Friday, December 14, 2012

Judge Rogers' dissent: image requirements are justified

R.J. Reynolds Tobacco Co. v. Food & Drug Admin., 696 F.3d 1205 (D.C. Cir. 2012)

Judge Rogers, in dissent, had a very different view of the government’s evidence and powers.  The majority wrongly disregarded both the tobacco companies’ history of deceptive advertising and the government’s stated goal: to effectively convey the negative health consequences of smoking.  Given that the warnings presented factually accurate information and combated misleading commercial speech, Zauderer was the appropriate test, and the government only had to show that the warning label requirement was reasonably related to its substantial interest in effectively conveying information to consumers.  Even treating Zauderer as limited to consumer deception cases, judicial, congressional, and FDA findings substantiated the interest in preventing deception here.

The FDA’s interests were complementary, but distinct: effectively conveying information about smoking’s dangers and decreasing smoking rates.  The majority didn’t buy the first interest, but it was there. “Although there are statements in the rulemaking record regarding the government’s interest in reducing smoking rates, nothing in that record, much less the White House press briefing cited by the court, suggests these statements were intended to override the clearly stated interest in effectively communicating information about the negative health consequences of smoking to consumers.”  Even the D.C. Circuit had recognized in the past that the government’s interest in preventing consumer confusion may take on added importance in the context of a product that affects health.  Tobacco is dangerous when used as intended, and highly addictive.  Given the informational interest involved, the government met its burden under Central Hudson or Zauderer, except for the requirement to put 1-800-QUIT-NOW in each label.

The evidence was clear that most people, especially teens, still lack a complete understanding of smoking’s harms and the difficulty of quitting once they start; over 80% of adult smokers became addicted at or under 18.  And tobacco use is the leading preventable cause of death in the US.  In addition, the tobacco companies engaged in a decades-long campaign of deception over these facts (of which more in a later post), and are still spending billions to attract new smokers and retain current ones.  The current warning labels, last revised in 1984, don’t work.

The FDA chose its nine graphic images based on measures of their salience to adults and teens, including “[e]motional reactions, cognitive reactions, and [reactions as to] whether the warning was difficult to look at.”  The reason, the FDA explained, is that “the literature suggests that risk information is most readily communicated by messages that arouse emotional reactions, and that smokers who report greater negative emotional reactions in response to cigarette warnings are significantly more likely to have read and thought about the warnings . . . .” Thus, the graphic warnings effectively communicated the negative health consequences of smoking.

The dissent turned to the “compelled speech” cases.  Zauderer makes clear that it is fine to prescribe “what shall be orthodox in commercial advertising,” as opposed to politics or religion.  Sorrell didn’t change that or overrule Central Hudson, and in any event the objection to the law in Sorrell—that the fear that people will make bad decisions with truthful information can’t justify content-based burdens on speech—was inapplicable here, where the tobacco companies were the ones trying to avoid truthful speech.  It defies Zauderer to say, as the majority did, that “attempts to regulate ‘what shall be orthodox in . . . matters of opinion’ — i.e., whether individuals should buy and use a lawful product — must be subject to strict scrutiny.” 
 
As the dissent pointed out, “because matters of opinion over whether individuals should buy and use a lawful product fall squarely within the domain of commercial advertising recognized by the Supreme Court, the regulation thereof is not … subject to strict scrutiny.”  Because the informational value of speech to consumers is the primary justification for protecting commercial speech at all, disclosure requirements don’t do as much harm to an advertiser’s legitimate interests as bans on commercial speech.  Thus, Zauderer holds that “an advertiser’s rights are adequately protected as long as disclosure requirements are reasonably related to the State’s interest in preventing deception of consumers.”  (Deception was sufficient, but not necessary; in view of the consumer deception and confusion about tobacco, there was no need for the dissent to determine whether some other interest might also support required disclosures.)

Zauderer, and subsequently Milavetz, requires only a possibility of deception or tendency to mislead.  Quoting Milavetz, “where ‘the likelihood of deception’ is ‘hardly a speculative one,’ the government need not produce ‘evidence that [the] advertisements are misleading,’ as the court may rely instead on experience and common sense.”  Milavetz involved law firm ads that the Supreme Court found inherently misleading because they “promise[d] . . . debt relief without any reference to the possibility of filing for bankruptcy, which has inherent costs.”  Without extrinsic evidence, the Court found this omission sufficiently misleading to trigger Zauderer.

Likewise, the DC Circuit in Spirit Airlines, Inc. v. Dep't of Transp., 687 F.3d 403 (D.C. Cir. 2012), applied Zauderer and upheld a rule requiring the most prominent number displayed in airline ads to be the total price, inclusive of taxes.  Even though the airlines had been complying with existing rules, the court accepted the Department of Transportation’s determination, “based on common sense and experience, ‘that it was deceitful and misleading when the most prominent price listed by an airline is anything other than the total, final price of air travel.’”  To the dissent, “[e]ven absent any affirmatively misleading statements, cigarette packages and other advertisements that fail to display the final costs of smoking in a prominent manner are at least as misleading” as airline ads, given the evidence that present warnings are ineffective.

Warner-Lambert Co. v. FTC, 562 F.2d 749 (D.C. Cir. 1977), also supported the FDA’s requirements.  The majority argued that the FDA wasn’t claiming that the images were remedial measures designed to counteract specific deceptive claims.  But even if Warner-Lambert were limited to remedial measures, Congress had provided the appropriate rationale.  Congress specifically found that “[t]obacco product advertising often misleadingly portrays the use of tobacco as socially acceptable and healthful to minors.” And “the FDA did frame its rule as a measure designed to counteract specific gaps in consumers’ knowledge of the health risks of smoking — gaps that align with specific deceptive claims made by the tobacco companies.”  Warner-Lambert also recognized that ads, though not misleading if standing alone, can become misleading in light of past ads.  Indeed, the D.C. Circuit also affirmed a finding that it was reasonably likely, notwithstanding the new restrictions in the Tobacco Control Act, that the tobacco companies would commit future RICO violations.  (More on this soon!)  Capitalizing on decades’ worth of prior deceptions, as the tobacco companies do every day, was enough to create a tendency to mislead justifying a disclosure. 

In addition, the nature of required disclosures, which do not suppress any information, inherently requires less scrutiny; though the tobacco companies argued that they wouldn’t be able to send their own messages on their packaging, “their objection rings hollow in the absence of any evidence of difficulty in conveying their desired messages, notwithstanding a decade of experience under a similar warning label requirement in Canada.”

The textual statements were factually accurate.  The question was whether the images rendered the warnings nonfactual or controversial.  The court should’ve viewed the images in connection with the text in order to answer that question.  The use of graphic images, “even if digitally enhanced, illustrated, or symbolic,” doesn’t necessarily make warnings nonfactual. Zauderer itself recognized that “[t]he use of illustrations or pictures in advertisements serves important communicative functions: it attracts the attention of the audience to the advertiser’s message, and it may also serve to impart information directly.”

The FDA concluded that the effects shown in the images are, in fact, accurate depictions of the effects of the sickness and disease caused by smoking, and the tobacco companies didn’t even argue otherwise.  “That such images are not invariably comforting to look at does not necessarily make them inaccurate.”  The consequences of smoking are disturbing, and the images reflect that fact.  The tobacco companies complained that the images were not chosen to convey information but to evoke negative emotions and discourage smoking.  It’s not a binary.  The evidence showed that warnings that elicit emotional reactions are better processed by consumers.  “Thus, the FDA’s reliance on salience measures was in the service of — not inconsistent with — the warnings’ informational purpose. Moreover, factually accurate, emotive, and persuasive are not mutually exclusive descriptions.”  This wasn’t like Entertainment Software Association v. Blagojevich, 469 F.3d 641 (7th Cir. 2006), which involved a much more subjective and exclusively nonfactual determination about labeling video games deemed to be “sexually explicit.”  The video game labels were nonfactual “because there were no facts to convey.”

It’s certainly true that comprehending the facts about smoking “is likely to provoke emotional reactions and also to discourage the use of cigarettes.”  But the tobacco companies’ position meant that

the more concerning the negative health effects of a particular product, the more constrained the government is in mandating disclosures of those facts. Unsurprisingly, the tobacco companies point neither to any case law in support of this argument nor to any legally significant distinction between fact and emotion. Rather, the greater the harms to public health, the greater the government’s interest in informing consumers of those harms.

This interest was particularly great in view of the tobacco companies’ history of deception.

The tobacco companies specifically objected to five of the nine selected images. They contended that the images of a man smoking through a tracheotomy hole in his throat and a man with chest staples on an autopsy table convey misleading messages about the consequences of smoking, and that the images of a man wearing a t-shirt reading “I QUIT,” a baby enveloped in smoke, and a woman crying convey no information about the consequences of smoking whatsoever.  These arguments ignored the accompanying text and didn’t address whether the overall message of the warning labels was nonfactual.  It wasn’t.  For example, the image accompanying the textual warning “Cigarettes are addictive” depicts a man smoking through a tracheotomy opening in his throat. “Viewed with the accompanying text, this image conveys the tenacity of nicotine addiction: even after under undergoing surgery for cancer, one might be unable to abstain from smoking.”  In fact, 50% of neck and head cancer patients continue to smoke—this image didn’t depict an extreme or unusual situation. 

Likewise, the autopsy image worked with the textual warning “Smoking can kill you.”  Though autopsies might not be “a common consequence of smoking,” per the majority, “neither are coffins or gravestones; yet the status evoked by images of an autopsy-scarred man, a coffin, or a gravestone — death — is a common consequence of smoking. The FDA might have opted for an image of a decaying cadaver or of a pile of ashes to portray the likely physical consequences of smoking, but it was not limited to such images in its representation of those consequences.”  The autopsy scar communicates that the man in the image is dead.

Similarly, “[t]he images of a baby enveloped in smoke and a woman crying both depict the significant harms of secondhand smoke.” They accompanied the textual warnings “Tobacco smoke can harm your children” and “Tobacco smoke causes fatal lung disease in nonsmokers,” respectively.  The former image clearly informed parents that smoking in the presence of children would lead the children to inhale toxins, while the latter highlighted the emotional suffering caused by fatal lung disease and other harms of secondhand smoke, which causes 3000 lung cancer deaths each year among nonsmokers in the US and 38,000 deaths overall. These labels conveyed the message that smoking poses risks not just to smokers but also to family members and others.

The dissent was less impressed by the image of a man wearing a t-shirt that reads “I QUIT,” though it addresses the benefits of quitting: as paired with the textual warning “Quitting smoking now greatly reduces serious risks to your health,” it conveys the message “I quit, and I am alive and healthy.”  Nothing in the image affects the factual, noncontroversial textual warning, and thus the overall label was a factually accurate, uncontroversial disclosure.

Under Zauderer scrutiny, the only question was whether the labels were reasonably related to the government’s interest in effectively conveying the health harms of smoking, and the dissent thought this was done: “In view of the scientific literature supporting the FDA’s reliance on the salience measures reported in its study, the warning label requirement is reasonably related to the government’s interest in effectively communicating information about the negative health consequences of smoking.”  Size and placement were also supported by the literature, many countries’ experience, and common sense.  The majority distinguished Zauderer because the warnings here involved “elements of compulsion and forced subsidization.”  But that’s a feature of commercial disclosure requirements.

However, the dissent would have found that the addition of “1-800-QUIT-NOW” on each warning label “does not directly disclose factual information about the health consequences of smoking.”  Since this requirement was directed towards making it easier to quit (how this differs from raising the salience of quitting is not 100% clear to me), it required analysis under Central Hudson, not under Zauderer. 

Even applying Central Hudson to everything, the majority should have looked at both government interests, conveying true information and decreasing smoking, not just at the latter. The informational interest in effectively conveying the negative health consequences of smoking was clearly substantial, especially given its health harms.  And the record had substantial evidence that the warnings would directly advance that interest, not least by “ensur[ing] that the health risk message[s] [are] actually seen by consumers in the first instance.” “[H]istory, consensus, and ‘simple common sense’” also showed that the requirement satisfied Central Hudson’s fit requirement.  The failure of past warnings, similar to the alternatives now suggested by the tobacco companies, were sufficient to show that the graphic warnings were not more extensive than necessary to serve the government’s substantial interest.

The only exception was “1-800-QUIT-NOW” telephone number. The government’s interest in reducing smoking rates was substantial.  Providing the number would directly advance this interest, since quitting on one’s own is very hard, while quitlines have been found to “significantly increase abstinence rates.” International experience referenced in the rulemaking, further supports the common sense proposition that informing smokers of cessation resources is likely to increase rates of successful quit attempts.   But including the number was more extensive than necessary to serve that interest, because it was presented in imperative terms: QUIT NOW, a command directly contradicting the tobacco companies’ desired message at the point of sale. This was too great a burden.  Alternative means, such as a package insert, had to be considered.

The dissent concluded by emphasizing the basis for First Amendment protection of commercial speech: the interest in truthful information, which the warnings served, especially given the tobacco companies’ history of deception.  “[N]othing in the Supreme Court’s commercial speech precedent would restrict the government to conveying these risks in ways that have already proved ineffective or would prohibit the government from employing the communication tools tobacco companies have wielded to great effect over the years.”

Final note: I understand that a fair number of my readers are more positive about the benefits of expansive trademark protection than I am (and thanks for reading anyway!). I’m enough of an Order Muppet that inconsistency can be worse, to me, than simple wrongness. Let’s set aside abortion warnings as too obvious and politicized (although their First Amendment offensiveness is thereby increased, not decreased!) and just consider trademark. Aside from counterfeiting, does anyone believe that the empirics of plain-vanilla infringement—much less dilution—can be proven to the standard applied by the majority here? Among other things, confusion studies routinely show a not insignificant baseline level of confusion even for the “control,” suggesting that there’s an irreducible level of consumer misunderstanding that trademark law can’t fix. How is that different from the weaknesses identified by the majority in the FDA’s evidence?

No comments: