Friday, March 13, 2020

What is “Government” “Speech”? A View from Charlottesville

Guest Blogger

For the Symposium on Helen Norton, The Government's Speech and the Constitution (Cambridge University Press, 2019).

Richard Schragger

Living in Charlottesville, I could not read Helen Norton’s excellent new book, The Government’s Speech and the Constitution, without thinking about the events of August 2017 and the controversies surrounding the removal of Confederate monuments in states across the South and elsewhere. Professor Norton’s book addresses a number of issues implicated by those events. Obviously, the Confederate monuments that stand in the middle of Charlottesville and that white supremacists were ostensibly “defending” when they rioted here are “government speech.” So, too, the previous vote by the Charlottesville city council to remove the monuments is a form of government expression as are the city’s filings in a lawsuit contesting its authority to do so. President Trump’s statement after the “Unite the Right” rally that there were “very fine people on both sides” is a form of government expression. And if the city does remove its monuments (the Virginia legislature recently lifted a ban on doing so), that too would be a form of government expression. Meanwhile, the white supremacists have asserted in a federal civil rights lawsuit that they were only engaged in “speech” and not in a conspiracy to commit violence—a claim about what kinds of communicative acts the government can punish.

I had not thought very hard about Confederate monuments before 2017, though I had thought some about crosses and Christmas displays; religious government speech has long been a contested topic in the First Amendment and one that Professor Norton discusses extensively. But the events in Charlottesville, and now Professor Norton’s book, have convinced me that the legal categories that we have heretofore used to analyze these kinds of expressive acts are mostly incoherent. Indeed, the very concept of “government speech” is itself troubled, something to which the title of Professor Norton’s book alludes. Professor Norton writes of the “government’s speech”—an activity—not of “government speech”—a doctrinal category. That is because the latter is both too narrow and too contested, as Professor Norton understands.

Nevertheless, I wonder how far Professor Norton would go in dismantling the conceptual apparatus that undergirds government speech. Her project is more constructive. I think she believes that the current doctrine can be applied given some sensible reforms. I am less sure. 


Start with the speech/conduct distinction, which Professor Norton sometimes seems to embrace but that her book also erodes. For example, early in the book, Professor Norton adopts a distinction between “soft” and “hard” power—drawing a distinction between the exercise of government power—speech—that is different from “its lawmaking and other regulatory actions that control behavior.”

Much of the argument of the book, however, seeks to assimilate speech to other forms of government conduct that are susceptible to constitutional regulation. Professor Norton argues strongly against the “sticks and stones” view of speech—the idea that “mere words” cannot do harm or that any harm of speech is not equivalent to kind of harm caused by conduct. And I take it that one of her strategies is to include within the “government speech” category a set of communicative acts—like police interrogation, threats, or government lies—that many of us might call acts.

I agree with Professor Norton’s view that communicative acts often cause harm and I agree with Fred Schauer (and Stanley Fish) that what we call “speech” is really a conclusion that reflects an implicit balancing of the harms of expressive conduct against its benefits in favor of the conduct. Professor Norton describes three kinds of constitutional problems that government speech might pose. The first is that the government’s speech might change a citizen’s choices or opportunities in material ways. The second is what she calls a dignitary or expressive harm: the government’s speech might treat a citizen or groups of citizens as less than equal. The third problem is that the government’s speech might be motivated by an improper purpose.

These constitutional problems are not exclusive to speech, of course. They could characterize the potential constitutional problems with any government act, whether purely communicative or regulative. Whether characterized as “speech” or “conduct,” we’d have to resort to substantive constitutional law to figure out how to proceed in all these instances.

I would slice things a little differently. Government speech, like conduct, can cause material harms, including dignitary ones. But the government’s communicative acts, like all kinds of conduct, might also be wrong because of their social meaning. Here I borrow from expressivists like Deborah Hellman, who consider the constitutionality of government acts—whether “speech” or not—by considering the meaning conveyed by the act. For expressivists, the speech/conduct distinction is irrelevant. The only thing that matters is whether the government conveys a demeaning or denigrating message through its activities—for example, whether it conveys a message of exclusion to one religious group by celebrating another. That test can be applied to the erection of a Confederate statue just as it can to any other government activity. On this account, an “expressive equal treatment” principle is not some outlier—secondary to the “real” harms that equal protection is trying to address. Instead, it is at the core of equal protection. 


Professor Norton also writes lucidly about the term “government” in “government speech.” Figuring out when the government is speaking is what Professor Norton calls a “first-stage problem.” She proposes a transparency principle, the purpose of which is ensure that citizens are not confused about who is speaking. This enables them to hold the government politically accountable for its speech.

She is fully aware of how the Court can manipulate the public/private distinction to suit its needs, however. If speech is private, it cannot be regulated for content; it is also not limited by the Establishment or Equal Protection Clauses. If speech is public, it need not abide by the First Amendment’s content neutrality requirements, though it might need to comport with the Establishment and Equal Protection Clauses. The decision about whether speech is public or private is often outcome determinative.   

There are lots of difficulties here and Professor Norton writes about them with considerable clarity. Consider prayers offered by volunteer clergy at the invitation of a town council. The Court has held that such prayers are more “private” than “public” and that it would be problematic for the government to modify or reject these prayers even if they are sectarian in nature. On the other hand, the Court has ruled that a Ten Commandments monument accepted as a gift and erected in the city’s public park is government speech and thus does not impose an obligation on the city to treat all speech equally by accepting other gifted monuments.

I don’t think the concept of “government speech” is doing much work in these cases. Instead, what is driving the Court’s public/private determinations is a substantive vision of the Establishment Clause. The conservatives on the Court reject an expressive equal treatment principle, which had come in the form of Justice O’Connor’s famous “endorsement test”—intended to ensure that the government did not denigrate religion or specific religions. In the recently decided Bladensburg Cross case, the Court has mostly rejected non-endorsement. It now looks increasingly plausible that the Court will dismantle most if not all restrictions on the government’s religious speech.

There is a deeper problem, however, with the very subject of “government speech” insofar as it takes the state action doctrine as a given. If we assume that speech, like conduct, causes serious harm, why are “private” speakers permitted to cause that harm when “public” speakers are not? Why not hold citizens to a principle of expressive equal treatment? Indeed, the muddiness of the categories provides opportunities for abuse. Donald Trump’s denigrating, demeaning, and dangerous tweets (some supporting white supremacy), for example, reside in that purposefully ambiguous space between public and private.

This space gives speech theorists fits; the maintenance of the public/private distinction is seen as necessary to allow the government to say things we want it to say while preventing the government from engaging in censorship. But many people not steeped in the state action doctrine do not readily distinguish between the marching neo-Nazis and a statue of Robert E. Lee in a city park, or between the government and Facebook, or between campaign donors and politicians. And they are not wrong to find these distinctions puzzling. The Robert E. Lee statue in the middle of Charlottesville was a gift funded by private donations, erected by whites in a whites-only park, and dedicated at a time when the Ku Klux Klan, a  paramilitary terrorist organization with deep ties to the ruling class, marched through black neighborhoods. No wonder so many people are disillusioned with the concept of “free speech.”

The Political Economy of Government Speech
Professor Norton’s book is a terrific tour through these many contradictions and complexities; I have and will continue to rely on it in my own work. She is acutely aware of the slipperiness of existing categories. I wonder, however, if the project is doomed—not because Professor Norton fails to offer correct approaches to difficult doctrinal problems, but rather because in this current political environment, government speech doctrine is not susceptible to principled application.  

The current Court’s fetishization of speech has a clear political valence. The Roberts Court is (to use Justice Kagan’s phrase) “weaponizing” the First Amendment to dismantle the regulatory state as it retreats from limits on the government’s religious (and other) speech. Meanwhile, the political donor class and white supremacists both march under the banner of “free speech.” Five justices think that the President’s clear anti-Muslim statements and (other expressions of animus) are not grounds for the invalidation of federal laws that target discrete ethnic, racial, or religious groups. At least four justices think that Confederate battle flags on state-issued license plates are protected speech. These positions seem to have more to do with the political sympathies of the justices than their decided views on government speech.

It also may be that the equality/speech conflict is intractable. Dangerous speech is not being addressed by the Supreme Court. The Court can’t or won’t constrain the giant social media companies, who now dwarf the government in their ability to control the speech environment—with huge public costs. And it is apparent that most of justices on the Court think that speech trumps equality all the time. That view is tough to take from my perspective sitting in Charlottesville, where the equality/speech trade-off looks increasingly one-sided.

Richard Schragger is Perre Bowen Professor at the University of Virginia School of Law. You can reach him by e-mail at

Older Posts
Newer Posts