 96 (4th Cir. 2011). The Fourth Circuit echoed the same exact Knibbs and Rhodes citations after the following sentence, but without any hint of the need for restraint or a conservative response: “Because there is an absence of controlling state law, we are charged with predicting what the State Supreme Court would conclude based on that state’s existing law.” 2025 WL 3009526, *10. The actual cited/quoted language from Knibbs and Rhodes (in order below without citation) supports the district court’s approach:
We “respond conservatively when asked to discern governing principles of state law” and take care to avoid interpreting that law in a manner that “has not been approved” by the Supreme Court of North Carolina.
* * *
In the absence of such action by the highest state court in West Virginia, our role in the exercise of our diversity jurisdiction is limited. A federal court acting under its diversity jurisdiction should respond conservatively when asked to discern governing principles of state law. See Day Zimmermann, Inc. v. Challoner, 423 U.S. 3, 4, 96 S. Ct. 167, 46 L. Ed. 2d 3 (1975) (per curiam). Therefore, in a diversity case, a federal court should not interpret state law in a manner that may appear desirable to the federal court, but has not been approved by the state whose law is at issue. See id. Mindful of this principle, we decline the plaintiffs’ invitation to predict that the West Virginia Supreme Court of Appeals would adopt the specific provisions of the Restatement advanced by the plaintiffs.
While both the Huntington district and appellate court recognized that the West Virginia Supreme Court had not ruled on the direct issue of whether West Virginia public nuisance law covered damages (not just traditional abatement) for harms allegedly caused by the distributors’ distribution of prescription opioids to pharmacies in the plaintiff jurisdictions, they came down on opposite sides of these issues. It is clear to us that the district court applied the proper restraint against expansion counseled by the Supreme Court and prior Fourth Circuit decisions—both as to the scope of conduct and the scope of remedies. The Fourth Circuit, which repeatedly framed its approach as refusing to restrict West Virginia’s common law on public nuisance, did not. The rest of the decision reversing and remanding the case for further proceedings was essentially a foregone conclusion.
Mind you, the Fourth Circuit had done the right thing by referring a certified question to the West Virginia Supreme Court and was put in an unusual, awkward position when the West Virginia Supreme Court denied its request on the grounds that it could not give an advisory opinion. That meant that an Erie prediction—with proper restraint, hopefully—was back on the Fourth Circuit’s plate. Of course, had it answered the question, the West Virginia Supreme Court very well could have expanded public nuisance to the extent needed to help out the case brought by a West Virginia county and city against out-of-state companies. After all, this is the same court that adopted medical monitoring without present injury in Bower v. Westinghouse Electric Co., 522 S.E.2d 424, 427 (1999), and rejected the learned intermediary doctrine for prescription drug product liability cases State ex rel. Johnson & Johnson v. Karl, 647 S.E.2d 899 (W. Va. 2007). The political pressure to provide such novel remedies in Huntington would surely have been substantial. But the Fourth Circuit, in its role on a diversity case, should not have “interpret[ed] state law in a manner that may appear desirable to the federal court, but has not been approved by the state whose law is at issue,” which was clearly the case here.
We could walk through the rest of the Fourth Circuit’s decision to highlight instances of questionable conclusions and suspect analysis throughout. Surely, others will, including on the likely cert. petition to the Supreme Court. Rather than do that, we will focus briefly on two interrelated recurring issues.
First, the district court, after recounting lots of evidence and many findings from bench trial, went ahead and addressed whether plaintiffs carried their burden as to the elements of public nuisance, assuming it encompassed products and monetary damages. The district court’s decisions, as factfinder, established layers of alternative grounds for affirming even if the Fourth Circuit disagreed on the scope of public nuisance. It was obvious from the opinion below that each factual determination that the Fourth Circuit revisited was supported by substantial evidence in the record.
Second, although questions of fact are supposed to be reviewed on a clear error standard, 2025 WL 3009526, *9 (“[T]his is a deferential standard, and we will not disturb the district court’s factual findings if they are plausible in light of the record reviewed in its entirety.”) (citation omitted), the Fourth Circuit appeared to undertake a de novo review of some of the district court’s many factual determinations. The Fourth Circuit held that a number of these factual determinations, such as the existence of multiple, fatal holes in plaintiff’s proximate cause case, “were based, at least in part” or “rested, in part” on purported legal errors in interpreting the distributors’ duties under the federal Controlled Substances Act. Even if the district court was wrong on that issue—it relied on plaintiffs’ witnesses’ testimony on those federal duties—this holding was quite a stretch. The Fourth Circuit did not identify improperly excluded evidence, hold that any factual determinations were not supported by substantial evidence in the record or were contrary to the weight of the evidence, or explain how there could be proximate cause given the many undisturbed factual determination on causation. For instance, this finding from the district court seems untouched by the Fourth Circuit’s criticisms and negates causation under the causation theory it resurrected:
Plaintiffs rely on a breach of a no-shipping duty to prove diversion and creation of an opioid epidemic. To the extent they also rely on the reporting requirement, plaintiffs failed to show that any alleged violations based upon a failure to report suspicious orders by defendants contributed to the volume of opioids distributed in Cabell/Huntington. Put another way, plaintiffs did not show that had defendants reported more suspicious orders that the DEA would have closed any of the pharmacies that defendants serviced in Cabell/Huntington.
609 F. Supp. 3d at 449 n.5. There are more examples, but we will leave it there.
Whether this goes to the Supreme Court or goes to the district court to redo its Findings of Fact and Conclusions of Law in light of the Fourth Circuit’s expansion of West Virginia public nuisance law, it is still far from a foregone conclusion that plaintiffs will win this case with the obvious holes in the evidence they offered at trial. As John Adams famously said, and some plaintiff lawyers like to quote:
Facts are stubborn things; and whatever may be our wishes, our inclinations, or the dictates of our passions, they cannot alter the state of facts and evidence.