Abstract
The government generally may not punish speakers based on the content of their speech. Or so the story goes. While American courts frequently describe content neutrality as a foundation stone of expressive liberty, the results do not track the recitations. A systematic analysis of free speech jurisprudence reveals that content-based laws remain acceptable across a host of situations.
The pervasiveness of content discrimination requires a reorientation of First Amendment doctrine. The core precept of expressive liberty under the U.S. Constitution is that the government may not punish speakers for their positions or perspectives. The cases reveal that so long as viewpoint neutrality prevails, content discrimination often poses no constitutional problem.
The Supreme Court’s 2021 Term was full of blockbusters. The Justices decided major cases involving abortion,[2] firearms,[3] state secrets,[4] and beyond. But the dispute that came before the Court on November 10 was of a decidedly different ilk. In an effort to preserve natural beauty and promote public safety, the City of Austin, Texas had enacted various restrictions on outdoor signage.[5] Among the restrictions was one that limited the digitization of signs.[6] Yet not all signs were relegated to the predigital era. They could be digitized—allowing for quicker, cheaper, and easier changes to their displayed messages—if they advertised happenings “on the premises where” they were located.[7] So, for example, a restaurant was free to digitize a sign in its own parking lot displaying its name and daily specials, but not a billboard ten miles down the road telling hungry drivers where to find it.
On some accounts, Austin’s regulation was unconstitutional. The law discriminated against certain messages: namely, those promoting activities away from the premises where they occur.[8] The City, though, had an explanation for its approach. Austin had not discriminated against disfavored messages, but rather had acted to protect the public interest in safety and natural beauty while giving people more leeway to express themselves on their own property.[9]
When the dispute over Austin’s law reached the Supreme Court, the key question became which of these stories best fit the facts. Had Austin punished disfavored messages? Or had it regulated in a neutral fashion, notwithstanding some incidental differences in terms of ability to digitize signs? After a wide-ranging oral argument, complete with hypothetical kale shops, coffee giveaways, and debates over the definition of “cheese,” the situation was no clearer than when the opening buzzer had sounded.[10] Justice Breyer captured the spirit of the argument when he suggested to counsel that “maybe you can’t explain” what makes a speech restriction discriminatory, in which case “I’ll go back to my state of confusion.”[11] An unassuming case about digital signs had turned into a conceptual quagmire. And when the Court issued its decision some five months later, uncertainty remained. The distinction between on-premises and off-premises signs was content neutral after all.[12] But the definition of content neutrality remained elusive.[13]
Underlying the Justices’ examination of digital signs is a key conceptual distinction. Content-based speech restrictions are “presumptively unconstitutional.”[14] A judge confronting such a restriction applies strict scrutiny, which in the realm of expressive liberty almost always spells doom for the law in question.[15] But content-neutral restrictions receive more favorable treatment in the courts.[16] Though such laws do not always survive, they have a fighting chance.[17] That explains the considerable attention directed at whether a speech restriction discriminates on the basis of content. If a law is content neutral, it might well survive. If a law is content based, it almost certainly will fail.[18]
The implications extend far beyond billboards. Restrictions that discriminate based on the content of speech are legion, covering everything from securities disclosures[19] to tax returns[20] to signs alerting motorists about hidden hazards.[21] Given the rarity with which speech restrictions survive strict scrutiny, the government’s best chance in defending these laws is to establish that, despite their ostensible focus on the messages being transmitted, they actually operate in a content-neutral fashion.[22] Establishing that point is necessary to invoke the more forgiving, intermediate brand of scrutiny suited to content-neutral laws.[23] From the standpoint of the government defending a speech restriction, it’s (usually) content neutral or bust.
Driving the focus on content discrimination are underlying concerns about government orthodoxy. Private actors can and do take peoples’ opinions into account in planning activities and managing relationships. As Justice Holmes put it a century ago, “[p]ersecution for the expression of opinions” is “perfectly logical.”[24] The rules are very different for government actors. All opinions, no matter how misguided, are insulated from official censure. Modern American jurisprudence prizes unfettered debate as the engine of democracy[25] and the bulwark against government overreach.[26] A wide-open marketplace of ideas may not be perfect, the argument goes, but it is better than the alternative.
Promoting robust debate entails protecting speakers’ choices about which topics to address. The government generally possesses no greater authority to play favorites based on speakers’ topics than it does their perspectives.[27] In most contexts, the First Amendment forbids the government from singling out a particular subject—such as labor relations, abortion, or military policy—for exceptional treatment, just as it forbids singling out speakers on one side of a given issue. Both types of government action trigger strict scrutiny and are presumptively invalid.[28] As a doctrinal matter, the takeaway is that when the government is restricting or compelling speech, discrimination on the basis of content runs parallel to discrimination on the basis of viewpoint.
Yet the doctrinal correspondence between viewpoint-based restrictions and content-based restrictions belies a theoretical asymmetry. The imperative of viewpoint neutrality prevents the government from marginalizing disfavored perspectives. The courts steadfastly defend even “the thought that we hate” as a means of staving off orthodoxy.[29] When the government welcomes all perspectives on a particular issue but gives favorable or unfavorable treatment to speech based on its subject matter, the state of the law is more difficult to explain. There may be good reasons to worry about the prohibition of, say, labor-related speech or abortion-related speech. But those reasons do not include concerns over burdening one side of a debate while privileging the other. Nor does the republic hang in the balance when a municipality restricts yard signs while making an exception for those that read “for sale”—an exception that depends on the content of speech.[30] The point of these examples is to illustrate that the pitch of constitutional concern can be markedly different when the government focuses on a particular perspective rather than a broader topic of discussion. Nevertheless, the legal effect is the same: content discrimination, like viewpoint discrimination, triggers the most demanding scrutiny.[31]
Except when it doesn’t. Notwithstanding judicial suggestions that strict scrutiny follows content discrimination as day follows night, there are numerous situations in which the link is severed. Laws against defamation and incitement need not satisfy strict scrutiny despite their content-based nature.[32] Neither must restrictions on commercial advertising,[33] nor zoning regulations that limit expression,[34] nor rules for the government’s use of public resources to encourage discussion and debate.[35] The doctrinal tests in some of these areas are quite protective of speech, even without the application of strict scrutiny. For example, speakers have ample discretion to criticize the manner in which public officials discharge their duties and to advocate for social and political change.[36] Applying a standard other than strict scrutiny might be chalked up to mere nomenclature or historical accident; terminology aside, the speech receives powerful protection. In other contexts, however, laws that depend on the content of a speaker’s message receive not just a different version of scrutiny, but a noticeably weaker version. Those cases bend the ordinary rules of free speech. No longer does the Constitution forbid the government from drawing distinctions based on content. Even so, the imperative of viewpoint neutrality remains.[37]
This Article seeks to reorient the relationship between viewpoint and content in the law of free speech. As a positive matter, the Article contends that First Amendment jurisprudence is more accepting of content discrimination than initial appearances suggest. As a normative matter, the Article defends a focus on viewpoint rather than content as consistent with the First Amendment’s core commitments.[38]
Though the case law often describes content discrimination and viewpoint discrimination in like terms, courts consistently respond more aggressively to the latter. To the extent that content discrimination is relevant, it is often as a proxy for, or facilitator of, viewpoint discrimination.[39] Asking whether a law depends on the content of speech is a means to an end, a way to determine whether “official suppression of ideas is afoot.”[40] Only rarely has the Supreme Court extolled content neutrality for its own sake. And the Court has done so without fully exploring the doctrinal foundation upon which it builds.
Though there is some support in the case law for the proposition that content discrimination leads ineluctably to strict scrutiny (and likely invalidation),[41] judicial practice over the past century suggests a more nuanced approach. Content-based restrictions trigger strict scrutiny when there is reason to believe that the government is disfavoring certain views in pursuit of a preferred vision. When there is no basis for such a belief—when, that is, “there is no realistic possibility that official suppression of ideas is afoot”[42]—restrictions based on the content of speech are not inherently suspect. The Supreme Court has already embraced this principle in a variety of contexts.[43] The remaining step is to recognize the principle’s applicability across the law of free speech. Asking whether a restriction depends on the content of speech is ancillary. It is viewpoint neutrality that stands at the center.
This Article begins in Part I by contextualizing the conventional account of content neutrality as integral to expressive liberty. Part II challenges the familiar narrative by examining numerous ways in which the law of free speech undermines the claim that content discrimination is presumptively invalid. In Part III, the Article explains how viewpoint neutrality, not content neutrality, has defined modern First Amendment jurisprudence. Part IV explores the implications of recognizing the primacy of viewpoint neutrality and the peripheral nature of content neutrality. Peripheral does not mean irrelevant; content discrimination remains an important consideration. But its function is to help identify risks of official suppression and ordained orthodoxy. Where content-based laws suggest the stifling of unwelcome views, strict scrutiny remains in order. Where no such risks are present, the simple fact of content discrimination does not justify the strong medicine of strict scrutiny.
The Article concludes by addressing the practical, transitional issues bound up with shifting the doctrinal focus from content to viewpoint.
I. Content Neutrality as Imperative
Black letter law provides that strict scrutiny applies to regulations that depend on the content of speech.[44] To survive, such regulations must be “narrowly tailored to serve compelling state interests.”[45] In the realm of free speech, strict scrutiny almost always leads to invalidation of the law at issue. Indeed, the Supreme Court has described content-based restrictions as “presumptively unconstitutional.”[46]
Sometimes laws focus not on the message that is conveyed, but rather on the circumstances of its conveyance. The government might prohibit the use of sound trucks in an effort to control noise levels, irrespective of the words being amplified.[47] Or it might forbid any spectator at a judicial hearing from interrupting proceedings with a diatribe, whether the speaker’s chosen topic is politics, sports, or weather. These sorts of limitations on the “time, place, or manner”[48] of speech do not depend on the speaker’s message. To the contrary, they are “content neutral.”[49] Courts accordingly treat them as less suspicious than content-discriminatory laws.[50]
Determining whether a speech restriction discriminates on the basis of content is a high-stakes proposition. When a court concludes that a law is content discriminatory, the rest is often academic; absent extraordinary circumstances, the law is destined to fail.[51] The effect has been to put immense pressure on the line between content discrimination and content neutrality. As Erwin Chemerinsky observed in 2000, “virtually every free speech case turns on the application of the distinction between content-based and content-neutral laws.”[52]
In 2015, the Supreme Court doubled down on the salience of content neutrality. The occasion was a dispute that appeared better suited to the back page of a local newspaper than the docket of the highest court in the land. The town of Gilbert, Arizona, had enacted a strikingly complicated set of regulations governing the public display of signs. The regulations covered everything from political signs to directional signs to signs advertising garage sales.[53] The type of sign at issue determined its permissible size, location, and duration of display.[54]
Reed v. Town of Gilbert arose from the application of this regulatory scheme to temporary directional signs advertising religious assemblies, which faced stringent limitations under the town’s regulatory scheme.[55] The onerous rules prompted a challenge by a pastor who, lacking access to a permanent location for his Sunday services, relied on temporary signs for promotion.[56] The constitutional question was whether the town’s regulatory approach violated the First Amendment.
Reed was not a hard case. All the Justices agreed that the regulatory system was invalid.[57] Justice Kagan went so far as to observe that the system could not pass “the laugh test,” given that it lacked “any sensible basis.”[58] What divided the Justices was not the outcome, but the mode of analysis. The regulatory scheme at issue depended, unabashedly, on the content of a given sign. The information that a sign conveyed determined its categorization. For the Reed majority, that link was enough to trigger strict scrutiny—and to seal the fate of the regulatory system.[59]
More important than the outcome of Reed were the principles that guided the Court’s analysis. It was indisputable that the town’s regulatory system operated based on the content of signs. Even so, there were grounds for concluding that the suspicion which sometimes attends content-discriminatory laws was unwarranted on the facts presented. True, the town had enacted a byzantine regulatory system.[60] But there was no indication that government officials were suppressing disfavored perspectives to push their own agenda. For the Reed majority, this distinction was immaterial.[61] When a law takes into account the content of speech, the proper rubric is strict scrutiny, regardless of whether the government appears to be penalizing disfavored views.[62] To survive, such a law must be narrowly tailored to a compelling interest—a nearly insurmountable hurdle.[63] In articulating this unflinching rule, the Reed Court did not delve into founding-era thought or the history of the First Amendment’s enactment. It did not elucidate what it means to “abridg[e] the freedom of speech.”[64] Nor did it offer a deep theoretical account of why a formal rule against content discrimination is vital to expressive liberty. Instead, Reed looked to precedent. It described the case law as charting a direct course from content discrimination to strict scrutiny, without any off-ramps.[65] To support its position, the Court cited cases on issues including public protests,[66] televised sexual content,[67] and commercial advertising,[68] among others.
Reed provides emphatic endorsement of the principle that distinctions based on the content of speech are subject to strict scrutiny (and probably invalid), regardless of whether the government appears to be imposing orthodoxy or stifling unwelcome perspectives.[69] The case warrants study for its unequivocal approach to the role of governmental justifications in determining the proper degree of First Amendment scrutiny. Under Reed, the government’s justification for enacting a speech restriction is a one-way ratchet. An illegitimate, suppressive justification can render a facially neutral law subject to strict scrutiny.[70] But a benign justification cannot absolve a law that draws content-based distinctions on its face.
Despite its youth, Reed already has found itself at the center of controversy. As noted above, the Supreme Court ruled in City of Austin v. Regan National Advertising of Austin, LLC, that a distinction between on-premises and off-premises signs does not necessarily discriminate on the basis of content.[71] The Court distinguished Reed on grounds that Austin’s law required “an examination of speech only in service of drawing neutral, location-based lines,” without regard to “topic,” “subject matter,” or “substantive message.”[72] As the dissent observed, however, it seems difficult to square the majority’s reasoning with Reed’s statement that restrictions are content based if they “target speech based on its communicative content.”[73] For now, two points are clear. First, the rationales of Reed and City of Austin are in considerable tension. Second, both cases are good law within their respective spheres. Which decision will resonate in the years ahead is anyone’s guess.
II. Challenging Content Supremacy
Despite their prevalence, descriptions of the link between content discrimination and strict scrutiny can be misleading. In a host of cases across a range of contexts, the commitment to content neutrality has proved more limited and contingent than it initially appears.
The following Sections explain how judicial practice complicates the depiction of content neutrality as a constitutional imperative. The Supreme Court has accepted content discrimination in numerous domains.[74] For present purposes, I set aside situations in which the government acts in a specialized role such as managing employees or educating students, in which content discrimination is ubiquitous. Whether an employee receives praise or punishment for statements made at work commonly depends on the content of those statements, and the same is true of students in public schools.[75] My project here is broader. I aim to show that when we venture beyond specialized contexts like public schools and workplaces to examine the government’s core regulatory activities, we encounter more and more evidence that content discrimination is an accepted facet of the law of free speech.[76] This lesson emerges from the treatment of nonpublic forums, the recognition of categories of speech that receive reduced protection, the curious doctrine of “secondary effects,” and the inquiry into unlawful suppression that occurs even when the government regulates speech of diminished value. I discuss each of these issues in turn.
Strict scrutiny applies to content-based laws in public spaces characterized by open discourse. Sometimes these spaces develop their character through usage over time, as in the case of parks and sidewalks that are quintessential traditional public forums.[77] In other situations, the government creates a designated public forum by establishing a locale as a venue for unfettered exchange.[78] In either situation, content-based laws are presumptively invalid as impediments to free-flowing discussion.[79]
Of course, the government frequently uses property for purposes other than the uninhibited exchange of ideas. Some of those purposes have little to do with facilitating speech per se, as when the government sets up a polling place to record votes or operates a courthouse to adjudicate legal disputes.[80] Though plenty of speech occurs within these nonpublic forums, that speech is incidental to the government’s operational goals. In other situations, the government is concerned with promoting expression, but in a cultivated environment. The result is a limited public forum, which is a type of nonpublic forum geared toward facilitating expression within specified bounds.[81] A university, for example, might offer meeting space to student groups, but not other speakers, in order to support its pedagogical objectives.[82] Or a city might make an auditorium available for theatrical productions, but not other forms of expression.[83] The government can pursue similar ends through the use of funding rather than physical space, for instance by subsidizing printing costs of student publications in an effort to spur commentary and exchange within the student body.[84]
The Supreme Court has recognized that a rule of content neutrality is an ill fit for speech restrictions within nonpublic forums. If the government is to accomplish its operational objectives, it needs discretion to allow certain types of speech while prohibiting others.[85] The same is true of limited public forums that are devoted to the facilitation of speech within articulated bounds. In responding to the unique dynamics of nonpublic forums, the Court has ruled that content discrimination is acceptable and unremarkable.[86] The government may discriminate on the basis of content as much as it wishes, so long as its restrictions are reasonable—and so long as the government remains neutral when it comes to viewpoint.[87] The idea is to ensure viewpoint neutrality while recognizing “[t]he necessities of confining a forum to the limited and legitimate purposes for which it was created.”[88]
The rules governing nonpublic forums undermine the notion of content neutrality as a First Amendment imperative. Whether those rules are sound or unsound,[89] they establish nonpublic forums as a realm in which content discrimination is perfectly lawful. Equally important is the recognition that nonpublic forums remain subject to the demand of viewpoint neutrality. The government’s authority to impose content-based speech restrictions in a nonpublic forum does not license the marginalization of disfavored perspectives.[90]
B. Exceptions to Full Protection
Even beyond the boundaries of nonpublic forums, content discrimination proceeds unabated in numerous contexts. A prominent set of examples fall under the heading of exceptions to full First Amendment protection.[91] These exceptions are notable for their departure from any pretense of content neutrality. Whether speech qualifies for full protection as opposed to diminished protection depends entirely on its content, and there is no suggestion that strict scrutiny is the applicable mode of analysis when the government restricts lesser-protected speech.[92]
For decades, it has been clear that certain types of speech occupy a lower rung of First Amendment salience. Illustrative categories include incitement, obscenity, and defamation.[93] While the dichotomy between fully-protected and lesser-protected speech is well established, its theoretical foundations have been fluid. In the mid-to-late twentieth century, the Supreme Court customarily described the exceptions to full First Amendment protection as the product of cost-benefit analysis conducted on a categorical basis.[94] More recently, the Court has explained that the exceptions are ultimately grounded in history and tradition. Going forward, the key question is whether a category of expression has been “historically unprotected” under American law.[95]
Whether they are understood in terms of costs and benefits or historical practice, the exceptions to full constitutional protection provide a vivid illustration of judicial acceptance of content discrimination. There is no doubt that laws against defamatory speech are driven by content.[96] The same is true of laws targeting obscenity, incitement, and other speech of diminished value.[97] Yet these laws do not face strict scrutiny.[98] The distinction sometimes can seem superficial, driven by differences in nomenclature rather than meaningful variations in constitutional solicitude. The modern test for incitement is strongly protective of speech, and its requirement of a likelihood of imminent lawless action might seem like strict scrutiny by another name.[99] Likewise, constitutional protection for speech is robust when a public official brings a defamation suit, leading to outcomes that might not differ markedly from those that would follow the application of strict scrutiny.[100] But in other contexts, the requirements for regulating speech within lesser-protected categories are less demanding. To return to the law of defamation, when a speaker levels false statements against a private figure, constitutional protection is fairly weak—even though the content of the speech is what triggers liability.[101] And though the doctrine of obscenity includes limits to avoid sweeping in too much speech,[102] it does not ask whether the government could have found a less speech-restrictive way to accomplish its goals—a question that is crucial within the realm of strict scrutiny.[103] Similarly, the Supreme Court has implied that speech causing intentional infliction of emotional distress on matters of private concern lacks constitutional protection, despite the fact that a cause of action based on such speech would flow directly from its content.[104]
The treatment of commercial speech underscores the point. It has been clear for decades that speech “propos[ing] a commercial transaction”[105] receives reduced constitutional protection.[106] This distinction has held in the face of strident criticism from those who would treat advertising as akin to other speech.[107] Regardless of who has the better of this argument, there is only one way to tell whether a message qualifies as commercial speech: by looking at what it says.[108] Commercial speech is not susceptible to regulation to the same extent as obscenity or incitement. The government still must demonstrate a “reasonable fit”[109] between the speech restriction and the government’s “substantial”[110] interest.[111] Yet the degree of constitutional protection is reduced as compared with fully protected speech, which the government may not suppress on the basis of content without triggering strict scrutiny.
The contrast between commercial advertising and other expression is even starker when speech is deceptive or illicit: misleading claims about commercial products are subject to restriction,[112] whereas misleading claims about matters such as religion and philosophy receive robust protection.[113] Moreover, advertisements for illegal products are unprotected, while ideological advocacy of illegal action is permissible within broad bounds.[114] In case after case, the government can and does restrict commercial speech on the basis of its content without triggering strict scrutiny.
If there is a nail in the coffin of the claim that content discrimination leads directly to strict scrutiny, it is the doctrine of secondary effects. That doctrine was on display in the mid-1970s as a governance challenge emerged in Detroit.[115] City officials sought to protect the quality of neighborhoods by preventing multiple adult businesses, such as adult movie theaters, from locating in the same area.[116] The rationale was that allowing those businesses to operate near one another could be “especially injurious to a neighborhood” by increasing crime and reducing property values.[117] But the zoning law that implemented the City’s policy raised a constitutional question, for it effectively excluded certain speech—including sexually explicit motion pictures—from particular locations.[118] What is more, the operation of the law depended on the content of speech. Whether a theater could operate in a given locale was determined by the types of movies it sought to display.[119]
Were content neutrality a constitutional imperative, striking down the zoning law would have been a simple matter. There are, after all, numerous ways to pursue goals like reducing crime and preserving property values without restricting speech; bolstering police presence and investing in physical infrastructure are two of the more obvious options. Nevertheless, the Supreme Court held that Detroit’s “interest in planning and regulating the use of property for commercial purposes” was sufficient to justify restrictions on the expressive activities of movie theaters.[120] A four-Justice plurality challenged the idea that content neutrality is an unflinching requirement.[121] The plurality acknowledged statements in the case law indicating that content discrimination is “absolutely preclud[ed].”[122] But those statements must not be “read literally and without regard” for context.[123] In reality, the protection afforded by the First Amendment “often depends on the content of the speech.”[124] The question is whether the government has played favorites based on the “social, political, or philosophical message” a speaker wishes to convey.[125] If the answer is no—if, in other words, the government merely seeks to regulate “secondary effect[s]” of speech such as neighborhood deterioration[126]—strict scrutiny is inapposite.
A decade would pass before the Justices transformed the secondary effects theory from a plurality rationale into a holding of the Court. Whereas Detroit sought to disperse adult businesses throughout the city, officials in Renton, Washington focused on preventing adult theaters from locating near places like residential zones, schools, and churches.[127] From a doctrinal perspective, the impact was the same. To determine where a theater was permitted to locate, one had to consider the types of films it sought to exhibit. The Renton law, like the Detroit law, was content based through and through.
As it had done in Detroit, the Supreme Court upheld Renton’s approach to zoning despite its discriminatory nature. In City of Renton v. Playtime Theatres, the Court accepted the proposition that content discrimination renders a law presumptively unconstitutional.[128] Yet it deemed that rule inapplicable to the case before it. Renton, the Court explained, was not attempting to restrict sexually explicit motion pictures because of their content. Instead, it was managing nonspeech effects associated with adult theaters.[129] Based on Renton’s goals and justifications, the Court treated the zoning law as if it were content neutral.[130] In making this conceptual move, the Court demonstrated the importance of pushing beyond superficial discussions to engage with underlying justifications for the rules of free speech. The First Amendment’s abiding concern is not content discrimination per se, but rather the possibility that the government might stack the deck for “people whose views it finds acceptable.”[131] When that concern is off the table, judicial skepticism dissipates.
The Supreme Court would engage with the doctrine of secondary effects again in yet another case involving adult businesses. No longer could a majority find common ground. In City of Los Angeles v. Alameda Books, Inc., four Justices adopted the Renton framework and treated the relevant ordinance as if it were content neutral.[132] Laws aimed at secondary effects, the plurality explained, furnish “less reason to be concerned” about attempts to “discriminate against unpopular speech.”[133] Writing separately, Justice Kennedy took issue with the depiction of zoning laws directed at adult materials as content neutral.[134] But he agreed with the plurality that such laws are valid if they are likely to reduce the deleterious secondary effects of speech without exacting a significant toll on the freedom of expression.[135] Not every law whose operation depends on the content of speech portends a looming governmental orthodoxy. Or, more simply: not all content discrimination is “impermissible content discrimination.”[136]
The secondary effects doctrine acknowledges that the government’s “paramount obligation of neutrality” does not foreclose discrimination on the basis of content.[137] Constitutional problems arise only when a law’s operation depends on the speaker’s “point of view.”[138] It does not matter whether a zoning ordinance prohibits speech “whose content is thought to produce distasteful effects.”[139] The crucial issue is viewpoint. So long as viewpoint neutrality persists, content discrimination is excusable.[140]
This principle runs counter to the Supreme Court’s approach to content discrimination in Reed.[141] Recall that in Reed, the fact that a law’s operation depended on the content of speech was sufficient to trigger strict scrutiny.[142] It was neither necessary nor appropriate to ask whether mitigating factors supported a less demanding form of review.[143] The lesson of the secondary effects cases is just the opposite, proceeding from the premise that not all content discrimination is created equal.[144] Justice Kagan noted this dissonance in her concurrence, which cited the doctrine of secondary effects as evidence that the Court had “been far less rigid than the majority admits in applying strict scrutiny to facially content-based laws.”[145]
Reed was not Justice Kagan’s first foray into the doctrine of secondary effects. During her time on the faculty of the University of Chicago, then-Professor Kagan depicted the secondary effects cases as revolving around government motive.[146] When the government acts to preserve property values or reduce crime rates, its regulations can be free of the suspicion that sometimes attends content discrimination.[147] The secondary effects doctrine, she suggested, avoids the “overkill” of strict scrutiny and represents a more nuanced approach to gauging the lawfulness of government action.[148] Justice Kagan’s previous scholarly observations serve as a counterpart to her concurrence in Reed.
One possible takeaway is that, notwithstanding Justice Kagan’s objections, the Reed majority undermined the logic of the secondary effects doctrine, foreshadowing its formal overruling at some future date.[149] Yet Reed did not purport to break any new ground. The Court framed its opinion as reflecting a familiar aversion to content discrimination that had emerged over the course of decades.[150] Reed’s incompatibility with the secondary effects cases does not suggest that the latter have been tacitly rejected, so much as it indicates that the doctrinal foundation Reed built upon requires additional examination.
The secondary effects cases are integral to understanding the role of content neutrality in the First Amendment ecosystem. Those cases advance the proposition that a brand of content discrimination—the kind where the government is pursuing goals other than the marginalization of disfavored views—is relatively benign. Even if zoning ordinances discriminate based on the content of speech, they are treated as if they were content neutral so long as they “are justified without reference to the content of the regulated speech.”[151] When the government punishes a disfavored viewpoint, the Court does not relax its scrutiny.[152] But when the issue is content rather than viewpoint, the constitutional calculus is different.
D. Segmentation of Low-Value Speech
The previous Sections explained how, notwithstanding statements to the contrary, the law of free speech does not impose a requirement of content neutrality. Some justifications for content discrimination are impermissible—namely, the stifling of disfavored viewpoints.[153] Others are less worrisome—for example, the regulation of secondary effects.[154] The crucial distinction is the law’s justification. Without the “why,” the “what” is indeterminate.[155]
The import of justification is underscored by a decision from 1992 dealing with the doctrine of fighting words. R.A.V. v. City of St. Paul[156] tested the validity of an ordinance criminalizing hate speech. As authoritatively construed by the state courts, the ordinance applied only to fighting words,[157] meaning words “which by their very utterance inflict injury or tend to incite an immediate breach of the peace.”[158] Fighting words stand alongside categories of expression such as defamation and obscenity as exceptions to full First Amendment protection.[159] As a result, the City of St. Paul would not have raised any constitutional concerns if it had enacted a complete ban on fighting words. But the City did something different. It singled out a subset of fighting words for illegality: those based on “race, color, creed, religion, or gender.”[160]
Rather than walking the line between fighting words and other forms of speech, the City drew distinctions among types of fighting words based on the nature of the underlying insult. In so doing, it stumbled from the realm of permissible content discrimination into the domain of unlawful viewpoint discrimination. The government must play by certain rules even when it regulates lesser-protected speech such as fighting words. It can restrict all speech within the category, or it can limit a subcategory on grounds that make “the entire class of speech . . . proscribable”—for example, by forbidding those fighting words which are most likely to provoke a fight, or those obscenities which are most prurient, or those threats which are most fear inducing.[161] What the government may not do is determine which words are unlawful by considering the viewpoint being expressed.[162] So long as the distinctions the government draws offer no suggestion that “official suppression of ideas is afoot,” discrimination on the basis of content may be permissible.[163]
Understood against this backdrop, the puzzle in R.A.V. was how to treat a speech restriction that could have gone further but didn’t. St. Paul’s ostensible disdain for fighting words should have led it to prohibit insults based on political affiliation and sexual orientation just as it did insults based on religion and gender.[164] By failing to go where logic should have carried it, St. Paul ran afoul of the First Amendment, despite the fact that it was dealing with a category of speech that receives reduced constitutional protection.[165] Moreover, the Supreme Court explained, the very “rationale of the general prohibition” against content discrimination is that it “raises the specter that the Government may effectively drive certain ideas or viewpoints from the marketplace.”[166] That is, we worry about content to protect viewpoint. The ultimate question is whether “official suppression of ideas is afoot.”[167] Discrimination on the basis of content may be nefarious, or it may be benign. The analysis is not one-size-fits-all.[168]
R.A.V. closes the case against content neutrality as a First Amendment imperative. The operation of nonpublic forums, the specialized rules in areas such as defamation and commercial speech, and the secondary effects cases reveal that there is nothing inherently problematic about content discrimination. R.A.V. doubles down, declaring that content discrimination is unobjectionable so long as “there is no realistic possibility that official suppression of ideas is afoot.”[169] Taken in combination, these doctrines demonstrate that the core concern of the law of free speech is not content.[170] It is viewpoint.
III. Content and Viewpoint Reconsidered
Part II disputed the account of content neutrality as central to the law of free speech. The notion of content discrimination as a problem in itself—as opposed to a problem insofar as it disguises or facilitates viewpoint discrimination—certainly has a basis in the Supreme Court’s opinions. That basis, though, is superficial, operating at the level of aphorism. The same opinions go on to describe the underlying concern as viewpoint discrimination.[171]
Within the American constitutional tradition, the engine of expressive liberty is the prohibition against compelled orthodoxy. Justice Jackson made the point some eighty years ago, observing that “[i]f there is any fixed star in our constitutional constellation,” it is that the government may not impose its preferred worldview upon the citizenry.[172] The principle has carried forward, with the Supreme Court reiterating that “the government may not prohibit the expression of an idea simply because society finds the idea itself offensive or disagreeable.”[173] Justice Alito recently described discrimination against disfavored views as nothing less than “poison to a free society.”[174]
Understood in light of this concern with compelled orthodoxy, judicial discussions of content discrimination take on a different complexion. Content-based restrictions are most problematic when they facilitate the imposition of orthodoxy.[175] There are exceptional circumstances in which content discrimination would be independently problematic, for instance if the government were to ban all discussion of an upcoming election or other issue of public concern.[176] That sort of ban would raise the risk of official orthodoxy despite appearing evenhanded. Outside of such exceptional circumstances, content discrimination raises a “red flag,” but it does not invariably trigger the application of strict scrutiny.[177] There must be some additional basis for suspicion.
This Part traces the evolution of content neutrality and viewpoint neutrality as interwoven First Amendment phenomena. Since the middle of the twentieth century, the identification of content discrimination has served primarily as a means of rooting out discrimination on the basis of viewpoint. The Supreme Court’s recent engagements with viewpoint discrimination—and the Court’s attendant efforts to develop a working definition of viewpoint neutrality—emphasize the point by training the law’s focus on whether the government is punishing speakers whose perspectives it disapproves.
A. Content Discrimination as Smoke, Not Fire
During the middle of the twentieth century, neutrality concerns began to play a leading role in the Supreme Court’s discussions of expressive liberty.[178] The Justices invalidated laws whose operation depended on “the content of the ideas expressed.”[179] The same fate awaited speech restrictions predicated on shielding listeners from offensive communications.[180] The doctrine that emerged called for the application of strict scrutiny to content-based restrictions on speech.[181] Content discrimination was lawful only if the government pursued a compelling interest and employed the narrowest and most careful of approaches to regulation.[182]
This history served as the basis of the Court’s recent reflections in Reed v. Town of Gilbert.[183] As Part I explained, Reed left no doubt that “[c]ontent-based laws” are “presumptively unconstitutional,” irrespective of whether the government appears to be disfavoring a particular view.[184] Reed grounded its aggressive skepticism of content discrimination in twentieth-century judicial practice. But while the cases upon which Reed builds speak the language of content discrimination, their primary focus is on ensuring viewpoint neutrality and fending off any efforts at the imposition of orthodoxy.
Let us begin with Police Department v. Mosley, decided in 1972.[185] A Chicago ordinance prohibited picketing near schools, but it made an exception for schools involved in labor disputes.[186] The ordinance unmistakably drew a distinction based on the content of speech. Labor picketing was permitted, while picketing related to other issues—for example, racial discrimination—was not.[187] The Court observed that the “central problem with Chicago’s ordinance is that it describes permissible picketing in terms of its subject matter.”[188] Yet the Court did not stop there. It went on to criticize the government for elevating some ideas over others based on implicit conclusions about their relative worth. The forging of hierarchies offends the constitutional ideal that so far as the government is concerned, “all points of view” deserve “an equal opportunity to be heard.”[189]
The Court’s reference to equality among “points of view” suggests that Mosley was not about content per se. Content discrimination was the smoke, but viewpoint discrimination was the fire. As a plurality would observe a few years later, “[i]f picketing in the vicinity of a school is to be allowed to express the point of view of labor,” the expression of other points of view likewise must be allowed.[190] The Chicago ordinance not only treated labor-related speech more favorably than speech on other topics; it effectively created a carve-out for people speaking from a pro-labor perspective.[191]
Much the same is true of Carey v. Brown, decided eight years after Mosley.[192] Carey, which Reed also cites,[193] was Mosley with a twist: While Mosley involved picketing near schools, Carey was about picketing near private residences.[194] The statute at issue in Carey imposed a restriction on protesting but made an exception for labor disputes.[195] The Supreme Court ruled that the statute violated the Constitution.[196] It used the language of content discrimination in explaining how the government’s approach had gone astray.[197] Content, though, was just a proxy. The statute allowed employees and contractors to picket homes owned by people against whom they had a grievance. Of course, it also permitted homeowners to respond, but the very point of protecting residential privacy was to remove the need for such engagement around the home. As a practical matter, the statute in Carey gave preferential treatment to those on one side of a particular type of dispute. Its constitutional flaw was the same as that of the ordinance in Mosley.
A comparable analysis applies to Reed’s invocation of Sorrell v. IMS Health Inc.,[198] a 2011 case involving doctors’ prescribing practices.[199] The State of Vermont limited the sale of certain prescriber data—data that was of particular interest to sellers of branded pharmaceuticals, who were willing to pay in hopes of enhancing their sales pitches.[200] The Supreme Court characterized the government as making things harder for branded pharma in order to support producers of lower cost, generic alternatives against which branded drugs compete. By pursuing this objective via impediments to the free flow of information, the government sought to “tilt public debate in a preferred direction.”[201] That tilting, the Court explained, represents the unlawful privileging of one viewpoint over another.[202]
Other cases that Reed cited are in accord, reflecting fundamental concerns about viewpoint and orthodoxy. Take Ward v. Rock Against Racism, which describes the constitutional question as whether the government has restricted speech due to “disagreement with the message it conveys.”[203] Or Turner Broadcasting System, Inc. v. FCC, which sounds the alarm about attempts to “suppress unpopular ideas or information or manipulate the public debate.”[204]
The decision that provides the greatest support for Reed’s focus on content neutrality is Consolidated Edison Co. v. Public Service Commission, decided in 1980.[205] The rule at issue in Consolidated Edison prohibited utility companies from including inserts on issues such as the use of nuclear power along with their customers’ bills.[206] The restriction did not advance one side of a debate, whether on nuclear power or on any other issue. It discriminated only on the basis of content, not viewpoint. In deeming the restriction unconstitutional, the Supreme Court observed that “[t]he First Amendment’s hostility” extends beyond viewpoint discrimination to the “prohibition of public discussion of an entire topic.”[207] Content neutrality appeared to be the utmost concern.
Still, it is worth noting the Court’s description of its inquiry. The Court explained that it must scrutinize content-based restrictions “more carefully” to make sure the government has not prohibited speech “merely because public officials disapprove the speaker’s views.”[208] That formulation certainly indicates that judges should be on guard against content discrimination to ensure it does not mask the imposition of orthodoxy. It also supports the conclusion that content discrimination can sometimes be enough to warrant the application of strict scrutiny. But it falls short of suggesting a reflexive rule that treats content-based laws as presumptively unconstitutional, irrespective of context and regardless of whether concerns about orthodoxy are in play.
In sum, Reed is on firm ground to the extent it counsels vigilance to ensure that content discrimination does not mask the imposition of orthodoxy. The case is also instructive in confirming that in some situations, a ban on all speech addressing a particular topic can manipulate the marketplace of ideas.[209] Yet we ought not interpret Reed as depicting all instances of content discrimination as equally nefarious, lest the case be read as revolutionizing the law in a manner it did not purport to do.[210]
Notwithstanding the Supreme Court’s pairing of concerns about content discrimination with concerns about viewpoint discrimination, separating the two concepts is crucial. This is true not only for the pursuit of analytical clarity but also for purposes of identifying situations in which content discrimination is permissible so long as viewpoint neutrality is preserved.[211]
Content discrimination arises from the regulation of speech based on its subject matter and “communicative impact.”[212] It is the converse of content neutrality, which the government maintains by regulating speech without regard to subject matter.[213] On the standard account, viewpoint discrimination is a strain—an especially virulent strain—of content discrimination.[214] A law discriminates on the basis of viewpoint when it “single[s] out a subset of messages for disfavor based on the views expressed.”[215] Laws that promote or burden unwelcome ideas are antithetical to the First Amendment, which recognizes viewpoint neutrality as integral to expressive liberty.[216]
For years, viewpoint neutrality received relatively little elucidation at the Supreme Court. The Justices affirmed the dangers of viewpoint discrimination without fully defining the concept. In 1987, Robert Post questioned the “cogency” of the distinction between content and viewpoint, remarking on its murkiness in application.[217] Until that time, the Court had not made much effort to crystallize the two ideas, let alone to chart the space between them. Like early accounts of obscenity, discussions of viewpoint discrimination reflected an “I know it when I see it” approach to constitutional jurisprudence.[218] There was widespread agreement over the importance of the enterprise, but less attention directed at defining its scope.[219] And though Justice Jackson’s canonical challenge to governmental efforts at declaring “what shall be orthodox” was elegant and powerful, it was not particularly substantive.[220] The difficult analytical work—of defining viewpoint, and of distinguishing viewpoint from content when that distinction matters—remained for another day, and indeed another generation. When answers finally began to come, they reflected a broad vision of viewpoint discrimination and a correspondingly narrow domain in which pure content discrimination carries a significant risk of suppression.
Before the 1990s, it was possible to describe a viewpoint as a reflection of one side of a debate. Understood in this way, speech that supports abortion rights expresses a viewpoint, and speech that supports restrictions on abortion expresses a different viewpoint. Speech in favor of affirmative action programs expresses a viewpoint, and speech critical of those programs expresses a different viewpoint. And the government discriminates on the basis of viewpoint when it hinders (or helps) speakers on one side of an issue.
Whatever its merits as a conceptual matter, this understanding of viewpoint discrimination has become untenable as a description of prevailing law. The foundational case is Rosenberger v. Rector & Visitors of the University of Virginia, which involved a university policy that subsidized printing expenses for student publications.[221] There is no categorical prohibition against the government’s commitment of resources to encourage discussion and debate.[222] Likewise, university officials certainly may provide money to support student activities. The wrinkle in Rosenberger was the scope of the University of Virginia’s funding policy. The school denied funding for certain types of publications: namely, those that “primarily promote[] or manifest[] a particular belie[f] in or about a deity or an ultimate reality.”[223] The University’s program did not purport to ban any publications; students were free to write whatever they wished. But if they wanted a subsidy, they had to play by the rules.
There is room to contend that policies like the University of Virginia’s do not even raise a constitutional question, let alone commit a constitutional violation. After all, the University had no obligation to subsidize printing costs. It could have withheld money from every student publication on campus. Moreover, students were under no compulsion to accept University funding. If student groups chose to write about topics that disqualified them from the subsidy program, the argument goes, they had no cause to complain.[224] But existing law runs counter to this position; courts review discretionary spending programs to ensure that the government does not overstep and that recipients are not bulldozed by public officials wielding nearly limitless resources.[225]
As noted above, when governments use public resources to facilitate speech in defined and targeted ways, courts generally apply forum analysis to evaluate the validity of the attendant speech restrictions.[226] Within a limited public forum, the government may discriminate on the basis of content as much as it wishes.[227] It need only behave reasonably[228] and remain neutral on the dimension of viewpoint.[229] In other words, once the government creates a forum, it must allow discussion to proceed in a free and fair manner, without putting a thumb on the scale in favor of preferred perspectives.
In Rosenberger, the Supreme Court employed forum analysis to address the tension between the University’s granting of subsidies and its imposition of restrictions.[230] The case turned on the exclusion of publications reflecting a religious perspective. If that exclusion served to define permissible subject matter within the forum, it was constitutionally permissible.[231] If the exclusion barred a particular view, it was constitutionally invalid.[232] Everything depended on whether singling out publications that focused on beliefs “about a deity or an ultimate reality” represented a restriction based on content or viewpoint.[233]
In the Court’s estimation, the University’s limitation penalized a disfavored viewpoint. The Court acknowledged the murkiness of the distinction between content and viewpoint.[234] On the facts before it, though, the Court was confident that the University had marginalized religious points of view. While religious thought is a “vast area of inquiry,” it nevertheless reflects a “specific premise” and “perspective.”[235] Justice Souter disagreed, emphasizing that the University had not burdened one side of any debate.[236] But the majority rejected the conclusion that the restriction of multiple perspectives demonstrates viewpoint neutrality. Under Rosenberger, religious speech is not simply a set of topics. It is a distinct, albeit capacious, viewpoint.[237]
In plotting the trajectory of viewpoint discrimination as a constitutional concern, analysis of Rosenberger offers two lessons. The first is that speaking from a religious perspective reflects a particular viewpoint. The second, broader lesson is that viewpoint discrimination can arise even when the government restricts a diverse array of speakers who advocate different, and competing, positions.
In the years immediately following Rosenberger, the Supreme Court found little occasion to grapple with the definition of viewpoint, even as it applied the concept from time to time.[238] That changed in 2015, in a dispute over trademark registration. Matal v. Tam involved a rock band called “The Slants” and an attempt to trademark the band’s name.[239] The band’s lead singer did not deny that the term “slant” was used as a slur against people of Asian descent. But he hoped to reclaim the word and provide it with a positive connotation.[240]
Whether to register the trademark was, in the first instance, a matter for the Patent and Trademark Office (PTO). The relevant statute prevented the PTO from registering any trademark that disparaged others.[241] Applying the disparagement bar to the application before it, the PTO denied the request for registration.[242] The band’s lead singer responded by challenging the PTO’s decision on constitutional grounds.[243] In a fractured decision, the Supreme Court found the disparagement bar to violate the First Amendment.[244]
Unlike Rosenberger, in which the Court was confident that the relevant analytical rubric came from the limited public forum doctrine,[245] Tam generated considerable uncertainty regarding the appropriate legal framework. One possibility was that federal registration of a trademark amounts to the government’s own speech; by registering a trademark, public officials effectively bless the trademarked phrase and ordain the mark as suitable.[246] Characterizing communications as the government’s own speech is significant from a doctrinal perspective, because when the government speaks or deputizes others to communicate on its behalf, it is free to formulate its message however it wishes—even by elevating certain views over others.[247] Some of the Court’s cases had embraced a capacious view of government speech, making it conceivable that trademark registration fit the mold.[248] But the Justices in Tam concluded that government speech doctrine is not the right fit for federal trademark registration.[249]
In concluding that the registration of trademarks does not amount to government speech, the Tam Court ensured that the prohibition against viewpoint discrimination—which applies when the government regulates other speakers’ messages—remained a salient concern. The Justices made little attempt to sweep up the conceptual debris thrown off by the collision of free speech and trademark law. Instead, they rested on the imperative of viewpoint neutrality, coupled with an expansive definition of viewpoint. For Justice Alito and the three others who joined him, the fact that the relevant restriction “evenhandedly prohibits disparagement of all groups” could not save it.[250] The point of view the government had singled out was broader: it was the very notion of “[g]iving offense.”[251] Justice Kennedy’s opinion, also joined by three Justices, was in accord. He concluded that the prohibition against viewpoint discrimination serves to “protect[] the right to create and present arguments for particular positions in particular ways.”[252] The disparagement bar impaired the exercise of that liberty.
Even with Rosenberger on the books, this outcome was not a foregone conclusion. It was possible to think of offensive, disparaging trademarks as defined by their content rather than any particular viewpoint. So long as the government remained evenhanded within the universe of offensive speech, it arguably steered clear of viewpoint discrimination. Tam rejected this position, enlarging the concept of viewpoint to encompass the degree of propriety with which messages are conveyed.[253] Though the galaxy of offensive speech is vast, a unified mindset connects it.
Tam was unique in that it afforded robust constitutional protection to disparaging speech, but in a case where no offense was intended. Though the applicant in Tam employed a racial slur, he did so in hopes of redirecting the term toward a positive purpose.[254] Whether or not one agreed with the applicant’s means, his approach implicated important and nuanced debates about the appropriate uses (if any) of racially disparaging words. There was no guarantee that future cases would raise similarly weighty issues.
Enter FUCT apparel. An entrepreneur started a clothing line known as FUCT.[255] When he sought to register FUCT as a federal trademark, the PTO rebuffed him based on the bar against “immoral[] or scandalous” marks.[256] That bar applied to FUCT given its vulgarity and offensiveness.[257] The dispute eventually found its way to the Supreme Court in Iancu v. Brunetti, the latest chapter in the saga of free speech and vulgar language.[258] Moving past the fractured opinions of Tam, the Justices rallied around the conclusion that the trademark system once again had disfavored a particular point of view.[259] To penalize immoral or scandalous marks, the Court explained, is to privilege perspectives “aligned with conventional moral standards” over “those hostile to them.”[260] The implication is that the universe of immoral or scandalous expressions is held together by hostility to established conventions. The viewpoint at issue is, in effect, the inclination to be outrageous, whether out of a sincere commitment to the sentiments conveyed, a provocateur’s desire to stir tumult, or the simple hope of making a buck. The Court added that in the trademark context, the threat of viewpoint discrimination had been borne out in practice via the government’s refusal to register marks taking controversial positions on issues such as “drug use, religion, and terrorism” even as it approved marks “expressing more accepted views on the same topics.”[261]
Some Justices in Brunetti noted a distinction between controversial or outrageous speech on the one hand and base vulgarity on the other. Justice Alito described Congress’s prerogative to draft a statute denying registration to “marks containing vulgar terms that play no real part in the expression of ideas.”[262] Chief Justice Roberts[263] and Justices Breyer[264] and Sotomayor[265] agreed, though they articulated their reasoning somewhat differently. But the majority opinion put aside the status of vulgarity to concentrate on discrimination against unconventional or controversial perspectives.[266] Brunetti caps off the trilogy that began with Rosenberger and ran through Tam in defining viewpoint discrimination as a concept broader than the punishment of one side of a debate.
Even with Rosenberger, Tam, and Brunetti on the books, it remains a challenge to furnish a comprehensive definition of viewpoint discrimination. We can, however, identify a few guiding principles:
- Restricting speech on one side of a debate while leaving the other side unencumbered is viewpoint discrimination.
- Restricting offensive speech while permitting more innocuous speech is viewpoint discrimination.
- Restricting speech that violates conventional moral standards while permitting speech that conforms to those standards is viewpoint discrimination.
These principles must be situated within the broader context of First Amendment case law. As we have seen, some types of offensive speech—such as fighting words, meaning personal insults that raise the prospect of an immediate breach of the peace—are subject to restriction as a categorial exception to full First Amendment protection.[267] Asking whether a restriction on fighting words is content based, or even viewpoint based, is beside the point. The same appears to be true of actions for intentional infliction of emotional distress when the underlying speech bears on matters of private concern.[268] Likewise, obscenity is speech that violates certain moral standards, yet restrictions on obscenity do not face the same scrutiny as other content-based or viewpoint-based laws. Rather, as another categorical exception to full protection, obscenity triggers a distinctive type of First Amendment analysis.[269]
Nothing in the Supreme Court’s recent cases on viewpoint discrimination calls these categorical exceptions into question. The exceptions play by specialized sets of rules for reasons of history and tradition.[270] Outside of the exceptions, the primacy of viewpoint neutrality remains intact. The most significant takeaway from the Court’s pronouncements is that viewpoint discrimination encompasses more than burdening one side of a debate. Whether viewpoint is at issue does not depend on the number of perspectives that a law restricts, nor even the relationship between them. The touchstone is commonality of perspective.
IV. The Viewpoint-Centric First Amendment
Part II described viewpoint neutrality as the heart of First Amendment jurisprudence. Part III added that under existing law, the judicial conception of viewpoint discrimination is vast. Taken together, the two Parts offer an important lesson: The First Amendment is most worried about viewpoint discrimination, and viewpoint discrimination is most of what the First Amendment is worried about.
This Part presents a revised account of the law of speech with viewpoint neutrality at the center. Content discrimination remains relevant, but not as an imperative whose violation ineluctably leads to strict scrutiny and likely invalidation. Instead, content-based laws within traditional public forums are a red flag. Content discrimination should continue to put judges on guard. But when a speech restriction raises no concerns about official suppression or ordained orthodoxy, a less demanding, intermediate standard of review ought to apply.
A. Demonstrating Viewpoint Centricity
Supporting or punishing speakers on one side of an issue in order to manipulate the marketplace of ideas is viewpoint discrimination in its most blatant form.[271] But the lesson of the past four decades is that viewpoint discrimination can also arise when the government punishes multiple speakers who stand on opposing sides of an issue.[272] A law against offensive speech is viewpoint discriminatory even if it applies in like fashion to those who support or oppose abortion restrictions, gun rights, labor strikes, military engagements, and so on. Whatever the speaker’s cause or topic, restricting offensive or immoral expression violates the precept of viewpoint neutrality.[273]
The prevailing definition of viewpoint discrimination suggests an abiding concern with government neutrality among standpoints, not subjects. Recall that in Mosley, the restriction at issue included an exception for “the peaceful picketing of any school involved in a labor dispute.”[274] Likewise, the carve out in Carey permitted “the peaceful picketing of a place of employment involved in a labor dispute.”[275] The Supreme Court has cited these cases for the proposition that content discrimination triggers strict scrutiny, irrespective of the reasons for or ramifications of that discrimination.[276] But Mosley and Carey are best understood as examples of viewpoint discrimination.[277] Speaking on behalf of employees in a labor dispute reflects not just a topic, but one side of a debate. The government may not restrict disfavored perspectives while exempting favored ones in order to “tilt public debate in a preferred direction.”[278]
More recent cases point toward the same conclusion. A California law limiting the purchase of ultra-violent video games responded not only to those games’ subject matter, but to their extreme and outrageous presentation in a manner at odds with established norms of propriety.[279] The issue was not content discrimination but viewpoint discrimination; the offensive trademarks in Tam and Brunetti find their parallel in bracingly violent games.
Canonical First Amendment cases are in accord, particularly when one views them in light of recent discussions of viewpoint discrimination. Consider Cohen v. California, the famous “Fuck the Draft” case decided in 1971.[280] Looking back through the lens of later cases like Rosenberger, Tam, and Brunetti, punishing Paul Robert Cohen for his “offensive” speech violated the prohibition against viewpoint discrimination.[281] Cohen was not simply about the content of the message carried on a jacket. It was about something more: the stifling of a message expressed in a convention-disrupting manner.[282] Similarly, when the Court invalidated mandatory flag salutes in West Virginia State Board of Education v. Barnette, it depicted the First Amendment as a bulwark against the government’s determination of “what shall be orthodox.”[283] So, too, in Texas v. Johnson, where the Court protected a constitutional right to burn the American flag rather than authorizing the government to “prohibit the expression of an idea simply because society finds the idea itself offensive or disagreeable.”[284] These landmark cases add force to the claim that it is viewpoint, not content, that matters most.
To be sure, some content-based laws trigger strict scrutiny—but not based on concerns about content discrimination per se. Consider the world of campaign-finance regulation, in which strict scrutiny applies to restrictions on electoral advocacy.[285] A prohibition against robust discussion of an upcoming election, though evenhanded in theory, might systematically favor incumbents and burden lesser-known challengers seeking to upend the status quo.[286] Other limitations on political advocacy might reflect a governmental judgment about which sources of information are reliable and which are not.[287] These concerns justify the application of strict scrutiny when the government seeks to suppress debate about elections, whether or not viewpoint discrimination is overt.
Likewise, strict scrutiny would be the proper rubric if the government were to stifle discussion of a pressing social issue. Imagine that Congress enacts a statute prohibiting, on pain of fine, any discussion of the efficacy of vaccines. The ban would be evenhanded in the sense of preventing statements that support vaccines as well as statements that cast doubt upon them. Even so, alarm bells would ring. For starters, the ban would deplete the marketplace of ideas as it relates to a matter of public concern. The constitutional primacy of public concern speech is an established feature of First Amendment jurisprudence,[288] and it can be instructive in informing the identification of “invidious” content discrimination.[289] Assessing the hypothetical ban on vaccine speech also requires taking stock of the speech that is left over: namely, speech by the government itself. Were the government to silence all private voices on an important issue, it could create a monopoly on discourse. The void left by the silencing of private speakers might be filled by public officials, opening a path to the promotion of state-sponsored orthodoxy. The same concerns would arise from a ban on discussing other matters of public concern, from abortion to firearms to immigration. Even if these concerns seem remote, it makes sense to leave open the prospect of subjecting some content-based laws to the aggressive scrutiny that applies to viewpoint-based restrictions.
Yet an automatic, across-the-board rule of strict scrutiny for content-based rules is problematic both doctrinally and theoretically. The foregoing examples of limits on political speech and other discussions of public concern are leagues apart from a city ordinance that restricts most signs but makes an exception for ones that announce hidden driveways. Where a content-based law avoids matters of public concern and offers no indicia of compelled orthodoxy, it might well be benign.[290] At very least, the judiciary must pause before jumping straight to strict scrutiny and likely invalidation. To conclude that every content-based regulation creates the same danger of suppression would be to extend the rule of strict scrutiny beyond its rationale.
This Section has focused on the treatment of content discrimination under a viewpoint-centric approach to constitutional adjudication. It is also worth clarifying the degree of scrutiny that applies to viewpoint discrimination itself.[291] One possibility is that viewpoint discrimination is invalid as a matter of law.[292] Yet the Supreme Court has been reluctant to embrace absolutes in the world of free speech.[293] In light of this tendency, the better understanding may be that viewpoint discrimination triggers a type of review that is structurally similar to the strict scrutiny that follows content-based laws, but with an even higher bar for sustaining the government’s action.[294] On this account, viewpoint discrimination leads to the application of strict scrutiny in a highly aggressive form. Speech restrictions that discriminate on the basis of viewpoint almost always will fail, but not due to a rule of per se unconstitutionality. The real driver is the surpassing difficulty even of imagining some suitable justification for overriding the core First Amendment principle that every speaker is entitled to express his or her view.
B Recognizing Content as Proxy
Accepting viewpoint neutrality as the First Amendment’s infrastructure leaves much of the doctrinal superstructure intact. As the previous Section explained, content discrimination sometimes provides a reliable proxy for the stifling of unwelcome views or the advancement of governmental orthodoxy.[295] Where it does, the application of strict scrutiny remains appropriate.[296] As Justice Kagan observed in Reed, this link between content and viewpoint “is by no means new.”[297]
A proxy, though, is different from an irrebuttable presumption.[298] It is possible to defend a rule of strict scrutiny for all content-based restrictions as a prophylactic. The rationale would be that because such laws sometimes reflect an effort to punish or marginalize disfavored perspectives, the “safest and most sensible course” is to treat the laws as constitutionally suspect across the board.[299] Overinclusive rules can prove useful when it is difficult to draw fine distinctions.[300] Maybe a rule requiring strict scrutiny for all content-based laws has something to recommend it, notwithstanding the possibility of false positives in the form of benign regulations that courts will strike down.[301]
The problem with this argument is that in the context of speech regulation, no such prophylactic rule is necessary. Courts commonly will be able to determine whether a content-based law raises concerns about the imposition of orthodoxy. Take Reed,[302] for example. The ordinance at issue, regulating the dimensions of various signs, was byzantine and arbitrary.[303] Those failings warranted its invalidation even under an intermediate standard of scrutiny. Still, it is easy to imagine other regulations that are reasonable, free of any orthodoxy-imposing tendencies, and still content based. Hence Justice Kagan’s recognition that laws across the country make exceptions for signs that announce hidden driveways, post a home’s address, identify historic sites, and so on.[304] Though he joined the majority opinion in Reed, Justice Alito offered similar sentiments in his concurrence, looking for a way to uphold reasonable efforts at regulation.[305]
At times it may be challenging to determine whether a content-based law distorts the marketplace of ideas in service of a privileged worldview. In those instances, it is both sensible and consistent with the case law to resolve doubts in favor of expressive liberty, such that close calls are treated as triggering strict scrutiny.[306] But in other instances, it will be plain enough that even though a regulation depends on the content of speech, there is “no realistic possibility that official suppression of ideas is afoot.”[307] The law of free speech need not treat innocent restrictions as engines of suppression.
Framed in this way, content discrimination becomes a red flag rather than a disqualifying feature. Content discrimination might mask or facilitate official suppression. But it might not. While it is important to remain on guard, it is equally important to acknowledge that some content-discriminatory laws will turn out to be benign—and thereby subject to a less demanding degree of scrutiny.
Inquiring into official suppression of ideas need not entail attempting to uncover government officials’ secret motivations.[308] The question is whether the government’s proffered justification for a speech restriction depends on the imposition of orthodoxy. Cohen, for instance, could not be jailed for wearing his vulgar jacket based on the government’s desire to maintain “a suitable level of discourse within the body politic.”[309] Situations may also arise in which a law’s orthodoxy-opposing derivation is plain on its face, and in those cases it is scarcely necessary to consider the law’s justifications. When Texas enacted a law protecting “[v]enerated [o]bject[s]” from “[d]esecration” in order to spare observers from offense, and when it applied that law to expressive conduct like flag burning, it promoted a preferred orthodoxy via the punishment of speech.[310] By contrast, in the more common case in which underlying governmental motive is a matter of speculation, the focal point will be the justification for the law at issue, which furnishes the best indication of what the law pursues and what its effects on expressive liberty will be.[311]
Reed once again is illuminating. The Town of Gilbert’s system of regulation, with its numerous distinctions among types of signs, lacked a reasonable basis, and it was vulnerable to judicial invalidation on that ground alone; intermediate scrutiny is not a rubber stamp.[312] For Justices Breyer, Ginsburg, and Kagan, there was no warrant for going further by subjecting the ordinance to strict scrutiny.[313] On their rationale, the fact that a law draws content-based distinctions is relevant, but not dispositive.[314] Even after finding that a law discriminates on the basis of content, a court must inquire into the law’s justifications.[315] Some content-based distinctions facilitate efforts to penalize certain viewpoints. Those distinctions are subject to strict scrutiny and face likely invalidation. But in other circumstances—like those presented in City of Austin, with its differential treatment of on-premises and off-premises signs[316]—governments regulate the content of speech in ways that present no realistic threat of suppressing disfavored ideas.[317] The powerful medicine of strict scrutiny unduly interferes with sensible regulations of the latter type. Just as intermediate scrutiny applies to regulations that are not based on the communicative impact of speech,[318] the same test can and should apply to benign content discrimination when the government makes no attempt to “drive certain ideas or viewpoints from the marketplace.”[319] Where the government has a reasonable basis for taking the content of speech into account, and where there is no “realistic possib[ility]” that the government is “favoring some ideas over others,” sensible regulations ought to stand a chance notwithstanding their curtailment of speech.[320]
C. Explaining the Ubiquity of Content Discrimination
Shifting the focus of First Amendment jurisprudence from content to viewpoint offers a way to explain why content discrimination continues to flourish without causing any evident judicial consternation. Frederick Schauer has noted that fields such as securities regulation, antitrust law, and products liability are awash in content-based speech restrictions.[321] Even so, the judiciary has not treated those restrictions as raising substantial First Amendment issues, let alone as infringing expressive liberty protected by the Constitution. Any claim that such matters do not involve the restriction of “speech” is unavailing. If speech is defined in the literal sense, the statement is obviously false. And if it is used as a legal term of art, it begs the question of what the First Amendment is properly understood to cover. In the years ahead, similar debates are sure to proliferate over issues such as data privacy[322] and sexual privacy,[323] both of which give rise to regulatory efforts that pursue vital objectives through content-based restrictions on speech. As Justice Breyer has observed, because so much regulation turns on the content of speech, inexorably linking content discrimination with strict scrutiny is a “recipe for judicial management of ordinary government regulatory activity.”[324]
In addition, as explained in Part II, the Supreme Court has crafted numerous First Amendment doctrines that depend on the content of speech but that do not entail the application of strict scrutiny.[325] If the First Amendment’s imperative is content neutrality, this phenomenon is puzzling. But from a viewpoint-centric perspective, it makes perfect sense. The government may treat commercial speech differently from other speech so long as it does not “tilt public debate” in the direction of favored viewpoints.[326] It likewise may forbid defamation or obscenity if it does not inject biases and orthodoxies—which is to say, if no “official suppression of ideas is afoot.”[327] These principles capture the state of First Amendment doctrine better than an unflagging insistence on content neutrality. Absent governmental imposition of orthodoxy, the law has a ready response to restrictions on speech: apply intermediate scrutiny. Intermediate scrutiny does not afford the government unchecked discretion. It simply means that reasonable and measured limitations on speech have a realistic prospect of survival.[328]
D. Harmonizing Secondary Effects
Recognizing viewpoint neutrality as the First Amendment’s touchstone provides a means of integrating the doctrine of secondary effects with other strands of the law of free speech. Secondary effects doctrine applies intermediate scrutiny to laws that reflect legitimate responses to the indirect consequences of speech.[329] On a content-centric approach, the doctrine of secondary effects is perplexing. The relevant cases involve distinctions based on the content of speech: whether a theater or bookstore was allowed to locate in a particular place depended entirely on the nature of its movies or books.[330] There is no denying the role of content in determining the application of those laws.[331] Nevertheless, the Supreme Court treated the laws as if they were content neutral, reasoning that the government was pursuing goals other than the stifling of disfavored ideas.[332]
The secondary effects cases have always tugged against the idea that content-based laws are presumptively invalid. The tension increased with Reed, which drew a direct line between content discrimination and the application of strict scrutiny.[333] But once we recognize that the problem with content-based laws is their facilitation of viewpoint discrimination, the role of secondary effects becomes clear. The government has discretion to address societal problems, even if its interventions cause difficulties for certain speakers, so long as its actions are reasonable and viewpoint neutral.[334] Whether a law draws distinctions based on the content of speech is, in a word, secondary. What matters is whether the government has suppressed disfavored ideas in pursuit of a preferred orthodoxy.
E. Distinguishing Government Speech
Most of the time, finding a law to be viewpoint discriminatory is the end of the case. Indeed, in Matal v. Tam[335] the Supreme Court could strike down a provision of trademark law as viewpoint discriminatory without even deciding which First Amendment rubric applied.[336] It sufficed that the Justices agreed on “a core postulate of free speech law: The government may not discriminate against speech based on the ideas or opinions it conveys.”[337]
Notwithstanding the gravity of viewpoint-discriminatory restrictions on speech,[338] the government possesses authority to promote preferred viewpoints when it advances its message in other ways. In Rust v. Sullivan, for example, the Supreme Court upheld a government initiative that provided public funds for family planning programs but prohibited doctors from giving advice about abortion while working within those programs.[339] The government was disfavoring the viewpoint that abortion is an appropriate topic in certain conversations about pregnancy, birth, and family planning.[340] Still, its approach was permissible.[341] When the government offers funds to support a preferred viewpoint, it is not required to allocate funds to advance other perspectives.[342] So long as the recipients of public money remain free to voice their opinions outside the program in question, the government’s decision to elevate some perspectives over others creates no constitutional problem. Only when the government takes a different approach by creating a forum for private expression does the requirement of viewpoint neutrality arise.[343] A viewpoint-centric approach to the First Amendment leaves this dichotomy undisturbed. The government retains the power to speak on its own behalf, and to deputize others to carry its message.
F. Rationalizing First Amendment Theory
The foregoing Sections explained how recognizing viewpoint neutrality as the First Amendment’s focal point solves several puzzles of free speech doctrine. Making the conceptual move from content to viewpoint is also useful in assessing the prevailing trends of First Amendment theory.
Concerns about content discrimination can be difficult to square with the notion that the First Amendment’s primary goal is to promote a robust marketplace of ideas.[344] In many cases, governments engage in content discrimination in order to reduce the amount of speech that a regulation curtails. A law might prohibit all residential picketing except labor picketing, decreasing the amount of speech that is restricted as compared with an all-out ban.[345] Or it might impose a general prohibition against robocalls, but create an exception for calls that relate to certain topics.[346] Or limit the digitization of some outdoor signs, but not others.[347] When the government carves out some speech for protection instead of enacting a broader prohibition, it can seem strange for courts to impose a penalty by applying strict scrutiny.[348] To be sure, it is possible that the government might sometimes lack the will or political capital to enact an across-the-board ban, such that the alternative to enacting a restriction with exceptions is enacting no restriction at all. Even if that argument holds in some circumstances, there is no reason to believe it is universally true. Absent concerns that the government is imposing orthodoxy through the suppression of speech, judicial skepticism of speech-protective exceptions can be counterproductive.
The more direct concern with content discrimination is that, irrespective of its effect on the amount of expression in circulation, it facilitates favoritism. The Supreme Court has indicated that the foundation of modern First Amendment law is distrust of government.[349] The point is not simply to populate the marketplace with as many ideas and utterances as possible. The robustness of the market is a byproduct of keeping the government where it belongs: on the sidelines, as opposed to in the middle of the action. This conception of the First Amendment is subject to dispute.[350] But as a descriptive matter, it continues to prevail.
A distrustful court might be skeptical of all content-based distinctions, even those that create exemptions from speech restrictions. Yet not every content-based rule reveals ideological foul play. Content-based distinctions can reflect good-faith (if occasionally misguided) efforts at addressing societal problems without suppressing more speech than is necessary.[351] Just as the Supreme Court has been willing to uphold underinclusive laws absent any “indication that the selective restriction of speech reflects a pretextual motive,” so, too, should it relax its scrutiny of content-based laws when there is no reason to fear that orthodoxy is around the corner.[352] Only in the face of viewpoint discrimination should the heavy artillery of strict scrutiny enter the fray.
By embracing this type of context-sensitive approach, the Court would bolster the coherence of both First Amendment theory and First Amendment doctrine. Distrust of government explains why judges ought to be skeptical of favoritism predicated on a speaker’s perspective. But when there is no reason to worry about privileging some speakers via the restriction of others’ speech, the need to keep government in check is less urgent. Reasonable efforts at regulation become potentially valid, precisely because there is no indication that even a distrusted government is stifling speech in pursuit of its preferred orthodoxy.
V. Making the Transition from Content to Viewpoint
As Part IV explained, a reorientation toward viewpoint neutrality leaves much of the law of free speech intact. Even so, the analytical shift carries important ramifications for certain rules as well as their rationales. That is particularly true given the Supreme Court’s aggressive stance toward content discrimination in Reed, which City of Austin did not disavow.[353] The transition accordingly implicates the doctrine of stare decisis, which counsels against disturbing established rules absent a “special justification” for departure.[354] The treatment of precedent, including constitutional precedent, has received considerable attention in the case law and scholarly literature, and this Article is not the occasion for a full exploration.[355] Nor is a full exploration necessary to the task at hand, for it turns out that the Court has recently brought about two doctrinal shifts in the law of free speech that provide ready models for initiating a similar move from content centricity to viewpoint centricity.
The first template involves the exceptions to full First Amendment protection described in Part II. I have emphasized the importance of those exceptions in challenging the narrative of content centricity. It is also instructive to consider the exceptions’ theoretical underpinnings, which have shifted over time.[356] The inflection point was United States v. Stevens, arising out of a conviction for distributing videos that depicted violence against animals.[357] In defending the constitutionality of the relevant statute, the government urged the recognition of a categorical exception for speech whose costs vastly outweigh its benefits.[358] That exception would allow the government to prohibit and punish depictions of extreme cruelty to animals, just as it may prohibit and punish defamation, obscenity, incitement, and fighting words.
The government’s theory had a basis in the case law. Over the years, the Supreme Court had described exceptions to full protection as imposing costs that seriously outweigh their benefits.[359] Even so, when the Court encountered the government’s argument in Stevens, it issued a forceful rebuke. The Court explained that it is not enough to find that a category of speech imposes costs in excess of benefits. To treat constitutional protection as dependent on that calculus would be “startling and dangerous.”[360] At the same time, the Court reaffirmed the status of the categorical exceptions that pepper its cases.[361] It explained that, notwithstanding various opinions’ language about costs and benefits, the exceptions should be understood as reflecting a tradition of denying full protection to certain categories of speech.[362] That approach would also guide the evolution of the law in the future. Going forward, requests to recognize categorical exceptions must rest on a showing that the relevant speech has been “historically unprotected.”[363] With those observations, the Court revised an important area of free speech jurisprudence based on an examination of underlying theory.
The concept of First Amendment underinclusivity offers a second template to inform the focal shift from content to viewpoint. In the parlance of the Supreme Court, a law is underinclusive if it fails to regulate certain speech that creates the same risk of harm as other speech being regulated.[364] So, for example, a restriction on residential protesting is underinclusive if it exempts labor-related speech, which raises the same concerns as residential protesting on other topics.[365] For years, underinclusivity posed a devastating problem for speech restrictions.[366] That changed in 2016 with Williams-Yulee v. Florida Bar.[367] Like many states, Florida selects judges via popular election.[368] Candidates are subject to the Florida Code of Judicial Conduct, which promotes the integrity and independence of the judiciary. Among the Code’s provisions was a prohibition against personal solicitation of donations by candidates or incumbent judges.[369] The ban on personal solicitation left open other options for fundraising, such as solicitation by campaign committees.[370]
When Williams-Yulee wound its way to the Supreme Court, the Justices pondered the disparity between personal solicitation and solicitation by campaign committees. According to the candidate who challenged the relevant restrictions, Florida’s decision to allow solicitation by campaign committees rendered the restrictions on candidate speech underinclusive.[371] If Florida really wanted to ensure the integrity of its judiciary, the argument went, it would not have tolerated such a simple workaround.
In addressing this argument, the Supreme Court redefined the nature of underinclusivity analysis. Underinclusivity may be cause for concern, but it is not necessarily a fatal defect. Rather than a road to automatic disqualification, underinclusivity is a “red flag.”[372] It puts the reviewing court on notice that the government may be disfavoring certain perspectives or pursuing an objective that is not particularly compelling.[373] Upon further inspection, if the court is convinced that those dangers are not present, the law may survive despite its underinclusivity. The rationale is straightforward: While underinclusivity sometimes cloaks the imposition of orthodoxy or suggests the lack of a compelling justification for government action, at other times it reflects a benign—perhaps even laudable—effort to pursue an important goal without restricting more speech than is necessary.[374] On the facts of Williams-Yulee, there was no reason to suspect any nefarious plan on the part of the government, nor was there any doubt about the importance of the interests the government was pursuing.[375] As a result, the law’s underinclusivity was excusable.[376]
Rethinking the role of content discrimination could follow the same pattern as recent innovations in thinking about the categorical exceptions to full protection and the doctrine of underinclusivity. Courts should continue to take a close look whenever the government is reacting to speech based on its content. But the reason for scrutiny is not the fear of content discrimination per se. The question is whether a given instance of content discrimination suggests bias against unwelcome viewpoints or imposition of a preferred orthodoxy. So long as viewpoint neutrality exists, content discrimination is not inherently problematic.[377] The Court’s recent cases teach that one cannot apply the exceptions to full constitutional protection or underinclusivity analysis without being sensitive to conceptual underpinnings. The move to viewpoint centricity entails the same transcendence of superficial labels to engage with the deeper theory behind the First Amendment.
Whether the government has restricted speech based on content is relevant, but ancillary. It is viewpoint neutrality that resides at the heart of the First Amendment. The law of free speech revolves around the liberty to express one’s views.
-
* Diane & M.O. Miller II Research Professor of Law, Notre Dame Law School. For helpful comments and conversations, thanks to William Baude, Joseph Blocher, Marc DeGirolami, Nicole Garnett, Richard Garnett, David Han, Marah McLeod, and Alexander Tsesis. Elisabeth Crusey and Joshua Lacoste provided excellent research assistance. ↑
-
. See Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022). ↑
-
. See N.Y. State Rifle & Pistol Ass’n, Inc. v. Bruen, 142 S. Ct. 2111 (2022). ↑
-
. See United States v. Zubaydah, 142 S. Ct. 959 (2022). ↑
-
. See City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464, 1469–70 (2022) (argued Nov. 10, 2021); Reagan Nat’l Advert. of Austin, Inc. v. City of Austin, 972 F.3d 696, 700 (5th Cir. 2020). ↑
-
. See Reagan, 972 F.3d at 699–700. ↑
-
. Austin, Tex., City Code § 25-10-4(9) (2017). ↑
-
. See Brief for Respondent at 12–13, City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464 (2022) (No. 20-1029). ↑
-
. See Brief for Petitioner at 10, City of Austin v. Reagan Nat’l Advert. of Austin, Inc., 142 S. Ct. 1464 (2022) (No. 20-1029). ↑
-
. See Transcript of Oral Argument at 46, 57, 70, City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464 (2022) (No. 20-1029). ↑
-
. Id. at 35. ↑
-
. See City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464, 1475–76 (2022). ↑
-
. Compare id. at 1471 (rejecting the view that “a regulation cannot be content neutral if it requires reading the sign at issue”), with id. at 1484 (Thomas, J., dissenting) (contending that a law is content based if it requires an official to “‘determine whether a sign’ conveys a particular message”) (quoting Reed v. Town of Gilbert, 576 U.S. 155, 170 (2015)). ↑
-
. Reed, 576 U.S. at 163. ↑
-
. See, e.g., Brown v. Ent. Merchs. Ass’n, 564 U.S. 786, 799 (2011) (noting that strict scrutiny “is a demanding standard,” and that “‘[i]t is rare that a regulation restricting speech because of its content will ever be permissible’”) (quoting United States v. Playboy Ent. Grp., Inc., 529 U.S. 803, 818 (2000)). ↑
-
. See, e.g., Hill v. Colorado, 530 U.S. 703, 725–30 (2000) (applying a less demanding version of scrutiny to a content-neutral restriction). ↑
-
. See id. at 730 (upholding the speech restriction at issue). ↑
-
. See Transcript of Oral Argument, supra note 9, at 25 (recognizing that the City of Austin had not contended that its regulatory scheme could survive strict scrutiny). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 177 (2015) (Breyer, J., concurring in the judgment) (citing 15 U.S.C. § 78l). ↑
-
. See id. (citing 26 U.S.C. § 6039F). ↑
-
. See id. at 180 (Kagan, J., concurring in the judgment) (citing Code of Athens-Clarke County, Ga., Pt. III, § 7-4-7(1) (1993)). ↑
-
. See id. (predicting that most laws “with subject-matter exemptions . . . will be struck down” under the logic of the majority opinion). ↑
-
. See, e.g., Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 662 (1994). ↑
-
. Abrams v. United States, 250 U.S. 616, 630 (1919) (Holmes, J., dissenting). ↑
-
. See id. (“[W]hen men have realized that time has upset many fighting faiths, they may come to believe . . . that the ultimate good desired is better reached by free trade in ideas.”). ↑
-
. See, e.g., Citizens United v. FEC, 558 U.S. 310, 340 (2010) (“Premised on mistrust of governmental power, the First Amendment stands against attempts to disfavor certain subjects or viewpoints.”); United States v. Alvarez, 567 U.S. 709, 751–52 (2012) (Alito, J., dissenting) (rejecting the claim that “there is no such thing as truth or falsity in” areas such as “philosophy, religion, history, the social sciences, the arts, and other matters of public concern,” but explaining “that it is perilous to permit the state to be the arbiter of truth”). ↑
-
. Ashutosh Bhagwat, In Defense of Content Regulation, 102 Iowa L. Rev. 1427, 1428 (2017) (observing that since 1972, “the central tenet of the Supreme Court’s free speech doctrine has been that . . . ‘government has no power to restrict expression because of its message, its ideas, its subject matter, or its content’”) (quoting Police Dep’t v. Mosley, 408 U.S. 92, 95 (1972)). ↑
-
. See, e.g., Williams-Yulee v. Fla. Bar, 575 U.S. 433, 444 (2015) (“We have emphasized that ‘it is the rare case’ in which a State demonstrates that a speech restriction is narrowly tailored to serve a compelling interest.”) (quoting Burson v. Freeman, 504 U.S. 191, 211 (1992) (plurality opinion)). The case law contains some suggestions that viewpoint-discriminatory laws might be flatly “prohibited.” Minn. Voters All. v. Mansky, 138 S. Ct. 1876, 1885 (2018). I discuss this issue in Section IV.A. ↑
-
. See Matal v. Tam, 137 S. Ct. 1744, 1764 (2017) (“[T]he proudest boast of our free speech jurisprudence is that we protect the freedom to express ‘the thought that we hate.’”) (alteration in original) (quoting United States v. Schwimmer, 279 U.S. 644, 655 (1929) (Holmes, J., dissenting)). ↑
-
. City of Ladue v. Gilleo, 512 U.S. 43, 45 (1994); cf. Leslie Kendrick, Content Discrimination Revisited, 98 Va. L. Rev. 231, 294 (2012) (observing that if the presumption against content discrimination “encompasses all communication-related classifications, it could reach a great deal of regulation which, in the modern regulatory state, might seem routine, and indeed necessary”). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 163–64 (2015). ↑
-
. See N.Y. Times Co. v. Sullivan, 376 U.S. 254 (1964) (defamation); Brandenburg v. Ohio, 395 U.S. 444 (1969) (incitement). ↑
-
. See Sorrell v. IMS Health Inc., 564 U.S. 552, 571–72 (2011). ↑
-
. See City of Renton v. Playtime Theatres, Inc., 475 U.S. 41, 48–49 (1986). ↑
-
. See Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 829 (1995) (recognizing the potential validity of content discrimination based on “[t]he necessities of confining a forum to the limited and legitimate purposes for which it was created”). ↑
-
. See Sullivan, 376 U.S. at 279–80 (affording protection to defamatory speech about the official conduct of public officials so long as the defendant did not speak with knowledge of falsity or reckless disregard for the truth); Brandenburg, 395 U.S. at 447–48 (articulating the line between protected advocacy and unprotected incitement). ↑
-
. See, e.g., R.A.V. v. City of St. Paul, 505 U.S. 377, 390 (1992) (recognizing concerns over viewpoint discrimination even in situations where the government is authorized to discriminate based on the content of speech). ↑
-
. See Iancu v. Brunetti, 139 S. Ct. 2294, 2299 (2019) (describing the “core postulate of free speech law” that “[t]he government may not discriminate against speech based on the ideas or opinions it conveys”). ↑
-
. Cf. Barr v. Am. Ass’n of Pol. Consultants, Inc., 140 S. Ct. 2335, 2361 (2020) (Breyer, J., concurring in the judgment with respect to severability and dissenting in part) (“[W]hen content-based distinctions are used as a method for suppressing particular viewpoints or threatening the neutrality of a traditional public forum, content discrimination triggering strict scrutiny is generally appropriate.”). ↑
-
. R.A.V., 505 U.S. at 390. ↑
-
. See, e.g., Reed v. Town of Gilbert, 576 U.S. 155, 163 (2015) (connecting content discrimination with strict scrutiny); Williams-Yulee v. Fla. Bar, 575 U.S. 433, 444 (2015) (noting the infrequency with which speech restrictions survive strict scrutiny). ↑
-
. R.A.V., 505 U.S. at 390. ↑
-
. See infra Part II. ↑
-
. See Reed, 576 U.S. at 163–64 (2015). ↑
-
. Id. at 163. ↑
-
. Id. ↑
-
. See Kovacs v. Cooper, 336 U.S. 77, 87–88 (1949) (finding no constitutional right “to gain the public’s ears by objectionably amplified sound on the streets”). ↑
-
. See, e.g., McCullen v. Coakley, 573 U.S. 464, 477 (2014) (noting the government’s “somewhat wider leeway” to impose restrictions on the “time, place, or manner” of speech). ↑
-
. See Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989). ↑
-
. See, e.g., City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 428 (1993) (“[A] prohibition against the use of sound trucks emitting ‘loud and raucous’ noise in residential neighborhoods is permissible if it applies equally to music, political speech, and advertising.”). ↑
-
. See, e.g., Williams-Yulee v. Fla. Bar, 575 U.S. 433, 444 (2015) (noting the rarity of situations in which a speech restriction survives strict scrutiny). ↑
-
. Erwin Chemerinsky, Content Neutrality as a Central Problem of Freedom of Speech: Problems in the Supreme Court’s Application, 74 S. Cal. L. Rev. 49, 53 (2000). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 159–61 (2015). ↑
-
. See id. ↑
-
. See id. at 161. ↑
-
. See id. ↑
-
. See id. at 173; id. at 174 (Alito, J., concurring); id. at 179 (Breyer, J., concurring in the judgment); id. at 184 (Kagan, J., concurring in the judgment). ↑
-
. Id. at 184 (Kagan, J., concurring in the judgment). ↑
-
. See id. at 171 (majority opinion). ↑
-
. See id. at 159–61 (describing the respective rules for various types of signs). ↑
-
. See id. at 165 (“A law that is content based on its face is subject to strict scrutiny regardless of the government’s benign motive, content-neutral justification, or lack of ‘animus toward the ideas contained’ in the regulated speech.’”) (quoting City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 429 (1993)). ↑
-
. See id. at 166 (“[A]n innocuous justification cannot transform a facially content-based law into one that is content neutral.”). ↑
-
. See id. at 172 (finding that the sign regulation scheme failed strict scrutiny). ↑
-
. U.S. Const. amend. I. ↑
-
. See Reed, 576 U.S. at 165. ↑
-
. See id. at 163 (citing Carey v. Brown, 447 U.S. 455 (1980), and Police Dep’t v. Mosley, 408 U.S. 92 (1972)). ↑
-
. See id. at 165–66 (quoting Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 642 (1994)). ↑
-
. See id. at 163 (citing Sorrell v. IMS Health Inc., 564 U.S. 552 (2011)). ↑
-
. See, e.g., Genevieve Lakier, Reed v. Town of Gilbert, Arizona, and the Rise of the Anticlassificatory First Amendment, 2016 Sup. Ct. Rev. 233, 235 (noting Reed’s “insist[ence] that strict scrutiny applies to all laws that treat speakers differently because of the content of their speech”). ↑
-
. See Reed, 576 U.S. at 166 (“[S]trict scrutiny applies either when a law is content based on its face or when the purpose and justification for the law are content based.”). ↑
-
. See City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464, 1475 (2022). ↑
-
. See id. at 1471–72. ↑
-
. Reed, 576 U.S. at 163; City of Austin, 142 S. Ct. at 1482–84 (Thomas, J., dissenting). ↑
-
. See Daniel A. Farber, Content Regulation and the First Amendment: A Revisionist View, 68 Geo. L.J. 727, 728 (1980) (observing that the Supreme Court had upheld numerous restrictions despite their being “clearly related to the content of the regulated speech”). The Court’s decisions in recent decades have underscored Professor Farber’s point, as explained in the pages that follow. ↑
-
. For cases involving content-based reactions to the speech of public employees, see, for example, Garcetti v. Ceballos, 547 U.S. 410 (2006) (dealing with reactions to an employee’s recommendations about the handling of a legal case). For cases involving content-based reactions to the speech of students, see, for example, Bethel Sch. Dist. No. 403 v. Fraser, 478 U.S. 675 (1986) (dealing with a student’s use of indecent language). On the latter score, Justice Alito recently observed that “[a]s a practical matter, it is impossible to see how a school could function if administrators and teachers could not regulate on-premises student speech, including by imposing content-based restrictions in the classroom.” Mahanoy Area Sch. Dist. v. B.L., 141 S. Ct. 2038, 2050 (2021) (Alito, J., concurring). ↑
-
. See Farber, supra note 73, at 730 (referring to content neutrality as a “mirage”). ↑
-
. See Minn. Voters All. v. Mansky, 138 S. Ct. 1876, 1885 (2018). ↑
-
. See id. ↑
-
. See id. ↑
-
. See id. at 1886 (treating a polling place as a nonpublic forum because it is “government-controlled property set aside” for a particular purpose). ↑
-
. See Christian Legal Soc’y Chapter of the Univ. of Cal., Hastings Coll. of the L. v. Martinez, 561 U.S. 661, 679 n.11 (2010) (describing the limited public forum doctrine). The Supreme Court in Mansky referred to “three types of government-controlled spaces: traditional public forums, designated public forums, and nonpublic forums.” 138 S. Ct. at 1885. That formulation suggests an understanding of limited public forums as a subset of nonpublic forums. ↑
-
. See Christian Legal Soc’y, 561 U.S. at 681. ↑
-
. See Se. Promotions, Ltd. v. Conrad, 420 U.S. 546, 555 (1975). ↑
-
. See Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 829–30 (1995). ↑
-
. See Christian Legal Soc’y, 561 U.S. at 681 (noting that the reservation for use by certain groups is a “defining characteristic” of a limited public forum). ↑
-
. See, e.g., Mansky, 138 S. Ct. at 1885. ↑
-
. See 1 Rodney A. Smolla, Smolla & Nimmer on Freedom of Speech § 8:8.50, Westlaw (database updated April 2022) (“Content-based restrictions that define the contours of a limited public forum are not subject to strict scrutiny, but rather are judged under a standard of reasonableness.”); Christian Legal Soc’y, 561 U.S. at 679 (noting the requirement of viewpoint neutrality within a limited public forum). ↑
-
. Rosenberger, 515 U.S. at 829; see also Mansky, 138 S. Ct. at 1885 (noting the government’s authority to reserve a nonpublic forum “‘for its intended purposes, communicative or otherwise’”) (quoting Perry Educ. Ass’n v. Perry Loc. Educators’ Ass’n, 460 U.S. 37, 46 (1983)). ↑
-
. For scholarly analysis of public forum doctrine, see, for example, Lyrissa Lidsky, Public Forum 2.0, 91 B.U. L. Rev. 1975 (2011); Ronald J. Krotoszynski, Jr., Celebrating Selma: The Importance of Context in Public Forum Analysis, 104 Yale L.J. 1411 (1995). ↑
-
. See, e.g., Rosenberger, 515 U.S. at 829 (recognizing the requirement of viewpoint neutrality in a limited public forum); cf. Iancu v. Brunetti, 139 S. Ct. 2294, 2299 (2019) (deciding a case on grounds of viewpoint discrimination without resolving whether the program in question was best understood as a limited public forum). ↑
-
. See United States v. Stevens, 559 U.S. 460, 468 (2010); cf. Genevieve Lakier, The Invention of Low-Value Speech, 128 Harv. L. Rev. 2166, 2168 (2015) (“It is widely accepted today that the First Amendment guarantee of freedom of speech does not apply—or applies only weakly—to ‘low-value’ categories of speech such as obscenity and libel.”). ↑
-
. Cf. Bhagwat, supra note 26, at 1447 (noting the Supreme Court’s recognition during the 1970s and 1980s that “not all speech is [created] equal”). ↑
-
. See Stevens, 559 U.S. at 468. ↑
-
. See, e.g., Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942) (describing a comparison between the “benefit that may be derived” from certain types of speech and the countervailing “social interest in order and morality”). ↑
-
. Stevens, 559 U.S. at 472. ↑
-
. See, e.g., Restatement (Second) of Torts § 559 (Am. L. Inst. 1977) (“A communication is defamatory if it tends so to harm the reputation of another as to lower him in the estimation of the community or to deter third persons from associating or dealing with him.”). ↑
-
. See Miller v. California, 413 U.S. 15, 24–25 (1973) (defining obscenity); Brandenburg v. Ohio, 395 U.S. 444, 447 (1969) (defining incitement). ↑
-
. See Barr v. Am. Ass’n of Pol. Consultants, Inc., 140 S. Ct. 2335, 2361 (2020) (Breyer, J. concurring in the judgment with respect to severability and dissenting in part) (drawing on the categorical exceptions to First Amendment protection and concluding that “it makes little sense to treat every content-based distinction Congress has made as presumptively unconstitutional”). ↑
-
. Brandenburg, 395 U.S. at 448–49 (1969). ↑
-
. See N.Y. Times Co. v. Sullivan, 376 U.S. 254, 279–80 (1964) (requiring a showing of actual malice to establish liability in such circumstances); Stephen A. Siegel, The Origin of the Compelling State Interest Test and Strict Scrutiny, 48 Am. J. Legal Hist. 355, 359 (2006) (noting that the Sullivan standard “substantially protects freedom of speech” but does so “through a metric that is entirely different from what is meant by strict scrutiny”). ↑
-
. See Gertz v. Robert Welch, Inc., 418 U.S. 323, 347 (1974) (describing the constitutional requirements for defamation actions by private individuals). ↑
-
. See Miller, 413 U.S. at 24 (articulating inquiries that include whether speech appeals to the prurient interest and lacks serious value). ↑
-
. See, e.g., McCullen v. Coakley, 573 U.S. 464, 478 (2014) (describing the elements of strict scrutiny analysis). ↑
-
. See Snyder v. Phelps, 562 U.S. 443, 451 (2011) (“Whether the First Amendment prohibits holding [the speaker] liable for its speech in this case turns largely on whether that speech is of public or private concern, as determined by all the circumstances of the case.”). ↑
-
. Bd. of Trs. of the State Univ. of N.Y. v. Fox, 492 U.S. 469, 482 (1989) (emphasis removed). ↑
-
. See, e.g., Cent. Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557, 562–63 (1980) (discussing the diminished constitutional protection afforded to commercial speech); City of Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 422 (1993) (same). ↑
-
. See 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484, 522 (1996) (Thomas, J., concurring in part and concurring in the judgment) (questioning the basis for treating commercial speech as less valuable than other types of speech). ↑
-
. See, e.g., Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 762 (1976) (defining commercial speech as speech that “propos[es] a commercial transaction”). ↑
-
. Discovery Network, 507 U.S. at 426. ↑
-
. See, e.g., Sorrell, 564 U.S. at 572 (“[T]he State must show at least that the statute directly advances a substantial governmental interest and that the measure is drawn to achieve that interest.”). ↑
-
. Discovery Network, 507 U.S. at 426. ↑
-
. See 44 Liquormart, Inc. v. Rhode Island, 517 U.S. 484, 496 (plurality opinion) (discussing misleading advertisements). ↑
-
. See United States v. Alvarez, 567 U.S. 709, 723 (2012) (plurality opinion) (denying the government’s “authority to compile a list of subjects about which false statements are punishable”); id. at 733–34 (Breyer, J., concurring in the judgment) (noting the danger of affording the government broad discretion to restrict or punish false speech across a range of topics); id. at 751 (Alito, J., dissenting) (“[T]here are broad areas in which any attempt by the state to penalize purportedly false speech would present a grave and unacceptable danger of suppressing truthful speech.”). ↑
-
. See Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 772–73 (discussing the proposal of illegal transactions). ↑
-
. See Young v. Am. Mini Theatres, Inc., 427 U.S. 50 (1976). ↑
-
. See id. at 52. Adult theaters were also prohibited from locating near residential areas. See id. ↑
-
. Id. at 54–55. ↑
-
. See id. at 52. ↑
-
. See id. at 53. ↑
-
. Id. at 62–63. The Court treated the government’s distinctions between different types of speech as raising concerns under the Equal Protection Clause. See id. at 72–73 (plurality opinion). ↑
-
. Justice Powell would have upheld the zoning law as an “innovative land-use regulation, implicating First Amendment concerns only incidentally and to a limited extent.” Id. at 73 (Powell, J., concurring). ↑
-
. Id. at 65 (plurality opinion). ↑
-
. Id. ↑
-
. Id. at 66. ↑
-
. Id. at 70. ↑
-
. Id. at 71 n.34; see also Bhagwat, supra note 26, at 1442 (“[T]he [plurality] declined to subject the law to strict scrutiny because the law purportedly did not single out for suppression any specific ‘social, political, or philosophical message.’”). ↑
-
. City of Renton v. Playtime Theatres, Inc., 475 U.S. 41, 43 (1986). ↑
-
. See id. at 46–47 (“[R]egulations enacted for the purpose of restraining speech on the basis of its content presumptively violate the First Amendment.”). ↑
-
. See id. at 48. ↑
-
. See id. at 49. ↑
-
. Id. at 48 (quoting Police Dep’t v. Mosley, 408 U.S. 92, 95–96 (1972)). ↑
-
. 535 U.S. 425, 440 (2002) (plurality opinion). ↑
-
. Id. ↑
-
. Id. at 448 (Kennedy, J., concurring in the judgment) (calling Renton’s characterization of the relevant law as content neutral “something of a fiction”); see also id. (“[I]f the statute describes speech by content then it is content based.”); cf. Chemerinsky, supra note 51, at 61 (“[I]t is simply wrong to say that a facial, content-based distinction is otherwise because it is based on a permissible purpose.”). ↑
-
. Alameda Books, 535 U.S. at 445 (Kennedy, J., concurring in the judgment) (“A zoning measure can be consistent with the First Amendment if it is likely to cause a significant decrease in secondary effects and a trivial decrease in the quantity of speech”). ↑
-
. Id. at 446. ↑
-
. See Young v. Am. Mini Theatres, Inc., 427 U.S. 50, 70 (plurality opinion). ↑
-
. Id. ↑
-
. Id. at 85 (Stewart, J., dissenting). ↑
-
. See id. at 67 (plurality opinion) (noting that the rule “which is generally described as a prohibition of regulation based on the content of protected communication” is grounded in the principle that the government’s “regulation of communication may not be affected by sympathy or hostility for the point of view being expressed”). ↑
-
. See supra Part I. ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 167 (2015). ↑
-
. See id. (“Innocent motives do not eliminate the danger of censorship presented by a facially content-based statute.”). ↑
-
. See Bhagwat, supra note 26, at 1443 (noting the “obvious and direct tension” between the secondary effects doctrine and cases like Reed). ↑
-
. Reed, 576 U.S. at 183 (Kagan, J., concurring in the judgment). ↑
-
. Elena Kagan, Private Speech, Public Purpose: The Role of Governmental Motive in First Amendment Doctrine, 63 U. Chi. L. Rev. 413, 487–88 (1996). ↑
-
. See id. (“[T]he harm in secondary effects cases derives from a thing only contingently related to expression . . . .”). ↑
-
. Id. at 490; see also id. (describing a potential defense of the secondary effects doctrine as calibrated to the importance of governmental motive). ↑
-
. See, e.g., Leslie Gielow Jacobs, Making Sense of Secondary Effects Analysis After Reed v. Town of Gilbert, 57 Santa Clara L. Rev. 385, 388–90 (2017) (raising, but rejecting, the possibility that the Reed Court “implicitly repealed its secondary effects precedents”). ↑
-
. See Reed, 576 U.S. at 165–66. ↑
-
. City of Renton v. Playtime Theatres, Inc., 475 U.S. 41, 48 (1986) (emphasis in original) (quoting Va. State Bd. of Pharmacy v. Va. Citizens Consumer Council, Inc., 425 U.S. 748, 771 (1976)). ↑
-
. See infra Section IV.A. ↑
-
. See Playtime Theatres, 475 U.S. at 48–49 (describing the “fundamental principle” that “government may not grant the use of a forum to people whose views it finds acceptable, but deny use to those wishing to express less favored or more controversial views”) (quoting Police Dep’t v. Mosley, 408 U.S. 92, 95–96 (1972)). ↑
-
. See id. at 52. ↑
-
. For a related discussion of the interaction between the purposes, effects, and validity of speech restrictions, see Larry Alexander, Is There a Right of Freedom of Expression? 40–46 (2005). ↑
-
. 505 U.S. 377 (1992). ↑
-
. Id. at 381. ↑
-
. Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942). ↑
-
. See R.A.V., 505 U.S. at 382–83. ↑
-
. Id. at 380. ↑
-
. See id. at 388. ↑
-
. See id. at 390–91. ↑
-
. Id. at 390. ↑
-
. See id. at 391; Elena Kagan, The Changing Faces of First Amendment Neutrality: R.A.V. v. St. Paul, Rust v. Sullivan, and the Problem of Content-Based Underinclusion, 1992 Sup. Ct. Rev. 29, 35 (recognizing R.A.V.’s determination that the government “may not . . . bar only those fighting words addressing a particular subject or expressing a particular viewpoint”). ↑
-
. See R.A.V., 505 U.S. at 391 (“The First Amendment does not permit St. Paul to impose special prohibitions on those speakers who express views on disfavored subjects.”). ↑
-
. Id. at 387 (quoting Simon & Schuster, Inc. v. Members of the N.Y. State Crime Victims Bd., 502 U.S. 105, 116 (1991)). ↑
-
. Id. at 390. ↑
-
. Cf. Reed v. Town of Gilbert, 576 U.S. 155, 182 (2015) (Kagan, J., concurring in the judgment) (asking whether “official suppression of ideas is afoot” in determining the degree of scrutiny with which courts should review speech restrictions). ↑
-
. 505 U.S. at 390; see also id. (noting that content discrimination within categories of lesser-protected speech may be valid even without need “to identify any particular ‘neutral’ basis”). ↑
-
. See Bhagwat, supra note 26, at 1447 (observing that the treatment of certain types of speech as carrying reduced value “tended to alleviate the effects of the content-neutrality doctrine, restricting its application to situations where the regulated speech was highly valuable, and the government’s reasons for limiting the speech suspect”). ↑
-
. Cf. Paul B. Stephan III, The First Amendment and Content Discrimination, 68 Va. L. Rev. 203, 233 (1982) (remarking on “how many of the Court’s post-Mosley decisions could have reached the same result if they had rested exclusively on a viewpoint neutrality rule”). ↑
-
. W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943); cf. Erwin Chemerinsky, Content Neutrality as a Central Problem of Freedom of Speech: Problems in the Supreme Court’s Application, 74 S. Cal. L. Rev. 49, 59 (2000) (“Nothing is more inconsistent with freedom of speech than for the government to use its awesome power to advance some views and suppress or disfavor others.”); Jed Rubenfeld, The First Amendment’s Purpose, 53 Stan. L. Rev. 767, 818–19 (2001) (describing the role of the “anti-orthodoxy principle” in First Amendment jurisprudence). ↑
-
. Texas v. Johnson, 491 U.S. 397, 414 (1989). ↑
-
. Iancu v. Brunetti, 139 S. Ct. 2294, 2302 (2019) (Alito, J., concurring). ↑
-
. See, e.g., Police Dep’t v. Mosley, 408 U.S. 92, 95 (1972) (“[A]bove all else, the First Amendment means that government has no power to restrict expression because of its message, its ideas, its subject matter, or its content.”); Stephan, supra note 170, at 233 (describing viewpoint neutrality as “an essential concomitant of any rational system of freedom of expression”). ↑
-
. See infra Section IV.A. ↑
-
. Cf. Williams-Yulee v. Fla. Bar, 575 U.S. 433, 449 (2015) (using the “red flag” formulation to describe the impact of underinclusive speech restrictions). I elaborate on the analogy between underinclusivity and content discrimination in the Conclusion. ↑
-
. See Jud Campbell, The Emergence of Neutrality, 131 Yale L.J. 861, 928 (2022) (describing this development). ↑
-
. Street v. New York, 394 U.S. 576, 592 (1969). ↑
-
. Id.; see also id. at 593 (underscoring the freedom to express “opinions which are defiant or contemptuous”). ↑
-
. See Richard H. Fallon, Jr., Strict Judicial Scrutiny, 54 UCLA L. Rev. 1267, 1275 (2007) (“Before the 1960s, there was no strict scrutiny as we know it today.”); Siegel, supra note 99, at 357 (discussing the emergence of the compelling state interest requirement as part of strict scrutiny analysis). ↑
-
. See Fallon, supra note 180, at 1270 (describing strict scrutiny as intended to “protect ‘preferred’ or fundamental rights that were too important to be enforced only by a rational basis test, but that the Supreme Court could not reasonably define as wholly categorical or unyielding”); Siegel, supra note 99, at 358 (describing “[t]he compelling state interest test” as reflecting the “bifurcat[ion of] judicial review into heightened protection for favored rights and minimal protection for the rest”). ↑
-
. 576 U.S. 155 (2015). ↑
-
. Id. at 163. ↑
-
. 408 U.S. 92 (1972). ↑
-
. See id. at 92–93. ↑
-
. See id. at 94–95. Given this distinction, the Court analyzed the issue under the Fourteenth Amendment’s Equal Protection Clause, see id., though in subsequent years it would treat the First Amendment as governing the validity of such restrictions. Cf. Williams-Yulee v. Fla. Bar, 575 U.S. 433, 470 (2015) (Scalia, J., dissenting) (referring to the First Amendment as “a kind of Equal Protection Clause for ideas”). ↑
-
. Mosley, 408 U.S. at 95; see also Stephan, supra note 170, at 233 ( “[T]he ordinance [at issue in Mosley] apparently allowed working teachers to express their views about conditions of employment, but denied access near schools to interested parents, prospective teachers, and others generally interested in school policies.”). ↑
-
. Mosley, 408 U.S. at 96. ↑
-
. Young v. Am. Mini Theatres, Inc., 427 U.S. 50, 64 (1976) (plurality opinion). ↑
-
. See Farber, supra note 73, at 735 (addressing the role of viewpoint discrimination in Mosley). ↑
-
. Carey v. Brown, 447 U.S. 455 (1980). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 163 (2015). ↑
-
. See Carey, 447 U.S. at 457. ↑
-
. See id. ↑
-
. See id. at 469–71. ↑
-
. See id. at 460–61. ↑
-
. See Reed, 576 U.S. at 163. ↑
-
. 564 U.S. 552 (2011). ↑
-
. See id. at 562 (describing the district court’s finding that representatives selling branded drugs made use of the relevant information). ↑
-
. Id. at 578–79; see also id. at 578 (explaining that while “Vermont may be displeased that detailers who use prescriber-identifying information are effective in promoting brand-name drugs,” the State may not respond by “hamstring[ing] the opposition”). ↑
-
. See id. at 580 (faulting Vermont for “burden[ing] a form of protected expression that it found too persuasive”). ↑
-
. 491 U.S. 781, 791 (1989); see also id. (“A regulation that serves purposes unrelated to the content of expression is deemed neutral, even if it has an incidental effect on some speakers or messages but not others.”). ↑
-
. 512 U.S. 622, 641–43 (1994). ↑
-
. 447 U.S. 530 (1980). ↑
-
. See id. at 532–33. ↑
-
. Id. at 537. ↑
-
. Id. at 536 (quoting Niemotko v. Maryland, 340 U.S. 268, 282 (1951) (Frankfurter, J., concurring in the result)). ↑
-
. See infra Section IV.A. ↑
-
. Cf. City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464, 1474 (2022) (reasoning that the case law is inconsistent with the idea that “any examination of speech or expression inherently triggers heightened First Amendment concern”). ↑
-
. See, e.g., Kent Greenfield, Trademarks, Hate Speech, and Solving a Puzzle of Viewpoint Bias, 2019 Sup. Ct. Rev. 183, 183 (“In the hierarchy of constitutional offenses to free speech principles, content discrimination is near the very top.”); Geoffrey R. Stone, Content-Neutral Restrictions, 54 U. Chi. L. Rev. 46, 46 (1987) (“The content-based/content-neutral distinction plays a central role in contemporary first amendment jurisprudence.”). ↑
-
. Turner Broad. Sys., Inc. v. FCC, 512 U.S. 622, 658 (1994). ↑
-
. Stone, supra note 210, at 47 (“Content-based restrictions limit communication because of the message it conveys.”). ↑
-
. See, e.g., Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 829 (1995) (“When the government targets not subject matter, but particular views taken by speakers on a subject, the violation of the First Amendment is all the more blatant.”). ↑
-
. Matal v. Tam, 137 S. Ct. 1744, 1766 (2017) (Kennedy, J., concurring in part and concurring in the judgment); see also Iancu v. Brunetti, 139 S. Ct. 2294, 2313 (2019) (Sotomayor, J., concurring in part and dissenting in part) (“Viewpoint discrimination is ‘an egregious form of content discrimination’ in which ‘the government targets not subject matter, but particular views taken by speakers on a subject.’” (quoting Rosenberger, 515 U.S. at 829)). ↑
-
. See, e.g., Texas v. Johnson, 491 U.S. 397, 414 (1989) (“If there is a bedrock principle underlying the First Amendment, it is that the government may not prohibit the expression of an idea simply because society finds the idea itself offensive or disagreeable.”). ↑
-
. Robert C. Post, Between Governance and Management: The History and Theory of the Public Forum, 34 UCLA L. Rev. 1713, 1751 n.155 (1987). ↑
-
. Jacobellis v. Ohio, 378 U.S. 184, 197 (1964) (Stewart, J., concurring); see also Miller v. California, 413 U.S. 15, 20–23 (1973) (describing the Supreme Court’s attempts to define obscenity). ↑
-
. See, e.g., Pac. Gas & Elec. Co. v. Pub. Utils. Comm’n., 475 U.S. 1, 20 (1986) (“[T]he State cannot advance some points of view by burdening the expression of others.”); Tinker v. Des Moines Indep. Cmty. Sch. Dist., 393 U.S. 503, 509 (1969) (observing in the educational context that disciplinary action against students based on their speech must be driven by “something more than a mere desire to avoid the discomfort and unpleasantness that always accompany an unpopular viewpoint”). ↑
-
. W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943). ↑
-
. 515 U.S. 819 (1995). Rosenberger expanded upon the Court’s decision two years earlier in Lamb’s Chapel v. Center Moriches Union Free School District, 508 U.S. 384 (1993), which dealt with the denial of permission to display a film series that addressed a particular set of issues from a religious perspective. See id. at 394 (“The film series involved here no doubt dealt with a subject otherwise permissible . . . , and its exhibition was denied solely because the series dealt with the subject from a religious standpoint.”). ↑
-
. See Rosenberger, 515 U.S. at 829. ↑
-
. Id. at 822–23. ↑
-
. See Randy J. Kozel, Leverage, 62 B.C. L. Rev. 109, 121–23 (2021) (discussing whether, and to what extent, the government possesses authority to enact speech restrictions as conditions on public benefits even if it could not impose those restrictions directly); Mitchell N. Berman, Commercial Speech and the Unconstitutional Conditions Doctrine: A Second Look at “The Greater Includes the Lesser,” 55 Vand. L. Rev. 693 (2002) (examining the argument that the government’s power to withhold a benefit implies the power to impose conditions on recipients). ↑
-
. See Kozel, supra note 223, at 131–40 (discussing the government’s use of financial resources to influence behavior). ↑
-
. See supra Section II.A.; Rosenberger, 515 U.S. at 829–30. ↑
-
. See, e.g., Christian Legal Soc’y Chapter of the Univ. of Cal., Hastings Coll. of the L. v. Martinez, 561 U.S. 661, 679 (2010) (noting that restrictions on speech within a limited public forum must be “reasonable and viewpoint neutral”). ↑
-
. See, e.g., Minn. Voters All. v. Mansky, 138 S. Ct. 1876, 1892 (2018) (finding that the law at issue was not “capable of reasoned application”). ↑
-
. See, e.g., Martinez, 561 U.S. at 679 (recognizing the requirement of viewpoint neutrality). ↑
-
. Rosenberger, 515 U.S. at 830. ↑
-
. See id. at 829–30. ↑
-
. See id. at 830. ↑
-
. Id. at 827, 830–31. ↑
-
. See id. at 831. ↑
-
. Id. ↑
-
. See id. at 895 (Souter, J., dissenting) (describing viewpoint discrimination as defined by the government’s “taking sides in a public debate”); see also id. (“[I]f government assists those espousing one point of view, neutrality requires it to assist those espousing opposing points of view, as well.”). ↑
-
. See id. at 831 (majority opinion). ↑
-
. See, e.g., McCullen v. Coakley, 573 U.S. 464, 485 (2014) (concluding that restrictions involving speech outside of abortion clinics did not discriminate on the basis of viewpoint); Christian Legal Soc’y Chapter of the Univ. of Cal., Hastings Coll. of the L. v. Martinez, 561 U.S. 661, 694–97 (2010) (concluding that a policy governing student organizations at a public law school did not discriminate on the basis of viewpoint); Good News Club v. Milford Cent. Sch., 533 U.S. 98, 109 (2001) (concluding that a school discriminated on the basis of viewpoint by prohibiting a club from using the school’s facilities for meetings); id. at 112 n.4 (“Religion is the viewpoint from which ideas are conveyed.”); Legal Servs. Corp. v. Velasquez, 531 U.S. 533, 548–49 (2001) (suggesting that a program for providing funds to support certain legal services discriminated on the basis of viewpoint); Bd. of Regents of Univ. of Wisc. Sys. v. Southworth, 529 U.S. 217, 235 (2000) (reasoning that viewpoint neutrality is not preserved by a system that renders “[a]ccess to a public forum” dependent “upon majoritarian consent”); Nat’l Endowment for the Arts v. Finley, 524 U.S. 569, 582 (1998) (rejecting the argument that certain conditions on funding for the arts “inevitably will be utilized as a tool for invidious viewpoint discrimination”). ↑
-
. 137 S. Ct. 1744, 1751 (2017). ↑
-
. See id. ↑
-
. See id. (citing 15 U.S.C. § 1052(a)). ↑
-
. See id. at 1754. ↑
-
. See id. ↑
-
. See id. at 1751. ↑
-
. See Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 830 (1995). ↑
-
. See Tam, 137 S. Ct. at 1758 (noting, and rejecting, the argument that “the content of a registered mark is government speech”). ↑
-
. See, e.g., Rust v. Sullivan, 500 U.S. 173, 193 (1991) (reasoning that the government’s selective funding of certain speech or activities does not constitute viewpoint discrimination); Tam, 137 S. Ct. at 1757 (confirming that the “Free Speech Clause does not require government to maintain viewpoint neutrality when its officers and employees speak about” a government entity’s course of action). On the evolution of government speech doctrine in the Supreme Court’s case law, see, for example, Helen Norton & Danielle Keats Citron, Government Speech 2.0, 87 Denv. U. L. Rev. 899, 903–10 (2010). ↑
-
. See Walker v. Tex. Div., Sons of Confederate Veterans, Inc., 576 U.S. 200, 208–09 (2015) (applying government speech doctrine to a state’s specialty license plate program). ↑
-
. Tam, 137 S. Ct. at 1760. ↑
-
. Id. at 1763 (Alito, J., plurality opinion). ↑
-
. Id.; see also id. (stating that the Supreme Court’s cases define the concept of viewpoint discrimination “in a broad sense”). ↑
-
. Id. at 1766 (Kennedy, J., concurring in part and concurring in the judgment); see also id. (observing that the principle of viewpoint neutrality “protects more than the right to identify with a particular side”). ↑
-
. See id. at 1763 (Alito, J., plurality opinion) (describing viewpoint discrimination as a “broad” concept); id. at 1767 (Kennedy, J., concurring in part and concurring in the judgment) (noting the importance of protecting “ideas or perspectives . . . a particular audience might think offensive, at least at first hearing”). ↑
-
. See id. at 1754 (majority opinion). ↑
-
. Iancu v. Brunetti, 139 S. Ct. 2294, 2297 (2019). ↑
-
. Id. at 2298 (quoting 15 U.S.C. § 1052(a)). ↑
-
. See id. ↑
-
. See Bethel Sch. Dist. No. 403 v. Fraser, 478 U.S. 675 (1986) (sexually suggestive speech by a student at school); FCC v. Pacifica Found., 438 U.S. 726 (1978) (expletive-laden comedy routine broadcast over radio); Cohen v. California, 403 U.S. 15 (1971) (jacket displaying an expletive). ↑
-
. Brunetti, 139 S. Ct. at 2299. ↑
-
. Id. at 2300. ↑
-
. Id. ↑
-
. Id. at 2303 (Alito, J., concurring). ↑
-
. See id. at 2303 (Roberts, C.J., concurring in part and dissenting in part) (concluding that when trademarks, including vulgar trademarks, are denied registration, “[n]o speech is being restricted” and “no one is being punished”). ↑
-
. See id. at 2306–07 (Breyer, J., concurring in part and dissenting in part) (describing the government’s interest in refusing federal registration of vulgar trademarks). ↑
-
. See id. at 2317 (Sotomayor, J., concurring in part and dissenting in part) (describing a bar against registering “obscene, profane, or vulgar marks” as content based but viewpoint neutral). ↑
-
. See id. at 2300 (majority opinion). ↑
-
. See Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942). ↑
-
. See Snyder v. Phelps, 562 U.S. 443, 451 (2011) (suggesting that tort claims for intentional infliction of emotional distress based on matters of purely private concern are not barred by the First Amendment). ↑
-
. See Miller v. California, 413 U.S. 15, 24–25 (1973). ↑
-
. See United States v. Stevens, 559 U.S. 460, 472 (2010). ↑
-
. Cf. Sorrell v. IMS Health Inc., 564 U.S. 552, 578–79 (2011) (“The State may not burden the speech of others in order to tilt public debate in a preferred direction.”). ↑
-
. See, e.g., Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 831 (1995) (rejecting the argument that “no viewpoint discrimination occurs” if a speech restriction “discriminate[s] against an entire class of viewpoints”). ↑
-
. See Iancu v. Brunetti, 139 S. Ct. 2294, 2299–2300 (2019). ↑
-
. Police Dep’t v. Mosley, 408 U.S. 92, 93 (1972). ↑
-
. Carey v. Brown, 447 U.S. 455, 457 (1980). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 163 (2015). ↑
-
. See supra Section III.A. ↑
-
. Sorrell v. IMS Health Inc., 564 U.S. 552, 578–79 (2011). ↑
-
. See Brown v. Ent. Merchs. Ass’n, 564 U.S. 786, 799 (2011). ↑
-
. 403 U.S. 15, 16 (1971). ↑
-
. Id. ↑
-
. See id. at 26 (worrying that “governments might soon seize upon the censorship of particular words as a convenient guise for banning the expression of unpopular views”). In Iancu v. Brunetti, Justice Sotomayor suggested that the problem in Cohen was “viewpoint-neutral content discrimination.”139 S. Ct. 2294, 2314–15 (Sotomayor, J., concurring in part and dissenting in part). She reasoned that an expletive-laden statement of support for the draft would have been equally problematic. See id. at 2314. Yet the Cohen Court described itself as rejecting the argument that “the States, acting as guardians of public morality, may properly remove this offensive word from the public vocabulary.” Cohen, 403 U.S. at 22–23. The Court went on to reject “the facile assumption that one can forbid particular words without also running a substantial risk of suppressing ideas in the process,” worrying that the creation of such a power would allow governments to “seize upon the censorship of particular words as a convenient guise for banning the expression of unpopular views.” Id. at 26. These passages raise concerns about the marginalization of unwelcome views and the imposition of orthodoxy. ↑
-
. 319 U.S. 624, 642 (1943). ↑
-
. 491 U.S. 397, 414 (1989). ↑
-
. See, e.g., Ariz. Free Enter. Club’s Freedom Club PAC v. Bennett, 564 U.S. 721, 734 (2011) (observing that strict scrutiny applies to restrictions on campaign-related speech). ↑
-
. Cf. McConnell v. FEC, 540 U.S. 93, 249 (2003) (Scalia, J., concurring in part, concurring in the judgment in part, and dissenting in part) (explaining how ostensibly evenhanded restrictions on campaign speech can favor incumbents). ↑
-
. See, e.g., Citizens United v. FEC, 558 U.S. 310, 356 (2010) (noting the dangers of governmental attempts to “command where a person may get his or her information”). ↑
-
. See, e.g., Snyder v. Phelps, 562 U.S. 443, 451–52 (2011) (recognizing the primacy of speech on matters of public concern); id. at 452 (“[S]peech on public issues occupies the highest rung of the hierarchy of First Amendment values, and is entitled to special protection.”) (quoting Connick v. Myers, 461 U.S. 138, 145 (1983)). ↑
-
. Cf. Nat’l Endowment for the Arts v. Finley, 524 U.S. 569, 587 (1998) (raising, in the context of a free speech challenge to a program subsidizing the arts, the possibility of more searching judicial review in the face of “invidious viewpoint discrimination”). ↑
-
. See Farber, supra note 73, at 736 (observing that speech restrictions based on subject matter can “be entirely innocuous”); cf. David S. Han, Middle-Value Speech, 91 S. Cal. L. Rev. 65, 69 (2017) (arguing for rules that acknowledge and respond to the nature of speech that falls “between clearly high-value speech, like political speech and truthful news reporting, and clearly low-value speech, like fraudulent speech and incitement”). ↑
-
. This discussion relates to the treatment of speech outside the categorical exceptions to full First Amendment protection. For analysis of those exceptions, see supra Section II.B. ↑
-
. See Minn. Voters All. v. Mansky, 138 S. Ct. 1876, 1885 (2018) (“In a traditional public forum . . . restrictions based on content must satisfy strict scrutiny, and those based on viewpoint are prohibited.”); Marc Rohr, Some First Amendment Implications of the Trademark Registration Decisions, 24 Marq. Intell. Prop. L. Rev. 105, 121 (2020) (connecting viewpoint discrimination with “per se invalidity, with rare exception”). ↑
-
. See, e.g., Fla. Star v. B.J.F., 491 U.S. 524, 532 (1989) (declining “to hold broadly that truthful publication may never be punished consistent with the First Amendment” and remaining “mindful that the future may bring scenarios which prudence counsels our not resolving anticipatorily”). Even the Court’s canonical statement of constitutional skepticism of ordained orthodoxy in Barnette was followed by a recognition that absolutes can be problematic. See W. Va. State Bd. of Educ. v. Barnette, 319 U.S. 624, 642 (1943) (“If there are any circumstances which permit an exception, they do not now occur to us.”). ↑
-
. See McCullen v. Coakley, 573 U.S. 464, 478 (2014) (noting that if the petitioners were correct that the statute in question “favors one viewpoint about abortion over the other,” then the statute “must satisfy strict scrutiny”); Matal v. Tam, 137 S. Ct. 1744, 1765 (2017) (Kennedy, J., concurring in part and concurring in the judgment) (noting that “rigorous constitutional scrutiny” applies to viewpoint-discriminatory laws); cf. Greenfield, supra note 210, at 185 (suggesting that, as compared with content discrimination, “the added constitutional affront of viewpoint discrimination does not increase the level of judicial scrutiny in the run of cases”). As Justice Breyer has observed, the Supreme Court has unequivocally permitted viewpoint discrimination in at least one context: a school may take disciplinary action based on student speech “that can reasonably be regarded as encouraging illegal drug use.” Iancu v. Brunetti, 139 S. Ct. 2294, 2304 (2019) (Breyer, J., concurring in part and dissenting in part) (quoting Morse v. Frederick, 551 U.S. 393, 397 (2007)). ↑
-
. See supra Section IV.A; Joseph Blocher, Viewpoint Neutrality and Government Speech, 52 B.C. L. Rev. 695, 704 (2011) (citing Richard H. Fallon, Jr., Implementing the Constitution, 111 Harv. L. Rev. 54, 94 (1997), and recognizing the possibility that judges might look to content discrimination as a proxy for viewpoint discrimination); cf. William D. Araiza, Invasion of the Content-Neutrality Rule, 2019 BYU L. Rev. 875, 880 (noting the “basic instinct” that content-based restrictions on speech “provide a tool for censorship-minded government officials” and may even “suggest a government motive” to distort debate). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 176 (2015) (Breyer, J., concurring in the judgment) (“To use content discrimination to trigger strict scrutiny sometimes makes perfect sense.”); see also Chemerinsky, supra note 51, at 50 (“Content-based restrictions risk the government targeting particular messages and attempting to control thoughts on a topic by regulating speech.”). ↑
-
. Reed, 576 U.S. at 183 (Kagan, J., concurring in the judgment). ↑
-
. See Barr v. Am. Ass’n of Pol. Consultants, 140 S. Ct. 2335, 2360–61 (2020) (Breyer, J., concurring in the judgment with respect to severability and dissenting in part) (describing the conclusion that content discrimination triggers strict scrutiny as “a rule of thumb applicable only in certain circumstances”). ↑
-
. Stone, supra note 210, at 55. ↑
-
. See Reed, 576 U.S. at 183 (Kagan, J., concurring in the judgment) (recognizing that not every content-based restriction that triggers strict scrutiny will “skew the public’s debate of ideas”). ↑
-
. Cf. Farber, supra note 73, at 745 (characterizing it as “extremely inefficient” to lump in “innocuous regulations” with dangerous restrictions simply because both are content discriminatory). ↑
-
. Reed v. Town of Gilbert, 576 U.S. 155 (2015). ↑
-
. See id. at 159–61. ↑
-
. See id. at 179–80 (Kagan, J., concurring in the judgment). For similar sentiments from Justice Breyer, see Barr v. Am. Ass’n of Pol. Consultants, 140 S. Ct. 2335, 2359 (2020) (Breyer, J., concurring in the judgment with respect to severability and dissenting in part) (contending that courts must refrain from using the First Amendment to “threaten the workings of ordinary regulatory programs posing little threat to the free marketplace of ideas”). ↑
-
. See Reed, 576 U.S. at 174–75 (Alito, J., concurring). Justice Kagan responded by contending that Justice Alito’s position regarding such signs was incompatible with the majority’s opinion. See id. at 181 n.* (Kagan, J., concurring in the judgment). ↑
-
. Cf. Stone, supra note 210, at 55 (noting the difficulty of separating content-based restrictions on speech “into those that do and do not seriously distort public debate,” and urging courts to “test all content-based restrictions of high-value speech with the same strict standards of review”). ↑
-
. R.A.V. v. City of St. Paul, 505 U.S. 377, 390 (1992). ↑
-
. See United States v. O’Brien, 391 U.S. 367, 383 (1968) (“It is a familiar principle of constitutional law that this Court will not strike down an otherwise constitutional statute on the basis of an alleged illicit motive.”). ↑
-
. Cohen v. California, 403 U.S. 15, 23 (1971). ↑
-
. Texas v. Johnson, 491 U.S. 397, 400 n.1 (1989) (quoting Tex. Penal Code Ann. § 42.09 (1989) (quotation marks omitted)). ↑
-
. Cf. Alexander, supra note 154, at 46 (arguing that while “the permissibility of all rules turns on their presumed effects over time,” the consideration of “[p]urpose” remains relevant to “determining what the rules actually are”). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 184 (2015) (Kagan, J., concurring in the judgment). Of course, the application of intermediate scrutiny can sometimes be complicated. Consider Barr v. American Association of Political Consultants, 140 S. Ct. 2335 (2020), in which the Court examined a statute that prohibited robocalling but made an exception for calls seeking to collect debts owed or guaranteed by the federal government. A plurality determined that strict scrutiny was the appropriate test. See id. at 2347 (plurality opinion). Justice Sotomayor and Justice Breyer each wrote separately to explain their choice to apply intermediate scrutiny instead—a choice that led them to divergent conclusions regarding the lawfulness of the exception. See id. at 2356–57 (Sotomayor, J., concurring in the judgment); id. at 2362–63 (Breyer, J., concurring in the judgment with respect to severability and dissenting in part). ↑
-
. See Reed, 576 U.S. at 176 (Breyer, J., concurring in the judgment); id. at 182–83 (Kagan, J., concurring in the judgment). ↑
-
. See id. at 176 (Breyer, J., concurring in the judgment); id. at 182–83 (Kagan, J., concurring in the judgment). ↑
-
. Id. at 183 (Kagan, J., concurring in the judgment) (“We can administer our content-regulation doctrine with a dose of common sense, so as to leave standing laws that in no way implicate its intended function.”). ↑
-
. See City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464, 1468–69 (2022). ↑
-
. See Reed, 576 U.S. at 183 (Kagan, J., concurring in the judgment) (suggesting that when there is no realistic possibility that the government is favoring some ideas at the expense of others, “we may do well to relax our guard so that ‘entirely reasonable’ laws imperiled by strict scrutiny can survive”); id. at 176 (Breyer, J., concurring in the judgment) (“[C]ontent discrimination, while helping courts to identify unconstitutional suppression of expression, cannot and should not always trigger strict scrutiny.”). ↑
-
. See Kendrick, supra note 29, at 244 (describing the position that “regulations that turn on the communicative impact of speech are content based”). ↑
-
. R.A.V. v. City of St. Paul, 505 U.S. 377, 387 (1992) (quoting Simon & Schuster, Inc. v. Members of the N.Y. State Crime Victims Bd., 502 U.S. 105, 116 (1991)); see also Reed, 576 U.S. at 182 (Kagan, J., concurring in the judgment) (quoting and endorsing this language). ↑
-
. Reed, 576 U.S. at 182 (Kagan, J., concurring in the judgment). ↑
-
. See Frederick Schauer, The Boundaries of the First Amendment: A Preliminary Exploration of Constitutional Salience, 117 Harv. L. Rev. 1765, 1770–71 (2004) (noting that “no First Amendment-generated level of scrutiny is used to determine” the constitutionality of speech restrictions connected to the Securities Act of 1993, the Sherman Act, the law of conspiracy, or products liability). ↑
-
. See, e.g., Jane Bambauer, Is Data Speech?, 66 Stan. L. Rev. 57, 61 (2014) (“[A]ny time the state regulates information precisely because it informs people, the regulation rouses the First Amendment.”); Neil M. Richards, Reconciling Data Privacy and the First Amendment, 52 UCLA L. Rev. 1149, 1151 (2005) (“The First Amendment critics overstate the First Amendment issues at stake in the context of most database regulation proposals.”); Bhagwat, supra note 26, at 1445–46 (“[S]ome privacy laws, such as bank and healthcare secrecy laws, might survive strict scrutiny . . . but many surely will not.”). ↑
-
. For an argument defending the lawfulness of regulating speech-related invasions of sexual privacy, see generally Danielle Keats Citron, Sexual Privacy, 128 Yale L.J. 1870 (2019). ↑
-
. Reed, 576 U.S. at 177 (Breyer, J., concurring in the judgment). ↑
-
. See supra Part II; Stephan, supra note 170, at 236 (contending that the Supreme Court “never has” embraced and “never will” embrace a “requirement of absolute content neutrality”); cf. Bhagwat, supra note 26, at 1429–30 (discussing “an unstated discomfort” among judges “with the implications of the all-speech-is-equal premise”). ↑
-
. Sorrell v. IMS Health Inc., 564 U.S. 552, 578 (2011). ↑
-
. R.A.V. v. City of St. Paul, 505 U.S. 377, 390 (1992). ↑
-
. See, e.g., City of Ladue v. Gilleo, 512 U.S. 43, 48 (1994) (“It is common ground that governments may regulate the physical characteristics of signs—just as they can, within reasonable bounds and absent censorial purpose, regulate audible expression in its capacity as noise.”). A focus on viewpoint neutrality and reasonableness already characterizes the constitutional rules for limited public forums. See, e.g., Christian Legal Soc’y Chapter of the Univ. of Cal., Hastings Coll. of the L. v. Martinez, 561 U.S. 661, 679 (2010) (noting that within a limited public forum, “[a]ny access barrier must be reasonable and viewpoint neutral”); Rosenberger v. Rector & Visitors of Univ. of Va., 515 U.S. 819, 829–30 (1995) (recognizing the importance of the distinction between content discrimination and viewpoint discrimination in the context of a limited public forum). ↑
-
. See, e.g., City of Los Angeles v. Alameda Books, Inc., 535 U.S. 425, 433–34, 439–40 (2002) (plurality opinion) (describing the judicial evaluation of zoning regulations that target the secondary effects of speech). ↑
-
. See id. at 444–45 (Kennedy, J., concurring in the judgment) (criticizing as “imprecise” the claim that regulations focused on the secondary effects of speech are necessarily content neutral). ↑
-
. See id. ↑
-
. See id. at 440 (plurality opinion); see also id. (“There is less reason to be concerned that municipalities will use these ordinances to discriminate against unpopular speech.”). ↑
-
. See Reed v. Town of Gilbert, 576 U.S. 155, 165 (2015). ↑
-
. Discretion is not boundless. See City of Renton v. Playtime Theatres, Inc., 475 U.S. 41, 50 (1986) (asking whether the ordinance in question was “designed to serve a substantial governmental interest and allow[ed] for reasonable alternative avenues of communication”). ↑
-
. 137 S. Ct. 1744 (2017). ↑
-
. See Iancu v. Brunetti, 139 S. Ct. 2294, 2298 (2019) (observing that in Tam, the Justices “could not agree on the overall framework for deciding the case”). ↑
-
. Id. at 2299. ↑
-
. See supra Section III.A. ↑
-
. 500 U.S. 173, 198–200 (1991). ↑
-
. See id. ↑
-
. Id. ↑
-
. See id. at 194 (“When Congress established a National Endowment for Democracy . . . it was not constitutionally required to fund a program to encourage . . . communism and fascism.”). ↑
-
. See Matal v. Tam, 137 S. Ct. 1744, 1757 (2017) (explaining that the requirement of viewpoint neutrality does not apply to the government’s own speech); Walker v. Tex. Div., Sons of Confederate Veterans, Inc., 576 U.S. 200, 215–17 (2015) (distinguishing the government’s own speech from its creation of forums for others’ speech). ↑
-
. See, e.g., N.Y. Times Co. v. Sullivan, 376 U.S. 254, 270 (1964) (recognizing “a profound national commitment to the principle that debate on public issues should be uninhibited, robust, and wide-open”). ↑
-
. See Carey v. Brown, 447 U.S. 455, 457 (1980) (examining a law that barred residential picketing but made an exception for speech related to labor disputes). ↑
-
. See Barr v. Am. Ass’n of Pol. Consultants, Inc., 140 S. Ct. 2335, 2343 (2020) (plurality opinion) (examining a law that gave favorable treatment to robocalls “made to collect debts owed to or guaranteed by the Federal Government”). ↑
-
. Cf. City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464, 1468–69 (2022) (involving a regulation that distinguished between off-premises and on-premises signs). ↑
-
. See Williams-Yulee v. Fla. Bar, 575 U.S. 433, 448 (2015) (“It is always somewhat counterintuitive to argue that a law violates the First Amendment by abridging too little speech.”). ↑
-
. See, e.g., United States v. Alvarez, 567 U.S. 709, 723 (2012) (plurality opinion) (“Our constitutional tradition stands against the idea that we need Oceania’s Ministry of Truth.”); Citizens United v. FEC, 558 U.S. 310, 340 (2010) (describing the First Amendment as “[p]remised on mistrust of governmental power”). ↑
-
. See, e.g., Ariz. Free Enter. Club’s Freedom Club PAC v. Bennett, 564 U.S. 721, 757 (2011) (Kagan, J., dissenting) (reasoning that the First Amendment does not prohibit reasonable measures that promote “a healthy, vibrant political system”). ↑
-
. Cf. Reed v. Town of Gilbert, 576 U.S. 155, 183 (2015) (Kagan, J., concurring in the judgment) (contending that when there is no reason to suspect governmental favoritism among ideas, “we may do well to relax our guard so that ‘entirely reasonable’ laws imperiled by strict scrutiny can survive”). ↑
-
. Williams-Yulee, 575 U.S. at 452. ↑
-
. See Reed, 576 U.S. at 165; City of Austin v. Reagan Nat’l Advert. of Austin, LLC, 142 S. Ct. 1464, 1471 (2022). One may question the conceptual compatibility of Reed with City of Austin, as noted in Part I. Still, the latter did not explicitly reject the former. ↑
-
. E.g., Kisor v. Wilkie, 139 S. Ct. 2400, 2422 (2019) (citation omitted). ↑
-
. For recent treatments of precedent and stare decisis, see, for example, Amy Coney Barrett, Originalism and Stare Decisis, 92 Notre Dame L. Rev. 1921 (2017); Richard M. Re, Narrowing Precedent in the Supreme Court, 114 Colum. L. Rev. 1861 (2014); Randy J. Kozel, Settled Versus Right: A Theory of Precedent (2017). ↑
-
. See supra Section II.B. ↑
-
. See United States v. Stevens, 559 U.S. 460, 466–67 (2010). ↑
-
. See id. at 469–70. ↑
-
. See id. at 470. ↑
-
. Id. ↑
-
. See id. at 468–69. ↑
-
. See id. at 471. ↑
-
. Id. at 472. ↑
-
. See, e.g., Brown v. Ent. Merchs. Ass’n, 564 U.S. 786, 802 (2011) (criticizing a statutory scheme that regulated violent video games but not violent content distributed by “booksellers, cartoonists, and movie producers”). ↑
-
. See Carey v. Brown, 447 U.S. 455, 465 (1980). ↑
-
. See, e.g., id.; Brown, 564 U.S. at 802; Fla. Star v. B.J.F., 491 U.S. 524, 540 (1989). There are some exceptions, such as the doctrine of secondary effects, pursuant to which the Supreme Court has described underinclusivity as less of a problem. See City of Renton v. Playtime Theatres, Inc., 475 U.S. 41, 52–53 (1986) (rejecting an underinclusivity challenge to an ordinance aimed at the secondary effects of certain speech-related businesses). ↑
-
. 575 U.S. 433 (2015). ↑
-
. See id. at 437–38. ↑
-
. See id. at 439. ↑
-
. See id. at 440. ↑
-
. See id. at 448. ↑
-
. Id. at 449. ↑
-
. See id. at 448–49. ↑
-
. See id. at 448; Holder v. Humanitarian L. Project, 561 U.S. 1, 31–32 (2010) (accepting the government’s decision to restrict speech deemed most detrimental without prohibiting related types of speech). ↑
-
. See Williams-Yulee, 575 U.S. at 448–49. ↑
-
. See id. at 449. ↑
-
. Cf. Barr v. Am. Ass’n of Pol. Consultants, 140 S. Ct. 2335, 2361 (2020) (Breyer, J., concurring in the judgment with respect to severability and dissenting in part) (contending that not all content-discriminatory laws warrant strict scrutiny). ↑