Plaintiff can't win false advertising claim because there's evidence on both sides

Korolshteyn v. Costco Wholesale Corp., No. 15-cv-709, 2017 WL 3622226 (S.D. Cal. Aug. 23, 2017)

Ugh. Costco’s TruNature Gingko labels represent that the product “supports alertness & memory,” that “Gingko biloba can help with mental clarity and memory,” and that “[i]t also helps maintain healthy blood flow to the brain to assist mental clarity and memory, especially occasional mild memory problems associated with aging.” Plaintiff alleged that these were false or misleading under the UCL and CLRA.  Relying on In re GNC Corp., 789 F.3d 505 (4th Cir. 2015), the court here found that this was an impermissible lack-of-substantiation claim, and that therefore a private plaintiff can never prove falsity “when a defendant offers scientific evidence and admissible expert testimony supporting an advertising claim about the efficacy of the product in question.”  To prove falsity, all reasonable experts in the field must agree that the representations are false.  The court found that California would follow GNC, because no state court cases have rejected it, but I doubt that given the state’s tradition of consumer protection and its occasional rejection of the Lanham Act analogy ostensibly underneath GNC (occasional because there’s so much California consumer protection law that the analogy is only rarely made).

Mullins v. Premier Nutrition Corp., 178 F.Supp. 3d 867 (N.D. Cal. 2016), held that a plaintiff could prove claims “literally false if a reasonable jury concludes that all reasonable scientists agree,” or that the claims are “misleading by showing that the vast weight of the competent evidence establishes that those health claims are false.” This is an attempt to draw off some of GNC’s poison, but as we will see it doesn’t work well.  The main reasons that this formulation isn’t a good one are that (1) the Lanham Act’s false/misleading distinction, from which GNC’s consumer protection law reasoning is supposedly drawn, doesn’t have anything to do with “all reasonable scientists” v. “vast weight of the competent evidence,” but rather with issues of consumer perception, and (2) the way scientific evidence works doesn’t generally allow for such a distinction; the most you can say about any given study is that, at various levels of quality, it provides support for an affirmative claim or it doesn’t provide support for that claim.  Mullins said that, for misleadingness, a plaintiff “can concede the existence of scientific studies substantiating a representation, but argue that those studies are poorly designed, incredible, or represent the view of a minority of scientists.” But now we are firmly detached from the idea of consumer perception, and the existence of poorly designed or “incredible” studies shouldn’t be allowed to avoid a finding literal falsity anyway!

It is one thing to say that “[i]nconclusive findings and unsettled science are insufficient to meet Plaintiffs’ burden of raising a question of fact on the issue of falsity,” and quite another to say, as if it were a mere restatement, that “mixed evidence demonstrates at most that the science on [the product’s] effectiveness is inconclusive.”  This is the fallacy of the excluded middle: in the former situation, even crediting all the evidence, the plaintiff couldn’t meet its burden of proof at trial on the merits; in the latter, it might be the case that, depending on which evidence you credited, the plaintiff could meet its burden.  This distinction may seem fine, but it makes a big difference because it is about whether courts will throw up their hands when truth is in question—the very essence of the factfinding project. And that’s not hyperbole on my part, that’s from the courts following GNC: “where there are studies demonstrating both the effectiveness and ineffectiveness of the Products, a reasonable jury could not find that the advertising claims are false.”

So, the court here reasoned, “when a defendant presents scientific studies supporting its advertising claim, a plaintiff must do more than present its own studies that do not support the advertising claim, thereby demonstrating that evidence is equivocal.” Mullins, with which the court here disagreed, held that the plaintiff could offer “principled, supported critiques” of the defendant’s studies, allowing the jury to disregard them or to find the plaintiff’s studies more persuasive and thus find the advertising claims misleading.

The court here also found that the plaintiff was only alleging falsity, not misleadingness, because it thought misleadingness had nothing to do with consumer perception.  Though she used the words “false and misleading,” she was really arguing that the claims were misleading because they were false.  Again, the court didn’t seem to understand misleadingness as something that happened on the consumer end.  In fact, consumers might—and, we know from other studies, in fact do—think that a claim on the label of a supplement has substantial scientific weight behind it, including FDA approval; they are thus misled, quite possibly materially, about the quality of the claim.

“Whether the Label Claims are true or false is a binary choice—they are true, or they are false. When the scientific evidence is equivocal, it is impossible to prove that an advertised claim is either literally true or literally false.”  Comment: This is (ironically?) at best misleading.  Not all cases in which there are factual disputes—even cases where there is enough on both sides to get past summary judgment—are “equivocal.”  Many of them are merely contested, something that courts see easily enough in the ordinary partisan context.  That doesn’t deprive the factfinder of the ability to determine which is the truth, and if it can’t do so, then the burden of proof does the necessary work.  GNC, as I feared, has led to a series of false equivalences, with slippages in meaning each time—from judge-made Lanham Act doctrine to consumer protection law, from false to misleading, from “equivocal” to “equipoise.”

“Essentially, Plaintiff is arguing that the Label Claims could be misleading because a jury could find that Defendants have not proven them to be literally true, which is little more than a ‘lack of substantiation’ claim.” Comment: No, the jury could find that the statement “this product aids memory” is false, because the plaintiffs’ evidence about the studies makes its falsity more likely than not.  If a good study would likely have shown an effect, and a good study exists that didn’t show this effect, that tends to make it more likely that there is no effect—indeed, that’s why you try to disprove a hypothesis.  “Literally false” and “lie” are not the same thing (even if “literally false” was a part of consumer protection law).

“In sum, when a plaintiff presents admissible expert testimony that scientific studies do not support an advertised claim, and a defendant presents admissible expert testimony that scientific studies support the advertised claim, the evidence is equivocal and all reasonable scientists do not agree. No jury conclusion would change either of these facts.” Comment: Note how misleadingness, in the ordinary sense of causing consumers to have false beliefs regardless of literal truth, has become conceptually impossible as well.  Under the court’s reasoning, attempting to prove misleadingness would also cause the impermissible “substantiation” problem—because the fact that a statement deceived consumers about the level of proof would also not change the fact that there was scientific evidence on both sides.

Missing from the court’s reasoning about substantiation is the role of the burden of proof.  A substantiation requirement means that the defendant has the burden of showing that its claim is true.  A falsity or misleadingness requirement means that the plaintiff has the opposite burden; scientific studies showing no effect are one way of meeting that burden.  In fact, they are the best way to do so, though whether they are sufficient in any given case may well be a matter for the jury. 

I think the courts who follow GNC have been confused about the concept of “scientific study” and its intersection with a very different mode of factfinding, the judicial trial. Consider the following claim: X sold the most houses in the county last year.  This might or might not be true; we have various means of getting at the truth; we might even have conflicting evidence (records, after all, may be imperfect; sometimes the best evidence comes from fallible human memory, or human testimony that might require a factfinder’s credibility evaluation).  Could anyone seriously maintain that “X sold the most houses in the county last year” cannot be false if the defendant possesses evidence sufficient to avoid a verdict on summary judgment?  What’s so different about “science”? 

What the GNC line of cases is really saying is that courts will not engage in the very process they’re constituted to engage in if a consumer protection case requires a factfinder, as the court makes clear when it says “under California law a Plaintiff cannot maintain a false advertising claim when the defendant offers admissible expert testimony and scientific evidence supporting the advertisement in question.”  That is, the court will only look at one side of the evidence; I would call that not very judicious.  And by the way, this formulation means that in fact there is a substantiation requirement under California law—just one in which the quality of the substantiation is judged by a minimal standard, that of admissibility. When a rationale for a rule makes the rule a failure on its own terms, there is something deeply wrong.

Because there was competent evidence on both sides, the defendant won on summary judgment.  If you see something wrong with that previous sentence, then you see the problem with GNC.