Tuesday, June 04, 2019

The Consumer “Right To Know” Versus the First Amendment

Panel II at the Abrams Institute's conference

Moderator: Jonah Knobler, Partner, Patterson Belknap Webb & Tyler: recent attempts to do things like require disclosure of child labor in manufacturing process, and related consumer protection cases arguing that failure to disclose was misleading.  Kavanaugh, then-Judge: gov’t argued that there was a substantial interest in providing info, but that would create endless potential burdens since consumers might want to know political affiliation of business owners etc. Consumer interest is not enough for a freewheeling power to mandate compelled commercial disclosures.  So we know what he thinks.


Rebecca Tushnet – Frank Stanton Professor of the First Amendment, Harvard Law School

Amestoy/6th Circuit cases striking down competing regulations of disclosure of rBGH, showing the current tensions in the law.  Amestoy says that consumer interest in a matter isn’t sufficient to compel disclosure and it’s just consumer interest because the FDA says that rBGH has no effect on milk; the 6th Circuit says that it’s not just consumer interest because research not accepted by the FDA indicates there might be an effect on milk.

A few points, so that we can get to doubtless robust discussion: Approaching strict scrutiny for regulations of nonfraudulent (as Mr. Abrams put it) commercial speech while applying rational basis scrutiny for regulations of false or misleading commercial speech puts incredible and I think unworkable tension on the line between truthful and nonmisleading and false or misleading speech, and we are beginning to see that tension manifest in weird and conflicting decisions about disclosure, consistent with Professor Post’s characterization of current doctrine as a hunting license.

Justice Stevens had it right in Central Hudson: where it’s a regulation of the economic relations between the parties, disclosure regimes should be subject to rational basis review as long as we don’t like Lochner. Of course judges are increasingly friendly to a return of Lochner, but the real justifications for that movement don’t distinguish compelled speech from compelled payment of a minimum wage.

I also agree with Professor Post that MPG and emissions labels for cars, nutrition labels, FDA regulation of drugs cannot be distinguished in type from these other regulations, which is why even the cases that strike disclosures down make the law more complex and hold out the prospect of approving mandatory disclosures. This may eventually change but I predict it will at least outlast Roe v. Wade—Central Hudson prong one (and relatedly Zauderer) are hard to kill because they’re really the only way that commercial deception stays actionable while false political speech is generally protected.

But consider a few recent cases: Mississippi Bd. of Engineers v. Tire Engineers: 5th Cir. case. New interpretation of decades old precedent about the line between potentially misleading and actually misleading speech.  The background is that the Supreme Court said a bunch of things in lawyer regulation cases about when you could ban speech versus when the regulator had to mandate a disclosure instead—this line of cases treated disclosures as speech protective and minimally invasive, and they haven’t been reconciled with the anti-disclosure cases.  Anyway, those cases say that the government can ban if the statement is false or actually misleading, but the government has to use a disclosure—or at least start there—if the statement is only potentially misleading.  This distinction doesn’t exist outside First Amendment law. 

According to the Fifth Circuit panel in the Tire Engineers case, surveys don’t show actual misleadingness, but an individual deceived consumer might.  This is a terrible distinction, wrong in both directions.  Given the survey showing 55% net confusion (an incredible number if you think about Lanham Act cases finding misleadingness), the panel suggested, a disclosure might be justified, but it could only be as minimal as necessary, which the panel thought meant it only needed to be disclosed when a consumer was actually at the facility—a classic bait and switch problem of sunk costs.  The case also indicates a broader issue with trying to treat disclosure as a separate issue from falsity: when a message is false or misleading, one response is to try to ban it, but the other is to say “fix it.”  If there is a false message, then there should also be a true message about the same thing.  And to say that you have to stick to the truth is to compel speech, under the current framework.

Another case worth noting: the Florida case barring pediatricians from asking about whether there are guns in the house.  A panel said this survived strict scrutiny, which I submit would be laughable if it weren’t a bellwether of where we are going in judicial politics.  And with abortion of course we’ve been there for decades: I agree with Professor Bambauer is that the glaring contradiction in NIFLA is on the many burdensome disclosures, neither factual nor uncontroversial, that the court has approved for abortion providers.

NIFLA “health and safety warnings long considered permissible.” We’ve already seen debate about that, and the quality of that debate has been terrible.  One concurrence in the American Beverage case thought that the specific health and safety warnings at issue had to be of long standing, meaning that for example no mandatory warnings about thalidomide would be constitutional because we didn’t know about thalidomide at common law. Even Justice Scalia didn’t think that: science can tell you things you didn’t know to which the common law principles about harm can then be applied.  (Lucas v. South Carolina Coastal Council, nuclear power plant that turns out to be on an earthquake fault line can be shut down without constituting a taking.)  Similarly, the dispute in Am. Bev. about whether the disclosure was “factual” b/c it only referenced “diabetes” without distinguishing between type 1 and type 2 diabetes shows how easy it will be to argue and hold that something isn’t factual.

Jonathan H. Adler – Johan Verheij Memorial Professor of Law and Director, Center for Business Law and Regulation, Case Western Reserve University School of Law

Zauderer should’ve been understood as a simple application of Central Hudson, not a separate test. Result: mess, especially in terms of explanation.  The rBGH cases from Sixth and First Circuits are fully reconcilable: gov’t intervention in information in the marketplace is something we should be skeptical of. Mandatory/conditional disclosures (if you say A, you must also say B; A means X) require justification, but should not be open season.  Gov’t has to be explicit and express about what it’s doing and why, which becomes a filter protecting against arbitrary gov’t action and against rent-seeking, which was clearly what’s going on in the milk cases.  [And why this is about Lochner.]  Similarly with land regulation, floodgates aren’t open [doubt Florida regulators would agree].  Fed gov’t will be able to defend the vast majority of its disclosures.  The Court hasn’t done a particularly good job of fleshing that out.  Nutritional labels: peanut allergy disclosures are strongly justified—the gov’t has an interest in protecting the uninformed consumer who can never know everything about a product from suffering harm from that product.  Probably satisfies both Central Hudson and strict scrutiny, provided that it’s no more intrusive than necessary.

Securities disclosures can be justified too—an expert agency can decide that consumers need to know that “you only pay fees if you win” doesn’t include costs which will be owed regardless. [Quite notable that it was the bar, not “experts” in the data collection/analysis sense, to which the SCt deferred in Zauderer—preference for anecdotes over data.]  SEC disclosures on conflict minerals: telling consumers to focus on this aspect of the product—he agrees the courts haven’t been effective on fleshing out intuitive judgments they then rationalize.  Conflict mineral disclosure is meaningfully different from other regulations on the books.  GMO disclosure is materially different from ingredient labels.  [See Doug Kysar, Preferences for Processes: this conclusion relies on paternalistic judgments about what consumers should care about, not what they do care about.]

In sugar sweetened bev. case, the gov. can warn about these things but it did more than necessary [unless you think that any of the concurrences might add some votes if SF came back with a 10% warning]. 

Jacob Spencer –Gibson, Dunn & Crutcher LLP
CTIA is still ongoing (remanded in light of NIFLA, still pending before 9th Circuit)—disclosure that cellphone retailers would need to provide at point of sale.  About RF exposure/exceeding FCC standards if you carry phone on and in pocket/in bra.  Epistemology isn’t litigated here: both parties appeal to what the FCC has determined about the science, accepted it as an accurate account of science and of what FCC was trying to do, and then disagreed about whether the message conveyed by Berkeley’s ordinance was consistent with the FCC.  CTIA: if you read it as a normal consumer, you would come away thinking there was a safety risk.  Uses “Safety” at beginning and end, and “radiation” is scary.  FCC has said there is no safety risk that’s scientifically provable from carrying cellphone.  Berkeley’s position: each sentence is true, and this is factual and uncontroversial. 

SF did a sugar sweetened beverage disclosure that was struck down as unduly burdensome. There was no definitive showing that SSBs uniquely contribute to obesity etc.  The parties didn’t have a fight about the science, though.  Even without controversy on facts, there’s a lot of disagreement. The trilogy of Zauderer, Ibanez, and Milovetz were all about lawyer marketing, where the Court is very confident about what it’s doing and what it knows about the facts.  Other cases, not so much.  [Which may say something about the judge as data scientist regulating the FDA.]  They were also all conditional disclosure cases: you say X, so you have to say Y.  NIFLA is a noncommercial speech case, and it’s weird to say stuff about Zauderer in that context—so none of these cases deal with run of the mill compelled commercial speech.

Knobler: safety disclosures: lack of clarity over when health and safety is a sufficiently important or substantial interest to justify disclosure.  Vaccine/autism?  Adler thinks precautionary disclosure might be allowable even if risks are hypothesized and not proven.  Tushnet thinks current epistemological uncertainties about truth expressed in 1A cases should imply that courts can’t rule on products liability cases, which is why the 1A cases are weird.

RT: [This is about scientific “proof” versus “proof.”]  It’s about probabalistic harm and whether regulators can act on probabilities, which I think they can.  If past experience w/ a class of drugs has indicated consumer misunderstanding (or safety problems), you should be able to predict that with a new entrant.

Adler: factual/uncontroversial is not a good framing.  Substantial gov’t interest is where we should be looking, and whether there are less burdensome ways of regulating. Uncontroversial in particular is prone to so many interpretations, ultimately incoherent and unhelpful.  We think that agency transparency about decisionmaking has an effect on the quality of decisionmaking. If you have a test that says the gov’t must explain why it’s regulating, that helps constrain the gov’t.  In Amestoy, the gov’t wasn’t willing to say it was protecting people’s health, and it wasn’t willing to do that.  In CFIA, the gov’t wasn’t willing to argue that it believed that there was an actual safety risk against which it was protecting people.  Requiring the agency to put its reputation behind things will discipline it some against rent-seeking.  [Except for abortion.]  The problem: how much scrutiny will we give the gov’t’s claims.  Most cases have tended to stay away from that question. Baltimore Gas, 1983 O’Connor opinion: when the Q is one of science, we should be more deferential to admin agencies than in any other context.  Judges know this is not their comparative advantage. Allowing agencies to, as they must, pick and choose b/t scientific research comes along w/having the agencies. But 1A context creates a tension that hasn’t been unpacked.  A candor rule relieves some of the pressure. 

Cases that come out of the states: deference comes from dealing w/expert agencies. But city council of SF may be different.  [so lying about federalism is part of the problem?]  A candor rule might be less effective with legislatures than w/admin agencies.  Courts have largely avoided this problem.

Knobler: most of these cases involve gov’t regulation, but what about private lawsuits under state tort law seeking to impose these disclosure requirements, and that’s even more problematic.

RT: I have a theory about this: argument made and ignored in Nike v. Kasky because it’s inconsistent w/the 1A’s distrust of gov’t.  If we believe in (1) judicial competence to find facts and (2) falsity/misleadingness, then these cases have to continue.  The California laws are not regulatory. They use the same general language of falsity/misleadingness as core FTC §5 does.

Knobler: when there is a safety risk dispute, what kind of evidence should be required to take it into the category factual/uncontroversial?  In CTIA before NIFLA, they said uncontroversial didn’t impose a separate requirement from purely factual; has to be phrased as a fact. NIFLA involved true facts but the Court still said it was controversial: on a subject that made people respond in an emotional way.  [Like whether kids should die of peanut allergies—and I’m not kidding about that.]

Spencer: Mistake to read Zauderer as distinct from Central Hudson.  He would read “controversial” very broadly. If it’s controversial: controversy in public; good faith scientific disagreement—the consequence is that you should then apply Central Hudson or heightened scrutiny to assess the gov’t’s reasoning. Level of gov’t is important: federal regulator v. municipality; core competency of gov’t (SEC lacks core competency on conflict minerals and whether disclosure would affect violence in Africa); whether this is a standard warning (he’d look at whether this is the sort of topic that looks like existing health and safety warnings).

Knobler: more exotic disclosure requirements: state interests grounded in consumer’s right to know facts for their own sake or facts of moral or ethical or aesthetic importance.  Can be phrased as economic harm (paid more than I would) as well as moral/ethical harm. Adler has argued right to know is not itself a consumer interest that justifies compelled speech, even if consumers do in fact care more about identity of producer than the other qualities of the product.  Where do you draw the line between mere consumer curiosity and a legitimate state interest?

Adler: If ungrounded consumer right to know is justification for compelling disclosure, without another gov’t interest, then there is no protection against mandatory disclosure. The fact that gov’t, legisl. or admin, has decided will always show that some nonzero number of consumers want to know. There is an infinite amount of information consumers might want to know, at least if prompted.  But these aren’t commercial concerns, but concerns about values, how we see ourselves as citizens of a polity.  This distinction b/t citizens and consumers has been obliterated, if it ever was a line.  [Then I really don’t see why these concerns should be distinguished from “commercial” concerns.]  Tilts the playing field to make decisions more about question A than question B. That should worry us.  Instead, the gov’t should have to articulate harms to people or property.  Squishier when the gov’t has other interests it’s trying to carry out, like strengthening a larger market, or AMI’s country of origin labeling.  These aren’t protecting consumers from harm but groups trying to achieve through regulation what they can’t get through the market: mandatory GMO labeling is about forcing producers to shame their own products.  If Vermont really wants this label on the milk, be willing to say up front that it thinks there’s a health risk.

RT: Doug Kysar was right; the distinction reflects a paternalistic view of what consumers ought to care about: the gov’t is already making a value laden choice about which characteristics should matter and Professor Adler’s point about the merger between commercial and political means that telling people to stay in your lane and care only about calories is already a political choice.  Another takings case, Eastern Enterprises v. Apfel: it’s possible that a combination of regulations would be so burdensome that you would go out of business.  That doesn’t make it ok for you to pick and choose which you want to comply with to avoid a taking. [Or as Jack Balkin says, the power to tax and the power to destroy can in fact be separated in practice.]

GMO labeling: empirics on this seem less dangerous than Prof. Adler make it sound. Consumers seem relatively welcoming.

Zauderer as application of Central Hudson: key difference is whether the burden is on the gov’t to show that the disclosure achieves the objective; since Zauderer our factual understanding of the world has changed and that provides us with a choice.  Bambauer and many others have noted that disclosures often don’t do what we want them to (though it depends very much on the disclosure and on dynamic effects especially effects on intermediaries).  Judges like Kavanaugh think that a disclosure should inform and so therefore it does, but that’s inconsistent with the evidentiary standards they apply to everything else. Neither side has taken seriously enough the objections to disclosure working: it would imply that many more speech bans would be acceptable because there is no happy medium of disclosure while preserving the truthful message, but many other regulations would fail unless we accept a 10% increase in awareness through disclosure as sufficient to show that the government has gone some way towards achieving its interest.

Adler: Kysar misses that consumers care about things but producers can come forward and identify themselves to groups that care.  Econ literature is largely ignored in legal literature: barring exceptional circumstances, where groups have strong preferences about characteristics, there is disclosure by firms to match preferences, and in a separating marketplace consumers tend to interpret nondisclosure as confession.  USDA has studied this for years.  It’s not that consumers shouldn’t care or can’t care, but when they do, the marketplace takes care of that.  If the requirement was that the gov’t simply had to articulate why the consumer demand wasn’t being met, that would eliminate the justification. Other areas of the law may make companies somewhat reluctant to be aggressive in making direct comparative claims [I don’t think this is true in the US]. If you want a GMO free product, you can find it.  Mandatory disclosures often differ from market disclosures; they get changed and less responsive to what consumers are demanding, but rather to what lobbyists are demanding.

[RT: this would all be fine if preferences were fixed.]

Spencer: agrees w/ Adler.  Right to know would be circular: always satisfied [unless we care whether disclosure works].  We have a separate set of standards for commercial speech b/c it deals only w/their economic interests, though that has never really made sense, but if the statement doesn’t deal w/consumers’ economic interests then it shouldn’t be relevant.  That said, you never want to be litigating whether the gov’t interest is substantial.

RT: if that were true, then you have to explain why it should be constitutional to bar false explicit statements about ethical practices because it’s not a commercial interest but instead is political.

Adler: that’s fraud.

RT: now we’re just debating price: what consumers understand from the unadorned statement.

No comments: