Thirty years ago, the Supreme Court created the government-speech doctrine to protect certain forms of government action against First Amendment challenges. The government must speak to govern, the thinking goes, and the First Amendment regulates private rather than governmental speech. Therefore, government action that is classified as “government speech” should not be vulnerable to free-speech claims. The doctrine has since been used accordingly, vaccinating certain government programs against the First Amendment. For example, the doctrine has insulated regulations banning federally funded medical providers from counseling patients about abortion; advertising campaigns spearheaded by the federal government; state-level choices about which specialty license plates to approve; and municipal decisions about which privately funded statues to install in public parks. The doctrine now features in dozens of decisions in federal and state courts, and the Supreme Court continues to deploy it—most recently in May 2022.
Commentators have complained about the government-speech doctrine for years. They have been especially critical of the Court’s use of the doctrine, which appears unnecessary at times and inconsistent or even unprincipled at others. And they have offered a range of suggested reforms meant to steady the ship, such as the adoption of certain transparency requirements for government speech. This Article offers the deepest, most sweeping critique yet of the government-speech doctrine: the doctrine cannot be saved. It is intrinsically unconstitutional, and it should be eradicated in its entirety. More specifically, this Article argues that the government-speech doctrine is anchored in a conceptual mistake: it is not that the government can sidestep its burdens under the First Amendment whenever it communicates, but rather that the benefits of the First Amendment do not extend to government communication. Moreover, although governments communicate ceaselessly, there is no such thing as government speech, and doing away with that fiction will clarify First Amendment jurisprudence considerably. Instead, the Article outlines a more elegant replacement for the government-speech doctrine—and one that remains true to the First Amendment: the recognition of a new type of forum for governmental communication that subsumes the traditional, limited, and nonpublic forums the Court has long recognized.
Introduction
Over the last thirty years, the Supreme Court has created and augmented a significant exception to the First Amendment—an exception this Article argues is inherently unconstitutional and doctrinally unsustainable. The Court started down this path by recognizing that its typical mechanism for handling the government’s suppression of private speech1 was inadequate.2 In previous instances where the government restricted private expression, the Court developed an approach for assessing whether the government had violated the First Amendment.3 The Court would begin by asking where this discrimination took place: a traditional public forum that historically served as a venue for robust private expression, like a public park;4 a designated public forum that the government elected to open for modes of private expression, like a municipal theater;5 or a nonpublic forum that lacked a history of being used (by private citizens) or opened (by the government) for private expression, like a polling place on election day.6 After categorizing the venue in question, the Court would then select some corresponding level of scrutiny for assessing whether the government’s actions violated the First Amendment.7
But in the early 1990s, the Court began to confront the possibility that sometimes the government impinges on private expression based on content or viewpoint8 through the unavoidable act of endorsing some policies or propositions over others.9 The government must speak to govern, the Justices have told us, so we must treat the government’s own speech differently.10 The Court began working this conclusion into its First Amendment jurisprudence in 1991, in Rust v. Sullivan, when it upheld a gag rule that conditioned the distribution of certain federal funds to medical providers on their refusal to counsel patients on abortion.11 Although Rust did not explicitly introduce the term “government-speech doctrine,”12 the Court has since made clear that the rationale behind its decisions rested on the special status of government speech.13 In essence, the Court attributed the silence that the Reagan administration demanded of medical providers to the government rather than to the medical providers.14 The Court invoked the necessity of government speech as the basis of its analysis.15
In the years since, the Court has built up its jurisprudence around the government-speech doctrine. The Court has extended the reach of the doctrine to cover, inter alia, compelled subsidies tied to certain television advertising schemes promoted by the federal government;16 state decisions concerning the approval of specialty license plates;17 municipal determinations about the installation of privately funded statues in public parks;18 and, arguably, the statements of government employees made in their professional capacities.19 Meanwhile, the Court has rejected the doctrine’s applicability to trademarks,20 a Rust-style gag rule for attorneys,21 and the flying of flags on one—but probably only one—of three flagpoles standing in front of Boston’s City Hall.22
The doctrine has proved unpopular with scholars, who question its completeness,23 consistency in application,24 and implications.25 Commentators have also tried to connect the dots left by the Court to glean the parameters of the doctrine using the various decisions that accept or reject it.26 This Article argues that such efforts are unavailing. The criticisms of the doctrine offered to date are fine as far as they go, but they are also needlessly glancing. Attempting to cobble together a consistent understanding of the doctrine only obscures its deeper theoretical incoherence. Although some commentators and Justices have noticed that the doctrine threatens to become hopelessly overbroad,27 they have not followed that observation to its natural conclusion: the doctrine simply cannot be reconciled with traditional First Amendment jurisprudence, and it must be discarded entirely.28
Part I of this Article unpacks the argument behind the government-speech doctrine and argues that the doctrine rests on a conceptual mistake. The Court anchors the government-speech doctrine in the conclusion that a key part of the First Amendment—the Free Speech Clause—does not apply to government speech.29 There are two salient ways of interpreting that seemingly obvious proposition. The first holds that, when the government communicates, it is exempted from the burdens of protecting certain First Amendment rights of the public (burden exemption). The second holds that, when the government communicates, it does not retain the benefits of First Amendment protections (benefit exemption). The Court has held that burden exemption serves as the foundation for the government-speech doctrine, but Part I argues that only benefit exemption is defensible.
Part II demonstrates just how pernicious burden exemption can be by showing that it is inherently unconstitutional. Although the Court has applied burden exemption uniquely in the context of government speech, there is no principled basis for that restriction. Burden exemption menaces individual rights both within the First Amendment context and beyond. Crucially, there is no workable principle that restrains the government-speech doctrine—and therefore burden exemption—from reaching the most quintessential form of governmental communication: the law itself. Part II also argues that the Court’s partial recognition of the threat of burden exemption and its ad hoc attempts to neutralize that threat thus far are inadequate.
Part III considers and rejects the two most limited responses to the threat of burden exemption: sustaining the government-speech doctrine with ad hoc exceptions (as the Court has done for three decades) or finding a principled basis for defining a narrower version of the doctrine. Finally, Part IV reveals the promise of a third option: replacing the government-speech doctrine outright. In fact, Part IV argues that there is no such thing as “government speech” at all, and abandoning that notion considerably clarifies First Amendment jurisprudence. Part IV closes by sketching out a replacement for the government-speech doctrine: the recognition of a new type of forum30 that subsumes traditional forum analysis and allows the government to abandon viewpoint neutrality and (at times) suppress counter-messaging from private citizens. Part IV concludes by offering basic rules within that forum to balance the government’s need to communicate with core First Amendment requirements.
I. Picking Apart the Argument for the Government-Speech Doctrine
Let us begin with the elephant in the room: how could the government-speech doctrine have persisted for the past three decades if it is inherently unconstitutional? The shortest answer is that the Court felt it had no choice but to introduce the doctrine when traditional forum analysis proved inadequate for analyzing certain types of cases.31 The more detailed answer is that a simple, superficially plausible, but ultimately invalid argument renders the doctrine enticing.32
A. Identifying the Argument
Embedded in the Court’s government-speech-doctrine jurisprudence is an argument with three premises. First, the government must speak to govern; it must endorse certain propositions and, in doing so, abandon viewpoint neutrality.33 It must also, at certain times and in certain places, suppress counter-messaging from private citizens, thereby discriminating based on viewpoint.34 As the Court has put it, “A government entity has the right to ‘speak for itself.’ ‘[I]t is entitled to say what it wishes,’ and to select the views that it wants to express. Indeed, it is not easy to imagine how government could function if it lacked this freedom.”35 We may call this the “Necessity Premise”: as a matter of necessity, we must make some allowances for government speech, for there is no practical alternative.
Second, all speech may be classified on a binary basis as emanating either from a governmental source or from a private one.36 We can refer to this as the “Exclusivity Premise.” One function of this exclusive approach to categorization is that the Court classifies some speech as governmental despite acknowledging its First Amendment implications for private citizens.37
Third, and finally, the doctrine relies significantly on the notion that the Free Speech Clause of the First Amendment does not apply to the government’s own speech.38 As the Court has put it, “[t]he Free Speech Clause restricts government regulation of private speech; it does not regulate government speech.”39 This claim seems obvious in some sense, and it captures something important. The First Amendment protects private interests against certain forms of governmental interference; its application to the government’s own communications must be different in some way.40 We may call this the “Exemption Premise” because it holds that we must exempt government action from certain First Amendment considerations.
Taken at face value, this trio of propositions would seem to suggest that important parts of the First Amendment should be set aside in the common and necessary scenarios where the government is speaking. In summary form, the argument runs something like the following:
- Necessity: Governance is impossible without allowances for government speech.
- Exclusivity: In any given instance, we can classify speech as emanating either from the government or from private parties.
- Exemption: The Free Speech Clause of the First Amendment does not apply to government speech.
Therefore, in those instances where we determine that the government is speaking, of necessity, we will set aside Free Speech Clause analysis altogether.
B. The Flaw in the Argument
The conclusion of the argument above takes us to absurd places. Fortunately, the conclusion does not actually follow from the premises.41 Take each premise in turn.
The Necessity Premise is perhaps the simplest, and it is also the most difficult to dispute. Although we may disagree about the extent of this necessity, it is genuinely hard to see how the government could operate without making some factual and value-laden endorsements at the expense of competing alternatives. I accept the Necessity Premise, and I am aware of nobody who rejects it. The main trouble with the Necessity Premise is that it pressures us toward an incorrect interpretation of the Exemption Premise.42 We may have no choice but to grant the necessity of certain forms of government viewpoint discrimination, but we should note that the Necessity Premise alone does not necessarily compel the government-speech doctrine specifically.
The second claim, the Exclusivity Premise, is neither here nor there. On one hand, it is arguably too strong. Some scholars have urged the recognition of “mixed” speech that is simultaneously private and governmental in material respects.43 Perhaps specialty license plates represent mixed speech because they bear a semi-tailored message approved by the state, yet are specifically selected from a range of options by private motorists.44 The Court’s embrace of the Exclusivity Premise forecloses the possibility of acknowledging mixed speech as such.45
But a weaker and less controversial version of the Exclusivity Premise would still function as a reasonable substitute in the argument for the government-speech doctrine. Even advocates of mixed speech should acknowledge that at least some communications emanate exclusively or entirely from the government. In other words, even if we recognize three categories of speech rather than two (namely, private, governmental, and mixed), we should be prepared to acknowledge that some speech will populate each category. A world with mixed speech does not necessarily foreclose purely governmental speech.46 Additionally, the primary virtue of acknowledging mixed speech is that doing so may facilitate a more nuanced or protective First Amendment jurisprudence.47 If we adopt the Exclusivity Premise but create a properly tailored government-speech doctrine that protects the private speech interests that arise in cases that could plausibly be regarded as mixed, we would arrive at substantially the same place.48 These distinctions serve a doctrinal function, and we can use them however we like. The Exclusivity Premise is not the source of the problem.
The third claim, the Exemption Premise, is the most important. What does it mean to say that the First Amendment (or, alternatively, the Free Speech Clause of the First Amendment) does not apply to the government?49 There are two salient interpretations of this claim: either the government is free from certain First Amendment burdens when it speaks (burden exemption), or it is free from certain First Amendment protections when it speaks (benefit exemption). In defending the government-speech doctrine, the Court does not even appear to recognize both options. Instead, it embraces burden exemption without hesitation or discussion. For instance, it has noted that “the Government’s own speech . . . is exempt from First Amendment scrutiny,”50 and that when “the State is speaking on its own behalf, the First Amendment strictures that attend the various types of government-established forums do not apply.”51 It has also endorsed the proposition that the “Government is not restrained by the First Amendment from controlling its own expression.”52
The pressure to interpret the Exemption Premise as exempting the government from First Amendment burdens flows naturally from the Necessity and Exclusivity Premises.53 If we think government speech is necessary and widespread, then we feel compelled to facilitate it by removing rather than imposing constitutional constraints when the government is speaking. And if we take the Court’s preferred view of the Exclusivity Premise, namely that speech sources are a binary matter, then government speech functionally precludes private speech. In cases where the government is speaking, then, it would seem there is no private person who could raise a First Amendment claim anyway. Thus, we might think that the Exemption Premise should mean the government is free from First Amendment burdens.
But that cannot possibly be right. Only benefit exemption can be correct, and benefit exemption gets us nowhere in justifying the government-speech doctrine.54 Constitutional rights are guaranteed to the people against the government, not vice versa.55 And there is no plausible way of reading the Establishment Clause as anything other than a First Amendment burden on the government.56 The more natural reading of the Exemption Premise is that the government’s own communications are not subject to First Amendment protection, not that the government can speak however it wants without worrying about harming the First Amendment interests of its citizens. The government does not need First Amendment benefits because those benefits would run against itself anyway. It is exempt from those benefits because they are not part of the text,57 superfluous, and conceptually ridiculous.58
By contrast, as the following Part illustrates, burden exemption conflicts with traditional First Amendment jurisprudence in numerous, irreconcilable ways. Burden exemption is quite different in principle from having a deferential standard of review for government action; it is quite literally a license to ignore constitutional limitations altogether.59 Moreover, the Court acknowledges that government speech frequently implicates the First Amendment rights of citizens,60 which means that burden exemption will necessarily result in the curtailment of constitutional rights. At minimum, before embracing burden exemption, the Court should have attempted to adopt clear and narrow parameters for what constitutes “government speech.” Yet the Court has struggled to do either of these things—presumably because, as I will argue below, they are impossible.
II. The Perils of Burden Exemption
The dangers of burden exemption flow, in large part, from the unavailability of a reasonable limiting principle: exactly how much government conduct is exempted from First Amendment burdens? To appreciate the difficulty of locating a reasonable limiting principle, we must confront the Court’s standard for identifying government speech.
A. What Qualifies as Government Speech?
To hear some Justices tell it, the Court has yet to provide a clear-cut test for government speech that reliably reveals when burden exemption applies.61 If true, that itself is remarkable for a thirty-year-old doctrine that is constitutionally minatory. To the extent that we have something like a test, it is also extremely broad: in holding that a state may reject citizen-nominated specialty license plates without infringing First Amendment rights, the Court found that a state’s decisions about whether to approve a design “are meant to convey and have the effect of conveying a government message, and they thus constitute government speech.”62 The Court drew this standard verbatim from an earlier decision, in which it upheld the power of a municipality to install some, but not other, privately funded statues in a public park.63 Noting well the Justices’ equivocation about whether they have ever completely articulated the test for government speech—as well as their view that the inquiry is “not mechanical” but rather “driven by [the] case’s context”64—I will still refer to this as the “test” because it is the closest articulation we have.
This test essentially comprises two elements. It captures action that (1) is intended to convey a government message (pursuant to substantial government control) and (2) has its intended effect. We might add a third criterion that factors in the historical significance of the form of government action in question,65 although that requirement may easily be understood to qualify the first two prongs by affecting our interpretation of the intentions or effects of any form of governmental activity. According to the Court, this test appears to capture most statements of governmental employees operating in their professional capacities, regardless of whether they are speaking to private citizens.66 The Court has also indicated that the test encompasses government speech that is open to multiple interpretations by civilian audiences—that is, speech that does not reliably convey any specific message endorsed by the government.67 Expanding the universe of government speech to garbled governmental communication arguably extends the doctrine beyond its original rationale, which is rooted in the necessity of the government endorsing specific messages.68
Even if the test we have is incomplete in some unspecified respect, its core structure is familiar in First Amendment jurisprudence.69 When applied to state action, however, this approach sweeps practically everything into the bucket of government speech. Much governance is simply communication, whether about the government’s own actions and plans, the state of the world, or what the populace should do and how we should do it.70 Governmental reports, proclamations, roadway signage, and curricular decisions by public educational institutions (if not all public educational instruction) are all obvious instances of government speech under this standard. More consequentially, the quintessential form of government speech would seem to be the law—the U.S. Constitution, state constitutions, statutes, administrative rules and regulations, judicial opinions, and so forth. Indeed, the government-speech doctrine is itself government speech on any plausible understanding of the term. Surely if license plate designs and statue selection have the intended effect of communicating a government message, the law does too.71 And if the law is government speech, then burden exemption becomes completely unsustainable.
B. The Logical End of the Government-Speech Doctrine
If the government must speak to govern,72 we can hardly be surprised to find that government speech is all around us. Taken to its logical conclusion, the doctrine would swallow the bulk of the First Amendment—or more. For example, absent an ad hoc exception introduced by the Court,73 the government-speech doctrine would wipe out the Establishment Clause—the very definition of a First Amendment burden on the government’s own speech.74 On its face, the Establishment Clause prohibits the government from enacting any law “respecting an establishment of religion,”75 and the Court has long held it to prevent the government from showing forms of favoritism toward certain religions76 or toward religion over nonreligion.77 Yet permitting governmental favoritism toward various propositions is the basis for establishing the doctrine in the first place.
The Court has only started to confront the perils of burden exemption.78 Consistently applied, the notion that the First Amendment does not apply when the government speaks simply precludes the possibility that a statute or regulation could ever violate the First Amendment. Presumably, the Court would not tolerate that result, but only ad hoc limitations can justify resisting it. Moreover, nothing about the structure of the government-speech doctrine—or the argument that leads to it—precludes the introduction of analogous doctrines that displace other constitutional rights. Why not apply burden exemption wherever the government has a need to engage in an activity that is analogous to a private activity protected as a matter of constitutional right? Yet I can locate no other context in which the government has carte blanche exemption from respecting such rights based on the presumption that the government’s own version of the rights in question trump our own.79 Consider the following examples of analogous doctrines in the Second and Fourth Amendment contexts.
Suppose a state encounters a gun shortage that makes it impossible to arm its police adequately, perhaps due to supply chain issues. The state enacts a statute preventing anyone within its jurisdiction from acquiring certain firearms until it deems the police adequately equipped with the enumerated makes and models. Aggrieved would-be gun owners sue to enjoin the statute as infringing on their rights under the Second Amendment. In its defense, the state invokes a new doctrine: the government-arms doctrine. It argues that the Second Amendment only protects the gun rights of private owners,80 but it claims that the government must also own firearms to govern effectively. Although private citizens may feel that the policy burdens their rights, this new statute ultimately derives from the government’s need to secure firearms for its law enforcement officials, which in turn flows from its broad police power.81 The state thus argues not that the ordinance survives the appropriate level of scrutiny under the Second Amendment,82 but rather that the Second Amendment simply does not apply.
The state’s argument is structurally identical to the arguments advanced under the government-speech doctrine: the government needs to deploy guns to govern (necessity); all things considered, it is the government’s interest in utilizing those guns that is primarily at issue here, despite its obvious Second Amendment implications (exclusivity); but the government’s own gun ownership is not burdened by the Second Amendment, so the measure evades Second Amendment scrutiny (burden exemption).83 Yet it is difficult to imagine a court approving this argument.84 If the state’s policy survives a constitutional challenge, it should be because it withstands some measure of constitutional scrutiny, not because we set aside constitutional scrutiny altogether. Arguably, the government-arms doctrine should capture an even more intrusive statute that confiscates all privately owned guns adequate to the task of equipping the state’s officers.85
Now imagine a Fourth Amendment variation. Suppose an intelligence agency finds that a set of highly classified documents has gone missing from one of its facilities. There is no evidence of a break-in, so agency officials surmise that the documents must have been taken by someone with access to the room where the documents were stored. Beyond that conclusion, the agency has no immediate leads, so it dispatches agents to search the homes of all fifty employees with access to the room in question. Several employees resist the searches on Fourth Amendment grounds, alleging that the agency has no probable cause to search their homes.86
The agency takes the following position: it is widely accepted that the government must be able to maintain secrecy around some of its operations, especially in national security contexts (necessity).87 The government effectively possesses a functional equivalent to the right to privacy that the Fourth Amendment provides for individual citizens. The agency therefore proposes the government-privacy doctrine: when government action reflects its acknowledged necessity to maintain the secrecy around certain of its operations, we set aside the manifest Fourth Amendment implications of its conduct (exclusivity), and Fourth Amendment inquiries need not apply (burden exemption). The agency therefore argues that it is not a question of whether its searches survive Fourth Amendment scrutiny;88 the Fourth Amendment is simply precluded by the government’s need to maintain its own privacy, to be secure in its “papers[] and effects.”89 It does not matter that private Fourth Amendment interests are implicated here; after all, in the government speech context, the Court acknowledges that exempted government activity may have First Amendment implications too.
Once again, the basis for the government-privacy doctrine parallels the basis for the government-speech doctrine. Necessity pushes us to make accommodations for the government’s right to classify information and protect it. Additionally, the searches undertaken by the agency derive from governmental privacy concerns rather than private ones. Should we not extend burden exemption to the context of the Fourth Amendment, at least under facts like these?
The mere possibility of such parallel doctrines underscores the risks of embracing burden exemption, but the absurd implications of burden exemption in fact reach even further. One oddity of the government-speech doctrine is that it only labels state action as government speech when that action arguably carries a First Amendment implication. In other words, the government invokes the doctrine only in response to First Amendment claims that it seeks to preempt.90
But why stop there? Our “test” for government speech is quite simply indifferent as to whether state action implicates First Amendment interests. Instead, as noted above, the Court tells us to consider whether the government intends to, and then actually does, communicate a governmental message. Sometimes the government will successfully seek to communicate a message in a manner that implicates First Amendment rights. But sometimes it will successfully seek to communicate a message that implicates Second or Fourth Amendment rights, or statutory rights, or no rights at all. As a matter of consistency, the hypothetical state and intelligence agency above should be able to invoke the government-speech doctrine to defend their actions.91 The government-speech doctrine tracks the necessity or inevitability of government speech, and the test for government speech is completely blind to the rights that government speech implicates.92 So why not recognize that the state and the agency both had to speak—to vindicate important state interests, no less—and, in doing so, they had to displace constitutional rights outside the First Amendment context? Once more, nothing about the doctrine itself or the argument behind it supplies the requisite limiting principle.
C. The Court’s Efforts at Limiting Burden Exemption
At some limited level, the Court is aware that the government-speech doctrine threatens to overwhelm our jurisprudence. Cases concerning the doctrine routinely advert to the doctrine’s propensity to metastasize. For instance, the Court has observed that, “while the government-speech doctrine is important—indeed, essential—it is a doctrine that is susceptible to dangerous misuse. . . . [If over-expanded], government could silence or muffle the expression of disfavored viewpoints. For this reason, we must exercise great caution before extending our government-speech precedents.”93 Justices writing for themselves or for minorities of the Court are especially prone to sounding the alarm about the unintended consequences of embracing the doctrine.94
The only reason the threat has not been fully realized is that the Court has imposed ad hoc restrictions on the use of the doctrine to prevent some of its most objectionable results.95 For example, as noted above, the Court has decided that the government-speech doctrine does not exempt government speech from scrutiny for Establishment Clause violations.96 Further, despite finding that a government prosecutor’s work product is essentially government speech, the Court has declared by fiat—again, it seems, of necessity—that this holding does not extend to “scholarship or teaching” at public educational institutions.97 The Court also takes pains to remind us that various other checks might also apply, whether legal or political. It has suggested (albeit vaguely) that certain constitutional and statutory provisions—aside from the Free Speech Clause—may constrain government speech,98 and so too may regulations and practice.99 And the Court reminds us that the government remains “accountable to the electorate” for its advocacy activities.100
The Court has also tried to read some of its government speech holdings narrowly to clip the doctrine’s wings even further. For instance, the Court distinguishes the government’s use of subsidies to control a government message from its funding of private speech,101 even in cases where the basis for that distinction is extremely thin, if present at all.102 The Court has even toyed with an internally inconsistent limitation on burden exemption, once suggesting that the Free Speech Clause still applies to restrict government speech when the government attempts to compel private citizens to express the government’s message.103 This suggestion betrays the Court’s discomfort with burden exemption, but it neither aligns with the cases nor makes sense to say government speech is exempt from the First Amendment only if it complies with the First Amendment.104
The impulse to introduce these restrictions derives from an obvious source. As we observed above, burden exemption is extremely dangerous, so it must be tightly controlled.105 If it cannot be sharply limited based on clear principles, then it must be limited in an unprincipled fashion. But that is hardly a satisfying result. Unprincipled restrictions that are introduced on an ad hoc, rolling basis render the doctrine vague, inelegant, and unpredictable at best. This is precisely why commentators find the doctrine so frustrating and inconsistently applied:106 the Court felt compelled to create the doctrine but has struggled to find a principled basis for restricting it.
III. Canvassing Our Options Vis-à-Vis the Government-Speech Doctrine
Three primary possibilities lie before us. First, we could proceed as we have been, allowing courts both to deploy the doctrine and to create exceptions as needed. A second option would be to improve the doctrine by installing new features, invoking principles, or making distinctions that the Court has avoided thus far. A third, more radical possibility is to replace the doctrine altogether. I ultimately endorse replacement, but let us consider each in turn.107
A. Staying the Course
Despite the foregoing criticisms of the government-speech doctrine—and of burden exemption in particular—perhaps the risks posed by the doctrine are relatively minor given the Court’s recognition that it needs to be contained. In any event, the single most likely outcome is for courts to continue deploying the doctrine as they have been, taking each case as it comes and fashioning a patchwork of additional limitations as new fact patterns present themselves.
This approach is suboptimal for multiple reasons. First, the doctrine has rendered an entire area of First Amendment jurisprudence needlessly confusing and unpredictable. Because it lacks a clear test for application,108 plausibly yields inconsistent holdings,109 and continuously invites new limitations,110 it seems to function less as a legal doctrine and more as a proxy for the collective judicial intuition of any given majority on the Court. As a result, it is ill-suited for guiding either private or governmental behavior.
Additionally, and for similar reasons, we can expect the doctrine to generate bad results and incorrect holdings. Just as the doctrine provides inadequate guidance to parties, it also provides inadequate guidance to judges.111 As discussed in more detail below, numerous government-speech cases have yielded questionable or clearly inappropriate outcomes,112 and we can expect more of that if the doctrine remains standing as is.
Finally, and in some respects most importantly, the doctrine enshrines burden exemption into our law even though burden exemption is fundamentally at odds with our approach to constitutional adjudication. At the highest level, discerning First Amendment violations has always been a two-step inquiry: the first question is whether the government is involved in the right way, for only the government bears the burdens of the First Amendment;113 the second question, which we reach only if government action is at issue, is whether the action infringes a protected First Amendment interest.114 The government-speech doctrine essentially negates the traditional understanding of how we go about enforcing the First Amendment by turning the first prong of that process on its head. Suddenly, the mere fact that government action exists forms the basis for rejecting First Amendment claims outright rather than proof that a litigant has cleared the first hurdle. Such a doctrine cannot coexist harmoniously with longstanding First Amendment precedent. It is intrinsically unconstitutional. Even those government-speech cases that yield a reasonable result—or a result that is defensible under other appropriate principles—are decided on the wrong grounds per a doctrine that is built around burden exemption.115
B. Refurbishing the Doctrine
Our remaining choice is between attempting to rescue the doctrine and replacing it outright. At least one sitting Justice has expressed significant skepticism about stare decisis,116 and six Justices have recently overturned popular, long-standing precedent.117 This Court may therefore prove unusually amenable to uprooting and replacing the doctrine wholesale. But it is reasonable to consider the viability of a less radical solution first—namely, what might be done to save the doctrine.
I argue above that the test for government speech, such as it is, threatens to capture so much government action that a principled application of the test would render the doctrine comically potent. Improving the doctrine would require substantially weakening it. In fact, there is a coherent and constitutional version of the doctrine, but it looks quite different from what the Court has created, and it is too narrow to serve as a functional replacement. It is the version of the doctrine that eschews burden exemption, along the lines of what Justice Alito proposes in his Shurtleff concurrence.118
Suppose a private litigant brings a First Amendment claim. The Court considers this claim and finds state action present—a prerequisite for any First Amendment claim, after all.119 But, despite giving no special weight to the government’s need to communicate, the Court also finds the absence of any First Amendment implications whatsoever.120 Thus, the challenge fails not because the state action survives the appropriate level of scrutiny for burdening First Amendment rights but because, upon consideration of the facts of the case, no First Amendment rights happen to be burdened in this situation at all. This version of the doctrine is plainly different from the actual doctrine, which purports to skip past the First Amendment screening exercise altogether.121
This variation on the government-speech doctrine raises fewer constitutional concerns, hangs together theoretically, and provides a reasonable basis for demarcating a distinctive subset of failed First Amendment claims: those that genuinely involve only government communication and no private free-speech rights.122 But it is too narrow to replace the current doctrine because it does not empower the government sufficiently to satisfy the consensus interpretation of the Necessity Premise. It would grant the government no special communicative prerogatives, and therefore it would not apply in most of the major government-speech cases that the Court has already decided.123 Indeed, the entire point of appealing to necessity is to create a doctrine that can override legitimate First Amendment interests; a version of the doctrine that happens to fit into the gaps already left by the First Amendment is a jurisprudential nullity. There would be no reason even to name such a doctrine.
If abandoning burden exemption renders the doctrine too narrow to serve its purpose, one might hope for some other principled limitations on the scope of burden exemption instead.124 Yet the Court’s inability to supply reasonable limiting principles is inauspicious, and I cannot locate a workable mechanism for segregating doctrine-eligible and doctrine-ineligible governmental communications. For example, one salient possibility is to distinguish “soft” government speech (such as the endorsement of some factual proposition by a politician) from “hard” government speech (such as the law).125 Slippery as those concepts may be—perhaps they occupy opposite ends of a spectrum rather than constituting a binary—readers might share an intuition that there is a meaningful difference between exempting from constitutional scrutiny nonbinding endorsements of propositions by government actors and exempting rules, regulations, statutes, and the like. At minimum, it seems inappropriate to exempt the law categorically from First Amendment—and possibly other constitutional126—strictures. Surely the Court would agree.127 There are various ways of articulating some version of this distinction; perhaps burden exemption should not apply to acts of legal significance, for instance.128
As appealing as this might seem at first blush, it will not work. Even leaving aside the possibility that “government speech often is a limitation . . . on private expression,”129 the foundational government-speech case, Rust v. Sullivan, sustained a regulation issued by the Secretary of Health and Human Services that denied certain medical providers federal funds if they counseled patients on abortion.130 The doctrine was literally born from a case that saved a federal regulation from a First Amendment challenge, a regulation restricting what doctors could say to their patients. To improve the doctrine without overturning its foundational case, we would need a more nuanced distinction that carves out only a subset of the law that qualifies for burden exemption. Perhaps the doctrine should apply to soft speech, as well as to hard speech that empowers the government to engage in soft speech; or perhaps it should apply to soft speech as well as to hard speech that carries certain penalties but not others. But even these variations are insufficiently nuanced because the Court is also reluctant to include all soft government speech within the scope of burden exemption—for instance, at least for now, refusing to extend the doctrine to teaching and scholarship by public educators.131 If there are sound principles that define these parameters, they are numerous and subtle at best.
Alternatively, perhaps improving the doctrine will require overturning some of the Court’s cases. For instance, perhaps Rust was a mistake because it saved hard rather than soft speech, and correcting that mistake clears the way.132 Yet that is no good either. Even if the distinction between “soft” and “hard” government speech were consistent with the Court’s cases,133 and despite its intuitive appeal, it is a completely artificial place to draw the line. There is no principled basis for exempting only “soft” government speech—and even then, not all of it134—from consistency with the First Amendment. After all, any reasonable conception of government speech must include the law—any law about any subject matter.
This conclusion works its way into our jurisprudence from multiple angles, including the holdings of the Court itself.135 For example, in the context of copyright law, the Court has created the “government edicts doctrine,” which establishes that “officials empowered to speak with the force of law cannot be the authors of—and therefore cannot copyright—the works they create in the course of their official duties.”136 An edict is simply a decree or authoritative order137—an important communication from the government to its people. Indeed, the rationale behind the government edicts doctrine is that “no one can own the law” because “‘[e]very citizen is presumed to know the law,’ and ‘it needs no argument to show . . . that all should have free access’ to its contents.”138 In other words, the law is an especially important message from the government that we expect citizens to receive and understand. It is the quintessential form of governmental communication; if burden exemption applies to government speech, it must attach to the law as well. No apparent principled distinction neutralizes that threat. If anything, a principled approach to the category of government speech expands the doctrine beyond what the Court has allowed through its ad hoc limitations.
IV. Replacing the Government-Speech Doctrine
Process of elimination compels us to consider the viability of locating a replacement for the government-speech doctrine. If we can find no workable replacement, we might have to stay the course. But this Part argues that there is a path forward that is practically and theoretically superior. To find that path, we should revisit the argument for the government-speech doctrine, now with a firm grasp on the problems with burden exemption.
A. Revisiting the Argument for the Government-Speech Doctrine
Part I of this Article presents three premises that seem to lead to the government-speech doctrine:
- Necessity: Governance is impossible without allowances for government speech.
- Exclusivity: In any given instance, we can classify speech as emanating either from the government or from private parties.
- Exemption: The Free Speech Clause of the First Amendment does not apply to government speech.
The foregoing analysis does nothing to weaken necessity, which forces us to accommodate government communication in our First Amendment jurisprudence. Additionally, even if we quibble with exclusivity, the weaker version discussed in Part I remains: at minimum, some communications emanate exclusively from a governmental source.139 But now it should be clear that “exemption” must mean benefit exemption rather than burden exemption. Burden exemption is both inherently unconstitutional and far too powerful. Unfortunately, benefit exemption does little to resolve the Court’s original government-speech conundrum, even as it reminds us of an important truth: whatever solution we locate must be compliant with the First Amendment.
Recall the puzzle before the Court: how to accommodate the fact that the government does, and must, endorse various propositions,140 often by clearing the field for its message and channeling that message through private citizens or private media.141 In other words, the Court needed a way to account for the inevitable failure of the government always and forever to remain viewpoint neutral in how its conduct affects private speech. That is a First Amendment problem in the sense that the Court has made viewpoint neutrality an important concept in First Amendment jurisprudence,142 but it is not a First Amendment problem per se: there is nothing in the First Amendment that demands the government itself remain viewpoint neutral always, everywhere, come what may.143 We need not actually exempt the government from First Amendment scrutiny when it speaks.144 We only need to mold First Amendment jurisprudence around the communicative demands of governance. More specifically, we need to relax the Court’s historic insistence that restrictions on private expression remain viewpoint neutral, but only in contexts where such restrictions follow directly from the government’s own (appropriate) expressive activity.
The better way forward has been in front of us from the start. Viewpoint neutrality operates in the background of forum analysis—common-law restrictions on when the government can favor certain private speakers over others.145 Recall that the limitations of forum analysis generated the pressure behind the government-speech doctrine in the first place.146 The best way of accommodating governmental communicative prerogatives is to develop a new sort of forum for situations where the government holds forth.147 For one, as I argue below, subsuming governmental communication within forum analysis should provide clearer results and guidelines than burden exemption, which (as we saw above) is necessarily riddled with numerous, unpredictable, ad hoc exceptions. Second, conceptualizing government communication within the context of a new forum will also be less threatening to First Amendment rights.148 And third, extending forum analysis to cover governmental communication stands to harmonize the Court’s recent jurisprudence in the area; it is simpler, more intuitive, and more elegant.149
A forum framing is therefore surprisingly helpful. I suggest two steps in operationalizing it. First, we need to create some distance between private speech and government speech as a constitutional, jurisprudential matter. That is, we need to get rid of the idea that government speech resembles private speech sufficiently to invite anything approaching analogous treatment for constitutional purposes.150 Second, we should start to sketch out the rules that govern the new type of forum. I address each step below.
B. Eliminating the Fiction of “Government Speech”
By reproducing the structure of the government-speech doctrine, the government-arms doctrine and the government-privacy doctrine proposed above151 highlight a critical misstep. All three doctrines characterize certain government action—communicating, wielding firearms, or protecting sensitive information—as meaningfully comparable to constitutionally protected private conduct. That characterization opens the door to mistakenly accepting burden exemption: the Constitution applies when private citizens do these things, but it lapses into silence when the government does them. I have argued above that the government-speech doctrine is both inherently unconstitutional and wildly excessive. But there is a third basis for rejecting it: state action is qualitatively different from private action. In other words, the government does not speak at all.152
Traditional scholarly engagements miss what is distinctive about government “speech,” and, in doing so, they facilitate acceptance of the government-speech doctrine. More specifically, scholars recognize that government speech can be dangerous,153 but many see it as similar in kind to private speech, albeit more powerful.154 Some see nothing special about it at all.155 Yet, although some governmental communication takes the form of oral or written expression emanating from a specific person, it is just as artificial to label such state action “speech” as it is to speak of the government’s rights to bear arms or to protect its own privacy. Government communication fundamentally differs from private speech, even when those communications take the form of writings or utterances of individual governmental representatives.
One essential difference between governmental and private communication concerns their respective functions. The law widely protects private speech interests—not just through the U.S. Constitution156 and various state constitutions,157 but also through international human rights instruments.158 Although scholars disagree on precisely why we value freedom of expression,159 they often associate it strongly with individual welfare,160 self-realization,161 and human agency,162 in addition to recognizing its importance for exerting influence over the government itself.163 By contrast, the values served by governmental communication are specifically parasitic on the interests of the government’s citizens. The fundamental purpose of governmental communication in a representative system—to govern better or more effectively—is about the wellbeing of those it governs.164 Whatever one’s view of the role of the government, empowering governments to communicate is not a matter of recognizing governmental dignity.165
Additionally, whereas a key basis for protecting private speech is that it allows individuals an important sphere for self-expression, government speech does not link substantially to any sentient entity, which seriously undermines the notion that the government “speaks” in a meaningful sense. In some cases, the government will communicate content “facelessly” (such as when the IRS issues regulations),166 or vicariously (such as when a federal department adopts certain ad campaigns).167 In those cases, the content communicated is functionally untethered from the perspective of any governmental agent.168 Even when a specific official is associated with government expression, it is not always clear if that person truly agrees with the content advanced, or how many others in government share the view.169
Thus, while honest statements from private speakers possess some stability because they spring from a web of interacting beliefs, government communication generally lacks such stability. Indeed, there are serious complications in attributing any specific viewpoint to the government in a number of important contexts, including the purpose of its laws and regulations.170 A substantial subset of governmental communication reflects the siloed, often-fleeting judgment of some official or group of officials about a specific subject of governance.171 Those decisions—about how to handle the taxation of some form of income, or the conditions under which one should get vaccinated against COVID-19—are subject to change not merely because the “speaker” changed her mind, but also quite frequently because another official, committed to another sort of policy, has rotated through the relevant chair in the relevant office. Even to the extent a government speaker speaks earnestly, reflecting her own views, the government’s commitment to that statement is subject to change for a broader set of reasons—not just if the official changes her mind, but also if she is replaced or otherwise ordered to change the policy.172
Additionally, unlike private speech, government communications draw on the sovereign power of the state and therefore take on a distinctive character.173 Governments can create juridical and even empirical facts simply by stating them, in part because the government’s position on a matter is often itself what is legally or empirically significant.174 Not only do governmental pronouncements restrict their subjects through the creation of laws and regulations—a power by and large unavailable to private citizens absent special circumstances, such as ballot initiatives175—but they also purport to restrict themselves through dictates both public176 and hidden.177
The government also possesses a distinctive power not to speak—or to stop others from speaking—which is especially pronounced in the domain of national security. The federal government, for example, holds unique power to classify (or over-classify178) information of public significance,179 to kill off lawsuits by invoking the state secrets privilege,180 to designate even public information as secret in certain contexts,181 and otherwise to restrict information that undermines its preferred messaging.182 Whether through affirmative statements or through withholding of information (or both at once), the government creates a reality that binds the rest of us. Citizens are constrained by governmental policies based on false factual predicates, even when citizens recognize that the predicates are false.183 Further, nontrivial segments of the population will often simply accept whatever representation the government makes, even if those representations are implausible or widely contradicted and debunked.184 This largely reflects the power of sovereign expression.
Because governments lack continuity across time and across personnel, the government can also contradict itself internally (whether through inter-branch disputes, like lawsuits, or intra-branch disputes, such as legislative divides) or change its positions (such as when one administration takes over for another, or one official replaces another) without serious damage to its credibility.185 Some scholars have concluded that this “balkanization” of government communications undermines the power of the government to engage in coordinated, mass media campaigns.186 But this flexibility is actually a great source of power. The government can tailor its message to its audience, even if the message is contradicted by other arms of the government, or by the same arm of the government operating in a different venue. That gives it remarkable potential to mold public opinion. The president can represent to his supporters that the 2020 election was stolen through widespread voter fraud while omitting that claim before the courts.187 Millions of supporters will still accept his statements as true.188 And millions more will be compelled, effectively, to engage with his views even knowing them to be false.189 In short, some of the very qualities that stand to weaken private speech—especially falsity and inconsistency, which we routinely treat as liabilities for witnesses in legal proceedings190—in fact heighten the power of governmental communication.
For these and additional practical reasons, governments require much less protection than private citizens to make themselves heard.191 Governments possess unique—and uniquely powerful—means of communicating. They tend to have far more money than individuals or private entities. They can create their own media outlets,192 and they collect and maintain massive amounts of proprietary information.193 They conduct wide-ranging assessments about the state of the world and its people, some of which they publish194 and some of which they keep hidden.195 They also have an unusual capacity to mask their role in conveying certain messages. For example, they selectively leak information to the press to channel preferred messages through independent news outlets.196 Some surreptitious communications efforts even garner protection under the government-speech doctrine.197 And, precisely because of their governing role—because the positions of various branches of the state set the terms by which their subjects live—officials’ pronouncements are also much more likely than typical private speech to be important and newsworthy. As a result, officials associated with the government can easily command the attention of independent press to amplify their messages.
Despite the typical scholarly perspective and the Court’s jurisprudence on government speech, there is also some independent intuitive appeal to the notion that government communication is sui generis. For example, that assumption would help to explain the historic association of governmental communication—much more so than private speech—with the concept of propaganda.198 Nevertheless, the extent of the power differential between governmental expressive conduct and private expressive conduct is often misunderstood as a matter of degree rather than a difference in kind. Although it can be difficult to ascertain the efficacy of any specific governmental informational campaign,199 governmental communication is simply a different kind of activity altogether from personal, private expression.200 It operates on a different plane and for different purposes, and different rules dictate its success. There is no good reason to maintain the fiction that the government “speaks,” especially knowing that the concept sets the stage for jurisprudential errors such as the embrace of burden exemption.
C. Introducing the Sovereign Forum
Let us therefore set aside the government-speech doctrine and the very notion of government speech. As we noted above,201 before the birth of the doctrine, the Court relied on forum analysis to establish rules about the restriction of private expression within governmental domains.202 When the government allegedly favored one private viewpoint over another or otherwise burdened the private rights of expression in some governmental space, the Court inquired into the nature of the venue where the state action in question took place.203 Was it in a traditional public forum, like a public park; a limited public forum, like a municipal theater; or a nonpublic forum, like a polling place on election day?204 The Court affixed different rules for compliance with the First Amendment to each of these categories,205 based partly on the historical significance of that sort of venue for private expressive activity,206 and it began to categorize various locales to resolve First Amendment questions that arose there.207 In each of these forums—even in nonpublic forums, which feature the most government-friendly rules208—the Court disfavors burdens on private expression that discriminate based on viewpoint. Restrictions may be struck down if they are not appropriately tailored to a sufficiently important interest.209
Although forum analysis and the government-speech doctrine frequently flirt with one another in the same opinions,210 and despite some scholarly interest viewing the doctrine as creating a forum,211 the Court does not regard the government-speech doctrine as an extension of forum analysis. I have found no instances of the Court characterizing the government-speech doctrine as a mechanism for identifying a new forum, and the Court has declined to name a new forum for which the doctrine would serve as a gateway.212 In fact, the Court has confronted—and rejected—golden opportunities to link these two jurisprudential mechanisms, so its decision not to make the connection appears deliberate. For instance, when ruling that specialty license plates are government speech, the Court canvassed the forum categories it has historically deployed and rejected them all in favor of invoking the government-speech doctrine—without characterizing the government-speech doctrine as a new element of forum analysis.213
The reason for keeping the doctrine distinct from forum analysis is burden exemption. Forum analysis involves subjecting state action to some measure of First Amendment scrutiny, which is precisely what the government-speech doctrine was intended to avoid.214 Having jettisoned burden exemption, however, we are free to harmonize forum analysis with the government’s communicative prerogatives. Of course, a full assessment of something as complex as a new forum warrants its own dedicated treatment, but it is possible to provide some surprisingly clear and helpful guidelines here—enough to show the solution is not just workable, but also superior to the government-speech doctrine.
Although we have some flexibility to define a new forum for governmental communication, its fundamental purpose is fixed: It is the domain within which the government may abandon viewpoint neutrality for the purposes of endorsing various propositions itself, potentially implicating private expression.215 It encompasses governmental property and assets—including locales that have been designated traditional, limited, and nonpublic forums—and it extends (in limited fashion) to private resources.216 Crucially, it need not be a First Amendment-free zone: the government cannot use this forum to engage in speech that itself violates the First Amendment, such as endorsing specific religious views or enacting statutes that unduly burden private expressive rights.217 In a nod to the government’s distinctive communicative power, we can call this domain the “sovereign forum.”
One of the most interesting features of the sovereign forum is that we have always operated as if it exists even without formally recognizing or naming it. There is no history of legal controversy over whether the First Amendment prohibits elected representatives from espousing substantive views; impedes the enactment of statutes in principle because they codify an exclusively governmental message; or provides a remedy because the Federal Register gives voice to the federal government but not private citizens.218 Such incidents of governance are so manifestly necessary that nobody could reasonably hold them to violate private rights. For this reason, the substantial majority of instances in which the government invokes the sovereign forum to communicate are unlikely to invite any litigation or scrutiny whatsoever.
Where the invocation of the sovereign forum generates controversy, however, is when it entails the regulation of a private expression in a manner that is not evenhanded, or involves the compulsion of private actors, or otherwise violates the expectations of civilians about their own rights to private expression. This is precisely why, even though the Court’s test for government speech has nothing to do with private expressive rights, government-speech doctrine cases typically arise when a private citizen has some colorable First Amendment claim.219 Nobody seriously objects to the power of the government to communicate unless that communication interferes somehow with their own power to do the same. This framing also reveals why forum analysis frequently circles overhead when the Court invokes the government-speech doctrine: within the sovereign forum, some of the places where private citizens have the strongest expectation of being able to communicate are the areas designated as traditional and limited public forums.220 Challengers to state action in those contexts naturally rely on forum analysis to frame their cases.
1. A Three-Factor Sovereign Forum Analysis
I suggest three primary, interrelated factors guiding sovereign forum analysis—factors the government must satisfy to defeat a First Amendment claim challenging its exercise of a communicative prerogative. I sketch out each factor here and briefly demonstrate how well they fit our government-speech cases before concluding. Notably, these factors can easily incorporate large chunks of the Court’s extant First Amendment jurisprudence,221 creating only modest jurisprudential disruption while also accounting for concerns that scholars have persistently raised about the government-speech doctrine. Although there will still be close cases, I argue that the factor-analysis recommended here is clearly superior to the government-speech doctrine because it centers the proper considerations, and it fits better with our standing First Amendment caselaw. Additionally, these factors leave space for differential interpretation, giving the Court some flexibility in how to apply them; they do not inexorably lead to decisions I favor in contested cases.
The first factor concerns the government’s invocation of the sovereign forum. Because private citizens have recognized First Amendment rights in certain governmental spaces, and more generally because we are designing special rules specifically to facilitate the government’s power to communicate, the government must do something to make clear it is abandoning viewpoint neutrality to communicate its own message rather than to favor a private message. At the same time, given the necessity driving our recognition of the sovereign forum,222 it should not be especially burdensome for the government to invoke this power. A natural standard presents itself, at least provisionally: the Court’s basic test for government speech. We can say that when the state controls a message intended to be governmental, and its message is likely to be received as such, it is attempting to exercise its communicative prerogative.223 That includes abandoning viewpoint neutrality for the purposes of endorsing propositions, and it may also entail the exclusion of competing, private voices—subject to other considerations addressed below. This factor guards against worries of governmental “ventriloquism,”224 and its adoption would affect cases such as Rust and Johanns.225
The second factor is whether the government’s mode of communication appropriately falls within the scope of the sovereign forum.226 We have some freedom to define the scope of the forum. As a rule of thumb, the question should be whether the government is using an appropriate resource to spread its message—a resource it is entitled to use—rather than inappropriately hijacking a private asset, for example.227 As I noted above, the Court has interpreted the government-speech doctrine to extend to its use of “private assistance” to spread its message.228 The Court therefore appears open to extending the sovereign forum not just to governmental property but also at times to private assets.229 The more broadly the Court draws the lines in that regard, the more tension230 there will be between the sovereign forum and compelled speech.231 Accounting for the compelled, private amplification of government speech allows us to address another standing concern critics have raised about the government-speech doctrine.232
Wherever we draw the lines for this factor, we should note that the boundaries of the sovereign forum are not plausibly defined entirely in physical terms. For example, in addition to real property owned by the government, the forum may include certain forms of media, like publications issued by the government, television ads, websites, and so forth. It may also encompass communications arising in certain social contexts, such as the utterances or writings of government employees in their official capacities, regardless of physical location.233 This conclusion should not seriously disrupt the Court’s jurisprudence. Although the Court often designates traditional, limited, and nonpublic forums by physical location, it has also recognized certain forums that qualify “more in a metaphysical than in a spatial or geographic sense.”234
The third and final factor comprises two related inquiries: whether the content of the communication itself directly infringes a protected First Amendment interest, or whether the government unreasonably restricts private speech to communicate that content.235 Once more, this criterion gets at broad concerns about the government-speech doctrine.236 Despite its duality, this is a relatively straightforward inquiry because we have generations of decisions that help point the way. The first prong is important because, contra the government-speech doctrine, the First Amendment must remain in effect when the government communicates. More specifically, we know that the government cannot use its communicative power to implement rules or regulations that directly burden protected expressive rights—rules that prohibit the criticism of public officials, for example,237 or compel citizens to attend church.238
We also have some natural guidelines for considering whether the government unreasonably suppresses private speech to communicate its message: forum analysis. The government is entitled to communicate in traditional, limited, and nonpublic forums because those are all governmental venues. They fall comfortably within any plausible definition of the scope of the sovereign forum. But the government’s latitude to suppress competing private speech differs across those three contexts because each offers a different gradation of free-speech rights to private speakers.239 The traditional, limited, and nonpublic forum rules thus provide a baseline for understanding what should happen when the government’s quest to communicate interferes with private expression in one of those locales. Sometimes suppressing competing private speech will be inappropriate because the public has an expectation that it can use a venue relatively freely for its own expressive purposes. And even where that is not the case, the government’s suppression of counter-speech—that is, its prevention of counter-messaging from private citizens—must be even-handed across viewpoints.240 The government’s power to restrict competing messages from private citizens should plausibly be at its highest ebb in government venues that are not traditional, limited, or nonpublic forums,241 and at its lowest ebb when it seeks to deploy its message in privately owned spaces.242
In short, the proposed test requires the government to invoke the sovereign forum (1) clearly and (2) appropriately, and (3) to use the forum in a way that respects First Amendment rights as strictly as possible. Although questions can arise at any of these levels,243 most governmental communications should easily survive the inquiry recommended here. Consider the enactment of law, for example. Statutes, rules, and judicial decisions are quintessential governmental communication,244 intentionally and effectively communicating a governmental message, and their distinctive function plainly satisfies the first criterion. Governments also tend to use a combination of governmental publications245 and voluntary publication by private entities246 to memorialize rules and regulations,247 thereby staying plausibly within the boundaries of the forum. Third, governmental rules and regulations often focus on quotidian matters of governance like procedural rules, or vehicular speed limits, or how to dispose of household recycling, rather than areas sheltered by the First Amendment. Nor will their publication generally involve indirect suppression of private expression, given that private entities can republish the law and add their own commentary as they choose.248 These criteria appropriately approve of the means and mechanisms governments typically use to disseminate the law.
This is a preliminary sketch of the key factors, so it naturally leaves some questions unanswered. For instance, we might assess each of these factors on a binary basis and insist that “passing” on all the factors is a necessary condition for the government to defeat a First Amendment challenge. But we could also evaluate some of the factors on a spectrum; perhaps we care how clearly the government has invoked its communicative prerogative, or how firmly the government’s mode of communication falls within the sovereign forum, and we give it only partial credit on either factor in marginal cases. Similarly, we could give weight to countervailing factors. Perhaps we care not just whether the government has invoked its communicative prerogative, but also whether a private citizen is also clearly attempting (and succeeding) to convey a message at the same time and through the same mode of communication.249 Such a move could empower us to account for “mixed speech” without relying on the Court’s formal acceptance of that category.
2. Test-Driving the Factors
My own view is that we should draw the boundaries of the sovereign forum narrowly and take seriously competing claims of private expression. The government owns a lot of tangible property and communications infrastructure outright, in addition to possessing unique power to spread its message.250 I am skeptical of its need for a tailwind where its communicative activities begin to clash with private speech interests. Regardless, it is worth highlighting the new light this framework casts on contested government speech cases. Consider three examples.
The Court’s decision in Rust has been controversial partly because it focuses almost exclusively on the government’s communicative interests in silencing medical providers who might otherwise discuss abortion with patients.251 The facts raise difficult questions under the proposed sovereign forum analysis: With respect to the first factor, it is unlikely that the audience in question—the patients—would interpret silence by their doctors as a government message against abortion.252 More generally, the transparency required by the first factor is difficult to meet when the government enforces silence rather than issuing an affirmative message, or even a disclaimer. Further, under the second factor, a private medical consultation is not obviously an appropriate part of the sovereign forum; although the Court has effectively suggested it may be,253 that result is neither obviously correct nor compelled by the analysis proposed here. And even if the Court were to find in favor of the government on both fronts,254 the third factor still looms. The entire premise of the program is to amplify the government’s message by suppressing contrary private speech—in a “private” zone where governmental power to discriminate on viewpoint should be especially weak. Rust may therefore be difficult to save under the proposed framework.
Johanns raises different kinds of difficult questions under proposed factors. Recall the key facts: The government compelled beef producers to pay a “checkoff” to the government, a portion of which was used to air television ads that some of the beef producers disagreed with—ads that neither identified the government as their source, nor captured the views of some beef producers that their own products were superior to the products of other producers also taxed to fund the ads.255 Because the government did not clearly identify the message of the ads as emanating from the government, it failed on the first factor (invoking the sovereign forum). Additionally, although a television ad could plausibly be part of the sovereign forum if paid for by government funds (the second factor), a nuanced analysis of the third factor might also account for the compelled-speech dimension, where some of the funds behind the ad come from private parties who reject its content and are therefore forced to subsidize and be associated with a message they reject.256 Thus, Johanns should also come out the other way under the new framework.
Consider one more case for good measure: Summum highlighted the deficiencies of the government-speech doctrine in a unique and powerful way because the setting for the case was a traditional public forum: a public park.257 Remember that, in Summum, a municipality successfully fended off a First Amendment challenge after it denied the request of a small religious group to install a statue in the park—despite having installed a statue donated by a different religious group in the very same park.258 It is plausible that a permanent statue installed in a public park meets the first factor,259 clearly conveying some sort of governmental message.260 It is also plausible that such a statue falls comfortably within the sovereign forum, the second factor, because it is a governmental installation on governmental land. But the facts raise a problem under the third factor. Even leaving aside the possibility of a direct First Amendment problem caused by the content of the message—namely, the possibility of trouble under the Establishment Clause261—the entire fact pattern centers around the municipality’s favoritism toward certain private speech in a governmental forum where private expressive rights are especially robust.262
The Summum Court permitted the municipality to launder its viewpoint discrimination by converting the statue into government speech.263 Additionally, because the government-speech doctrine sits uncomfortably alongside traditional forum analysis, the Court rejected the latter despite the obvious classification of the venue as a traditional public forum.264 By contrast, under the factor analysis proposed here, there is a legitimate question about whether the municipal government’s communication in this traditional public forum unreasonably suppressed private expression because the government did not communicate sua sponte but rather on the prompting of a private party.265 Its rejection of the statue certainly was not viewpoint neutral, and it therefore chose not just to convey its own message but also to amplify one group’s speech and not another’s—in a part of the sovereign forum where it has historically faced the highest hurdles to restricting private speech, no less.266 Accordingly, under the factor analysis, the better holding would acknowledge the impracticality of demanding that municipalities accept all statues for installation in public parks, but hold instead that municipalities should reject all private submissions.
Different Justices could well apply the factors identified above somewhat differently—proof the factors are not rigged—but note two important points. First, despite some latitude for interpretation, the factors count against some of the key holdings the Court has reached under the government-speech doctrine.267 Second, even for those who might use the factors to affirm Rust, Johanns, Summum, or other questionable government-speech cases, the reasoning for sustaining those holdings is much more nuanced and compatible with traditional First Amendment analysis. I would argue that many government-speech cases are wrongly decided, but even if they are not, their reasoning—based on combining burden exemption with a binary classification of communications as either private or governmental—is erroneous. Instead, we now have the outlines of a framework for analyzing governmental communication that centers the First Amendment, conforms with forum analysis, and rejects burden exemption.
Conclusion
This Article urges the complete eradication and replacement of the government-speech doctrine. Although it is only thirty years old, the doctrine already features in dozens of federal and state decisions. Each passing year, its roots grow deeper, despite its reliance on the dangerous and unconstitutional notion of burden exemption and its advancement of the confused concept of “government speech.” This Article argues that we can abandon both these ideas and replace them with an upgraded forum analysis—a sovereign forum that allows the government to exercise its necessary communicative prerogatives while honoring the First Amendment and accommodating decades of precedent on governmental forums.