Thursday, July 31, 2014

COOL story, part 3: dissents

Judge Henderson dissented, believing the question of Zauderer’s appropriate scope to have been resolved in R.J. Reynolds and unfairly revisited.

Judge Brown also dissented, quite vigorously. Under the majority’s reasoning, “a business owner no longer has a constitutionally protected right to refrain from speaking, as long as the government wants to use the company’s product to convey ‘purely factual and uncontroversial’ information.” This ignored the government’s unique interest in preventing commercial deception and applied a standard that was more relaxed than rational basis review.  The majority searched “sua sponte through the underlying statute’s legislative record, desperately seeking justifications while ignoring the agency’s actual rulemaking record.”  Zauderer is about deception, and, worse than “nonsense on stilts,” “the court’s analysis in this case can best be described as delirium on a pogo stick.”  (I’m going to try to limit further outraged quotes until I get to the end.)

Zauderer said that, where deceptive advertising could be cured by more speech, the government may choose between requiring disclosure and directly prohibiting the advertisement. (I find this to be a weird description; the Court has sometimes suggested that the government is required to try disclosure if the speech is only potentially misleading, but that doesn’t seem like the same thing.)  But compelled speech can be as bad as banned speech, “and the government faces a heavy burden to justify involuntary affirmation (being forced to carry the government’s message).”  Zauderer contrasted the imposition of orthodoxy “in politics, nationalism, religion, or other matters of opinion” with regulation of deceptive commercial advertising.  The latter is ok as long as disclosure requirements are reasonably related to the State’s interest in preventing deception of consumersThe Court wasn’t just distinguishing between disclosure and bans. “[T]he state’s option to require a curative disclosure cannot be disconnected from its right to entirely prohibit deceptive, fraudulent, or misleading commercial speech.”  The option to ban instead is what justifies requiring advertisers to provide more information than they’d otherwise be inclined to present.  (The greater includes the lesser?)  But Zauderer doesn’t allow a commercial speaker to be forced to speak factual and noncontroversial information “in the first instance.”  Even potentially deceptive commercial speakers have “minimal First Amendment protections,” and when deception isn’t involved, “constitutional protections remain robust and undiminished.”

The Constitution adopted the principle of natural law “that an adult human being, as a free moral agent, cannot be coerced without good reason,” but also that “no one ever had a natural right to do wrong.” That’s the basis of commercial speech doctrine.  Disclosure wasn’t an exception to the First Amendment’s stringency, but rather an acknowledgement that sellers had no right to “wrongly deceive” consumers; the state can’t control opinion, but can require accuracy for commercial facts.  (Apparently there’s a right to rightly deceive consumers though.  Unclear why NYT v. Sullivan comes out the way it does on this reasoning, unless it's not a natural wrong to deceive someone in a political context, but maybe it doesn't come out the same way in this scheme.)

Supreme Court jurisprudence routinely refers to deception as the justification for controls on commercial speech, and its clear trajectory is to more protection, not less.  It doesn’t matter how small AMI’s interest in remaining silent is; the government has to show that its interest justifies the regulation.  Zauderer is, as the majority says, just a particular application of Central Hudson.  Central Hudson requires a substantial interest, and preventing “inherent or actual” deception is always substantial.  When a mandated disclosure is reasonably related to preventing deception, as Zauderer requires, it will always directly advance that substantial interest.  And it will be less restrictive than an outright ban, and no more restrictive than necessary.

(The Central Hudson test part 1 explicitly exempts false and misleading speech from the test set out in parts 2-4; to then run what Judge Brown says are the only acceptable compelled disclosures, those that fight false and misleading speech, through Central Hudson analysis seems odd.  The reference to “inherent or actual” deception is a red flag—this is language from cases that say that the government has to try disclosures first if the speech is only “potentially” misleading, whereas it can outright ban speech that is “inherently or actually” misleading.  So in fact, Judge Brown has not applied Central Hudson to the key Zauderer class of “potentially misleading” speech; however, she might well concede that the prevention of “potentially” misleading speech is a substantial government interest. 

A separate issue: Judge Brown argues that the remaining Central Hudson steps are pre-answered for false/misleading speech in terms of government interest and fit.  But Judge Brown puts the question in terms of Zauderer and not the ability to ban false/misleading speech outright.  Why not run all bans on commercial speech, including bans on false or misleading speech, through Central Hudson factors two through four?  Being explicit about that would probably raise the question of how an outright ban could ever be a proper fit, since disclosure requirements would regularly be an alternative, but her analysis depends on outright bans being okay.  Navigating this tangle would have to depend on some very thick theories of when consumers actually understand disclosures (spoiler: rarely); query how good courts are compared to regulators at identifying when disclosures would be sufficient.)

Anyhow, “Central Hudson—without any shortcuts— applies to disclosures that target interests other than deception.”  But the majority “disembowel[ed]” that case as well by holding such “amorphous” interests to be substantial.  And it relied on “interests the agency never asserted and even denied were rationales for the rule.”  As a matter of administrative law, this was wrong, and in any event heightened scrutiny couldn’t be satisfied by “hypothesized justifications” “based on a few scattered comments in the legislative record.”  Nor did the congressional record show that the interest, even if substantial, was properly fitted to this restriction.

The government only asserted the vague interest in providing consumers with information. But the government never explained why origin information matters.  (Contrast discourse about geographic indications, where vague assertions about terroir matter a lot, and proponents happily claim that the differences can’t be defined but are real.)  And even if it did, the government didn’t explain why coerced speech was the only solution.  (And we’re back to less restrictive alternatives!  Why isn’t counterspeech a less restrictive alternative to suppressing false commercial speech, as it is for false political speech?)

The result was “rational basis review minus any legitimate justification.”  Then res ipsa loquitur meant that disclosure was self-evidently likely to convey the information to recipients, thus advancing the government’s interest.  “Seriously? With logic like this, who needs a Ministry of Truth?”  (Okay, but what is the structural difference between this logic and Judge Brown’s reasoning above that requiring disclosures for misleading speech is self-evidently likely to further the government’s interest?)

Judge Brown also rejected reliance on the long history of country of origin labeling, because for a long time the First Amendment wasn’t enforced to protect commercial speech and litigants usually didn’t bother to argue about it, raising substantive due process claims instead. If the test of time were enough to protect a law, there’d be no commercial free speech at all.

The majority concluded that protectionism or patriotism was the true justification for COOL, “even if it is only acknowledged with a sly wink by the government.” But the government didn’t assert protectionism as a justification, and that would be a substantial justification for coerced speech only if “voluntary action and direct government speech were obviously inadequate.”  (Least restrictive means again.  Again, why doesn’t that also apply to false or misleading commercial speech?)  Anyway, the agency persistently denied protectionist motives, stating that COOL wouldn’t necessarily change aggregate consumer demand for products of any given origin and pointing to lack of participation in voluntary labeling programs as evidence that consumers don’t have a strong preference for COOL. 

Likewise, the government didn’t assert any health or safety interests, maintaining instead that food safety regulations governed all food and that traceability wasn’t the intent of the rule.  Indeed, the agency said that, though some evidence suggested that consumers used COOL as a proxy for safety information, that wasn’t a valid inference.  This undercut the idea that it was reasonable for Congress to anticipate consumer preferences.  (The government’s litigation position really is silly; by asserting factual premises that contradict consumer beliefs—those beliefs being that country of origin matters for various reasons, including safety risks—the government manages to be paternalistic in both directions.  Its position seems to be: “this information should be disclosed to you, consumers, even though you don’t know what you’re doing and your preferences are wrong.”)  The anecdotes in the legislative history didn’t suggest that COOL would be useful in a health crisis.  Rather, the agency determined that prevention and recall measures are the means to be used to protect health.

And now to the beating, Lochner-esque heart of the matter: this case isn’t about COOL, or patriotism or protectionism, or health and safety.  “[T]his is a case about seeking competitive advantage.”  The rule “benefits one group of American farmers and producers, while interfering with the practices and profits of other American businesses who rely on imported meat to serve their customers. Such a disproportionate burden ‘stands in sharp conflict with the First Amendment’s command that government regulation of speech must be measured in minimums, not maximums’” (citation omitted).  (But wait!  How, you might well ask, did an economic burden on conditions of production become a First Amendment harm?  Money may be speech in campaign finance, but how is money spent to segregate animals by country of origin speech?  I think the dissent’s point must be that an interest in helping one group over another can’t be “substantial.”  But stated that way it’s a very broad claim, and one I don’t think the case law supports.)

Judge Brown predicted that today’s victors will live to regret this when other objectionable disclosure requirements burden  them, like disclosures about cattle feeding or raising practices, environmental effects of beef production, “or even the union status or wage levels of their employees.”  

If “Made in the USA” worked, producers and sellers would happily and noisily disclose, and consumers’ desires to buy American could be satisfied by voluntary action.  But mandatory COOL just facilitated rent-seeking.  By accepting such flimsy, nebulous interests, the court allowed the government to “commandeer the speech of others” on any ground, including motives “in aid of any sort of crony capitalism or ideological arm-twisting.”  The government’s alleged interest in providing information will result in higher prices because of the cost of monitoring the supply change, taking away the price advantage currently enjoyed by some producers.  “Query whether the protections of the First Amendment should be abrogated for some businesses in order to benefit other businesses.” (Query whether the First Amendment analysis should turn on which businesses are benefited by a regulation.  I think there are good non-speech-related reasons we might want to limit big producers’ comparative advantage in the meat production process; I also think that the “crony capitalism” objection may well be valid.  I merely doubt that the D.C. Circuit Court of Appeals is the appropriate governmental body to resolve it, and in particular that the First Amendment is the appropriate mechanism by which to do so.)

Judge Brown concluded that “[t]he First Amendment ought not be construed to allow the government to compel speech in the service of speculative or hypothetical interests for purely private benefits.”  (What counts as a public benefit?  If there’s no such thing as society, aren’t all benefits private?)  Under this reasoning, there’s no limit to what government could compel.  “[I]f this example of cronyism is okay, who will balk at any other economic or ideological discrimination?” This result dissolved “the whole idea of a right not to speak. … And it does so to facilitate coercion and the imposition of orthodoxy. What is more uncontroversial than orthodoxy?”

(I’m unclear on the fact/orthodoxy relationship here.  I’m also unclear about the orthodoxy being enforced by this regulation.  I will absolutely buy that, in the context of American society, COOL implicitly carries a “buy American” message.  But compared to, say, Christian prayer at the beginning of a legislative session, or mandatory ultrasounds and “disclosures” that abortion is associated with suicide, I find it hard to identify ideological coercion here.  There’s something here about the fever pitch of American politics of late, I think.  If putting country of origin labeling on meat is tyranny, what do we call it when the government jails journalists for reporting, or collects all our private communications in case they might be useful later?)

And the final heights:

There can be no right not to speak when the government may compel its citizens to act as mouthpieces for whatever it deems factual and non-controversial and the determination of what is and what is not is left to the subjective and ad hoc whims of government bureaucrats or judges. In a world in which the existence of truth and objective reality are daily denied, and unverifiable hypotheses are deemed indisputable, what is claimed as fact may owe more to faith than science, and what is or is not controversial will lie in the eye of the beholder.

On one reading of this claim, SEC disclosures, the FTC’s consumer protection side, and most of what the FDA does are equally unconstitutional, since only the speaker should decide for herself what facts are “true.” On another, judges (and bureaucrats?) can decide some truths, but not this one—mandatory disclosures are ok but have to address real deception, however narrowly Judge Brown would define that. I think the dissent does itself a disservice by mushing this all together.  The objection to allowing the government to find “facts” is itself not a First Amendment objection, but rather an argument—Lochner again!—that the government should not be allowed to regulate at all.  (Also, and consistent with Judge Brown’s dislike of commercial speech doctrine, note how commercial speakers have become unmodified “citizens,” as if the majority’s holding allowed disclosure mandates in noncommercial speech as well.)

COOL story, part 2: concurrences

AMI continued: concurrences

Judge Rogers concurred in part.  She wrote to disassociate herself from the suggested collapse of Central Hudson and Zauderer.  “Viewing Zauderer as simply an application of Central Hudson to special circumstances … finds support in neither Supreme Court precedent nor the precedent of this court or our sister circuits.”  Blurring the lines would be unnecessarily confusing, especially in other cases where the key issue could be evidentiary support for, or “fit” of, the regulation.

Both the majority and the dissent contravened Zauderer and didn’t give sufficient weight to the purposes of First Amendment protection for commercial speech.  Zauderer “was not tracing a shortcut through Central Hudson but defining a category in which the interests at stake were less threatened.”  There are material differences between disclosure requirements and speech bans.  Mandatory disclosure of beneficial consumer information is “consistent with the reasons for according constitutional protection to commercial speech and therefore justifies less than strict review.” 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484 (1996) (plurality).  This is consistent with the Supreme Court’s longstanding focus in commercial speech cases on consumers’ interests in the free flow of commercial information.  Truthful disclosure “furthers, rather than hinders, the First Amendment goal of the discovery of truth and contributes to the efficiency of the ‘marketplace of ideas.’”  Because of the difference between disclosures and bans, Zauderer scrutiny is less exacting than Central Hudson scrutiny, regardless of the government interest asserted.

Judge Kavanaugh concurred, doing just what Judge Rogers didn’t want: applying Central Hudson, these particular regulations passed.  A substantial government interest was required, and was present here. 

“The Government has long required commercial disclosures to prevent consumer deception or to ensure consumer health or safety. Those interests explain and justify the compelled commercial disclosures that are common and familiar to American consumers, such as nutrition labels and health warnings.”  But the requirements here weren’t traditional anti-deception, health, or safety interests, because COOL “obviously does not serve those interests.”  Instead, the regulation provided consumers with information. 

Providing information alone “plainly” isn’t enough, because it would be true of any and all disclosure requirements.  (Why is that in itself problematic?  Even if we accept that there must be some disclosure requirements that flunk Central Hudson, must there be some disclosure requirements that do not further a substantial interest?)  That would allow any compelled commercial disclosure and give governments “a free pass to spread their preferred messages on the backs of others.”  Some consumers might want to know “whether their U.S.-made product was made by U.S. citizens and not by illegal immigrants,” “whether a doctor has ever performed an abortion,” or “the political affiliation of a business’s owners.”  “These are not far-fetched hypotheticals, particularly at the state or local level.”  History and tradition provide no justification for “free-wheeling” power of this sort.

However, COOL was justified by the “historically rooted interest in supporting American manufacturers, farmers, and ranchers as they compete with foreign manufacturers, farmers, and ranchers.”  Economists can debate protectionism, but Congress has long sought to support US industries against foreign competitors; that gives this interest a sufficient historical pedigree to be substantial. COOL serves this interest because it causes many American consumers, for varied reasons, to buy more American-made products. 

But has the government actually asserted this interest, as required under Central Hudson?  The executive branch didn’t expressly do so during the litigation, “apparently because of the international repercussions that might ensue.”  But it’s still obvious, and Congresspeople did articulate it, which was enough.

Then the question was whether there was a sufficient fit between the disclosure requirement and the government interest. Judge Kavanaugh read Zauderer to apply Central Hudson on this point. Requiring mandatory disclosures be “purely factual,” “uncontroversial,” not “unduly burdensome,” and “reasonably related to” the government’s interest was just an application of Central HudsonZauderer just explains what sufficient tailoring is in the context of compelled commercial disclosures.  This is more than rational basis review, which “would undoubtedly tolerate government mandates of moral or policy-laden messages, of controversial messages, of burdensome labels, of disclosures that are only indirectly related to the Government’s interests.”  Zauderer, by contrast, “tightly” limits mandatory disclosures to “a very narrow class.”  Judge Kavanaugh commends the majority for recognizing that Zauderer was more than rational basis review—a beautiful example of attempting to influence future readings of a majority opinion.

Here, Zauderer’s stringent requirements were met: the disclosures here were purely factual, not unduly burdensome, and reasonably related to the government’s interest.  (In my reading, the one thing really missing from these “stringent” requirements is any consideration of alternatives.  Here, that would seem to be outright protectionism, if the interest is as described.  By contrast, an interest in informing consumers so they can make their own choices really isn’t satisfied by tariffs on foreign meat.)  Controversiality is an unclear standard, but not here “given the factually straightforward, even-handed, and readily understood nature of the information, as well as the historical pedigree of this specific kind of disclosure requirement.”

COOL story, part 1: DC Circuit upholds country of origin labeling rule

American Meat Institute v. U.S. Dep’t of Agriculture, No. 13-5281 (D.C. Cir. July 29, 2014)

The D.C. Circuit here, en banc, upholds country of origin labeling (COOL) requirements for meat, and in the process holds that Zauderer’s standard for compelled commercial disclosures does not require the government interest at issue to be the prevention of deception. This holding, however, reveals deep divides among the judges about compelled commercial disclosures, with both concurrences and dissents that try to frame the issue differently—looking ahead to the inevitable cert petition.

Congress required COOL on many foods, including some meat products, and passed a law defining “country of origin” for meat to be based on where the animal had been born, raised, and slaughtered—the three major production steps. After a WTO panel found the first implementing regulations to violate our international obligations (they allowed multiple countries to be listed, and further allowed commingling of meat from animals of different origins as long as the label listed all the countries from which the animals might have come), the Secretary promulgated a rule requiring more precise information on the location of each production step.  The new rule also eliminated the flexibility allowed in labeling commingled animals, which promises increased costs for certain operators.  (Here’s another variant of the First Amendment as Lochner: it’s a labeling rule, but it will demand changes in production practices so that producers can actually identify the thing that needs to be labeled; this is an economic regulation, but its output is information, so now it’s a First Amendment problem.  But the producers’ objection isn’t really to the compelled speech; it’s to the expense of implementing the new production practices required of them.)

AMI argued that Zauderer didn’t apply to government interests other than preventing consumer protection, and that this regulation flunked Central Hudson. The parties agreed that Zauderer applied to mandatory disclosure of “purely factual and uncontroversial information” appropriate to prevent deception in the regulated party’s commercial speech. What of purely factual and uncontroversial disclosures serving other government interests?  AMI argued that even if Zauderer did extend that far, the government had no substantial interest in COOL.

Zauderer itself isn’t clear; the government’s interest in that case, and the subsequent Milavetz case was in preventing deception, and thus the language could simply have conformed to that.  However, the language Zauderer used to justify its approach was far broader.  Zauderer rejected Central Hudson analysis as unnecessary in light of the “material differences between disclosure requirements and outright prohibitions on speech.” The Court continued: “the First Amendment interests implicated by disclosure requirements are substantially weaker than those at stake when speech is actually suppressed.” The speaker’s interest was “minimal”: “Because the extension of First Amendment protection to commercial speech is justified principally by the value to consumers of the information such speech provides, appellant’s constitutionally protected interest in not providing any particular factual information in his advertising is minimal.” These reasons seem “inherently applicable beyond the problem of deception,” and the en banc majority noted that other circuits have so found.

Okay then.  Is the government’s interest in COOL adequate?  AMI disparaged the interest at stake as merely satisfying consumers’ “idle curiosity.”  The Supreme Court hasn’t told us whether a Zauderer-passing interest has to qualify as “substantial” under Central Hudson, and anyway what counts as a substantial interest “itself seems elusive.”  However, several aspects of COOL for food combined to make a substantial interest: “the context and long history of country-of-origin disclosures to enable consumers to choose American-made products; the demonstrated consumer interest in extending country-of-origin labeling to food products; and the individual health concerns and market impacts that can arise in the event of a food-borne illness outbreak.”

COOL has “an historical pedigree that lifts it well above ‘idle curiosity,’” dating to 1890.  “[T]he ‘time-tested consensus’ that consumers want to know the geographical origin of potential purchases has material weight in and of itself.”  The value of this information is a matter of common sense.  Congresspeople identified the statute’s purpose as “enabling customers to make informed choices based on characteristics of the products they wished to purchase, including United States supervision of the entire production process for health and hygiene.”  Some believed that COOL would lead to consumers choosing to buy American.  “Even though the production steps abroad for food imported into the United States are to a degree subject to U.S. government monitoring, it seems reasonable for Congress to anticipate that many consumers may prefer food that had been continuously under a particular government’s direct scrutiny.” 

Congresspeople also indicated that people “would have a special concern about the geographical origins of what they eat.”  The legislative history refers to “the collapse of the cantaloupe market when some imported cantaloupes proved to be contaminated and consumers were unable to determine whether the melons on the shelves had come from that country.” This anecdote “more broadly suggests the utility of these disclosures in the event of any disease outbreak known to have a specific country of origin, foreign or domestic.”  (This discussion is conducted in a weird vacuum, with no mention of just how few USDA inspectors there are and just how little food gets inspected—here and elsewhere.  Is it ironic that many of the people who might buy American based on assumptions about superiority may well have voted for representatives who have made American origin less and less a guarantee of safe and hygenic production?)

Surveys in the record indicated that 71-73% of consumers would be willing to pay for COOL.  Though consumers tend to overstate their willingness to pay, these studies, plus the many favorable comments received during rulemaking, reinforced the historical basis for treating such information as valuable.  

This will be important later: The interests served by the rule were those advanced by Congress when it adopted the statute, even if the agency didn’t discuss those interests.  In defense of this position, the en banc majority stated that it didn’t want to allow “perfectly adequate legislative interests properly stated by congressional proponents” to be “doomed by agency fumbling (whether deliberate or accidental),” because that rule would “allow the executive to torpedo otherwise valid legislation simply by failing to cite to the court the interests on which Congress relied,” and allow the next President to reinstate a regulation by citing the right interests. “We do not think the constitutionality of a statute should bobble up and down at an administration’s discretion.” 

Anyway, the agency sufficiently invoked the relevant interests. Agency statements from prior rulemakings claiming that COOL serves no food safety interest weren’t inconsistent with the government’s litigation position.  “Simply because the agency believes it has other, superior means to protect food safety doesn’t delegitimize a congressional decision to empower consumers to take possible country-specific differences in safety practices into account.”  (Compare this idea to the Pom Wonderful statement that the FDA isn’t as expert about consumer understanding as market participants are.  When regulators fail to protect us, who can take up the slack, and how?)

Given an adequate interest, the remaining Zauderer question is whether the regulation fits the interest.  The en banc majority comments that the Zauderer test “differs in wording, though perhaps not significantly in substance,” from the Central Hudson test, at least on these facts.  Central Hudson would require the court to determine whether “the regulatory technique [is] in proportion to [the] interest,” an inquiry comprised of assessing whether the chosen means “directly advance[s] the state interest involved” and whether it is narrowly tailored to serve that end.  A Central Hudson analysis commonly requires evidence of a regulation’s effectiveness, but “such evidentiary parsing is hardly necessary when the government uses a disclosure mandate to achieve a goal of informing consumers about a particular product trait.”  Zauderer is like res ipsa loquitur: “by acting only through a reasonably crafted disclosure mandate, the government meets its burden of showing that the mandate advances its interest in making the ‘purely factual and uncontroversial information’ accessible to the recipients.”  The disclosure must however relate to the good or service offered by the regulated party.

The majority also commented that “[t[he self-evident tendency of a disclosure mandate to assure that recipients get the mandated information may in part explain why, where that is the goal, many such mandates have persisted for decades without anyone questioning their constitutionality.”  COOL isn’t the only required disclosure about product attributes—others include fiber content, care instructions for clothing items, and listing of ingredients. 

“Notwithstanding the reference to ‘narrow tailoring,’ the Court has made clear that the government’s burden on the final Central Hudson factor is to show a ‘reasonable fit,’ or a ‘reasonable proportion,’ between means and ends.”  When the government’s interest is to ensure that consumers receive particular information, this means-ends fit is “self-evidently satisfied when the government acts only through a reasonably crafted mandate to disclose ‘purely factual and uncontroversial information’ about attributes of the product or service being offered.”  Mandatory disclosures will therefore almost always have a reasonable means-ends relationship, unless disclosure is unduly burdensome in a way that chills protected speech.

Applying Zauderer here, the disclosures were “purely factual and uncontroversial information” about the good or service being offered.  AMI objected to the word “slaughter,” but the court didn’t need to address its objections because the rule allows retailers to use “harvested” instead, and AMI didn’t object to that.  AMI also didn’t disagree with the truth of the facts to be disclosed, so they weren’t controversial “in that sense.”  Nor was COOL controversial “in the sense that it communicates a message that is controversial for some reason other than dispute about simple factual accuracy.”  Some required factual disclosures could be so one-sided or incomplete that they wouldn’t qualify as “factual and uncontroversial” (note: but not abortion disclosures; never those), but not these.  (Note the tension with the idea that consumers could believe that foreign countries might provide dangerous food.)  COOL doesn’t require corporations to carry messages biased against or expressly contrary to the corporation’s views. Nor is it so detailed that it effectively rules out ordinary advertising methods.  As a result, the government’s interests here were sufficient to sustain COOL under Zauderer.

Second Circuit reiterates presumptions of harm, damages in literal falsity case

Merck Eprova AG v. Gnosis S.p.A., No. 124218 (2d Cir. July 29, 2014) 

Merck won a false advertising case against Gnosis, and the court awarded it over $500,000 in damages, over $2 million in attorneys’ fees and costs, and prejudgment interest, and also ordered corrective advertising.  The court of appeals affirmed in full.  The key holdings are both clear and relatively narrow: in a two-player market, literal falsity and deliberate deception justify legal presumptions of both consumer confusion and injury for liability purposes.  Willful deception also justifies an award of defendant’s profits, and enhanced damages in circumstances such as those present here.

The parties produce folate, a critical prenatal supplement.  The predominant naturally occurring form is known as 5MTHF.  Chemical compounds that differ only in their arrangement around a carbon atom, as are relevant here, are known as stereoisomers (either L and D or S and R).  In folate, L is the naturally occurring isomer; synthetically produced folate will have both and would thus be labeled “D,L” or “R,S.”  Merck was the first to manufacture a pure S isomer (aka L5MTHF), which it sold under the name Metafolin to customers who use it in vitamins and supplements.  This was an important and expensive-to-develop product for Merck.

Gnosis also makes raw dietary ingredients, and sold a product called Extrafolate (the current version of which was not at issue here).  Extrafolate was a mixture of the S and R isomer, also known as a “D,L5MTHF product.  This mixture doesn’t occur in nature and hasn’t been found to have the same nutritional benefits as L5MTHF.  As a mixture, it sells for significantly less than Metafolin—on the order of $10,000 less per kilogram.  But from 2006-2009, Gnosis used product specification sheets, brochures and other marketing materials that used the chemical descriptions, terms, and formulas attributed to the pure S-isomer for the sale of the R,S mixture.  Merck sued Gnosis in 2007.

In a bench trial, the district court found that Gnosis recognized that the mixture and the S-isomer had different chemical names, but deliberately used the S-isomer’s common name or abbreviation, and described the chemical properties of the pure product in its promotional material.  Gnosis continued this advertising until nearly two years after the litigation began.  The district court found that the use of the S-isomer’s common name and abbreviation was literally false, and that the use of descriptions of the chemical properties of the S-isomer in brochures, material safety data sheets, and certificates of analysis was literally true but implicitly, and intentionally, false.

“In essence, Gnosis was accurately describing a product it was not selling.”  Its statements were literally true in that they described the chemical makeup of a pure S-isomer product, but they were implicitly false because they were used in marketing a mixed product.  Comment: Hunh?  Literal falsity isn’t limited to claims that aren’t true of anything.  If I truthfully describe something that exists in the world, but I don’t have it to sell, selling something with that description isn’t literally true.  It’s literally false.  At the very least, the necessary implication of what Gnosis did was a literal falsehood.  No linguistically competent speaker of English would understand the materials as anything other than a representation that the product being sold was the pure S-isomer.  But because the literal/implicit falsity line has colonized other aspects of the doctrine, as we’ll see, this holding takes on exaggerated importance.

The district court found that Gnosis engaged in a “concerted and organized campaign to deceive customers.” Gnosis’s own expert testified that the pure isomer name is never used to refer to anything but the S-isomer.  Privately and in its patent application, Gnosis used the correct nomenclature.  The district court found Gnosis’s explanation for the labeling “simply fanciful—and false—and discount[ed] it entirely.”  Gnosis’s witness had testified that he’d read “two or three” articles that referred to the mixture substance as L5MTHF, but he could not recall the names of the articles.  The next trial day, Gnosis’s counsel informed the court that he’d reviewed the relevant documents, and that [t]hey do not . . .  exactly have the L5MTHF. The court concluded that “Gnosis’s use of the common name and abbreviation in its marketing efforts was a calculated decision to copy Merck’s advertising and capture a portion of Merck’s market share, knowing full well that its 6R,S Mixture Product was materially distinguishable from Merck’s pure 6S Isomer Product.”

On appeal, Gnosis challenged the district court’s presumption of consumer confusion and injury.  But under Second Circuit doctrine, a literally false claim may be enjoined without further evidence of impact on consumers.  And even in implied falsity cases, where a plaintiff shows intentional deception and egregious conduct, a presumption of deception arises.  Given the unchallenged factual findings of literal falsity and intentional deception as to the implicit falsity, a presumption of confusion was justified.  The intention to mislead was clear: “Gnosis put a description of the chemical properties of the Pure Isomer 6S product on its Extrafolate materials in order to mislead consumers into believing that they were, in fact, purchasing a Pure Isomer 6S product rather than the 6R,S Mixture Product, Extrafolate.”  (Query why such a presumption of confusion wouldn’t also be justified if a middleman was unaware that its supplier had deceived it.  The labeling and promotional materials would the same: a mixed product misidentified as pure.  At the very least, wouldn’t you want to call that literally false?)

When a plaintiff shows deceptive intent, the burden shifts to the defendant to show absence of confusion.  While the district court should’ve explained why Gnosis didn’t rebut the presumption, the record strongly supported a finding of actual consumer confusion.  There were only a few customers in this market, and there was evidence that several of them (as well as their own customers) were confused. Even assuming that some of Gnosis’s direct consumers weren’t confused, the record readily supported the conclusion that a significant number were misled.  Plus, the court of appeals noted, the district court’s finding of confusion could be based on literal falsity alone.

What about presuming injury to Merck?  Gnosis argued that this was only appropriate in cases of comparative advertising.  But in literal falsity cases, the Second Circuit doesn’t require extrinsic evidence of injury to consumers or to the plaintiff.  In McNeilab, Inc. v. American Home Prods. Corp., 848 F.2d 34 (2d Cir. 1988), the court of appeals held that misleading comparisons necessarily harm the victim in consumers’ minds, but misleading positive claims by the defendant injure all competitors equally, thus requiring some evidence of actual injury and causation.  The Time Warner case found that disparaging references to “cable” fell in the former category because Time Warner was cable in the relevant market.  The court of appeals now concluded that this rationale extends beyond disparagement when there is only one competitor in the market and when the deception is intentional.  “Because its only competitor for such a pure product at the time was Merck, it follows that Merck was damaged by Gnosis’s false advertising of a mixed product as a pure one.”  There was no risk that injury to Merck would be too speculative.  (I can’t see why intentional falsity is required here. Falsity would seem to work the same damage in a two-player market.)

Under the Lanham Act, a prevailing plaintiff can, subject to the principles of equity, recover defendant’s profits.  Damages can be trebled, as long as the award is compensatory and not punitive.  Willfulness is a prerequisite for a profit award in the Second Circuit.  And an award of profits may be made based on deterrence rationales, even though that’s not compensatory.  Under these standards, the district court didn’t abuse its discretion.  It reasoned that awarding profits was necessary to deter future unlawful conduct, prevent Gnosis’s unjust enrichment, and compensate Merck for its lost business, all of which are acceptable goals.  And where the parties directly competed in a two-player market, and literal falsity and willful deception have been proved, no more evidence was needed of injury and consumer confusion other than the resulting presumptions. “‘Having established falsity, the plaintiff should be entitled to both injunctive and monetary relief, regardless of the extent of impact on consumer purchasing decisions’” (citation omitted).

The district court also found that a profits award didn’t sufficiently reflect the total harm to Merck and trebled the damages.  This wasn’t a punishment or penalty, but reflected the intangible benefits to Gnosis, in particular its usurpation of Merck’s market share. The court of appeals found no abuse of discretion.  Although the statutory provision was intended to deal with hard-to-prove damages, deterrence of willful infringement is also an acceptable rationale. The facts of the case—including Merck’s exclusive control of the market prior to Gnosis’s entry, Gnosis’s continued false advertising for two years after the lawsuit began and after Gnosis’s sales agent settled Merck’s lawsuit against it and stopped distributing the mixture product—made this case “particularly appropriate” for enhanced damages. “Gnosis was unjustly enriched as a result of its false advertising, and, in light of Gnosis’s demonstrated deceptive and willful conduct—manifested by its stubborn persistence—the court’s conclusion that enhanced damages were needed to deter Gnosis from any future willful infringement was not an abuse of discretion.”  Other cases with less egregious, willful conduct might not warrant such an award.

The court of appeals also affirmed the grant of prejudgment interest (justified by willfulness again) and mandatory corrective advertising.  The corrective ads had to disclose that the campaign was court-ordered, but not that Gnosis was found to have acted willfully.  The ads had to link to the court’s opinion for context, and had to run on Gnosis’s homepage as well as product sale pages, as well as on thirdparty industry websites and in trade magazines where the offending products were or are presently advertised by Gnosis. Gnosis argued that this was unfair double recovery, but the court didn’t award Merck any damages for corrective advertising.

The court of appeals also affirmed the district court’s finding that this was an exceptional case warranting a fee award of nearly $2 million. The district court not only found that the false advertising was willful, but that “Gnosis’s litigation strategy was conducted in bad faith, with senior officials, including [CEO] Berna, frustrating the litigation process at every turn, from withholding documents in discovery and obstructing depositions to testifying falsely under oath at the bench trial in this action.”  The court of appeals noted that Gnosis had been found to engage in “egregious discovery violations,” including coaching of a witness during a deposition by the CEO, and commented that the litigation was clearly prolonged by Gnosis’s conduct.  This award was appropriate even though the fees outstripped the damages; the district court found the hours (and rates, after a reduction) reasonable.  The district court also pointed out that the actual stakes of the case (market share for one of Merck’s flagship products), as well as the level of success counsel achieved, helped justify a fee award that was substantially larger than the award of profits.  Merck didn’t get fees and costs on appeal, however, because Gnosis conducted itself appropriately in the appeal and its arguments were nonfrivolous.