Caroline Mala Corbin, Compelled Disclosures
Discussant: Amy Kapczynski: Brennan-Marquez’s case starts with graphic warnings for cigarettes, struck down as inappropriate compelled speech because these images were emotionally laden and not limited to factual information. Thesis: false symmetry between right to speak and right to be free from speaking. Right to free speech is democratic and autonomy, whereas right not to speak rests only on autonomy values. The main payoff: corporate speakers lack autonomy/personhood interests and therefore the tobacco case is wrongly decided. Autonomy as the capacity for conscience: can only be fulfilled by entities that have that capacity; sometimes people who aren’t corporations will not be speaking in a matter that implicates this capacity for conscience as well.
Concludes: only sustain challenge to compelled corporate speech when it constitutes functionally a ban on the right to speak.
Corbin covers the tobacco case but also abortion “informed consent” requirements that make doctors show women ultrasounds, listen to fetal heartbeat, disclose “facts” as well as disclosure requirements (struck down) that pregnancy crisis centers disclose that they don’t provide abortion services. Sets out a four part test for when compelled speech can’t be allowed: chilling speech; distorting speech (introducing falsehoods or misleadingness; it might not be clear it’s gov’t speech; compelling professional speech is commandeering professional credibility in order to have an impact on the speech market); autonomy of listeners—paternalism or exploitation of cognitive limitations to work our biases; gov’ts are allowed to be paternalistic if they’re autonomy-enhancing.
Corbin argues that the cases are all wrong: tobacco disclosures should have been fine; mandatory abortion disclosures should not have been; pregnancy crisis center disclosures should have been fine. Critical moves: gov’t isn’t making false statements in the tobacco context, nor working cognitive biases impermissibly. Courts disagreed with Corbin’s presentation of what counted as “true” in abortion and tobacco cases. Constructed a tortured version of the truth value of the abortion statements, and disagreed about the emotional value of the images—might have to go image by image to decide. Addiction as autonomy-degrading, making intervention on tobacco more justifiable.
Questions: Autonomy, as Post has argued, doesn’t make a lot of sense in the 1A context. Gov’t compels speech all the time—honor code as student; disclosing certain info about patients.
Who counts as a commercial speaker? Commercial speech v. speaker—corporate entity v. person. Corbin argues that corporations can’t have autonomy interests, but Brennan-Marquez may have noncorporations without autonomy interests in speaking because of their roles. So who doesn’t have autonomy? Can a corporation never have an autonomy interest? Are there other collectives that do have autonomy interests?
Compare interests about paternalism: there is a universe of arguments about protecting women from abortion as enhancing their autonomy because the world is pushing them to abort. Why not agree with the autonomy-enhancing structure of the argument, if we assume (as the court did) that the statements required to be made were true.
Corbin: two separate reasons why autonomy isn’t implicated in cigarette disclosures. Corporations don’t have a conscience. Commercial speech is protected for the benefit of the listener, not for the speaker. Might also extend to other entities, though more complicated for nonprofits.
Cigarettes: 90% of smokers wish they could quit; can’t say the same with abortion, that women who want an abortion would be better off without having one.
In terms of manipulating viewers, court’s assumption in smoking case that appeals to emotion were automatically problematic was wrong. Appeals to emotion aren’t automatically problematic—turns on assumption that ideal decisionmaker is completely rational. That’s inaccurate descriptively as a matter of the science of decisionmaking, nor is it normatively sensible. That doesn’t mean all uses are problematic. So when is provoking emotion ok? Distinction between persuasive and manipulative. Not wed to the line I try to draw: appeals to emotion are ok unless trying to exploit cognitive biases—intention to take advantage of systematic errors in decisionmaking.
Brennan-Marquez: using Reynolds as a departure point to conceive of the relationship between different kinds of rights to speak/not speak. Even if a speaker has a right to produce meaning, doesn’t necessarily confer a right to resist meaning.
Compelled physician speech: often runs into issues of what’s factual and not; when the content seems nonfactual and false, it’s not clear we need to reach the analytical question of whether there’s a 1A problem conscripting the doctor—the state lacks any interest in broadcasting false information to women.
To the extent that commercial speakers have a right to speak, chilling commercial speech isn’t the same as a regulation that compels them to broadcast the gov’t message—that begs the question of the shape of the right, so he’s comfortable distinguishing Sorrell.
James Grimmelmann: Analogy between corporations and computers? Entity not capable of autonomy, but used by people to communicate ideas. Operated by single person/small group—when gov’t says you have to disclose commercial ties; the gov’t requires you to place a particular ad in a particular place; the gov’t requires you to advertise ham though it’s a Jewish search engine.
Corbin: it’s not the only question you ask about 1A: chill and distortion and listeners’ autonomy may be implicated even if the computer has no autonomy right to speak. Do you impute programmers’ rights to the computer? That’s a separate question. Legally, a corporation is not the alter ego of the owners. Because we’ve insisted on distinction for liability purposes, it makes sense to say these are distinct legal entities for legal rights.
Brennan-Marquez: compelling speech of the machine seems unproblematic to him. But there’s a difference between compulsion in commercial context that requires carrying gov’t message and invites responses from that commercial speaker itself and the search engine context where speech is zero-sum in a sense. Compelling results first means that the programmer no longer has the autonomy to run its own algorithm. So this is about when a compulsion circles around to constructive prohibition.
RT: I wish I could believe that we can all agree that women who want abortions can’t be told you’d be better off without one, but the concerns expressed are not entirely for their later health, or at least not just for their physical health. Scrutiny of gov’t’s ability to find facts is where all the action is, and where we might want deference unless we can found the scrutiny in some specific constitutional right (like reproductive rights). Many women experience abortion as a regrettable necessity—what makes that different from addiction? Especially since the graphic image isn’t really designed to get the addict to quit—Christine Jolls has a very good paper empirically investigating what the image can and can’t do.
Persuasive/manipulative also impossible to sustain—it’s heuristics all the way down. Look at studies on how risks are presented: presented one way, people pay more attention to the risk, the other way they pay less attention. Which is the persuasive way and which is the manipulative way?
Similarly, disclosures are zero-sum: less package real estate to create the emotional connections the TM owner wants to create, so Brennan-Marquez’s distinction seems unsustainable to me. Reynolds made that very same constructive chill argument.
Corbin: paternalism concern—if gov’t is just requiring disclosures clarifying misleading ads/facts otherwise unavailable, most of us unworried. More uncertain when facts might be persuasive, when persuasive here means trying to make you act in a particular way (as opposed to informing you, presumably with the awareness that it might make you act in a particular way). But not all attempts to persuade are the same. So which are acceptable and unacceptable? Is it trying to persuade you to do something you’d do anyway? Paternalism = gov’t knows better than you do. If soup can ingredients are persuasive—gov’t mandated disclosure lets you avoid soup you wouldn’t have want to buy anyway.
Kapczynski: If you take gov’t facts as true, the fetus does have a heartbeat.
Corbin: but if the gov’t is trying to convince you something you disagree with, that’s the source of resentment—people resent the cigarette warnings too (I think that cuts against her argument). General social consensus that smoking is better avoided (I think that’s also true of abortion in many circumstances, though). Some people may still find the smoking disclosures paternalistic, if their smoking is rebellious. Can’t make the argument that abortion disclosures are nonpaternalistic—there’s no consensus that abortion is always in the worst interests of women. Not all paternalism is problematic.
Kapczynski: so can the gov’t require disclosures that big sodas make you fat?
Corbin: can draw the line in different places.
Some heuristics are more problematic than others. Advertisers are masters at manipulative heuristics. Affective priming. Associate product with preexisting positive association—sell you a toaster w/beautiful toaster v. sell you a toaster by draping beautiful woman over toaster. That latter is exploiting affective heuristic. It’s intentional. Example of exploiting errors that we make.
RT: but that’s b/c there’s no fact-based connection between the woman and the toaster. And there is between the cigarette and the diseased lung and between the fetus and the sonogram.
Corbin: but the wanted pregnancy is different—it’s like connecting the woman and the toaster. Ultrasound has a very particular social meaning in our culture, that is neither inevitable nor universal—it has the meaning of a wanted baby, though not in other cultures.
Jane Bambauer: don’t diseased lungs also have social meaning?
Corbin: the sonogram is not a picture of fetal development, but culturally it is a representation of a wanted baby; heartbeat means life/love. So the attempt is to create the fetus as wanted baby, mapping social meaning onto the fetus—affective priming. To Corbin, that amounts to manipulation.
Brennan-Marquez: Zero-sum: spatial and temporal constraints means that there will be some tradeoff no matter what. But there are lines to be drawn. Pruneyard: giving space to pamphleteers in shopping mall takes up some space that now can’t be used for other messages, but we have conception of information production through time that allows for proliferation of ideas and messages—inherent corrective in way that might not be true of search engine.
Felix Wu: Autonomy of listeners—is that a collective interest or an individual interest? Couldn’t quite see how we connected consensus to autonomy. Maybe cigarettes are easy, but what sort of consensus are we looking for?
Corbin: thinks of it as individual. If gov’t is trying to convince you to do something you want to do anyway, that’s not paternalistic.
Balkin: most schoolchildren aren’t Jehovah’s Witnesses, so can you make them salute the flag?
Corbin: implicates other concerns about forced speech. How do you know when the individual interest has been implicated? Can’t be done on person by person basis, so best alternative is to look for social consensus.
Brennan-Marquez: what if there was a consensus on abortion being bad for women? Would it change your analysis?
Corbin: would still find distortion of discourse—other factors would still matter. State is imposing its own ideology as it doesn’t do for any other medical procedure. (This strikes me as a reproductive rights objection rather than a 1A objection.) Would still be paternalistic even if there was a consensus.
Jason Mazzone: what if the gov’t’s message was “there’s an absence of consensus on the long-term effects of abortion on women’s health”?
Corbin: still thinks that’s wrong.
Mazzone: Congress found it.
Corbin: some of these claims are just made up; others are incredibly misleading. There’s an association between abortion and suicidal ideation—it’s ridiculous to say that this is just a disclosure of a correlation, since no one receives it that way. Claim about you’re taking the life of a whole human being—that’s an ideological statement. Some “facts” are not facts.
Mazzone: so what about “a substantial number of people believe that you’re about to murder a human being.”
Corbin: that’s a factual statement, but that doesn’t make it ok. It’s a persuasive statement, and there are some circumstances where the gov’t shouldn’t be doing that. (But they do a lot with drugs/supplements/etc.—again, this strikes me as an objection founded in the constitutional right to choose abortion.)
Mazzone: this reflects a slipperiness that seems related to normative commitments that show in the paper; thinks it needs more.
For Brennan-Marquez—how do categories of corporations and noncorporations map onto how speech is delivered in the modern context? Who is it you’d allow the gov’t to conscript to disseminate its message? If I write about politics and review products on my blog and there’s an affiliate link and I advertise and I incorporate, can I be required to speak in particular ways?
If you allow gov’t to conscript people, what are the mechanisms to make clear that this is gov’t speech? Should we require gov’t to pay for the ad space?
Brennan-Marquez: distinguish between your capacity as political/expressive speaker and as a commercial speaker. It depends on what kind of compulsion! Gov’t can require you, near your ads for a regulated product, to put the gov’t disclosure related to that product—burdens you only in your capacity as a commercial speaker. We can imagine websites where this is unclear, but it turns on the type of compelled disclosure. There might be some that has quasi-ideological content and quasi-factual. We can distinguish in first instance between you speaking as expressive agent and you speaking commercially.
Kapczynski: incorporation isn’t the key here.
Brennan-Marquez: the corporation doesn’t need a conscience to have an interest in expressing viewpoints. But it can be compelled to speak commercially at least. It’s because there’s commercial speech going on (and therefore not about incorporation?).
Mazzone: the blog that accepts ads is compelled as long as there’s a nexus between the commerciality and the compelled speech.
Brennan-Marquez: yes. That’s the meaningful difference between commercial disclosure and gov’t’s ideological opinion. The test is to ask whether a compelled commercial disclosure is acceptable.
(But what about the unincorporated blogger with affiliate links who gets valuable free stuff in return for reviews? The FTC regulations purport to require disclosure of at least some of her activities. Is that constitutional under this scheme?)
Kapczynski: so a corporate speaker can’t be compelled to speak about noncommercial issues. But it still lacks a conscience. So why does the corporate speaker have some right against compulsion?
Brennan-Marquez: if we’re thinking about compelled ideological disclosure—bloggers have to have a banner saying “guns kill people,” we’d ask whether this was commercial or ideological speech. As compelled ideological speech, does the speaker have a claim to the capacity for and exercise of conscience, and my first claim is that a corporation doesn’t. In blogging, we can tie it back to the human speaking not the corporation. If the gov’t wanted the same mandate on the cigarette cartons, that’s a difficult case.
Kapczynski: but I know that Rachel Maddow’s blog is not just written by Rachel Maddow.
Brennan-Marquez: there are also concerns about audience interest and distortion problems.
Corbin: would encourage disclosure of speaker’s objection to gov’t message—that helps correct distortions in that people trust doctors more than the gov’t.
RT: A fact is a fact. It is not a government fact. The FDA requires certain disclosures about supplements—if you said “we disagree with the FDA” on the label, I would expect some objection. You don’t get to advertise shark cartilege as good for cancer, no matter how much you disagree with the FDA. Compare the classic lawyer advertising case: it would be bizarre to say that the only allowed disclosure is one that says “the gov’t requires us to tell you that ‘fees’ don’t include ‘costs,’ which you’ll be liable for even if you lose.” The problem was that the use of “fees” misled consumers.
Kapczynski: so isn’t there tension b/t that stance that disclosure of gov’t source isn’t required and deference to gov’t in finding facts?
RT: no, it’s just a big gov’t.
Balkin: Nexus requirement? Disclosure should be germane to the particular transaction.
Kapczynski: if you’re an NGO delivering safer sex information, can you be required to give the gov’t’s message that prostitution is wrong?
Brennan-Marquez: esp. in the funding context, might be bad policy, but not unconstitutional.
Leslie Kendrick: regret as a legally significant thing? Kennedy says women’s putative later regret of abortion has constitutional meaning. We don’t do that with the right to vote just because one might regret a vote. Corbin’s argument is that in compelled disclosure it might be relevant; if people regret they started smoking, they might want some disclosures about risks up front. (So then we can further concentrate fault in them and not in the providers of the relevant products/services? This is a really interesting point and I don’t think Kendrick is going this direction, but it makes me think about individualism and the way that we use disclosure to offload responsibility on individuals when the overall conditions are created by social structures.) Regretting abortion in the sense of regretting having the abortion v. having a child they didn’t plan to have or regretting the necessity of the abortion—this is very hard to judge. Even if turns out that women regret having an abortion, that might not be relevant—that might be a good argument for putting an ultrasound on a condom package or a wine bottle, but the abortion is the wrong point to intervene in the nature of the regret.
Corbin: there’s no study showing women regret abortion more than they regret giving birth to an unwanted child. Instead it’s an assumption that women are incapable moral decisionmakers. Can imagine a world in which informed consent for all medical procedures included discussion of nonmedical consequences and moralizing about the nature of the procedure, but that’s not our world. Adding layers for one procedure is reflecting a belief about women’s capacity and a belief that regret would only be natural. (These strike me as further good arguments about reproductive rights, but not 1A rights since gov’t can address different kinds of problems—cigarettes are unique too and that’s often given as a special reason to regulate them.)
Kendrick: open-ended persuasiveness and end-determined persuasiveness. Disclosure is always when the gov’t thinks it will have some impact on decisions. But you want to distinguish where the gov’t wants you to make a particular decision. Gov’t has no interest in what decision you make, but just wants you to make the best decision for you. (But tobacco and abortion disclosures slot into the same end-determined category as far as I can tell; you can tell a story about wanting people to understand the risks before they proceed in both cases, but in neither is it particularly persuasive.) Other contexts like sterilization do define informed consent—reproductive rights organizations were instrumental there b/c they thought it was important to think hard about the consequences of sterilization.
Bryan Choi: commercial speech v. speakers. If artificial entities don’t consciences, commercial speech is directed at the public, but other areas are also implicated—incorporation documents require certain disclosures. 5th Amendment: artificial entities don’t have a privilege against self-incrimination, also rooted in conscience. Anything can be ideology; if a sale happens to be involved, is that the test about which you’re concerned?
Corbin: there are lots of considerations in compelled speech, such as chill. Interests of speaker are sometimes present and sometimes not. Doctrine is unclear about commercial/noncommercial, fact/ideology. Look at first principles: why do we care about compulsion? Fewer reasons to care with corporation.
Brennan-Marquez: definitely what I’m concerned about. Distinction between types of entities rather than speech is important. The hard case for him: Reynolds being required to put ideological message on cigarette carton. They have no right of conscience to resist this, but 1A is implicated maybe because of the audience distortion if messages aren’t disclaimed as coming from the gov’t. Another problem is constructive chill on the part of the corporate speaker. If FDA says Reynolds can’t put anything on the carton, that is too much. (Plain packaging?) If the amount of disclaiming required to distance the speaker from the message requires too much diversion from the speaker’s own speech, that’s too much—that’s a way to explain Zauderer—it’s much easier to distance yourself/you don’t need to distance yourself since it’s just a factual disclosure. But that means a fuzzy line between ideological and factual/commercial disclosures.