major beer battle turns on Mead (Johnson)

Long post, lots of stuff to cover in this opinion.

MillerCoors, LLC v. Anheuser-Busch Cos., No. 19-cv-218-wmc (W.D. Wisc. May 24, 2019)

There’s an apocryphal ad story about a cannery stuck with unmarketable pale salmon that turned its disadvantage around by labeling the product as “Guaranteed not to turn pink in the can!” (The competitors supposedly responded by advertising that their products contained no bleach.)  The point is that the advertising context itself conveys a lot of meaning: in particular, aspects of one’s own product that one presents proudly are good, while aspects of the competition’s product that one brays loudly about are bad, even (or especially) if the consumer has no other information about how to identify good/bad characteristics.  AB took advantage of this standard feature of human communication, and got sued for it.

There’s a lot of detail here (and really good lawyering on both sides), but: AB started an ad campaign highlighting MC’s use of corn syrup in brewing Miller Lite and Coors Lite, whereas AB uses rice in Bud Light (but corn syrup in various other AB beverages).  MC sued for false advertising and trademark dilution; the court granted a preliminary injunction limited to ads that indicated in some way that Miller Lite and Coors Lite actually contained corn syrup when consumers drank it, specifically: (1) Bud Light contains “100% less corn syrup”; (2) Bud Light in direct reference to “no corn syrup” without any reference to “brewed with,” “made with” or “uses”; (3) Miller Lite and/or Coors Light and “corn syrup” without including any reference to “brewed with,” “made with” or “uses”; and (4) describing “corn syrup” as an ingredient “in” the finished product.  In addition, “the court would strongly encourage advance clearance before adopting any phrase or creative license in an attempt to maneuver around the prohibitions.”

The most disappointing part of the opinion for me was the court’s refusal to dismiss the trademark dilution claim because fair use is an affirmative defense. First, that’s new law as far as I’m aware, and given the First Amendment implications of applying dilution law to comparative advertising, it’s not a well-justified decision.  (The Ninth Circuit has held that in nominative fair use cases (which this is), the defendant’s only burden is to show that the use is referential (which MC pled), after which the burden shifts back to the plaintiff to show lack of entitlement to the defense.) Second, even so, the complaint clearly pled itself out of court; there is no way that this is not comparative advertising/nominative fair use.

On to the main event: The yeast that makes beer needs a nutrient substrate with sugars.  Sugars can come from malt, “or from a combination of malt and starchy grains like corn or rice.” The parties agreed that the sugar source relates to style and taste characteristics, though cost may also be a factor: corn syrup is cheaper than rice.  “There is no meaningful difference between using rice or corn syrup as an ingredient in terms of health or safety of the resulting beer product.”  Fermentation “converts the corn syrup sugars into ethanol, flavors, aromas, carbon dioxide, heat and a next generation of yeast cells, leaving a small amount of residual sugars.” No corn syrup remains in the Coors Light and Miller Lite products at the end of fermentation, and none is added.

Nonetheless, AB launched a huge ad campaign featuring claims that Miller Lite and Coors Light are “made with” or “brewed with” corn syrup. Here’s one Super Bowl ad.  Suffice it to say that the gist—central to the ads and much repeated—is that Bud Light doesn’t use corn syrup and Miller Lite/Coors Light do (“made with” and “brewed with” are frequently repeated).  AB billboards touted “100% less corn syrup” than the two competitors. An AB Twitter account posted an image displaying a Miller Lite can next to a Karo corn syrup bottle as if in a family portrait, and a Facebook user posted a similar image, referencing the campaign:

Karo Lite (syrup) and Miller Lite "family portrait"

user-generated content matching BL/ML with ears of corn and Aunt Jemima syrup

Another ad showed the following frames:

"Corn syrup"/"No corn syrup" by bottles of the parties' products

Another ad mocked MC for complaining and an AB character suggested “also imitating us by putting an ingredients label on your packaging. People want to know what ingredients are in their beer. But what do I know? I’m just the king of a kingdom that doesn’t brew beer with corn syrup.”

Beer Business Daily reported that according to Andy Goeler, AB’s head of marketing for Bud Light, “[Anheuser-Busch] did focus-group the heck out of this [Special Delivery] ad, and found consumers generally don’t differentiate between high fructose corn syrup and corn syrup, and that it is a major triggering point in choosing brands to purchase, particularly among women.”  In another interview, in response to the question, “What is wrong with corn syrup?,” Goeler responded: “People started to react to corn syrup, they started to react to no preservatives, and they started to react to no artificial flavors. There are things that consumers on their own had perceptions -- for whatever reason -- that there were ingredients they preferred not to consume if they didn’t have to.”  The interviewer noted that AB focused on corn syrup and not preservatives and asked why Goeler thought “consumers see corn syrup as something they don’t want.” He called it “an ingredient some prefer not to consume is the simple answer. . . . [S]ome consumers -- for their own personal reasons -- have concluded that they prefer not putting something like corn syrup, if they had a choice, into their body.” 

Though this is in one sense an ordinary bullshit nonanswer, it contains the key conflation at issue here: treating the brewing substrate as if it were what consumers “put[] … into their body,” which it is not.  Goeler’s answer recognizes the obvious fact that consumers have no particular reason to think about the distinction between beginning ingredients and the stuff in the can, and the ads are self-evidently encouraging consumers to consider the stuff in the can as if it still contained corn syrup, which is a reason to buy Bud Light instead.  What would be the point of the ads otherwise?  [See discussion of Grice below.]  Anyway, in another interview, Goeler stated that AB would continue to focus on “ingredient transparency,” and AB’s senior director of corporation communications echoed this, concluding that “Knowing what is and isn’t in your beer -- whether you paid for it or not -- can only be good for the beer industry overall.”  Again: there’s no corn syrup in MC beer, but of course that’s the message; the use of “made with” and “brewed with” does not tell consumers otherwise.

MC submitted a survey in which the control ad was the same as the test ad, but with an added, prominent disclaimer: “While corn syrup is used during the brewing of Miller Lite and Coors Light, there is NO corn syrup in the Miller Lite and Coors Light you drink.” Its expert concluded that “61 percent of respondents who saw the test stimulus believe that corn syrup is in the Miller Lite and/or Coors Light you drink,” compared to 26 percent in the control, for 35% net confusion. The expert also concluded based on an analysis of open-ended responses that the phrase “‘made with’ used in the Bud Light commercial to describe the relationship between Miller Lite and/or Coors Light and corn syrup is ambiguous.”

MC also argued that AB’s ads exploited or furthered misconceptions about corn syrup and high fructose corn syrup (HFCS).  They are not the same thing, but MC’s industry expert identified “a common point of confusion about corn syrup and HFCS, despite the fact that they are very different products.” The survey expert concluded that “more respondents who saw the test stimulus (24 percent) than who saw the control stimulus (19 percent) believed that the commercial says, suggests, or implies that corn syrup and high fructose corn syrup are the same.”

Unsurprisingly, after the ad campaign was launched, MC received a bunch of consumer communications related to corn syrup, up from basically none. The survey expert also analyzed these, and found that 18% indicated that the presence of corn syrup would end or decrease their purchases of MC beers and 4% indicated they’d start or increase purchases.  Relatedly, social media responses indicated that consumers were reacting to the ads as if corn syrup were a bad thing.  Of consumer posts that mentioned corn syrup and at least one of the relevant brands, 22% expressed negative sentiments and 9.8% expressed positive sentiments. Similarly, 28.6% of consumer posts indicated a mistaken belief that corn syrup used in brewing is present in the final product (the beer itself), while only 6.9% of posts demonstrated understanding that corn syrup wasn’t present in the final beverage.

Social media posts indicating that corn syrup is in the beer and that's bad

For the misleadingness alleged here, at the preliminary injunction stage, a consumer survey or other “hard evidence of actual consumer confusion” is not required. Instead, the Seventh Circuit, consistent with its self-understanding as a court of common-sense realists who are good at interpreting the ordinary meaning of ads, instructs that district courts should analyze “the ads themselves, the regulatory guidance, and the evidence of decreased demand.”  The court mentioned using intent to deceive as the basis of a presumption that the deception succeeded, but pointed out that courts that list this possibility never seem to find the requisite egregious intent, and it wasn’t going to be the first.  [The evidence here seems to me to support a finding that MC intended consumers to take away the idea that “brewed with” and “made with” is the same as “ingredient in the beverage.”  Is that “egregious”?  I think it’s about the natural meaning of an advertising statement.]

Statements that Miller Lite and Coors Light “use” or are “made with” or “brewed with” corn syrup were literally true, but allgedly deceived consumers into believing that MC’s products actually contained corn syrup and thus were unhealthy and inferior to Bud Light. MC relied on Abbott Laboratories v. Mead Johnson & Company, 971 F.3d 6 (7th Cir. 1992), in which the Seventh Circuit concluded that the name Ricelyte “implies more than the permissible message that Ricelyte is produced from rice or contains rice syrup solids derived from rice carbohydrates. It also implies that Ricelyte actually contains rice and rice carbohydrates—or at least we can say that Abbott has established a strong likelihood of so proving at trial.” 

But there is no affirmative duty to disclose or disclaim if the advertisement at issue is not false or misleading. And, unlike in Abbott Laboratories, the court thought that, viewing the “made with,” “brewed with” or “uses” statements in their full context, there were no express or implicit messages that the corn syrup is actually in the finished product. [So the court rejects the survey finding that indeed this was the message received by consumers from the ads.]  Specifically, the ads using the “made with” or “brewed with” language didn’t show corn syrup being added to the finished MC products.

MC also relied on Eli Lilly and Company v. Arla Foods, Inc., 893 F.3d 375 (7th Cir. 2018), in which the Seventh Circuit affirmed a preliminary injunction against a cheese manufacturer’s ads implying that milk from recombinant bovine somatropin (“rbST”)-treated cows was unwholesome. The ads stated that “Arla cheese contains no ‘weird stuff’ or ‘ingredients that you can’t pronounce’—in particular, no milk from cows treated with [rbST],” and depicted rbST as “a cartoon monster with razor sharp horns and electric fur.”  Though the explicit statements were true, the campaign “centers on disparaging dairy products made from milk supplied by rbST-treated cows” and drew “a clear contrast between Arla cheese (high quality, nutritious) and cheese made from rbST-treated cows (impure, unwholesome)” with monster imagery, “weird stuff” language, and child actors “to colorfully communicate the message that responsible consumers should be concerned about rbST- derived dairy products.”

But the ads here didn’t “disparage” corn syrup “or otherwise expressly draw attention to any negative health consequences.” (Even so, the court apparently agreed with MC that an express link “may have been unnecessary here, since at least some consumers appear to associate corn syrup, and particularly high fructose corn syrup, with harmful health consequences, or certainly defendant hoped,” but apparently didn’t consider that to be enough.)  Instead of “analyzing the commercial or the advertising campaign more broadly in order to demonstrate that these statements are misleading,” MC relied on the responses to open-ended survey questions, and that wasn’t enough.

The problem with using the survey to find the ads misleading (even though misleadingness is, even in the Seventh Circuit, supposed to be shown with consumer reaction evidence) is that surveys can’t be used to bring statements that are merely “susceptible to misunderstanding” within the scope of the Lanham Act. Mead Johnson & Company v. Abbott Laboratories, 201 F.3d 883 (7th Cir. 2000). In that case, “the Seventh Circuit rejected an attempt to solely rely on survey evidence to demonstrate that language is misleading.” On a petition for rehearing in Mead Johnson, the court of appeals stated that “interpreting ‘misleading’ to include factual propositions that are susceptible to misunderstanding would make consumers as a whole worse off by suppressing truthful statements that will help many of them find superior products.” The court of appeals also removed its initial statement that “[a] ‘misunderstood’ statement is not the same as one designed to mislead.” The court here commented that this statement would have supported plaintiff’s position that intent to deceive could be a basis for finding a statement misleading, but it’s gone now.  [I find it ironic that an amendment made to recognize that the Lanham Act is strict liability is used here to contract liability; the district court’s reasoning also makes a logical error about the contrapositive: the fact that intentional deception isn’t required for liability doesn’t mean that intent to deceive isn’t probative of whether an ad is misleading.]

“The court notes that Mead Johnson is not without its critics, and understandably so: how does one draw the line between an advertising statement that is susceptible to misunderstanding but not misleading?”  The court cites me:

One scholar has suggested that this holding requires materiality in addition to a misunderstanding for a claim to be misleading. Rebecca Tushnet, Running the Gamut from A to B: Federal Trademark and False Advertising Law, 159 U. Pa. L. 1305, 1349 (2011). To illustrate, Tushnet explains, “One may misunderstand a fact in the abstract: I could be wrong about the size of a computer’s hard drive. If I am misled, however, I am being led: induced, or at least potentially induced, to change my position based on my misunderstanding, as when I am more likely to buy a computer because of my belief about the size of the hard drive.”

But even this formulation “suggests that something more overt is required on the part of the advertiser to find a true statement misleading.” [I respectfully disagree.  The “something more” is materiality, which shouldn’t have to be “overtly” claimed in the ad, but rather should be based on evidence about the kinds of things consumers care about.  In my hard drive example, nothing more should be required than a misstatement about the hard drive’s size plus evidence that consumers care about the size of hard drive, rather than also requiring the rest of the ad to say “you should care about hard drive size!”]

“The Third Circuit -- the only court to discuss substantively the Mead Johnson decisions -- described the Seventh Circuit’s essential holding as follows: ‘there are circumstances under which the meaning of a factually accurate and facially ambiguous statement is not open to attack through a consumer survey.’”  [But exactly what are those circumstances?  I obviously share the district court’s frustration with the vagueness of the existing Seventh Circuit formulation.  The other possibility, besides materiality, that I discuss in my advertising law class is cost-benefit analysis: given how utilitarian the Seventh Circuit is, that might be even more plausible, especially given its language in response to the petition for rehearing.  The difference between true but misleading and true but misunderstood might be whether one can convey the truth some other, less misleading way. If a statement is useful to some substantial number of consumers and it can’t be appropriately qualified without confusing/failing to inform that subset, then confusion among another subset ought to be accepted as irreducable and unfortunate “misunderstanding.”  But if there is a way to convey the truth without also producing a big chunk of materially confused consumers, then the consumer protection purpose of the Lanham Act demands that we label the inefficient presentation of the truth “misleading.”  In that case, the survey here apparently shows that the existing ads are misleading, not just misunderstood, given the possibility shown by the control ad of conveying the truth without confusing nearly as many consumers.  In other words: some people may still misunderstand the control ad. But if it’s as good as we can get while still allowing AB to convey the truth, for whoever cares about it, then it’s the ad that should be used.]

The district court reasoned that “requiring plaintiff to point to some other aspect of the ad besides a truthful statement” was consistent with the caselaw—other cases finding misleadingness “consistently seem to rely on disparaging or derogatory references to the ingredients in the competitor’s product, or references that suggest a quality not present in a product is in fact in the product.” The fact that AB made a bunch of noise about corn syrup was not itself disparaging. 

MC relied on Grice’s “maxim of relevance,” H.P. Grice, Logic and Conversation, in 3 Syntax and Semantics: Speech Acts, 47 (P. Cole and J.L. Morgan, eds. 1975). As formulated by MC’s counsel:

if a speaker says something that is capable of more than one meaning and knows that . . . one of those meanings is something that’s relevant and meaningful information to the listener and that the other meaning is something that is irrelevant and not meaningful to the listener, the listener will assume that the speaker means the meaningful, relevant statement because why would you be telling me something that’s not important.

The court found this argument “[i]ntuitively” appealing, despite AB’s argument that there are other reasons to highlight rice versus corn syrup distinction, namely that consumers may align Bud Light with the “farm-to-table” movement or using a “real ingredient versus a syrup.” The court thought that was “counter to the weight of the evidence as to defendant’s intent that associating Miller Lite and Coors Light with corn syrup would motivate consumer’s health concerns with consuming corn syrup or worse, high fructose corn syrup.”  But even so, the scale of a campaign wasn’t enough to find a “neutral, truthful statement” misleading. 

I think the court’s reasoning does not mesh with how advertising actually works, and encourages pollution of the information environment. As Eric Goldman and I have written in our text:

[W]hen an advertiser prominently claims some feature for its product, we are likely to assume that the feature is relevant and desirable. The advertiser is using the cooperative principle of relevance to imply these things without stating them directly. The only direct statement is that the feature exists. Researchers have studied irrelevant claims used to tout brands, such as claims that instant coffee has flaked crystals. When some consumers saw ads featuring relevant attributes and others saw ads featuring irrelevant attributes, the latter group preferred the product more. Even when the researchers told consumers which attributes were irrelevant in advance, consumers still preferred the product more when shown ads using the irrelevant attribute. See Gregory S. Carpenter et al., Meaningful Brands From Meaningless Differentiation: The Dependence on Irrelevant Attributes, 31 J. Mktg. Res. 339 (1994).

The fact that it’s in an ad instructs the viewer about the value to be assigned to the advertised characteristic—and consumers know this.

MC had another possible argument: the “something else” might come from the rest of the campaign—the ads that didn’t limit themselves to “made with” or “brewed with” but were (otherwise) focused on exactly the same message. “In other words, consumers could reasonably interpret ‘made with’ and ‘brewed with’ to mean corn syrup is in the final product because of defendant’s other ads describing Bud Light as having 100% less corn syrup, or describing Miller Lite as ‘corn syrup’ and Bud Light as ‘no corn syrup’ without reference to the brewing process.”  But the court was concerned that this was too much of a stretch because most of the ads at issue, and the ones most widely disseminated, used “made with,” “brewed with,” or “uses.”

The court also declined to follow FTC cases saying that defendants can’t exploit preexisting misunderstandings, because FTC precedent “is largely inapplicable to Lanham Act cases.” [While each of the district court’s conclusions here is understandable on its own, I have to say that the net impression is: MC needs to show “something more,” but the court systematically excludes any possible “more” that is outside the four corners of a challenged ad—which does not fit my understanding about how human communication works.]

Thus, there was no likelihood of success on the merits as to ads solely using the language “brewed with,” “made with,” or “uses.”

Next, the court turned to ads touting that Bud Light has “100% less corn syrup than Miller Lite or Coors Light” or that it has “no corn syrup.” “[T]hese statements, while also literally true, support a reasonable interpretation that Miller Lite and Coors Light contain corn syrup,” and MC showed a likelihood of success on the merits.  AB argued that, given the other ads, consumers would understand these to be shorthand references to brewing, but “a reasonable jury could and likely would find that these ads encouraged consumers to draw the wrong inference from the original ads.”

“Ingredient” claims: There aren’t labeling requirements for beer ingredients, and it’s not clear what an “ingredient” is when it comes to beer. For food, “ingredient” “appears to cover items that are used in the production of a food product, even if not in the end product (e.g., leavening items).” In the past (and apparently even now), MC has listed corn syrup as an “ingredient” on its website, without any reference to it being used solely in the brewing process and not present in the end product.  In context, in most of the ads, “ingredient” was used in conjunction with the “made with” or “brewed with” language, except for one ad, the one that says “Look if you’re this set on imitating our kingdom, may I suggest also imitating us by putting an ingredients label on your packaging. People want to know what ingredients are in their beer. But what do I know? I’m just the king of a kingdom that doesn’t brew beer with corn syrup.”  That language crossed the line “to encourage a reasonable consumer to believe that corn syrup is actually contained in the final product.”

MC also challenged AB’s statements that Miller Lite and Coors Light selected corn syrup to “save money” or because it is “less expensive.” The court refused to hold this misleading based on MC’s theory that a consumer would conclude that cheaper substrates are less healthy—that was just too much of a stretch.  [This result, by the way, seems reasonably consistent with my proposed cost-benefit reading of Mead Johnson.]

The court then returned to MC’s “strongest evidence,” AB’s own statements of its intent that “it was both aware of and intended to exploit consumer concerns about corn syrup (and high fructose corn syrup in particular).” Though this evidence indicated AB’s hope that consumers would interpret statements about “made with” or “brewed with” as meaning that corn syrup was actually in the finished products, the court declined to presume misleadingness, “though perhaps this evidence could serve to color the ‘made with’ and ‘brewed with’ statements to push them across the line to allow a reasonable finding that they are misleading, and not simply susceptible to misunderstanding.” [Which suggests that a jury could consider the issue as to all the ads, though the court then indicates that it’s most interested in guidance from the court of appeals.  Again, one might be able to fit evidence of intent into a cost-benefit analysis as well as into a materiality analysis: an intent that consumers believe something beyond the literal statement could itself suggest that there are other, less confusing ways to convey the truth, and an intent that consumers be alarmed by a statement is pretty good evidence that they might act on it, if we believe that a major manufacturer like AB generally has some idea about what its consumers care about.]

The court then, somewhat confusingly, turned to whether there was evidence that a “substantial segment” of the audience was deceived by the challenged ads: the survey, consumer communications to MC, and social media evidence.  AB contested the survey and the 35% net confusion finding.

First, it was ok to use a test commercial that wasn’t the ad that launched the ad campaign and ran the most times/was seen by the most people. Both ads compared the beers and highlighted corn syrup, and AB didn’t articulate a basis for treating them differently.

AB argued that the questions were leading because a consumer who chose the answer “made with corn syrup” to a communication/filter question was then told: “Being “made with” corn syrup may mean a number of different things” and then asked “Which, if any, of the following statements does the TV commercial say, suggest, or imply …?” (The possible answers included used only during brewing, in the beverage you drink, both, neither, and don’t know/unsure.)  AB argued that the follow-up improperly instructed respondents that the term “made with” is ambiguous. This went to weight, rather than admissibility, and there was “at least some likelihood of persuading a reasonable jury of the risk of confusion.” Anyway, survey evidence isn’t required for a preliminary injunction.

As for the consumer communications, there weren’t that many of them—32 total, out of the approximately 100 million people who watched the Super Bowl, and many accurately described use or brewing with corn syrup. Viewed in isolation, these statements were insufficient to prove that a substantial segment of consumers were likely to be deceived, “but it is some anecdotal support for the survey results.” The social media reactions were treated similarly.

So the survey, which tested a “made with”/“brewed with” ad, helped support a finding of likely success on the merits, but only as to those ads that didn’t use the “made with”/“brewed with” language and thus crossed the line from misunderstanding to misleading.  That sounds weird, but I guess I see how the court got there: if “made with”/“brewed with” ads are confusing but nonetheless nonactionable, then surely the ads that go even further are deceptive.

Irreparable injury:  “[I]t is well settled that injuries arising from Lanham Act violations are presumed to be irreparable, even if the plaintiff fails to demonstrate a business loss.” Promatek Indus., Ltd. v. Equitrac Corp., 300 F.3d 808, 813 (7th Cir. 2002). The Seventh Circuit hasn’t revisited this in light of eBay (though it should, but I understand why a district court wouldn’t, even though I think it could also have read the writing on the wall/in the statute). Anyway, no presumption was necessary given the evidence of irreparable harm.  AB argued that failure to prove harm to sales prevented a finding of irreparable harm, but MC showed reputational harm through its survey and anecdotal social media and consumer comments, which was irreparable even with conflicting evidence about whether MC’s sales or market share had been harmed.  Indeed, the evidence showed that MC’s market share was “stable at best, while defendant’s sales and market share is growing, albeit apparently at the expense of other, smaller competitors. This begs the question why plaintiff’s share of these lost sales has not grown, at least in proportion to its share of the market for light beers.”

However, the court qualified its finding of harm, which affected the scope of the injunction.  As in the SDNY Danone case, the court thought that there were truthful things that AB can say, and harm caused by truth couldn’t be actionable; relevant harm comes from the difference between the truth and what AB actually said. “Here, if plaintiff’s reputation was injured solely, or even principally, by consumer awareness that its products are brewed with or made with corn syrup, this may not serve as a basis for reputational injury.” [In the Danone case, however, there was no evidence that the potential truthful statement—Chobani actually had less sugar than the Dannon product, just not as much less as it claimed—was misleading, whereas MC’s evidence tends to show that the truthful “made with”/“brewed with” statements, presented in the way AB presents them, imply misleading factual claims to consumers.]

 

package with "no corn syrup" and image prominently on sides

The injunction thus didn’t disturb ads that only use the words “brewed with,” “made with,” or “uses.” MC also sought to cover AB’s packaging that prominently featured the claim “no corn syrup,” but this wasn’t part of the original written PI submissions.  The court allowed MC to submit an argument why the injunction should cover noncomparative packaging.  [The answer is Grice, or maybe xkcd.]