En banc court again strikes down sugar-sweetened beverages warning, with divided and divisive reasoning

American Beverage Ass’ v. City & County of San Francisco, No. 16-16072 (9^th Cir. Jan. 31, 2019)

After the panel opinion striking down SF’s sugar-sweetened beverages (SSB) disclosure was reheard by the en banc court in light of National Institute of Family & Life Advocates v. Becerra (NIFLA), 138 S. Ct. 2361 (2018), the en banc court again struck it down, though not with enough force for a vigorous special concurrence in the judgment.  To the majority, the disclosure was an “unjustified or unduly burdensome disclosure requirement[] [that] might offend the First Amendment by chilling protected commercial speech” under Zauderer v. Office of Disciplinary Counsel, 471 U.S. 626 (1985).

The ordinance required certain SSB ads within the city to state: “WARNING: Drinking beverages with added sugar(s) contributes to obesity, diabetes, and tooth decay. This is a message from the City and County of San Francisco.” Traditional media ads and some other things, like small logos, were excluded. The warning had to occupy at least 20% of the advertisement and be set off with a rectangular border.

Under Zauderer, the government may compel truthful disclosure in commercial speech as long as the compelled disclosure is “reasonably related” to a substantial governmental interest. The Supreme Court vacated and remanded CTIA–The Wireless Ass’n v. City of Berkeley, 854 F.3d 1105 (9th Cir. 2017), in light of NIFLA, but that didn’t make a difference to this case.  CTIA “examined a similar health and safety warning and held squarely that Zauderer provides the proper analytical framework for considering required warnings on commercial products,” and is not limited to disclosures that correct falsity. 

“NIFLA did not address, and a fortiori did not disapprove, the circuits’ precedents, including CTIA, which have unanimously held that Zauderer applies outside the context of misleading advertisements.”

Indeed, “NIFLA preserved the exception to heightened scrutiny for health and safety warnings. The Supreme Court made clear that it was not calling into ‘question the legality of health and safety warnings long considered permissible.’” Judge Ikuta’s special concurrence read this language to mean that only health and safety warnings of ancient origin could avoid heightened scrutiny, but “the most natural reading of this passage” was as a reference to health and safety warnings in general, especially given that the NIFLA majority was answering Justice Breyer’s dissent listing examples such as warnings about seat belts and about the availability of whooping cough vaccine. “There is no indication that either of those required disclosures is of ancient origin or that the majority intended some health and safety warnings (if accurate, uncontroversial, and not unduly burdensome) to be precluded merely because the knowledge that the warnings convey is new.”

Under Zauderer, a compelled notice must be (1) purely factual, (2) noncontroversial, and (3) not unjustified or unduly burdensome. SF failed to meet its burden on (3). SF’s argument that the border and 20% size requirements adhered to best practices for health and safety warnings was unpersuasive. Some tobacco and prescription warnings must occupy at least 20% of those products’ labels or advertisements, and SF’s expert concluded that larger warnings are more effective. But the record here showed that a smaller warning would achieve SF’s stated goals; SF’s expert cited and discussed a study that examined a warning similar to that required by the ordinance, but covered only 10% of the ad. SF failed to show that the contrasting rectangular border containing a warning that covers 20% of the advertisement didn’t “drown[] out” Plaintiffs’ messages and “effectively rule[] out the possibility of having [an advertisement] in the first place.”  A 10% warning wouldn’t necessarily be valid, and a 20% warning isn’t necessarily invalid, but on this record, SF didn’t carry its burden.

Judge Ikuta dissented from most of the reasoning. NIFLA superseded Zauderer by holding that government-compelled speech is a content-based regulation of speech. (The majority is obviously right that this isn’t new; what’s relatively new is the suggestion that there are some commercial speech regulations that are other than content-based, always contrasted with the content-based commercial speech regulation actually before the court in a given case, combining poisonously with the post-Reed idea that all content based regulations are the same variety of bad.)  Content-based regulations are presumptively unconstitutional. After NIFLA, governments may not “impose content-based restrictions on speech without persuasive evidence of a long (if heretofore unrecognized) tradition to that effect.”

Zauderer now only applies to “some laws that require professionals to disclose factual, noncontroversial information in their ‘commercial speech.’”  Anyway, the factual accuracy of this warning wasn’t undisputed. (Will it ever be?)  The FDA says that added sugars are “generally recognized as safe,” and “can be a part of a healthy dietary pattern” when not consumed in excess amounts.  The disclosure was also not “uncontroversial” (will it ever be?) because the record showed this was a controversial topic.  Nor did the warning “relate to the terms on which the advertisers provide their services,” which Judge Ikuda took to be a NIFLA-added limit on the reach of Zauderer, because soda is a product and not a service and the warning doesn’t address the terms of the deal. [This reasoning is just bad; (1) there is no reason to distinguish disclosures for services from disclosures for products for constitutional purposes (I think for Judge Ikuda this is a collateral consequence of the idea that Zauderer is now only for professional speech and not for ordinary commercial speech, but professionals can sell you products as well as services), and (2) the “terms” here should include “what you’re getting for your money,” and the disclosure relates to the components of SSBs.] Judge Ikuda did agree that the warning was too burdensome, as well.

And the majority here was wrong to say that “NIFLA preserved the exception to heightened scrutiny for health and safety warnings,” because only “health and safety warnings long considered permissible” were excepted. “NIFLA did not specify what sorts of health and safety warnings date back to 1791, but warnings about sugar-sweetened beverages are clearly not among them.” [Look, the majority clearly has the better of this argument; the sensible meaning is that health and safety warnings have long been considered permissible, not that warnings about poxy air would be ok because they’re old but warnings about PCBs would be problematic because they’re new. Even Justice Scalia was very clear (in his Lucas v. South Carolina Coastal Council opinion) that newly discovered facts about the nature of the physical world can trigger old doctrines allowing regulation.  However, the not unrelated question about what constitutes a “factual and uncontroversial” disclosure may thwart the majority’s attempt to preserve health and safety warnings, insofar as it’s usually possible to find someone to contest any factual proposition.  [But of course, when it comes to abortion/ “informed consent,” the government can require the disclosure of all sorts of things, without a “factual/uncontroversial” requirement, because that’s different.]]

Without Zauderer’s cover, the mandatory disclosure flunked intermediate scrutiny because the warming wasn’t sufficiently tailored to the government interest in warning of the public health dangers from drinking SSBs.  It was underinclusive—it didn’t apply to all SSBs, much less all sugar-sweetened products, and it didn’t apply to many ads for covered SSBs. Moreover, “San Francisco could disseminate health information by other, less burdensome means, such as a less intrusive notice or a public health campaign.”

Judge Ikuda also objected that  the majority failed “to provide any guidance regarding when a warning is unjustified or unduly burdensome.” Instead of considering the “totality” of the requirements, the majority considered only whether a smaller warning would accomplish the city’s stated goals. It should have tested the regulation against Central Hudson scrutiny, rather than reasoning that if it couldn’t meet Zauderer’s exception then it couldn’t pass Central Hudson.

Judge Christen and Chief Judge Thomas concurred in part and in the judgment. They would apply Zauderer, but would have reasoned that SF couldn’t show that the speech at issue was “purely factual.”

 “[W]here, as here, the parties disagree about the veracity of compelled speech, the court should begin by asking whether the government’s message is objectively true,” because an assessment of controversiality or burdensomeness will often entail much more subjective judgments. 

To the concurrence, there was no evidence that SSBs “contribute[] to” obesity, diabetes, and tooth decay, because “diabetes” is an umbrella term referring to both type 1 diabetes and type 2 diabetes and the causes of type 1 diabetes are unknown, so the research linking SSBs to type 2 diabetes is irrelevant.  [This reasoning is unimpressive.  Drinking and driving contributes to car accidents even though it doesn’t contribute to every car accident or even every kind of car accident.]  The message was therefore also controversial. That provided a more objective basis for invalidating the law than its burden.  [The concurrence doesn’t notice that its subjectivity comes in when interpreting the semantic meaning of the required disclosure—a weakness that also came up in the case invalidating the initial tobacco warning images for putatively conveying a factual message that smoking always causes autopsies.]

SF argued that a reasonable person would understand the reference to diabetes as “the kind of diabetes that can be caused by overconsumption of sugar,” but that argument was “in tension with the goal of having a public health message understood by the maximum number of consumers, not just those with sophisticated levels of health literacy. Because the message would be conveyed to sophisticated and unsophisticated consumers, we must read it literally.”  [I guess in the next round of this we get to debate whether “contributes to” communicates something like “always causes” or “uniformly raises everyone’s risks the same amount” or something like that.  The point of the message was not to educate consumers about the different types of diabetes; it was to provide them with actionable information about the relationship between SSBs and the disease state of diabetes.  If they end up knowing the latter without being aware of the former, they’re not misled.  I find it odd to treat the warning as a purely abstract item of information rather than information that might affect actions, as was its point.]

The concurrence defends its “persnickety” insistence on factuality not only because of the constitutional rules, but because of the social consequences of government-mandated messages, given the pervasive stigmatization of people with type 1 diabetes as having a disease that is the result of a lack of personal responsibility.  [Again, it’s completely unclear to me how the warning would worsen this problem; assume it said (correctly, as far as the concurrence is concerned) that SSBs are associated with type 2 diabetes.  People would still need to learn the difference between type 1 and type 2 independently of the warning.  Unless the thought is that the mandatory warning should also educate them about type 1 diabetes and perhaps caution them against shaming anyone with diabetes?]

Further, the warning was “problematic” because it suggested that sugar was always dangerous for diabetics, whereas consuming SSBs can be medically indicated for a type 1 diabetic when there are signs of hypoglycemia, to raise blood sugar levels quickly.

Judge Nguyen concurred also, arguing that Zauderer is only for disclosures that correct false/misleading speech and that SF’s warning flunked Central Hudson.  Applying Zauderer was particularly bad because the SSB warning was a content-based regulation.  [Sigh.]