Our recent post about the First Amendment decision in American Beverage Ass’n v. City & County of San Francisco, ___ F.3d ___, 2019 WL 387114 (9th Cir. Jan. 31, 2019) (en banc) (“ABevA”), holding unconstitutional a purported product “safety warning” was more than enough to set the old First Amendment juices flowing.  Ninth Circuit en banc decisions involve the largest panels of judges in the country, and ABevA featured no less than four different interpretations of the recent United States Supreme Court decision, Nat’l Institute of Family & Life Advocates v. Becerra, 138 S. Ct. 2361 (2018) (“NIFLA”), so we thought we’d start by examining NIFLA, looking for First Amendment gold.

Briefly, NIFLA addressed the constitutionality under the First Amendment of a statute aimed at certain unlicensed “crisis pregnancy centers” run by anti-abortion activists who were more interested in interfering with their “client”s best interests than helping them.  138 S. Ct. at 2368. The State of California enacted a statute requiring those clinics to post notices containing information about the availability of competing entities offering real services that also included abortion as an option.  Id. at 2369.  The challenged law also compelled disclosure of these clinics’ unlicensed status.  Id. at 2370.

A five-justice majority in NIFLA held the statute unconstitutional as mandating “compelled speech” – in the form of supposed health-related warnings.  Citing, inter alia, one of our favorite cases, Reed v. Town of Gilbert, 135 S. Ct. 2218 (2015), the majority found “[t]he licensed notice [to be] a content-based regulation of speech.  By compelling individuals to speak a particular message, such notices alter the content of their speech.”  138 S. Ct. at 2371.  The Court then rejected the idea of “professional speech” – that is, persons subject to state licensure − as a new category of less protected speech.  Id. at 2371-72.  “This Court’s precedents do not permit governments to impose content-based restrictions on speech without persuasive evidence of a long (if heretofore unrecognized) tradition,” and the concept of “professional speech” was not “such a tradition.”  Id. at 2372 (citations and quotation marks omitted).  Further, “[t]his Court has been reluctant to mark off new categories of speech for diminished constitutional protection.” Id. (citations and quotation marks omitted).

An exception had been made for compelled speech in attorney advertisements.  Id. (citing Zauderer v. Office of Disciplinary Counsel, 471 U.S. 626 (1985).  For compelled speech to fall within that exception, however, requires four elements (and here’s where it gets interesting):  (1) “commercial advertising”; (2) a “purely factual” disclosure; (3) “uncontroversial information”; and (4) about the “terms for acquiring” services or (we presume) products.  138 S. Ct. at 2372.  Our usual interest in the First Amendment concerns off-label use, but we have little doubt that an FDA-mandated disclaimer, “this information concerns an off-label use,” would fall within the Zauderer exception.  Our interest here, is as product liability litigators faced with “warning” claims demanding the inclusion of information that is neither “purely factual” nor “uncontroversial.”  As we stated in an earlier post, “FDA Takes It on the Chin Again in First Amendment Case”:

[T]he court’s discussion of the Zauderer “purely factual and uncontroversial information” test for compelled speech leaves us with some interesting hypotheticals to consider.  Remember, after New York Times v. Sullivan, [376 U.S. 254 (1964),] private tort claims are on par with other forms of governmental restrictions on speech for purposes of the First Amendment. Query whether plaintiffs demanding warnings about supposed causal relationships that are anything but “uncontroversial” might be running afoul of the First Amendment.

For one thing, NIFLA makes clear that the Zauderer exception applies to compelled speech generally.  138 S. Ct. at 2372-73.  For another thing, NIFLA, held that Zauderer “does not apply” to compelled speech that “in no way relates to the services that [the speakers] provide.”  Id. at 2372.  That comes up fairly frequently in product liability litigation, when plaintiffs claim that a defendant has a “duty” imposed by (usually) state law to include information not only about its product, but about the supposedly “safer” attributes of some competing product or treatment option.  We’ve discussed before reasons why so-called “comparative” warning claims should fail, but since such compelled warnings don’t involve the attributes of the defendant’s own product, there may also be a constitutional angle to work with here.  Nor can compelled comparative warnings for prescription medical products be considered “traditional,” since the FDA does not allow them (much less mandate them) except under strict conditions.  See 21 C.F.R. §201.57(c)(3)(v).

Content-based regulations, such as compelled speech, “are presumptively unconstitutional and may be justified only if the government proves that they are narrowly tailored to serve compelling state interests.”  138 S. Ct. at 2371.  Although NIFLA did not “question the legality of health and safety warnings long considered permissible,” id. at 2376, the Court had a lot to say about governmental manipulation of medically-related information:

Throughout history, governments have manipulated the content of doctor-patient discourse to increase state power and suppress minorities. . . .  [W]hen the government polices the content of professional speech, it can fail to preserve an uninhibited marketplace of ideas in which truth will ultimately prevail.  Professionals might have a host of good-faith disagreements, both with each other and with the government, on many topics in their respective fields.

Id. at 2374-75 (citations and quotation marks omitted).  The “benefit” of products (NIFLA offered the example of “medical marijuana”) can be a constitutionally protected dispute.  Id. at 2375. “Speaker-based laws run the risk that ‘the State has left unburdened those speakers whose messages are in accord with its own views.’”  Id. at 2378 (quoting Sorrell v. IMS Health, Inc., 564 U.S. 552, 557 (2011)).

Thus, think about what may fail to qualify as “noncontroversial” or “strictly factual” information in our product liability sandbox – thereby triggering strict (or at least “heightened”) scrutiny.  Those elements of the Zauderer test weren’t really issues in NIFLA; “abortion [is] anything but an ‘uncontroversial’ topic.”  138 S.Ct. at 2372.  But what about product liability?  Well, we have the holding in the aforementioned Sorrell decision that “[s]peech in aid of pharmaceutical marketing, however, is a form of expression protected by the Free Speech Clause of the First Amendment.”  564 U.S. at 557.  We have long been of the view that one side of a scientific conflict should not be able, under the First Amendment, to sue the other side into silence through tort litigation.

Thus, we believe that under NIFLA, there may be a constitutional dimension to the kind of scientific controversies normally addressed these days by Daubert.  Simply looking through our recent “best of” lists, we’ve seen quite a few examples of plaintiffs pursuing, on a massive scale, warning claims about purported product “risks” that were ultimately debunked on scientific grounds.  See In re Lipitor (Atorvastatin Calcium) Marketing, Sales Practices & Products Liability Litigation (No II) MDL 2502, 892 F.3d 624 (4th Cir. 2018) (Lipitor and diabetes); In re Zoloft (Sertraline Hydrochloride) Products Liability Litigation, 858 F.3d 787 (3d Cir. 2017) (Zoloft and birth defects); In re Mirena IUS Levonorgestrel-Related Products Liability Litigation, 341 F. Supp.3d 213 (S.D.N.Y. 2018) (Mirena and a particular form of high blood pressure); In re Mirena IUD Products Liability Litigation, 169 F. Supp.3d 396 (S.D.N.Y. 2016) (Mirena and perforation); In re Accutane Litigation, 191 A.3d 560 (N.J. 2018) (Accutane and bowel problems). There are many more such examples in the law books.

While Daubert is a perfectly fine way of dealing with some litigation over scientifically (and sometimes generally) controversial claims of product risk, perhaps the post-NIFLA First Amendment protection of compelled scientific speech would be another.  Turning back to ABevA, at least three concurring judges would have decided the case on the grounds that the compelled speech about sugary drinks contributing to “diabetes” was not “purely factual” (such drinks have no effect on certain types of diabetes) and thus “controversial.”

This warning does not provide “purely factual and uncontroversial information about the terms under which services will be available.”  The factual accuracy of the warning is disputed in the record.  Among other things, the warning is contrary to statements by the FDA. . . .  Moreover, . . . the disclosure is literally false with respect to Type I diabetes. Although NIFLA did not define “uncontroversial,” the warning here requires the advertisers to convey [the government’s] one-sided policy views. . . .  The record shows this is a controversial topic, and therefore, the ordinance does not qualify as “uncontroversial information” under the third prong of NIFLA.

ABevA, 2019 WL 387114, at *8 (citations omitted) (concurring opinion); see, id. at *11 (“it cannot be doubted that the government’s proposed message is controversial:  it would require that manufacturers, retailers, and advertisers include in their ads, under a banner that begins “Warning,” the message that their product contributes to diabetes even though the causes of type 1 diabetes are not actually known”) (another concurring opinion).  Query: If the claimed link between diabetes and sugary drinks in ABevA is sufficiently “controversial” to require heightened First Amendment scrutiny of a compelled warning, how can the at least equally scientifically bogus demand for a diabetes warning on Zoloft not be equally unconstitutional?

The question then becomes, what is the threshold for “controversial” under NIFLA?  It may well be considerably less stringent than the Daubert standard for lack of valid scientific support.  After all, abortion is “controversial” despite – indeed because of – there being two vigorously asserted, conflicting points of view.  A lot of scientific disputes may well be “controversial” under NIFLA, and thus First Amendment protected against warnings purportedly compelled by state law, even though neither side could prevail over the other side on Daubert grounds.  Such examples might include Proposition 65 carcinogenic warnings concerning cola drinks, coffee, beer, and soy sauce, see Riva v. Pepsico, Inc., 82 F. Supp.3d 1045, 1062 (N.D. Cal. 2015), and similarly dubiously supported cancer claims involving herbicides.  See National Ass’n of Wheat Growers v. Zeise, 309 F. Supp.3d 842, 846 (E.D. Cal. 2018).  Perhaps the opinion of a single outlier group (as was alleged in Zeise) may be enough to survive Daubert, but is questionable whether, when viewed as governmentally compelled speech about a scientifically contested risk, demands for such warnings could survive First Amendment analysis of the sort mandated by NIFLA and conducted by ABevA.  Indeed, a demand that a product warning adhere to a outlier scientific viewpoint would seem to be the epitome of “controversial,” at least within the relevant scientific community.

It seems clear to us that after NIFLA and ABevA, the kind of claim that the California Supreme Court allowed to withstand First Amendment challenge in Kasky v. Nike, Inc., 45 P.3d 243, 251-52 (Cal. 2002), compelling disclosure of a defendant’s controversial labor practices, could not occur now.  Moreover, given availability of rapid First Amendment-based relief to defendants under so-called SLAPP statutes, it may be cheaper and easier to seek First Amendment dismissal of state-law-based demands for controversial, or even untrue, safety warnings.  See Vess v. Ciba-Geigy Corp. USA, 317 F.3d 1097, 1109-1024 (9th Cir. 2003) (SLAPP relief appropriate against groundless tort suit alleging scientifically doubtful drug-related risks).

One thing that was not decided, either in NIFLA or ABevA, is what degree of scrutiny applies to cases where the Zauderer exception allowing rational basis analysis is unavailable.  Both cases resolved First Amendment claims without clarifying that point, NIFLA holding that the compelled speech statute “cannot survive even intermediate scrutiny,” 138 S. Ct. at 2375; and ABevA striking down the compelled warning as “unduly burdensome” even under Zauderer.  2019 WL 38711, at *5.

But that’s why the already discussed holding in NIFLA that compelled speech is inherently “content-based,” 138 S. Ct. at 2371, is key.  That holding brings compelled speech within the rubric of Reed and Sorrell, holding that strict scrutiny, or at minimum “heightened” scrutiny, is required.

In sum, we don’t think that “compelled speech” rationale of NIFLA or ABevA is likely to affect the FDA very much. The Agency’s “valid scientific evidence” standards, and similar requirements for scientific support of label changes would appear to provide the basis for meeting the elements Zauderer exception.  The FDA’s prohibition on off-label communication has its own well-known First Amendment problems, but they don’t involve compelled speech, nor would the kind of disclaimers and “balance “ that might be alternative approaches to the FDA’s ban probably be anything but factual and non-controversial.  Finally, compelled speech analysis would not have much impact on product liability litigation where the risk of a prescription medical (or other) product is well-established and the only question is whether the plaintiff did or did not encounter it.

Where First Amendment compelled speech principles could have significant impact – and where we urge our readers to consider whether a compelled speech defense is available − is in the kinds of cases described above, where the existence of the risk itself is scientifically controversial and would attract a Daubert challenge.  In those situations, the standards for First Amendment protections may be easier to meet than Daubert, so the impact of cases like NIFLA and ABevA is something that may be helpful in the long run.  The defense could also rely on arguments (including FDA statements) against overwarning to support the constitutional defense.  At the very least, they provide a basis for avoiding “a proliferation of warnings and disclosures compelled by local municipal authorities that have only a tenuous link to a more than trivial government interest.”  ABevA, 2019 WL 387114, at *13 (concurring opinion).