Introduction

In 2020, the U.S. Agency for Global Media (USAGM) was sued by several of its employees. USAGM oversees U.S.-funded international broadcasting outlets, including the Voice of America (VOA). The plaintiffs, five USAGM senior managers and VOA’s program director, alleged that USAGM CEO Michael Pack, who was appointed by President Trump in 2020, “[had] sought to interfere in the newsrooms of the USAGM networks, in violation of their eighty-year practice … of journalistic autonomy.” Plaintiffs accused Pack of “seek[ing] to quash … coverage that is insufficiently supportive of President Trump,” as well as “any coverage, unless unfavorable, of President Trump’s political opponents.” These actions, the plaintiffs charged, ran afoul not only of statutory commands but of the First Amendment. USAGM responded that VOA and the other networks speak on behalf of the government and lack any First Amendment rights in so doing. In taking the actions that he did, Pack was simply “exercis[ing] his [own] authority to ‘direct and supervise’ and to ‘assess the quality, effectiveness, and professional integrity of’ USAGM” reporting.

The First Amendment arguments in this case, Turner v. USAGM, reflect a broader tension in the case law concerning the government’s role as “knowledge producer”—that is, its role in producing or conveying information or otherwise fostering knowledge. From the plaintiffs’ perspective, the government ties itself to a mast when it purports to produce journalism. That mast is comprised of the norms of professional journalism, including a strict separation between an operation’s business or political commitments and its journalistic endeavors. This argument is consistent with several strands of Supreme Court case law. For example, the Court repeatedly has held that, although government is not required to subsidize private speech or create speech forums, once it does so, it may not impose restrictions that are based on viewpoint or that are incompatible with the very nature of the speech subsidized or forum created. The defendants, on the other hand, invoked aspects of free speech doctrine that emphasize the government’s broad discretion to control the speech that it produces. This includes the Garcetti rule—stemming from the 2006 Supreme Court case of Garcetti v. Ceballos—whereby government employees generally are unprotected by the First Amendment for their work product speech, meaning speech that they produce as part of their job duties. Garcetti itself arguably is in tension with the Court’s acknowledgment elsewhere to the effect that “speech by public employees on subject matter related to their employment holds special value precisely because those employees gain knowledge of matters of public concern through their employment.”

Similar First Amendment questions are raised by battles currently raging over state legislative proposals to curtail discussions of race and racism in another site of knowledge production: public colleges and universities. The laws’ opponents argue that they are antithetical to the very nature of higher education. They suggest that states tie themselves to the mast of academic freedom norms—including rules of faculty and intradisciplinary governance on matters of scholarship and pedagogy—when they create colleges and universities. The laws’ proponents, on the other hand, emphasize the “public” in public education, suggesting that schools effectively belong to the public, are funded partly by their tax dollars, and that members of the public, through their representatives, must have a say in what is taught and studied at the schools.

These First Amendment controversies are layered on top of major cultural and political tensions. This is unsurprising, as public knowledge institutions often are sites of cultural contestation. If one reviews the public debates alongside the legal arguments about these matters, one can find illuminating overlaps between the two. For example, public outcries against the press and against “critical race theory” are often framed as fights against indoctrination by elites. From this perspective, it is not journalism or higher education that is under siege. Rather, it is ordinary folk who are seeking simply to right the balance and to reclaim neutrality in public spaces. These arguments parallel legal arguments to the effect that government employees, or persons carrying out government-subsidized functions, have no constitutional right to speak freely while carrying out their government-supported roles. Their words effectively belong to the people.

The shared populist core of the arguments for broad political control of public knowledge institutions betrays the arguments’ fundamental failings. First, the notion that political might should govern knowledge production runs counter to the very idea of discipline-based knowledge and expertise; it would strip knowledge production of its meaning and value. Worse still, it would mislead consumers of any “knowledge” so produced, because the knowledge would purport to stem from disciplinary best practices and expertise. Such deception poisons the speech marketplace and is antithetical to core First Amendment values. Second, the conceit that political controls protect against indoctrination and support neutrality is belied by the nature of the power that proponents of political control seek: the power to bar or require certain speech content in public schools and in other public knowledge institutions.

In this essay, I explore the nature and value of government’s knowledge producers in our constitutional order and the legal, cultural, and political threats that they face. In Part I, I explain that public knowledge producers are an essential part of a democratic society, and that their worth depends partly on their having some insulation from political pressure. In Part II, I use the example of international broadcasting, with an emphasis on the USAGM case to argue that such insulation is called for not only as a matter of good policy but as a matter of First Amendment theory. I acknowledge, however, that First Amendment doctrine is more mixed; one can find support for this position, as well as contrary indicia in judicial precedent. I also make the case for more robust doctrinal support to insulate public knowledge producers going forward. In Part III, I explore the broader legal, political, and social contexts. With respect to law, I observe that legislation plays at least as crucial a role in protecting knowledge producers as does the First Amendment. Yet such legislation increasingly is under threat by the Supreme Court’s growing allegiance to unitary executive theory. I also explore parallels between judicial reasoning in some of the First Amendment case law, unitary executive theory, and cultural and political movements against knowledge producers. Finally, I apply some of my earlier analyses to one last set of examples: ongoing legal and political controversies concerning the topic of race in public higher education.

I. The Nature and Value of Public Knowledge Producers … and Their Independence

In a time of epistemic crisis and creeping authoritarianism, it is only fitting that some scholars have turned their attention to the institutional forces that produce knowledge. In a sprawling history of the historical relationship between democracy and truth, for example, Sophia Rosenfeld cites the democratic importance of “mediating and educational institutions,” or “knowledge-producing institutions.” Vicki Jackson, in recent work, compares “knowledge institutions” across constitutional democracies. She deems such institutions “fundamental to the success of constitutional democracy.” Jackson offers a helpful definition of knowledge institutions: “ongoing entit[ies,]” either public or private, “whose principal purpose is knowledge production or dissemination … according to disciplinary norms.”

I adopt the term “public knowledge producers” in this essay. My definition overlaps in large part with Jackson’s, but my focus is narrower in one respect and broader in others. It is narrower in the sense that I focus solely on government-supported actors. To be sure, private institutions are an essential part of any ecosystem of democratic knowledge. My emphasis, however, is on the special opportunities and challenges presented by public knowledge producers, including the unique constitutional questions that they raise. I use the term “public” to refer to individuals and entities who are funded in whole or in part by the government. This includes government agencies and employees, as well as entities and individuals who operate at least partly through government subsidies or contracts. To be clear, this criterion is not sufficient to distinguish public from private actors for all purposes. In the context of this essay, however, the definition is fitting for two reasons. First, I stress the value of government’s subsidizing the production of knowledge. Any amount of public support furthers that value, even if more support is preferable to less. Second, the problems that I tackle in this essay arise when the government leverages such support—whether in the form of employment, subsidies, or otherwise—to restrict the speech of a subsidized entity or person. The concerns that I address thus are implicated whenever the government imposes restrictions on speech as a condition of its largesse, in whatever amount and form that largesse takes.

Within the confines of public entities and individuals, however, my definition extends somewhat further than Jackson’s. First, I use the term “producers” rather than “institutions” to make clear that knowledge creators can include individuals within their larger institutions, as well as the institutions themselves. Indeed, as USAGM and Garcetti both illustrate, some difficult constitutional and policy questions pit the speech of individuals within institutions against the institutions themselves, with the latter represented by supervisors or others atop the institutional hierarchy. Furthermore, even entities that do not themselves constitute knowledge institutions might assign one or more individuals to play a knowledge-producing role. For example, internal watchdogs—most notably inspectors general—are quintessential knowledge producers. The essence of their jobs is to gather and report information in service of accountability.

Second, and this may be a small quibble as a practical matter, I would not limit the definition of knowledge producers to those “whose principal purpose is knowledge production or dissemination.” Rather, I would take a somewhat more liberal view, asking whether the individual or entity’s tasks, in the ordinary course of their job, include “knowledge production or dissemination … according to disciplinary norms.” The value of discipline-based knowledge production is no less significant in cases where that production constitutes an ordinary, if not a principal part of one’s job responsibilities. To the contrary, it is essential to incorporate knowledge production throughout public institutions and roles, rather than cordoning it off as a specialized project. More importantly, politicization efforts can target discrete knowledge-production tasks—for example, political appointees might retaliate against a public employee for expertise-based reporting that is a regular but nonprincipal part of their job—just as they can target entire knowledge institutions or knowledge-based jobs.

Public knowledge producers include government scientists, economists, inspectors general, as well as public universities and their faculty members. As Jackson and others have concluded, such organizations and actors are essential to a well-functioning democratic society. They are necessary, in part, for the same reasons that knowledge producers on the whole are democratically crucial. For one thing, they help to build and maintain a “shared epistemic foundation” on which officials can base public policy decisions and citizens can meaningfully judge those decisions. They also foster the critical thinking skills that empower citizens to evaluate information, to assess candidates and policy questions, and to “resist manipulations by those in high office” and elsewhere. It is no surprise that autocracies seek to destroy the independence of knowledge producers, if not eliminate their roles entirely.

Public knowledge producers also add value that stems specifically from their public status, although that value is contingent on their having some insulation from political pressures. First, their location within government agencies helps to keep government itself honest and accountable. For example, career scientists and economists with civil service protections can produce reports based on expertise and disciplinary best practices. Such reports can present a challenge to political actors who wish to misrepresent factual information or expert consensus. Other knowledge producers, such as inspectors general, can suss out government lies and misdeeds more directly.

Public knowledge producers—including disciplinary experts found in public universities and in agencies—also fill informational and educational gaps that the market alone cannot be relied on to fill. They can provide services that are central to democracy but that are insufficiently profitable for private actors to provide consistently on their own. Relatedly, public funding liberates knowledge producers from market demands that might otherwise compromise the integrity of their research and educational programs.

Publicly supported journalism illustrates the importance of “public options” to supplement private knowledge production. Although the United States provides far less financial support per capita for public journalism than do other democracies, it has provided some support—mostly in the form of postal subsidies for newspapers and magazines—since the nation’s founding. In the 20th century, it created the Corporation for Public Broadcasting, which to this day provides partial funding for Public Broadcasting Service (PBS) and National Public Radio.

Public opinion surveys reflect the relatively high esteem in which viewers and listeners hold public news organizations. For example, respondents in an annual survey have rated PBS “the most trustworthy institution among nationally known organizations for [18] consecutive years.” This trust stems partly from perceptions of PBS’ political independence. Another survey conducted in 2021 found that “the political leanings of PBS viewers span the spectrum from extremely liberal to extremely conservative.” A majority of that survey’s 1,500 respondents “rank[ed] PBS as neutral when asked about ‘bias in news.’”

Apart from the substantive value of a public news option, such an option is increasingly necessary to ensure that all Americans can access fact-based, professional news services, particularly local news and resource-intensive investigative reporting. The current economic landscape for news media is very bleak. More than a decade ago, researchers at University of Southern California’s Annenberg School for Communication and Journalism concluded that “the financial model for news is facing the greatest crisis in decades.” The situation has only worsened since then. Last year, Victor Pickard and Timothy Neff observed that “[t]he newspaper industry … has seen its number of employees reduced by well over fifty percent in the last two decades. As newspapers downsize or close altogether, news deserts are spreading rapidly across the country, and hundreds of communities no longer have any local media coverage whatsoever.” Nor, they add, will the market right the problem. “As newspapers’ advertising revenue continues to shrink, local news is no longer commercially viable in many markets and communities. In the meantime, hedge funds and other parasitic buyers continue to devour and dismantle what’s left of the local media landscape.”

Confronting this crisis calls for a variety of initiatives, including philanthropy and indirect government support as in subsidies and tax incentives. Public journalism, too, is a crucial part of any such efforts. The U.S. provides far less financial support for public reporting than do many other democratic nations. Although the U.S. has the world’s largest gross domestic product (GDP), it is 25th out of 27 countries examined by Pickard and Neff in percentage of GDP spent on public media. In contrast, “[t]he world’s strongest democracies, such as Norway, Sweden, and Finland, also have the best-funded public media.” The U.S. funding model for public broadcasting is also an unstable one that leaves news producers vulnerable to political pressure. When PBS was first created, for example, the Carnegie Commission on Educational Television proposed funding it through taxes on television set sales. Instead, the U.S. chose to fund, and continues to fund, PBS through direct congressional appropriations.

Despite these limitations, public news broadcasting has been remarkably successful, as suggested by the surveys cited above. We need more of it, and it is overdue for changes in its funding model and its delivery systems. With respect to the latter, for example, a more robust print and internet presence seems essential. But public journalism itself is a crucial democratic presence, more so today than ever.

Apart from funding, the key concern cited by public media critics is the risk of state capture, or undue influence by political or partisan interests. Though this concern is a legitimate one, the answer is not to give up on the very idea of public media. As we have seen, even under current conditions, public media is widely viewed as less ideologically skewed than other news sources. Rather, we must maintain—and, where lacking, build—a public media infrastructure that enables its journalists to operate within the confines of disciplinary norms and professional journalistic judgments, rather than political pressure. This is achieved partly through policy decisions, such as funding sources and legislative and regulatory directives. The governing constitutional framework, too, is no small matter.

In Part II, I explore the First Amendment principles and doctrine that ought to govern publicly supported journalism. Using the recent USAGM case as a jumping-off point, I explain that government programs that purport to deliver services associated with a communicative discipline—journalism, in this case—should be understood to bind themselves to the mast of that discipline’s norms as a matter of First Amendment law. I situate this conclusion within existing case law. Specifically, I observe that important strains of precedent already take this view, but that they are in tension with other, more restrictive areas of judicial doctrine.

II. Public Knowledge Production and the First Amendment: The Example of USAGM

A. Background: International broadcasting and the firewall

The United States’ longstanding involvement in international broadcasting presents an illuminating twist on the concept and value of public media. The oldest and best known of the U.S. international broadcasting ventures is VOA, which was created in 1942 to broadcast to Europe throughout World War II. VOA continued to broadcast after the war and today reaches an estimated weekly audience of roughly 300 million people around the globe through digital, television, and radio content. The U.S. also broadened its international broadcasting offerings after World War II. Today, it boasts two federal broadcasting networks—VOA and the Office of Cuba Broadcasting. It also started three networks—Radio Free Europe, Radio Free Asia, and the Middle East Broadcasting Networks—that are formally headed by nonprofit corporations but that the U.S. supports through grants.

These media ventures reflect two core goals: advancing U.S. foreign policy interests and transmitting independent journalism that adheres to the highest professional standards. The latter commitment is epitomized by VOA’s first transmission in 1942: “‘The news may be good or bad; we shall tell you the truth.’” It has also been expressed through various statutory directives over the years. For example, the current statutory framework governing the networks requires both the secretary of state and the CEO of USAGM to “respect the professional independence and integrity of [USAGM], its broadcasting services, and the grantees of [USAGM].” It also demands that U.S. international broadcasting “be conducted in accordance with the highest professional standards of broadcast journalism.” The former commitment—advancing U.S. foreign policy interests—is spelled out in a statutory directive to the effect that U.S. international broadcasting “shall … be consistent with the broad foreign policy objectives of the United States.”

It is a delicate task, undoubtedly, to strike a balance between advancing journalistic professionalism and independence and guarding U.S. foreign policy interests. But doing so is both possible and worthwhile. Through the statutory provisions governing USAGM and the networks, the U.S. has evinced the view that its foreign policy interests intrinsically are served by exporting journalism that foreign audiences can trust on the basis of its independence and commitment to journalistic best practices. At the same time, the relevant statutes impose some content directives beyond “professional independence and integrity” and the “highest professional standards of broadcast journalism.” They require, for example, that VOA news “be accurate, objective, and comprehensive,” and “present a balanced and comprehensive projection of significant American thought and institutions.” Furthermore, U.S. international broadcasting should provide “a variety of opinions and voices from within particular nations and regions prevented by censorship or repression from speaking to their fellow countrymen.” Trickier than these fairly open-ended directives, which in themselves are compatible with professional journalistic standards, is the matter of who shall enforce these standards and how shall they do so—how, in particular, USAGM shall reconcile the directives of journalistic professionalism and independence with its CEO’s statutory authority to “direct and supervise all broadcasting activities”; “assess the quality, effectiveness, and professional integrity of, all such activities within the context of the broad foreign policy objectives of the United States”; and ensure that U.S. international broadcasting adheres to the requisite “standards and principles,” including the criteria, such as professionalism and objectivity, cited above.

The most obvious means by which to reconcile these various directives and authorities is by building a firewall between the networks’ day-to-day journalistic operations and USAGM management, as well as other upper-level executive branch personnel, including the presidentially appointed Broadcasting Advisory Board. Such a firewall itself parallels the separation practiced in private journalistic enterprises between day-to-day editorial decision-making and commercial and institutional considerations. VOA Director Amanda Bennett, one of the plaintiffs in USAGM, signed a declaration to this effect based on her decades of experience as a journalist. Referencing her previous places of employment, she wrote:

The Wall Street Journal, The Oregonian, The Lexington Herald-Leader, The Philadelphia Inquirer, and Bloomberg News all had a strong system in place—a de facto firewall—to protect the independence and integrity of their journalists. At each of these institutions, though the word “firewall” is not always used, this separation prevented the business side of the network, which has a directed interest in maximizing profits, from interfering with editors and reporters, who were responsible for producing journalism.

B. Free speech and the anti-distortion principle

Striking this balance is not solely a question of statutory interpretation and policy judgment. It has a constitutional dimension as well. Specifically, one can draw an anti-distortion principle from existing First Amendment case law and from the normative commitments associated with free speech. Such a principle constrains the government from restricting speech in a manner that would distort the nature of a communicative good that it purports to provide. In other words, government must tie itself to the mast of the disciplinary norms of institutions that it purports to establish or to subsidize.

There are two core justifications for the anti-distortion principle, one negative and one positive. The negative justification involves avoiding the harm caused when the government commits to deliver a type of communicative good that is associated with expertise and disciplinary norms—in this case, journalism that is explicitly billed, by statute, as independent and adhering to the “highest professional standards” —but undermines such expertise or norms in a way that skews the good’s communicative output. Such distortion runs counter to some of the most significant concerns underlying constitutional free speech protections. This includes the notion that free speech is a prerequisite of democratic government, enabling the public to exchange ideas and information. It also includes a healthy distrust of government’s capacity and motivation to abuse censorial powers to skew information and debate in its favor. Although these concerns relate most naturally to communications geared toward domestic audiences—for instance, to public broadcasting that airs predominantly in the United States—they apply to U.S. operations geared toward foreign audiences as well. As a practical matter, it is simply not realistic in this age of digital media and global interconnectedness to proceed as though U.S. audiences will not consume media created by the U.S. government for foreign audiences. Even if such airtight division were plausible, the effects of distortion on the speakers—those hired, ostensibly, to conduct journalism and then used as propaganda vessels—would remain a serious concern.

There is also a positive justification for the anti-distortion principle, one based on its affirmative furthering of First Amendment interests. By protecting the integrity of public knowledge production, the principle helps to support the existence of a shared epistemic foundation. Much more than a core of shared facts, such a foundation includes lessons, if only by example, in building knowledge and thinking critically about the information that one receives. Indeed, some scholars, ranging from Alexander Meiklejohn in the mid-20th century to Martha Minow in more recent days, have argued that the First Amendment can be construed to impose affirmative obligations on government to maintain institutions that foster public knowledge. Although I take no position here on the constitutional imperative of public knowledge production, I agree that such production at minimum is essential from a policy perspective. Such production fills gaps that the market leaves open and adds unique value by virtue of its noncommercial nature. Much of this value is lost, however, if public knowledge production is readily subject to political manipulation. Indeed, politically skewed communications that are presented in the guise of independent, discipline-based knowledge do not simply fail to capture the advantages of public knowledge production; they affirmatively cause harm.

C. First Amendment doctrine: Anti-distortion and competing forces

There is considerable support for the anti-distortion principle in existing First Amendment case law. However, the doctrine also contains substantial counterforces. In this section, I use the USAGM case as a jumping-off point to explore this tension in two closely related areas of First Amendment law: that involving public programs and subsidies, and that involving government employees. In both realms, I explain, the judicial precedent most consistent with the anti-distortion principle gets things largely correct. That precedent includes the district court’s decision in USAGM, which was not appealed in light of the change in administrations that followed it.

1. On Garcetti’s reach, the government speech doctrine, and the anti-distortion principle

After finding that at least one plaintiff had standing, and that the plaintiffs’ First Amendment claims were not preempted by the Civil Service Reform Act (CSRA), Chief Judge Beryl A. Howell addressed the applicability of the 2006 case of Garcetti v. Ceballos. Howell held that Garcetti—which deemed public employees categorically unprotected for speech that constitutes work product—does not apply to “the core editorial or journalistic functions of government-employed journalists.” Howell analogized journalism to academic research and teaching, which the Garcetti Court itself had suggested might fall beyond Garcetti’s reach. Writing for the Court in Garcetti, Justice Kennedy had acknowledged that “[t]here is some argument that expression related to academic scholarship or classroom instruction implicates additional constitutional interests that are not fully accounted for by this Court’s customary employee-speech jurisprudence.” In the years since Garcetti, “at least two circuits have found that Garcetti does not apply to the teaching and writing activities, undertaken ‘pursuant to their official duties,’ of faculty members at public universities.” Judge Howell cited these developments with approval, noting “the special status of teachers and academics in the First Amendment tradition.” Turning to journalism, she observed that “[f]reedom of the press holds an equally exalted place in the First Amendment firmament.” It follows, she concluded, that “Garcetti does not apply to the core editorial or journalistic functions of government-employed journalists.”

An approach like Judge Howell’s—one that takes seriously the Garcetti Court’s suggestion that academic research and teaching might retain some First Amendment coverage and extends this carve-out to analogous activities—can help to preserve the integrity of public knowledge production. Going forward, however, courts should place such carve-outs on firmer footing by relying explicitly on the anti-distortion principle. They should make clear, in other words, that public academics’ speech retains some protection because the alternative—unfettered state control of such speech—is incompatible with the very nature of the academic enterprise. The same can be said of unfettered state control of speech produced by publicly funded journalists or by other public knowledge producers. The problem, in short, is not solely the infringement on a highly valued type of speech, such as academic writing or journalism. The problem is the government’s purporting to fund the production of such speech while distorting central features of that production.

To glean an anti-distortion principle from precedent, one can begin within the public employee speech cases themselves. Since its 1968 decision in Pickering v. Board of Education —the first in the line of modern public employee speech precedents—the Court has stressed that the First Amendment covers public employee speech not only because such speech is valued by the speaker but because of its social significance. In the 2004 case of San Diego v. Roe, for example, the Court acknowledged that “[u]nderlying the decision in Pickering is the recognition that public employees are often the members of the community who are likely to have informed opinions as to the operations of their public employers, operations which are of substantial concern to the public.” More recently, in Lane v. Franks, the Court rejected the argument that speech about one’s work is necessarily work product that falls within the Garcetti rule. Writing for the Lane Court, Justice Sotomayor explained that “speech by public employees on subject matter related to their employment holds special value precisely because those employers gain knowledge of matters of public concern through their employment.” Implicit in this “special value” rationale are notions of both access and expertise. Public employees bring something special to the speech marketplace partly because they are privy to information about the government that the average citizen cannot access on their own. Equally important is their professional expertise, including their ability to make sense of and convey the information to which they have special access. For such expertise to add value, however, it must remain undistorted by political or other extrinsic pressures and reflect the speaker’s best professional judgment.

Concerns about distortion shape the doctrine more overtly in several cases involving government programs, subsidies, and public forums. Among these precedents is the 2001 case of Legal Services v. Velazquez, which involved a federal statutory restriction on the use of congressionally appropriated Legal Services Corporation (LSC) funds. The restriction prohibited attorneys, in the course of LSC-funded representation, from challenging the constitutionality of state or federal welfare laws or the consistency of state welfare laws with federal statutes. LSC attorneys were free to argue that agents had interpreted or applied welfare statutes incorrectly. They were forbidden, however, from challenging the legality of the statutes themselves. The Supreme Court held that the law violated the First Amendment. Its fundamental failing, said the Court, was that it attempted to “use an existing medium of expression and to control it, in a class of cases, in ways which distort its usual functioning.” Specifically, Congress had sought to fund activities within the legal system under conditions that distorted the role of LSC lawyers. This disadvantaged clients and courts, who depend on attorneys to “present all the reasonable and well-grounded arguments necessary” in a case. To make matters worse, the statutory restriction shielded the government from zealous advocacy against itself. “We must be vigilant,” the Court admonished, “when Congress imposes rules and conditions which in effect insulate its own laws from legitimate judicial challenge.”

In several other cases involving government funding or property, the Court similarly observed that the government may not impose conditions on an expressive medium that it funds or creates where those conditions would distort the nature of the medium. In the context of limited public forums, for example, the Court has explained that once the government opens such a forum, “[it] must respect the lawful boundaries it has itself set. The State may not exclude speech where its distinction is not ‘reasonable in light of the purpose served by the forum,’ nor may it discriminate against speech on the basis of its viewpoint.”

To be sure, there are significant counterforces in the case law as well. The most important one is government speech doctrine. This is the notion that, “[w]hen government [itself] speaks, it is not barred by the Free Speech Clause from determining the content of what it says.” Thus, if a court finds that a government-funded position or program is a vehicle for government messaging, the government is free to shape the resulting expression however it wishes. Were government speech doctrine limited to cases involving expression that is transparently conveyed as a scripted message from the government, it would not be problematic. Indeed, such expression might even further government accountability. As Justice Alito suggests, for example, in the 2015 case of Walker v. Sons of Confederate Veterans, “it is the democratic electoral process,” rather than the First Amendment, “that first and foremost provides a check on government speech.” However, democratic checks cannot be counted on when the government’s impact on speech is not transparent. This is a growing concern, as the Court has decided several cases in which it characterized as government speech expression that the government played a nontransparent role in shaping.

The distorting effect of an expansive government speech doctrine is especially worrisome when speech is presented to readers and listeners as the product of expertise. Such a scenario is epitomized by Rust v. Sullivan, the 1991 case that was the first in the line of expansive government speech cases. In Rust, the Court upheld federal regulations barring family planning clinics from mentioning abortion in the course of providing federally subsidized counseling. Rust strikes me as incorrect from an anti-distortion perspective. Although the program that the Court upheld purported to subsidize health care, it conditioned that subsidy in a manner that distorted the expressive relationship between patient and health care provider. As Robert Post put it, the role of a physician requires the exercise of “independent professional judgment.” Patients, in turn, “expect and rely upon that judgment.”

Although one can draw distinctions between Rust and Velazquez—indeed, the Supreme Court did just that in Velazquez and has distinguished the two cases from one another elsewhere—I do not see much material difference between the two with respect to distortion. I raise this point only to stress, again, that these two threads of case law—Garcetti and the government speech cases on the one hand, and those aspects of employment, subsidy, and public forum cases that protect subsidized speech from distortion—are in tension with one another. For those concerned about distortion, then, the task at hand is to limit the reach of the former and to broaden the latter’s scope.

This brings us back to Garcetti. Garcetti’s rationale echoed the government speech doctrine in part, with the Court stating that restrictions on work product speech “simply reflect[] the exercise of employer control over what the employer itself has commissioned or created.” The problem with this rationale is its overbreadth. Certainly, some government employees are hired to convey messages crafted by the government. But as noted earlier, and as others have pointed out in response to Garcetti, much public employment does not entail such a role. To the contrary, many government employees are hired to engage in knowledge production in the ordinary course of their jobs. Treating knowledge production as scripted government speech is a recipe for distortion. Worse still, it invites distortion that insulates the government from speech that might criticize it. Such criticism might be direct—for example, it could take the form of an inspector general’s damning findings. Or it might be indirect—say, a scientific report that casts doubt on the premises underlying a president’s policy goals.

Yet Garcetti also contains the seeds of its own containment through the anti-distortion principle. First, Garcetti explicitly embraced the reasoning of the Pickering line of cases to the effect that public employee speech has special, expertise-driven value. Indeed, in the Supreme Court’s first major public employee speech case following Garcetti—the 2014 case of Lane v. Franks—the Court rejected the argument that speech about one’s public employment is necessarily work product speech that is not covered by the First Amendment. Writing for the Lane Court, Justice Sotomayor observed that “our precedents dating back to Pickering have recognized that speech by public employees on subject matter related to their employment holds special value precisely because those employees gain knowledge of matters of public concern through their employment.” Second, Garcetti floats the possibility of a special exception for academic speech. Although Justice Kennedy’s opinion does not elaborate much on this idea, it logically rests, as we have seen, at least partly on anti-distortion concerns. For reasons already explored, the same concerns support a broader exception to Garcetti’s reach.

2. Balancing harms and considering who decides

In addition to the government speech rationale, the Garcetti Court also relied on the notion of ordinary managerial discretion. That is, the idea that supervisors require leeway to evaluate the quality of employees’ work product. I refer to this as ordinary managerial discretion to distinguish it from the government speech rationale. The latter assumes a supervisorial prerogative to dictate employees’ speech content. The former is considerably less aggressive, envisioning a managerial right not to script employees’ work product speech but to evaluate its quality.

Ordinary managerial discretion is entirely appropriate, and I know of no serious objections to Garcetti that say otherwise. The dissenters in Garcetti themselves took the position that any managerial concerns can be addressed under the “Connick-Pickering balance test” that courts ordinarily apply to review public employees’ free speech claims. Under that test, courts ask whether the speech at issue involves a matter of public concern. If it does not, then there is no First Amendment protection. If it does involve a matter of public concern, then courts balance employees’ free speech interests against the “interests of the state, as an employer, in promoting the efficiency of the public services it performs through its employees.”

As I detailed in earlier work, I urge courts to take a somewhat different approach to assess ordinary managerial discretion arguments. Specifically, “[j]udicial review in the work product context can and should be designed not to second-guess supervisor assessments of work product quality, but to smoke out retaliation against work product speech for reasons other than quality.” Under this approach, for example, “a government scientist’s superior would be free to discipline her for sloppy research methods or poorly written reports.” On the other hand, it would not “constitute a work quality-based decision were a government scientist’s supervisors to discipline her for reaching scientific conclusions in tension with an administration’s policy agenda.” Courts are well equipped to manage such inquiries. They “have considerable experience conducting inquiries designed to smoke out illegitimate decision-making bases and to distinguish them from permissible rationales.”

Improper political interference may take the form not only of individual disciplinary decisions but of broader policies and structures. This brings us back to USAGM. The plaintiffs’ First Amendment claims stemmed from a series of actions, including some directed at individuals. They all centered, however, on an alleged structural failing—specifically, defendants’ breach of the firewall between the networks’ day-to-day journalistic operations and upper-level management. One firewall-breach claim, for example, entailed “defendants’ requests to participate in news coverage meetings and efforts to directly oversee journalists’ assignments.”

I raise this aspect of USAGM to highlight two points. First, because the firewall-breach claims concerned “‘generally applicable’ policies and practices,” Judge Howell applied a somewhat heightened version of the Connick-Pickering test, asking “whether the restrictions defendants have allegedly imposed on [plaintiffs’] speech are no more restrictive than ‘reasonably necessary to protect’ various government interests.” This higher scrutiny level originated in the 1995 Supreme Court case of United States v. National Treasury Employees Union (NTEU). The NTEU Court suggested that the “widespread impact” of the challenged statutory provision gave “rise to far more serious concerns than could any single supervisory decision. In addition, unlike an adverse action taken in response to actual speech,” a general restriction can “chill[] potential speech before it happens.” It makes good sense to apply this somewhat tougher and more forward-looking approach to general policies and practices for the reasons described by the NTEU Court. Indeed, one can see the wider perspective to which such scrutiny lends itself in Judge Howell’s observation that defendants’ efforts created a chilling effect. As she put it, “[i]n response to the knowledge that defendants are monitoring VOA and network coverage and reaching out to individual journalists directly with questions about coverage, journalists and editors have already refrained from engaging in certain speech and are likely to continue doing so.”

Second, the firewall breaches illustrate how decision-makers’ identities themselves can evidence improper political interference. In the case of USAGM, there is no single, correct resolution to each of the myriad coverage, editing, and other journalistic decisions that must be made on a day-to-day basis. Yet political distortion was evidenced by high-level management’s intrusion into the daily processes through which those decisions are made.

III. Broader Legal, Political, and Social Contexts

A. Legislative protections for knowledge production, and the threat posed by unitary executive theory

Though necessary, First Amendment protections alone do not suffice to safeguard public knowledge production. Legislation that fosters expertise-driven public knowledge production and shields it from undue political influence is essential as well. Indeed, the merit-based civil service itself is a creation of the political branches. As Jon Michaels writes, the 1883 Pendleton Act “marked a sharp turn away from premodern public employment, sweeping away the longstanding spoils system.” In the many years since its passage, subsequent “Congresses and presidents [have] extend[ed] coverage to more and more federal employees and [have] thicken[ed] the protections enjoyed by all civil servants.” For example, civil service laws today protect federal employees who blow the whistle on certain workplace problems including lawbreaking, gross mismanagement, and abuses of authority.

Civil service protections are especially crucial in the face of judicial precedent narrowing the scope of the First Amendment’s safeguards. In addition to substantive limits, the Supreme Court for decades has barred federal employees from bringing damages actions against their employers for infringing their First Amendment rights. In the 1983 case of Bush v. Lucas, the Court held that the CSRA—which today includes the amendments of the Whistleblower Protection Act and the Whistleblower Protection Enhancement Act, among other statutes—is the exclusive vehicle for federal employees to complain of First Amendment violations by their employers. Given this precedent, Judge Howell deemed it a close question whether the CSRA precluded the USAGM plaintiffs’ First Amendment claims. Howell allowed the claims to proceed, however, because they alleged general changes in plaintiffs’ working environment, rather than targeted acts of retaliation like those covered by CSRA.

Federal knowledge producers’ statutory safeguards—including civil service laws—themselves are increasingly at risk of judicial invalidation. One of the most significant threats is the Supreme Court’s widening embrace of unitary executive theory. As I have detailed elsewhere, unity is a “decades-long project of the conservative legal movement” that “posits that the President must control all discretionary activity within the executive branch.” Taken to its logical conclusion, unity could obliterate tenure protections for knowledge producers and for those charged to safeguard their independence. Such threats currently loom over civil servants, inspectors general, and members of the Office of Special Counsel and the Merit Systems Protection Board. The latter two groups are tasked with reviewing federal employees’ CSRA complaints.

Although an extensive discussion of unity is beyond this essay’s scope, four brief points are in order. First, unitary executive theorists—including justices who embrace unity—rely partly on arguments from accountability. For example, writing for a majority of the Court in the 2020 case of Seila Law LLC v. Consumer Financial Protection, Chief Justice Roberts touted the unique political accountability of the presidential office: “Only the President (along with the Vice President) is elected by the entire Nation.” Roberts cautioned, however, that the people cannot hold the president to account for that which he cannot control. Accountability thus is undermined by checks on presidential power in the administrative state, including “for cause” removal restrictions.

Second, this pro-unity argument relies on a thin, formalistic vision of accountability. As a result, it overlooks unity’s capacity to undermine meaningful accountability by enabling the president and his allies to manipulate the information by which the public can judge them. As I have put it elsewhere:

[E]xecutive branch accountability has both formal and substantive components. Formally, the people and the other branches must have mechanisms to respond to executive branch successes and failures alike. Substantively, the public and the other branches must have the means to discover such misdeeds in the first place. Unity is partly responsive to formal accountability, concentrating power in one nationally elected and highly visible figure who is subject to one re-election opportunity and to the possibility of impeachment. Yet a categorical unity directive can gravely damage accountability’s substantive aspects. It can do so by enhancing the President’s ability, directly or through subordinates, to shield or manipulate the very information against which the public and the other branches may judge his actions.

A key means through which unity can undermine accountability is by undermining knowledge producers’ independence from partisan political pressures.

Third, there is a striking parallel between the accountability-based arguments of unitary executive theorists and those made to advance broad applications of government speech doctrine. Recall, for example, Justice Alito’s assurance that “the democratic electoral process … first and foremost provides a check on government speech.” As we have seen, however, democratic checks do not work where the government’s impact on speech is not transparent. In the case of speech that purportedly stems from expertise, nontransparent government pressure can have a distorting effect that undermines accountability.

Fourth and finally, although the Supreme Court has increasingly embraced unitary executive theory over the past few years, it can and should refrain from extending this line of precedent further. Given the many flaws in unity’s underlying reasoning —including but not limited to the accountability-related problems just referenced—the Court would be well advised to exercise restraint going forward. At minimum, the Court should defer to legislative choices to protect public knowledge producers. As I detail in another recent paper, it is possible for the Court to do so within the confines of existing precedent.

B. Cultural and political hostility to knowledge production, and parallels in legal reasoning

A certain populist sensibility underlies the Supreme Court’s accountability-based justifications in both the unitary executive theory and government speech contexts. Each justification is premised, first, on the assumption that control by elected officials—whether of public employees or of publicly funded speech—ultimately equals control by the people. Second, each treats popular will as the appropriate determinant of the speech output in question, including public knowledge production.

The implications of these premises, particularly the second, is starkly illustrated by a 1988 memorandum opinion by the Justice Department’s Office of Legal Counsel (OLC). The opinion took the position that Congress could not constitutionally require the director of the Centers for Disease Control and Prevention to mail “AIDS information flyers” to the public “without necessary clearance” by the president. OLC deemed its conclusion to follow from the unitary structure of the executive branch, which itself stemmed partly from the founders’ desire to ensure politically accountable governance. OLC explained that the “highly scientific” nature of the information in question had no bearing on its analysis. Rather, it stressed, the “President’s supervisory authority encompasses all of the activities of his executive branch subordinates, whether those activities be technical or non-technical in nature.” Similar consequences follow when Garcetti and government speech doctrine are applied to permit control by elected officials of public knowledge production.

The notion that “the people” have a right to determine the content of publicly subsidized speech, including that which purports to reflect disciplinary expertise, plays an influential role in American culture and politics as well as in the law. One contemporary manifestation of this view can be seen in right-wing attacks on the so-called “deep state,” which generally refers to career bureaucrats in nonpartisan, expertise-driven roles. It is no coincidence that the same politicians who frequently invoke the deep state also use the term “fake news” to describe reported information that reflects poorly on them or their political allies. In wielding these terms, politicians convey the notion that “truth” is a product of power and loyalty. From this perspective, public knowledge production is among the spoils of political power. There is little if any place, in this view, for publicly supported speech that reflects disciplinary expertise, shielded from partisan politics.

Suspicion of expertise is nothing new in American life, nor is it invariably a negative force. To the extent that American law and culture reflect epistemic humility—including the notions that even the loftiest are fallible, that leaders must be kept in check by the people, and that individuals should recognize the limits of their own perspectives—that is largely a good thing, in my view. Trouble follows, however, when epistemic humility shades into epistemic relativism, or the idea that “any idea or claim is just as good as any other.” Epistemic relativism, in turn, lends itself to the notion that truth is always up for grabs, the mere spoils of political conquest.

C. A final example: The critical race theory bogeyman

1. Cultural and political aspects of the controversy

In political discourse, epistemic relativism often takes the form of complaints to the effect that experts—or “elites” more broadly—are attempting to indoctrinate others. Those making the complaints sometimes pair them with calls to regulate the speech of the offending elites. There is tension between these positions, of course—that is, between championing free speech and advocating using the power of the state to control speech content. Yet they are consistent in two respects. First, each is compatible with the notion that there is no objective reality that matters; there is only power. From that perspective, and as elaborated in the previous section, truth is rightfully determined by those with political power. Second, each position makes sense as part of an appeal to populism—that is, as a piece of a larger narrative that pits elite experts against “real Americans.” Historically, populist appeals against academic elites and other disciplinary experts have had a right-wing bent in the United States. This is reflected in much political discourse today, including in assertions of an anti-Trump deep state and in rampant, false claims of a fraudulent 2020 presidential election that could not possibly have been won by anyone other than Donald Trump, the true voice of the people.

Perhaps the most consistent target of right-wing populist outcries in the United States is public higher education. The stated concerns typically include some variation on the theme that universities are hotbeds of left-wing radicalism that seek to brainwash their students. These worries are often framed as fears that universities are trampling free speech through “political correctness” or “cancel culture.” Though couched as anti-authoritarian in nature, these concerns often take the form of calls for the state to control or retaliate against curricular choices, faculty hiring, or pedagogical practices at state universities.

This phenomemon is far from new. As I wrote in 2017: 

For much of the past century, universities have been among the major targets in culture wars between so-called liberal elites and conservatives invoking populist rhetoric. … Teachers and administrations have been accused in multiple forums over the years—from the congressional hearings of the 1950s to the online watchdog groups of today —of threatening free speech and free thought by indoctrinating students. Yet such concerns themselves can lead to actions—from the loyalty oath requirements of the mid-twentieth century to present-day state legislative responses to campus protests —that threaten academic freedom.

In our current moment, calls to regulate higher education in the name of freedom have become a near-constant drumbeat. Particularly ubiquitous is a growing furor over “critical race theory.” CRT, which arose in American law schools in the 1970s and 1980s to examine systemic racism in American law and society, is taught predominantly in elective law school and graduate school courses. It had been relatively obscure until conservative activists, by their own admission, began a deliberate campaign to energize prospective voters by convincing them that CRT is being taught pervasively in grade school as well as in colleges and universities, and that it indoctrinates students into a divisive, far-left ideology. CRT’s antagonists depict its content and prevalence in ways that range from incoherent to false; nonetheless, their efforts have borne fruit. Legislation and other official actions targeting CRT—or taking aim at other topics that activists have conflated with CRT (hereinafter “educational gag orders”)—have been enacted or proposed in multiple states. Although most of these measures initially were directed at K-12 public education, they increasingly target public higher education as well. According to the American Association of University Professors (AAUP), eight states passed educational gag order laws in 2021, with two targeting higher education. As of June 2022, AAUP reports that it has identified “over 120 educational gag orders in state legislatures, including 47 that explicitly target higher education.”

A few examples help to illustrate the scope of the various bills. Oklahoma Senate Bill 1141 would bar any “institution within the Oklahoma State System of Higher Education” from requiring, either as a general education requirement or as a “core requirement for a degree program,” enrollment “in a course … that addresses any form of gender, sexual, or racial diversity, equality, or inclusion curriculum.” South Carolina House Bill (HB) 4605, applicable to public and private colleges alike that receive any state funding or tax benefits, would bar schools from “subject[ing] individuals to, or require completion of instruction, presentations, [or] … discussions” affirming or promoting particular concepts, including the existence of “unconscious or implicit bias” or the desirability of affirmative action, even where the “instruction, presentation [or discussion] … is part of a lesson, assigned or suggested materials made available in any format or setting.”

Several bills fixate on The New York Times’ 1619 Project. As PEN America summarized in February 2022:

Alaska’s HB 228 and Oklahoma’s HB 2988 would ban the use of the 1619 Project as a college classroom assignment; New York’s A[ssembly]B[ill] 8253 would ban it as a required assignment, even when paired with competing perspectives; Iowa’s H[ouse]File] 222 would ban the 1619 Project “or any similarly developed curriculum.” Missouri’s HB 1634 would ban seven different curricula, including the 1619 Project, the Southern Poverty Law Center’s Learning for Justice Curriculum, and the Zinn Education Project. Students training to be teachers would not be able to examine or discuss these curricula in a college classroom setting.

A striking aspect of educational gag order campaigns and related public discourse is how much they reflect right-wing populism and the notion that political power, rather than expertise or a search for truth, should determine the content of public education. As Nate Hochman put it in the Manhattan Institute’s City Journal, “[w]hat one teaches and does not teach in [the public school] setting is an unavoidably political decision, using public funds to favor certain concepts and theories at the necessary exclusion of others.” Christopher Rufo, the person most directly responsible for igniting the nationwide anti-CRT frenzy, took Hochman’s point still further: He deemed it “totalitarian” to apply First Amendment protections to anti-CRT bills. Writing in City Journal, Rufo explained that those who see anti-CRT campaigns as matters of free speech “fail to understand the basic premise of our constitutional rights.” He went on:

The First Amendment protects the right of individuals to speak freely … the government itself … does not have a First Amendment right as such. This is the proper context for evaluating critical race theory, which is an almost entirely government-created and government-sponsored ideology, developed in public and publicly-subsidized universities, formulated into policy by public bureaucracies, and transmitted to children in the public school system. The critical race theorists and their enablers … want the right to enshrine their personal ideology as official state dogma. They prioritize the “freedom of the state” over the “freedom of the individual”—the prelude, whether deliberate or accidental, to any totalitarian system.

The rhetoric of the anti-CRT movement also relies on a populist division between the people—those (presumably white) students whom activists imagine will feel guilt or offense at learning of structural racism and its role in America’s past and present—and all others, including those whose experiences and ancestral histories might be illuminated by challenged materials and teachings. For example, several of the educational gag order bills, as well as the Trump executive order on which many of them are based, target teachings and materials that might cause students to feel guilt or discomfort. Similarly, a common refrain in anti-CRT rhetoric is that offending lessons or materials are oppressive, even discriminatory against white students. From these accusations, it is a short leap to the notion—also trumpeted by anti-CRT activists—that educational gag orders do not infringe on freedom. To the contrary, they positively advance it.

2. First Amendment implications

The First Amendment principles and doctrinal tools explored in Part II map on to anti-CRT controversies in a few ways. First, where educational gag order laws target publicly funded higher education, Velasquez and the limited public forum cases provide the right starting point. That is, such laws ought not to pass muster where they are viewpoint-based or would distort the nature of higher education. When the government funds higher education, it implicitly claims to support an enterprise that is largely self-governed by disciplinary experts on matters of academic merit, pedagogical method, and curriculum. When elected officials supplant such decision-making through legislative mandates—for example, by prohibiting “‘concepts related to … racial diversity, equality, or inclusion’” in the classroom or micromanaging how professors may teach the concept of slavery —they distort the very nature of academic pursuits. Specifically, they seek to use the university—and public expectations that its scholarly and curricular output will reflect disciplinary expertise—to launder political messaging.

A more complicated situation arises where restrictions on teaching or research come not from legislatures but from university administrations—say, in the form of professional consequences for poorly received pedagogy or scholarship. This scenario pits the academic freedom of professorial knowledge producers against the academic freedom of their knowledge producer institutions. Yet here, too, the doctrinal tools explored in Part II point us toward a solution. They call for a limited type of judicial review—one “designed not to second-guess supervisor assessments of work product quality, but to smoke out retaliation against work product speech for reasons other than quality.” This is similar to the approach advocated by Paul Horwitz, who suggests that courts should defer to universities’ disciplinary decisions, so long as they are “‘genuinely academic decision[s].’” Drawing on due process case law involving public universities, Horwitz explains that “an academic decision is a professionally arrived at academic judgment that can only be judicially overridden if the university has engaged in such a ‘substantial departure from accepted academic norms as to demonstrate that the person or committee responsible did not actually exercise professional judgment.’”

This analysis brings us full circle, back to our initial look at Garcetti and the potential academic exception that it raises. As I observed in that discussion, Garcetti’s academic exception holds promise, but it would be more principled, and potentially more robust, if it were framed as part of a broader anti-distortion rule. There is also a second reason why an anti-distortion rule is a more satisfying approach than an exception based on academic freedom alone. Specifically, an exception grounded in the nature of academic freedom could be invoked as readily to defend as to protest Garcetti’s application to professorial speech. Academic freedom, the argument would go, demands special deference not to individual professors but to academic institutions and their decision-making. An anti-distortion approach, on the other hand, alerts us to the dangers of political manipulation posed to professorial and institutional speech alike. It also lends itself to the doctrinal approach outlined above, one that protects institutional autonomy for legitimate academic decision-making, while remaining on guard against judgments that are mere pretexts for distortion.

Conclusion

These are dangerous times. According to polls, more than half of the members of one of our two major political parties continue to believe that the last presidential election—an election uniformly hailed as free, fair, and secure by election security experts—was fraudulent and that President Biden is not the rightful president. The ongoing COVID pandemic, too, has revealed the susceptibility of so many Americans to scientific misinformation and disinformation. Conspiracy theories abound regarding public schools as well, ranging from lies about tragic school shootings to fearmongering about curricular choices and library books.

Public knowledge production can serve as a crucial counterweight against such phenomena, supporting a shared epistemic base and fostering the critical thinking skills necessary to interpret, challenge, and build on that base. Yet public knowledge production itself is under siege by some of the same cultural and political forces that support misinformation and disinformation. In this push and pull of events, First Amendment law plays an important but complicated role. Aspects of the law bolster the politicization of public knowledge production. But there are important countertendencies in the doctrine as well, parts of the case law that evince an understanding of and protectiveness toward public knowledge production. These salutary pieces of our legal tradition are among the tools that we must draw upon as we strive to shore up public knowledge production and to forge a path out of the epistemic crisis that envelops us today.

 

I am very grateful to the Knight First Amendment Institute at Columbia University for holding the Lies, Free Speech, and the Law symposium for which I wrote this piece, and especially grateful to Katy Glenn Bass and Genevieve Lakier for inviting me to participate. For their thoughtful comments and questions, I am also indebted to Floyd Abrams, Jack Balkin, Mark Blitz, Joe Blocher, Erin Carroll, Alan Chen, Caroline Mala Corbin, Claudia Haupt, Vicki Jackson, Cort Kenney, Christina Koningisor, Francesca Procaccini, Tyler Valeska, and other participants at the Knight Institute symposium and at the 2022 Freedom of Expression Scholars Conference at Yale Law School.

 

© 2022, Heidi Kitrosser.

 

Cite as: Heidi Kitrosser, Protecting Public Knowledge Producers, 22-17 Knight First Amend. Inst. (Dec. 16, 2022), https://knightcolumbia.org/content/protecting-public-knowledge-producers [https://perma.cc/QY2V-RG52].