FDLI Symposium: Constitutional Challenges to FDA Law & Regulation
Session 1: Compelled Speech
Moderator: Allison Zieve, Director, Public Citizen Litigation Group, Vice-Chair, FDLI
First Amendment Limits on Compulsory Labeling
Nigel Barrella, Sole Practitioner, Washington DC
Review of commercial speech doctrine. Product labels are commercial: propose a transaction. Other FDA “labeling”? Labeling not sold direct to consumers, such as Rx drugs, vaccines? Package inserts? Medical literature? (Caronia, Amarin case finding that it was unconstitutional to regulate off-label promotion by the manufacturer.)
DC Circuit’s rejection of tobacco images trying to promote an emotional response; 6th Circuit pointed out that facts can provoke emotions, but that doesn’t make them opinions.
Ongoing dispute: what does “uncontroversial” mean? At a minimum, controversial can’t mean that someone was willing to sue.
Ongoing cases: genetic engineering compulsory labeling for foods; on appeal in 2d Cir. Argument: topic is simply too controversial for labeling. Vt. says the test is whether the disclosure is factual, not opinion-based, not reasonably disputable. Abortion regulations provide precedent for saying that extremely controversial subject matter can’t be subject to labeling, no matter how true it is. If that’s the test, though, then tobacco labeling shouldn’t have been allowed; total fat etc. labeling shouldn’t be allowed.
Some cases suggest you can’t tell consumers what to do: “try to quit smoking.” But what about “Surgeon General’s Warning”? What about “This statement hasn’t been evaluated by the FDA”?
Future of Zauderer: uncontroversial seems to be gravitating to being fact-based; can’t tell consumers what to do but can provide them with consensus. Substantial interest in things other than preventing deception as justification for disclosure?
Discussant: Bert Rein, Founding Partner, Wiley Rein
Zauderer: easy cases make bad law. Deceptive by omission. If the alternatives are ban or cure with disclosure, disclosure is what makes sense. Listerine remedy case: disclosure that it doesn’t kill germs corrects the falsehood you’ve been disseminating.
Safety warnings: state’s interest is truth in market/health and safety of consumers. That’s a reasonable interest, and the temptation is to say “why not”? Why not poison, electrical hazard, flammability warning—state has a compelling interest no matter what standard you apply. [Hunh? Why isn’t public education a less restrictive alternative?] But you can’t use that foundation of health, safety, fairness to go beyond and start expressing opinions. Also important to have standardized ways of defining terms, like “gallons” or the names of foods—fairness/fair choice is an important interest. Another interest: state using compelled speech to carry out a regulatory regime. Labeling on OTC drugs—comprehensive responsibility to take care of health and safety in drugs. Regulatory interest can be very important, and as part of that it may need to regulate labeling. That is a generally recognized legit interest, but what is the limit on that? 44 Liquormart and compounding cases: you can use speech regulation as part of a regulatory regime, but only as a last resort. Is this essential/can’t be done another way? Is it a reasonable advancement of the interest the gov’t is serving?
The AMI case is therefore wrong. There is no general regime governing safety [of meat?]. Same with SEC disclosure of conflict minerals—no relation to the overall regime. [What counts as a general regime?] Parsing the conflict minerals disclosure is itself very difficult.
Suppose the gov’t was worried that people were ignoring Christmas. Could the gov’t require all food sold in November to say “remember Xmas is coming on Dec. 25th”? [No, because of the Establishment Clause, not b/c of a general restriction on gov’t position-taking.] That’s inconsistent w/1st Amendment b/c inconsistent of your right to determine what you want to say, though it’s truthful and beyond dispute (unless you are Greek Orthodox). [Wow, where to start. Okay: let me just say that the “you” is an important issue here. Does the commercial speaker have interests of its own, or just interests in informing consumers?] Gov’t shouldn’t instruct people what to buy. State has no legit interest in telling you (food seller) to put that message on food. If the state can require GMO disclosure, then it can also require you to disclose that there’s no benefit to GM-free food that’s been shown. [Yes, I believe that’s the teaching of the modern understanding post-Lochner: the legislature in general gets to make these calls. I’m not clear on why that’s the troublesome outcome of the slippery slope.]
Authors: You Want a Warning with That? Sugar-Sweetened Beverage Warnings and the Constitution
Sabrina Adler, ChangeLab Solutions, Oakland, CA (co-authored)
Increase in obesity; 135% increase in calorie intake from sugar-sweetened beverage, half of Americans over age 2 drink at least one a day. 63% of HS students. Connection between liquid sugar consumption and greater health problems v. solid food. Safety warning issues? First Amendment, preemption under NLEA, and dormant commerce clause.
Paper argues that the science justifies special treatment for SSBs, and thus that warnings should be analyzed under Zauderer. Current agreement under courts of appeals: Zauderer is not limited to the potential for consumer deception, but can extend to the protection of public health. Even if substantial interest is required, health is such an interest. No disclosure has ever been struck down under Zauderer.
However, the evidence that labels decrease consumption (much less improve health outcomes) is weaker than the evidence that labels increase awareness of the risk. Would be at risk if strict scrutiny were applied; there are less restrictive means of conveying the info such as public education. (Or we could, you know, ban SSBs without running afoul of the First Amendment.) What about the possibility of gov’t speech? Forcing message on someone else’s label—boundaries have not been spelled out, but most proposed labels do attribute message to the gov’t.
NLEA has express preemption, no implied preemption. No definition of what constitutes “nutrition labeling.” They argue that it should mean the quantitative info required on the label, not qualitative info. Thus shouldn’t be preempted by nutrition labeling. Nutrient claim/Health claim: relationship between ingredient and health conditions. This is technically what an SSB warning would do. But we argue that “claims” refers to positive statements made by producers to increase sales—that was what Congress was concern about. Regs specify a list of permissible health claims, and they’re all positive. Safety warning exception to preemption? Lactose (not a warning); BPA in baby bottles (FDA specifically exempted from labeling regulation, and thus warning exception couldn’t apply); coloring in colas that California added to Prop. 65/potential carcinogen list (FDA specifically found it to be GRAS/generally recognized it as safe—court found that unrelated to labeling, but predicate to labeling; thus this type of warning shouldn’t be preempted and was subject to safety warning exception to preemption).
Stuart Pape, Shareholder, Polsinelli
First Amendment never came up in the 1970s at the FDA—pre Virginia Pharmacy. Doesn’t envy the task now. How do you make sense out of current doctrine combined with regulations that have been around for decades? How do you advise a client? The principles aren’t clear enough.
If SSB warning has to be on 20% of billboard/other ads, how does that work? It’s not scientifically apparent that two products—12 oz. of 100% organic apple juice, which SF favors b/c it’s not regulated, and equivalent amount of cola, which does get regulated—are different; the apple juice even has more calories. So how does the city conclude there’s clear evidence that one product should get a warning and the other shouldn’t? That is part of what makes the warning constitutional. FDA is currently considering an “added sugars” line on the nutrition panel, and establish a daily reference value. But the evidence to support that is weak/unscientifically justified, according to past FDA heads. If there’s serious controversy over mere disclosure of amount/reference point that’s subject to the warning, the warning itself is controversial—singles out one type of beverage and forces them to say “this is a bad product.”
Court rejected disclosure about radiation from cellphones for similar reasons.
Also hard to demonstrate substantial interest b/c you require warnings for some things/not other things (the apple juice)—swiss cheese approach makes it harder to defend. Also, it’s just an opinion: we have to do something about obesity; the scientific evidence about health f/x is compelling (he thinks both of these are opinions).
Discussion among panelists: anti-regulationists say alcohol label may be ok b/c of the scientific evidence that it is really bad, not like other things, which is an interesting thing to be assessing under First Amendment standards.
Pape: Amarin case: the court observes that the entire premarket approval system for pharma is at risk under current standards. While pharma industry would like more freedom to communicate w/doctors, they’re not looking to have a system in which pharma reg is more like dietary supplement reg, which is a free for all. Bears watching. (Careful what you wish for? The First Amendment doesn’t enact Mr. Herbert Spencer’s Social Statics, and neither does it enact Bayer’s wishlist. Welcome to your new competitor, GNC.)
Zieve: Congress, in response to health tragedies, organized the regulatory structure: illegal until approved, and then approved only for conditions it’s safe and effective for according to FDA. FDA doesn’t regulate the practice of medicine; FDA’s authority is only over manufacturers. If you promote it for an unapproved use, that’s illegal.
Rein: asymmetry: unlawful to tout, but lawful to use—that doesn’t work.
Zieve: but the doctor isn’t engaging in off-label promotion.
Rein: what bothered ct in Amarin was that the manufacturer was disseminating information based on well-controlled studies. If studies had been done by independent third parties, would have been able to disseminate them. May have promotional effect, but is scientific speech. Disparity of regulation/intervention. Nike v. Kasky: the same asymmetry. [No, a different one.] Kasky is regulated only by defamation standard when he says Nike uses child labor; Nike’s speech is commercial when it denies it does.
Zieve: In your remarks before you mentioned a valid warning as integral to regulatory process—so did you actually mean that? Vt. did a study showing 90% of consumers wanted to know whether the food they bought had GMO ingredients—would that matter?
Barrella: the whole reason that we extended protection to commercial speech was consumers’ interest in getting truthful information. DC Circuit en banc did say that consumer interest was substantial interest in itself for country of origin information.
Rein: overwhelming support is poor indicator of constitutionality; NH residents “overwhelmingly supported” putting “Live Free or Die” on license plates.
Adler: commercial speech also makes a difference from Live Free or Die.
Q: What about the standards of identity? Hampton Creek’s “Just Mayo”—barred on standards of identity, which is integral to a regulatory process, but what’s the compelling interest in saying you can’t label your product “Just Mayo” just because it isn’t mayo according to the standards of identity.
Barrella: it’s a Central Hudson question.
Zieve: this is false and deceptive advertising. People think mayo means a certain thing; it’s like calling Coke orange juice. Standard of identity makes the deception stronger, but calling it just mayo divorces “mayonnaise” from its meaning.
Pape: maybe historically people knew what mayo was; standard of identity developed b/c producers were making things that people had once made at home. Congress wanted consumer to be able to go to store and pick up what he/she would have previously made in the kitchen. The standards of identity are definitions. “Cheese food.” It’s not cheese. Signal that it’s not cheese. Same thing with mayo. There is something called “mayonnaise dressing”: like mayo, but not. Interesting to consider whether standards of identity are now vulnerable to the First Amendment. [Yeah, interesting’s one word.]
Q: soy milk = milk? Industry was laughed out—why isn’t this the same question? Soy milk has an ingredient list.
Zieve: why would you read the ingredient list? Soy milk isn’t called “just milk.”
Rein: State puts up these markers to provide a basis for comparison in the marketplace. It’s meaningless unless it’s enforced. If std of identity for mayo is legit exercise of gov’t power, enforcement is ancillary to that fundamental Q.
Q: where in the First Amendment analysis do we deal with false/misleading speech, trademark, Lanham Act?
Barrella: that’s just outside the First Amendment.
Adler: there may be affirmative misrepresentations in our areas, but we’re more worried here about a lack of information. Slightly different issue.
Rien: Zauderer is a case of fraud by omission.
Zieve: Also consider influence of POM v. FTC, where DC Circuit said that a 2 clinical trial standard was too onerous. Pom says that if it has one study, it can make those claims, even if other studies show no effect. Can drug companies take advantage of this? Keep watching—cert petition just filed.