court rejects survey indicating consumers think "white chips" have white chocolate

Cheslow v. Ghirardelli Chocolate Co., No. 19-cv-07467-PJH, --- F.Supp.3d ----, 2020 WL 4039365 (N.D. Cal. Jul. 17, 2020) 

Plaintiffs brought the usual California claims against Ghirardelli’s “Premium Baking Chips Classic White Chips.” They sought to bolster the plausibility of the complaint by including a consumer survey about whether the labeling misleads consumers into believing that the product contains white chocolate. 1,278 respondents were asked about one of four products: Ghirardelli’s Classic White Chips, Nestle Toll House’s Premier White Morsels, Target’s Market Pantry White Baking Morsels, and Walmart’s White Baking Chips, e.g., “Based on your review of this package, do you think that this product contains white chocolate?” 

According to the results, 91.88% of respondents indicated that they believed the product contained white chocolate. In response to “If, after purchasing this Product, you learned that the Product contained no white chocolate or chocolate of any kind, would you be less or more satisfied with you purchase?” 64.69% of respondents answered that they would either be “much less satisfied” or “somewhat less satisfied.” Similar percentages responded that they would be much or somewhat less likely to purchase the product again. 

Each named plaintiff also alleged with greater specificity than in the first complaint the reasons why they were deceived by the packaging and why they relied on the product’s package. “For example, Cheslow desired white chocolate chips to bake holiday cookies, bars, and brownies and found the product in a section of a Target store labeled ‘chocolate chips.’” Plaintiffs also alleged facts about the history of chocolate and the attributes of white chocolate to support the allegation that chocolate is perceived to be a unique, irreplaceable product and that reasonable consumers do not think they are purchasing a “cheap knock-off pretending to be chocolate.” 

None of this worked.

In dismissing the previous version of the complaint, the court held that “white” in “white chips” was a color reference, so “it would not be appropriate to base liability off of a misunderstanding of that word.” [Ah, those silly consumers—not misled, but just misunderstanding.] Likewise, for the images of baking chips and cookies with chips, “it would be unreasonable to draw a specific qualitative message about a product from an image on that product.” Since there was nothing deceptive on the front of the package, plaintiffs couldn’t reasonably ignore the ingredients label, which didn’t include the words chocolate or cocoa. The placement of the product in the grocery aisle wasn’t alleged to be under defendant’s control, and anyway it wasn’t reasonable to draw any conclusion from its placement. 

The court declined to revisit its holding that “white” definitely means a color, no matter what the context (I assume it would be happy to get a “flat white” that was a smooth cup of milk), and so considered the new allegations in isolation. 

Citing Becerra v. Dr. Pepper/Seven Up, Inc., 945 F.3d 1225 (9th Cir. 2019), which dealt with the meaning of “diet,” Ghirardelli argued that the court should reject the survey. The court agreed. Just as, survey notwithstanding, “a reasonable consumer would still understand ‘diet’ in this context to be a relative claim about the calorie or sugar content of the product,” it was still the case that understanding “white” in “white chips” to mean “white chocolate” was unreasonable. So Becerra posed a legally identical question: “whether a consumer survey can shift the prevailing reasonable understanding that white chips does not include chocolate.” [Where does the “prevailing” understanding come from? Prevailing here means winning, not widespread—reasonableness as courts implement it is a normative concept. But being explicit about that is a bit embarrassing for a system that is supposedly dedicated to consumer protection. My own view about the reasonable consumer construct is that it should be more empirical in cases like this one, as “[a] drunken man is as much entitled to a safe street as a sober one, and much more in need of it,” Robinson v. Pioche, Bayerque & Co., 5 Cal. 460 (1855).]  

The court noted that a survey can bolster a plausible claim but can’t make an implausible claim plausible. [No epistemological modesty here, even on a motion to dismiss! I should be clear: this is a readily available reading of Becerra; the court here isn’t going off on a frolic of its own. That’s just a problem with a standard that says that likely deception has to be alleged but then rejects evidence of actual consumer reaction—the magic happens where the court posits, without saying outright, that a substantial--here, nearly unanimous--majority of consumers are unreasonable in what they think words mean, and therefore marketer exploitation of their “misunderstanding” is acceptable.] 

Anyway, the survey was bad because it only showed respondents the front panel of the packages; consumers aren’t entitled to disregard the ingredient list in the absence of a deceptive front panel.