SUPREME COURT OF THE UNITED STATES

_________________

No. 20–1029

_________________

City of Austin, Texas, PETITIONER v. Reagan National Advertising of Austin, LLC, et al.

on writ of certiorari to the united states court of appeals for the fifth circuit

[April 21, 2022]

Justice Thomas, with whom Justice Gorsuch and Justice Barrett join, dissenting.

In Reed v. Town of Gilbert, 576 U.S. 155 (2015), we held that a speech regulation is content based—and thus presumptively invalid—if it “draws distinctions based on the message a speaker conveys.” Id., at 163. Here, the city of Austin imposes special restrictions on “off-premise[s] sign[s],” defined as signs that “advertis[e] a business, person, activity, goods, products, or services not located on the site where the sign is installed, or that direc[t] persons to any location not on that site.” Austin, Tex., City Code §25–10–3(11) (2016). Under Reed, Austin’s off-premises restriction is content based. It discriminates against certain signs based on the message they convey—e.g., whether they promote an on- or off-site event, activity, or service.

The Court nevertheless holds that the off-premises restriction is content neutral because it proscribes a sufficiently broad category of communicative content and, therefore, does not target a specific “topic or subject matter.” Ante, at 8. This misinterprets Reed’s clear rule for content-based restrictions and replaces it with an incoherent and malleable standard. In so doing, the majority’s reasoning is reminiscent of this Court’s erroneous decision in Hill v. Colorado, 530 U.S. 703 (2000), which upheld a blatantly content-based prohibition on “counseling” near abortion clinics on the ground that it discriminated against “an extremely broad category of communications.” Id., at 723. Because I would adhere to Reed rather than echo Hill’s long-discredited approach, I respectfully dissent.

I

A

The First Amendment, applicable to the States through the Fourteenth, prohibits laws “abridging the freedom of speech.” U. S. Const., Amdt. 1; see also Stromberg v. California, 283 U.S. 359, 368 (1931). “When enforcing this prohibition, our precedents distinguish between content-based and content-neutral regulations.” National Institute of Family and Life Advocates v. Becerra, 585 U. S. ___, ___ (2018) (slip op., at 6). A content-based law is “presumptively invalid,” United States v. Playboy Entertainment Group, Inc., 529 U.S. 803, 817 (2000) (internal quotation marks omitted), and may generally be upheld only if the government proves that the regulation is narrowly tailored to serve compelling state interests, R. A. V. v. St. Paul, 505 U.S. 377, 395 (1992).[1]

In Reed v. Town of Gilbert, we held that courts should identify content-based restrictions by applying a “commonsense” test: A speech regulation is content based if it

“target[s] speech based on its communicative content.” 576 U. S., at 163. Put another way, a law is content based “ ‘on its face’ [if it] draws distinctions based on the message a speaker conveys.” Ibid. While we noted that “[s]ome facial distinctions based on a message are obvious,” we emphasized that others could be “more subtle, defining regulated speech by its function or purpose.” Ibid. In all events, whether a law is characterized as targeting a “topic,” “idea,” “subject matter,” or “communicative content,” the law is content based if it draws distinctions based in any way “on the message a speaker conveys.” Id., at 163–164.[2]

Applying this standard, we held that the town of Gilbert’s sign code was “a paradigmatic example of content-based discrimination” because it classified “various categories of signs based on the type of information they convey[ed], [and] then subject[ed] each category to different restrictions.” Id., at 169, 159. For instance, Gilbert defined “ ‘Temporary Directional Signs’ ” as any sign that “convey[ed] the message of directing the public to [a] ‘qualifying event,’ ” and permitted their display for no more than 12 hours before and 1 hour after the event occurred. Id., at 164, 161. Meanwhile, “ ‘Ideological Sign[s],’ ” defined as any sign (not covered by another category) that “ ‘communicat[ed] a message or ideas for noncommercial purposes,’ ” were subject to no temporal limitations. Id., at 159–160. In short, the restrictions on any given sign depended “on the communicative content of the sign.” Id., at 164. Gilbert’s sign code was thus facially content based and presumptively unlawful. See id., at 159.

In contrast to Reed’s “commonsense” test, Gilbert urged us to define “content based” as a “term of art that ‘should be applied flexibly’ with the goal of protecting ‘viewpoints and ideas from government censorship or favoritism.’ ” Id., at 168. Such a functionalist test, Gilbert argued, could ferret out illicit government motives while obviating the need to subject reasonable laws to strict scrutiny. See ibid. We rejected Gilbert’s attempt to cast the phrase “content based” as a “term of art” because “[i]nnocent motives do not eliminate the danger of censorship presented by a facially content-based statute.” Id., at 167. We noted that “one could easily imagine a Sign Code compliance manager who disliked [a] Church’s substantive teachings deploying the Sign Code to make it more difficult for the Church to inform the public of the location of its services.” Id., at 167–168. Thus, we concluded that “a clear and firm rule governing content neutrality is an essential means of protecting the freedom of speech, even if laws that might seem entirely reasonable will sometimes be struck down because of their content-based nature.” Id., at 171 (internal quotation marks omitted).

We also rejected the Ninth Circuit’s reasoning that Gilbert’s sign restrictions were content neutral because they depended on “the content-neutral elements of . . . whether and when an event is occurring.” Id., at 169 (internal quotation marks omitted). That is, whether a temporary directional sign was permissible depended, in part, on its temporal proximity to a “ ‘qualifying event.’ ” Id., at 164. This partial dependence on content-neutral elements was immaterial, we explained, because the restrictions also depended on the signs’ communicative content. Gilbert officials still had to examine a sign’s message to determine what type of sign it was, and this “obvious content-based inquiry d[id] not evade strict scrutiny simply because an event [was] involved.” Id., at 170.

B

Under Reed’s approach for identifying content-based regulations, Austin’s off-premises sign restriction is content based. As relevant to this suit, Austin’s sign code imposes stringent restrictions on a category of “off-premise[s] sign[s].” §25–10–3(11). The code defines “off-premise[s] sign[s]” as those “advertising a business, person, activity, goods, products, or services not located on the site where the sign is installed,” or as signs “direct[ing] persons to any location not on that site.” Ibid. This broad definition sweeps in a wide swath of signs, from 14- by 48-foot billboards to 24- by 18-inch yard signs. The sign code prohibits new off-premises signs and makes it difficult (or impossible) to change existing off-premises signs, including by digitizing them. See ante, at 3.

Like the town of Gilbert in Reed, Austin has identified a “categor[y] of signs based on the type of information they convey, [and] then subject[ed that] category to different restrictions.” 576 U. S., at 159. A sign that conveys a message about off-premises activities is restricted, while one that conveys a message about on-premises activities is not. See id., at 171 (regulating signs based on “a particular message” about “the time and location of a specific event” is content based). And, per Reed, it does not matter that Austin’s code “defin[es] regulated speech by its function or purpose”—i.e., advertising or directing passersby elsewhere. Id., at 163. Again, all that matters is that the regulation “draws distinctions based on” a sign’s “communicative content,” which the off-premises restriction plainly does. Ibid.

This conclusion is not undermined because the off- premises sign restriction depends in part on a content- neutral element: the location of the sign. Much like in Reed, that an Austin official applying the sign code must know where the sign is does not negate the fact that he also must know what the sign says. Take, for instance, a sign outside a Catholic bookstore. If the sign says, “Visit the Holy Land,” it is likely an off-premises sign because it conveys a message directing people elsewhere (unless the name of the bookstore is “Holy Land Books”). But if the sign instead says, “Buy More Books,” it is likely a permissible on- premises sign (unless the sign also contains the address of another bookstore across town). Finally, suppose the sign says, “Go to Confession.” After examining the sign’s message, an official would need to inquire whether a priest ever hears confessions at that location. If one does, the sign could convey a permissible “on-premises” message. If not, the sign conveys an impermissible off-premises message. Because enforcing the sign code in any of these instances “requires [Austin] officials to determine whether a sign” conveys a particular message, the sign code is content based under Reed. Id., at 170.

In sum, the off-premises rule is content based and thus invalid unless Austin can satisfy strict scrutiny. See Playboy Entertainment Group, 529 U. S., at 813. Because Austin has offered nothing to make that showing, the Court of Appeals did not err in holding that the off-premises rule violates the First Amendment.

II

To reach the opposite result, the majority implicitly rewrites Reed’s bright-line rule for content-based restrictions. In the majority’s view, the off-premises restriction is not content based because it does not target a specific “topic or subject matter.” Ante, at 8. The upshot of the majority’s reasoning appears to be that a regulation based on a sufficiently general or broad category of communicative content is not actually content based.

Such a rule not only conflicts with Reed and many pre-Reed precedents but is also incoherent and unworkable. Tellingly, the only decision that even remotely supports the majority’s rule is one it does not cite: Hill v. Colorado. There, the Court held that an undeniably content-based law was nonetheless content neutral because it discriminated against “an extremely broad category of communications,” supposedly without regard to “subject matter.” 530 U. S., at 723. The majority’s decision today is erroneous for the same reasons that Hill is an aberration in our case law.

A

The majority concedes that “[t]he message on the sign matters” when applying Austin’s sign code. Ante, at 8. That concession should end the inquiry under Reed. But the majority nonetheless finds the sign code to be content neutral by recasting facially content-based restrictions as only those that target sufficiently specific categories of communicative content and not as those that depend on communicative content simpliciter.

For example, while Reed defined content-based restrictions as those that “dra[w] distinctions based on the message a speaker conveys,” 576 U. S., at 163 (emphasis added), the majority decides that Austin’s sign code is not content based because it draws no distinctions based on “[a] sign’s substantive message,” ante, at 8 (emphasis added). Elsewhere, the majority speaks not of “substantive message[s]” but of “topic[s] or subject matter[s],” which the majority thinks are sufficiently specific categories of communicative content. Ibid. As a result, the majority contends that a law targeting directional messages concerning “events generally, regardless of topic,” would not be content based, but one targeting “directional messages concerning specific events” (e.g., “religious” or “political” events) would be. Ante, at 12, n. 6, 8 (emphasis added).[3] Regardless of the label, the majority today excises, without a word of explanation, a subset of supposedly non-substantive or unspecific messages from the First Amendment’s protection against content-based restrictions.

This understanding of content-based restrictions contravenes Reed, which held that a law is content based if it “target[s] speech based on its communicative content”—not “specific” or “substantive” categories of communicative content. 576 U. S., at 163; see also, e.g., Norton v. Springfield, 806 F.3d 411, 412 (CA7 2015) (“Reed effectively abolishes any distinction between content regulation and subject-matter regulation. Any law distinguishing one kind of speech from another by reference to its meaning now requires a compelling justification”). Only by jettisoning Reed’s “commonsense” definition of what it means to be content based can the majority assert that the off-premises rule is strictly “location-based” and “agnostic as to content,” ante, at 6, even though the law undeniably depends on both location and communicative content, supra, at 5–6.

Moreover, the majority’s suggestion that laws targeting broad categories of communicative content are not content based is hard to square with the sign categories that Reed invalidated. For instance, we found Gilbert’s expansive definition of “Ideological Sign[s]” to be content based even though it broadly covered any “sign communicating a message or ideas for noncommercial purposes” that did not already fall into one of the other categories. 576 U. S., at 159 (internal quotation marks omitted). Nor did we suggest that the outcome in Reed would have been different if the sign categories were defined even more generally.

The majority answers that it is not “fashion[ing] a novel ‘specificity test,’ ” but instead “simply” “quoting the standard repeatedly enunciated in Reed.” Ante, at 13. The majority finds this alleged specificity test in a paragraph near the end of Reed, where we noted that a law “targeted at specific subject matter is content based even if it does not discriminate among viewpoints within that subject matter,” and then affirmed that Gilbert’s sign code “single[d] out specific subject matter for differential treatment.” 576 U. S., at 169.

These statements never purported to endorse a specificity test of the sort now suggested by the majority. Read in context, Reed’s two references to “specific subject matter” naturally address laws that target a “subject matter,” however broadly defined, as opposed to some other subject matter; they did not refer only to laws targeting some sufficiently “specific” category of “subject matter.” Moreover, the concept of “specificity” or “generality” appears nowhere in the part of Reed that set forth its “commonsense” test for content neutrality. See id., at 163–164. If Reed’s content-neutrality test turned on specificity, we would have said so explicitly when stating the test. Finally, even crediting the majority’s strained reading of Reed’s passing references to “specific subject matter,” the paragraph where they appear made clear that it was describing only “a paradigmatic example of content-based discrimination.” Id., at 169 (emphasis added). That part of Reed never professed to announce a comprehensive rule with respect to all laws targeting speech based on its communicative content.

Our pre-Reed precedents likewise foreclose a construction of “content based” that applies only to some content. We have held many capacious speech regulations to be content based, including restrictions on “ ‘advice or assistance derived from scientific, technical or other specialized knowledge,’ ” Holder v. Humanitarian Law Project, 561 U.S. 1, 12–13 (2010); “ ‘advertising, promotion, or any activity . . . used to influence sales or the market share of a prescription drug,’ ” Sorrell v. IMS Health Inc., 564 U.S. 552, 559 (2011); “editorializing,” FCC v. League of Women Voters of Cal., 468 U.S. 364, 382–383, and n. 14 (1984); “ ‘[publication] for philatelic, numismatic, educational, historical, or newsworthy purposes,’ ” Regan v. Time, Inc., 468 U.S. 641, 644 (1984); and “anonymous speech,” McIntyre v. Ohio Elections Comm’n, 514 U.S. 334, 348, 357 (1995). These speech categories are no more “specific” or “substantive” than messages regarding off-premises activities. And some of these examples, like “editorializing” or publishing “newsworthy” information, are clearly less so. What unites these speech restrictions is that their application turns “on the nature of the message being conveyed,” Carey v. Brown, 447 U.S. 455, 461 (1980), not whether they regulate specific or general categories of speech, or whether they address substantive or non-substantive categories of speech.

We have defined content-based restrictions to include all content-based distinctions because any other rule would be incoherent. After all, off-premises advertising could be considered a “subject” or a “topic” as those words are ordinarily used. See L. D. Management Co. v. Gray, 988 F.3d 836, 839 (CA6 2021) (off-premises billboard restriction “turns on the ‘topic discussed’ ” (emphasis added)). And, in any event, there is no principled way to decide whether a category of communicative content is “substantive” or “specific” enough for the majority to deem it a “topic” or “subject” worthy of heightened protection. Although off-premises advertising is a more general category of speech than some (e.g., off-premises advertising of religious events), it is a more specific category than others (e.g., advertising generally). The majority offers only its own ipse dixit to explain why off-premises advertising is insufficiently specific to qualify as content based under Reed. Worse still, the majority does not explain how courts should draw the line between a sufficiently substantive or specific content-based classification and one that is insufficiently substantive or specific.

On this point, Austin suggests there is no need to worry because our cases provide “guideposts” from which one can divine what “level of generality” renders a speech regulation content based. Tr. of Oral Arg. 18, 24. To be sure, that is the sort of inquiry the majority’s opaque test invites. But Reed directed us elsewhere—to the text of the law in question and whether that law “ ‘on its face’ draws distinctions based on the message a speaker conveys.” 576 U. S., at 163. The majority’s holding that some rules based on content are not, as it turns out, content based nullifies Reed’s clear test.

B

The majority offers several reasons why its approach is consistent with Reed and other cases. None of these arguments is persuasive. Instead, they only serve to underscore the Court’s ill-advised departure from our doctrine.

1

The majority first suggests that deeming Austin’s sign code content based would require us to adopt an “extreme” reinterpretation of Reed. Ante, at 6. Specifically, the majority faults the Court of Appeals for concluding that Austin’s regulation was content based because, to enforce the off-premises rule, “ ‘[a] reader must ask: who is the speaker and what is the speaker saying’ ”? Ibid. (quoting 972 F.3d 696, 706 (CA5 2020)). In the majority’s view, Reed cannot stand for such a simplistic read-the-sign test.

The majority’s skepticism is misplaced. We have often acknowledged that the need to examine the content of a message is a strong indicator that a speech regulation is content based. One year before Reed, for example, we stated that an abortion clinic buffer-zone law “would be content based if it required enforcement authorities to examine the content of the message that is conveyed to determine whether a violation has occurred.” McCullen v. Coakley, 573 U.S. 464, 479 (2014) (internal quotation marks omitted). That statement was not an outlier. See, e.g., Arkansas Writers’ Project, Inc. v. Ragland, 481 U.S. 221, 230 (1987) (tax exemption for periodicals “uniformly devoted to religion or sports” was content based because it required state officials to “examine the content of the message” (internal quotation marks omitted)); Forsyth County v. Nationalist Movement, 505 U.S. 123, 134 (1992) (regulation requiring parade organizers to pay a fee depending on the security costs anticipated for the event was content based because “[i]n order to assess accurately the cost of security for parade participants, the administrator must necessarily examine the content of the message that is conveyed” (internal quotation marks omitted)); League of Women Voters, 468 U. S., at 366, 383 (law forbidding public broadcasting stations from “engag[ing] in editorializing” was content based because it required “enforcement authorities [to] necessarily examine the content of the message that is conveyed” (internal quotation marks omitted)).

Ultimately, the majority’s objection to the Court of Appeals’ reliance on a read-the-sign test is a red herring; its real objection is to Reed’s rule that any law that draws distinctions based on communicative content is content based.

2

The majority next argues that Austin’s sign code is content neutral under our precedents. See ante, at 8–10. But none of the cases the majority cites supports its crabbed view of what constitutes a content-based restriction.

First, in Heffron v. International Soc. for Krishna Consciousness, Inc., 452 U.S. 640 (1981), the Court upheld, as content neutral, an ordinance providing that the “[s]ale or distribution of any merchandise, including printed or written material,” could occur only from certain booths at the fairgrounds. Id., at 643 (internal quotation marks omitted). Such a statute is facially content neutral under Reed because it does not “ ‘on its face’ dra[w] distinctions based on the message a speaker conveys” when selling or distributing merchandise subject to the ordinance. 576 U. S., at 163. True, the Court construed the ordinance also to limit “fund solicitation operations,” 452 U. S., at 644, but that was not, as the majority claims, a prohibition on “asking for donations,” ante, at 9. Rather, anyone was free to “as[k] for donations” wherever he liked, because the ordinance did “not prevent respondents from wandering throughout the fairgrounds and directing interested donors or purchasers to their booth.” 452 U. S., at 664, n. 2 (Blackmun, J., concurring in part and dissenting in part). Then, once “at the booth,” the donor could “make a contribution.” Ibid.

Second, in Cantwell v. Connecticut, 310 U.S. 296 (1940), the Court invalidated a licensing system for religious and charitable solicitation while acknowledging in dicta that a State could regulate the time, place, and manner of solicitation. Id., at 304, 307. But here, we are not faced with a true time, place, or manner restriction, as even the majority concedes. See ante, at 8.[4] And, in any event, Cantwell did not suggest that a content-based restriction could be sustained as a time, place, or manner restriction; its analysis focused predominantly on the plaintiff ’s free exercise claim; and the case predated our modern content-neutrality doctrine by nearly three decades. Thus, nothing in Heffron or Cantwell supports the majority’s narrow approach to identifying content-based restrictions.

Finally, the majority argues that we have “previously understood distinctions between on-premises and off-premises signs . . . to be content neutral.” Ante, at 9–10. To be sure, in both Suffolk Outdoor Adv. Co. v. Hulse, 439 U.S. 808 (1978), and Metromedia, Inc. v. San Diego, 453 U.S. 490, 503–512 (1981) (plurality opinion), this Court suggested that some restrictions on off-premises advertising were constitutional. And later, in Members of City Council of Los Angeles v. Taxpayers for Vincent, 466 U.S. 789 (1984), the Court described Metromedia as upholding “a content- neutral prohibition against the use of billboards.” 466 U. S., at 807 (emphasis added). But the statement in Vincent was dictum, and, as the majority concedes, both our summary decision in Suffolk and the plurality opinion in Metromedia sanctioned off-premises restrictions only insofar as they applied to commercial speech. Ante, at 10. That is, the “Court did not need to decide”—and did not decide—“whether the off-premises prohibition was content based” because restrictions on commercial speech are “subject to intermediate scrutiny in any event.” Ibid.

3

The majority also claims that finding Austin’s sign code to be content based “would render the majority opinion in Reed irreconcilable with” Justice Alito’s Reed concurrence. Ante, at 12, n. 6. In particular, Justice Alito identified nine different types of sign regulations that he believed “would not be content based,” including “[r]ules distinguishing between on-premises and off-premises signs” and “[r]ules imposing time restrictions on signs advertising a one-time event.” 576 U. S., at 174–175. The majority evidently believes that these two types of sign regulations necessarily turn on a sign’s communicative content, like the off-premises sign restriction at issue here.

That reading of the Reed concurrence makes little sense. First, there is no reason to interpret the concurrence as referring to off-premises or one-time-event rules that turn on a sign’s communicative content. Doing so would make those two rules categorically different from the other seven, none of which would ever turn on message content. See, e.g., id., at 174 (“Rules distinguishing between lighted and unlighted signs”). And although off-premises and one-time-event rules could be drafted in terms of a sign’s communicative content, as is true here, they need not be. “There might be many formulations of an on/off-premises distinction that are content-neutral.” Thomas v. Bright, 937 F.3d 721, 733 (CA6 2019); see also ante, at 2, n. 1 (Alito, J., concurring in judgment in part and dissenting in part) (explaining that “[i]n ordinary usage” an “off- premises” sign is one that is not “attached to or located in close proximity to a building”). For instance, a city could define “ ‘an o[n]-premise[s] sign as any sign within 500 feet of a building,’ ” 937 F. 3d, at 732, or a sign that is installed by “ ‘a business . . . licensed to occupy . . . the premises where the sign is located,’ ” Brief for Summus Outdoor as Amicus Curiae 10. As for regulations of one-time-event signs, Austin itself amended its sign code, at the behest of its lawyers, specifically to make its ordinance content neutral. See Austin, Tex., City Code §25–10–102(D) (2021); App. 152. Thus, interpreting Justice Alito’s concurrence as referring to rules that turn on communicative content, as opposed to rules that are content neutral, is unwarranted.

Second, it would be strange to interpret the concurrence as proclaiming that all off-premises sign restrictions are content neutral considering the longstanding dispute over that question. In fact, 20 years before Reed, then-Judge Alito opined that there was “no easy answer to [the] question” whether “exceptions for ‘for sale’ signs and signs relating to on-site activities” would render a sign code content based. Rappa v. New Castle County, 18 F.3d 1043, 1080 (CA3 1994) (concurring opinion); see also, e.g., Ackerly Communications of Mass., Inc. v. Cambridge, 88 F.3d 33, 36, n. 7 (CA1 1996) (“In ‘commonsense’ terms, the distinction surely is content-based because determining whether a sign must stay up or must come down requires consideration of the message it carries”); Norton Outdoor Adv., Inc. v. Arlington Heights, 69 Ohio St. 2d 539, 541, 433 N.E.2d 198, 200 (1982) (“In prohibiting all forms of offsite billboard advertising, the ordinance is thus inescapably directed to the content of protected speech”). Ultimately, it seems quite unlikely that Justice Alito’s quick recital of some content-neutral rules purported to pre-emptively decide an issue that had long perplexed federal and state courts.

4

Near the end of its analysis, the majority invokes an allegedly “unbroken tradition of on-/off-premises distinctions” that it claims “counsels against” faithful application of Reed. Ante, at 12. To be sure, history and tradition are relevant to identifying and defining those “few limited areas” where, “[f ]rom 1791 to the present,” “the First Amendment has permitted restrictions upon the content of speech.” Brown v. Entertainment Merchants Assn., 564 U.S. 786, 791 (2011) (internal quotation marks omitted); see supra, at 2, n. 1. But the majority openly admits that off-premises regulations “were not present [at] the founding.” Ante, at 12. And while it asserts that “large outdoor advertisements proliferated in the 1800s,” ibid., it offers no evidence of any content-based restrictions from that period, let alone off-premises restrictions on noncommercial speech. The earliest example of an off-premises restriction that the majority cites arose in Packer Corp. v. Utah, 285 U.S. 105 (1932), but that case involved a restriction on commercial advertising and did not even feature a First Amendment claim. See id., at 108–112.

Ultimately, the majority’s only “historical” support is that regulations like Austin’s “proliferated following the enactment of the Highway Beautification Act of 1965.” Ante, at 2. The majority’s suggestion that the First Amendment should yield to a speech restriction that “proliferated”— under pressure from the Federal Government—some two centuries after the founding is both “startling and dangerous.” United States v. Stevens, 559 U.S. 460, 470 (2010). This Court has never hinted that the government can, with a few decades of regulation, subject “new categories of speech” to less exacting First Amendment scrutiny. Id., at 472.

Regardless, even if this allegedly “unbroken tradition” did not fall short by a century or two, the majority offers no explanation why historical regulation is relevant to the question whether the off-premises restriction is content based under Reed and our modern content-neutrality jurisprudence. If Austin had met its burden of identifying a historical tradition of analogous regulation—as can be done, say, for obscenity or defamation—that would not make the off-premises rule content neutral. It might simply mean that the off-premises rule is a constitutional form of content-based discrimination. But content neutrality under Reed is an empirical question, not a historical one. Thus, the majority’s historical argument is not only meritless but misguided.

C

Despite asserting that the Court of Appeals’ analysis under Reed would “contravene numerous precedents,” ante, at 11, the majority identifies no decision of this Court supporting the idea that a speech restriction is not content based so long as it regulates a sufficiently broad or non- substantive category of communicative content. In fact, there is only one case that could possibly validate the majority’s aberrant analysis: Hill v. Colorado. That Hill is the majority’s only support underscores the danger that today’s decision poses to the First Amendment.

Hill involved a law that prohibited persons outside abortion clinics from knowingly approaching within eight feet of another person without consent “for the purpose of . . . engaging in oral protest, education, or counseling.” 530 U. S., at 707 (internal quotation marks omitted). Hill concluded, implausibly, that this regulation was content neutral.

The majority’s reasoning in this case is just as implausible. The majority asserts that the off-premises rule is not content based because it does not target a sufficiently “specific” or “substantive” category of communications. Ante, at 8. Hill correspondingly held that restrictions on “protest, education, or counseling” were not content-based classifications because they cover “an extremely broad category of communications.” 530 U. S., at 723. The majority also tries to disguise its redefinition of content neutrality by characterizing Austin’s rule as a “neutral, location-based” restriction. Ante, at 6. So too did Hill try to conceal its doctrinal innovation by characterizing the buffer-zone law as a neutral “place restriction.” 530 U. S., at 723. Finally, the majority finds it immaterial that Austin’s rule can be enforced only by “reading a [sign] to determine whether it” contains an off-premises message. Ante, at 8. Hill likewise found it irrelevant that “the content of the oral statements” would need to “be examined to determine whether” the prohibition applied. 530 U. S., at 720.

The parallel between the majority’s opinion and Hill should be discomforting given that Hill represented “an unprecedented departure” from this Court’s First Amendment jurisprudence. Id., at 772 (Kennedy, J., dissenting). Its content-neutrality analysis was, as Justice Scalia explained, “absurd” given that the buffer-zone law was “obviously and undeniably content based.” Id., at 742–743 (dissenting opinion). First Amendment scholars from across the ideological spectrum agree. See, e.g., M. McConnell, Professor Michael W. McConnell’s Response, in K. Sullivan, Sex, Money, and Groups: Free Speech and Association Decisions in the October 1999 Term, 28 Pepperdine L. Rev. 723, 748 (2001) (“The Court said that this statute is content-neutral. I just literally cannot see how they could possibly come to that conclusion”); Colloquium, id., at 750 (Laurence Tribe stating Hill “was slam-dunk simple and slam-dunk wrong”); R. Fallon, Strict Judicial Scrutiny, 54 UCLA L. Rev. 1267, 1298, and n. 174 (2007) (Hill “unconvincingly . . . maintain[ed] that a content-based restriction on speech [was] not really content-based”). And, since Hill, this Court has all but interred its flawed content-neutrality analysis in both McCullen, see supra, at 11, and Reed. See Price v. Chicago, 915 F.3d 1107, 1118 (CA7 2019) (“In the wake of McCullen and Reed, it’s not too strong to say that what Hill explicitly rejected is now prevailing law”).

The majority’s refusal to acknowledge Hill simply underscores the decision’s defunct status. Again, Hill is the only case that could support the majority’s ill-conceived content-neutrality analysis, and yet the majority disclaims reliance on it. Lower courts should take the majority’s disclaimer at face value: Hill is “a decision that we do not cite.” Ante, at 13. And today’s decision amounts to little more than an ad hoc exemption for the “location-based” and supposedly “content-agnostic on-/off-premises distinction.” Ibid.

Even so, the majority’s approach should offer little comfort because arbitrary carveouts from Reed undermine the “clear and firm rule governing content neutrality” that we understood to be “an essential means of protecting the freedom of speech.” 576 U. S., at 171. The majority’s deviation from that “clear and firm rule” poses two serious threats to the First Amendment’s protections.

First, transforming Reed’s clear definition of “content based regulation” back into an opaque and malleable “term of art” turns the concept of content neutrality into a “vehicl[e] for the implementation of individual judges’ policy preferences.” Tennessee v. Lane, 541 U.S. 509, 556 (2004) (Scalia, J., dissenting). Hill exemplifies this danger. See 530 U. S., at 742 (Scalia, J., dissenting) (“I have no doubt that this regulation would be deemed content based in an instant if the case before us involved antiwar protesters, or union members seeking to ‘educate’ the public about the reasons for their strike”). The majority’s approach in this case is cut from the same cloth. As the majority transparently admits, it seeks to “apply [our] precedents to reach the ‘commonsense’ result” and avoid what it perceives as a “bizarre result.” Ante, at 13 (emphasis added). But Reed mandates a “commonsense” test for content neutrality even if the result is that “laws that might seem entirely reasonable will sometimes be struck down.” 576 U. S., at 163, 171 (internal quotation marks omitted).

Second, sanctioning certain content-based classifications but not others ignores that even seemingly reasonable content-based restrictions are ready tools for those who would “suppress disfavored speech.” Id., at 167; see also Hill, 530 U. S., at 743 (Scalia, J., dissenting) (“ ‘The vice of content-based legislation . . . is not that it is always used for invidious, thought-control purposes, but that it lends itself to use for those purposes’ ”). This is because “the responsibility for distinguishing between” permissible and impermissible content “carries with it the potential for invidious discrimination of disfavored subjects.” Cincinnati v. Discovery Network, Inc., 507 U.S. 410, 423–424, n. 19 (1993). That danger only grows when the content-based distinctions are “by no means clear,” giving more leeway for government officials to punish disfavored speakers and ideas. Ibid.

The content-based distinction drawn by Austin’s off-premises speech restriction is “by no means clear,” ibid., and plainly lends itself “to suppress[ing] disfavored speech,” Reed, 576 U. S., at 167. As the Court of Appeals noted, Austin’s “prepared counsel” “struggled to answer whether” signs conveying messages like “ ‘God Loves You,’ ” “ ‘Vote for Kathy,’ ” or “ ‘Sally makes quilts here and sells them at 3200 Main Street’ ” would be regulated as off- premises signs. 972 F. 3d, at 706. Before us, Austin’s counsel had similar difficulties, and amici have proposed dozens of religious and political messages that would be next to impossible to categorize under Austin’s rule. See, e.g., Brief for Alliance Defending Freedom et. al. as Amici Curiae 15–19; Brief for Institute for Justice as Amicus Curiae 3–9. These pervasive ambiguities offer enforcement officials ample opportunity to suppress disfavored views. And they underscore Reed’s warning that “[i]nnocent motives do not eliminate the danger of censorship presented by a facially content-based statute.” 576 U. S., at 167.

*  *  *

Because Reed provided a clear and neutral rule that protected the freedom of speech from governmental caprice and viewpoint discrimination, I would adhere to that precedent rather than risk resuscitating Hill. I respectfully dissent.

2.  In Reed, we acknowledged that some prior decisions had skipped over this facial analysis and applied a justification-focused test. See 576 U. S., at 165–167. But we explained that the justification-focused test implicated a “separate and additional category of laws that, though facially content neutral, [are] content-based regulations [because they] cannot be ‘ “justified without reference to the content of the regulated speech,” ’ or . . . were adopted by the government ‘because of disagreement with the message [the speech] conveys.’ ” Id., at 164 (quoting Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989)). All agree that this second type of content-based regulation is not at issue here.