In re GNC Corp., -- F.3d --, 2015 WL 3798174 (4th Cir. June 19, 2015)
Well, this is a terrible opinion that makes no sense. In the Fourth Circuit, if reasonable experts disagree, it’s now impossible for one to be wrong, and therefore impossible for there to be literal falsity. No more preponderance of the evidence. Perhaps worse, this standard is stated as being the standard for the Lanham Act in a state law consumer protection case, with citation of but no apparent comprehension of the difference between literal falsity and literal truth that is nonetheless misleading. Of course class action jurisprudence of late represents a general hostility to consumer protection, but that’s mostly done procedurally—this is a deep substantive wound.
Appellants bought joint health supplements produced and sold by GNC and Rite Aid. The accused products all contain glucosamine and chondroitin, and most contain additional purportedly active ingredients. Appellants alleged violations of various state consumer protection laws because, they alleged, many scientific studies have shown that glucosamine and chondroitin are no more effective than a placebo in treating the symptoms of osteoarthritis, contradicting the labels’ health claims. Those claims include that various products will, e.g.: “promote[ ] joint mobility & flexibility,” “protect[ ] joints from wear and tear of exercise,” “rebuild[ ] cartilage and lubricate[ ] joints,” and provide “[m]aximum strength joint comfort.” The product label for TriFlex Fast–Acting represents that the product was “[c]linically studied” by means of a “12–week multi-center, randomized, double-blind, placebo controlled study of 60 adults ... taking 250 mg/day of the GNC TriFlex Fast–Acting Blend” and was “shown to improve joint comfort and function,” and also promises 20% improvement in joint function and 25-30% improvement in joint flexibility.
The court of appeals affirmed the dismissal of the complaint because “marketing statements that accurately describe the findings of duly qualified and reasonable scientific experts are not literally false.”
I’m going to go Scalia one better and insert some reaction gifs here.
Plaintiffs alleged that “the vast weight of competent and reliable scientific evidence” indicated that glucosamine and chondroitin do not provide the promised health benefits. They cited a number of peer-reviewed studies to this effect, which experts in the field allegedly deemed to be appropriate proxies for users. However, the complaint didn’t contain any allegations about the ineffectiveness of herbal compounds found in some of the products.
“[T]he district court held that a manufacturer cannot be liable for false advertising so long as at least one qualified expert opines that the representations made are truthful, even if the overwhelming weight of scientific evidence is to the contrary.”
Procedural wrangling omitted.
The court of appeals found the district court’s ruling accurate “while incorrect in its specific formulation.” The complaint’s reference to specific studies didn’t matter, since a plaintiff doesn’t need to “forecast” evidence sufficient to prove its case. Studies cited in the complaint may be statements of fact that make a claim plausible, but don’t comprise the full body of evidence that would ultimately be coinsulted. “[T]he question of falsity hinges on the existence (or not) of scientific consensus and not on the conclusions that hypothetical scientists might draw from those studies referenced in the [complaint].”
Though the Lanham Act didn’t control here, Lanham Act precedent was “instructive” in construing the state laws here. (This is not so, for a variety of historical and drafting reasons; state UDAP/UCL acts have different purposes—specifically and most notably consumer protection directly, with consumer standing to sue and often to receive statutory/enhanced damages—and state consumer protection cases have essentially never used the Lanham Act explicit/implicit falsity division except when cases are brought by competitors asserting both state and federal claims and the parties don’t dispute that the state test ought to be the same for competitor plaintiffs.) Lanham Act false advertising cases divide into literally false claims and literally true but misleading claims, the latter of which require extrinsic evidence of actual consumer confusion. So too here. “Every statute at issue here imposes liability for misleading representations. Although each state supplies its own test for determining whether true statements are misleading, statements that are literally false are necessarily misleading within the meaning of all of the statutes.” Plaintiffs consistently argued literal falsity, not misleadingness.
Plaintiffs’ theory of the case was a syllogism: “(i) the Companies represent that the products improve joint health; (ii) scientific evidence will show that glucosamine and chondroitin do not improve joint health more than a placebo would; (iii) therefore, the representations must be false because the products do not and cannot improve joint health.” But that’s not enough, because the complaint failed to allege that “all scientists agree that glucosamine and chondroitin are ineffective at providing the promised joint health benefits” (emphasis added). Instead, it alleged that the “vast weight” and the “overwhelming weight” of the evidence favored plaintiffs. This necessarily concedes that some reasonable experts believe that glucosamine and chondroitin can provide the promised symptom relief. This means that the scientific evidence of efficacy is equivocal. And that means that plaintiffs failed to allege literal falsity.
Plaintiffs said, quite reasonably, that the court shouldn’t resolve a battle of the experts on the pleadings. But the court rejoined that it didn’t have to: having determined that there was a battle, the court was confident that the complaint therefore failed to plead literal falsity:
When litigants concede that some reasonable and duly qualified scientific experts agree with a scientific proposition, they cannot also argue that the proposition is “literally false.” Either the experts supporting the Companies are unreasonable and unqualified (in which case, there is no real battle of the experts to begin with) or they reflect a reasonable difference of scientific opinion (in which case the challenged representations cannot be said to be literally false).
False advertising cases could still survive if plaintiffs can, after due diligence, plead that no reasonable scientist would agree with the challenged representations. “Moreover, plaintiffs remain protected from dubious experts by the Federal Rules of Evidence, which ‘ensure that any and all scientific testimony ... is not only relevant, but reliable.’”
And the court promised that it wasn’t insulating supplement makers from liability. “A manufacturer may not hold out the opinion of a minority of scientists as if it reflected broad scientific consensus.” (But what does it mean to “hold out the opinion of a minority of scientists as if it reflected broad scientific consensus”? Apparently it can’t be false to state that minority’s opinion as if it were established fact.) Anyway, the court need not decide whether the claims could be misleading, because plaintiffs never pled misleadingness.
“In sum, we hold that in order to state a false advertising claim on a theory that representations have been proven to be false, plaintiffs must allege that all reasonable experts in the field agree that the representations are false.”
[That's not what "proof" means.] Also, the complaint failed to allege that the other herbal ingredients, present in almost all the products, didn’t work. This failure was independently fatal, since most of the challenged representations referred to the products as a whole.
The court did reject some grounds for affirmance: defendants argued that the studies cited in the complaint weren’t specific enough to the challenged products to raise any plausible inferences about their efficacy. This argument was “unpersuasive and inconsistent with notice pleading.” Defendants argued that the studies only looked at people with arthritis, not people who only experience joint pain and stiffness, but the court had to take as true the complaint’s allegation that experts in the field deemed such studies to be appropriate proxies for non-arthritic people as well. “It may well be that glucosamine and chondroitin work differently in people with arthritis than in people without arthritis, but such a factual dispute is not susceptible to resolution at the motion-to-dismiss stage.”
Plus, the defendants argued that the scientific studies weren’t sufficient evidence of falsity because they didn’t assess “the specific formulations used in the products or the synergistic effects between the products’ ingredients.” But that wasn’t an appropriate inquiry at the motion to dismiss stage.
Okay, deep breaths.
The court of appeals made a category error: reasonability and correctness are not the same, and deliberately so in most of the legal system. We recognize that reasonable and scientifically accepted methods are sometimes applied wrongly, or sometimes simply produce wrong results—that’s what it means to have an error rate. Having at least one admissible expert opinion in your defense means that you should be able to fight the charge of falsity in front of a jury, not that you should be able to get the claim dismissed before anyone evaluates the competing expert opinions. If a reasonable expert is nonetheless wrong in a particular case, the claim that the expert supports is not misleading, it is untrue, that is, false.
Compare: based on intelligence that meets ordinary standards for action, a drone strike hits a building that turns out not to have the target inside it. Was the statement “our target is inside this building” merely misleading, or was it false? If you don’t like “false” because of its implication of intent, you’ve already lost that battle in the consumer protection context, but would you accept “that statement was not true, though we reasonably thought it was at the time we said it”? By contrast, misleadingness relies on inference to complete or extend a claim that is true on its face. Suppose the claim on the supplement at issue here was “Tired? Iron can remedy anemia that causes tiredness.” Surely most if not all reasonable scientists would agree that this is true. Now suppose that most tiredness does not come from anemia. Is the claim “iron can remedy anemia that causes tiredness” misleading on a product marketed to consumers in general? Of course it is, because of implicature, specifically the rule of relevance: connecting it with the nonfactual question “tired?” implies that iron will treat tiredness in general.
The difference between falsity and misleadingness is not whether reasonable experts can disagree—in the drone/falsity example, they might have disagreed, given how the relevant evidence is interpreted, whereas in the iron example, they are unlikely to disagree at all. The difference is whether the statement requires consumers who accept that the statement is true to make further inferential leaps to reach the conclusion that benefits the advertiser. In some cases, those inferential leaps may be so likely that we say that a claim is false by necessary implication. (Relatedly, the court of appeals’ new standard makes an unintelligible hash of falsity by necessary implication, recognized in every circuit to consider it. Under the new standard, if lots of experts disagree with a challenged claim but one agrees with it, is that falsity by necessary implication? Is it a percentage standard now? This is one reason that (a) Lanham Act cases shouldn’t have a rigid falsity/misleadingness divide, and (b) one ought to be careful before importing Lanham Act doctrines into consumer protection cases, given the different considerations regularly at issue.)
At a minimum, this holding on literal falsity puts the 4th Circuit in conflict with everyone else’s definition of literal falsity, where juries are routinely allowed to resolve battles of the experts in finding literal falsity. E.g., Hot Wax, Inc. v. S/S Car Care, 1999 WL 966094 (N.D. Ill. 1999) (where a product containing mineral seal oil was sold as “wax” for cars, a jury found the claim literally false, rejecting several experts’ testimony that “wax” could be defined broadly, in favor of the plaintiff’s expert testimony that mineral seal oil was plainly not “wax” because it lacked the basic characteristics of wax); Eastman Chemical Co. v. Plastipure, Inc., 775 F.3d 230 (5th Cir. 2014) (taking one side in a scientific debate can be literally false when the ads are directed to the general public, not to experts to whom the details and limitations of a study are disclosed); Kraft General Foods, Inc. v. Del Monte Corp., 28 U.S.P.Q.2d 1457 (S.D.N.Y. 1993) (featuring competing expert definitions of the word “gelatin”). Literal falsity should be determined by a two step process: (1) what unambiguous message does the ad convey, if any? (2) is that unambiguous message false? E.g. United Indus. Corp. v. Clorox Co., 140 F.3d 1175 (8th Cir. 1998) (ambiguity is the key divide between explicit and implicit falsity). That has nothing to do with the presence of scientific debate.
Indeed, the very case that the GNC court here cites to explain literal falsity holds exactly that:
Castrol presented affirmative evidence to prove the literal falsity of Pennzoil’s claims and … [the district judge] did not find Pennzoil’s evidence to rebut Castrol’s proof persuasive. The dissent asserts, however, that a defendant need only establish a reasonable basis to support its claims to render the advertisement literally true. We disagree. Rather, the test for literal falsity is simpler; if a defendant’s claim is untrue, it must be deemed literally false.
Castrol v. Pennzoil Co. 987 F.2d 939, 944 (3rd Cir.1993). GNC cites Castrol to define literal falsity, but neglects its explanation of what literal falsity is; Design Resources, Inc. v. Leather Industries of America, -- F.3d ----, *7 (4th Cir. June 18, 2015), also relies on Castrol for falsity by necessary implication.
Other consumer protection cases as well have easily recognized that falsity means wrongness, not global consensus on wrongness. For a sampling of cases where a consumer plaintiff’s allegations sufficed by alleging falsity, not total scientific consensus: Zakaria v. Gerber Products Co., No. 15–00200, 2015 WL 3827654 (C.D. Cal. June 18, 2015) (affirmative misrepresentation is actionable when its falsity is properly alleged, as opposed to lack of substantiation for the claim); Bezdek v. Vibram USA Inc., 2013 WL 639145, No. 12–10513 (D. Mass. Feb. 20, 2013); In re Clorox Consumer Litigation, 894 F. Supp. 2d 1224 (N.D. Cal. 2012); Cardenas v. NBTY, Inc., 870 F. Supp. 2d 984 (E.D. Cal. 2012); Prohias v. Pfizer, Inc., 490 F. Supp. 2d 1228 (S.D. Fla. 2007).
Oh, and—what exactly would the extrinsic evidence of confusion be directed to? The logic of the ruling implies that it would have to be “consumers are confused into thinking there’s a scientific consensus about this claim,” but that doesn’t make sense. (Consider what the court of appeals could want the survey to show: the reception of the explicit messages on the packages, that the supplements improve joint health etc.? But that’s just a reading test: those are the explicit messages on the packages. The reason that explicitly false claims don’t require further evidence of consumer reception is that it’s pointless and wasteful to require a survey to see if consumers can read.) The thing that consumers are deceived by, and likely to think about, is the underlying truth of the claim, not the degree of consensus about it. Those two perceptions are probably reasonably close together to the extent that consumers consider consensus at all, but consensus is a proxy for the fact of interest: truth or falsity of the health claim. The only way this holding can be coherent (if still wrong) is if it is no longer false or misleading to make any claim against which there is no “consensus,” no matter what message consumers receive.
To put it another way: under this rule, what is the remaining place of a battle of the experts in a false advertising case? We regularly expect that some experts are wrong and some are right. But that’s a matter of evaluating the evidence presented by the experts. Evidence of consumer perception, collected via a survey, would not help us evaluate the evidence presented by the experts in any way. But it’s the survey requirement that divides explicit from implicit falsity. Is the court's idea that we should do a pointless survey so we can then have the actual battle of the experts?