Tuesday, September 08, 2020

I can't believe it's not (vegan) butter: First Amendment invalidates standard of identity for butter as applied to vegan product

Miyoko’s Kitchen v. Ross, No. 20-cv-00893-RS (N.D. Cal. Aug. 21, 2020)

Eric Goldman’s far more anti-regulatory take is here.



The California Department of Food and Agriculture told Miyoko’s not to call its “vegan butter” product “butter” (it challenged certain other aspects of the marketing but some of those challenges have been abandoned). Miyoko’s sued and won a ruling that California law—and thus the federal law that it adopts—violates the First Amendment insofar as the standard of identity for butter prevents the use of “vegan butter” as a name for food that clearly signals its nondairy status. “Lactose free” and “cruelty free” also were literally true and could not be regulated on this record. However, Miyoko’s did not receive a preliminary injunction against enforcement that challenged its use of “hormone free” and  “revolutionizing dairy with plants,” both of which the court thought were likely misleading.

“For nearly a century, the standard of identity for butter has required a product ‘made exclusively from milk or cream, or both . . . and containing not less than 80 per centum by weight of milk fat.’”

Both parties agreed that Central Hudson applied. If the speech at issue is false or misleading, it can simply be banned under part 1 of the original CH test, which was the result for “hormone free” and  “revolutionizing dairy with plants.”

First, “because plants contain naturally-occurring hormones, and because Miyoko’s vegan butter is made of plants, it necessarily contains hormones as well. That the claim is literally false places it beyond the bounds of protected commercial speech, fatally undermining Miyoko’s right to use it without fear of reprisal.” Miyoko’s argued that consumers would understand the phrase, in context, to mean that it didn’t contain “the artificial hormones that are sometimes added to animal-based dairy products.” [The last I saw, though there may be new evidence, the artificial hormones given to cattle have not been shown to exist—much less be “added”—n the resulting dairy products.] The court was unpersuaded. “Central Hudson insists, at the threshold, that commercial speech be true, and provides no exception for falsities made true by the target consumer’s supposed contextual awareness. … [N]o court has ever repudiated a regulator’s authority to demand that products claiming to lack hormones actually lack hormones.” It wouldn’t be illogical for any consumer to believe that a product labelled “hormone free” does not contain hormones.

“Lactose free” and “cruelty free” were literally true, but also had to be analyzed in the context of their “quasi-close” proximity to “vegan butter” in case they helped create or maintain deception.  Miyoko’s highlighted the “numerous phrases on its label that point in an unmistakably non-dairy direction”: “vegan,” “made from plants,” “cashew cream fermented with live cultures,” and “cashew & coconut oil spread.” Without more, the state’s evidence of deceptiveness (discussed in greater detail below) wasn’t enough to show deception.

However, “revolutionizing dairy with plants” was false enough to regulate. Using the dictionary definition of “revolutionize,” the phrase “denotes direct interaction with animal-based milk products in a way that leaves them ‘fundamentally’ different than they were before.” But dairy replacements do not revolutionize dairy; they leave it behind. Similarly, you can’t “revolutionize leather with cotton,” or “revolutionize whiskey with seltzer.” This claim was “plainly misleading,” relying purely on dictionary definitions. [Which we all know are always uncontroversial and relevant.]

Miyoko’s, however, succeeded in its core aim: “vegan butter” was not inherently misleading. California pointed to the federal definition of butter as being made “exclusively” from milk and cream, in place since 1923, which the court found “circular”; Miyoko’s pointed to the long history of apple, nut, and similar butters. Neither party’s use of definitions was itself dispositive, so the court turned to other evidence, focusing on whether Miyoko’s specific use of the word “butter,” “in immediate or close proximity” to terms like “vegan,” “made from plants,” “cashew cream fermented with live cultures,” and “cashew & coconut oil spread,” was misleading.

“Butter” was the most prominent word on the front of the label, preceded by the smaller “European style cultured vegan” and directly followed by the smaller “made from plants.” “Cashew cream fermented with live cultures” was in even smaller text on the middle of the label’s bottom-left quadrant, and “cashew & coconut oil spread” was tinier still.

The state didn’t meet its burden of showing misleadingness. As to precedents: Courts have found that soymilk and almond milk aren’t deceptive, and “even the least sophisticated consumer[] does not think soymilk comes from a cow.” [This is empirically untrue, by the way, but then again there is a tranche of consumers who say that peanut butter has dairy in it.] And plant-based “meat” products have also been allowed to use “meat” terms when framed with “ample terminology to indicate [their] vegan or vegetarian nature . . . .” Nor was it convincing to argue that butter was different because “[n]either milk nor meat has a plant-based analog as well-known as margarine” given the current prevalence of products like “soymilk,” “almond milk,” and “coconut milk.” Although eggless “mayonnaise” might be deceptive, Duran v. Hampton Creek, 2016 WL 1191685 (N.D. Cal. Mar. 28, 2016), that case involved an “uniquely unrepresentative label,” which was “Just Mayo” and the picture of an egg, without any accompanying qualifiers such as “vegan” or “made from plants.”

As to consumer evidence: the state’s showing of confusion around plant-based dairy alternatives was “empirically underwhelming.” Although some research exists—I used some for my advertising law midterm last year—the state didn’t submit any. Miyoko’s submitted a study indicating, in relevant part, that the public “accurately identifie[s] the source of animal-based milk products 84% of the time, plant-based milk-products 88% of the time, animal-based cheese products 81% of the time, and plant-based cheese products 74% of the time.” Even if 26% of the participants were confused by plant-based cheeses, 19% were confused by animal-based cheeses; while 12% misidentified plant-based milks, 16% misidentified animal-based milks. At most, consumers are “perhaps a bit better at identifying traditional cheeses than vegan cheeses,” and perhaps the opposite for vegan v. traditional milks. This “modest” result  “hardly cuts in favor of finding Miyoko’s use of ‘butter’ inherently misleading.”

Finally, the court found “using the government-issued dictionary” to be “troublingly self-fulfilling.” I find this objection overblown. Especially where there is no clear definition or where consumers might just not know the technical details, government definitions can consolidate consumer expectations and, more importantly, insufficiently signaled deviations can be deceptive. Suppose Miyoko’s just called its product “just butter” and only explained what was going on in the ingredients list—as in Just Mayo. Is there any question but that the government definition, enforced for nearly a century, would be helpful in figuring out whether consumers were likely to be deceived?

But the court found that the federal standards—which distinguish between “butter” (dairy), “margarine” (which Miyoko’s product would have to be called if it was slightly fattier), and “spreads”—have to reflect “identifiable linguistic norms” to justify regulatory power over commercial speech. But: Where are such norms to come from? Why wouldn’t they plausibly come from nearly 100 years of regulation? What happens when a margarine producer wants to call its product “butter” and argues—quite plausibly—that no reasonable consumer knows the exact legal definition of butter or margarine? Maybe the answer is that this is really about labeling, and vegan butter labeling is cost-justified (helps more people than it hurts) while labeling margarine as butter would not be. But those are usually considered to be ordinary regulatory judgments, which is why First Amendment commercial speech doctrine has latterly been accused of Lochnerism. Anyway, the court didn’t credit the state’s “opinion” of what “butter” was.

Without misleadingness, CH factors two through four also favored Miyoko’s. Although the court thought that, in the abstract, government definition of terms might not be “more extensive than is necessary to serve [the state’s] interest” in consumer protection, this restriction didn’t directly advance the state’s interest. “In the narrow context of an as-applied constitutional challenge, the State might satisfy this requirement with a moderate showing of the ban’s tendency to redress harms caused by Miyoko’s vegan butter in particular, if not the dairy-alternative market writ large.” But the state didn’t show that this label had tricked anyone or otherwise show a “unique[]” threat from the label. [Unique can’t possibly be the right standard. Ordinary deceivers—which I’m not saying Miyoko’s is—are far from unique. I think the court has to mean “something about the label’s use of ‘butter’ increases deception potential more than other uses of ‘vegan butter.’”]

 

No comments: