Monday, June 10, 2019

Free Speech and Boycotts Revisited

by Michael C. Dorf (cross-posted at The Volokh Conspiracy)

In Arkansas Times v. Waldrip, the US Court of Appeals for the Fourth Circuit will decide whether an Arkansas law requiring public entities (including contractors with the state) to certify that they do not boycott Israel or companies that do business with Israel violates the First Amendment. The district court held that it does not. Various champions of free speech have filed briefs arguing that political boycotts are protected free speech. Three scholars who also fashion ourselves champions of free speech -- Professors Andrew Koppelman, Eugene Volokh, and I -- take the opposite view. You can read our brief here. You can read a summary of our argument in a blog post by Prof Volokh here. His blog post also contains links to the briefs on the other side. Here I'll add a few words that go beyond what I wrote on this topic back in February and what we say in the brief.

Some substantial portion of the argument is doctrinal. We cite Rumsfeld v. FAIR for the proposition that a boycott--as opposed to speech accompanying a boycott--is not speech. The other side cites NAACP v. Claiborne Hardware for the proposition that political boycotts are protected speech even apart from the speech that accompanies them. For the reasons we lay out in our brief and that Prof Volokh summarizes in his blog post, I think we have the better doctrinal case. That said, I'll concede for the sake of argument that there is sufficient wiggle room in this and other constitutional doctrines, that one could say that it is an open question whether boycotts themselves--in contrast to speech accompanying boycotts--should be deemed protected speech.

So the question then becomes this: Should boycotts be treated as speech? Answering that question requires both conceptual analysis (more about that below) and some sense of the consequences of treating boycotts as speech. One apparent consequence--as we highlight in our brief--would be the gutting of public accommodations laws. If refusal to buy goods from Israel is speech because one's motive is opposition to Israeli policy (or Israel's existence), then  ideologically motivated refusal to sell goods to LGBT customers or for same-sex weddings is also speech.

In the briefs for the other side and my discussions with people who come down on the other side, I have encountered two main arguments why one supposedly needn't worry about anti-discrimination law. First, they say that there is a compelling interest in public accommodations laws forbidding anti-LGBT discrimination but not in laws like the Arkansas anti-boycott-of-Israel law. Second, they distinguish between purchaser boycotts (protected they say) and seller boycotts (unprotected they say). Neither contention is persuasive.

(1) I agree that there is no compelling interest justifying the Arkansas law or others like it. Indeed, I think such laws are unwarranted. I oppose them on policy grounds. I also agree that there is a compelling interest in public accommodations laws. However, one must think strategically about such issues. The question is not what some liberal law professors regard as a compelling interest but what a majority of the Supreme Court will ultimately regard as compelling. I have no confidence that the Court would find a compelling interest in forbidding discrimination on the basis of LGBT status.

Indeed, we have good reason to think the opposite based on Justice Alito's majority opinion in Hobby Lobby v. Burwell. Here's what I wrote last month:
Dissenting in Hobby Lobby, Justice Ginsburg charged that the Court's aggressive use of [the Religious Freedom Restoration Act (RFRA)] could undercut laws that protect against discrimination based on race and sex, including sexual orientation. Not to worry, responded Justice Alito for the majority: RFRA allows that other federal statutes -- such as anti-discrimination laws -- can override religious objections where those other federal statutes amount to the least restrictive means of advancing compelling interests. And, he added, "[t]he Government has a compelling interest in providing an equal opportunity to participate in the work-force without regard to race, and prohibitions on racial discrimination are precisely tailored to achieve that critical goal." Notably, despite the fact that Justice Ginsburg expressly pointed to LGBT discrimination, Justice Alito responded with respect to racial discrimination only, thereby implying that the Court's now-even-more-conservative majority might think there is no compelling interest in addressing LGBT discrimination.
As a practical matter, my many friends on the other side of the Arkansas Times case are wrong that their position poses no threat to anti-discrimination law.

(2) What about the claim that purchaser boycotts are different from seller boycotts? The claim is problematic in at least two respects.

(a)  In an economic sense, there is no real difference between a purchase and a sale. When Henry sells Georgia a wedding cake, Henry exchanges the wedding cake for money from Georgia. Henry had a cake and Georgia had some money. They trade. It's hard to see why one side of the transaction is expressive and the other isn't. It's even harder when we think about the function of money.

Imagine a barter economy in which people meet at the market to exchange goods. Jim is a baker and Sheila is a farmer. Jim trades his baked goods for raw materials he uses to make more baked goods as well as for finished products that his family needs but that he doesn't produce. Sheila trades the wheat she grows on her farm for items she needs. Now let's suppose Sheila is trying to decide whether to trade some of her wheat with Jim in exchange for a cake. If, in the money economy of the prior paragraph, Georgia, as the "buyer," is engaged in speech by refusing to purchase cakes from Henry (perhaps because Henry obtains some of the ingredients for his cakes from Israel), then Sheila is equally engaged in speech when she refuses to hand over a sack of wheat to Jim in exchange for a cake. (Sheila doesn't want to trade with Jim because Jim trades cakes for sugar with Ophelia, whose beet farm sits on land that Sheila believes rightly belongs to indigenous people.) But if refusing to exchange wheat for a cake is speech, then obviously so is refusing to trade a cake for wheat also speech, assuming some ideological motive, such as opposition to the use of the cake in a same-sex wedding ceremony. And because there is nothing more or less expressive about barter than monetary transactions, it follows that selling is every bit as expressive (or non-expressive) as buying. And that makes sense intuitively.

Economic transactions either are or are not inherently expressive. Thus, if the plaintiff and its amici are right that purchase boycotts are speech, so are seller boycotts. The purchaser/seller distinction does not work as a basis for protecting public accommodations laws against the consequences of treating purchaser boycotts as speech.

(b) Meanwhile and also troublingly, a rule that treats boycotts -- even if only purchase boycotts and not seller boycotts -- as inherently expressive threatens to Lochnerize the First Amendment. Suppose that before Congress zeroed out the tax due for failure to maintain minimum coverage under the Affordable Care Act, Tea Partiers organized a boycott of mandated health insurance and defended on free speech grounds. Or suppose that right now, with the employer mandate still in force, employers with ideological objections to the ACA refuse to purchase health insurance for their employees or to pay the associated financial penalty. These would not need to be religiously motivated refusals. Rather, if refusal to purchase a good or service based on a political view is inherently expressive, then the individual and employer mandates could only be validly applied if they satisfy strict scrutiny as applied by the increasingly conservative federal judiciary. Even the very creative lawyers who challenged the ACA on multiple grounds lacked the temerity to suggest that it violates free speech. But if the challengers of the Arkansas law prevail, such a claim could be coming next. So could a wide range of other economic-liberty-as-free-speech claims challenging progressive legislation.

* * *
Here's another issue: The ACLU, which has filed on the other side of this case, describes the Arkansas law as "suppressing one side of a public debate." After all, Arkansas does not require state contractors to certify that they do not boycott goods from or firms that do business with Palestine or anywhere else. Isn't that a problem?

No, unless one assumes the conclusion. If boycotts are speech, then sure, a law targeting boycotts targets speech, and a law that targets only some boycotts would be "content-based." But the very question under discussion is whether to treat a boycott as speech. As I have explained, there are conceptual and practical reasons not to.

To be sure, a law targeting a boycott could implicate free speech under special circumstances. As I observed in my February post and as Profs Koppelman, Volokh, and I acknowledge in our brief, the Arkansas law could implicate the First Amendment if the record contained evidence of censorial motivation on the part of the Arkansas legislature. Suppose that a majority (or a decisive minority) of Arkansas legislators voted for the legislation at issue for the purpose of suppressing the message sent by boycotts of Israel rather than because of what they regarded as the economic impact of boycotts of Israel. That might well violate the First Amendment.

But absent such evidence of illicit subjective motive, the fact that a law treats economic transactions involving one foreign country or territory differently from economic transactions with others is not a First Amendment issue, even if the difference implicates a longstanding conflict. Purchases of certain goods from Iran but not the same goods from Saudi Arabia are forbidden. Tariffs differ by country and region. Etc. Nobody thinks these differences implicate the First Amendment, even if someone wants to make forbidden purchases of goods from North Korea or Venezuela to make a point.

Hold on, you say. Laws governing tariffs and sanctions are enacted at the federal level. What business does the state of Arkansas have in regulating economic transactions with foreign firms and governments? That's a fair question. Perhaps there could be a challenge to the Arkansas law under some conception of a "dormant foreign commerce clause" doctrine. The Supreme Court avoided ruling on whether there is such a doctrine in Crosby v. National Foreign Trade Council with respect to a Massachusetts law that forbade state entities from purchasing goods and services from companies doing business with Burma. But even if there were a dormant foreign commerce clause claim available, that would have nothing to do with free speech.

Even assuming the doctrinal question is open, there are powerful conceptual and pragmatic grounds to conclude that boycotts, absent more, are not expression, and that therefore, absent proof of censorial legislative motive, laws forbidding boycotts do not infringe free speech.


Joe said...

This sort of thing seems discriminatory and that can be an issue beyond the speech issue.

As to the speech issue, I think at some point public accommodations laws override that or we are in trouble. Let's take some business that simply doesn't really have an expressive content even to the extent people try to make cake or flowers (for weddings and such) to fit there. Fixing eyeglasses or something. (Noting if you really try, probably can fit basically everything here; but let's put that aside.)

Laws that require business owners not to discriminate (likewise allow owners to so require their employees) in practice require some sort of speech. Let's say greeting customers. What if some owner decides, "fine, I'm forced to serve 'x' group, but I won't say "Thanks for your business" to them! That's speech!" Being a member of the wait staff involves communication with customers. etc.

Is this sort of thing deemed "not really speech" like common booking questions aren't considered testimonial for purposes of the self-incrimination clause? Or, will you be allowed to do that for gays but not based on race?

Michael C. Dorf said...

"Seems discriminatory" is among the reasons I oppose such laws on policy grounds. Proving that it violates equal protection would be considerably harder, because the law on its face does not employ any suspect classifications. Maybe it has a disparate impact on Muslims, although maybe not. Lots of Christians and many Jews are active in BDS, and I don't know whether there's anything like the record of Trump's statements for what became the Travel Ban, which had a much starker disparate impact.

Second point raises interesting questions, but I think they would fall within a possible as-applied challenge, as Eugene elaborates in his post.

Joe said...

The discriminatory thing also includes the last part of the post, which as noted, isn't speech related, but it can be a problem regardless.

I don't know if strict scrutiny is required. Being able to not do business with every single country in the world but one seems (in a legal sense) irrational to me. Would a state be allowed to only protect French customers from discrimination?

But, that is for another day.

Marty Lederman said...

As Mike, Andy and Eugene know, I tend to side with them on this issue--and to agree that the First Amendment argument here is very unwise because of its implications for the marriage/vendor cases. How I arrive at that conclusion, however, might be slightly different (although I think they would all agree with my principal argument, below):

It strikes me that some boycotts are "expressive," and thus entitled to at least nominal 1A scrutiny, whereas others--those principally motivated by a desire either to deny funds to the disfavored nation, or to prevent the putative buyer from being complicit in the nation's disfavored conduct -- are not expressive. So it's a mistake to argue that boycotts aren't expressive, just as it's a mistake to argue that refusals to sell cakes or burn draft cards) are categorically beyond constitutional scrutiny.

Even as to those boycotts that are expressive, however, the key question (cf. O'Brien, or Texas v. Johnson) is whether the state's *regulation* of boycotts is itself designed to suppress a message that the boycott might send. And I have a hard time believing that the Israel-specific laws at issue here are ill-motivated in that way. Their principal, and perhaps exclusive, object is to affect Israel's *economic* well-being, by encouraging business to be done with Israeli concerns. And if I'm right about that, there isn't any Free Speech problem, even if a particular boycott *is* intended to be--and is--expressive.

Michael C. Dorf said...

I agree with Marty entirely. We don't deny that a boycott CAN be expressive. Our claim is simply that it isn't necessarily so, and that therefore, even when a boycott is expressive, it doesn't follow that a law disadvantaging those engaged in the boycott targets the boycott because of its expressive character.

Marty Lederman said...

I figured you agreed, Mike, and am glad to hear it. The brief, unfortunately, tends to leave the impression that whether or not a boycott is "protected" (a vague term) depends upon whether it's "inherently expressive," as opposed to--as you and I agree--why the state is prohibiting such boycotts.

Joe said...

I'm not as convinced that such laws aren't also partially based on opposition of the message as well. This area is ideologically freighted.

Michael C. Dorf said...

A promos of Joe's last comment: I'm also not CONVINCED they're not based on opposition to the message, but that's not the issue; the burden of persuasion rests with the plaintiff, which, as far as I can tell, offered no evidence of illicit censorial motive.

Michael C. Dorf said...

"propos," not promos.

Samuel Rickless said...

I'm confused. Why is this being treated as a freedom of *speech* issue, rather than a freedom of *action* issue? Under substantive due process precedents, Arkansas is placing an arguably undue burden on the exercise of freedom of action (boycotting Israel or companies that do business with Israel) by stipulating that public entities must certify that they are not exercising their freedom of action to engage in such boycotts. Imagine, then, that Arkansas decides to require every public entity (including contractors with the State) not to get married to someone of the same sex, and not to employ gays or lesbians, and not to march in a Gay Pride parade, and not to attend a same-sex wedding, and not to go to a gay bar, and so on. In fact, if Arkansas is permitted to place this particular BDS-related restriction on public entities, what is to stop the State from placing a whole bunch of other restrictions, unrelated to LGBTQ rights, on them? What is the point of freedom if one has to choose between (a) going out of business or completely retooling one's business at great personal cost and (b) not exercising one's (constitutionally protected) freedom of action? I am reminded of Justice Harlan's argument that Homer Plessy should not be pointing to the *property* (in the reputation of being white) aspect of the DP clause, but to the *liberty* aspect instead. What am I missing?

John Marcher said...

Those are some interesting things you're imagining there, but I think the hornbook answer is that there isn't much if any of a substantive due process right to buy or not buy goods, much less goods from country x but not country y, unless the good you're purchasing has something to do with some fundamental right (e.g., contraceptives), while there are, doctrinally, substantive due process rights to marry and, probably, date members of the same sex, or attempt to do so at bars. Marching in a Gay Pride parade = speech; your employment hypothetical probably poses an equal protection problem more than an SDP problem. Then there's an unconstitutional conditions problem with tying contracting to all the conditions you pose.

I'm not quite sure that what Lederman's saying is consistent with the combination of FAIR and Town of Gilbert. Once you suppose that these boycotts are some way of symbolically saying that one doesn't like Israel, I don't know that it matters if the state's purpose in conditionally banning them isn't suppression of that message; it seems, perhaps, enough that boycotts with certain symbolic content are banned and boycotts with other symbolic content are not. That said, though some boycotts are undoubtedly intended to express certain things, they are not expressive in FAIR's (perhaps unduly crabbed, but the law) sense, as no one can appreciate this expression, or even perceive the inaction that makes it up, without a lot of pure speech calling attention to it and explaining it.

JimM47 said...

(cross-posting the comment I left on the Volokh Conspiracy re-post of Professor Dorf's post)

Professors Volokh and Dorf, I appreciate that you continue to engage with criticism.

Professor Dorf criticizes a distinction between purchaser boycotts and seller boycotts, saying there are unworkable elements. Fair enough. Perhaps that isn’t the right place to draw the line. I’ve pondered a slightly different distinction between public and private transactions, and Professor Volokh has pointed out (in comments at VC) that this distinction is in tension with current doctrine. Perhaps a third option is business transactions versus end-consumer transactions.

But maybe the more honest way for me to respond is that I have a particular fact scenario in mind. Imagine that a private bus company has imposed racially segregated seating. It seems like (1) it should be constitutional to ban that practice, and (2) it should be unconstitutional to ban private citizens from boycotting the bus company until it changes its practices. Now, I don’t know exactly what kind of line needs to be drawn to permit this — i.e. seller-purchaser, etc. But my strong prior is that it must be possible to draw some principled line here. And I am going to retain that prior unless defenders of the Israel-boycott law start forthrightly saying “no, if you want the anti-segregation law to be constitutional, then the law forbidding the anti-segregation boycott has to be constitutional too.”

Its very hard to give up on the feeling that Claiborne supports the intuition that the anti-segregation boycott is constitutionally protected. I know that Professor Volokh has argued that Claiborne protects only the speech adjacent to the boycott. But Claiborne is littered with the word “conspiracy.” Where else in the law do we say that an underlying act can be criminalized, but that a conspiracy to commit that same act cannot? Cf. New York v. Ferber, 458 U.S. 747, 761 n.14 (1982). It seems like a necessary implication of Claiborne that the boycott itself had some protected status — whether that was under the First Amendment or (more likely) as a matter of substantive due process.

Samuel Rickless said...

Here's a two part post. Apologies for the length.

That's helpful, John, thanks. I guess I should have said that I depart from the hornbook this far: I think that the right to engage in economic activities is (or, at least, can be, depending on the activities) fundamental. In case you are worried about Lochner, there is an answer (which I've now published--, at 473-474) to why Lochner was wrongly decided, consistent with the proposition that liberty of contract is a fundamental right.

But here are two other problems, both of which bring us back to the free speech issue.

1. First, the Arkansas law requires public entities to *certify* that they do not engage in a boycott of Israel or companies that do business with Israel. I suggest that the question whether boycotts are a form of speech is a red herring. The relevant question is whether the *certification* is a form of speech. Now *certification* presumably occurs when one signs a document that states something like the following: "I certify that I am not engaged in a [BDS-type] boycott." Signing such a document is a form of speech, and is no different from *saying* (by uttering vocables) that one is not engaging in the relevant kind of boycott. Thus, the Arkansas law implicates freedom of speech, but not via the assumption that boycotts themselves are forms of speech.

Samuel Rickless said...

2. Second, the Arkansas law is completely useless as a tool to prevent boycotts of Israel or of companies that do business with Israel, even though that is its stated purpose ( It's useless because, as long as one does not say that one is engaging in a boycott, there is no practicable way for someone else to prove (by an appropriate legal standard) that one is engaging in the boycott. The reason for this is that every boycott can be represented as a mere decision, for reasons that have nothing to do with a boycott, not to do business with Israel or with this or that company: the goods and services are not of the right sort, or not of the right quality, or whatever. So if I am a contractor who does not want to do business with Israel or with any company that does business with Israel, then I can still engage in a BDS-type boycott and sign the certification without fear of negative repercussions. All of this is obvious, and we may presume that the legislators who passed the measure and the governor who signed it recognized this. But this does not mean that the law has no effects. By its own terms, the law says this:

"A company’s statement that it is participating in boycotts of Israel, or that it has taken the boycott action at the request, in compliance with, or in furtherance of calls for a boycott of Israel, can be considered by the Arkansas Development Finance Authority as a type of evidence, among others, that a company is participating in a boycott of Israel."

So, as the law itself contemplates, its effect will be to prevent public entities from *stating* that they are boycotting Israel or companies that do business with Israel. (For if the public entities so state, then this will be taken as evidence that they are violating the express terms of the law.) This is the *only* predictable effect of the law, and the effect is on *speech*.

Now for some reasoning. Suppose (i) that a law requires that one not do A, (ii) that everyone knows (and knows that everyone knows) that it is possible for one to do A despite the law without negative repercussions, and hence that the law will not have the effect of preventing anyone from doing A, (iii) that (everyone knows, and everyone knows that everyone knows that) the law's evidential requirements with respect to A entail that doing B is sufficient evidence of doing A, and hence that the only practical effect of the law is to get people not to do B, then (c) the law's point or purpose is to get people not to do B.

Mike says:

"To be sure, a law targeting a boycott could implicate free speech under special circumstances. As I observed in my February post and as Profs Koppelman, Volokh, and I acknowledge in our brief, the Arkansas law could implicate the First Amendment if the record contained evidence of censorial motivation on the part of the Arkansas legislature. Suppose that a majority (or a decisive minority) of Arkansas legislators voted for the legislation at issue for the purpose of suppressing the message sent by boycotts of Israel rather than because of what they regarded as the economic impact of boycotts of Israel. That might well violate the First Amendment."

Mike goes on to say that the law would not implicate 1A "absent such evidence of illicit subjective motive". But it seems to me that we don't need "special circumstances", nor is it necessary for us to peer into the minds or brains of the legislators to find sufficiently probative evidence of the law's real purpose. All we need is the reasoning I've described.

JimM47 said...

Professor Lederman says in the comments:

"I have a hard time believing that the Israel-specific laws at issue here are [designed to suppress a message]. Their principal, and perhaps exclusive, object is to affect Israel's economic well-being, by encouraging business to be done with Israeli concerns."

That seems like a slight-of-hand that would make it very hard for the O'Brien test to ever have bite. Virtually every restriction on expressive conduct can be recast as having the purpose of preventing the harm that the communicative impact of that expressive conduct seeks to bring about.

Put another way: what other threat to Israel's economic well-being is addressed by this law other than the threat of people acting to send a message?

If this law is not designed to suppress a message, then it seems both over-broad and under-inclusive. First, it extracts a piece of speech from people who have no practical ability to affect Israel with their personal conduct (overbreadth). Second, it does not cover purely economic decisions that are not "boycotts" but may ultimately have greater effect on Israel's economic well-being (under-inclusiveness). Third, it has a 20% escape clause (under-inclusiveness).

Joe said...

The law has this provision:

"a public entity shall not: Enter into a contract with a company to acquire or dispose of services, supplies, information technology, or construction unless the contract includes a written certification that the person or company is not currently engaged in, and agrees for the duration of the contract not to engage in, a boycott of Israel; or

(2) Engage in boycotts of Israel."

[there is an exception for those who provide a significant cost reduction or the contract is relatively minor in monetary value]

Why should "the person" not be engaged in a boycott of Israel? The provision requires a loyalty oath that does not appear to be related to the public contracting involved. Why should a personal decision to boycott be relevant here? And, "boycott" can include expressive boycotts.

I would be interested to know the need for such a singular provision. What is the economic stakes of the BDS boycott here? I would think that there are a variety of boycotts out there. Why is Israel specifically protected by so many states? I don't think it is merely money but use qualifiers like "seems" since I didn't study the matter closely.

Such speech + action mechanisms, especially as a form of regulation of government speech, might be allowed. But, I am wary. Appreciate the discussion.

David Bernstein said...

Mike, you are neglecting a salient point. Many advocates of anti-BDS laws quite reasonably believe that such laws are antisemitic, both in the sense that boycotts of Israel-related entities in the U.S. have a huge and obvious disparate effect on Jews, and because the history of the BDS movement is rife with antisemitism. Thus, your public accommodations point is even stronger than you acknowledge. Regardless of what anyone else thinks, if the Arkansas legislature believes that its anti-BDS law is preventing discrimination against Jews, and that belief is not wholly unreasonably, to invalidate the law on First Amendment grounds would directly be a holding that antidiscrimination laws that prohibit refusal to deal, i.e., all public accommodations and employment discrimination laws.

Joe said...
This comment has been removed by the author.