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.” ⁽⁷⁾ As Justice Thurgood Marshall phrased it, “above all else, the First Amendment means that government has no power to restrict expression because of its message, its ideas, its subject matter or its content.” ⁽⁸⁾ The importance of all this is widely taken for granted.

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.” ⁽⁹⁾ Perhaps, “the size of the audience [is] wholly irrelevant to First Amendment issues,” ⁽¹⁰⁾ but government possesses a unique capacity to amplify its speech, which carries with it a unique capacity to affect political, social, and economic discourse. Government begins with significant advantages over other speakers: it has an unparalleled capacity to get its messages to the public; it generates enormous amounts of information and has access to an incomparable range of information sources, some of them classified secret; portions of the population may regard its messages as inherently believable; and it may be able to use its monopoly over the use of legitimate force to intimidate, silence, or weaken opponents. In the words of Stephen Gardbaum, “A way of life that the state endorses and promotes, even through symbolic or persuasive means, is an ‘authorized’ way of life. … . individuals may defer to the state's authority, just as we normally wish them to do in the case of general obedience to the law.” ⁽¹¹⁾ On many subjects, not least its own preservation, government is not neutral. Sometimes, as with public schools, it speaks to a captive audience. ⁽¹²⁾ In all this, it is well to remember that government cannot speak; persons speak in its name, and these persons inevitably have their own interests to protect and advance. “Government” may present an abstract appearance, but its reality will have a personal dimension

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. ⁽¹³⁾

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. ⁽¹⁴⁾

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, ⁽¹⁵⁾ the pursuit of truth, ⁽¹⁶⁾ self-realization, ⁽¹⁷⁾ and participation in self-government. ⁽¹⁸⁾

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. ⁽¹⁹⁾ The First Amendment, however, generally protects even cruel speech. The British philosopher and historian of ideas, Isiah Berlin, believed that value conflicts are “an intrinsic, irremovable element of human life,” and that there is no moral hierarchy or common measure that commands a consensus that would allow us to resolve these conflicts to the satisfaction of all. ⁽²⁰⁾ Accordingly, Berlin recommended not value relativism, which in the end leads to nihilism, but value pluralism, which enforces a kind of live-and-let-live toleration. This approach may most closely resemble the workings of the First Amendment. But in the context of government speech, how to enforce this attitude in the face of officials convinced that their speech will do good? As Charles Fried observed, “The greatest enemy of liberty has always been some version of the good.” ⁽²¹⁾

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 ⁽²²⁾ and also treated in a scattering of academic publications, ⁽²³⁾ but the Supreme Court had not considered it in any depth. In West Virginia Board of Education v. Barnette, in the midst of World War II, the Court struck down government speech in the form of a pledge of allegiance mandated for school children. ⁽²⁴⁾ In Speiser v. Randall ⁽²⁵⁾, the Supreme Court reversed a statute that made a veteran's property tax exemption conditional on taking a loyalty oath. Speaking for the Court, Justice Brennan found it “a discriminatory denial [that] necessarily will have the effect of coercing the claimants to refrain from proscribed speech.’” ⁽²⁶⁾ Thus, “to deny an exemption to claimants who engage in certain forms of speech is in effect to penalize them for such speech.” ⁽²⁷⁾ In dicta in CBS v. Democratic National Committee (1973), Justice Stewart, concurring, declared that the First Amendment “protects the press from government interference; it contains no analogous protection for the government.” ⁽²⁸⁾ In Keller v. State Bar of California, ⁽²⁹⁾ Chief Justice Rehnquist observed, “if every citizen were to have the right to insist that no one paid by public funds express a view with which he disagreed, debate over issues of great concern to the public would be limited to those in the private sector, and the process of government as we know it radically transformed.” ⁽³⁰⁾ If government speech was not limited by the Free Speech Clause, however, it remained limited by other constitutional provisions, including the Establishment Clause and the Equal Protection Clause.

In the development of the notion of government speech, by common consent ⁽³¹⁾ a key case is Rust v. Sullivan (1991). Like the earlier cases, it did not turn on government itself speaking, but on its compelling or designating others to give voice to a government message. For some years, the Department of Health and Human Services’ family planning regulations had prohibited institutions receiving federal funds from using them to perform abortions, ⁽³²⁾ but had permitted physicians to discuss abortion with their patients. This changed, when the Reagan administration published regulations in 1988 that denied physicians permission to “encourage, promote or advocate abortion as a method of family planning,” ⁽³³⁾ even if the woman specifically requested the information. The key phrase, “method of family planning,” was not defined. From the outset, the new policy was highly controversial. ⁽³⁴⁾

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. ⁽³⁵⁾ The term “government speech” did not appear in the Court's opinion. ⁽³⁶⁾ The law also required that if the clinic performed abortions, those facilities must be “physically and financially separate” from other facilities. ⁽³⁷⁾

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. ⁽³⁸⁾ A federal district court and the second circuit both found for the government, and the Supreme Court granted certiorari to hear Rust's appeal.

Chief Justice Rehnquist, speaking for a five to four majority, conceded that the “family planning” language of the statute was ambiguous, ⁽³⁹⁾ but following Chevron deference, ⁽⁴⁰⁾ thought the interpretation of the department was plausible and did not conflict with Congress’ expressed intent. The legislative history was also so unclear as not to be determinative. The law provided that “None of the funds appropriated ... shall be used where abortion is a method of family planning,” ⁽⁴¹⁾ and Rehnquist thought the department was within its authority to ban counseling, referral, and advocacy under this heading. Nor did separating the abortion facilities from the remainder of the program compromise the program's integrity, for the law was so ambiguous that it was not clear that Congress intended that the health care system be integrated: in any event, the separation did “not represent a deviation from past policy.” ⁽⁴²⁾ The department's new regulation helped the public avoid the misimpression that federal funds were being used improperly for abortion activities. Bad patient experiences prior to the changed policy plus a public opinion that now was “against the elimination of unborn children by abortion” ⁽⁴³⁾ also supported the ruling.

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, ⁽⁴⁴⁾ itself a corollary of the counter-majoritarian nature of judicial review. Following this, his principal point was that there “is a basic difference between direct state interference with a protected activity and state encouragement of an alternative activity consonant with legislative policy.” ⁽⁴⁵⁾ Government need not be neutral in encouraging activities, but may choose some and reject others. Here, “the government ... has merely chosen to fund one activity to the exclusion of the other.” ⁽⁴⁶⁾ In rejecting the funding of abortion counseling, it was under no obligation to subsidize the opposing view. To hold otherwise would be to embrace the absurd conclusion that when Congress created a National Endowment for Democracy, it also was “constitutionally required to fund a program [encouraging] communism and fascism.” ⁽⁴⁷⁾ Doctors were free to counsel abortions, as were medical agencies, but they simply were not entitled to have the government pay for their efforts. That women had a constitutional right to choose to have an abortion did not imply an affirmative duty on the part of government to fund that choice by paying for counseling. Thus, their free speech rights were unimpaired. Mandating government viewpoint neutrality would “render numerous government programs constitutionally suspect.” ⁽⁴⁸⁾

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.” ⁽⁴⁹⁾ Unconstitutional conditions, he added, would “involve situations in which the government has placed a condition on the recipient of the subsidy rather than on a particular program or service, thus effectively prohibiting the recipient from engaging in the protected conduct outside the scope of the federally funded program.” ⁽⁵⁰⁾ Here, however, “The employees’ freedom of expression is limited during the time that they actually work for the project; but this limitation is a consequence of their decision to accept employment in a project.” ⁽⁵¹⁾ The government is free to advance its own message that abortion is an inappropriate means of family planning, and it does so by blocking contrary messages. By agreeing to accept federal funds, Rust implicitly agreed to follow this rule. Sometimes, government speech is intended to protect a manipulable audience imperiled by powerful speakers, for example, by mandating warning statements on cigarette packs and providing informational material with prescription drugs. Rehnquist, perhaps dismissing this practice as the acts of a nanny state, did not apply the principle to Rust.

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. ⁽⁵²⁾ If indigent women suffered as a consequence of the policy, the fault lay with their indigence, not the policy. ⁽⁵³⁾

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,” ⁽⁵⁴⁾ he thought Chevron deference would not apply; instead, Congress must “express ... intent in explicit and unambiguous terms.” ⁽⁵⁵⁾ In the absence of this clear intent, as in this case, he thought it plain that the views of a politically accountable Congress should prevail over those of an unaccountable agency. He thus turned Rehnquist's point on the antidemocratic character of judicial review on its head.

The administrative law arguments seemed to Blackmun a “disingenuous” ⁽⁵⁶⁾ distraction from the central First Amendment issue: “the extent to which the government may attach an otherwise unconstitutional condition to the receipt of a public benefit.” ⁽⁵⁷⁾ For “the first time,” he wrote, the Court sanctioned “viewpoint-based suppression of speech simply because the suppression was a condition upon the acceptance of public funds.” ⁽⁵⁸⁾ The regulations compel the provider “to facilitate access to parental care and social services, including adoption services ... while making it abundantly clear that the project is not permitted to promote abortion by facilitating access to abortion through the referral process.” ⁽⁵⁹⁾

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.” ⁽⁶⁰⁾ This clearly amounted to a government effort to suppress what it took to be dangerous speech. But as the statute did not mention discussing abortion, the entire First Amendment issue could have been avoided simply by pointing this out. ⁽⁶¹⁾ In support of this, he cited a canon different from Rehnquist's, one that would have the Court “avoid passing unnecessarily upon important constitutional questions.” ⁽⁶²⁾ At the very least, the Court should have insisted that the interest of government in suppression be forced to compete with the interests of the doctor to speak and the patient to listen. The law did not even require doctors to inform patients that their inability to discuss abortion was the government's decision, and one they might not have agreed with.

As in his famous opinion a few years earlier in Roe v. Wade, ⁽⁶³⁾ Blackmun's chief concern seems to have been the regulation's impact on the doctor-patient relationship. ⁽⁶⁴⁾ Suppressing talk of abortions denies the parties essential information they would need in making their decisions, weakens the patient's confidence and trust in her doctor, and impairs the doctor's ability to meet his professional ethical responsibilities. Patients, not understanding that the doctors’ advice is limited by law and not by medicine, may be misled into ignoring the abortion alternative, thus suffering a “constitutional injury.” ⁽⁶⁵⁾ The government was using doctors, a highly credible group assumed to be quite apolitical, to sell its political message—and not informing women of the subterfuge. ⁽⁶⁶⁾ Thus did the government contrive to have its message misattributed to a more trusted and valued source. For Blackmun, who as a youth seriously considered a medical career and later counted his years as general counsel for the Mayo Clinic as “the happiest of his professional life,” ⁽⁶⁷⁾ all this may have grown out of personal experience.

Oddly absent from Blackmun's extensive discussion of the rights of women ⁽⁶⁸⁾ was a realistic challenge to Rehnquist's central point concerning consent: employees and clients who take federal funds have implicitly consented to the applicable conditions. It may well be true that a doctor, who finds the ban on discussing abortion sufficiently onerous, might withdraw his consent and evade the ban by joining a private practice and offering abortion counseling there. But the indigent women who use the agencies’ facilities probably have no such option, as they could not afford to see private physicians. For them, Rehnquist's advice is akin to Anatole France's famous observation: “The law, in its majestic equality, forbids rich and poor alike to sleep under bridges, to beg in the streets, and to steal bread.” ⁽⁶⁹⁾

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? ⁽⁷⁰⁾ Put more prosaically, can government use the government speech doctrine to force private speakers to back government policies? Government here is not joining a debate on the medical or ethical aspects of abortion, which some women might find illuminating, but by suppressing one side of the discussion, actually discouraging discussion. Government speech, in this sense, is here antithetical to private speech. As Randall Bezanson put it, “the government speech doctrine is mis-named. It is not just immunity for the government's act of speaking, but also for the government's exclusion of unwelcomed speech ... . government possesses the constitutional power of the censor.” ⁽⁷¹⁾

Government, from this perspective, in Rust used its power to monopolize the marketplace, crowding out other voices and engaging in deliberate deception. ⁽⁷²⁾ Three decades later, monopolization may seem an unfounded fear in a society brimming with speakers taking every imaginable position on every imaginable topic, their energy amplified by the marvels of contemporary communications technology. But in Rust, the question of monopoly was joined: doctors were barred from discussing abortion, while government insisted that there was no monopoly as abortion information was available outside the venue of the agencies receiving federal funds. Dr. Rust insisted that the clinics enjoyed a de facto monopolization. His point was that it was the monopolization that made government deception so potent. If there are no easily and cheaply available alternative sources of information, the audience will find it difficult to critically evaluate government speech.

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; ⁽⁷³⁾ in Rust, it was exactly the reverse.

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.” ⁽⁷⁴⁾ However, as Martin Redish noted, “In a democratic society, a government may seek to influence the choices of the populace not by means of selective suppression, but rather by making its own contributions to the debate.” ⁽⁷⁵⁾

Rust’s progeny have produced mixed results. National Endowment for the Arts v. Finley ⁽⁷⁶⁾ concerned a federal statute that required the NEA to take “into consideration general standards of decency and respect for the diverse beliefs and values of the American public” in its decisions to fund artists. ⁽⁷⁷⁾ On these grounds, Karen Finley, a performance artist, was denied funding. She argued that the policy discriminated on the basis of viewpoint and was excessively vague.

Justice O’Connor, speaking for the Court, thought that as the NEA was forced to make esthetic judgments, the choice was “inherently content-based” ⁽⁷⁸⁾ and “absolute neutrality [was] simply inconceivable.” ⁽⁷⁹⁾ Quality judgments unavoidably entail a good measure of subjectivity. These esthetic judgments that were not challenged on other grounds, moreover, were inherently at least as vague as the decency requirement; if we accept one, we cannot reject the other. Out of about 100,000 awards, she noted, “only a handful ... have generated formal complaints.” ⁽⁸⁰⁾As to the “decency and respect” criteria, they helped to define artistic excellence and were not applied separately and on their own. ⁽⁸¹⁾ Nor was Finley silenced, for she remained able to produce her art and seek funding elsewhere. ⁽⁸²⁾ In preserving the nation's artistic heritage, Congress may “selectively fund a program to encourage certain activities it believes to be in the public interest, without at the same time funding an alternative program which seeks to deal with the problem in another way.” ⁽⁸³⁾

In a scathing concurrence, ⁽⁸⁴⁾ Justice Scalia declared that the statute “unquestionably constitutes viewpoint discrimination [but that] it makes not a bit of difference” ⁽⁸⁵⁾ because the message is government speech, and thus free from First Amendment neutrality concerns.

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.” ⁽⁸⁶⁾ The NEA falls within the second category and is bound by the First Amendment.

Departing from Rust was Legal Services v. Velazquez, ⁽⁸⁷⁾ which involved a federal statute that prevented lawyers paid by federal funds to represent welfare benefits claimants from challenging existing welfare benefits laws. ⁽⁸⁸⁾ Like Rust, it involved government limiting the advice funded professionals were permitted to dispense to their clients. The Court through Justice Kennedy thought the lawyers’ speech was not government speech because they were the government's adversaries. ⁽⁸⁹⁾ Where Rust had been an agent of government disseminating its message, Legal Services was a private speaker, facilitated by government, but with its own message. As “constitutionally protected expression,” their speech was covered by the ban “against viewpoint discrimination.” ⁽⁹⁰⁾

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. ⁽⁹¹⁾ However, both Velazquez and Rust saw professionals arguing against statutes that, by limiting their speech, also limited their ability to serve their clients and meet their professional ethical obligations. ⁽⁹²⁾ When government subsidizes private speech, the result (as in Rust) is government speech, except (as in Velazquez) when it is not.

Justice Scalia, dissenting, labeled the law subsidized speech, which, like government speech which it closely resembled, granted government freedom from First Amendment restraints. ⁽⁹³⁾ It seemed to him “embarrassingly simple” ⁽⁹⁴⁾ that Velazquez was covered by Rust.

Pleasant Grove City v. Summum ⁽⁹⁵⁾ concerned a Christian organization, Summum, which sought to erect a monument containing the chief tenets of its beliefs in a city park that already contained a monument featuring the Ten Commandments donated by a private organization decades earlier. The city turned Summum down, whereupon it brought suit, alleging viewpoint discrimination.

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. ⁽⁹⁶⁾ He concluded that the facts indicated that these criteria were met; a “monument is, by definition, a structure that is designed as a means of expression,” and the city granted approval, owned the monuments, and arranged for their upkeep. ⁽⁹⁷⁾ Thus, the city's denial constituted government speech and need not be viewpoint neutral. However, Alito held that formally determining the message would be a “pointless exercise that the Constitution does not mandate,” ⁽⁹⁸⁾ perhaps because if it were promoting religion, it might have activated church-state constitutional prohibitions. The refusal of the monument was an instance of government speech, though strangely the Court was unwilling to say what was said.

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.” ⁽⁹⁹⁾ He thought it did not.

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.” ⁽¹⁰⁰⁾ He evidently believed that this artificial construct might provide a workable solution to this problem.

Walker v. Texas Division, Sons of Confederate Veterans ⁽¹⁰¹⁾ asked whether specialty license plates could be considered government speech. Nearly forty years earlier, the Court had denied New Hampshire the power to require car owners to display license plates with the state's motto, Live free or die. The state had sought to “communicate to others an official view as to proper appreciation of history, state pride, and individualism,” but the Court concluded that here “such interest cannot outweigh an individual's First Amendment right to avoid becoming the courier for such message.” ⁽¹⁰²⁾ Though the decision preceded Rust and the acceptance of government speech, the Court used similar terms in declaring that the state could not force a driver to “use [his] private property as a ‘mobile billboard’ for the state's ideological message.” ⁽¹⁰³⁾

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.” ⁽¹⁰⁴⁾

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;” ⁽¹⁰⁵⁾ the messages on the plates “are often closely identified in the public mind with” the state; ⁽¹⁰⁶⁾ the board had direct control of the licensing system; and the state maintains “direct control over the messages conveyed on the specialty plates.” ⁽¹⁰⁷⁾ It made no difference whether the idea for a specialty plate came from a private source. The final approval decision came from government. The law allowed the board to deny requests that were vulgar or obscene, and it could deny offensive plates, too. In this, Breyer echoed the government's attorney at oral argument in Summum, when he had declared that government speech “turns on control, right? So once the government takes control of something, then it's the government speaking.” ⁽¹⁰⁸⁾ This speech is not limited by the First Amendment. Thus, once government claims to be speaking, private speakers cannot shut it up or even have a right to voice their disagreement in the same venue. In these cases, government speech directly abridges private speech.

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. ⁽¹⁰⁹⁾ He thought the plates contained some government speech (the name of the state and the identifying letters and numbers) and some private speech (the specialty message), and this differentiates it from Summum.

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? ⁽¹¹⁰⁾

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. ⁽¹¹¹⁾ The Supreme Court has taken the position that mixed is not an option, in other words, that a choice must be made, either favoring government ⁽¹¹²⁾ or private speech. ⁽¹¹³⁾ Erring on the side of freedom and accountability suggests preferring private speech categorization, but this will not fit every case. Even to speak of these considerations is to raise the troubling issue of content discrimination. Yet it may be hard to avoid.

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, ⁽¹¹⁴⁾ stated the central issue: did Boston [reserve] the pole to fly flags that communicate governmental messages, or instead [open] the flagpole for citizens to express their own views”? Following Summum, Breyer adopted a “holistic” approach, examining “the history of the expression at issue; the public's likely perception as to who (the government or a private person) is speaking; and the extent to which the government has actively shaped or controlled the expression.” Here he found that though the flags “usually convey the city's messages,” Boston did “not at all” control the flag raisings, and “that is the most salient feature of this case.” Boston had never bothered to review flag requests in the past, having approved them all. Thus, government's role was much smaller than in Summum or Walker. The flags represented speech by private groups, not government, so the neutrality principle prevailed and Boston's excuse for viewpoint discrimination evaporated.

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. ⁽¹¹⁵⁾ Alito, concurring, insisted that neither control nor final approval was dispositive, maintaining that the true issue was whether the speech was “the purposeful communication of a governmentally determined message by a person exercising a power to speak for government,” and whether government “did not rely on a means that abridges the speech of persons acting in a private capacity.” A prominent constitutional lawyer, writing before the Supreme Court heard the case, dismissed it as “an interesting moment in the ongoing culture wars, but [of] almost no practical legal importance . . because it can so easily be limited to its extraordinarily peculiar facts.” ⁽¹¹⁶⁾

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. ⁽¹¹⁷⁾

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.” ⁽¹¹⁸⁾ A typical description states that propaganda must meet four criteria: It “must be said by the government [; it must] assert a verifiably false or misleading statement of fact [; it must] concern a matter of public interest [; and it must] be made with actual malice,” ⁽¹¹⁹⁾ that is, the speaker must be aware that the statement is false or utter it with a reckless disregard for whether it is false. ⁽¹²⁰⁾ In this telling, we are surely revolted by the thought that we might be unknowing puppets controlled by some master government puppeteer.

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. ⁽¹²¹⁾ But what of values? Sacrificing for the fatherland? The glorification of the Aryan race? “Verifiably false”? According to whom? The speakers may sincerely believe the statements to be true. Do we fault them for not employing sophisticated statistical analysis or ignoring the latest research? Suppose (as is normally the case) that their statements are partially true? Suppose the propaganda is entirely (and uncomfortably) true, like, for example, China's pointing to the Gulf of Tonkin incident as an example of American war escalation ⁽¹²²⁾ and Operation Mockingbird as illustrating American efforts to shape world public opinion? ⁽¹²³⁾ Indeed, the definition's implicit outrage at lying seems uncomfortably misplaced, partly because propaganda need not be lies, but also because it ignores the widespread cliché of the crooked politician that suggests that the public may not be so easily taken in. The outrage also runs counter to the stance of the Supreme Court, which in United States v. Alvarez noted the pervasiveness of lying in human affairs, the useful functions it sometimes serves, and the futility of imagining that it could be eliminated. ⁽¹²⁴⁾ “Misleading”? It is a commonplace of political rhetoric (in fact, of rhetoric generally) to generalize from unrepresentative examples. The government's presentation of facts will always be incomplete, and thus always vulnerable to the charge that the facts omitted would have undermined the point made. When it urges the public not to smoke, must government also note that many smokers live long, healthy lives? And what of “public concern”? Does it also include matters of private concern, like gender identification, that much of the public seems concerned about? Like Justice Stewart and obscenity, we may think we know propaganda when we see it. ⁽¹²⁵⁾ But this subjective confidence can hardly support a First Amendment rationale. The popular notion of propaganda as dishonest opinion manipulation has a slipperiness about it that undercuts its utility.

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. ⁽¹²⁶⁾ There is still, however, a potent tradition that views public opinion, autonomous and active, as a key factor rendering government policies legitimate. ⁽¹²⁷⁾ Americans still recoil from what Walter Lippmann a century ago called the “manufacture of consent.” ⁽¹²⁸⁾ But where a later writer condemned government creation of a “falsified majority,” ⁽¹²⁹⁾ Lippmann, ever the realist, acknowledged both that the public needs experts to explain the world to them and that the experts will often be corrupted by self-interest. How, then, can government remain neutral? Even if government merely gathers and publishes data, it will have to decide on questions and criteria. If it decides to acquire information on race, for example, it must first address contentious questions. Is a black person a person with at least “one drop” of black blood? Is “Asian” a viable category, considering the immense variety it covers? More fundamentally, it is certainly plain that government does not passively respond to an agenda conceived elsewhere, but instead seeks out issues it believes merit attention or contribute to its support. The point is to get the public, or at least the significant portion of the public, to see the issue as the government does. ⁽¹³⁰⁾

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. ⁽¹³¹⁾ As A.V. Dicey observed in 1905: “Laws foster or create law-making opinion. This assertion may sound, to one who has learned that laws are the outcome of public opinion, like a paradox; when properly understood it is nothing but an undeniable though sometimes neglected truth outside the pale.” ⁽¹³²⁾ This interaction between government and public policy, present even in Dicey's era preceding modern communications technology, is far more salient over a century later. Its supporters call it leadership, which is almost universally regarded as a virtue necessary for effective government.

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.” ⁽¹³³⁾ Private virtues, that is, may not transfer to the public arena; indeed, the two sets of virtues will always be to some significant degree irreconcilable. The leader no longer merely looks out for himself and his family; his obligation now extends to the entire society, and in protecting and advancing its interests, Machiavelli teaches that the leader may have to resort to deception or force, choices that he might find repugnant on a personal level. He must choose evil in order to avoid a greater evil, and he must do this not merely in isolated times of extraordinary emergencies, but regularly on a day by day basis. “Truthfulness,” as Hannah Arendt observed, “has never been counted among the political virtues.” ⁽¹³⁴⁾ The resulting government speech will sacrifice some veracity.

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; ⁽¹³⁵⁾ often, the public is apathetic and ignorant; ⁽¹³⁶⁾ occasionally public opinion will favor an unworkable or clearly immoral path. These options may each give rise to official deception, which we may justify, as necessary to sustain public support for a desirable policy, or reject, as merely serving to protect officials’ reputation or position.

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.” ⁽¹³⁷⁾ But as government will always wish to prevail—why else try to persuade?—it will always be tempted to monopolize, coerce, or ventriloquize. In other words, if government seriously limits itself, there will be no problem, which hardly reassures us that there will be no problem.

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.” ⁽¹³⁸⁾ As one of the foremost students of government speech put it, “the government should pay for its ability to invoke the government speech defense by transparently taking political responsibility for its expressive choices.” ⁽¹³⁹⁾ How is the public to hold government accountable for its speech if it is unaware that government is speaking? This is not a new problem, as in nineteenth century America, “newspapers conducted many if not most of the opinion-shaping activities we now call campaigning.” ⁽¹⁴⁰⁾ Today, it is widely believed that attribution has become more difficult, as social media have greatly facilitated anonymous and pseudonymous messaging, offering countless opportunities to hide. If the public understands that government is the source of the information, presumably it may grasp the conflict of interest and approach the message with some skepticism. When President Trump tells us that covid-19 is nothing to worry about ⁽¹⁴¹⁾ or President Nixon's press secretary dismisses the Watergate break-in as a “third-rate burglary” ⁽¹⁴²⁾ or President Clinton denies having sex with Monica Lewinsky ⁽¹⁴³⁾ or President Kennedy rails against a “missile gap,” ⁽¹⁴⁴⁾ we at least retain the power to match their claims with the source of the message.

But suppose the role of the government in the messaging is hidden, so that the level of deception is taken to another level? ⁽¹⁴⁵⁾ During both world wars, the government utilized mass media to raise morale and support the war effort, sacrificing truth whenever it seemed advisable. In World War I, George Creel's Committee on Public Information distributed thousands of articles to newspapers and magazines that published them, either to avoid trouble with the government or from a patriotic impulse. ⁽¹⁴⁶⁾ In World War II, the Office of War Information's Bureau of Motion Pictures reviewed two-thirds of the 2,500 movies released from 1942–1945, altering nearly three-quarters of them; the head of OWI preferred working with movies because the people “do not realize they’re being propagandized.” ⁽¹⁴⁷⁾ At the same time, the government, appealing to the patriotism of reporters, withheld important information throughout the war. ⁽¹⁴⁸⁾ In these wars, then, government deceived us in pursuit of a goal, victory, we nearly all shared. Balancing means and ends, doubtlessly few of us would complain. In still more obvious circumstances, we understand that the government may deceive us because it cannot tell us the truth without sharing it with the enemy. Who would quarrel with President Roosevelt's decision to keep the upcoming D-Day invasion secret? Candor is not always our top priority. Sometimes we demand neutrality, and sometimes we don’t.

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,” ⁽¹⁴⁹⁾ and thus justify military cooperation in the form of advice, equipment or personnel. ⁽¹⁵⁰⁾ A result was a proliferation of movies that glorified the military, including The Longest Day, The Green Berets, and Top Gun. Did these movies and television programs contribute to an ambience supporting military adventures? We likely will never know.

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. ⁽¹⁵¹⁾ Instead of assuming an obligation of transparency, the government thought it sufficient to leave disclosure to the television executives, knowing that they had no reason to come forward with revelations. Hundreds of local televisions stations also broadcast news segments produced and distributed by the federal government on a wide range of topics without any acknowledgement of government's role. ⁽¹⁵²⁾ A broadcast commentator was also paid $240,000 by the Department of Education to support the administration's policies. ⁽¹⁵³⁾

Courts have repeatedly held that, however valuable transparency may be to democracy, it is not a constitutional requirement. ⁽¹⁵⁴⁾ In support of this, Greene observed that “citizens will often know speech is dictated by the government even if no disclosure is made,” adding that the question is “more a concern of political theory than of constitutional law.” ⁽¹⁵⁵⁾ However, that some citizens will know that government is speaking hardly argues against giving everyone the opportunity to know, particularly, as this can be easily achieved by simply identifying the government authorship or influence. Is the subject more a concern of political theory than constitutional law? It is hard to know exactly what this means, how it could be demonstrated, or what its relevance is.

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, ⁽¹⁵⁶⁾ even if the audience for so-called fake news websites is far smaller than popular accounts would indicate. ⁽¹⁵⁷⁾ All this is said to be magnified by the “post-truth” tenor of the times.

The government may be “entitled to say what it wishes,” ⁽¹⁵⁸⁾ but is it entitled to hide its authorship? The answer must be “it depends,” for transparency, like nearly everything else, is decidedly a mixed blessing. An emphasis on accountability, for instance, may contribute to a dysfunctional “culture of suspicion.” ⁽¹⁵⁹⁾ The Brandeis cliché that “sunlight is said to be the best of disinfectants” ⁽¹⁶⁰⁾ takes no account of the practicalities of negotiation and compromise, which may be made impossible by transparency. ⁽¹⁶¹⁾ Moreover, as Michael Gilbert has pointed out, transparency makes public information that may be used for corrupt purposes. ⁽¹⁶²⁾ Transparency, a means and not an end in itself, can only be evaluated in specific contexts. ⁽¹⁶³⁾

How, then, to counter potential abuse? The answer, the Court advances, is “accountability to the electorate and the political process for its advocacy.” ⁽¹⁶⁴⁾ As Justice Kennedy explained, “When the government speaks, for instance, to promote its own policies or to advance a particular idea, it is, in the end, accountable to the electorate and the political process for its advocacy. If the citizenry objects, newly elected officials later could espouse some different or contrary position.” ⁽¹⁶⁵⁾ And again, “it is the democratic electoral process that first and foremost provides a check on government speech.” ⁽¹⁶⁶⁾

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. ⁽¹⁶⁷⁾ Is the public sufficiently interested in the subject to read, analyze, and remember what the government said? Is the subject sufficiently important to the public for it to form opinions on this government speech and support or oppose government on that account? ⁽¹⁶⁸⁾ As consumers, our choices likely have direct and immediate effects on our lives; we buy chocolate ice cream because we love how it tastes. But political choices usually are so remote that we typically do not bother to assume that they have any practical effect. We may conclude that a candidate is on our side, but even if we are correct in this (which is necessarily problematical in a context of wholesale dissembling), we will almost certainly be incapable of adequately evaluating his policy proposals? Will trade protectionism help us? Reducing immigration? Taxing capital gains at a higher or lower level? As politicians obsessively seek to claim credit and avoid blame and so much in politics takes place behind closed doors, it is commonly hard to know to what extent they are even responsible for the policy. In such a context, as Anthony Downs observed, for most people most of the time, the rewards of becoming politically informed do not exceed the costs, in time, effort, rejected alternative activities, and so forth. ⁽¹⁶⁹⁾ Given this widespread rational ignorance, most of the public will not enforce accountability in the idealized fashion: they might never have heard of the issue to which accountability applies, they may have misunderstood the subject, they may have failed to act on their understanding of the subject. The voter may react simply on the basis of “whatever makes him feel best. When a person puts on his voting hat, he does not have to give up practical efficacy in exchange for self-image, because he has no practical efficacy to give up in the first place.” ⁽¹⁷⁰⁾ Most people, in any case, will not usually approach issues with a truly open mind.

The influence of government is also undercut by the two-step flow phenomenon. ⁽¹⁷¹⁾ Many of us belong to small groups, in which public affairs is not very important; as we care little, we may rely on a member of the group who cares more to act as an intermediary, passing on information and opinions from outside sources and altering them in the process because he is also subject to selective perception, confirmation bias, and other inherent distorting filters. What we learn from the opinion leader may be quite different from what the government or the media intended to convey. ⁽¹⁷²⁾ All of this is to say that the Court's implicit assumption that voters reach decisions on the basis of a specific government speech is, except in the most extreme possible examples, preposterous.

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. ⁽¹⁷³⁾

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, ⁽¹⁷⁴⁾ they believe these sources to be more credible and less biased, ⁽¹⁷⁵⁾ their partisan identification likely slants their evaluation of political information, ⁽¹⁷⁶⁾ and they share the information they acquire with people who share their beliefs. ⁽¹⁷⁷⁾ People may imagine that they are searching for truth, but they may merely be searching for confirmation. Academics and others influenced by fashionable postmodernism, furthermore, are uncomfortable with the notion of objective truth, holding instead that “all judgments are contingent cultural products.” ⁽¹⁷⁸⁾ Thus, “there can be no such thing as knowledge of reality; what we think is knowledge is always belief and can apply only to the context within which it is asserted.” ⁽¹⁷⁹⁾

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,” ⁽¹⁸⁰⁾ he was basically correct, in that most of his supporters would subconsciously find a way to reduce the dissonance and continue their positive evaluation of him and his message. ⁽¹⁸¹⁾

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.” ⁽¹⁸²⁾ Nor do well informed people necessarily resolve their differences or agree; differences typically turn on values and preferences, which are not susceptible to confirmation or disconfirmation.

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. ⁽¹⁸³⁾ Still, accountability has a hit-or-miss quality about it, and doubtless the vast bulk of government speech will provoke no reaction that can be translated into accountability. Transparency, widely considered “the sine qua non of good governance” ⁽¹⁸⁴⁾ and cloaked with a “quasi-religious significance,” ⁽¹⁸⁵⁾ cannot by itself solve the problem.

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, ⁽¹⁸⁶⁾ for example, the Warren Commission Report, ⁽¹⁸⁷⁾ and some informational speech, like a political press conference, has an obviously mixed governmental and partisan rationale. ⁽¹⁸⁸⁾ All this suggests that government speech, like government streets, is beset by potholes that may undo the unwary.

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. ⁽¹⁸⁹⁾ Paul Stephan dates the change in the 1930s, ⁽¹⁹⁰⁾ Genevieve Lakier counters that neutrality “has been a feature of free-speech law ... since the eighteenth century.” ⁽¹⁹¹⁾ Whenever the origin, there is no question that the First Amendment has taught us that, in general, ⁽¹⁹²⁾ “government has no power to restrict expression because of its message, its ideas, its subject matter or its content.” ⁽¹⁹³⁾ We take its obligation of neutrality for granted. But we also take for granted that government will inform and educate us on a wide range of topics, and understand that this responsibility is incompatible with neutrality. Hence the conclusion of one careful observer, “It is plausible to view the development of the government speech doctrine in large part as an effort to relieve the government of the suffocating demands of the prohibition on viewpoint discrimination.” ⁽¹⁹⁴⁾ “When government speaks,” as the Court put it, “it is not barred by the Free Speech Clause from determining the content of what it says.” ⁽¹⁹⁵⁾ How to distinguish speech from government that requires neutrality from speech from government that does not?

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.” ⁽¹⁹⁶⁾ The most obvious problem is accepting the power of consensus. As Mill famously wrote, “If all mankind minus one were of one opinion, and only one person were of the contrary opinion, mankind would be no more justified in silencing that one person, than he, if he had the power, would be justified in silencing mankind.” ⁽¹⁹⁷⁾ Kagan's is the path to conformity and timidity, not liberty and curiosity; too many consensuses have vanished for us to suppose them to be infallible. Even on its own terms, the proposal makes little sense. Consider her topic, smoking. Stopping smoking may be a question of weighing risks and rewards, but it is not a matter of right and wrong; nearly sixty years after the Surgeon General's famous report, ⁽¹⁹⁸⁾ the claim of consensus is still challenged by tens of millions of smokers; and the question of harm arguably requires individual answers. Kagan found a consensus in 1992; fifty years earlier, the consensus favored smoking. If government, seeking to combat the harmful effects, had initiated a stop smoking campaign in 1942, would it have had to submit to the neutrality requirement, perhaps featuring Chesterfield's claim that it was the brand that doctors smoke?

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.” ⁽¹⁹⁹⁾ His formula sounds straightforward and sensible until we recall that government will always defend its speech as within its functioning ambit. Indeed, as government can rationalize almost any but the most brazenly partisan speech as tied to a governmental function, it is hard to see this as serving as much of a practical limitation.

Helen Norton, having devoted an entire book to government speech, devised some useful criteria. ⁽²⁰⁰⁾ Her three questions, however, are actually six questions, and most of the questions themselves are subdivided further. The result is a scheme that, however sensible if considered bit by bit, is simply too complicated to be usefully applied.

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”? ⁽²⁰¹⁾ Under what conditions does it extend to others acting as proxies of the government, even unwilling proxies? Has burgeoning privatization meant that firms performing public functions, like running schools or prisons, should have their speech characterized as government speech? If government speech reflects a consensus, what voice is left for those who dissent? If the consensus dissipates, will the government speech label dissipate with it? Does the Constitution that famously protects our right to speak against government interference also protect the government's right to silence us? Government silencing of private speech is not per se bad. To accomplish their work, school teachers may force their students to be quiet. Should this rationale also keep Dr. Rust from telling his patients about abortions? What to do about mixed government-private speech? If the defining quality of government speech is that the message is effectively controlled by the government, do we ignore considerations of intent and effect? The questions come in crowds.

The Court's answers, sad to say, have not been very helpful, hence their routine description as “unprincipled,” ⁽²⁰²⁾ “nefarious,” ⁽²⁰³⁾ or “intellectually undeveloped.” ⁽²⁰⁴⁾ Perhaps, one reason for this failing is the set of cases the Court has used to construct the doctrine. Except in peripheral ways, none of these cases addresses the core problem: How to distinguish between government's power to inform, educate, and persuade and its power to manipulate? How to prevent government speech from abridging private speech? Apart from Rust, few people even care about the key cases’ practical consequences. It is hardly surprising that the parched soil of NEA grants, license plates, park monuments, and city hall flags produced such scrawny plants.

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.” ⁽²⁰⁵⁾ Consistency is undercut, when Velazquez, a case similar to Rust, produced the opposite result. Alito in Summum thought the key questions were whether government had a history of using the particular means of communication and whether it had direct control over the message. In an era when new means of communication arise like mushrooms after a rainfall, an emphasis on history seems misplaced. It also assumes that past practice was acceptable, which may sometimes be quite dubious. And the term “direct control” contains a hidden ambiguity: does it refer to the government's composing the message or merely approving it, which may in practice be pro forma? In Summum, Alito ruled that approving monuments in parks constituted government speech, though he was unwilling or unable to say what the message was, because final government approval was required. In Walker, he maintained that final approval counted less than the authorship of the message. Breyer, concurring in Summum, was even less helpful. He wrote of government “disproportionately” burdening private speech, but the word is so hopelessly vague and subjective that adopting it as a standard could only breed disagreement and confusion. Justice Scalia was surely correct that all these cases saw government decisions driven by viewpoint considerations, but in rejecting the relevance of that fact, he also added little to doctrine.

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.” ⁽²⁰⁶⁾ This approach, which in a way mimics the tort law's concept of the reasonable person of ordinary prudence, might encourage transparency and objectivity, while generating precedents that might offer some clarity. ⁽²⁰⁷⁾

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, ⁽²⁰⁸⁾ but it might be close enough to be cause for concern.

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.” ⁽²⁰⁹⁾ By freeing government from the usual obligation of viewpoint neutrality, government speech imperils the traditional marketplace of ideas because government is so much stronger than all the other participants. If the bedrock rationale of the First Amendment is, in Justice Brennan's famous words, “a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open,” ⁽²¹⁰⁾ government speech would seem sometimes to protect and elevate speech whose chief impact is profoundly hostile to the freedom of speech. Of course, sometimes government speech that silences private speech is a phenomenon we take for granted, as with a Marine drill instructor and his cadets. But government speech, amoeboid in its boundaries, is not confined to these unexceptional cases. Why endorse government speech that includes government's power to prevent private persons from speaking?

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.