Government Speech

The importance of the category, government speech, lies in its exemption from the general rule that the First Amendment mandates government neutrality in its treatment of speech, that is, content neutrality as to subject matter and viewpoint neutrality as to how normatively the subject matter is addressed. There are a few isolated exceptions, like true threats, but otherwise content-based restrictions must meet the tough strict scrutiny test that requires that government justify abridging free speech by pointing to a compelling interest and a narrowly tailored law. The importance of the general rule of government neutrality is evident in the celebrated words of Justice Jackson: “If there is any fixed star in our constitutional constellation, it is that no official, high or petty, can prescribe what shall be orthodox in politics, nationalism, religion, or other matters of opinion.”

W. Va. State Bd. Educ. v. Barnette, 319 U.S. 624, 642 (1943).

Why the fixation on speech neutrality? The central argument has always been that when the government speaks, its voice is not like that of everybody else, for it “is likely to be the biggest, loudest, best-funded speaker on the block—by far.”

Steven D. Smith, Why Is Government Speech Problematic? The Unnecessary Problem, the Unnoticed Problem, and the Big Problem, 87 Denv. U. L. Rev. 945, 950 (2010).

However, government does not possess only advantages, for it is also true that unlike private persons, government may be held politically accountable for its speech. Too, a portion of the population regards government as inherently corrupt and hardly believes it at all. Still, the examples of abusive government speech in authoritarian societies—think Goebbels in Nazi Germany—have sensitized us to its extraordinary potential impact. In America, too, the periods of major attacks on political dissent—the sedition controversy under John Adams, the repeated repression of abolitionists prior to the Civil War, the suppression of anti-war speech during World War I, and the Red Scare of the 1940s and ‘50s—saw government speech playing a dominant role. Other high profile incidents, like the 1996 official harassment of a heroic security guard, Richard Jewell, as an Olympic bomber, have added to the public skepticism toward government speech.

The episode was made the subject of a 2019 movie. See Richard Jewell (Warner Bros. Pictures 2019).

To the extent that government reflects the will of elected officials, it is not supposed to be neutral. To take an example, if elected officials decide to fund public schools, but not charter schools or parochial schools, government will favor public schools and not charter schools and parochial schools, and government speech will reflect that decision. Indeed, if government ignored the will of officials and in its speech was neutral as to charter schools, parochial schools, and public schools, we would say something was deeply wrong. Similarly, if government instructs those public schools to train students to be good citizens and instead schools were neutral on the question, we would also say that something was deeply wrong. If we do not expect government to be neutral, why would we expect (or demand) that government speech be neutral? The purpose of speech generally is to affect the conduct and beliefs of others; why should government speech be different?

The noted polemicist, Stanley Fish, answers that valorizing neutrality

will always and necessarily proceed from the vantage point of some currently unexamined assumptions about the way life is or should be, and it is these assumptions, contestable in fact but at the moment not contested or even acknowledged, that will really be generating the conclusions that are supposedly being generated by the logic of principle. … . Judgment without partiality … is not an option for human beings.

Stanley Fish, The Trouble with Principle 3, 113 (1999). Jedediah Britton-Purdy et al. argue that the ideal of neutrality should be replaced by that of equality. Jebediah Britton-Purdy et al., Building a Law-and-Political-Economy Framework: Beyond the Twentieth-Century Synthesis, 129 Yale L. J. 1784, 1824 (2020).

Neutrality, like other principles, is a rhetorical trick, a con that both justifies and hides the play of self interest. Yet there is a circularity here; of course, the self pursues what it is interested in and prefers. What else could it pursue? The real issue (which Fish dodges) is what the interest is. That I pursue my self interest need not be a bad thing or hostile to the public good, however this is conceived. Moreover, even if, arguendo, one accepts that abstractions are mere rationalizations, may they not be rationalizations for beliefs in tolerance and fair play? Taken seriously, Fish seems to reject the whole enterprise of rational justification, urging us to rely on our own moral commitments. But suppose these commitments are also mere rationalizations for self-interest? Why should they be exempt from this general phenomenon? Wouldn’t examining these commitments through the lens of neutrality at least introduce the possibility of discouraging abuse? In the end, one is reminded of Voltaire's warning that the best can be the enemy of the good; the neutrality principle, like other human constructs, is radically imperfect, but this is not cause for its rejection. Indeed, it contributes to free expression's multiple goals, including political accountability,

Vincent Blasi, The Checking Value of First Amendment Theory, 2 Am. Bar Found. Res. J. 521 (1977).

Are there, then, too many goals? The political theorist, Judith Shklar, concluded that individualism and pluralism had ruled out a consensus on virtues, leaving a consensus only on a single goal: condemning cruelty as the worst vice. Hence, a political skepticism built on fear, in place of America's congenital optimism inferred from its exceptionalism.

Judith N. Shklar, Ordinary Vices (1984).

Until fairly recently, government speech received little attention. It had been addressed three-quarters of a century ago in a well publicized report from the Commission on Freedom of the Press

Zacharia Chafee, Jr., Government and Mass Communications (1947).

In the development of the notion of government speech, by common consent

See e.g., Legal Services Corp. v. Velazquez, 531 U.S. 533, 541 (2001); Alyssa Graham, The Government Speech Doctrine and Its Effect on the Democratic Process, 44 Suffolk L. Rev. 703, 707 (2011); Jessica Pagano, The Elusive Meaning of Government Speech, 69 Ala. L. Rev. 997, 1002–4 (2018); Andy G. Olree, Identifying Government Speech, 50 Conn. L. Rev. 365, 374, 412 (2009); Caroline Mala Corbin, Mixed Speech: When Speech Is Both Private and Governmental, 83 N.Y.U. L. Rev. 605, 617 (2008); Alan C. Hake, The States, a Plate, and the First Amendment: The “Choose Life” Specialty License Plate as Government Speech, 85 Wash. U. L. Rev. 409, 422 (2007); David S. Ardia, Government Speech and Online Forums: First Amendment Limitations on Moderating Public Discourse on Government Websites, 2010 B.Y.U. L. Rev. 1981, 2011 Carl G. DeNigris, When Leviathan Speaks: Reining in the Government-Speech Doctrine Through a New and Restrictive Approach, 60 Am. U. L. Rev. 133, 140 (2010)..

At issue was whether Dr. Irving Rust, the medical director of a Planned Parenthood clinic receiving Department of Health and Human Services family planning funds, could be prevented by law from discussing abortion with his patients. The speech was Rust's, not the government's, but by funding him, government in effect designated him as its agent.

This was the interpretation made in subsequent cases, e.g., Velazquez, 531 U.S. 533, 541.

Rust maintained that the regulation on its face violated the free speech rights of both doctors and patients. They were allowed to discuss continuing the pregnancy till birth but not the option of abortion, even though the Supreme Court had earlier established the woman's right to make that choice as a fundamental right.

Roe v. Wade, 410 U.S. 113 (1973). The Court has since famously revisited this case in Dobbs v. Jackson Women's Health Organization, _U.S._, 142 S.Ct. 2228 (2022).

Chief Justice Rehnquist, speaking for a five to four majority, conceded that the “family planning” language of the statute was ambiguous,

Rust v. Sullivan, 500 U.S. 173, 184 (1991).

Only after these excursions into administrative law did Rehnquist address First Amendment concerns. He made explicit the importance of the canon requiring courts to seek constitutional interpretations of statutes,

Rust, 500 U.S. 173, 190.

There still remained the question as to whether the regulation conditioned the funding on relinquishing the constitutional right to speak. Rejecting this claim, Rehnquist maintained that “the government is not denying a benefit to anyone but is instead simply insisting that public funds be spent for the purposes for which they were authorized.”

Id. at 196.

As to the regulation's impact on the doctor-patient relationship, Rehnquist thought it was not significant. The program did not offer post-conception medical services, and so patients would not expect that it would discuss abortions.

Id. at 200. Actually, the regulations did not confine the clinic only to family planning, but permitted it also to address breast cancer, sexually transmitted diseases, and gynecological concerns.

In a lengthy and angry dissent, Justice Blackmun had little patience with the majority's reliance on administrative law. In a case where “Congress intends to press the limits of constitutionality,”

Rust, 500 U.S. 173, 207.

The administrative law arguments seemed to Blackmun a “disingenuous”

Id. at 205.

The implication, Blackmun thought, was that courts would “tolerate any governmental restriction upon an employee's speech so long as that restriction is limited to the funding workplace.”

Rust, 500 U.S. 173, 212. Immediately after, one observer reported that Rust “sent shock waves through the arts, scientific, humanities, research, foundation, and university communities.” James F. Fitzpatrick, The Spread of Rust, Colum. Journalism Rev. 1, 53 (Nov/Dec 1991).

As in his famous opinion a few years earlier in Roe v. Wade,

Roe, 410 U.S. 113.

Oddly absent from Blackmun's extensive discussion of the rights of women

Rust, 500 U.S. 173, 214–20.

From Blackmun's perspective, Rust posed an issue of grave constitutional importance: can the government, by asserting government speech, effectively prevent people from accessing information essential to their exercise of a fundamental right? If the effect of the regulation was to induce indigent women to waive their constitutional right to choose an abortion, did Rust violate the principle that such waivers must be knowing and voluntary?

Johnson v. Zerbst, 304 U.S. 458, 464–66 (1938).

Government, from this perspective, in Rust used its power to monopolize the marketplace, crowding out other voices and engaging in deliberate deception.

Steven Shiffrin has made a similar argument concerning California referenda. The ballot contains only the government's view of the proposition. Other views may be available elsewhere, but voters may well encounter the issue only in the voting booth. To the extent that this is the case, the government will effectively monopolize the issue. Steven Shiffrin, Government Speech, 27 U.C.L.A. L. Rev. 565, 639 (1980).

Attribution may play a significant role here. If the government's speech is presented as speech from independent experts, in this instance, doctors, it may well appear more believable. Hence, the irony that angered Blackmun: the wording and history of the First Amendment point to its protecting private speech against government action, not protecting government speech;

David Fagundes argues that the First Amendment does protect government speech. David Fagundes, State Actors as First Amendment Speakers, 100 Nw. L. Rev. 1637, 1664–78 (2006).

Not with standing these issues, Rust established the principle that government may exercise viewpoint discrimination when it utilizes private speakers to communicate government messages. These messages constituted government speech, even though the speakers were private parties. Caroline Mala Corbin has called this phenomenon First Amendment capture: “the government, which is supposed to be regulated by the First Amendment, gains control of speech.”

Caroline Mala Corbin, Government Speech and First Amendment Capture, 107 Va. L. Rev. Online 224, 226 (2021).

Rust’s progeny have produced mixed results. National Endowment for the Arts v. Finley

National Endowment for the Arts v. Finley, 524 U.S. 569 (1998).

Justice O’Connor, speaking for the Court, thought that as the NEA was forced to make esthetic judgments, the choice was “inherently content-based”

Finley, 524 U.S. 569, 573.

In a scathing concurrence,

Id. at 593, 598. This is not the only instance of Scalia attacking O’Connor with thinly veiled contempt. See, e.g., Webster v. Reproductive Health Services, 492 U.S. 490, 532–34 (1989); Hamdi v. Rumsfeld, 542 U.S. 507, 573–77 (2004).

Justice Souter, dissenting, drew a distinction between the government as speaker and the government as patron, where it “expends funds to encourage a diversity of views from private speakers.”

Id. at 610–11, 613.

Departing from Rust was Legal Services v. Velazquez,

Legal Services Corp. v. Velazquez, 531 U.S. 533 (2001).

Perhaps favoring Legal Service's lawyers over Rust's doctors reflects the Justices’ natural sensitivity to circumstances affecting the practice of their own profession. Lawyers might be unable to provide a zealous defense for their clients if they were prevented from pursuing all potential arguments.

Id. at 546.

Justice Scalia, dissenting, labeled the law subsidized speech, which, like government speech which it closely resembled, granted government freedom from First Amendment restraints.

Velazquez, 531 U.S. 533, 554.

Pleasant Grove City v. Summum

Pleasant Grove City v. Summum, 555 U.S. 460 (2009).

The ruling was unanimous, though it also produced no fewer than three concurring opinions. Justice Alito, speaking for a majority, thought the case pivoted on whether government had a history of using park monuments as a means of communication and whether government had direct control over the message conveyed.

Id. at 462.

As the park is limited in space, the city cannot approve every proposed monument; there simply would not be enough room to accommodate them all. But imagine that the park exists only in cyberspace on the city's website, where space scarcity issues would disappear. The city might still retain discretionary approval powers—it would presumably refuse monuments to Ted Bundy or the ebola virus—but it would necessarily have only a qualitative rationale. Does government even need a quantitative rationale?

Justice Breyer, concurring, asked “whether a government action burdens speech disproportionately in light of the action's tendency to further a legitimate government objective.”

Id. at 484.

Justice Souter, also concurring, urged that the determinative question was “whether a reasonable and fully informed observer would understand the expression to be government speech.”

Id. at 487.

Walker v. Texas Division, Sons of Confederate Veterans

Walker v. Texas Division, Sons of Confederate Veterans, 576 U.S. 200 (2015).

In Walker, the Sons requested a specialty plate featuring the Confederate battle flag, and the Texas Department of Motor Vehicles Board twice turned them down, explaining that it had invited public comments and that “public comments had shown that many members of the public find the design offensive, and because such comments are reasonable.”

Walker, 576 U.S. 200, 206. The Sons’ plate was far less extreme than a Virginia plate, ZYKLON B, that referenced the poison gas used to murder Jews in the Holocaust. Leef Smith, Va. Man's License Tags Recall Holocaust Horror, Wash. Post, May 13, 1997.

The Court speaking through Justice Breyer for a five to four majority determined the plates to be government speech; plates “long have communicated messages from the states;”

Walker, 576 U.S. 200, 211.

Justice Alito, dissenting, did not quarrel with the majority's version of government speech, but denied that it applied in this case. The message came from the Sons, he insisted, and denying it constituted viewpoint discrimination. He disagreed that the message on the plates would likely be attributed to the state. When Texas issues a “Rather Be Golfing” plate, it is not indicating that golfing is an official state policy or that Texas prefers golf to tennis.

Walker, 576 U.S. 200, 222.

Which raises the question as to whether Souter's reasonable observer would view the license plate as government or private speech? If the driver attached the Confederate flag to the trunk of his car, a reasonable observer would understand the message as private speech. But if the message were on a license plate, would this imply government approval and thus constitute government speech? After all, the very existence of specialty plates was due entirely to a series of government decisions.

But Alito's driver might insist that the message was entirely his, and that the whole reason government created specialty plates was to facilitate the expression of private messages. Accordingly, the driver might well reject claims for neutrality with the jibe: If the plate bothers you, look somewhere else.

However, if the speech were considered entirely private, would the speaker necessarily avoid the neutrality requirement? Suppose the driver displaying the Confederate flag plate was in charge of overseeing a diversity program for a government agency, and his supervisor decided on that account to punish him. A court would ask whether the Confederate flag plate was actually likely to interfere with government operations, and it would insist that the driver was not being singled out in retaliation for the content of the message. If his supervisor concluded that the plate would lead his colleagues and clients to question his commitment, impeding his performance and justifying some disciplinary action, would the court agree?

Cf. Locurto v. Guiliani, 447 F.3d 159, 179 (2d Cir. 2006).

If the speech is a mixture of private and governmental, which should prevail? If the mixture is treated as private speech, the government's role will be obscured; if treated as government speech, the private speaker's role will be obscured. In either case, the audience will face barriers in apprehending the source of the message.

Cf. Corbin, supra note 31, at 605, 650, 654–55, 663–65. She would replace the binary categorization with a third mixed option to be determined by a five-pronged test. Government would be barred from viewpoint discrimination unless it passes an intermediate scrutiny test. Intellectually defensible, the proposal may be too cumbersome and complex to be workable, particularly, since the difficult five-pronged test would vastly expand the mixed category.

Which raises the matter of the message itself. While Alito noted that a given plate might well contain both government and private messages, the Court did not consider the possibility that different specialty plates might have different statuses. Perhaps, some plates, like New York's Yankee logo, would qualify as mostly private speech, on the theory that the state does not prefer baseball teams; others, like Louisiana's Choose Life featuring the state bird holding a baby, might qualify as mostly government speech, on the theory that it appears to be government endorsement of opposition to prevailing abortion policy.

Shurtleff v. Boston (2022) was a government speech case with church-state elements. Boston's city hall features three flag poles, one for the United States, a second for Massachusetts, and a third that displays either the city's flag or, temporarily, the flags of foreign countries or private organizations. Over the preceding dozen years, none of the 284 requests concerning the third flag pole had been turned down. Harold Shurtleff, head of a religious organization called Camp Constitution, asked to display a religious flag for only one hour, but Boston denied the request, saying that this might be seen as “an endorsement by the city of a particular religion” in violation of the First Amendment's establishment clause. Boston also feared that if it were forced to fly Shurtleff's flag, it might lose all control and later perhaps be compelled to fly flags of Nazis or terrorists. Boston also claimed that “all (or at least most) of the 50 unique flags it approved reflect particular city-approved values or views,” like gay pride. The First Circuit Court of Appeals held for Boston, seeing the flag displays as a form of government speech.

The Supreme Court ruled unanimously in favor of Shurtleff, though the result was splintered with three concurring opinions. Speaking for the Court, Justice Breyer in one of his last opinions before retiring,

The opinion had a characteristically impish quality, with asides on the ugly architecture of city hall and the Boston Red Sox.

In Walker, Breyer had argued that Texas’ final approval of specialty plates made them government speech; in Shurtleff, he chose to emphasize the salience of Camp Constitution's input. Since it was not government speech, it did not amount to government endorsement of a religion.

In concurring opinions, Justices Gorsuch, Alito, and Kavanaugh expressed their displeasure with a major establishment precedent that requires that a “principal or primary effect” of a government policy can be one that neither “advances nor inhibits religion.” Lemon v. Kurtzman, 403 U.S. 602, 612 (1971) (Burger, C.J.).

We could no more imagine government without government speech than breathing without air. At the same time, government speech is also inextricably tied to government power, and for this reason lends itself to a wide range of abuses, perhaps nowhere as clearly as when that speech takes the form of propaganda.

The issue is not peculiarly American. For example, the Conservative British government mailed anti-Brexit leaflets to every home, provoking an outcry from the opposition. Taxpayer to Fund Anti-Brexit leaflets, Daily Telegraph, Apr. 7, 2016.

Historically, “propaganda” derives from the Latin propagare, which means “to spread,” and was first used by the Roman Catholic Church in the seventeenth century to refer to spreading or propagating the Gospel. The term has long since lost its theological connotation, and instead has taken on a decidedly unsavory character, most often described as psychological manipulation that “intentionally undermin[es] reasoned analysis.”

E.g., Jacques Ellul, Propaganda: The Formation of Man's Attitudes 61 (Konrad Kellen & Jean Lerner trans. 1965); Caroline Mala Corbin, The Unconstitutionality of Government Propaganda, 81 Ohio St. L. J. 815, at 818, 825.

Yet a moment's reflection reveals the inadequacy of this definition. Must propaganda be said by government? What if government speaks through private actors, as in Rust? Is propaganda only about statements of fact? Certainly, governments lie; hence, the ubiquitous “credibility gap” charge.

William M. Hammond, Public Affairs: The Military and the Media, 1962–1968 (1988).

The importance of government propaganda today is vastly greater than it was in 1791, when the First Amendment was adopted, mainly because the role of government in society has increased enormously and now touches nearly every aspect of life. The expanded role of government has naturally vastly expanded its role as a marketer of ideas and policies.

Steven Smith suggests that the controversy over government speech reflects “the collapse of any working consensus about the proper domain and functions of government.” Smith, supra note 9, at 946.

This involves not merely identifying an issue, but shaping its contours, for how questions are framed will heavily influence the answers proposed. Policymakers, for instance, may agree that energy is an important issue that government should address, but some may speak of it in terms of jobs and prosperity and others in terms of climate change and impending catastrophe; each side will try to define the various positions to its advantage. If government chooses to address energy (an issue it can hardly avoid), it will have to decide how to frame the issue (or, perhaps, to adopt multiple framings). Thus, as a practical matter, government will feel compelled to act as a marketer of ideas and policies, and this government speech will invariably be seen by its opponents as propaganda.

Hanna Pitkin tries to distinguish manipulation, which seeks to undermine public opinion, from leadership, which “succeeds only so long as [the people] are willing to follow,” but it is hard to see how this could be applied to the real world. If the leader successfully manipulates the people, they will willingly follow him. Hanna Fenichel Pitkin, The Concept of Representation 233 (1967).

Our natural reaction as consumers of government speech is to demand that government tell us the truth. To be lied to, we feel, is to be cheated and insulted. Machiavelli, however, cautions us that “Any man who tries to be good all the time is bound to come to ruin among those who are not good. Hence, a prince who wants to keep his authority must learn how not to be good.”

Niccolo Machiavelli, The Prince 42 (Robert M. Abrams trans. & ed. 2d ed.) (1992/1532). Max Weber distinguished between the ethics of responsibility (which stresses the consequences of choices) and the ethics of ultimate ends (which stresses the purity of intentions). From Max Weber 120 (trans. & ed. Hans H. Gerth & C. Wright Mills 1946/1923).

Nearly 2000 years before, Aeschylus observed that in war, truth is the first casualty. Sometimes, government will be convinced that an urgent response to some challenge will be required before public opinion can form;

Thomas Halper, Foreign Policy Crises: Appearance and Reality in Decision-Making (1971).

We don’t like manufacturing consent and bitterly resent the idea that government is entitled to manipulate us, via the government speech doctrine. Yet if it produces results we like, the manipulation may not bother us greatly. We object to the means. But if we approve of the ends, would our objections melt away? Would we rule out government sponsored anti-smoking television commercials that utilize dying, disfigured smokers—even if this manipulative emotional approach is more efficacious than a presentation of graphs and tables? In the end, is the distinction between, say, education and propaganda simply a matter of whether one supports or opposes the speech? “So long as government neither monopolizes, coerces, or ventriloquizes,” wrote Abner Greene, “its voice will be one of many, it will be one of persuasion not coercion, and the speech will clearly be in the government's voice.”

Abner S. Greene, Government of the Good, 53 Vand. L. Rev. 1, 25 (2000).

Even if the goal is benign, the principle that government can intimidate, cajole, or surreptitiously pay off private parties to influence the public can only be disturbing. Good intentions—and they emphatically are not always good—do not rinse the paternalistic, antidemocratic stain from the practice. The government speech doctrine enables this practice, yet leaves us with a puzzle: government speech's potential power derives from its exemption from the neutrality principle; however, any court-imposed rule on government propaganda speech would necessarily be viewpoint based, and thus violate the same principle. How to formulate a rule distinguishing the acceptable from the unacceptable?

Which raises the matter of transparency, for “the identity of the speaker is an important component of many attempts to persuade.”

Ladue v. Gilleo, 512 U.S. 43, 56 (1994).

But suppose the role of the government in the messaging is hidden, so that the level of deception is taken to another level?

Abner Greene has called this practice, when government surreptitiously uses others to speak for it, ventriloquism. Greene, supra note 136, at 49–52.

Some situations, however, are not so clear cut. During the Cold War, the Defense Department's Motion Pictures Production Branch reviewed movie and television scripts to determine if they would be “in the interest of the Department of Defense or otherwise in the national interest,”

Department of Defense Instruction 540.15.

Nor is this practice of co-opting private speech confined only to wartime. In domestic policy, to take another widely reported example, in the interest of fighting drug abuse, the government reviewed scripts of over 100 episodes of several popular television programs to ensure that they carried appropriate messages; in return, networks were freed from broadcasting a portion of their required public service announcements.

Don Van Natta, Jr., Drug Office Will End Scrutiny of TV Scripts, N.Y. Times, Jan. 20, 2000; Howard Kurtz & Sharon Waxman, White House Cut Anti-Drug Deal with TV; Ad Credits Given for “Proper Message,” Wash. Post, Jan. 14, 2000. Similarly, in Britain, the popular television series, Coronation Street, included stories to further the government's National Year of Reading. Bob Franklin, Packaging Politics 90 (2d. ed. 2004).

Courts have repeatedly held that, however valuable transparency may be to democracy, it is not a constitutional requirement.

McBurney v. Young, 569 U.S. 221, 233–34 (2013); Houchins v. KQED, Inc., 438 U.S. 1, 14 (1978). In NAACP v. Alabama, the Supreme Court rejected the state's demand for the organization's membership list as a potential intrusion into the members’ freedom of speech and assembly. Past disclosures of NAACP membership had made possible “economic reprisal, loss of employment, threat of physical coercion, and other manifestations of public hostility.” NAACP v. Alabama, 357 U.S. 449, 462 (1958). But where the NAACP membership was vulnerable, government as the most powerful actor on the scene is largely immune from the organization's travails.

There is an ad hominem quality to this distaste with deception, as it targets the source of the message and not the message itself. Yet we feel that knowing the speaker may help us understand his motives and intentions, and thus aid us in evaluating what he says. When government speaks or pays for someone else to speak, we naturally expect transparent disclosure. Why, it must be asked, would government insist on anonymity?

It is as obvious as water in a thunderstorm that the typical reason for government anonymity is to evade accountability. Transparency does not guarantee rigorous or even minimal accountability, but its absence greatly complicates achieving it. Furthermore, a lack of transparency not only is said to impede accountability. Also, the inevitable revelations gnaw away at the public's trust in government, generally, and the administration in power, specifically. It confirms the cynic's worst fears and leaves its supporters looking gullible, complacent, or stupid. Conspiratorial thinking is far from a recent development,

Richard Hofstadter, The Paranoid Style in American Politics, Harper’s Magazine 77 (Nov. 1964).

The government may be “entitled to say what it wishes,”

Rothenberger v. Univ. of Va., 515 U.S. 819, 833 (1995).

How, then, to counter potential abuse? The answer, the Court advances, is “accountability to the electorate and the political process for its advocacy.”

Bd. of Regents of the Univ. of Wis. System v. Southworth, 529 U.S. 217, 235 (2000).

At the outset, a logical problem emerges like a punch in the face. If accountability depends on information made available to the public and government speech effectively controls that information, how can the public hold government accountable? When government is most abusive—that is, most effective at controlling information— t is also most impervious to accountability.

More broadly, much of the rhetoric on accountability is built on fanciful presumptions.

The conventional view thus resembles the idealized mode of opinion change advanced by certain fashionable theorists of deliberative democracy, in which open minded people at the periphery of civil society seeking the common good come together in pursuit of rational solutions to social problems, in the process gradually influencing the administrative state to adopt their ideas. See, e.g., Seyla Benhabib, Toward a Deliberative Model of Democratic Legitimacy, in Democracy and Difference 67 (Benhabib ed. 1996); Joshua Cohen, Deliberation and Democratic Legitimacy, in Deliberative Democracy 67 (James Bohman & William Rehg eds. 1997); Jurgen Habermas, Between Facts and Norms ch. 8 (William Rehg trans. 1996).

The influence of government is also undercut by the two-step flow phenomenon.

Elihu Katz & Paul F. Lazarsfeld, Personal Influence: The Part Played by People in the Flow of Mass Communications (1955).

Accountability is also weakened and made much less coherent by the whole package principle. When we vote on a candidate, we do not have the option of approving his position on one issue and disapproving his position on another. We must accept or reject the whole package. I may approve his position on the abortion gag rule, the NEA grant criteria, the Legal Services prohibition, the park monuments policy, the license plate standards, and the flag pole decision, yet vote against the official for other reasons that I find more salient. Retrospective voting, in any event, is likely tainted by ignorance.

Christopher H. Achen & Larry Bartels, Democracy for Realists: Why Elections Do Not Produce Responsive Government (2016).

The Court's assumption that the public's appetite for information is dictated by truth seeking is also hard to sustain. People look for information sources that share their beliefs,

Sylvia Knobloch-Westerwick & Jingbo Meng, Looking the Other Way: Selective Exposure to Attitude-Consistent and Counterattitudinal Political Information, 36 Communication Res. 426 (2009).

Likely, a plethora of factors—childhood socialization, socioeconomic status, partisan affiliation, and many more—predetermine our political positions. Most of us do not gather evidence and then make up our minds; we make up our minds and then gather evidence. Thus, when President Trump said that he “could stand in the middle of Fifth Avenue and shoot somebody and I wouldn’t lose any voters,”

Qtd. in Colin Dwyer, Donald Trump: “I Could Shoot Somebody, and I Wouldn’t Lose Any Voters,” NPR, Jan. 23, 2016. This phenomenon is by no means confined to Trump.

The major misinformation problem, in a word, remains demand, not supply. We demand information congenial to our pre-existing beliefs, and ignore the vast supply of uncongenial information. That accurate information is clearly more readily available today than ever before has not banished the problem, which like a drunk at a wedding, simply will not go away. Nor is it a recent phenomenon. There is no shortage of blatant and consequential examples of misinformation in years past, for example, the white South's antebellum belief in the happy slave and many Germans’ belief that World War I was lost on account of a “stab in the back.” Even becoming well informed does not guarantee sound opinions; research on the 2016 presidential election revealed that “visits to fake news websites are highest among people who consume the most hard news.”

Andrew M. Guess et al., Exposure to Untrustworthy Websites in the 2016 U.S. Election, 4 Nature Human Behav. 472 (2020).

This is not to argue that accountability is a mirage. On high profile, large issues—war/peace, inflation/depression—it may engage large numbers of the public, and on small targeted issues—threats to a particular industry, taxes directed at a particular group –it may engage small but perhaps decisive numbers of the public. Government speech may here take on central importance. It is widely believed, for example, that government speech during the covid pandemic in favor of closing schools and having children wear masks was sufficiently potent to affect election results.

E.g., Michael Hartney & Renu Mukherjee, Closures and Consequences, City Journal, Dec. 8, 2021.

Another practical problem bedeviling accountability is the principle of separation of powers. When the executive and legislative branches engage in government speech, they invariably regard judicial interference as intrusions into areas the Constitution has designated as their province. Courts, as the least dangerous branch wielding counter-majoritarian judicial review, may be loathe to enforce constitutional accountability.

Some government speech would appear controversial only to the lunatic fringe, for example, informational speech in the form of a School Crossing sign intended to safeguard children or details on when and where a civil service examination would be given. But even informational government speech may be controversial. Consider the Miranda warning given to criminal suspects, bathroom instructions for transgender persons, or pleas to get vaccinated during a pandemic. An ostensibly informational message may also have a subtext that presents policies or officials in a favorable or unfavorable light,

Justin Grimmer et al., The Impression of Influence 28–29 (2015).

How, then, do we regard government speech? Jud Campbell maintains that the Court did not embrace neutrality until the late 1960s and early 1970s, when it displaced the focus on toleration.

The Emergence of Neutrality, 131 Yale L. J. 861, 865 (2022).

Academics, it must be conceded, have not been very helpful. Elena Kagan, writing before she joined the Supreme Court, focused on the element of controversy, presumably having in mind a practical solution. In considering whether a government's no-smoking campaign would be bound by the neutrality requirement, she offered three criteria: does the topic offer “the hope of right and wrong answers ... subject to verification and proof;” has “society ... reached a shared consensus on the issue;” and does “one side of the debate ... do great harm.”

Elena Kagan, The Changing Faces of First Amendment Neutrality: R.A.V. v. St. Paul, Rust v. Sullivan, and the Problem of Content-Based Under-Inclusion, 1992 Sup. Ct. Rev. 29, 75.

Thomas Emerson, in his magisterial The System of Freedom of Expression, concedes that “government has a broad right to engage in expression as part of its regular functions,” but believes that this right “does not extend to any sphere that is outside the governmental function.” A problem with this formula, he admits, is that “the governmental function certainly covers an extensive area.”

Emerson, supra note 2, at 699.

Helen Norton, having devoted an entire book to government speech, devised some useful criteria.

Norton, supra note 1, at 6–10, 68–211.

What, then, exactly are the contours of government speech? How do we know when it is government that is speaking, given the myriad problems of attribution, which, in any case, is “easily manipulated”?

Richard C. Schragger, Of Crosses and Confederate Monuments: A Theory of Unconstitutional; Government Speech, 63 Ariz. L. Rev. 1, 45 (2021).

The Court's answers, sad to say, have not been very helpful, hence their routine description as “unprincipled,”

Donald W. Park, Government Speech and the Public Forum: A Clash between Democratic and Egalitarian Values, 45 Gonz. L. Rev. 113, 1134 (2010).

Surveying the Court's efforts is not very uplifting. Rehnquist in Rust, downplayed First Amendment concerns, presenting the case mostly as a matter of administrative law. The impact of governmental coercion was dismissed with the hollow observation that if doctors found the regulations unacceptable, they could seek employment elsewhere. Finley also saw no real effort to devise a workable jurisprudence, O’Connor agreeing that if artists find the regulations unacceptable, they can also seek funding elsewhere. These rationales not only place heavy burdens on those affected. They also amount to granting “government nearly carte blanche ability to exclude speakers and speech on the basis of viewpoint so long as the government can show that it ‘effectively controlled’ the message being conveyed.”

Ardia, supra note 31, at 1983–84.

Perhaps, the most ambitious Justice was Souter, who did not speak for the Court in any of these principal decisions. In Finley, he differentiated between government as patron and government as speaker. In Rust and Velazquez, government, like a Renaissance aristocrat, acted as patron, indirectly affecting speech in the public marketplace by granting or denying subsidies to the speakers. Government itself was not speaking; private health care providers and public defenders were. In these situations, Souter found that the speech was really not government speech, and so it should not be exempt from the standard obligation of neutrality. On the other hand, if government were speaking, maybe in a pamphlet or online, this would constitute government speech and avoid the neutrality issue altogether. The problem is, however, that the test does not address the central issue, government speech manipulating public opinion.

The other distinction Souter drew was in Summum, when he asked how a reasonable, fully informed observer would understand the speech: would it seem to him government speech or private speech? This approach has the advantage of taking into account the audience for the speech, even if in an idealized fashion. Of course, it poses the question: how to describe such a person? He or she is not the average person drawn from empirical or statistical data nor a person markedly superior in morality, temperament and perception. Rather a reasonable person is a kind of hybrid of the two, that is, common enough to apply generally, but normatively prone to do the correct or prudent thing, so he or she acts as an ordinary person should act. Presumably, how such a person would behave would gradually be mapped out as courts would fill in the details as they ruled in particular cases. The result would be a set of more or less objective criteria that could be applied without inquiring into “the infinite variations of temperament, intellect, and education which make the internal character of a given act so different to different men.”

Oliver Wendell Holmes, Jr., The Common Law 108 (1881).

Of course, applying the observer construct would not always be easy. Would a reasonable observer understand the constraints imposed on Dr. Rust? The Court could not agree. Even in Summum, the answer might be in doubt. Are the specialty license plates so small and near the ground that literally there rarely would be observers? Unlike, say, billboards, plates do not thrust themselves on an indifferent or unwilling audience, but must be actively sought out. Ordinarily, viewing them closely enough to read the message requires a specific decision and action to do so. More fundamentally, do we really want to so empower the observer that he might override the First Amendment or spare government from meeting the neutrality standard? The observer, from this perspective, would not exactly wield the notorious heckler's veto,

Feiner v. New York, 340 U.S. 315 (1951).

What is so troubling about government speech is that its purpose and effect is to empower government, already the most powerful set of institutions in the nation, “in the sense that [government speech] is used to fend off other First Amendment claims by private speakers and government employees.”

Steven G. Gey, Why Should the First Amendment Protect Government Speech When the Government Has Nothing to Say? 65 Iowa L. Rev. 1259, 1261 (2010).

What this suggests is that both academics and courts may well have sailed off in the wrong direction. The real issue is not, as they imagine, how to distinguish government speech from non-government speech. Rather, it is, how to distinguish government speech that abridges free speech from that which does not. This second question may appear merely a rewording of the first. But its import lies in its focus on the consequences of the speech, not the nature of the speaker. Put differently, the choice is between government political speech, on which there are nearly always major differences of opinion, and government nonpolitical speech, on which there is typically consensus. It is, after all, the political consequences that justify our interest in the topic, not the point of origin.

The basic problem remains. If courts treat government speech as covered by the First Amendment, the practical utility of government speech disappears. But if courts deny that government speech is covered by the First Amendment, government speech may overwhelm or silence private speech and much of the practical utility of the First Amendment may disappear. Government speech may stimulate private speech, enhance tolerance, promote health, security, and prosperity. Or it may not.