Freedom of Thought, Compelled Speech, and Expressive Businesses

I spent Friday of last week at a one-day symposium called Revitalizing Freedom of Thought hosted by the Cornell Law Review. There was a great slate of presenters and stimulating discussion throughout the day. I moderated and offered some of my own thoughts for a panel on Free Thought and Free Exercise, but in today's essay I want to offer an idea that grew out of remarks by USC law & philosophy professor Erin Miller during the first panel. Professor Miller's paper builds on and thus led me to revisit an idea I first encountered many years ago when I was editing the chapter on West Virginia State Board of Educ. v. Barnette for my book Constitutional Law Stories.

In Barnette, the Supreme Court invalidated a mandatory public school flag salute. The chapter was co-authored by Professors Vincent Blasi and Seana Shiffrin; the idea that serves as my current point of departure appears in some of Professor Shiffrin's solo writings as well. It goes like this: Avoiding misattribution is not the only, or even the primary, reason to prevent the government from compelling speech. Rather,

The focus of constitutional concern should be turned inwards towards speakers, not outwards towards their audiences. We suggest that what underpins Barnette is the First Amendment interest in the speaker's freedom of thought and freedom of conscience. The main constitutional defects with the mandatory pledge lie in the attitude and the message the recitation requirement conveys toward the speaker and the risks that such a requirement will exert an inappropriate influence on the speaker's freedom of thought.

There is evidence from psychology to support the worry that by repeating a proposition, a person will come to believe it. Mandatory flag salutes and the like thus raise the troubling prospect of indoctrination rather than education. Meanwhile, Shiffrin and Blasi are also onto something in thinking that the internal effects of compelled speech may be more substantial than the risk of external misattribution, especially because misattribution can often be counteracted with a disclaimer.

What follows? For one thing, we might want to reconsider Wooley v. Maynard, which upheld the right of New Hampshire motorists who disagreed with the state's motto ("Live free or die") to tape over it on their license plates. It's very unlikely that anyone would have attributed that sentiment--which, at the time at least, appeared on all New Hampshire license plates--to the drivers or owners of New Hampshire cars. Thus, misattribution was a low risk in Wooley.

But so was the risk of internal indoctrination. Just as external observers will not likely attribute the state motto to any particular New Hampshire motorist, so drivers of cars with New Hampshire plates are unlikely to experience getting behind the wheel as an affirmation of the state motto.

Consider that as I wrote that last sentence, I realized that I wasn't sure what message New York State has placed on my own license plate. I guessed that it was either "Empire State" or "Excelsior" but had to go look to verify that it's the latter. I then looked that up and discovered that Excelsior has been the state motto since 1778 and means "ever upward," which seems like the wrong message to put on a car, which should head upward only when the road goes uphill. Otherwise, cars should head forward, backward, to the right, or to the left. Nonetheless, I won't tape over "Excelsior."

In the end, however, I don't have a strong view about whether Wooley was correct as an original matter, but if it's wrongly decided, it doesn't seem so wrong as to warrant overruling. In their chapter in my book, Professors Shiffrin and Blasi acknowledge that Wooley is a harder case than Barnette. They also suggest (in footnote 135 at page 445) that the case might divide people based on how important driving is to them, noting that one of them lives in Los Angeles and the other (at the time) in Manhattan.

Let's move on to a more recent and consequential case.

Next consider 303 Creative v. Elenis, in which the Supreme Court upheld the free speech right of a web designer to refuse to create a wedding webpage for a same-sex wedding notwithstanding Colorado's public accommodations law. Justice Gorsuch's opinion for the Court treated the law as applied to 303 owner Lorie Smith as compelling her to create bespoke expression contrary to her beliefs and announced a rule under which the right Smith won applies to all those engaged in "expressive" businesses. Because Colorado had stipulated to the proposition that bespoke web design is expressive, the majority said it did not need to set forth a detailed test for distinguishing expressive from non-expressive businesses.

Some of the commentary criticizing 303 Creative (including by me) points to likely trouble down the road, when the Court will need to sort out which businesses are expressive. Photographers? Probably yes. Florists? That seems more borderline, as it is hard to see how the arranging of flowers--which is certainly creative--carries any articulable message regarding same-sex marriage or anything else. What about bartenders, who often banter with customers? Etc.

I want to suggest, however, that the category of expressive businesses might be over-inclusive. If the harms of compelled speech are either misattribution by the audience or internal effects, perhaps compulsion is only problematic when there is a risk of at least one of these. In 303 Creative, the thrust of Justice Sotomayor's dissent was that Colorado forbade the conduct of discriminating based on sexual orientation, with only an incidental impact on Smith's expression. But during the oral argument, Justice Justice Sotomayor pressed Smith's lawyer on whether the websites Smith designs are really her speech at all. Smith "isn't inviting" guests directed to the website "to the wedding. Lilly and Mary are." That example shows that there is not much risk of misattribution.

What about the more serious risk--that the compelled speech will affect Smith's thoughts in the way that the compelled flag salute does in Barnette? That does strike me as possible but it matters how.

There are two pathways by which compelled expressive services contrary to one's thoughts could change those thoughts. Suppose that the website Smith is asked to create for Lilly and Mary involves creative work in the design elements but that none of the design elements or information conveyed carries an articulable message about same-sex marriage. In that circumstance, Smith will have provided an expressive product for a same-sex marriage, and she may in so doing come to see herself as more supportive of same-sex marriage than previously, but the change in her thought will not have been mediated in any way by any message she expressed. A vendor providing non-expressive products like folding chairs could equally come to see himself as more supportive of same-sex marriage in virtue of having performed the act. But the chair vendor wouldn't have a valid free speech claim, and, I'm suggesting, in this scenario neither should Smith.

So consider the second pathway. Suppose that Smith is asked to create a website that incorporates statements in support of same-sex marriage. "Love is love," for example, or "may God bless the union of Philip and Steven." In those circumstances, the risk identified by Professors Shiffrin and Blasi would be present, because the potential change in thoughts is mediated by the content of compelled expression.

Accordingly, I would support a narrowing of the rule announced in 303 Creative: purveyors of expressive goods and services may be able to object on compelled speech grounds but only when they are asked to provide expressive goods and services with an articulable message that is contrary to their views.

-- Michael C. Dorf