Commercial Speech & the First Amendment

From Yale Law School’s Abrams Institute for Freedom of Speech: Commercial Speech and the First Amendment

Floyd Abrams: does commercial speech protection affect non-commercial speech law, for example by decreasing protection for political speech?  We’ve found that political speech proponents now cite commercial speech cases given how robust the protection is.  So the concern about defining scrutiny downwards for noncommercial speech hasn’t materialized.  SCt now believes that commercial speech is more important than some thought, and that limitations on nonfraudulent speech are more threatening than many thought.

Commercial Speech Post-NIFLA v. Becerra:  Legitimate Check on Compelled Speech or Weaponization of the First Amendment?

Moderator: Joel Kurtzberg – Partner, Cahill Gordon & Reindel LLP

Calls to overturn Central Hudson at least for truthful/nonmisleading speech. Sorrell v. INS Health applied “heightened” scrutiny. What does that mean?  Our part of the bar argued for strict scrutiny. Sorrell went on to invalidate the law at issue under Central Hudson, so it wasn’t clear. Content-based: not subject to Central Hudson intermediate scrutiny?  But what does that mean, given that commercial speech as a category is content based?  Breyer issued the first of many dissents: opened the floodgates to dangerous litigation.  Regulatory programs necessarily draw distinctions based on content.  Electricity regulators oversee company statements about electricity: specific to content. The Federal Reserve regulates interest rate disclosures, but only by financial institutions. The FDA oversees labeling and advertising for drugs, but not furniture. It’s all content based. 

Reed v. Town of Gilbert: Reed found everything to be content based and everything content based subject to strict scrutiny?  If law applies b/c of topic discussed, then it’s content based according to Reed regardless of gov’t motive and content neutrality of justification. Breyer concurred in judgment but warned again, you can’t possibly mean this is true.  [It just means the distinctions will be less predictable and more subject to judicial reevaluation of the facts.]  The entire regulatory state is built on laws that target specific content. [Which is pretty much the point for some: they don’t like the regulatory state and think that common law is enough.]

NIFLA: compelled speech case.  He litigated the graphic cigarette warning cases.  Zauderer: rational basis, essentially.  If Zauderer doesn’t apply (purely factual/uncontroversial—calories on menus) what is the standard?  Strict scrutiny? Central Hudson? NIFLA answers that question by saying Zauderer doesn’t apply when it’s content based and it’s strict scrutiny. [This wouldn’t be my summary, but ok.]  Struck down California’s disclosures required for anti-abortion clinics (licensed clinics had to post disclosures about California’s own services for pregnant people; unlicensed clinics had to post disclosure that it wasn’t a licensed medical clinic).  Court struck down both provisions: altered the content of clinics’ speech and was therefore content based.  Rejected professional speech concept as to the clinics b/c failed under intermediate scrutiny; no traditional exception of lesser protection for professional speech.  However, Court said it wasn’t calling into question legality of health and safety warnings long considered permissible.

Panelists: Jane Bambauer, Professor of Law, James E. Rogers College of Law, The University of Arizona: Zauderer apparently survives, but strict scrutiny for everything else, and it’s an acidic test.  Any error of over or underinclusion will lead to flunking even intermediate scrutiny. She’s more ok with that than others are.

Two big limits of NIFLA: Zauderer. Not totally unambigous in the opinion, but looks largely intact. The first thing the majority does is acknowledge that these statements, while factual, are not uncontroversial. Doesn’t explain why it’s not uncontroversial, but everything to do with abortion is controversial.  [If it’s litigated, it’s controversial in a trivial sense.] Court probably isn’t ready to define controversial—her own work has begun to look at how this might be done.  This is probably not the right set of facts.

Second: informed consent in the medical context.  The majority treats it as so traditional that it’s basically a mere conduct regulation. NIFLA does some damage to the work the Court has been doing to clarify the separation between what’s speech and what’s conduct.  “Talk therapy” as conduct and not speech—older case.  Led eventually to NIFLA; this Court seems to suggest you can’t take an entire category of speakers and describe what they’re doing as conduct. At the same time the Court characterized informed consent as regulating the conduct of providing medical care.  Biggest internal inconsistency is treatment of Casey: Bambauer can’t see daylight b/t disclosures required of abortion providers under Casey and disclosures unconstitutional in NIFLA.  [Which is why having courts decide what’s uncontroversial/factual is about arrogating political power to them.]  The increased scrutiny may turn out to be messy and subjective—a problem of epistemology: which branch of gov’t is better at sifting messy evidence.  Often framed as whether courts are better than legislatures; this is why Breyer cautions modesty.  Courts are interested in epistemic modesty, but they think regulators and legislatures [RT: but not private parties] should be more modest and commercial speech is a case study of why this might be.  Assumptions used by regulators in 1970s were challenged and found to be lacking—limiting optometry offices from listing prices was bad for consumers. Gov’t will mess up even the goals it intends to pursue.  [… and we’re at Lochner.]  Regulation shouldn’t apply to information transmission unless two branches of gov’t have looked at the same body of evidence and agreed it would be good for consumers. Mandatory disclosures haven’t been good for consumers and getting rid of that style of regulation may be good for them. 

Kurtzburg: off-label promotion of drugs: is this truthful and nonmisleading?  [Amarin, I think.] The judge didn’t want to have to answer that question, but he did, and found the FDA went too far.

Robert Post – Sterling Professor of Law at Yale Law School: Why wasn’t commercial speech protected before Central Hudson? Because we had free speech for a reason: to make the gov’t responsive to us, as required for a democracy.  Trying to form public opinion=state is responsive to us. That logic gives rise to speaker rights, b/c we want state to be responsive to people who are talking. Why not approve content discrimination? We want public opinion to set the agenda for the gov’t, not the other way around in choosing this not that topic.  Content discrimination ban preserves that priority. Often described in terms of marketplace of ideas, as Thomas does in NIFLA. Epistemological justification. 

Any institution that produces truth does not have a marketplace of ideas. That’s not how you run a university—any institution that produces truth makes judgments about competence and incompetence. There are better and worse ideas.  The notion that there’s no such thing as a false idea implies that there’s no such thing as a true idea, which is the opposite of the idea of truth winning a competition in the market. The equality here is political, not epistemological: politically, we are all equal, even though there is epistemological truth.

When Court extended protection to commercial speech, it was an audience based justification.  Right to receive information (as it is known in Europe).  Not a subordinate set of protections: a different kind of right for different reasons, and the Court was very clear about that b/c it had in the back of its mind Lochner. The Court has lost this distinction, putting autonomy into the commercial speaker and back into Lochner, second-guessing any commercial regulation.  This is cool if you like Lochner, not cool if you don’t. Those are the stakes.

Does NIFLA make any sense in its own terms? Thomas says content discrimination prevents a market for ideas in which truth will prevail, so every mandatory disclosure is subject to strict scrutiny.  Imagine: you give a lousy opinion letter and get sued for malpractice.  Is this subject to strict scrutiny? Why isn’t this just a marketplace of ideas?  Doctor falsely tells you you don’t have cancer: is this subject to strict scrutiny? Professional speech is supposed to be governed by standards of competence of the profession.  Doesn’t make sense to say all compelled speech is subject to strict scrutiny: in commercial real estate, endless statutes require disclosures in leases. Under NIFLA, that’s subject to strict scrutiny?  Does it make sense to have a court epistemologically determining the truth of these disclosures?  Mandatory disclosure of latent defects subject to strict scrutiny. The opinion is crudely written as “speech content regulation bad.”   If you say there are exceptions, the exceptions contradict the general logic.  A hunting license for conservative judges to decide what social regulations they don’t like.  It has nothing to do with truth or self-government but with entrenchment of a political party.  If courts are busy with striking down the SEC, they will not be rigorous when the gov’t tries to lock up terrorists.

Coleen Klasmeier – Partner and Global Coordinator of the Food, Drug and Medical Device Regulatory Practice Area Team at Sidley & Austin LLP

NIFLA’s antecedents: Sorrell, Reed [but not Casey!] have a lot of resonance on the same intellectual level, and Central Hudson is still being applied. Her expectation is that courts of appeal will continue to use those rubrics. Not sure that Central Hudson or Zauderer were ever particularly helpful analytical tools. Rules of thumb: gov’t’s view of what’s smart decision doesn’t count (graphic tobacco disclosures).  Don’t burden speech b/c you find it too persuasive (Sorrell).  Don’t manipulate consumers with neuroscience (tobacco again) [which is totally fine for private sellers]. But there are exceptions, historically accepted health/safety warnings—she has no idea what that might include. Thomas might use that for FTC rules. Another rule of thumb: Don’t create a stupid record—9^th Cir. American Bev. case—had record ev. that 10% of ad devoted to warning would be enough but required a 20% warning. CMS proposed rule requiring disclosures of list price in TV ads: the entire record says we’re doing this because we think it might work but we’re not sure.  That’s a bad record.  [Having read the proposed rule, I disagree with this characterization of the evidence and its characterization, but that’s not surprising.]  Breyer understands that FDA is threatened by the language in these cases.  All of the labeling rules raise these issues.  There’s a drug warning on a diabetes drug saying it causes thyroid cancer, but it’s not based on human data. The FDA wants doctors to know of the risk; it works in judgments, not in facts. The agency needs facts, but it judges those facts.  Judges make decisions about data all the time.  The epistemology question does go to competence—FDA’s primary jurisdiction is an issue, but judges do have a role b/c they’re good at judging too. 

Amarin: the FDA didn’t want offlabel marketing even though pharmaco proved that its claim was accurate; not allowed on label and thus FDA said marketing wasn’t allowed, b/c other drugs indicated that early data of the type you generated don’t necessarily correspond to longterm cardio benefit that consumers would expect from a drug of this type—an extrinsic factor having nothing to do with the drug itself [um, that conclusion doesn’t logically follow at all—especially if you believe in misleadingness]. You can debate whether it’s sound public health policy, but it’s not a great rationale for a speech regulation.  Hijacking labels for the agency’s view of public health, which is a pathology.  Agency has now internalized some of the lessons of the cases.

Kurtzburg: Why isn’t having exceptions enough?  Zauderer: Can compel if speech is factual, noncontroversial, and not unduly burdensome?  Gov’t can also regulate conduct w/incidental impact on speech.  Two other exceptions he think exists: under precedents, gov’t may not impose content based restrictions w/o persuasive evidence of a long, if heretofore unrecognized, tradition to that effect. If there’s a long history of requiring disclosures in commercial real estate transactions, that’s ok.  The opinion says it’s not calling into question health and safety warnings long considered permissible.  Emphasis on tradition, not just about health and safety. Doesn’t the analysis of professional speech suggest that if the gov’t could come forward and show the long tradition, it would be ok?  Why is Post so worried?  It can’t be that every regulation is really going to be called in question.  [Which is also the hunting license point: we’re now in the business of randomly targeting things.]

Post: he has a bridge he’d like to sell you.  How much work has “tradition” done in public forum doctrine done?  None.  It depends on who gets to say what tradition is.  Also, redefines Zauderer by putting the burden on the state, which is not what Zauderer was. And changes the meaning of “controversial.”  If it means “can’t be about controversial subject” then it takes the disclosure away when it is needed most.  How do you distinguish between speech/conduct when regulating the language of a contract?  We’ve tried that for 50 years and failed. It’s an invitation to ad hoc judgments reflecting political prejudices of courts.  SEC required disclosures: is that a long enough tradition?  Why shouldn’t the gov’t be able to experiment w/ new forms of disclosure? This is about ending the development of regulation: command and control wasn’t good, but information forcing disclosures became preferred. It’s not traditional so you can’t do it: but why does that make any sense given evolving knowledge?  Also, if you want to talk tradition, commercial speech protection was unknown to the Framers; it’s not traditional at all! 

Compelled speech serves the very constitutional value for which we created commercial speech doctrine in the first place, and NIFLA treats getting more information as a bad thing.

Health/safety as tradition: commercial leases and SEC disclosures aren’t health/safety.  Is that in the exception?  What does it mean?  Thomas is just carving stuff arbitrarily as it comes into his mind, not thinking about how markets are formed and regulated in the US.  Amarin: missed the main justification for why you’d want safety & effectiveness data for offlabel use: that’s an information forcing device to require producers to come forth with information. Judge had such a poor understanding of basic foundations of regulatory scheme. Now, you can get safety & effectiveness approval for one indication and market it for something else—that’s law made by a judge who admits he can’t operate his own toaster.

Klasmeier: there is appropriately a burden on manufacturers to substantiate safety/effectiveness for initial marketing. We don’t have fully publicly funded research. Offlabel use has been allowed for decades; FDA has accepted that. It can’t regulate doctors b/c they’re ungovernable; tried to regulate manufacturer speech instead. Not a foundational principle of food & drug law that it’s illegal to disseminate info: an artifact of 1970s controversy.

Post: it was illegal to market to general public.  [What makes it a drug is how you advertise it.]

Klasmeier: the policies are generally agnostic to audience.  [Certainly Amarin suggest you could market directly to consumers.] Patients are entitled to receive information about offlabel uses, so they can go to doctors.  Consistent w/listener-oriented rationale.

Bambauer: Post said regulatory state used information as better version of regulation.  80% agree that more information is generally better, but that’s too facile.  It’s really hard to use facts to convince people.  If you try to regulate directly, like taxing soda, the politics show the public [or relevant interest groups] isn’t ready, but a speech regulation can be slipped in w/o as much political awareness.  That may be good or bad, but it doesn’t mean that the 1A check is undemocratic.

Kurtzburg: en banc 9^th Circuit American Bev. Ass’n v. SF applied NIFLA to sugar-sweetened beverage warning.  Majority applied NIFLA.  Maybe NIFLA is much ado about nothing (concurrence said it would have gone further, because health/safety didn’t traditionally include any warnings about sugar sweetened beverages).

Bambauer: the issue is there are so many requirements that you can go in any order and likely invalidate a regulation.  Makes more sense to look at factual/uncontroversial as prerequisites for Zauderer, rather than undue burden.  9^th Cir. seems to say that warning may be factual and uncontroversial; only the burden part indicates that NIFLA might have put more teeth in the analysis. Worries that NIFLA will be construed in a way that will always find a way for the gov’t regulation to fail, but only in undue burden.  Case suggests that factual/uncontroversial disclosures will often be ok.  [But see the weird discussion of diabetes, where the argument is that it’s not “factual” b/c it doesn’t sufficiently distinguish type 1 and type 2 diabetes.]

Post: DC Circuit has been at war w/Zauderer for a while now.  Misleadingness correction only?  DC Circuit recently went en banc and said it wasn’t just for misleadingness, but was basically just Central Hudson.  9^th Circuit similarly set up a situation where Zauderer can always be rejected in favor of strict scrutiny. When you can’t predict how the opinion should be applied, and when there is so much commercial speech regulation, you’re giving a hunting license to courts.  Much less worried about the difference b/t Central Hudson & strict scrutiny—not interested in fetishizing strict v. intermediate scrutiny, which is an evasion of the central Q: what constitutional values are you trying to protect?

Being forced to carry a calorie label is not being forced to affirm that one believes “Live free or die.”  We now have gov’t speech doctrine saying gov’t can say anything, no matter its falsity: how does that go along with this other problem?

Klasmeier: A defendant in a wire fraud case was subject to a DOJ press release that was concededly inaccurate. The defendant sued over DOJ’s statutory authority; DC Circuit rejected idea that DOJ should have to account for its falsehoods under Data Quality Act. We litigated it as a case about fairness and equality; disappointing result.

Post: the gov’t is absolutely privileged to defame: Paul v. Davis.

Kurtzburg: we normally defer to agencies w/in their area of expertise.  Should we defer to FDA on safety & effectiveness judgments?  [Note how this is a Lochner Q.]  Does the 1A tip too far in the other direction?

Klasmeier: trends seem to be converging against deference.  A lot of these cases are Chevron step 1 cases: deference is not appropriate.  Reluctant to live in apocalyptic world Post describes because that’s awful, but Thomas may be trying to move towards idea that we don’t know legislative motives of yore, while we have information about today, and we see bad records from today [so yesterday’s bad decisions are fine and can’t be changed, especially if they were from the common law, I guess]. Moving forward, can’t politicize questions of conscience and value [like whether to defraud consumers?]—but doesn’t know what to make of Reed in that light, since that wasn’t a bad record [or a question of conscience and value].

Kurtzburg: false/misleading can be regulated under commercial speech: it’s a big carveout.  Consistent w/NIFLA’s holding: there’s a long tradition of allowing the regulation of commercial speech that is potentially false or misleading, and no one could argue to the contrary. [Wait for it: I’m about to discuss a 5^th Circuit case that doesn’t just argue but holds to the contrary.]  It’s an exception to the ban on content discrimination. 

Post: sign regulation was a longer tradition, and now there’s Reed.  Street sign, house number, those distinctions have been made at least as long as misleadingness.

Bambauer: whether we talk about long tradition or deference to regulatory agencies, neither really make sense for constitutional law: we had a long tradition of racial discrimination [and more to the point, of defamation law]. The disagreement is more about what we’re trying to accomplish. Even the libertarians seem to recognize that fraud and things close to fraud are unprotected not just b/c of tradition but for actual substantive reasons.

Post: but you could never have that for political speech: false statements about Brexit. We do apply that rule to commercial speech. Why would that distinction matter? Could it be that not all speech is the same?