HM FLORIDA ORL LLC v. Secretary of the Florida Department of Business and Professional Regulation, Defendant-Appellant. (2025)

Justice Potter Stewart famously offered a non-definition of obscenity: “I know it when I see it.” Jacobellis v. Ohio, 378 U.S. 184, 197, 84 S.Ct. 1676, 12 L.Ed.2d 793 (1964) (Stewart, J., concurring). Many know Justice Stewart's quip. But it's not, in fact, the law.

The Constitution demands specificity when the state restricts speech. Requiring clarity in speech regulations shields us from the whims of government censors. And the need for clarity is especially strong when the government takes the legally potent step of labeling speech “obscene.” An “I know it when I see it” test would unconstitutionally empower those who would limit speech to arbitrarily enforce the law. But the First Amendment empowers speakers instead.

Yet Florida's Senate Bill 1438 (the “Act”) takes an “I know it when I see it” approach to regulating expression. The Act prohibits children's admission to “live performances” that Florida considers obscene for minors. But by providing only vague guidance as to which performances it prohibits, the Act wields a shotgun when the First Amendment allows a scalpel at most. And Florida's history of arbitrarily enforcing other, similar laws against performances that are far from obscene only deepens our concerns. We therefore hold that the Act is likely unconstitutional on its face and affirm the lower court's injunction against its enforcement.

I. Background

A. Senate Bill 1438

In 2023, Florida enacted Senate Bill 1438, known also as the Protection of Children Act. See Fla. Laws ch. 2023-94. The Act makes it a misdemeanor to “knowingly admit a child to an adult live performance.” Fla. Stat. § 827.11(3). An “adult live performance” is

any show, exhibition, or other presentation in front of a live audience which, in whole or in part, depicts or simulates nudity, sexual conduct, sexual excitement, or specific sexual activities as those terms are defined in s. 847.001, lewd conduct, or the lewd exposure of prosthetic or imitation genitals or breasts when it:

1. Predominantly appeals to a prurient, shameful, or morbid interest;

2. Is patently offensive to prevailing standards in the adult community of this state as a whole with respect to what is suitable material or conduct for the age of the child present; and

3. Taken as a whole, is without serious literary, artistic, political, or scientific value for the age of the child present.

Id. § 827.11(1)(a). The preexisting Section 847.001, part of a chapter entitled “Obscenity,” in turn defines “nudity,” “sexual conduct,” “sexual excitement,” and “specific sexual activities” in great detail. Id. § 847.001(11), (19), (20), (23). A “child” is “any person ․ younger than 18 years of age.” Id. § 847.001(10).

The Act also allows the Florida Department of Business and Professional Regulation (“FDBPR”) to fine or revoke the licenses of eating, drinking, and lodging establishments that admit a child to an adult live performance. Fla. Laws ch. 2023-94 §§ 1–3 (codified at Fla. Stat. §§ 255.70, 509.261, 561.29).

B. The Act and Drag Shows

Though the Act applies to a range of “adult live performances,” its enactors focused on how it would restrict one particular type of performance: drag shows. When signing the Act into law, Florida's governor described it as being about “adult performances ․ like those drag shows.” FOX 13 Tampa Bay, Full Press Conference: Governor Ron DeSantis Signs Education Bills in Tampa, YouTube, AT 8:22 (May 17, 2023), https://www.youtube.com/watch?v=t1kIP2dd2xc [https://perma.cc/U3LC-K4S8]. One of the Act's legislative sponsors said it would “protect our children by ending the gateway propaganda to this evil—‘Drag Queen Story Time.’ ” State Representative Randy Fine, Facebook (Mar. 3, 2023), https://www.facebook.com/voterandyfine/posts/761831661970637 [https://perma.cc/5ENU-FPTD].

Florida has a history of efforts aimed at restricting drag shows and venues. Before the Act's passage, FDBPR brought administrative proceedings to revoke the liquor licenses of several drag venues. One administrative action alleged a drag show had violated Florida's law against “lewd and lascivious exhibition in the presence of a minor”—despite the government inspectors’ report that “agents did not witness any lewd acts.” Nicholas Nehamas & Ana Ceballos, Florida Undercover Agents Reported No “Lewd Acts” at Drag Show Targeted by DeSantis, Tampa Bay Times (Mar. 20, 2023), https://www.tampabay.com/news/florida-politics/2023/03/20/desantis-drag-show-lewd-liquor-license-complaint-lgbtq [https://perma.cc/GX43-42HR]. Another FDBPR complaint cited “graphic depictions of childbirth and/or abortion” as an example of “sexual conduct, simulated sexual activity, and lewd, vulgar, and indecent displays.”

C. Hamburger Mary's

Plaintiff-Appellant HM Florida-ORL, LLC (“Hamburger Mary's”), operates Hamburger Mary's Restaurant and Bar. According to its initial, verified complaint, Hamburger Mary's regularly hosted a variety of drag performances before Florida passed the Act. On Sundays it presented drag performances it considered “family friendly” and invited children to attend. These shows featured no “lewd activity ․ or anything inappropriate for a child to see.” The restaurant also hosted other “drag-centric performances, comedy sketches, bingo, trivia, and dancing.” Among these, Hamburger Mary's warned its patrons that some were “not suitable for children.”

In response to the Act, though, Hamburger Mary's canceled its family drag shows and barred children from attending any of its other shows. The restaurant feared losing its business or liquor licenses if it violated the Act. After Florida adopted the Act, Hamburger Mary's lost twenty percent of its bookings.

II. Procedural History

Shortly after Florida's governor signed the Act, Hamburger Mary's sued Florida, its governor, and FDBPR Secretary Melanie Griffin in her official capacity. The parties later agreed to the dismissal of all defendants except Defendant-Appellant Griffin.

Hamburger Mary's sued under 42 U.S.C. § 1983 and alleged that the Act was void for vagueness, overbroad, and was a content-based speech regulation that failed strict scrutiny, in violation of the First and Fourteenth Amendments. Based on these allegations, Hamburger Mary's sought an injunction against then-Defendants’ enforcement of the Act. The restaurant also moved for a temporary restraining order and preliminary injunction against enforcement. Then-Defendants responded and moved to dismiss the case.

After considering the filings, the district court issued a preliminary injunction and denied the motion to dismiss. The court enjoined Griffin's enforcement of the Act, against Hamburger Mary's or anyone else.

Griffin appealed and moved in this Court to stay the injunction pending appeal as to any entities other than Hamburger Mary's. A split panel denied that motion for a partial stay. HM Florida-ORL, LLC v. Governor of Fla., No. 23-12160, 2023 WL 6785071 (11th Cir. Oct. 11, 2023) (per curiam). Then Griffin applied to the Supreme Court for a partial stay of the injunction. The Court denied that application. Griffin v. HM Florida-ORL, LLC, ––– U.S. ––––, 144 S. Ct. 1, 217 L.Ed.2d 227 (2023) (mem.).

III. Standard of Review

We review for abuse of discretion a district court's grant of a preliminary injunction. Gonzalez v. Governor of Ga., 978 F.3d 1266, 1270 (11th Cir. 2020). In conducting our review, we evaluate legal conclusions de novo and factual findings for clear error. Id.

A party seeking injunctive relief must also show that it has standing to bring the case. “To establish standing, the plaintiffs must demonstrate a substantial risk that, in the near future, they will suffer an injury that is traceable to a Government defendant and redressable by the injunction they seek.” Murthy v. Missouri, 603 U.S. 43, 49–50, 144 S.Ct. 1972, 219 L.Ed.2d 604 (2024). The strength of proof required differs at separate stages of the litigation; “[a]t the preliminary injunction stage, ․ the plaintiff must make a ‘clear showing’ that she is ‘likely’ to establish each element of standing.” Id. at 58, 144 S.Ct. 1972 (quoting Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 22, 129 S.Ct. 365, 172 L.Ed.2d 249 (2008)).

IV. Discussion

Our discussion proceeds in four parts. First, we explain why Hamburger Mary's has standing to sue here. Second, we show that the restaurant's relocation does not moot its claim. Third, we examine whether Hamburger Mary's First Amendment claim is likely to succeed on the merits. We explain how the Act's “depicts ․ lewd conduct” restriction and its age-variable obscenity standard make it overbroad. Based on that discussion, we conclude that the district court correctly enjoined the Act's enforcement. Last, we discuss the scope of the injunction and show that statewide relief is warranted here.

A. Hamburger Mary's has standing to bring this case.

A threshold issue here, as in every suit, is standing. Under Article III of the Constitution, for us to enjoy jurisdiction over this suit, Hamburger Mary's must have standing. Standing helps assure us that we are hearing a “case[ ]” or “controvers[y].” See Wilding v. DNC Servs. Corp., 941 F.3d 1116, 1124 (11th Cir. 2019) (citing Spokeo, Inc. v. Robins, 578 U.S. 330, 338, 136 S.Ct. 1540, 194 L.Ed.2d 635 (2016)). In particular, standing doctrine has been “developed primarily to ensure that the person seeking to litigate a claim is the ‘right’ person to advance the claim.” Wooden v. Bd. of Regents of Univ. Sys. of Ga., 247 F.3d 1262, 1280 (11th Cir. 2001).

To establish standing, a plaintiff must show three familiar things: “(1) ․ an injury in fact, (2) that is fairly traceable to the challenged conduct of the defendant, and (3) that is likely to be redressed by a favorable judicial decision.” Spokeo, 578 U.S. at 338, 136 S.Ct. 1540 (citing Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992)). We address each element in turn. But our analysis concentrates on the injury-in-fact requirement because it is the parties’ main point of contention.

i. Hamburger Mary's has sufficiently alleged injury in fact to establish standing for its pre-enforcement suit.

The injury-in-fact requirement could create a dilemma for anyone seeking to challenge a statute: either violate the law to provoke an enforcement action and establish standing or steer clear of a violation and avoid injury but lose standing to sue. Because of this conundrum, special standing rules govern pre-enforcement suits like this one.

A plaintiff bringing a pre-enforcement challenge to an enacted law must show it intends to “engage in a course of conduct” that the Constitution arguably protects but a law prohibits or otherwise unconstitutionally burdens, and “a credible threat of prosecution” exists under that law. Babbitt v. United Farm Workers Nat'l Union, 442 U.S. 289, 298, 99 S.Ct. 2301, 60 L.Ed.2d 895 (1979). Of course, no party wants to tell a court that it intends to violate a law, and we do not ask parties to do so. Indeed, “[n]othing in [the Supreme] Court's decisions requires a plaintiff who wishes to challenge the constitutionality of a law to confess that he will in fact violate that law.” Susan B. Anthony List v. Driehaus, 573 U.S. 149, 163, 134 S.Ct. 2334, 189 L.Ed.2d 246 (2014) (citing Babbitt, 442 U.S. at 301, 99 S.Ct. 2301).

“[W]e apply the injury-in-fact requirement most loosely when First Amendment rights” are at stake, “lest free speech be chilled even before the law or regulation is enforced.” Harrell v. The Fla. Bar, 608 F.3d 1241, 1254 (11th Cir. 2010) (citing Hallandale Prof. Fire Fighters Loc. 2238 v. City of Hallandale, 922 F.2d 756, 760 (11th Cir. 1991)). So self-censorship—that is, “forgo[ing] expression in order to avoid enforcement consequences”—can constitute an injury for standing purposes. Id. (quoting Pittman v. Cole, 267 F.3d 1269, 1283 (11th Cir. 2001)).

In that respect, a plaintiff need not show that its self-censored speech would definitely run afoul of a challenged law, but only that its speech is “arguably proscribed.” Driehaus, 573 U.S. at 162, 134 S.Ct. 2334. In other words, self-censorship establishes standing when “ ‘the operation or enforcement’ ․ of the government policy would cause a reasonable would-be speaker to ‘self-censor.’ ” Speech First, Inc. v. Cartwright, 32 F.4th 1110, 1120 (11th Cir. 2022) (cleaned up) (quoting Ga. Latino All. for Hum. Rts. v. Governor of Ga., 691 F.3d 1250, 1257 (11th Cir. 2012), and Wollschlaeger v. Governor, Fla., 848 F.3d 1293, 1305 (11th Cir. 2017) (en banc)). That is, we ask whether a state policy “objectively chills” protected expression. Id. (citing Speech First, Inc. v. Fenves, 979 F.3d 319, 330–35 (5th Cir. 2020)).

As for the requirement of a credible threat of prosecution, “[i]f a challenged law or rule was recently enacted ․ an intent to enforce the rule may be inferred.” Harrell, 608 F.3d at 1257 (citing Int'l Soc'y for Krishna Consciousness of Atlanta v. Eaves, 601 F.2d 809, 821 (5th Cir. 1979)). Put simply, a credible threat of prosecution inheres in any recently passed law because we “assume that law enforcement agencies will not disregard ․ a recent expression of the legislature's will.” Id. (alteration in original) (quoting Eaves, 601 F.2d at 821).

Finally, vagueness challenges under the First Amendment have their own standing standards. To establish an injury from self-censorship because of a statute's vagueness, a plaintiff must show “(1) he seriously wishes to [speak]; (2) such [speech] would arguably be affected by the rules, but the rules are at least arguably vague as they apply to him[;] and (3) there is at least a minimal probability that the rules will be enforced, if they are violated.” Id. at 1254 (citations and footnote omitted).

Despite these lenient standards, Griffin argues that Hamburger Mary's lacks standing for a preliminary injunction. The crux of Griffin's argument is that Hamburger Mary's has not shown any intent to engage in “arguably proscribed” speech. In support of its contention, Griffin points to the restaurant's statement that its family-friendly performances contained “no lewd activity, sexually explicit” content, or anything else inappropriate for a child. And so, Griffin reasons, no reasonable person could think such performances fall within the ambit of the Act. By canceling these performances, Griffin argues, the restaurant is doing what I have previously characterized as “flopping”—faking an injury. See Am. All. for Equal Rts. v. Fearless Fund Mgmt., LLC, 103 F.4th 765, 780 (11th Cir. 2024) (Rosenbaum, J., dissenting).

We disagree; Hamburger Mary's has shown enough on this record to establish a real injury under governing standards. We base our conclusion on four things: the facts Hamburger Mary's alleged in its verified complaint, the vague text of the statute, Florida officials’ prior enforcement actions and statements about drag shows, and the lenient standards for standing in First-Amendment suits.

To be sure, Hamburger Mary's believes its family-friendly performances contain no “lewd” or “sexually explicit” content. And the restaurant would surely argue exactly that if it had to defend against an enforcement action under the Act. But Hamburger Mary's also makes clear that its drag shows, like almost all shows in the genre, involve performers wearing “clothing more conventionally worn by the other sex” and often “[p]rosthetic breasts.” And it notes that some people might consider even “a man in a dress” reading to young children to violate the statute.

Bearing in mind the speech Hamburger Mary wishes to undertake, we ask whether it was “reasonable” for the restaurant to self-censor. Speech First, 32 F.4th at 1120. And based on the considerations our Court has looked to before, we believe it was.

First, the Act's vagueness makes Hamburger Mary's self-censorship more reasonable. A speech restriction's “imprecision exacerbates its chilling effect.” Id. at 1121. This consequence of vague speech laws of course implicates their constitutionality. See, e.g., Grayned v. City of Rockford, 408 U.S. 104, 109, 92 S.Ct. 2294, 33 L.Ed.2d 222 (1972). And it also affects our self-censorship standing analysis by making a broader range of self-censorship a “reasonable” response. Speech First, 32 F.4th at 1121.

As we explain in greater detail below, the Act's vague terms like “depicts ․ lewd conduct” and “value for the age of the child present” make the law's scope ambiguous and potentially inclusive of even Hamburger Mary's “family-friendly” drag performances. Fla. Stat. § 827.11(1)(a). After all, no less than the federal government has recently indicated its intent to police firm “sex-based distinctions,” including those related to people's “appearance.” Exec. Order No. 14168, 90 Fed. Reg. 8615 (Jan. 20, 2025). The possibility that FDBPR might view a gender-bending but chaste drag performance as “lewd” and lacking “value” is far from an unreasonable conclusion.

In evaluating the reasonableness of self-censorship, we have also looked to the severity of the potential consequences. In one case involving a school board's speech policy, we described potential penalties of “up to 60 days in jail and a $500 fine” as “severe.” Moms for Liberty - Brevard Cnty. v. Brevard Pub. Schs., 118 F.4th 1324, 1330 (11th Cir. 2024). If those are severe, the penalties for violations of the Act—business-license suspension and fines of up to $5,000 for the first offense or a first-degree misdemeanor prison sentence of up to one year—are grievous. See Fla. Stat. §§ 509.261(10), 561.29(1)(l), 775.082(4)(a), 827.11(4).

When we determine whether a plaintiff faces a “credible threat” of enforcement, we also look to the statements and prior actions of state officials that shed light on potential enforcement plans. For example, if those responsible for enforcing a challenged law say clearly and definitively that a would-be plaintiff's planned speech would not violate the law, the plaintiff may lack standing (or a lawsuit may be mooted). See, e.g., Graham v. Butterworth, 5 F.3d 496, 500 (11th Cir. 1993); Wilson v. State Bar of Ga., 132 F.3d 1422, 1428–29 (11th Cir. 1998).

On the other hand, anything less than a “clear disavowal of enforcement” does not divest a plaintiff of standing. Brown v. Kemp, 86 F.4th 745, 769 (7th Cir. 2023). See also Wollschlaeger, 848 F.3d at 1306 (nonbinding letter disavowing enforcement in general terms did not deprive plaintiff of standing, especially given defendant's previous contradictory positions); Speech First, 32 F.4th at 1121–22 (counsel's equivocation as to whether particular statements violated speech code illustrated credible threat of enforcement).

Griffin has done little to mitigate the threat of enforcement. A declaration from FDBPR's deputy general counsel asserted that the drag shows FDBPR had previously deemed problematic “differ significantly” from those Hamburger Mary's describes. But the declaration pointedly did not clarify the boundaries of impermissible performances nor commit to nonenforcement. And it would defeat decades of First Amendment jurisprudence to allow states to impose laws whose “breadth—and slipperiness” obfuscate their reach, then dodge lawsuits by saying plaintiffs are confused as to a law's scope. Speech First, 32 F.4th at 1122. The government cannot shroud rules in foggy language and then blame would-be speakers for their fears of what may lurk in the fog.

And FDBPR's prior enforcement actions against other drag venues do little to assuage a reasonable speaker's fears. True, as FDBPR has argued, those shows seemed to be somewhat more risqué than the Sunday “family” shows at Hamburger Mary's. And of course, the FDBPR's prior administrative actions alleged violations of different statutes. Still, two things about those enforcement actions reinforce the “credible threat” of FDBPR's enforcement of the Act.

First, in at least one instance, FDBPR alleged that a drag show featured “lewd” acts despite its own inspector's determination to the contrary. See Nehamas & Ceballos, supra. FDBPR leadership need not agree with its own inspector's conclusion. But this internal disagreement points to the malleability of the speech statutes regulating “lewd” conduct (like the Act) that the FDBPR enforces.

Second, FDBPR's prior enforcement actions reinforce a reasonable observer's sense that when the Department watches drag shows, it is likely to see something it thinks lewd, regardless of its actual content. Twice, FDBPR alleged a particular drag performance featured “lewd, vulgar, and indecent displays” including “graphic depictions of childbirth and/or abortion.”

But based on FDBPR's photographs (provided as exhibits), the offending depiction seems to have been a performance by a drag artist named “Jimbo.” One of Jimbo's signature acts involves donning Marcel Marceau-like makeup, a prosthetic stomach and backside, and a stretchy, full-body white suit (leaving no skin or prosthetic skin visible other than the face). Jimbo dances and prances onstage, lip-syncing to Björk's cover of Betty Hutton's 1951 song “It's Oh So Quiet,” before undoing a hidden zipper on the stomach's underside and pulling from within ․ a pile of baloney. See Kathy Sparkles, Baloney with Jimbo from Drag Race, YouTube (Nov. 27, 2022), https://www.youtube.com/watch?v=wk5H4LvHmFU [https://perma.cc/SRP8-DP8B]. Perhaps some may consider Jimbo's baloney birth a bit odd (and hammy in every sense of the word). But FDBPR seems to think the act “outrage[s] the sense of public decency,” is “nasty, suggestive, and indecent,” or is “obscene,” even for adults.¹ If FDBPR believes Jimbo's baloney birth was legally problematic even before the Act, it is reasonable for Hamburger Mary's to fear that FDBPR might enforce the new law—presumably passed to expand the range of prohibited performances—against even tame drag shows. And standing law doesn't require a party in Hamburger Mary's pickle to take a chance and find out it was right to fear FDBPR's enforcement.

Legislators’ descriptions of the Act also show that Hamburger Mary's fear is reasonable. One of the Act's key legislative proponents declared the law was meant to “end[ ] ․ ‘Drag Queen Story Time.’ ” State Representative Randy Fine, supra. Drag Queen Story Time, also called Drag Queen Story Hour, involves drag performers simply reading children's books to young audiences. See generally Jaweed Kaleem, How Drag Queen Story Hour Became a Battle over Gender, Sexuality and Kids, L.A. Times (Feb. 22, 2023, 3:00 AM PT), https://www.latimes.com/world-nation/story/2023-02-22/drag-queen-story-hour [https://perma.cc/4DKD-Y8YC]. If any legally kid-friendly drag performances exist under the Act—and Griffin insists they do—Drag Queen Story Time would seem to be one. But at least one key legislator disagrees. This lends credibility to Hamburger Mary's fear of an enforcement action.²

Hamburger Mary's briefing before this Court also provides details that would strengthen our standing conclusion if Hamburger Mary presented record evidence of them in the district court. The restaurant explains that before the Act's passage, its longstanding policy was to label some drag performances “18+” but to allow minors to attend those shows with adult supervision. Hamburger Mary's relied on an “honor system” to enforce this policy.

But after the Act's passage, the restaurant barred children from all its shows—including those not previously labeled “18+”—regardless of adult supervision. It hired private security guards to conduct ID checks and enforce this policy. If shown by record evidence, these facts—suggesting the Act led Hamburger Mary's to ban minors entirely from its adult-oriented performances and pay guards to enforce the ban—would leave little reason even to question the restaurant's standing. But we don't consider these assertions because the record does not contain evidence of them.

Still, even relying on only Hamburger Mary's verified complaint, we are satisfied that the restaurant has shown injury in fact “with the manner and degree of evidence required” at this early, preliminary-injunction stage of litigation. Bischoff v. Osceola County, 222 F.3d 874, 878 (11th Cir. 2000) (quoting Lujan, 504 U.S. at 561, 112 S.Ct. 2130). The restaurant has alleged that it hosts drag shows it considers child-friendly but that FDBPR and proponents of the Act may view very differently. Given the Act's severe penalties and vague scope, Hamburger Mary's self-censored to avoid the threat of litigation. This threat was credible, Hamburger Mary's actions were objectively reasonable, and its injury in fact from self-censorship was real.

Griffin resists this conclusion. She compares Hamburger Mary's situation with those of the plaintiffs in Driehaus, whom the Supreme Court found enjoyed standing, and the would-be intervenors in Younger v. Harris, 401 U.S. 37, 91 S.Ct. 746, 27 L.Ed.2d 669 (1971), whom the Supreme Court concluded lacked standing. As Griffin sees things, Hamburger Mary's situation is more like that of the would-be intervenors in Younger than like that of the plaintiffs in Driehaus. We disagree.

In Driehaus, the Supreme Court held that a plaintiff had standing for a pre-enforcement challenge to an Ohio statute that regulated statements about election candidates because (1) the plaintiff had been the subject of past proceedings before the enforcing agency and (2) the plaintiff “alleged an intent to engage in the same speech” that, when made by a different party, had triggered an enforcement action. Driehaus, 573 U.S. at 166–67, 134 S.Ct. 2334.

Like the Driehaus plaintiff, Hamburger Mary's claim of standing rests in part on FDBPR's prior enforcement actions against similar speakers. Id. But Griffin argues that this history is irrelevant because FDBPR's prior enforcement actions were against different parties and involved different statutes and different performances than the ones at issue here. In Griffin's view, these differences put Hamburger Mary's in the same situation as the would-be plaintiffs in Younger.

We don't see things that way. The would-be intervenors in Younger sought to intervene in the prosecution of a different individual under an anti-leafletting statute. 401 U.S. at 41, 91 S.Ct. 746. The Court found they lacked standing because they claimed only that they “fe[lt] inhibited” by the statute, not that they “have ever been threatened with prosecution, that a prosecution is likely, or even that a prosecution is remotely possible.” Id. at 42, 91 S.Ct. 746.

In contrast to the Younger would-be intervenors, Hamburger Mary's unambiguously points to the threat of prosecution as the reason for its self-censorship. It doesn't allege a generalized sense of feeling “inhibited” like the Younger would-be intervenors. The restaurant says it “cannot take the chance that [its] business or liquor licenses would be suspended,” let alone risk the Act's “criminal penalties.”

True, Hamburger Mary's, unlike the Driehaus plaintiff, does not allege its intent to engage in the “same speech” as a defendant in another FDBPR action. But nothing in Driehaus comes close to suggesting that pre-enforcement standing requires an intent to engage in the exact speech that has already provoked one prosecution. Driehaus, 573 U.S. at 164–67, 134 S.Ct. 2334. Indeed, the Court said only that prior enforcement against the same conduct is “good evidence”—not necessary evidence—of a credible enforcement threat. Id. at 164, 134 S.Ct. 2334.

Not only that, but Griffin's argument that FDBPR took its prior enforcement actions under different statutes does not necessarily weigh in her favor. To start, Florida passed the Act presumably to solve a problem that the other statutes didn't already address. So whatever the scope of the Act's proscriptions, they likely extend beyond the obscenity statutes that FDBPR already previously enforced against drag shows. See, e.g., Fla. Stat. §§ 847.011 (“Prohibition of certain acts in connection with obscene, lewd, etc., materials”), 847.0133 (“Protection of minors; prohibition of certain acts in connection with obscenity”). Otherwise, the Act would have “no job to do.” Doe v. Chao, 540 U.S. 614, 623, 124 S.Ct. 1204, 157 L.Ed.2d 1122 (2004). For that reason, FDBPR's enforcement of other obscenity statutes against drag performances suggests that the Act only increases the threat of enforcement.

Plus, FDBPR's enforcement history casts a shadow over Hamburger Mary's pleadings here. Hamburger Mary's pleadings or evidence about its past performances or admissions policy could become fodder for an FDBPR enforcement action—even if Hamburger Mary's succeeded in obtaining an injunction against enforcement of the Act.³ But as we've explained, a plaintiff seeking to challenge a law's constitutionality need not “confess that he will in fact violate that law.” Driehaus, 573 U.S. at 163, 134 S.Ct. 2334. For similar reasons, we won't ignore a plaintiff's alleged self-censorship injury when doing so would effectively require a plaintiff to disclose facts that might provoke a prosecution under other, related statutes.

At bottom, we conclude that Hamburger Mary's has sufficiently shown injury from its reasonable self-censorship in the face of a vague statute and broad enforcement history.

ii. Hamburger Mary's injuries are traceable to the Secretary and can be redressed by relief against her.

Secretary Griffin presents no reason that the traceability and redressability prongs of our standing analysis should come out differently than the injury-in-fact prong. She contends, only briefly, that for “the same reasons” the restaurant has suffered no injury, its claimed harms are neither traceable to Griffin nor redressable by an injunction against her. But as we have explained, Hamburger Mary's has shown it has experienced an injury in fact. And for those same reasons, that injury is traceable to Griffin and redressable by an injunction against her.

Amicus America First Legal Foundation (“America First”) makes a more detailed argument against Hamburger Mary's bid for standing. We've already addressed many of America First's arguments above and don't rehash them here.

America First does present one argument, though, that Griffin didn't make (perhaps tellingly so). In the interest of completeness, we address it now.

America First notes that we must analyze Hamburger Mary's standing in relation to its suit against Secretary Griffin, the only current defendant. So Hamburger Mary's must show that its asserted injury is traceable to Griffin's conduct, not just the Act in the abstract.

But we have made clear that when a plaintiff's asserted injury flows from steps it takes to comply with a law, that injury is generally traceable to government officials with the authority to enforce that law. To be sure, we have previously denied standing on traceability grounds when plaintiffs challenge statutes by suing government officials who aren't responsible for enforcement. See, e.g., Support Working Animals, Inc. v. Governor of Fla., 8 F.4th 1198, 1203 (11th Cir. 2021) (denying standing to sue state Attorney General because she “neither has the authority to enforce” challenged provision “nor has done anything else to cause the plaintiffs’ harm”); Lewis v. Governor of Ala., 944 F.3d 1287, 1299 (11th Cir. 2019) (en banc) (denying standing to sue state Attorney General in challenge to statute that “provides for no enforcement mechanism whatsoever, and ․ certainly envisions no role for the Attorney General”).

But here, the Act's text unambiguously reveals that the Secretary is the proper defendant for this suit because portions of the Act's enforcement proceed under her authority. Florida law gives the Division of Hotels and Restaurants and the Division of Alcoholic Beverages and Tobacco authority to enforce the Act's provisions against licensed businesses. See Fla. Stat. §§ 509.261(10)(a), 561.29(1)(l). Each division is part of the FDBPR. Id. § 20.165(2). And each division's director is appointed by the Secretary of FDBPR—Griffin. Id. § 20.165(3). The directors are “responsible” to Secretary Griffin. Id.

And when FDBPR filed administrative complaints against drag venues in the past, each bore a director's name. Hamburger Mary's credible threat of prosecution emanates from FDBPR, which Griffin leads and others “under her control” carry out. Cf. Support Working Animals, 8 F.4th at 1204. In sum, the injury here is traceable to Griffin and we can redress it with injunctive relief against her.

B. Hamburger Mary's suit is not moot.

Before reaching the merits, we address one more justiciability concern: mootness. “Because a case or controversy must exist throughout all stages of litigation, we must ensure—up until the moment our mandate issues—that intervening events have not mooted the appeal ․” Norwegian Cruise Line Holdings Ltd v. State Surgeon Gen., Fla. Dep't of Health, 55 F.4th 1312, 1315 (11th Cir. 2022) (cleaned up) (quoting Vital Pharms., Inc. v. Alfieri, 23 F.4th 1282, 1288 (11th Cir. 2022)). “A case becomes moot—and therefore no longer a ‘Case’ or ‘Controversy’ for purposes of Article III—when the issues presented are no longer live or the parties lack a legally cognizable interest in the outcome.” Id. (cleaned up) (quoting Already, LLC v. Nike, Inc., 568 U.S. 85, 91, 133 S.Ct. 721, 184 L.Ed.2d 553 (2013)). Viewed from another angle, “a case is moot when it no longer presents a live controversy with respect to which the court can give meaningful relief.” Soliman v. United States ex rel. INS, 296 F.3d 1237, 1242 (11th Cir. 2002) (per curiam) (quoting Fla. Ass'n of Rehab. Facilities, Inc. v. Fla. Dep't of Health & Rehab. Servs., 225 F.3d 1208, 1216–17 (11th Cir. 2000)).

The potential for mootness here stems from Hamburger Mary's closure and planned reopening during the pendency of this suit. In May 2024, Hamburger Mary's announced that it planned to close its Orlando location but hoped to reopen at a nearby location soon. Hamburger Mary's identified a new location in Kissimmee and sent the building's owner a letter of intent to lease the space. At the time, Hamburger Mary's hoped to reopen in the new location in mid-July. The restaurant publicized the planned Kissimmee move on social media while it continued to produce several drag events at other venues. Repairs at the Kissimmee location are taking longer than expected, but as of December 2024, the location's owner confirmed that repair work was ongoing.

Secretary Griffin argues that Hamburger Mary's temporary closure and relocation moot its case. We disagree. Two key Supreme Court cases explain why.

In City of Erie v. Pap's A.M., the Court considered a strip club's challenge to a town's public-indecency ordinance. 529 U.S. 277, 120 S.Ct. 1382, 146 L.Ed.2d 265 (2000). The Pennsylvania Supreme Court partially enjoined enforcement of the ordinance, and the city appealed. Id. at 286–87, 120 S.Ct. 1382. After the Supreme Court granted certiorari, the club submitted an affidavit stating that it had closed, and it moved to dismiss the case as moot. Id. at 287, 120 S.Ct. 1382.

The Court denied the motion. It reasoned that while the plaintiff had closed the club, as a business entity it was “still incorporated ․ and it could again decide to operate a nude dancing establishment.” Id. at 287, 120 S.Ct. 1382. Plus, the Court found that the fact that the plaintiff—who brought the case initially and had won below—was the one seeking mootness represented a potential attempt to “manipulate the Court's jurisdiction to insulate a favorable decision from review.” Id. at 288, 120 S.Ct. 1382. Denying mootness prevented such game-playing.

The Court clarified Pap's A.M. the next year. In City News & Novelty, Inc. v. City of Waukesha, the plaintiff, an “adult-oriented shop,” sued the city after it declined to renew the shop's “adult business license[ ].” 531 U.S. 278, 281–82, 121 S.Ct. 743, 148 L.Ed.2d 757 (2001). State courts upheld the denial, and the shop petitioned the Court for certiorari. Id. at 282, 121 S.Ct. 743. While the petition was pending, the store informed the city that it would “withdraw its renewal application and close its business” because the city had granted a license to a competitor business that would have made it hard for the store to succeed. Id. at 282–83, 121 S.Ct. 743.

After the Court granted certiorari, the city argued the case had become moot. The Court agreed. Although it noted that the adult store was like the Pap's A.M. plaintiff in that it had not firmly foreclosed the possibility of reopening, the Court recognized that mooting the case would not have left intact an adverse judgment against the city, which won below. Id. at 283–84, 121 S.Ct. 743. So City News clarified that mere “speculation” about a closed business's reopening would not overcome mootness in the absence of possible attempted jurisdictional manipulation as in Pap's A.M. Id. at 285, 120 S.Ct. 1382.

The key word in City News, Pap's A.M., and their progeny is “speculation.” “Although a ‘party need not show with certainty that the situation will recur,’ a ‘speculative possibility is not a basis for retaining jurisdiction over a moot case.’ ” In re Operation of Mo. River Sys. Litig., 421 F.3d 618, 631–32 (8th Cir. 2005) (cleaned up) (quoting McCarthy v. Ozark Sch. Dist., 359 F.3d 1029, 1036 (8th Cir. 2004)). When a business's closure means that a court order “will neither presently affect the parties’ rights nor have a more-than-speculative chance of affecting them in the future,” the case is moot. Munsell v. Dep't of Agric., 509 F.3d 572, 583 (D.C. Cir. 2007) (quoting 21st Century Telesis Joint Venture v. FCC, 318 F.3d 192, 198 (D.C. Cir. 2003)) (internal quotation marks omitted in original).

In cases involving businesses that pause operations but may resume them, courts take a common-sense approach to evaluating mootness. The business's own actions and the concreteness of its reopening plans carry significant weight. Future injuries may be speculative when a plaintiff expresses a “desire” to go back into business but shows “no clear plans” to do so. Id. at 582–83. But a company that files for bankruptcy yet shows its ability and intent to resume operations after a corporate reorganization retains an interest in an ongoing suit (even when the reopening depends on the suit's outcome). Id. at 583 (discussing Supreme Beef Processors, Inc. v. USDA, 275 F.3d 432, 436–37 (5th Cir. 2001)).

So while future events “up until the moment our mandate issues” could moot this case like any other, Vital Pharms., Inc., 23 F.4th at 1288, at this time, Hamburger Mary's continuing interest prevents mootness. The restaurant publicly expressed its plans to reopen even before it closed its Orlando location. In fact, Hamburger Mary's described itself as going through not a closure but a “move.” Plus, the proprietors engaged in “many conversations” with the owner of Hamburger Mary's intended new location and ultimately sent a letter of intent to lease the property. Not only that, but repairs to the restaurant are ongoing. As with the bankrupt plaintiff that convincingly showed its ability and desire to resume operations, Hamburger Mary's reopening is not mere speculation.

Hamburger Mary's also has a continuing interest in this suit because it has remained involved with drag performances while its restaurant is closed. The company has continued “produc[ing] ․ drag events in partnership with other venues.”

Secretary Griffin insists this fact cannot give rise to liability because the Act bars “admit[ting] a child” to adult live performances—not “producing” them—and an “establishment's proprietor is the entity admitting guests to its show.” But it is hardly evident that a producer's responsibilities do not include controlling admission within the scope of the Act. Cf. What Does a Producer Do?, Get Into Theatre (June 4, 2024), https://perma.cc/H3JJ-AXLE (listing producer's responsibilities such as “[s]etting ticket prices” and “[e]nsuring legal compliance”). If anything, the Secretary's argument only highlights the Act's use of yet another vague term that obscures its scope.

And even if the Secretary's argument in this respect had legs, that wouldn't change the fact that Hamburger Mary's is actively trying to open a new location where it intends to continue its performances if they don't violate the law.

Proving mootness is a “heavy burden” that falls on the movant—in this case, the Secretary. Norwegian Cruise Line, 55 F.4th at 1314. The Secretary has not carried that burden here.

C. Hamburger Mary's is likely to succeed on the merits of its claim.

Now that we've resolved any justiciability concerns, we turn to the merits of this case. Griffin appeals the district court's decision to preliminarily enjoin enforcement of the Act.

A preliminary injunction generally requires a movant to establish four things: (1) it is substantially likely to succeed on the merits; (2) it will suffer an irreparable injury unless the court grants the injunction; (3) “the harm from the threatened injury outweighs the harm the injunction would cause the opposing party; and (4) the injunction would not be adverse to the public interest.” Gonzalez, 978 F.3d at 1270–71 (citations omitted). The third and fourth factors merge when a party seeks an injunction against the government. Id.

Here, the district court found that all the factors supported granting a preliminary injunction. On appeal, Griffin challenges only Hamburger Mary's likelihood of success on the merits. So we accept the district court's findings as to the other factors.

After careful review of the Act and with the benefit of oral argument, we conclude that Hamburger Mary's facial challenge is likely to succeed. As we explain, the Act is substantially overbroad. Two provisions—a vague restriction on prohibitions of certain “depict[ions of] ․ lewd conduct” and a fine-grained yet ambiguous standard of what speech is appropriate for which children—make it so. Fla. Stat. § 827.11(1)(a). The provisions’ vagueness threatens a broad range of protected speech, even if the law has some permissible applications at its core. These provisions turn the Act into an “I know it when I see it” law. But the Constitution requires more clarity.

i. To succeed on its facial challenge to the Act, Hamburger Mary's must show its unconstitutional applications substantially outweigh its constitutional ones.

Before we analyze the Act, we first explain the applicable standard for Hamburger Mary's facial challenge to the Act. Because facial challenges have a broad impact, they are “hard to win.” Moody v. NetChoice, LLC, 603 U.S. 707, 723, 144 S.Ct. 2383, 219 L.Ed.2d 1075 (2024).

But the stringent standard for facial challenges softens somewhat in First Amendment litigation, to secure “breathing room for free expression.” Id. (quoting United States v. Hansen, 599 U.S. 762, 769, 143 S.Ct. 1932, 216 L.Ed.2d 692 (2023)). In First Amendment facial challenges, we ask whether the challenged law “prohibits a substantial amount of protected speech relative to its plainly legitimate sweep.” Id. (quoting Hansen, 599 U.S. at 770, 143 S.Ct. 1932). In other words, laws abridging First Amendment rights must be sufficiently precise: “[p]recision of regulation must be the touchstone in an area so closely touching our most precious freedoms.” NAACP v. Button, 371 U.S. 415, 438, 83 S.Ct. 328, 9 L.Ed.2d 405 (1963); see also Rodney A. Smolla, Smolla & Nimmer on Freedom of Speech § 6.1 (2024). “I know it when I see it” is not acceptable.

The reasons for demanding precision are so well known that they have leapt from legal language into the layperson's lexicon. Speech regulations can produce a “chilling effect,” discouraging people from speaking their mind even when their speech does not actually fall within the four corners of a prohibition. Counterman v. Colorado, 600 U.S. 66, 75, 143 S.Ct. 2106, 216 L.Ed.2d 775 (2023).

Laws can chill speech in at least three ways. Id. A would-be speaker may be unable to tell whether the law actually prohibits her speech, so she may stay silent. Id. She may intend to speak in ways that she knows to be legally permissible but fear that the judicial system will “err, and count speech ․ that is permissible as instead not.” Id. (citing Philadelphia Newspapers, Inc. v. Hepps, 475 U.S. 767, 777, 106 S.Ct. 1558, 89 L.Ed.2d 783 (1986)). Or she may be confident in winning a potential legal challenge but stay silent to avoid the time and expense of litigation. Id.

The desire to avoid chilling protected speech animates several related doctrines that can invalidate speech laws on their face. Each flows from another way that the Constitution demands precision when the government regulates speech.

The first doctrine, overbreadth, permits “facial invalidation” of a speech law whose “unconstitutional applications ․ [are] substantially disproportionate to the statute's lawful sweep.” Hansen, 599 U.S. at 770, 143 S.Ct. 1932 (2023).

The second doctrine, vagueness, demands that statutes give “fair notice” to speakers about what speech will run afoul of the law. FCC v. Fox Television Stations, Inc., 567 U.S. 239, 253, 132 S.Ct. 2307, 183 L.Ed.2d 234 (2012). Put another way, a law is impermissibly vague “if it fails to provide people of ordinary intelligence a reasonable opportunity to understand what conduct it prohibits.” Hill v. Colorado, 530 U.S. 703, 732, 120 S.Ct. 2480, 147 L.Ed.2d 597 (2000).

The void-for-vagueness rule is a principle of criminal law generally. But it applies with “heightened” vigor to laws touching on protected speech. See Brown v. Ent. Merchs. Ass'n, 564 U.S. 786, 793, 131 S.Ct. 2729, 180 L.Ed.2d 708 (2011) (citing Winters v. New York, 333 U.S. 507, 517–19, 68 S.Ct. 665, 92 L.Ed. 840 (1948)); Smith v. Goguen, 415 U.S. 566, 573, 94 S.Ct. 1242, 39 L.Ed.2d 605 (1974). Vague speech laws violate the First Amendment because they lead speakers to “avoid the risk ․ by restricting their conduct to that which is unquestionably safe. [But f]ree speech may not be so inhibited.” Baggett v. Bullitt, 377 U.S. 360, 372, 84 S.Ct. 1316, 12 L.Ed.2d 377 (1964).

Vagueness and overbreadth interplay in several ways. Most relevant here, vague statutory language exacerbates overbreadth concerns. Overbreadth challenges hinge on the range of a statute's permissible applications in comparison to its impermissible ones. See Moody, 603 U.S. at 723–24, 726, 144 S.Ct. 2383. A vague statute can lead those whose protected speech the statute may not in fact prohibit to silence themselves anyway, effectively increasing the statute's range of impermissible applications. See, e.g., Reno v. ACLU, 521 U.S. 844, 874, 117 S.Ct. 2329, 138 L.Ed.2d 874 (1997) (“Given the vague contours of the coverage of the statute, it unquestionably silences some speakers whose messages would be entitled to constitutional protection.”). So when we assess a statute's sweep in an overbreadth challenge, we “evaluate the ambiguous as well as the unambiguous scope of the enactment. To this extent, the vagueness of a law affects overbreadth analysis.” Vill. of Hoffman Ests. v. Flipside, Hoffman Ests., Inc., 455 U.S. 489, 494 n.6, 102 S.Ct. 1186, 71 L.Ed.2d 362 (1982); see also Am. Booksellers v. Webb, 919 F.2d 1493, 1505–06 (11th Cir. 1990) (same).

A third doctrine that can facially invalidate a speech law, distinct from a failure to give fair notice of prohibited conduct, is the failure to provide “explicit standards for those who apply” speech laws. Grayned, 408 U.S. at 108, 92 S.Ct. 2294. The problem with standardless statutes, which we sometimes evaluate under the rubric of vagueness, is that they “authorize[ ] or even encourage[ ] arbitrary and discriminatory enforcement.” See Hill, 530 U.S. at 732, 120 S.Ct. 2480 (2000) (citing Chicago v. Morales, 527 U.S. 41, 56–57, 119 S.Ct. 1849, 144 L.Ed.2d 67 (1999)). See also Fox, 567 U.S. at 253, 132 S.Ct. 2307 (“[P]recision and guidance are necessary so that those enforcing the law do not act in an arbitrary or discriminatory way.”).

Laws without discernible standards threaten enforcement that is “impermissibly based on content or viewpoint.” Barrett v. Walker Cnty. Sch. Dist., 872 F.3d 1209, 1221 (11th Cir. 2017) (citing City of Lakewood v. Plain Dealer Publ'g Co., 486 U.S. 750, 757–59, 108 S.Ct. 2138, 100 L.Ed.2d 771 (1988)). After all, if the courts can't identify where the statute draws the line between legal and illegal speech, those who enforce the law can choose to impose its purported limitations how they please. And when discriminatory enforcement does occur, it may be hard to prove and to prevent: “the difficulties of proof and the case-by-case nature of ‘as applied’ challenges” can make speech regulators’ actions “effectively unreviewable.” Lakewood, 486 U.S. at 758–59, 108 S.Ct. 2138.

Yet the fear of improper governmental motives in regulating speech—such as the desire to suppress politically disfavored speech or speakers—lies at the heart of the First Amendment. See generally Elena Kagan, Private Speech, Public Purpose: The Role of Governmental Motive in First Amendment Doctrine, 63 U. Chi. L. Rev. 413 (1996). So to protect against discriminatory enforcement, speech laws with insufficient standards to cabin enforcement discretion are liable to facial challenge.

Now that we've set forth the modes of First Amendment facial challenges, we turn to the Act itself. We lay out the distinction between protected and unprotected speech under the First Amendment and show how the Act restricts the protected speech. We then explain why Hamburger Mary's is likely to succeed on its facial challenge to the Act.

ii. The Act reaches First-Amendment-protected speech, not just unprotected obscenity.

The First Amendment, as incorporated against the states, bars the government from “abridging the freedom of speech.” U.S. Const. amend. I; Gitlow v. New York, 268 U.S. 652, 655, 45 S.Ct. 625, 69 L.Ed. 1138 (1925)). Still, some speech is “unprotected” and does not have, “in and of itself, a claim upon the First Amendment.” R.A.V. v. City of St. Paul, 505 U.S. 377, 386, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992). These “well-defined and narrowly limited classes of speech” have historically fallen outside the First Amendment's protection because they are “of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality.” United States v. Stevens, 559 U.S. 460, 468–69, 130 S.Ct. 1577, 176 L.Ed.2d 435 (2010) (quoting R.A.V., 505 U.S. at 383, 112 S.Ct. 2538 (quoting Chaplinsky v. New Hampshire, 315 U.S. 568, 571–72, 62 S.Ct. 766, 86 L.Ed. 1031 (1942))).

While so-called unprotected speech is not “entirely invisible to the Constitution,” it is substantially more amenable to regulation than its protected counterpart. R.A.V., 505 U.S. at 383, 112 S.Ct. 2538. Unprotected speech includes defamation, fighting words, true threats, and, most important for our purposes, obscenity. See Counterman, 600 U.S. at 74–76, 143 S.Ct. 2106.

1. Obscene representations are unprotected only when they meet the requirements of the Miller test and are “specifically defined.”

In the words of one authority, “[o]bscenity remains one of the more colorful, controversial, and confounding areas of First Amendment jurisprudence.” David L. Hudson, Jr., The First Amendment: Freedom of Speech § 4:2 (2012). As we've mentioned, the “confounding” difficulty of defining obscenity led Justice Stewart to give up on the effort: “I shall not today attempt further to define the kinds of material I understand to be embraced within that shorthand description; and perhaps I could never succeed in intelligibly doing so. But I know it when I see it ․”⁴ Jacobellis, 378 U.S. at 197, 84 S.Ct. 1676 (Stewart, J., concurring).

The difficulty of defining obscenity—and the problems vague speech proscriptions create—led to the Supreme Court's approach in Miller v. California, 413 U.S. 15, 93 S.Ct. 2607, 37 L.Ed.2d 419 (1973). The Miller test defines speech as obscene if it satisfies three requirements:

(a) whether the average person, applying contemporary community standards would find that the work, taken as a whole, appeals to the prurient interest; (b) whether the work depicts or describes, in a patently offensive way, sexual conduct specifically defined by the applicable state law; and (c) whether the work, taken as a whole, lacks serious literary, artistic, political, or scientific value.

Miller, 413 U.S. at 24, 93 S.Ct. 2607 (internal quotation marks and citations omitted). See also United States v. Bagnell, 679 F.2d 826, 835 (11th Cir. 1982) (quoting the Miller test). All three of Miller’s prongs must be satisfied for content to qualify as “obscene.” In this way, the Miller test creates both substantive and procedural limits on state-imposed obscenity regulations.

Miller substantively limits obscenity regulations by defining, in broad strokes, what content they may deem obscene. Obscene speech must appeal to the “prurient interest,” depict or describe “sexual conduct” in a “patently offensive way,” and must lack “serious literary, artistic, political, or scientific value.” Miller, 413 U.S. at 24, 93 S.Ct. 2607.

The Court has since clarified the permissible reach of obscenity regulation. Obscene content is always sexual, see Stevens, 559 U.S. at 478–80, 130 S.Ct. 1577; Brown, 564 U.S. at 792, 131 S.Ct. 2729—but not all sexual content is obscene. The line between obscene and non-obscene sexual content can be fuzzy, but the Court has given some guidance.

Material that provokes only “healthy sexual desires,” as opposed to “shameful or morbid” ones, is not obscene. Brockett v. Spokane Arcades, Inc., 472 U.S. 491, 498–99, 105 S.Ct. 2794, 86 L.Ed.2d 394 (1985). And whether sexual content is “patently offensive” cannot hinge on gender-based views of what representations are appropriate. See Manual Enters. v. Day, 370 U.S. 478, 490, 82 S.Ct. 1432, 8 L.Ed.2d 639 (1962) (“[T]hese portrayals of the male nude [in a magazine aimed at gay men] cannot fairly be regarded as more objectionable than many portrayals of the female nude that society tolerates.”).

Nor is the Constitution a prude: material may be crude, vulgar, or offensive without rising to the level of obscene. For example, the Court has distinguished “obscene” materials from those that are just “indecent.” FCC v. Pacifica Found., 438 U.S. 726, 740, 98 S.Ct. 3026, 57 L.Ed.2d 1073 (1978). “ ‘[I]ndecent’ merely refers to nonconformance with accepted standards of morality.” Id. So indecency includes a broader range of materials, including some sexual expression that falls short of obscenity. Reno, 521 U.S. at 874, 117 S.Ct. 2329 (citing Sable Commc'ns of Cal., Inc. v. FCC, 492 U.S. 115, 122, 109 S.Ct. 2829, 106 L.Ed.2d 93 (1989)). As for obscenity, only depictions of “ ‘hard core’ sexual conduct” can satisfy that term. Miller, 413 U.S. at 27, 93 S.Ct. 2607.

This last point means that the prongs of the Miller test limit which types of “hard core” sexual materials are obscene (and therefore unprotected), and they don't permit restrictions of other speech that might be considered simply “patently offensive” or lacking “value.” In other words, Miller’s “serious value” exceptions clause is not a formula whose invocation strips First Amendment protection from nonsexual speech. In the Court's words, “Miller did not determine that serious value could be used as a general precondition to protecting other types of speech in the first place.” Stevens, 559 U.S. at 479, 130 S.Ct. 1577. Rather, as we've noted, content must satisfy all three of Miller’s prongs to amount to obscenity.

Besides its substantive limits on obscenity regulation, Miller creates a procedural requirement. Speech that meets Miller’s substantive requirements for obscenity remains protected unless it is “specifically defined by the applicable state law [regulating obscenity], as written or authoritatively construed.” Miller, 413 U.S. at 24, 93 S.Ct. 2607. Miller offered examples of what would qualify as sufficiently specific definitions: “representations or descriptions of ultimate sex acts, normal or perverted, actual or simulated” or “representation or descriptions of masturbation, excretory functions, and lewd exhibition of the genitals.” Id. at 25, 93 S.Ct. 2607.

The reasons for this specificity requirement are the same as those motivating much of contemporary First Amendment doctrine: providing “fair notice” to speakers as to the possibility of prosecution, id. at 27, 93 S.Ct. 2607, preventing arbitrary enforcement, and not chilling protected speech. This “specifically defined” requirement is not throwaway language but instead a “critical” part of the Miller test. See Reno, 521 U.S. at 873, 117 S.Ct. 2329 (“The second prong of the Miller test ․ contains a critical requirement ․ that the proscribed material be ‘specifically defined by the applicable state law.’ ” (quoting Miller, 413 U.S. at 24, 93 S.Ct. 2607)).

And though it resembles the more general prohibition of vague legislation, the specificity requirement stands on top of that prohibition. True, the specific-definition requirement furthers the same values as void-for-vagueness doctrine. But because of the drastic legal and political implications of deeming content “obscene” and thus beyond First Amendment protection, the need for specific and definite legislation in this context is greater than it is generally or even elsewhere in the First Amendment realm.

2. Obscenity restrictions for minors still are subject to the Miller test.

One other thread of obscenity doctrine we must consider here is its application to minors. The history of speech standards that vary by age dates back at least to Ginsberg v. New York, 390 U.S. 629, 88 S.Ct. 1274, 20 L.Ed.2d 195 (1968).

In Ginsberg, the Court upheld a New York statute that barred the sale to minors under seventeen of any materials that depicted “nudity” and were “harmful to minors”—that is, “appealing to the prurient, shameful, or morbid interest of minors,” “patently offensive” to adult standards regarding what is appropriate “for minors,” and completely without redeeming value “for minors.” Ginsberg, 390 U.S. at 631–33, 88 S.Ct. 1274.⁵ The Court held that minors’ access to speech bordering on obscene need not be as extensive as adults’: “[m]aterial which is protected for distribution to adults is not necessarily constitutionally protected from restriction upon its dissemination to children.” Id. at 636, 88 S.Ct. 1274 (quoting Bookcase, Inc. v. Broderick, 18 N.Y.2d 71, 271 N.Y.S.2d 947, 218 N.E.2d 668, 671 (1966)).

“In other words, the concept of obscenity or of unprotected matter may vary according to the group to whom the questionable material is directed or from whom it is quarantined.” Id. (quoting Bookcase, Inc., 271 N.Y.S.2d 947, 218 N.E.2d at 671). Noting that the statute still allowed “parents who so desire” to provide the prohibited materials to their children, the Court reasoned that the statute was rationally related to the state's interest in helping parents control their children's exposure to sexual material. Id. The Court also noted the statute's role in furthering the state's “independent interest in the well-being of its youth.” Id. at 640, 88 S.Ct. 1274.

Today, many statutes use what we will call a “Miller-for-minors” test, also called a “harmful for minors” standard or a “variable obscenity” standard. The Miller-for-minors test takes the Miller prongs but adjusts the second and third standards (and sometimes the first) “for minors.” See, e.g., Fla. Stat. § 847.001(7); Ala. Code § 13A-12-200.1(11); Ga. Code Ann. § 20-2-324.6(a). So the “patently offensive” requirement becomes “patently offensive for minors,” and the “serious value” exception becomes “serious value for minors.”

But the Miller-for-minors test does not do away with Miller’s procedural requirement that obscene content must be specifically identified to remove it from First-Amendment protection. After all, “[i]t is ․ essential that legislation aimed at protecting children from allegedly harmful expression—no less than legislation enacted with respect to adults—be clearly drawn and that the standards adopted be reasonably precise.” Interstate Circuit, Inc. v. City of Dallas, 390 U.S. 676, 689, 88 S.Ct. 1298, 20 L.Ed.2d 225 (1968) (quoting People v. Kahan, 15 N.Y.2d 311, 258 N.Y.S.2d 391, 206 N.E.2d 333, 335 (1965) (Fuld, J., concurring)). And content that is obscene as to minors may differ from its adult equivalent only in quality, not in kind: as with adults, only “sexual material” can be obscene for minors. Brown, 564 U.S. at 793, 131 S.Ct. 2729.

Finally, a regulation of speech “harmful to minors” may not overly “suppress[ ] ․ speech that adults have a constitutional right to receive and to address to one another.” Reno, 521 U.S. at 874, 117 S.Ct. 2329. Even when a statute only indirectly limits protected adult speech—such as by making it more difficult for adults to access, Webb, 919 F.2d at 1506–08, or by forcing speakers to “err on the side of caution” because of statutory ambiguities, id. at 1505–06—the law still must do so narrowly and in proportion to the government's interest in shielding children from speech that is unprotected as to minors. Id. at 1501.

3. The Act restricts speech that is protected as to minors.

With these principles in mind, our first task is to assess the Act's burden on protected speech, if any.

Secretary Griffin argues that the Act dutifully observes Miller’s requirements so it reaches only unprotected speech. The Act, she argues, “specifically defines” the activities that constitute an adult live performance: anything “depict[ing] or simulat[ing] nudity, sexual conduct, sexual excitement, or specific sexual activities as those terms are defined in [Section] 847.001, lewd conduct, or the lewd exposure of prosthetic or imitation genitals or breasts.” See Fla. Stat. § 827.11(1)(a). She continues, asserting the Act then bars children's admittance to performances that feature such representations and meet all three prongs of the Miller test: appeal to the prurient interest, patent offensiveness, and the lack of serious value. Id. § 827.11(1)(a)1–3. And, Griffin says, the statute adjusts the standards for the second and third prongs of the Miller test to be “for the age of the child present.” Id. That is, prohibited performances must lack serious “value for the age of the child present” and be “patently offensive ․ with respect to what is suitable material or conduct for the age of the child present.” Id.

But as it turns out, the Act's specific definitions aren't all that specific. Undoubtedly, the statute meticulously explains the first four terms, whose definitions cross-reference a preexisting statute. Anyone curious about the “specific sexual activities” whose depiction may be obscene can peruse Florida Statutes, Section 847.001(23). Here, it's enough to say that Florida law amply details which body parts, doing what, constitute “nudity,” “sexual conduct,” and so on. Fla. Stat. § 847.001(10), (11), (19), (20), (23).

We can't say the same for the meaning of “lewd conduct” under the statute, though. Indeed, the Secretary's brief is revealingly honest when it omits “lewd conduct” when listing the terms the Act “specifically define[s].”

Still, Griffin argues that “lewd conduct” is sufficiently definite to prevent the statute from being void for vagueness. She notes that several state and federal statutes employ the term “lewd.”

And she points to the Florida Supreme Court's definition of “lewd” as meaning the same thing as “lascivious ․ that is, an unlawful indulgence in lust, eager for sexual indulgence.” Chesebrough v. State, 255 So. 2d 675, 677 (Fla. 1971). Chesebrough upheld against a vagueness challenge a statute barring “lewd or lascivious acts” in the presence of a child under fourteen years old. Id. at 678.

Griffin also highlights a Florida model jury instruction, defining “lewd” and “lascivious” as “a wicked, lustful, unchaste, licentious, or sensual intent on the part of the person doing the act.” Fla. Sup. Ct. Comm. on Standard Jury Instructions in Criminal Cases, Florida Standard Jury Instructions in Criminal Cases 322 (2024), https://perma.cc/NJX6-FRFW.

But the Secretary's arguments ignore two points. First, by focusing on whether the term “lewd conduct” is unconstitutionally vague as a general matter, Griffin skirts the more pressing concern: whether a restriction of depictions of “lewd conduct” meets Miller’s more stringent test for defining obscene speech. Unlike the lewdness statute in Chesebrough, which barred “lewd or lascivious act[s] in the presence of” children under fourteen, Chesebrough, 255 So. 2d at 676 (emphasis added) (quoting Fla. Stat. § 800.04), the Act here—which focuses on “depict[ions of] ․ lewd conduct”—squarely targets speech. Fla. Stat. § 827.11(1)(a) (emphasis added).

Miller is clear that when a state seeks to punish speech as obscene, it must “specifically define[ ]” the forbidden depictions. Miller, 413 U.S. at 24, 93 S.Ct. 2607. One of the Miller Court's example definitions—“patently offensive representations or descriptions of ․ lewd exhibition of the genitals”—proves the point. Id. at 25, 93 S.Ct. 2607. “Lewd exhibition of the genitals” obviously belongs to the larger category of “lewd conduct.” So if “lewd conduct” were sufficiently specific to meet Miller’s test, the Court would have had no reason to provide a more precise description. Instead, Miller shows that states can't define obscenity by taking a broad descriptor like “lewd” and applying it to the entire universe of “conduct.” Doing so would eviscerate Miller’s “specific definition” requirement and amount to little more than an “I know it when I see it” test for obscenity, which Miller rejects.

Second, the Secretary doesn't read the “lewd conduct” provision in its statutory context. Rather, she relies on judicial interpretations of “lewdness” plucked from statutes that don't specifically describe any prohibited depictions. But as we explain, doing so is misleading here.

Griffin's chief case in point is Miller’s companion case, United States v. 12 200-Ft. Reels of Super 8mm. Film, 413 U.S. 123, 130 n.7, 93 S.Ct. 2665, 37 L.Ed.2d 500 (1973). The 12 200-Ft. Reels Court dealt with a challenge to a pre-Miller statute that barred the importation of “any obscene book ․ or other representation, figure, or image.” Id. at 124, 93 S.Ct. 2665 (quoting 19 U.S.C. § 1305(a)). After announcing Miller’s new test for obscenity, the Court remanded 12 200-Ft. Reels for a lower court to evaluate whether the materials at issue were obscene under the newly announced Miller test. Id. at 130, 93 S.Ct. 2665.

In a footnote, the Court anticipated a potential vagueness challenge to the statute restricting “obscene” material and other similar laws. Id. at 130 n.7, 93 S.Ct. 2665. It said, “[i]f and when such a serious doubt is raised as to the vagueness of the words ‘obscene,’ ‘lewd,’ ‘lascivious,’ ‘filthy,’ ‘indecent,’ or ‘immoral’ as used to describe regulated material,” the Court was “prepared to construe such terms as limiting regulated material to patently offensive representations or descriptions of that specific ‘hard core’ sexual conduct given as examples in Miller.” Id. (cleaned up).

The next year the Court heard exactly that challenge and resolved it as promised. In Hamling v. United States, the Court interpreted a different pre-Miller statute that regulated “obscene, lewd, lascivious, indecent, filthy or vile” content as covering only the “permissibly proscribed depictions described in Miller.” 418 U.S. 87, 110–16, 94 S.Ct. 2887, 41 L.Ed.2d 590 (1974) (quoting 18 U.S.C. § 1461). So Miller required specificity when regulating obscenity and gave examples of how to satisfy that requirement. And 12 200-Ft. Reels and Hamling showed that speech statutes with otherwise vague “lewdness” prohibitions could be saved by reading the term to encompass only Miller’s specific prohibitions. Put simply, the Court saved “lewdness” statutes by reading the term to be coextensive with Miller’s specific examples of obscenity.

But no such interpretation is available for “lewd conduct” here. That's because the Act's extensively detailed prohibitions—of nudity, sexual conduct, and so on—essentially exhaust the types of “hard core” depictions that Miller described as potentially obscene. Miller, 413 U.S. at 25, 93 S.Ct. 2607. That means the Act's bar on “depict[ions of] ․ lewd conduct” must mean something different than the Miller-prohibited depictions, or it would be “mere surplus-age.” Aspen Am. Ins. Co. v. Landstar Ranger, Inc., 65 F.4th 1261, 1271 (11th Cir. 2023) (quoting United States v. Canals-Jimenez, 943 F.2d 1284, 1287 (11th Cir. 1991)). And so we assume it does mean something different.

But neither Miller and its progeny nor the Florida materials Griffin presents even suggest a way to interpret “lewd conduct” that is distinct from the other prohibited depictions but accords with Miller’s specificity requirement. And while we “tolerate some redundancy” in statutory terms, see United States v. Conage, 50 F.4th 81, 88 (11th Cir. 2022) (mem.), we will not run roughshod over Miller’s requirement that laws condemning speech as unprotected by the First Amendment do so with precision.

The dissent laments our use of the surplusage canon to understand the scope of the “lewd conduct” provision. Instead, the dissent urges us to certify a question about the Act's interpretation to the Florida Supreme Court.

The dissent's suggestion must come as news to both parties. Indeed, the State asked us to accept its proffered interpretation of the Act (which, in suggesting we certify the Act's interpretation, the dissent seems to implicitly recognize suffers from constitutional problems). And neither party requested or briefed the idea of certifying questions to the Florida Supreme Court. Nor did we ever ask about it during oral argument. Yet “normally ․ the crucible of adversarial testing is crucial to sound judicial decisionmaking. We rely on it to yield insights (or reveal pitfalls) we cannot muster guided only by our own lights.” Sessions v. Dimaya, 584 U.S. 148, 190, 138 S.Ct. 1204, 200 L.Ed.2d 549 (2018) (Gorsuch, J., concurring) (cleaned up). And in any case, as we explain, the surplusage canon is helpful here; certification would not be. That's because nothing in the interpretive toolkit would cut “lewd conduct” down to constitutional size.

First, the ground rules for certification. The Florida Supreme Court may rule on outcome-determinative state-law questions we certify to it. Fla. R. App. P. 9.150. But “the decision to certify ․ ‘rests in the sound discretion of the federal court.’ ” Minn. Voters All. v. Mansky, 585 U.S. 1, 22 n.7, 138 S.Ct. 1876, 201 L.Ed.2d 201 (2018) (quoting Expressions Hair Design v. Schneiderman, 581 U.S. 37, 58, 137 S.Ct. 1144, 197 L.Ed.2d 442 (2017) (Sotomayor, J., concurring in the judgment)). In Mansky, the Supreme Court deemed certification unnecessary in a First Amendment challenge to a state law where (1) the state's “request for certification comes very late” and (2) the state “has not offered sufficient reason to believe that certification would obviate the need to address the constitutional question.” Id. at 22 n.7, 138 S.Ct. 1876.

The same is true here. As we've noted, the Secretary has not even requested certification. That's presumably because the Secretary is confident that “lewd” has the same broad meaning it did when the Florida Supreme Court interpreted the term in Chesebrough, 255 So. 2d at 677. And as in Mansky, “[o]ur analysis today reflects the State's proffered interpretation; nothing in that analysis would change if the State's interpretation were also adopted by the [Florida] Supreme Court.” Mansky, 585 U.S. at 22 n.7, 138 S.Ct. 1876. Neither Florida nor the dissent proposes “a viable alternative construction that the [Florida] Supreme Court might adopt instead.” Id. And while we of course agree with the dissent that certification is permissible even when parties do not seek it, we know too that certification has value only when a statute is “readily susceptible” to a “proffered narrowing construction.” Virginia v. Am. Booksellers Ass'n, Inc., 484 U.S. 383, 397, 108 S.Ct. 636, 98 L.Ed.2d 782 (1988). That no such construction has been “proffered” here is a good clue to the fact that the statute is not “readily susceptible” to such a construction. Federalism principles rightly counsel us to steer clear of friction with states. But they do not instruct us to detour into state court just to postpone a constitutional collision that, as we show, is unavoidable.⁶

The dissent suggests that “lewd conduct” is “merely a catchall phrase, rather than a significant expansion of the statute's scope.” Dissenting Op. at ––––. This seems to us precisely the problem: “lewd conduct” is indeed a “catchall phrase” that “catches” much non-obscene speech. (Of course, if “lewd conduct” really were pure surplusage—totally duplicative of the Act's other terms—an injunction of that part of the Act would have no effect: no harm, no foul.)

Still, we understand the dissent's point to be that “lewd conduct” has some meaning that is distinct from the Act's other prohibitions, but still is sufficiently limited to content that may be proscribed as obscene. Or at least, that the Florida Supreme Court could decide as much.

The dissent argues that we miss this point because rely on the surplusage canon to the exclusion of other interpretive tools. The noscitus a sociis canon instructs us that “a word is known by the company it keeps.” United States v. Dawson, 64 F.4th 1227, 1237 (11th Cir. 2023) (quoting Yates v. United States, 574 U.S. 528, 537, 135 S.Ct. 1074, 191 L.Ed.2d 64 (2015)). And the related canon of ejusdem generis teaches that a final, general term at the end of a list of specific items should be interpreted “in light of any common attributes shared by the specific items.” Bissonnette v. LePage Bakeries Park St., LLC, 601 U.S. 246, 252, 144 S.Ct. 905, 218 L.Ed.2d 204 (2024) (cleaned up). The dissent would have us therefore take “lewd conduct” to encompass anything—or at least something—that is distinct from the Act's more specific prohibitions, yet still similar enough that it can constitutionally be defined as obscene.

But these canons fire blanks here. First, we note that the ejusdem canon is not strictly on target. The list of prohibited depictions ends not with “lewd conduct” but with something extremely specific: “the lewd exposure of prosthetic or imitation genitals or breasts.” Fla. Stat. § 827.11(1)(a). This furthers our suspicion that “lewd conduct” is not a narrow, conclusory “catchall” but a broad term hidden between two commas.

And even if a court were to apply the ejusdem canon, we don't see how it would help. The Act's more specific prohibitions do little to shed light on the meaning of “lewd conduct.” To apply the ejusdem canon, a court would have to read the statute as barring only “lewd conduct” that is somehow similar to the “nudity, sexual conduct, sexual excitement, or specific sexual activities” prohibited by the statute—but still distinct from them. Call us unimaginative, but we are unsure what occupies that territory.

It is particularly perplexing to propose taking the term “sexual conduct” as a guide to interpreting “lewd conduct.” If there is any bedrock in obscenity doctrine, it is that obscenity must be “sexual” in nature. Brown, 564 U.S. at 793, 131 S.Ct. 2729 (quoting Miller, 413 U.S. at 24, 93 S.Ct. 2607). So to be proscribed under Miller, “lewd conduct” would need to be “sexual.” But to have any content in the context of the Act, “lewd conduct” would need to be distinct from “sexual conduct.” So the ejusdem canon would instruct a court to identify sexual, “lewd conduct” that is not “sexual conduct.”

We're stumped. And so, it seems, are the Secretary and the dissent, neither of which provide a single example of “lewd conduct” that might be constitutionally deemed obscene (at least for minors) but is not already covered by the Act's other terms. Even the Secretary waffles on this point, arguing only that “it is unclear whether” the Act's other terms “extend[ ] to the full reach of what is regulable as obscenity.” Yet Florida's determination that performances like Jimbo's are “lewd” suggest the state views that term broadly indeed.

Other considerations increase our skepticism. The dissent suggests “lewd conduct” is meant to cover “known unknowns”: depictions the legislature did not think to include specifically but would have if it had known of such exotic material while drafting, Dissenting Op. at –––– (quoting Republic of Iraq v. Beaty, 556 U.S. 848, 860, 129 S.Ct. 2183, 173 L.Ed.2d 1193 (2009))—in other words, the legislature and the dissent would know it when it sees it. Aside from that problem, this theory strains against the fact that Florida's other statutes protecting minors from harmful content do not reference “lewd conduct.” See, e.g., Fla. Stat. §§ 847.012–.013. The legislature presumably had something in mind when it added the new “lewd conduct” prohibition. Yet no one has shared what that might be.

Rather than speculate wholesale or attempt to turn the “lewd conduct” provision into something it's not, we think it best to adopt “the State's proffered interpretation” based on the broad construction in Chesebrough and analyze the case accordingly. Mansky, 585 U.S. at 22 n.7, 138 S.Ct. 1876. And nothing in our “analysis would change if the State's interpretation were also adopted by the [Florida] Supreme Court.” Id.

Finally, we note that even if “lewd conduct” could be cut down to constitutional size, doing so would not cure all its problems. The term still falls far short of Miller’s requirement that prohibited depictions be “specifically defined by the regulating state law, as written or construed.” Miller, 413 U.S. at 27, 93 S.Ct. 2607. The dissent seems to suggest that the Supreme Court's decision in Hamling offers a way around the need for specific definition. But Hamling cannot be so read to undermine that requirement, which the Supreme Court has since reiterated is “critical.” Reno, 521 U.S. at 873, 117 S.Ct. 2329.

Hamling upheld an obscenity conviction under a federal statute that barred sending “obscene, lewd, lascivious, indecent, filthy, or vile” material through the mail. Hamling, 418 U.S. at 98, n.8, 94 S.Ct. 2887 (quoting 18 U.S.C. § 1461). The Court interpreted the statute's scope as “limited to material such as that described in Miller.” Id. at 115, 94 S.Ct. 2887. The material at issue was “a form of hard-core pornography well within the types of permissibly proscribed depictions described in Miller.” Id. (We will spare readers the details and just say that the material involved “a horse.” Id. at 93, 94 S.Ct. 2887.) Essentially, the content was so clearly one of the “obnoxiously debasing portrayals of sex” described in Miller that the defendants could not plausibly shelter under a vagueness claim, regardless of the specific-definition requirement. Id. at 112, 94 S.Ct. 2887.

But Hamling never erased that requirement. Indeed, while noting that Miller’s sample statutory language did not necessarily exhaust the constitutional limits of obscenity, the Court stated that “Congress could always define other specific ‘hard core’ conduct.” Id. at 113, 94 S.Ct. 2887 (emphasis added) (quoting Miller, 413 U.S. at 25 n.7, 93 S.Ct. 2607). And while Miller suggested what Hamling confirmed—that courts may supply specific definitions of nebulous statutory language—the thrust of those cases is that doing so meets the goals of the specific-definition requirement only when courts effectively adopt the definitions in Miller. Those definitions have the benefit of both being specific and describing material that is indubitably obscene. Hamling permits courts to treat obscenity statutes as incorporating the specific definitions in Miller. It is not a free pass for legislatures to tack on broad terms to the Miller prohibitions and allow courts—or state officials enforcing the laws where these terms appear—to fill in the blanks.

In short, we understand the Act's prohibition on depictions of lewd conduct to reach speech that is constitutionally protected, even as to minors. That's so because Miller’s test for unprotected obscenity contains both substantive and definitional requirements. Even if some of the “lewd” speech falls within the substantive scope of speech that can, under Ginsberg and Miller, be deemed obscene as to minors, it remains protected until the state clearly defines it not to be so.⁷ And that the state has not yet done.

That brings us to whether the Act is overbroad—that is, whether it “punishes a ‘substantial’ amount of protected free speech, ‘judged in relation to the statute's plainly legitimate sweep.’ ” Virginia v. Hicks, 539 U.S. 113, 118–19, 123 S.Ct. 2191, 156 L.Ed.2d 148 (2003) (quoting Broadrick v. Oklahoma, 413 U.S. 601, 615, 93 S.Ct. 2908, 37 L.Ed.2d 830 (1973)).

iii. The Act is unconstitutionally overbroad.

That the “lewd conduct” provision restricts protected speech does not alone determine the Act's constitutionality. But it informs our review of Hamburger Mary's facial challenge.

As we've mentioned, on a facial challenge, we examine whether a law “prohibits a substantial amount of protected speech relative to its plainly legitimate sweep.” Moody, 603 U.S. at 744, 144 S.Ct. 2383 (quoting Hansen, 599 U.S. at 770, 143 S.Ct. 1932).⁸ A statute's impact on speech includes its “direct and indirect burdens.” Webb, 919 F.2d at 1499–500. While overbreadth and vagueness are distinct, we've explained that vagueness can contribute to overbreadth because our overbreadth analysis accounts for a law's “ambiguous as well as ․ unambiguous scope.” Id. at 1505–06.

As we show below, the Act is overbroad. At least two elements of the Act make it so.⁹ First is the “depict[ions of] ․ lewd conduct” provision. The problem is that the vague restriction on “depict[ions of] ․ lewd conduct” threatens protected speech. Second is the age-variable obscenity standard, which purports to narrow the Act's scope but in fact just obscures its meaning.

1. The Act's “lewd conduct” provision is overbroad.

The “depict[ions of] ․ lewd conduct” provision restricts a substantial amount of speech, including both speech that falls within its scope and speech that does not but it threatens to chill by its vagueness. “Lewd” speech, particularly as the FDBPR construes the term, includes much that is protected even for minors.

For example, the Supreme Court in Bethel School District No. 403 v. Fraser weighed a public school's punishment of a student whose “lewd speech” at a student-government campaign forum relied on “an elaborate, graphic, and explicit sexual metaphor.” 478 U.S. 675, 677, 678, 106 S.Ct. 3159, 92 L.Ed.2d 549 (1986). The Court approved of his punishment in the school context only and noted later that the “lewd speech” “would have been protected” outside that setting. Morse v. Frederick, 551 U.S. 393, 405, 127 S.Ct. 2618, 168 L.Ed.2d 290 (2007) (discussing Fraser, 478 U.S. at 682–83, 106 S.Ct. 3159).

Jimbo's baloney birth, too, might be “lewd” to the FDBPR, but it cannot be deemed “obscene” even for minors. “There is no exhibition whatever of the actor[’s] genitals, lewd or otherwise.” Jenkins v. Georgia, 418 U.S. 153, 161, 94 S.Ct. 2750, 41 L.Ed.2d 642 (1974). The act is “simply not the ‘public portrayal of hard core sexual conduct for its own sake’ ” that falls within Miller’s scope. Id. (quoting Miller, 413 U.S. at 35, 93 S.Ct. 2607).

One of the Act's sponsor's stated intent to target “Drag Queen Story Time” also helps show the potential breadth of a term like “lewd conduct.” See Ashleigh Walters, Florida Lawmaker Wants to Stop Children from Attending Drag Events, WPTV (Mar. 24, 2023, 8:26 PM), https://www.wptv.com/news/lgbtq/florida-law-maker-wants-to-stop-children-from-attending-drag-events [https://perma.cc/3YNR-NPTJ]. Of course, one legislator's interpretation of the Act does not an authoritative construction make. But it does betray how much protected speech may fall within the Act's “ambiguous as well as [its] unambiguous scope.” Webb, 919 F.2d at 1506.

The vagueness of the “lewd conduct” term only exacerbates its breadth. As we've explained, the term “lewd conduct” is not a model of clarity. Consider the Florida Supreme Court's definition of “lewd” in Chesebrough: “the unlawful indulgence of lust, signifying that form of immorality which has a relation to sexual impurity.” Chesebrough, 255 So. 2d at 677. This definition offers little assurance to would-be speakers and to this Court that content “about birth control practices, homosexuality ․ or the consequences of prison rape”—all content the Supreme Court has held is not necessarily obscene—for example, would not fall on the wrong side of the Act's proscriptions. Reno, 521 U.S. at 871, 117 S.Ct. 2329.

Plus, Chesebrough’s definition of “lewd” shows how “lewd conduct”—unlike the Act's other, more specific prohibitions—is decidedly in the eye of the beholder. Do same-sex relations “signify[ ] that form of immorality which has a relation to sexual impurity”? Chesebrough, 255 So. 2d at 677. What about out-of-wedlock sex? Or most pertinent here, dressing in clothes typically associated with the non-birth-assigned gender? In the context of an otherwise quite specific statute, the “lewd conduct” provision is a prosecutorial skeleton key allowing “policemen, judges, and juries” to unlock penalties and punish speech on “an ad hoc and subjective basis.” Grayned, 408 U.S. at 109, 92 S.Ct. 2294.

Finally, the incantation of the Miller factors does not satisfactorily limit the breadth of the “lewd conduct” provision.¹⁰ That's because the Act doesn't actually adopt the Miller factors, which include the “critical requirement” to specifically define prohibited speech. Reno, 521 U.S. at 873, 117 S.Ct. 2329.

Contrary to the Miller specific-definition requirement, “lewd conduct” essentially does nothing to define the prohibition beyond the Miller factors themselves. “Conduct” includes nearly anything. The Secretary's proposed definitions of lewd—“eager for sexual indulgence” or “a wicked, lustful, unchaste, licentious, or sensual intent”[Blue Br. at 40]—just duplicate or perhaps expand Miller’s “prurient interest” requirement. See Prurient Interest, Black's Law Dictionary (12th ed. 2014) (“A morbid, unhealthy fixation with sex, nudity, or obscene or pornographic materials.”). The Act takes the Miller test as its starting point but amorphously expands its scope rather than limiting it as Miller requires. Accepting the Secretary's argument would require us to ignore Miller’s specific-definition proviso, a “critical” part of the rule. Reno, 521 U.S. at 873, 117 S.Ct. 2329.

A facial overbreadth challenge requires us to assess a law's “legitimate sweep.” Broadrick, 413 U.S. at 615, 93 S.Ct. 2908. We freely acknowledge the difficulty in doing so here. We are unsure which depictions of “lewd conduct,” if any, would satisfy the Miller-for-minors test but not already be included within the Act's more specific prohibitions. We presume that something falls into that category, otherwise the term would be wholly duplicative. But if Miller means anything, it is that an obscenity regulation's sweep is only “legitimate” if it is specific. The Act's “lewd conduct” provision is the opposite. The result is that venues like Hamburger Mary's are prone to restrict minors from consuming speech that they are within their constitutional rights to access.

Not only that, but the Act's sweep risks indirectly squelching adults’ access to nonobscene speech. Of course, the likely outcome of the Act—that venues require adults to show identification for admittance to certain live performances—is not a tremendous burden. But still, such a requirement would “completely bar” those adults who do not have proof of age. See Reno, 521 U.S. at 856, 117 S.Ct. 2329.

Enforcing an ID policy would also threaten to silence some performances entirely. Whether online or not, age-verification requirements “impose costs” on speakers or the venues that host them: the costs of checking the identification of those who have it and keeping out those who don't. Id. This cost would, on the margin, make some performances financially nonviable. Id. And of course, the law's vagueness, FBDPR's apparently quite broad view of the term “lewd,” and the Act's harsh penalties expand the range of venues that are likely to implement admissions restrictions and the stringency with which they are likely to enforce them.

So the “lewd conduct” provision, given its vagueness and its context, directly threatens protected speech. Its “legitimate sweep” is marginal at best. And it indirectly burdens, or even jeopardizes outright, a range of protected speech. Based on the record at this preliminary stage, we think that Hamburger Mary's challenge to the “lewd conduct” regulation is likely to succeed on the merits.

2. The Act's “age-variable” obscenity standard is overbroad.

Were the “lewd conduct” provision the Act's only constitutional infirmity, we might be able to sever the clause and preserve the rest of the statute. But the Act contains another legally innovative—and constitutionally problematic—feature: an “age-variable” obscenity standard that purports to narrow the Act's scope but in fact just expands it.

Dating back to Ginsberg, “harmful to minors” statutes have typically defined which content is harmful with reference to minors as a whole. See, e.g., Ginsberg, 390 U.S. at 646, 88 S.Ct. 1274. For example, Florida law generally defines “harmful to minors” as that which “is patently offensive ․ for minors” and “is without serious value ․ for minors.” Fla. Stat. § 847.001(7).¹¹

Of course, “minors” are not an undifferentiated group. What is “harmful” for a child just learning to read may be very different than what is obscene for an adolescent on the edge of adulthood.

This poses two related challenges for “harmful to minors” statutes. The first is interpretive: when a statute regulates material that lacks serious value and is patently offensive “for minors,” which minors are the reference group? The second relates to overbreadth: if the obscenity standard for minors is based on the youngest group of children or the average child across all ages, it will cover much more material, and older minors could lose access to harmless material on account of younger children's sensitivities.

This problem and related overbreadth issues led the Supreme Court in Virginia v. American Booksellers Association, Inc., 484 U.S. 383, 398, 108 S.Ct. 636, 98 L.Ed.2d 782 (1988), to ask Virginia's high court to clarify the state's statute barring the display to juveniles of material “harmful to minors.” The Virginia Supreme Court interpreted the state's statute narrowly to mitigate the overbreadth threat. Commonwealth v. Am. Booksellers Ass'n, Inc., 236 Va. 168, 372 S.E.2d 618 (1988). Speech had “serious value” for minors, the court said, if it “has serious value for a legitimate minority of juveniles, ․ consist[ing] of older, normal (not deviant) adolescents.” Id. at 623. See also Am. Booksellers Ass'n Inc. v. Virginia, 882 F.2d 125, 127 (4th Cir. 1989) (upholding Virginia's statute as interpreted by the Virginia Supreme Court).

We soon adopted a similar (though arguably more speech-protective) approach in American Booksellers v. Webb. There, we analyzed a Georgia law restricting the display of materials “harmful to minors.” Webb, 919 F.2d at 1500. We held that “[a]s applied to a Ginsberg-type adaptation of the adult obscenity test ․ if any reasonable minor, including a seventeen-year-old, would find serious value, the material is not ‘harmful to minors.’ ” Id. at 1504–05. See also Davis-Kidd Booksellers, Inc. v. McWherter, 866 S.W.2d 520, 527 (Tenn. 1993) (adopting Webb’s reasoning with respect to a Tennessee statute); ACLU v. Ashcroft, 322 F.3d 240, 254 n.16 (3d Cir. 2003) (rejecting Webb’s narrow interpretation of “harmful to minors” and thus enjoining enforcement of federal “Miller for minors” statute as not narrowly tailored), aff'd on other grounds, 542 U.S. 656, 661, 124 S.Ct. 2783, 159 L.Ed.2d 690 (2004).

Our interpretation of typical “harmful to minors” statutes protects older children's rights. But it means that younger children may encounter material suitable for kids a few years older.

Responding to this potential underinclusion, the Act takes a different (and to our knowledge, novel) approach to protecting minors from harmful material. The Act adjusts the Miller standards for what is “patently offensive” and what has “serious value” to be “for the age of the child present.” Fla. Stat. § 827.11(1)(a)(2)–(3). On paper, the Act is the Goldilocks of speech regulation, ensuring each child can access only that speech that is “just right” for their age. Seventeen-year-olds have access to speech that would be obscene as to sixteen-year-olds but not eighteen-year-olds, sixteen-year-olds can see content that would be obscene as to fifteen-year-olds but not seventeen-year-olds, and so on.¹²

But the Act's strategy to avoid overbreadth problems introduces other ones. The age-by-age maturity test is impossibly vague. At oral argument, we asked the Secretary's counsel how to determine what might be acceptable for a twelve-year-old but not an eight-year-old. Even when pressed, he could provide no guidance and instead simply acknowledged that there were “edge cases.” Oral Argument at 10:28–45. Of course, most laws have edge cases, and standard “harmful to minor” statutes are no exception. But as interpreted in Webb, these have only one “edge”: the line between minority and adulthood.

In contrast, the Act has, at a minimum, eighteen (and perhaps as many as 6,575). It asks speech purveyors to make judgements about what is appropriate for children year-by-year (or maybe month-by-month, week-by-week, or day-by-day—the Act is not clear). This results in all the usual harms of statutory vagueness. The Act will chill more speech relative to the typical “harmful to minors” statute that Webb contemplated: rather than risking a chill for only speech at the border of adult obscenity, the Act threatens speech that might clearly be appropriate for seventeen-year-olds (so not “harmful to minors” under a statute like the one in Webb) but questionable for children of younger ages. Worse still, the Act's vague standards create ample room for discriminatory enforcement.

Not only that, but which speech is appropriate for children of different ages under the Act is left extraordinarily vague. The Act provides no guidance. That contrasts mightily with other instances when Florida has identified content it deems appropriate for children based on their age: grade-level educational standards. Take Florida's grade-by-grade standards for K–12 education. They're extraordinarily detailed, running to 229 pages—and that's just for math class. Fla. State Bd. of Educ., Florida's B.E.S.T. Standards: Mathematics (2020), https://cpalmsmediaprod.blob.core.windows.net/uploads/docs/standards/best/ma/mathbeststandardsfinal.pdf [https://perma.cc/BSZ5-4727]. But here, the Act does little to teach speakers, performance venues, parents, or anyone else who might “admit” a child to a performance what is acceptable for children of any given age.

Consider this example: Miami is home to an historic, thirty-five-foot-tall billboard for Coppertone sunscreen, which features the brand's historic logo—a girl, perhaps age seven, or so, with a dog pulling at her swimsuit, revealing her pale posterior and its contrast with her tanned skin. See Tim Swift, After Irma, Miami Coppertone Girl Gets a Facelift, Local10.com (Dec. 12, 2017, 6:16 PM), https://www.local10.com/news/2017/12/12/after-irma-miami-coppertone-girl-gets-a-facelift [https://perma.cc/LB8H-MWD8]. Clearly, some have objected to this cheeky logo: Coppertone once removed the “Coppertone girl” from the brand's logo, then reinstated her with partial, then full coverage of her backside. Coppertone Logo, Logolook.net (Jan. 3, 2024), https://logolook.net/coppertone-logo [https://perma.cc/Z2TM-X3Y9]. Would a depiction like the Coppertone logo be “patently offensive” for a five-year-old? An eight-year-old? How about a seventeen-year-old? We don't know, and we don't think the burden should be on speakers to find out.

We do not decide today whether minors’ First Amendment rights and the speech that may be deemed obscene for them is the same at all ages. Obviously, children mature as they age (at least, their parents hope so). Many laws distinguish between children based on their age. Driving privileges or children's ability to work often phase in gradually over the course of adolescence. See, e.g., Fla. Stat. §§ 322.16 (restricting sixteen- and seventeen-year-olds’ driving hours), 450.081 (restricting sixteen- and seventeen-year-olds’ work hours). Some privileges, like buying alcohol, are age-restricted even among those who are at least eighteen. Fla. Stat. § 562.11.

But these rules provide very clear guidance as to what they permit, when. The Act doesn't, even as it regulates speech, where “standards of permissible statutory vagueness are strict.” Button, 371 U.S. at 432, 83 S.Ct. 328. The resulting uncertainty as to what is permissible for children of different ages creates an obvious chilling effect that increases the Act's effective breadth.

True, the distinction between seventeen and eighteen—the one we enshrined in Webb—is not inherently less arbitrary or less vague than the distinction between any other two ages.¹³ But much of our law and culture are oriented around the singular age of majority.¹⁴ It is one thing to have a line between obscenity for adults and for minors: it's just one line, drawn in parallel to the countless social norms and legal rights that distinguish between adults and children. This rich social context gives meaning and relative clarity to the line between that which is within minors’ rights to access and that which is “adults-only.”

So from Ginsberg forward, we have recognized states’ ability to distinguish in general terms between materials obscene for “a willing ‘adult’ one month past the state law age of majority and a willing ‘juvenile’ one month younger.” Miller, 413 U.S. at 27, 93 S.Ct. 2607. This kind of binary rule has tradeoffs. When that age cutoff comes at a point less than the state's age of majority, it leaves younger children able to access speech essentially for the benefit of older children as to whom it's not obscene. But it avoids burdening older children's rights based on what might be inappropriate for younger ones. It also mitigates speakers’ need to make a series of difficult judgment calls about what is obscene for children of different ages.

And even if the Act's unique age-variable standard were completely clear, its implementation is likely to be more burdensome than a binary rule like in Webb. That's because admittance to some shows may depend on proof of age from minors, not just adults. Suppose a performance venue's proprietors think a show is acceptable for anyone at least twelve years old, but not younger. How are they to determine with confidence that a child is old enough to attend? Most adults have some readily accessible photo identification showing their age; fewer children do, especially those under driving age. When material is divided into two categories—okay for all and adult-only—only adults need to prove their age, so the burden on people's access to protected speech is relatively low. Under the Act's regime, which relies on children needing to prove their age to access speech, the burden is greater.

Finally, our overbreadth analysis must account for the fact that other Florida law already covers much of the material the Act prohibits.¹⁵ Florida Statutes, Section 847.013(3)(a), for example, makes it a misdemeanor to “knowingly admit a minor for a monetary consideration to premises whereon there is exhibited a motion picture, exhibition, show, representation, or other presentation which, in whole or in part, depicts nudity, sexual conduct, sexual excitement, sexual battery, bestiality, or sadomasochistic abuse and which is harmful to minors.”

The Act's primary extensions beyond this existing law, other than the broad “lewd conduct” provision we've discussed, are (1) that it covers performances attended without “monetary consideration,” and (2) that it uses a gradated age standard rather than the “harmful to minors” standard we took in Webb to mean “obscene for seventeen-year-olds.” So the content the statute newly restricts includes (1) free productions of material “harmful to minors” and (2) content obscene for younger children but not for seventeen-year-olds.

The statute's effective “legitimate sweep,” then, is a narrow slice of speech: the bulk of it is performances that are obscene for some minors but not for some others. Yet the Act's vagueness at every age means it is likely to stifle a substantial amount of protected speech. At oral argument, the Secretary's attorney could not explain the difference even between performances that would be acceptable for an eight-year-old versus a twelve-year-old—a four-year-gap that can be the difference between primary-age kids and adolescents. Oral Argument at 10:28–45. If the Secretary's attorney can't articulate the difference, it's hard to imagine how we could expect performance proprietors to know what the Act means.

But let's assume they could somehow magically discern what even the Secretary's attorney can't define or provide guidance for. In that improbable case, if we follow the Secretary's lead, performance proprietors would give at least a four-year buffer when determining which performances are appropriate for which ages. Every child may enjoy only performances the proprietors think appropriate for children four years younger, and each performance venue will be that much more limited in whom they can admit. We think the gap between the speech that is protected for children of various ages and the speech they will likely be allowed to consume easily comprises “a ‘substantial’ amount of protected free speech, ‘judged in relation to the statute's plainly legitimate sweep.’ ” Hicks, 539 U.S. at 118–19, 123 S.Ct. 2191 (quoting Broadrick, 413 U.S. at 615, 93 S.Ct. 2908).

The dissent insists that the age-variable standard requires only an ordinary degree of “judgment at the margins,” similar to what any foray into obscenity regulation will require. Dissenting Op. at ––––. But we cannot agree that the age-variable standard is just an ordinary regulation of content “harmful to minors.”

For one, as we've explained, no one can have perfect knowledge of the Act's parameters. Recognition of that uncertainty is perhaps the heart of modern First Amendment doctrine, especially in the realm of obscenity. The Act recognizes that uncertainty only to exploit it by demanding covered parties exercise a great degree of “judgment at the margins.” Cf. Dissenting Op. at ––––. But the age-variable standard means it's “margins” all the way down. The dissent rejects the notion that the “rich social context” marking the age of majority makes it relatively easy to discern what content is “adults-only” compared to determining the line between, say, content for ten-year-olds versus eleven-year-olds. Id. But in the very next paragraph, the dissent acknowledges that “context” informs all obscenity judgments. Id. at ––––. We agree. The Secretary simply not shown how context would help Floridians make the judgments the Act requires.

Second, the age-variable standard would deprive Floridians of much of the value of the traditional type of context that informs legal judgment: precedent. Courts explicating a single standard provide useful guidance to regulated parties. If one person steps over the line for what is acceptable for minors as a whole, it creates a benchmark for everyone else. Precedent under the age-variable standard would be much less helpful. A high court might rule in a case that some speech has “serious ․ value” for sixteen-year-olds. Fla. Stat. 827.11(1)(a)(3). But what about fifteen-year-olds? Fourteen-year-olds?

Third, the Act's distinctive scienter regime creates more dangers than limits. The dissent is quick to conclude that the Act's scienter requirement shields Floridians from strict liability. Id. at 39. And to be sure, the Act purports to impose liability only when a defendant knows of or has reason to know of or inquire into a child's age. Fla. Stat. § 827.11(1)(b). But what it gives with one hand it takes away with the other. The Act also provides that “ignorance of a child's age, [or] a child's misrepresentation of his or her age ․ may not be raised as a defense.” Id. § 827.11(2). In other words, this is strict liability for anybody who admits a minor to a performance after seeing a convincing fake ID. Even Florida's laws for serving minors alcohol are more forgiving. See Fla. Stat. § 562.11(1)(d). (And drinking or serving alcohol is not a constitutional right.)

Fourth, and finally, the dissent does not address the fact that the Act's age-variable standard forecloses Webb’s path: reading a “harmful to minors” statute as covering only content that is obscene for all minors, even those who are almost of age. Doing so dramatically narrows the potential scope and impact of a statute by limiting it to material that is strictly adults-only. In Webb, we noted that “only a minimal number of works will have serious value for reasonable adults but not for reasonable [older] minors.” Webb, 919 F.2d at 1506. The burdens on speech purveyors and consumers alike were minimal.

We cannot read the Act this way. The dissent does not acknowledge this fact, let alone address the difference it makes. Even if Floridians had perfect knowledge of which performances were obscene for which ages under the Act, carrying out its requirements would greatly burden speech that is impermissible for younger minors but not older ones. Every performance with an age cutoff somewhere between zero and eighteen would need to check IDs at the door. Minor patrons—including those far too young to have a driver's license—would need to obtain and present proof of age to access even performances are that constitutionally protected speech for their age. The problem with the Act is not just that it “lacks clarity,” Dissenting Op. at ––––, but that even if its substance were unambiguous, it would still burden protected speech far beyond anything this Court has previously sanctioned.

To conclude our discussion of the Act, we note that nothing we say today decides whether, in principle, some “age-variable” obscenity standards can pass constitutional muster. Eight-year-olds are different from thirteen-year-olds who are different from seventeen-year-olds. If Florida wishes to describe in detail which depictions it considers obscene for which ages, it is welcome to do so.¹⁶ Nor does anything we say touch on Florida's many other laws protecting children from harmful content. See, e.g., Fla. Stat. §§ 847.012; 847.0125; 847.0133; 847.0134; 847.0138; 847.01385; 847.0141.

Florida speakers and parents also still retain the freedom to decide what speech is appropriate at which ages. Nothing we say impinges on parents’ ability to determine what performances are appropriate for their children. And nothing we say limits the ability of performers or performance venues to choose to welcome only patrons of a certain age. But the Act would take those decisions out of citizens’ hands and instead give them to the Secretary and other State officials—all without clear notice of where the standards begin and end.

The Act's age-variable standard poses as a well-tailored limit on its application. But in practice it is yet another “I know it when I see it” provision. For these reasons, the age-variable standard sweeps more broadly than the obscenity statute we upheld in Webb yet offers less guidance about what it covers. And the Act's effective coverage expands significantly beyond its “legitimate sweep.” So Hamburger Mary's appears likely to succeed on the merits of its facial challenge.

D. A broad injunction of the Act is proper.

Last, we address the Secretary's argument that the district court erred by entering a so-called “universal” injunction prohibiting the Act's enforcement against anyone, not just Hamburger Mary's.

The contours of this argument are familiar. Injunctive relief, the Secretary argues, has traditionally been limited to that which is “no more burdensome to the defendant than necessary to provide complete relief to the plaintiffs.” Madsen v. Women's Health Ctr., Inc., 512 U.S. 753, 765, 114 S.Ct. 2516, 129 L.Ed.2d 593 (1994) (quoting Califano v. Yamasaki, 442 U.S. 682, 702, 99 S.Ct. 2545, 61 L.Ed.2d 176 (1979)). So, the Secretary urges, enjoining a statute's enforcement against parties not before this Court exceeds our equitable powers. In support, the Secretary highlights cases like United States v. National Treasury Employees Union, 513 U.S. 454, 115 S.Ct. 1003, 130 L.Ed.2d 964 (1995). There, the Court agreed that the relief from a facial First Amendment challenge to a statute barring federal employees from making paid speeches should be limited to the lower-ranking federal employees who formed the plaintiff class, and not extended to the “high-level employees” who presented a “different constitutional question than the one” the Court decided. Id. at 477–78, 115 S.Ct. 1003.

This argument is especially familiar because we have addressed it before, in this very case. The Secretary raised similar arguments when she asked us to stay the district court's preliminary injunction to the extent it prevented enforcement against nonparties. HM Florida-Orl, LLC, 2023 WL 6785071, at *2–3. A majority of the panel held that the district court did not abuse its discretion by issuing universal relief. Id. at *4. We noted that while that some cases in our circuit and the Supreme Court supported the Secretary's position, “they are not the only authorities on point.” Id. And we observed that universal injunctions can play a role that can take on special importance when it comes to First Amendment cases. Id. at *3–4.

We briefly review the relevant authorities. The Supreme Court has long recognized an “expansive remedy” for First Amendment overbreadth challenges for the same reasons behind the overbreadth doctrine: a fear that the threat of enforcement will chill protected speech. Hicks, 539 U.S. at 119, 123 S.Ct. 2191 (2003). See also Broadrick, 413 U.S. at 612, 93 S.Ct. 2908 (Overbreadth litigants “are permitted to challenge a statute not because their own rights of free expression are violated, but because of a judicial prediction or assumption that the statute's very existence may cause others not before the court to refrain from constitutionally protected speech or expression.”). We have, too. See, e.g., FF Cosmetics Fl, Inc. v. City of Miami Beach, 866 F.3d 1290, 1303-04 (11th Cir. 2017) (declaring enforcement of a city ordinance “totally forbidden” (quoting Broadrick, 413 U.S. at 612, 93 S.Ct. 2908)); HM Florida-Orl, LLC, 2023 WL 6785071, at *4 (collecting cases “where a law has been found to be overbroad in violation of the First Amendment [and] we have affirmed injunctions preventing enforcement of a law or ordinance against nonparties as well as parties”).

Nothing has changed since our previous decision.¹⁷ In fact, the Supreme Court denied the Secretary's later application for a partial stay on the same grounds the Secretary presented to us. Griffin, 144 S. Ct. at 1. Six Justices agreed, with two observing that whatever the merits of the Secretary's objection to universal injunctions as a general matter, “the context of a First Amendment overbreadth challenge” involves distinct “doctrinal complexities.” Id. at 2.

We also note that the concerns we've expressed about the “drastic form of relief” known as the nationwide injunction are largely absent in this case involving a state law. Georgia v. President of the U.S., 46 F.4th 1283, 1303 (11th Cir. 2022). Of course, even nationwide injunctions are acceptable “in appropriate circumstances.” Florida v. Dep't of Health & Hum. Servs., 19 F.4th 1271, 1281–82 (quoting Texas v. United States, 809 F.3d 134, 188 (5th Cir. 2015), as revised (Nov. 25, 2015)). Still, we have sometimes been skeptical of nationwide injunctions of federal policies because “[b]y cutting off parallel lawsuits, nationwide injunctions frustrate foundational principles of the federal court system” and encourage forum shopping. Georgia, 46 F.4th at 1305–06.

But given that nationwide injunctions are nevertheless acceptable to “protect similarly situated nonparties,” Florida, 19 F.4th at 1282 (citing City of Chicago v. Barr, 961 F.3d 882, 916–17 (7th Cir. 2020), statewide injunctions, which reach only within our Court's geographic jurisdiction and do not silence other federal courts who might otherwise speak on the matter, enjoy an even stronger claim to permissibility in appropriate circumstances. See also Mila Sohoni, The Lost History of the “Universal” Injunction, 133 Harv. L. Rev. 920, 958–79 (2020) (analyzing the history of “statewide” injunctions and concluding it was likely “well understood [in the early part of the twentieth century] that when a federal district court declared a state law unconstitutional, it could properly enjoin the law's enforcement against nonparties”).

The Secretary raises one new argument in this challenge to the scope of the district court's injunction, which she grounds in courts’ supposed obligation to consider as-applied challenges before facial ones. More specifically, the Secretary contends that the district court either determined that the Act was unconstitutional as applied to Hamburger Mary's so the court should have issued relief as to only it, or that the district court improperly weighed in on the Act's facial overbreadth (and so issued broad relief) without considering whether the statute was unconstitutional as applied.

To the first point, we do not read the district court's order as addressing the Act's constitutionality only as applied to Hamburger Mary's. The part of the order the Secretary cites in support of her argument begins with a header, “The Act is a Facially Content-Based Regulation,” (emphasis added) and the analysis not surprisingly then considers a facial, not as-applied challenge.

As to the second argument, we disagree that courts cannot reach an overbreadth challenge until they have evaluated an as-applied one. The Supreme Court did exactly that in Stevens, when it held a federal statute “invalid under the First Amendment.” Stevens, 559 U.S. at 482, 130 S.Ct. 1577. In reaching its conclusion, the majority rejected the dissent's argument that “because there has not been a ruling on the validity of the statute as applied to Stevens, our consideration of his facial overbreadth claim is premature.” Id. at 473 n.3, 130 S.Ct. 1577. Like here, neither the parties nor the lower courts in Stevens thought the case involved an as-applied challenge. Id. But the absence of an as-applied challenge or ruling did not prevent the Court from deciding the case on facial overbreadth grounds. Id.

The Court's decision in Americans for Prosperity Foundation v. Bonta is even more illustrative, as the case clearly presented an as-applied challenge but the Court opted to conduct only a facial analysis. The suit concerned petitioners’ as-applied and facial First Amendment challenges to a California law requiring certain nonprofit organizations to disclose donors’ names and addresses. Ams. for Prosperity Found. v. Bonta, 594 U.S. 595, 611, 141 S.Ct. 2373, 210 L.Ed.2d 716 (2021). Although the Court noted the as-applied challenge, it did not consider that challenge, instead holding California's policy facially unconstitutional. Id. at 618, 141 S.Ct. 2373. In doing so, the majority rejected the dissent's argument that only “as-applied relief” was warranted because in the majority's view, “the pertinent facts ․ are the same across the board” and because “First Amendment freedoms need breathing space to survive.” Id. (quoting Button, 371 U.S. at 433, 83 S.Ct. 328). That's precisely the case here.

Americans for Prosperity and Stevens show that whether or not a plaintiff brings an as-applied challenge, a court may evaluate a law's constitutionality on its face. And as we have explained, when a court holds a law facially unconstitutional, broad-based relief may follow.

The district court's ruling is therefore affirmed.

AFFIRMED.

“In litigation generally, and in constitutional litigation most prominently, courts in the United States characteristically pause to ask: Is this conflict really necessary?” Arizonans for Off. Eng. v. Arizona, 520 U.S. 43, 75, 117 S. Ct. 1055, 1072, 137 L.Ed.2d 170 (1997). Here, the Majority fails to ask this question and, by skipping it, puts the First Amendment on a collision course with core principles of federalism and judicial restraint. Because this conflict was entirely avoidable, I respectfully dissent.

This case presents a pre-enforcement challenge to Florida's new statute regulating the admission of minors to live, sexually explicit performances. The statute adopts the familiar Miller v. California, 413 U.S. 15, 93 S. Ct. 2607, 37 L.Ed.2d 419 (1973), three-part obscenity test, adjusted only to account for the age of the child—an approach apparently inspired by Supreme Court cases such as Ginsberg v. New York, 390 U.S. 629, 88 S. Ct. 1274, 20 L.Ed.2d 195 (1968). In other words, whatever individual lawmakers may have claimed, Florida's legislature has passed yet another run-of-the-mill obscenity statute.

But before Florida's courts had any chance to interpret the law, the District Court issued a sweeping injunction, holding that the statute was likely unconstitutionally vague and overbroad. On appeal of the injunction, we had two good options: we could apply ordinary tools of statutory construction to read the statute narrowly and avoid unnecessary constitutional conflict, or we could certify the unsettled state-law questions to the Florida Supreme Court, allowing the state's highest court to speak first. Either course would have honored federalism, respected judicial modesty, and kept this Court within its Article III limits.

Instead, the Majority chooses a third, unwarranted path: it reads the statute in the broadest possible way, maximizes constitutional conflict, and strikes the law down wholesale. That decision rests on two flawed premises. First, the Majority “assume[s]” that the statutory phrase “lewd conduct” must “mean something different” than what courts have understood it to mean for decades. Maj. Op. at ––––. Second, the Majority objects to the statute's “age-variable” obscenity standard, apparently preferring an arbitrary, one-size-fits-all age cutoff. See Maj. Op. at –––– – ––––. But as I will explain, neither argument justifies facial invalidation of the statute.

My dissent proceeds as follows: Part I provides the factual and procedural background. Part II explains the constitutional framework governing obscenity, overbreadth, and vagueness. Part III addresses the statute's lewd-conduct provision, showing why, under proper construction, it avoids constitutional conflict—or at minimum, why the question should have been certified to the Florida Supreme Court. Part IV examines the statute's age-based obscenity standard, explaining why that framework is neither vague nor overbroad and why the Majority's concerns should have been resolved through narrowing interpretation or certification. Part V explains why the District Court's injunction sweeps too broadly. And Part VI concludes.

I. Background

In 2023, Florida's Legislature enacted Fla. Stat. § 827.11, a statute making it a misdemeanor for a person to knowingly admit a child to an “adult live performance.” The statute defines that term, in part, to mean a sexually explicit show that would be obscene in light of the child's age. The statute provides:

(1) As used in this section, the term:

(a) “Adult live performance” means any show, exhibition, or other presentation in front of a live audience which, in whole or in part, depicts or simulates nudity, sexual conduct, sexual excitement, or specific sexual activities as those terms are defined in s. 847.001, lewd conduct, or the lewd exposure of prosthetic or imitation genitals or breasts when it:

1. Predominantly appeals to a prurient, shameful, or morbid interest;

2. Is patently offensive to prevailing standards in the adult community of this state as a whole with respect to what is suitable material or conduct for the age of the child present; and

3. Taken as a whole, is without serious literary, artistic, political, or scientific value for the age of the child present.

(b) “Knowingly” means having general knowledge of, reason to know, or a belief or ground for belief which warrants further inspection or inquiry of both:

1. The character and content of any adult live performance described in this section which is reasonably susceptible of examination by the defendant; and

2. The age of the child.

(2) A person's ignorance of a child's age, a child's misrepresentation of his or her age, or a bona fide belief of a child's consent may not be raised as a defense in a prosecution for a violation of this section.

(3) A person may not knowingly admit a child to an adult live performance.

(4) A violation of subsection (3) constitutes a misdemeanor of the first degree, punishable as provided in s. 775.082 or s. 775.083.

Fla. Stat. § 827.11.

Shortly after the statute's enactment, an Orlando restaurant known as Hamburger Mary's sued Melanie Griffin, in her official capacity as Secretary of the Florida Department of Business and Professional Regulation. The restaurant alleged that the statute violated the First Amendment, asserting that the law prohibited it from admitting minors to its drag performances. Hamburger Mary's also claimed that the statute was unconstitutionally broad and vague. For example, it argued that “[t]he terms ‘predominately,’ ‘shameful or morbid’ are vague terms subject to the interpretation of the reader and not subjective.” The restaurant contended that “[t]he uncertainty about what specific conduct this law prohibits ․ is precisely what concerns the Plaintiff in this case.” It requested that “the Defendants be permanently enjoined from enforcing the [statute].”

A day after filing its complaint, Hamburger Mary's moved for a temporary restraining order and preliminary injunction. It expanded on its concerns about overbreadth and vagueness and asked the Court to block the statute from taking effect.

The District Court granted the preliminary injunction. See HM Fla.-ORL, LLC v. Griffin, 679 F. Supp. 3d 1332 (M.D. Fla. 2023). It found that Hamburger Mary's had shown a substantial likelihood of success on the merits of its First Amendment claim, concluding that the statute was likely unconstitutionally overbroad and vague. See id. at 1341–44. The Court enjoined Secretary Griffin from enforcing the statute statewide—not just against Hamburger Mary's, but against any person or entity. Id. at 1345. The Court also included a footnote expanding the injunction to “proceedings instituted, maintained, or prosecuted under the statutes.” Id. at 1345 n. 17.

Four days later, the Secretary appealed. In the District Court, she moved for a partial stay of the injunction, arguing that even if enforcement were barred against Hamburger Mary's, the statute should remain enforceable against others during the appeal. The District Court denied that motion, classifying the request as an attempt “to neuter the Court's injunction.” See HM Fla.-ORL, LLC v. Griffin, No. 6:23-CV-950, 2023 WL 11257409, at *1 (M.D. Fla. July 19, 2023). Because the Court had concluded that the Act was facially unconstitutional, it held that the injunction “necessarily must extend to protect all Floridians.” Id. at *4.

The Secretary then sought a stay from our Court. A divided panel denied the request.¹ See HM Fla.-Orl, LLC v. Governor of Fla., No. 23-12160, 2023 WL 6785071 (11th Cir. Oct. 11, 2023). The panel majority reasoned that the District Court had not abused its discretion in crafting an injunction that extended beyond the immediate parties, especially given the First Amendment overbreadth concerns. See id. at *3–4.

Judge Brasher dissented. He offered a careful and principled analysis of why the injunction's scope exceeded constitutional and remedial limits. Id. at *4–6. He emphasized that under Article III, established remedial principles, and Circuit precedent, federal courts may grant only the relief necessary to redress the plaintiff's injury. Id. at *4–6. A universal injunction—one barring enforcement against nonparties—was inappropriate, he explained, because an injunction limited to Hamburger Mary's would have fully remedied its asserted harms. Id.

The Secretary then sought a stay from the Supreme Court, which denied the application. Griffin v. HM Fla.-ORL, LLC, ––– U.S. ––––, 144 S. Ct. 1, 1, 217 L.Ed.2d 227 (2023). Justice Thomas, Justice Alito, and Justice Gorsuch would have granted the stay. Justice Kavanaugh, joined by Justice Barrett wrote separately to clarify that the stay request raised only the scope of the District Court's injunction—not the underlying First Amendment merits—and that the Secretary had not shown a sufficient likelihood that the Supreme Court would grant certiorari on that procedural issue.² Id. at 1–2.

II. The First Amendment Framework

A. Obscenity

The First Amendment provides that the government “shall make no law ․ abridging the freedom of speech.” U.S. Const. amend. I. Generally, this means that government regulations of speech must survive heightened judicial scrutiny. For instance, “[c]ontent-based laws—those that target speech based on its communicative content—are presumptively unconstitutional and may be justified only if the government proves that they are narrowly tailored to serve compelling state interests.” Reed v. Town of Gilbert, 576 U.S. 155, 163, 135 S. Ct. 2218, 2227, 192 L.Ed.2d 236 (2015).

But not all speech receives full constitutional protection. Since its inception, the First Amendment has allowed the regulation of certain “well-defined and narrowly limited classes of speech,” such as defamation, fraud, incitement, speech integral to criminal conduct—and obscenity. United States v. Stevens, 559 U.S. 460, 468, 130 S. Ct. 1577, 1584, 176 L.Ed.2d 435 (2010) (quoting Chaplinsky v. New Hampshire, 315 U.S. 568, 571–72, 62 S. Ct. 766, 769, 86 L.Ed. 1031 (1942)). These forms of speech “are of such slight social value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality.” Chaplinsky, 315 U.S. at 572, 62 S. Ct. at 769.

It “has been categorically settled ․ that obscene material is unprotected by the First Amendment.” Miller, 413 U.S. at 23, 93 S. Ct. at 2614. Still, “statutes designed to regulate obscene materials must be carefully limited.” Id. at 23–24, 93 S. Ct. at 2614. That is why, in Miller v. California, the Supreme Court “confine[d] the permissible scope of such regulation” by establishing a three-pronged test:

(a) whether the average person, applying contemporary community standards would find that the work, taken as a whole, appeals to the prurient interest; (b) whether the work depicts or describes, in a patently offensive way, sexual conduct specifically defined by the applicable state law; and (c) whether the work, taken as a whole, lacks serious literary, artistic, political, or scientific value.

Id. at 24, 93 S. Ct. at 2614–15 (citation omitted).

Among other things, this test limits the definition of obscenity “to works which depict or describe sexual conduct.” Id. And it requires that regulated conduct “be specifically defined by the applicable state law, as written or authoritatively construed.” Id. But the Court in Miller made clear that “existing state statutes, as construed heretofore or hereafter, may well be adequate.” Id. at 24, 93 S. Ct. at 2615 n.6. Indeed, in a case decided the same day as Miller, the Court noted its own willingness to construe federal obscenity statutes so as to avoid constitutional doubts. See United States v. 12 200-Foot Reels of Super 8mm. Film, 413 U.S. 123, 130, 93 S. Ct. 2665, 2670 n.7, 37 L.Ed.2d 500 (1973).

To be clear, Miller does not demand complete precision in how states define obscenity. As the Court in Miller noted, “[i]f the inability to define regulated materials with ultimate, god-like precision altogether removes the power of the States or the Congress to regulate, then ‘hard core’ pornography may be exposed without limit to the juvenile, the passerby, and the consenting adult alike.” Miller, 413 U.S. at 27–28, 93 S. Ct. at 2617. Consequently, for a statute to provide fair notice, the Constitution requires only “that the language conveys sufficient[l]y definite warning as to the proscribed conduct when measured by common understanding and practices.” Roth v. United States, 354 U.S. 476, 491, 77 S. Ct. 1304, 1312, 1 L.Ed.2d 1498 (1957) (internal quotation marks omitted).

Miller also recognizes the states’ “legitimate interest” in preventing obscene material from being “expos[ed] to juveniles.” Miller, 413 U.S. at 18–19, 93 S. Ct. at 2612. Miller echoes sentiments expressed in an earlier case, Ginsberg v. New York, where the Court upheld a statute specifically targeting the sale of obscene materials to minors. See id. at 36, 93 S. Ct. at 2621 n.17; Ginsberg, 390 U.S. at 631–34, 88 S. Ct. at 1275–77.

In Ginsberg, the Court explained that the First Amendment's protections were not identical for adults and minors, reiterating that “the power of the state to control the conduct of children reaches beyond the scope of its authority over adults.” Ginsberg, 390 U.S. at 638, 88 S. Ct. at 1280 (quoting Prince v. Massachusetts, 321 U.S. 158, 170, 64 S. Ct. 438, 444, 88 L.Ed. 645 (1944)). And in cases following Ginsberg and Miller, the Supreme Court continued to emphasize the government's special interest in protecting children from exposure to otherwise protected speech. See, e.g., F.C.C. v. Pacifica Found., 438 U.S. 726, 749–50, 98 S. Ct. 3026, 3040–41, 57 L.Ed.2d 1073 (1978) (citing “the concerns recognized in Ginsberg” as justifying the regulation of “indecent broadcasting”).

B. Overbreadth and Vagueness

“For a host of good reasons, courts usually handle constitutional claims case by case, not en masse.” Moody v. NetChoice, LLC, 603 U.S. 707, 723, 144 S. Ct. 2383, 2397, 219 L.Ed.2d 1075 (2024). So plaintiffs wishing to challenge a law on its face usually must prove either “that no set of circumstances exists under which the [law] would be valid” or “that the law lacks a ‘plainly legitimate sweep.’ ” Id. (alteration in original) (quoting United States v. Salerno, 481 U.S. 739, 745, 107 S. Ct. 2095, 2100, 95 L.Ed.2d 697 (1987); Wash. State Grange v. Wash. State Republican Party, 552 U.S. 442, 449, 128 S. Ct. 1184, 170 L.Ed.2d 151 (2008)).

This high bar is lowered only slightly for First Amendment challenges. In this unique context, “to provide breathing room for free expression,” the Supreme Court has “substituted a less demanding though still rigorous standard.” Id. (internal quotation omitted). When a plaintiff challenges a law as facially overbroad, the question for courts becomes whether “a substantial number of [the law's] applications are unconstitutional, judged in relation to the statute's plainly legitimate sweep.” Id. (alteration in original) (quoting Ams. for Prosperity Found. v. Bonta, 594 U.S. 595, 615, 141 S. Ct. 2373, 210 L.Ed.2d 716 (2021)).

The law's unconstitutional effect must be substantial because facial invalidation is still “strong medicine” that “destroys some good along with the bad.” United States v. Hansen, 599 U.S. 762, 770, 143 S. Ct. 1932, 1939, 216 L.Ed.2d 692 (2023). Just recently, the Supreme Court admonished us to be more rigorous in our handling of overbreadth claims. See Moody, 603 U.S. at 724, 144 S. Ct. at 2397–98. Moody reiterates that a court considering an overbreadth challenge must ask: “What activities, by what actors, do the laws prohibit or otherwise regulate?” Id. And once the court answers this question, it “must explore the laws’ full range of applications—the constitutionally impermissible and permissible both—and compare the two sets.” Id. at 726, 144 S. Ct. at 2398. Only then can the court properly declare that a law is, in fact, overbroad.

Alternatively—or in addition to overbreadth—a plaintiff may challenge a law on its face because it is unconstitutionally vague. The vagueness doctrine is technically an outgrowth of due process, but it features prominently in First Amendment challenges. See, e.g., United States v. Williams, 553 U.S. 285, 304, 128 S. Ct. 1830, 1845, 170 L.Ed.2d 650 (2008). In general, a law is void for vagueness if it “fails to provide a person of ordinary intelligence fair notice of what is prohibited, or is so standardless that it authorizes or encourages seriously discriminatory enforcement.” Id. Once again, however, “perfect clarity and precise guidance have never been required even of regulations that restrict expressive activity.” Id.

III. The Lewd-Conduct Provision

In reaching its conclusion that Fla. Stat. § 827.11 is both vague and overbroad, the Majority relies on a key premise—that the statutory phrase “lewd conduct” must sweep in at least some protected speech. But that is wrong for at least four reasons:

• First, the phrase, as appropriately construed, satisfies Miller’s specificity requirements.

• Second, the phrase is not “mere surplusage,” see Maj. Op. at ––––, and labeling it as such does not end our statutory analysis.

• Third, even if the text remained unclear after a full analysis, certification to the Supreme Court of Florida was the proper solution.

• And fourth, even if we assume the Majority's reading to be correct, its facial overbreadth analysis would be woefully inadequate to invalidate the law.

A. Specificity and Narrow Construction

The Majority's analysis starts with the claim that phrases like “lewd conduct” are too broad to satisfy Miller’s specificity requirement. Maj. Op. at ––––. It takes the example definitions in Miller as proof that the Supreme Court would not have allowed anything less precise. That claim is remarkable because the Court in Miller resisted that argument itself, and in later cases outright rejected it.

Miller holds that the conduct regulated by an obscenity statute “must be specifically defined by the applicable state law, as written or authoritatively construed.” Miller, 413 U.S. at 24, 93 S. Ct. at 2615 (emphasis added). The Majority emphasizes this specificity requirement, but it gives short shrift to the courts’ role in statutory construction. Miller plainly contemplates an active role for courts in providing narrowing constructions of obscenity statutes. And in later cases applying Miller, the Supreme Court has led by example. See, e.g., Hamling v. United States, 418 U.S. 87, 115, 94 S. Ct. 2887, 2906, 41 L.Ed.2d 590 (1974).

In Miller, the Court “emphasize[d] that it [wa]s not [its] function to propose regulatory schemes for the States.” Id. at 25, 93 S. Ct. at 2615. Rather, it sought only “to give a few plain examples of what a state statute could define for regulation.” Id. (emphasis added). Those examples included “[p]atently offensive representations or descriptions of ultimate sexual acts, normal or perverted, actual or simulated,” and “[p]atently offensive representation or descriptions of masturbation, excretory functions, and lewd exhibition of the genitals.” Id. Far from an exhaustive list, these examples were meant to illustrate the sort of “ ‘hard core’ sexual conduct” with which Miller was concerned. See id. at 27, 93 S. Ct. at 2616.

In 12 200-Foot Reels, decided the same day as Miller, the Court remarked on its own role in construing federal obscenity statutes. See 12 200-Foot Reels, 413 U.S. at 130, 93 S. Ct. at 2670 n.7. While it “le[ft] to state courts the construction of state legislation,” the Court noted its own “duty to authoritatively construe federal statutes where ‘a serious doubt of constitutionality is raised’ and ‘a construction of the statute is fairly possible by which the question may be avoided.’ ” Id. (quoting United States v. Thirty-Seven Photographs, 402 U.S. 363, 369, 91 S. Ct. 1400, 1404, 28 L.Ed.2d 822 (1971)). To do so, the Court was prepared to construe terms such as “ ‘obscene,’ ‘lewd,’ ‘lascivious,’ ‘filthy,’ ‘indecent,’ or ‘immoral’ ” as compatible with the examples Miller provided. Id.

In Hamling v. United States, the Court delivered as promised: it construed 18 U.S.C. § 1461—prohibiting “obscene, lewd, lascivious, indecent, filthy[,] or vile” mailings—as consistent with Miller’s requirements. Hamling, 418 U.S. at 112, 94 S. Ct. at 2905. At the same time, the Court made clear that Miller’s examples “were not intended to be exhaustive.” Id. at 114, 94 S. Ct. at 2906. Again, their purpose was merely to emphasize a line that legislators could not cross: obscenity regulation must be limited to depictions of sexual conduct. See id.; see also Stevens, 559 U.S. at 479–80, 130 S. Ct. at 1591 (declining to extend Miller to depictions of animal cruelty); Brown v. Ent. Merchs. Ass'n, 564 U.S. 786, 792–93, 131 S. Ct. 2729, 2734–35, 180 L.Ed.2d 708 (2011) (declining to extend Miller to depictions of violence).

If it were possible to define the phrase “lewd conduct” with perfect clarity, courts presumably would have done so already. The Court in Miller was, as the Majority notes, concerned with specificity and fair notice. See Miller, 413 U.S. at 27–28, 93 S. Ct. at 2616–17. Yet the Court in Miller readily acknowledged the limits of human language, refusing to require “ultimate, god-like precision” of lawmakers. Id. Rather, as long as states abided by Miller’s three prongs, the Court was “satisfied that these specific prerequisites w[ould] provide fair notice” to would-be defendants. Id.

Miller and its progeny “were intended neither as legislative drafting handbooks nor as manuals of jury instructions.” Hamling, 418 U.S. at 115, 94 S. Ct. at 2906. Legislators are not expected to be omnipotent, but Courts are expected to construe statutes, whenever possible, to comply with the limitations outlined in Miller. The key limitation is that obscene speech depicts sexual acts, of which Miller provides examples. Here, § 827.11—including its reference to “lewd conduct”—can and should be construed to satisfy Miller’s specificity requirement. The Supreme Court has shown us how to do so.

B. Canons of Construction

1. “Mere Surplusage”

The Majority's analysis continues with the claim that the phrase “lewd conduct” might be “mere surplusage” because § 827.11 also describes some more specific conduct falling within its prohibition. The Majority's reasoning is as follows: The statute defines adult live performances as depicting “lewd conduct, or the lewd exposure of prosthetic or imitation genitals or breasts.” Maj. Op. at ––––; Fla. Stat. § 827.11(1)(a). But the statute also lists “nudity, sexual conduct, sexual excitement, [and] specific sexual activities as those terms are defined in [Section] 847.001.” Id. Section 847.001, in turn, contains a laundry list of sexual acts. Maj. Op. at ––––; Fla. Stat. § 847.001. Therefore, “lewd conduct,” if defined consistently with Miller, could not possibly refer to acts that are not already described elsewhere in the statute. Id.

Of course, the Majority acknowledges that courts have, in the decades following Miller, upheld statutes prohibiting “lewd” acts by defining the term consistently with our Constitution. Maj. Op. at –––– – ––––. For instance, in Chesebrough v. State, 255 So. 2d 675, 677 (Fla. 1971), the Supreme Court of Florida explained that “[l]ewdness may be defined as the unlawful indulgence of lust, signifying that form of immorality which has a relation to sexual impurity.”³ And in 12 200-Foot Reels and Hamling, the United States Supreme Court explained how those and similar terms could be read consistently with the examples of “ ‘hard core’ sexual conduct” provided in Miller. See 12 200-Foot Reels, 413 U.S. at 130, 93 S. Ct. at 2670 n.7; Hamling, 418 U.S. at 110–14, 94 S. Ct. at 2904–06.

So why does the Majority not rely on these cases? It says it cannot do so because Florida has defined away all permissible applications of “lewd conduct” elsewhere in § 827.11’s text. That claim is remarkable for two reasons: First, it requires us to adopt precisely the same “I know it when I see it” approach the Majority condemns. And second, it suggests that our primary role in interpreting statutes is avoiding redundancy—rather than, say, avoiding constitutional conflict.

On the first point, courts should not be in the business of deciding whether a state has so perfectly captured the essence of hard-core sexual conduct. It is ironic that the Majority repeatedly jabs at Justice Stewart's “I know it when I see it” quip when its own analysis boils down to the same claim. See Jacobellis v. Ohio, 378 U.S. 184, 197, 84 S. Ct. 1676, 1683, 12 L.Ed.2d 793 (1964) (Stewart, J., concurring). Trust us, the Majority says, we know exactly what depictions are implicated in Miller, and Florida has exhausted them all. See Maj. Op. at ––––. That is a dubious claim because Miller does not even attempt to define “ ‘hard core’ sexual conduct”—it explicitly leaves that to the states. See Miller, 413 U.S. at 27–28, 93 S. Ct. at 2616–17. We should do the same. To faithfully apply Miller, we must enforce its limitations while at the same time respecting states’ ultimate authority over the act of legislation.

On the second point, the Majority mistakes our Court's role for that of an editor reviewing text for stylistic error. Contrary to the Majority's analysis, redundancy is not such a great evil that it singlehandedly precludes us from upholding a statute. A court's role in statutory construction is to consult the myriad interpretive tools available to glean the statute's proper meaning. By disregarding all these tools save for one—the surplusage canon—the Majority short-circuits this process and gives Fla. Stat. § 827.11 an artificially problematic scope.

By default, we construe statutory text in accordance with its ordinary meaning. Heyman v. Cooper, 31 F.4th 1315, 1319 (11th Cir. 2022). To do so, we consult and apply various interpretive principles—often called “canons of construction.” Id. These canons are “useful tools,” but they are not strict rules. See Facebook, Inc. v. Duguid, 592 U.S. 395, 410–13, 141 S. Ct. 1163, 1173–75, 209 L.Ed.2d 272 (2021) (Alito, J., concurring). Rather, they are best treated as instruments in a “statutory-interpretation toolbox,” and it would be a mistake to rely on one to the detriment of all others. See West Virginia by & through Morrisey v. U.S. Dep't of the Treasury, 82 F.4th 1068, 1071–72 (11th Cir. 2023) (Rosenbaum, J., dissenting from denial of rehearing en banc).

Yet the Majority does exactly that: it treats a single tool, the surplusage canon, as dispositive of the statute's meaning. But “[r]edundancy is not a silver bullet.” Rimini St., Inc. v. Oracle USA, Inc., 586 U.S. 334, 346, 139 S. Ct. 873, 881, 203 L.Ed.2d 180 (2019). Although courts try to read text so that each word carries independent meaning, “our preference for avoiding surplusage constructions is not absolute.” Lamie v. U.S. Trustee, 540 U.S. 526, 536, 124 S. Ct. 1023, 1031, 157 L.Ed.2d 1024 (2004). Often, we must tolerate some redundancy even if we think it in poor taste:

So like all other canons, this one must be applied with judgment and discretion, and with careful regard to context. It cannot always be dispositive because (as with most canons) the underlying proposition is not invariably true. Sometimes drafters do repeat themselves and do include words that add nothing of substance, either out of a flawed sense of style or to engage in the ill-conceived but lamentably common belt-and-suspenders approach.

Antonin Scalia & Bryan A. Garner, Reading Law: The Interpretation of Legal Texts 176–77 (2012).

Examples of the “belt-and-suspenders approach” abound: “Execute and perform—what satisfies one but not the other? Rest, residue, and remainder—could a judge interpret these as referring to three distinct things? Peace and quiet—when is peace not quiet?” Id. at 177. I would add the phrase “obscene, lewd, lascivious, indecent, filthy[,] or vile.” See 18 U.S.C. § 1461. Under the Majority's logic, the word “obscene” alone should “exhaust the types of ‘hard core’ depictions that Miller described.” See Maj. Op. at –––– (quoting Miller, 413 U.S. at 25, 93 S. Ct. at 2615). After all, the First Amendment exception detailed in Miller pertains exclusively to “obscene materials.” See Miller, 413 U.S. at 23–24, 93 S. Ct. at 2614–15. If something is not obscene, it cannot be regulated under Miller. So, by the Majority's reasoning, the statutory words “lewd, lascivious, filthy, [and] vile” presumably must cover conduct beyond Miller’s permitted scope.

But that is not what the Supreme Court concluded in Hamling, and it is not how courts interpret statutes. In Hamling, the Court addressed this statutory language and held that it could be construed entirely consistent with Miller. See Hamling, 418 U.S. at 114, 94 S. Ct. at 2906. Indeed, the Court noted its “duty to authoritatively construe federal statutes where a serious doubt of constitutionality is raised and a construction of the statute is fairly possible by which the question may be avoided.” Id. at 113, 94 S. Ct. at 2905. Because such a construction was possible, the Court interpreted the statute to abide by Miller. Id. at 114, 94 S. Ct. at 2906.

2. The Constitutional Doubt Canon

Hamling demonstrates the application of the constitutional-doubt canon. Put simply, the canon provides that a statute should be interpreted so as to avoid placing its constitutionality in doubt. See Zadvydas v. Davis, 533 U.S. 678, 689, 121 S. Ct. 2491, 2498, 150 L.Ed.2d 653 (2001). As the Supreme Court recently put it, “It bears emphasis that even if the Government's reading were not the best one, the interpretation is at least ‘fairly possible’—so the canon of constitutional avoidance would still counsel us to adopt it.” Hansen, 599 U.S. at 781, 143 S. Ct. at 1946 (quoting Jennings v. Rodriguez, 583 U.S. 281, 296, 138 S. Ct. 830, 200 L.Ed.2d 122 (2018)). That is because “[w]hen legislation and the Constitution brush up against each other, our task is to seek harmony, not to manufacture conflict.”⁴ Id.

The constitutional-doubt canon is of special significance here, where a federal court sits in review of a state statute. In general, the canon respects the separation of powers by avoiding conflict between courts and legislatures. But here it also furthers principles of federalism and comity because it avoids friction between the states and our federal government. We should remember that we are not the court charged with authoritatively construing § 827.11—that would be the Supreme Court of Florida. So we should be especially reluctant to strike down the statute before that Court has had the chance to read it.

3. Other Canons

But the list goes on: the Majority's strained reading of § 827.11 bypasses too many interpretive principles to count. Consider the noscitur a sociis canon, which provides that “a word is known by the company it keeps.” United States v. Dawson, 64 F.4th 1227, 1237 (11th Cir. 2023) (quoting Yates v. United States, 574 U.S. 528, 537, 135 S. Ct. 1074, 191 L.Ed.2d 64 (2015)). That is, “words grouped in a list should be given related meaning.” Third Nat'l Bank in Nashville v. Impac Ltd., 432 U.S. 312, 322, 97 S. Ct. 2307, 2313, 53 L.Ed.2d 368 (1977) (emphasis added). This principle “avoids ascribing to one word a meaning so broad that it is inconsistent with the company it keeps.” Fischer v. United States, 603 U.S. 480, 487, 144 S. Ct. 2176, 2183–84, 219 L.Ed.2d 911 (2024) (internal quotation marks omitted). Here, the Majority would flip the canon directly on its head: it singles out the phrase “lewd conduct” as meaning something fundamentally different from its associates.

Consider also the ejusdem generis canon, which states that courts “interpret a general or collective term at the end of a list of specific items in light of any common attributes shared by the specific items.” Bissonnette v. LePage Bakeries Park St., LLC, 601 U.S. 246, 252, 144 S. Ct. 905, 911, 218 L.Ed.2d 204 (2024) (cleaned up). The canon applies here, where the phrase “lewd conduct” follows after the statute lists “nudity, sexual conduct, sexual excitement, or specific sexual activities as those terms are defined in s. 847.001.” See Fla. Stat. § 827.11. Indeed, the statute incorporates § 847.001’s “specific sexual activities” before inserting the “lewd conduct” phrase. See id. The statute's enumeration of terms is not perfectly sorted by specificity, but its ordering still lends credence to the idea that “lewd conduct” was intended merely as a catchall phrase, rather than a significant expansion of the statute's scope.

Both the canons of noscitur a sociis and ejusdem generis “track the common sense intuition that [a legislature] would not ordinarily introduce a general term that renders meaningless the specific text that accompanies it.” Fischer, 603 U.S. at 487, 144 S. Ct. at 2184. This approach prevents redundancy—the Majority's desired end—but it does so by reading the text harmoniously, rather than pitting the statute against itself. Here, § 827.11’s enumeration of specific terms—especially the “specific sexual activities” incorporated from § 847.001—serves a useful purpose in limiting the statute's “lewd conduct” phrase to similar, hard-core depictions of sexual conduct.⁵

4. Implications

The Majority's approach creates a perverse incentive for state legislatures to regulate with less specificity—not more. A reasonable legislature wants its statutes to cover all the conduct it considers undesirable. But it also prefers that the public has as much notice as possible as to which forms of conduct are prohibited. That way, people will be more likely to avoid that conduct in the first place, and there will be less need for enforcement by prosecution. The natural solution, then, is for legislatures to enact statutes containing both specific terms and catchall provisions. A “catchall” addresses “known unknowns,” conduct that was not specifically contemplated by legislators but that falls naturally within the statute's ambit. Republic of Iraq v. Beaty, 556 U.S. 848, 860, 129 S. Ct. 2183, 2191, 173 L.Ed.2d 1193 (2009). But in statutory interpretation, “ ‘known unknowns’ should be similar to known knowns,” so any catchall is limited to the kinds of conduct that the legislators did contemplate, which are exemplified by the statute's more specific terms. See Yates, 574 U.S. at 551, 135 S. Ct. at 1089 (Alito, J., concurring).

The Majority's reasoning turns this common legislative technique into a hazardous balancing act: each example the statute provides brings it one step closer to overbreadth. Rather than guess at what a court might find acceptable, a legislature might prefer the safer route of omitting examples altogether. Indeed, that may very well happen here: if this Court ultimately holds that § 827.11 is unconstitutional, Florida's legislature could redraft or amend the statute. This time, it might define an “adult live performance” purely by reference to “lewd conduct,” omitting all the other terms that, according to the Majority, deprived that phrase of any permissible meaning. Could the Majority object to that? Not if it is being faithful to Miller and its progeny. But what would that accomplish? Nothing. The statute would be less specific yet perfectly constitutional when construed in light of Miller. This sort of fruitless exercise shows why our interpretive canons often err on the side of judicial restraint: we are not legislators, and our meddling in legislation risks undesirable results.

Simply put, the question before us is not whether § 827.11 is stylishly and elegantly written. The question is whether the statute violates the Constitution, and our review requires us to engage with the statutory text, as written, in good faith and with the presumption that the legislature did not intend to infringe on constitutional rights. By applying the aforementioned principles and reading the statute harmoniously, we can and should conclude that the statute reaches only speech that would be considered obscene under Miller. I would follow the lead of the Supreme Court in Hamling and interpret “lewd conduct” as “limited to the sort of patently offensive representations or descriptions of that specific hard core sexual conduct given as examples in Miller.” See Hamling, 418 U.S. at 114, 94 S. Ct. at 2906 (cleaned up).

5. Federalism, Certification, and Comity

There is a more fundamental problem with the Majority's decision today: even if every member of this Court were satisfied that the Majority has read the statute perfectly, we would still not have the final say on what § 827.11 actually means. Even the United States Supreme Court lacks that authority. See Gooding v. Wilson, 405 U.S. 518, 520, 92 S. Ct. 1103, 1105, 31 L.Ed.2d 408 (1972). The power to construe state statutes in our federalist system belongs to the state courts—in this case, the Supreme Court of Florida. So, despite the Majority's overbreadth ruling, state courts may still offer a narrowing construction, rendering this whole exercise futile. See Younger v. Harris, 401 U.S. 37, 50, 91 S. Ct. 746, 753, 27 L.Ed.2d 669 (1971).

Of course, if we had no other option but to invalidate the statute now and let the state courts cure it later, the Majority's decision might seem more reasonable. But that is not the case. For decades, in the interests of comity, federal courts have exercised two related doctrines—Pullman abstention and certification—when our cases have called for the resolution of state-law questions. See R.R. Comm'n of Tex. v. Pullman Co., 312 U.S. 496, 61 S. Ct. 643, 85 L.Ed. 971 (1941); Arizonans for Off. Eng. v. Arizona, 520 U.S. 43, 117 S. Ct. 1055, 137 L.Ed.2d 170 (1997). By allowing state courts to rule first on matters of state law, these doctrines can “avoid the waste of a tentative decision as well as the friction of a premature constitutional adjudication.” Pullman, 312 U.S. at 500, 61 S. Ct. at 645.

a. Pullman Abstention

In Pullman, the Texas Railroad Commission issued a regulation requiring that all sleeper cars on trains be supervised by a conductor. Id. at 497–98, 61 S. Ct. at 644. The Pullman Company sued in federal court, arguing that the regulation was “unauthorized by Texas law as well as violative of the Equal Protection, the Due Process and the Commerce Clauses of the Constitution.” Id. Pullman porters—lower-level employees who were predominantly Black, unlike Pullman conductors—intervened as complainants and alleged that the regulation discriminated against them in violation of the Fourteenth Amendment. Id.

The Supreme Court acknowledged the “substantial constitutional issue” raised by the porters, but it observed that the issue could “be avoided if a definitive ruling on the state issue would terminate the controversy.” Id. That is, if the Texas Railroad Commission lacked the statutory authority to issue the regulation, then the regulation would fail without the need to consider its constitutional implications. See id. at 498–99, 61 S. Ct. at 644. And the three-judge panel below had, in fact, concluded that the Commission had exceeded its authority. Id. The only problem, as the Supreme Court observed, was that federal courts lacked the final say on the matter:

Had we or they no choice in the matter but to decide what is the law of the state, we should hesitate long before rejecting their forecast of Texas law. But no matter how seasoned the judgment of the district court may be, it cannot escape being a forecast rather than a determination. The last word on the meaning of ․ the Texas Civil Statutes, and therefore the last word on the statutory authority of the Railroad Commission in this case, belongs neither to us nor to the district court but to the supreme court of Texas.

Id. at 499–500, 61 S. Ct. at 644–45.

The Court reasoned that the district court's holding might ultimately be displaced by a state court ruling, resulting in a premature waste of judicial resources. Id. Therefore, after reviewing “[t]he history of equity jurisdiction,” the Court articulated an abstention doctrine “whereby the federal courts, exercising a wise discretion, restrain their authority because of scrupulous regard for the rightful independence of the state governments and for the smooth working of the federal judiciary.” Id. at 500–01, 61 S. Ct. at 645 (citations and internal quotation marks omitted). Concluding, the Court remanded for the district court to “stay[ ] its hands” while the Texas state-law issue was litigated in state court. Id. at 501–02, 61 S. Ct. at 645–46.

b. Certified Questions

In the decades since Pullman, states have increasingly allowed federal courts to certify questions of state law directly to state courts, rather than requiring litigants to bring parallel state actions. This practice of certification is now preferred to Pullman abstention, which “proved protracted and expensive in practice, for it entailed a full round of litigation in the state court system before any resumption of proceedings in federal court.” Arizonans for Off. Eng., 520 U.S. at 76, 117 S. Ct. at 1073. “Certification procedure, in contrast, allows a federal court faced with a novel state-law question to put the question directly to the State's highest court, reducing the delay, cutting the cost, and increasing the assurance of gaining an authoritative response.” Id.

In Arizonans for Official English, the Supreme Court considered a challenge, brought in federal court under the First and Fourteenth Amendments, to a 1988 amendment to Arizona's Constitution that declared English “the official language of the State of Arizona” and “the language of ․ all government functions and actions.” Id. at 48–50, 117 S. Ct. at 1059–61 (alteration in original). The Court noted that the federal litigants, “proceeding without benefit of the views of the Arizona Supreme Court, expressed diverse opinions on the meaning of the amendment.” Id. The plaintiff, an Arizona employee who handled medical malpractice claims against the state, feared that she could be fired or punished for using Spanish in communications with the public. Id. An official opinion by Arizona's Attorney General, meanwhile, concluded “that government employees remained free to use other [non-English] languages ‘to facilitate the delivery of governmental services.’ ” Id. at 51–52, 117 S. Ct. at 1061.

The district court considered the Arizona Attorney General's opinion but found it to be “simply at odds with [the amendment's] plain language.” Id. at 54–55, 117 S. Ct. at 1062. The court interpreted the amendment as “impos[ing] a sweeping ban on the use of any language other than English by all of Arizona officialdom, with only limited exceptions.” Id. It believed that the text of the amendment, codified at Arizona Constitution Article XXVIII, “left no room for a moderate and restrained interpretation,” and so the court “decline[d] ‘to allow the Arizona courts the initial opportunity to determine the scope of Article XXVIII.’ ” Id. at 55, 117 S. Ct. at 1063. Therefore, the court concluded that Article XXVIII was fatally overbroad under the First and Fourteenth Amendments. Id.

On appeal, a Ninth Circuit panel affirmed. Id. at 61, 117 S. Ct. at 1066. The court then reheard the case en banc and, by a six-to-five vote, upheld the panel's opinion. Id. at 62, 117 S. Ct. at 1066. The en banc court interpreted the amendment as generally prohibiting “the use of any language other than English by all officers and employees of all political subdivisions in Arizona while performing their official duties.” Id. (internal quotation marks omitted). “Because the court found the ‘plain language’ dispositive, it rejected the State Attorney General's limiting construction and declined to certify the matter to the Arizona Supreme Court.” Id. (citations omitted). The court again “condemned the provision as manifestly overbroad.” Id. at 63, 117 S. Ct. at 1067.

In its opinion vacating the Ninth Circuit's judgment, the Supreme Court began by noting that “[f]ederal courts lack competence to rule definitively on the meaning of state legislation.” Id. at 48, 117 S. Ct. at 1059. It observed that the Ninth Circuit had “lost sight of” this fact, as well as Article III standing limitations.⁶ Id. The Court reiterated,

In litigation generally, and in constitutional litigation most prominently, courts in the United States characteristically pause to ask: Is this conflict really necessary? When anticipatory relief is sought in federal court against a state statute, respect for the place of the States in our federal system calls for close consideration of that core question.

Id. at 75, 117 S. Ct. at 1072–73 (footnote omitted).

The Court noted that Arizona's Attorney General had asked both lower courts to pause the litigation and certify a question to the Arizona Supreme Court. Id. It observed that, “[t]hrough certification of novel or unsettled questions of state law for authoritative answers by a State's highest court, a federal court may save ‘time, energy, and resources and hel[p] build a cooperative judicial federalism.’ ” Id. at 77, 117 S. Ct. at 1073 (quoting Lehman Bros. v. Schein, 416 U.S. 386, 391, 94 S. Ct. 1741, 1744, 40 L.Ed.2d 215 (1974)). And it criticized the lower courts for their overconfidence in concluding that the amendment “was not fairly subject to a limiting construction.” Id. at 77, 117 S. Ct. at 1074.

The Court explained that “[f]ederal courts, when confronting a challenge to the constitutionality of a federal statute, follow a ‘cardinal principle’: They ‘will first ascertain whether a construction ․ is fairly possible’ that will contain the statute within constitutional bounds.” Id. at 78, 117 S. Ct. at 1074 (quoting Ashwander v. Tennessee Valley Auth., 297 U.S. 288, 348, 56 S. Ct. 466, 483–84, 80 L.Ed. 688 (1936) (Brandeis, J., concurring)). The Court noted that “[s]tate courts, when interpreting state statutes, are similarly equipped to apply that cardinal principle.” Id. And it added that certification is particularly appropriate when state courts have yet to interpret a new statute:

Warnings against premature adjudication of constitutional questions bear heightened attention when a federal court is asked to invalidate a State's law, for the federal tribunal risks friction-generating error when it endeavors to construe a novel state Act not yet reviewed by the State's highest court. “Speculation by a federal court about the meaning of a state statute in the absence of prior state court adjudication is particularly gratuitous when ․ the state courts stand willing to address questions of state law on certification from a federal court.”

Id. at 79, 117 S. Ct. at 1076 (citations omitted) (quoting Brockett v. Spokane Arcades, Inc., 472 U.S. 491, 510, 105 S. Ct. 2794, 2805, 86 L.Ed.2d 394 (1985) (O'Connor, J., concurring)).

Relevant here, Florida has long authorized its courts to receive state-law questions from our federal courts. See Fla. Stat. § 25.031; see also Clay v. Sun Ins. Off. Ltd., 363 U.S. 207, 212, 80 S. Ct. 1222, 1226, 4 L.Ed.2d 1170 (1960) (praising the “rare foresight” of Florida legislators in providing for certification). The Florida Rules of Appellate Procedure allow a federal appellate court to “certify 1 or more questions of law to the Supreme Court of Florida if the answer is determinative of the cause and there is no controlling precedent of the Supreme Court of Florida.” Fla. R. App. P. 9.150(a).

Just recently, this Court certified a question to the Supreme Court of Florida in another case challenging a statute for vagueness and overbreadth. See Dream Defs. v. Governor of Fla., 119 F.4th 872 (11th Cir. 2024). In Dream Defenders, just as here, the district court entered a preliminary injunction because a Florida statute “criminaliz[ed] constitutionally protected First Amendment activity.” Id. at 874. On appeal, “we certified a question to the Supreme Court of Florida asking it to provide an authoritative interpretation of [the law,] Florida's amended criminal riot statute, Fla. Stat. § 870.01(2) (2021).” Id. We explained that:

Certification in this circumstance allows us to avoid the friction that could arise if we, as a federal court, addressed the merits of the plaintiffs’ pre-enforcement constitutional challenge without first giving the Florida Supreme Court an opportunity to interpret its State's law. See Arizonans for Off. Eng., 520 U.S. at 75, 117 S. Ct. 1055. As we have explained, certification “give[s] the highest court of a state an opportunity to ․ attempt to interpret [state law] in such a way as to make it constitutional.” Pittman v. Cole, 267 F.3d 1269, 1289–90 (11th Cir. 2001) (internal quotation marks omitted). Providing this opportunity is “especially important, because it may well be that the courts of the relevant state are less constrained than is the federal judiciary with respect to statutory interpretation.” Id. at 1290 (alteration adopted) (internal quotation marks omitted). We conclude that certification is consistent with “respect for the place of the States in our federal system.”

Dream Defs. v. Governor of the State of Fla., 57 F.4th 879, 893–94 (11th Cir. 2023), certified question answered sub nom. DeSantis v. Dream Defs., 389 So. 3d 413 (Fla. 2024).

Indeed, the Supreme Court of Florida's interpretation of the statute ultimately resolved the case, obviating the need for us to strike it down on constitutional grounds. Dream Defs., 119 F.4th at 872. That is what the Majority should have done here. Instead, the Majority sidesteps the very tools our system provides—tools designed to respect state authority, foster comity, and avoid unnecessary constitutional rulings. By casting aside those safeguards, today's decision stretches this Court beyond its proper role and departs from the humility and restraint that federal courts owe when state law is in question.

Still, the Majority responds that “neither party requested or briefed the idea of certifying questions to the Florida Supreme Court.” Maj. Op. at ––––. That assertion is as irrelevant as it is revealing. Florida Rule of Appellate Procedure 9.150 says that “On either its own motion or that of a party ․ a United States court of appeals may certify one or more questions of law to the Supreme Court of Florida.” Fla. R. App. P. 9.150(a) (emphasis added). The rule entrusts the decision to us—not just to the litigants. And we have recognized time and again that certification is not a procedural courtesy owed to the parties; it is a structural duty owed to our state counterparts when state law is unsettled and constitutional adjudication looms. See, e.g., Dream Defs., 57 F.4th at 893 (quoting Am. Booksellers Ass'n, 484 U.S. at 393, 108 S. Ct. at 643) (internal quotation marks omitted).

By brushing aside certification simply because the parties did not ask for it, the Majority discards our obligation to exercise restraint, respect federalism, and avoid unnecessarily resolving constitutional questions. It implies that comity and judicial humility are optional—and only if the parties remember to request them. That has never been the law. And it is no answer to say the parties were silent when it is this Court's responsibility to avoid needless constitutional adjudication when a narrowing construction by the state's highest court may well resolve the dispute.

IV. The Age-Variable Standard

The Majority's second basis for facial invalidation is its discomfort with the statute's so-called “age-variable” obscenity standard. According to the Majority, Florida's law is both vague and overbroad because it requires courts to evaluate obscenity in light of a child's age, rather than imposing a fixed cutoff. But that objection misunderstands both the statute and the governing constitutional standards.

Properly framed, the Majority's concern is that the law lacks clarity: that it requires speakers and courts to guess whether material suitable for a fifteen-year-old would offend a twelve-year-old. In its view, that uncertainty chills protected expression and invites arbitrary enforcement. But that framing demands a precision the First Amendment has never required, especially in the context of obscenity.

A. Vagueness and Line Drawing

A statute is unconstitutionally vague if it “fails to provide people of ordinary intelligence a reasonable opportunity to understand what conduct it prohibits” or “authorizes or even encourages arbitrary and discriminatory enforcement.” Hill v. Colorado, 530 U.S. 703, 732, 120 S. Ct. 2480, 2498, 147 L.Ed.2d 597 (2000) (citation omitted). That inquiry is especially demanding when a law restricts speech, because vagueness risks chilling protected expression. See Reno v. ACLU, 521 U.S. 844, 871–72, 117 S. Ct. 2329, 2344–45, 138 L.Ed.2d 874 (1997). But the test is not whether every term is defined with mathematical precision. “[P]erfect clarity and precise guidance have never been required even of regulations that restrict expressive activity.” Ward v. Rock Against Racism, 491 U.S. 781, 794, 109 S. Ct. 2746, 2755, 105 L.Ed.2d 661 (1989) (citation omitted).

That principle governs Miller itself. The Miller test asks whether a work appeals to the prurient interest, whether it is patently offensive, and whether it lacks serious literary, artistic, political, or scientific value—all judged by contemporary community standards. Miller, 413 U.S. at 24, 93 S. Ct. at 2614–15. None of those terms is self-defining. Each requires judgment informed by context. Miller did not impose mathematical lines. It offered a framework that invites discretion, not rigidity.

That is equally true when the audience includes minors. Indeed, Miller cited Ginsberg v. New York with approval. See id. at 19, 27, 93 S. Ct. at 2612, 2617. In Ginsberg, the Court upheld a statute prohibiting the sale of sexually explicit magazines to minors under the age of seventeen, even though the material in question would not have been obscene for adults. Ginsberg, 390 U.S. at 642, 88 S. Ct. at 1282. The Court explained that legislatures could “adjus[t] the definition of obscenity to social realities by permitting the appeal of this type of material to be assessed in term of the sexual interests ․ of ․ minors.” Id. at 638, 88 S. Ct. at 1279 (citations and internal quotation marks omitted). In other words, the Court specifically condoned what it called “variable obscenity.” See id. at 635, 88 S. Ct. at 1278 n.4.

That doctrine has persisted. In Erznoznik v. City of Jacksonville, the Court reaffirmed that “[i]t is well settled that a State or municipality can adopt more stringent controls on communicative materials available to youths than on those available to adults.” 422 U.S. 205, 212, 95 S. Ct. 2268, 2274, 45 L.Ed.2d 125 (1975) (citing Ginsberg, 390 U.S. 629, 88 S. Ct. 1274, 20 L.Ed.2d 195). And in New York v. Ferber, the Court again emphasized that “[i]t is evident beyond the need for elaboration that a State's interest in safeguarding the physical and psychological well-being of a minor is compelling.” 458 U.S. 747, 756, 102 S. Ct. 3348, 3355, 73 L.Ed.2d 1113 (1982) (citation and internal quotation marks omitted). The same point appeared in F.C.C. v. Pacifica Foundation, where the Court cited Ginsberg to confirm that the government may restrict speech that is appropriate for adults but harmful to children. 438 U.S. 726, 749–50, 98 S. Ct. 3026, 3040–41, 57 L.Ed.2d 1073.

In sum, the Supreme Court has long recognized that obscenity standards may vary based on the age of the audience. The Majority resists that principle not by denying it, but by insisting that Florida's law lacks a fixed age cutoff. That resistance is not grounded in constitutional doctrine.

To its credit, the Majority concedes that “the distinction between seventeen and eighteen ․ is not inherently less arbitrary or less vague than the distinction between any other two ages.” Maj. Op. at ––––. But it insists that the “much of our law and culture are oriented around the singular age of majority.” Id. According to the Majority, “[t]his rich social context gives meaning and relative clarity to the line between that which is within minors’ rights to access and that which is ‘adults-only.’ ” Id. That is not a constitutional standard.

The true objection is not to vagueness—it is to flexibility. The Majority worries that tailoring the obscenity test to a child's age introduces too much discretion. But every obscenity statute calls for judgment at the margins. “[T]he mere fact that close cases can be envisioned [does not] rende[r] a statute vague.” United States v. Williams, 553 U.S. 285, 305–06, 128 S. Ct. 1830, 1846, 170 L.Ed.2d 650 (2008).⁷ And Miller itself requires case-by-case context: prurient appeal, patent offensiveness, and value, all gauged by contemporary community standards. That framework demands reasoned application, not rigid rules.

In the end, the Majority sees vagueness where precedent sees discretion. It imagines a constitutional flaw that doctrine does not support—and manufactures a rule that no court has ever imposed.

B. The Statute's Guideposts

Even if the Majority were right to worry about vagueness in the abstract, that concern cannot survive contact with the statute's safeguards. Florida has not invited arbitrary enforcement. It has structured discretion through familiar legal guideposts.

First, the statute does not impose strict liability. A person violates the statute only if he “knowingly admit[s] a child to an adult live performance.” Fla. Stat. § 827.11(3). And “knowingly” is defined to require “general knowledge of, reason to know, or a belief or ground for belief which warrants further inspection or inquiry” into both the nature of the performance and the child's age. Id. § 827.11(1)(b). That definition incorporates a culpable mental state and protects those who act reasonably and in good faith.

Indeed, the statute's requirements mirrors that of the statute in Ginsberg. As the Supreme Court there said, “[t]he constitutional requirement of scienter, in the sense of knowledge of the contents of material, rests on the necessity to avoid the hazard of self-censorship of constitutionally protected material and to compensate for the ambiguities inherent in the definition of obscenity.” Ginsberg, 390 U.S. at 644, 88 S. Ct. at 1283 (internal quotation marks omitted) (emphasis added). So too here.

Second, the definition of “adult live performance” is tethered directly to the Miller framework. A covered performance must appeal to a prurient interest, be patently offensive to prevailing standards, and lack serious value—all “for the age of the child present.” Id. § 827.11(1)(a). These are familiar, judicially tested standards that apply across federal and state obscenity laws. And they apply here with a single adjustment: they are measured in light of the actual child admitted—not an abstract viewer or generalized audience.

Third, the statute requires that the performance be judged “as a whole,” and applies only if all three elements of Miller are met. Id. That structure constrains enforcement to performances that satisfy the constitutional definition of obscenity for the age group in question.

The Majority finds none of this sufficient. But its preferred alternative—a rigid age threshold—would not resolve the uncertainty it identifies. It would merely relocate it to a different point on the spectrum, imposing a fixed line that lacks any constitutional foundation and would likely chill more speech than it protects.

C. Certification and Constitutional Avoidance

Even if the Florida statute presented some ambiguity, that ambiguity concerns the meaning of a state-law term. And as discussed above, the proper course in such cases certification to the state courts. That is especially true here, where Florida's judiciary stands ready to clarify the statute's scope and application.

The Majority's own opinion illustrates the need for that approach in relation to the variable age standard. The Majority poses a series of rhetorical questions: Whether speakers must tailor their message “year-by-year, ․ month-by-month, ․ week-by-week, ․ or day-by-day,” see Maj. Op. at ––––, or how “[a proprietor is] to determine with confidence that a child is old enough to attend [a show,]” see Maj. Op. at ––––. These are precisely the types of questions state courts are best positioned to resolve. But it appears the Majority instead deployed these hypotheticals not to clarify the law, but to manufacture constitutional error. That move disregards well-established tools of restraint. When state law is reasonably susceptible to a narrowing construction, federal courts must allow the state's judiciary to supply it. Certification, not invalidation, was the correct response.

V. Scope of the Injunction

Even if the Majority were correct on the merits (it is not), it still errs in upholding the scope of the District Court's injunction. Less than a year ago, the Supreme Court admonished this Court that “[e]ven in the First Amendment context, facial challenges are disfavored, and neither parties nor courts can disregard the requisite inquiry into how a law works in all of its applications.” Moody, 603 U.S. at 744, 144 S. Ct. at 2409. Our opinion was reversed because, in the words of Justice Jackson, “The Eleventh Circuit failed to appreciate the nature of this [facial] challenge.” Id. at 748, 144 S. Ct. at 2411 (Jackson, J., concurring). That instruction could hardly have been clearer. Yet today, the Majority sidesteps it.⁸

As Moody explains, facial challenges carry a notoriously heavy burden. To succeed, the plaintiff must show that “a substantial number of [the law's] applications are unconstitutional, judged in relation to the statute's plainly legitimate sweep.” Id. at 723, 144 S. Ct. at 2398 (quoting Americans for Prosperity Found., 594 U.S. at 615, 141 S. Ct. at 2387 (internal quotation marks omitted)). A court must first determine what the law regulates: “What activities, by what actors, do the laws prohibit or otherwise regulate?” Id. at 724, 144 S. Ct. at 2398. Then the court must identify the law's “full range of applications—the constitutionally impermissible and permissible both—and compare the two sets.” Id. at 726, 144 S. Ct. at 2398. Only if the unconstitutional applications are both substantial and disproportionate may the court preclude enforcement of the law on its face. Id. at 743–44, 144 S. Ct. at 2408–09.

The Moody framework reflects the judiciary's structural limits. Facial invalidation is “strong medicine” because it risks sweeping away constitutionally valid applications not before the court. See Williams, 553 U.S. at 293, 128 S. Ct. at 1838 (citations and internal quotation marks omitted). As Justice Alito explained,

[Facial] challenges are strongly disfavored. See Washington State Grange, 552 U. S. at 452 [128 S.Ct. 1184]. They often raise the risk of “ ‘premature interpretation[n] of statutes’ on the basis of factually barebones records.” Sabri v. United States, 541 U. S. 600, 609 [124 S.Ct. 1941, 158 L.Ed.2d 891] (2004). They clash with the principle that courts should neither “ ‘anticipate a question of constitutional law in advance of the necessity of deciding it’ ” nor “ ‘formulate a rule of constitutional law broader than is required by the precise facts to which it is to be applied.’ ” Ashwander v. TVA, 297 U. S. 288, 346–347 [56 S.Ct. 466, 80 L.Ed. 688] (1936) (Brandeis, J., concurring). And they “threaten to short circuit the democratic process by preventing laws embodying the will of the people from being implemented in a manner consistent with the Constitution.” Washington State Grange, 552 U. S. at 451 [128 S.Ct. 1184].

Facial challenges also strain the limits of the federal courts’ constitutional authority to decide only actual “Cases” and “Controversies.” Art. III, § 2. “[L]itigants typically lack standing to assert the constitutional rights of third parties.” United States v. Hansen, 599 U. S. 762, 769 [143 S.Ct. 1932, 216 L.Ed.2d 692] (2023). But when a court holds that a law cannot be enforced against anyone under any circumstances, it effectively grants relief with respect to unknown parties in disputes that have not yet materialized.

For these reasons, we have insisted that parties mounting facial attacks satisfy demanding requirements

Moody, 603 U.S. at 777–78, 144 S. Ct. at 2428–29 (Alito, J., concurring in the judgment).

In other words, courts always favor as-applied relief, not across-the-board invalidation. See, e.g., id. at 744, 144 S. Ct. at 2409 (“Even in the First Amendment context, facial challenges are disfavored.”). Our own Circuit decisions express the same sentiment. Just last month we applied Moody to highlight that facial challenges are “a demanding standard, and for good reason.” See Henry v. Sheriff of Tuscaloosa Cnty., Alabama, No. 24-10139, ––– F.4th ––––, ––––, 2025 WL 1177671, at *43 (11th Cir. Apr. 23, 2025) (per curiam). We recognized that “the Supreme Court has made facial challenges hard to win.” Id.

The District Court here skipped the entire facial challenge analysis. It declared the statute unconstitutional as to Hamburger Mary's and later said that it thought “the injunction necessarily must extend to protect all Floridians.” See HM Florida-ORL, 2023 WL 11257409, at *4 (emphasis added). In doing so, the District Court ignored the presumption against facial relief. Rather than examine the statute's full scope or its various permissible applications, the Court focused narrowly on drag performances and raised a singular concern:

[T]he Act's focus on “prosthetic or imitation genitals or breasts” raises a host of other concerns not simply answered—what are the implications for cancer survivors with prosthetic genitals or breasts?

HM Florida-ORL, 679 F. Supp. 3d at 1344.

That was the full extent of the District Court's analysis of the law's “full range of applications—the constitutionally impermissible and permissible.” See Moody, 603 U.S. at 726, 144 S. Ct. at 2398. Needless to say, under Moody that is nowhere near enough.

The Majority compounds this error. It does not map the statute's reach. It does not weigh how many of its applications are constitutional. It does not explain why narrower, as-applied relief would fall short. Instead, it affirms a sweeping, statewide invalidation—the exact shortcut Moody forbids.

Indeed, the Majority concedes that it has not assessed the statute's permissible scope, insisting that the law is too vague and overbroad to map fully. But that is no excuse—it is the very reason Moody’s inquiry is required. Courts cannot strike down a law on its face without first attempting to determine how it operates in practice.⁹ If the statute truly fails that test, the constitutional defect must emerge through the process Moody mandates—not in place of it. To skip that analysis because the law might be vague in some applications flips the burden and invites the very kind of premature invalidation Moody prohibits.

To be sure, the District Court fashioned its injunction before Moody was decided. But this Court did not. Moody is now binding, and its framework is not optional. At a minimum, we should remand this case with instructions to apply Moody’s framework: examine the law's full set of applications, assess which are constitutional and which are not, and determine whether the unconstitutional ones are sufficiently substantial to justify facial relief. Anything less ignores what the Supreme Court and Article III require.

VI. Conclusion

The judicial power carries limits. We must read statutes narrowly, not manufacture conflict. We must respect the balance between state and federal authority. And we must confine our role to deciding the dispute before us, not pronouncing broad constitutional rules unnecessarily.

Today, the Majority disregards those limits. It passes over the tools meant to harmonize state and federal law, overlooks the state courts’ interpretive role, and leaps to constitutional invalidation prematurely. That overreach upends the careful balance our system preserves—and takes this Court beyond the authority Article III grants.

I respectfully dissent.

FOOTNOTES

1.   In its complaints about drag revues featuring Jimbo, FDBPR alleged that the defendant drag venues featured a variety of problematic displays. FDBPR cited Jimbo's performance as presenting “sexual conduct, simulated sexual activity, and lewd, vulgar, and indecent displays, including ․ graphic depictions of childbirth.” FDBPR then charged defendants with violating a number of Florida statutes, ranging from prohibitions of “lewd or lascivious exhibition,” Fla. Stat. § 800.04(7), and vulgar or indecent “exposure of sexual organs,” Fla. Stat. § 800.03, to bans on “obscene show[s] ․ by a live person before an audience,” Fla. Stat. § 847.011(4), distribution of “obscene material to a minor,” Fla. Stat. § 847.0133, and “public nuisances” or “disorderly conduct.” Fla. Stat. §§ 823.05, 877.03.

2.   To be clear, we aren't looking to the legislative history to interpret the Act's meaning. See Polycarpe v. E&S Landscaping Serv., Inc., 616 F.3d 1217, 1224 (11th Cir. 2010). But considering statements by legislators and agencies is appropriate when we evaluate the threat of an enforcement action. After all, laws can chill a person's speech either because he fears actually losing in court or because “he may simply be concerned about the expense of becoming entangled in the legal system.” Counterman v. Colorado, 600 U.S. 66, 75, 143 S.Ct. 2106, 216 L.Ed.2d 775 (2023). Even if legislative history does not shed light on the meaning of a statute's text and what speech it actually prohibits, legislative history may indicate who the enacting coalition would like to see prosecuted under the law. So legislative history that suggests a speaker's risk of prosecution may support his standing to challenge a statute even if that history would not guide our interpretation of the law.

3.   Indeed, amicus American First Legal Foundation urges us to hold that if we were to vacate the preliminary injunction, Hamburger Mary's could be retroactively held liable for violations of the Act even while the injunction was in effect. Because our holding makes that issue a hypothetical one, we do not consider it here.

4.   More precisely, Justice Stewart was discussing the attempt to define the category of “hard-core pornography” that could constitutionally be regulated under the rubric of obscenity. Jacobellis, 378 U.S. at 197, 84 S.Ct. 1676 (Stewart, J., concurring).

5.   The New York statute modified a pre-Miller obscenity test that deemed content obscene only when it was “completely without redeeming value,” rather than the less stringent “without serious value” prong in Miller. See Ginsberg, 390 U.S. at 635, 88 S.Ct. 1274; Roth v. United States, 354 U.S. 476, 485, 77 S.Ct. 1304, 1 L.Ed.2d 1498 (1957). By contrast, here, the Act bases its age-variable test on Miller. Shortly after Miller, the Court noted that it had yet to decide “what effect Miller will have on the Ginsberg formulation.” Erznoznik v. City of Jacksonville, 422 U.S. 205, 213 n.10, 95 S.Ct. 2268, 45 L.Ed.2d 125 (1975). It still has not done so. But we don't think updating the Ginsberg test of obscenity for minors to reflect the new standard for adult obscenity from Miller affects Ginsberg’s reasoning or conclusion.

6.   And even if we certified interpretation of the “lewd conduct” provision, that would not resolve the distinct constitutional concerns with the Act's age-variable obscenity standard.

7.   We emphasize that our holding does not doom the many state and federal statutes that use the term “lewd.” As in Hamling, speech regulations targeting “lewd” speech may be interpreted to apply to only the specific types of “hard core” content Miller described. Miller, 413 U.S. at 27, 93 S.Ct. 2607. But when, as here, a statute lists that content in great detail and then adds a vague, broad descriptor, we see no interpretive route to keep the statute within Miller’s substantive and procedural limits.

8.   America First urges us to use a “lopsided ratio” test, purportedly announced in Hansen, 599 U.S. at 770, 143 S.Ct. 1932, to determine whether the Act's unconstitutional applications so outnumber its constitutional ones that the Act is overbroad. But whatever test Hansen uses, we don't think it differs much, if at all, from the longstanding requirement that overbreadth “not only be real, but substantial as well, judged in relation to the statute's plainly legitimate sweep.” Broadrick, 413 U.S. at 615, 93 S.Ct. 2908. After all, Hansen invokes the “substantial” test in its opening paragraph. Hansen, 599 U.S. at 766, 143 S.Ct. 1932 (quoting Broadrick, 413 U.S. at 615, 93 S.Ct. 2908). And in a post-Hansen First Amendment case, the Court required the plaintiff to show that “the law at issue ․ ‘prohibits a substantial amount of protected speech relative to its plainly legitimate sweep.’ ” Moody, 603 U.S. at 744, 144 S.Ct. 2383 (2024) (quoting Hansen, 599 U.S. at 770, 143 S.Ct. 1932).

9.   We do not address the Act's provision concerning “lewd exposure of prosthetic or imitation genitals or breasts.”

10.   The Act's requirement to judge performances’ appropriateness by statewide community standards, rather than local ones, creates additional difficulties. Miller required that obscenity be judged by “community standards,” as “[i]t is neither realistic nor constitutionally sound to read the First Amendment as requiring that the people of Maine or Mississippi accept public depiction of conduct found tolerable in Las Vegas, or New York City.” Miller, 413 U.S. at 32, 93 S.Ct. 2607. And to be sure, Miller upheld the use of statewide community standards in the underlying prosecution, which dealt with a man who had mailed brochures advertising raunchy books for sale. Id. at 17–18, 32, 93 S.Ct. 2607.More recently, though, the Court considered a challenge to a statute applying “community standards” (with no clarification of local, state, or national ones) when assessing whether Internet content was “harmful to minors.” Ashcroft v. ACLU, 535 U.S. 564, 122 S.Ct. 1700, 152 L.Ed.2d 771 (2002). The Court held that the potential for local standards to restrict speech on an inherently national medium did not, by itself, make the statute facially overbroad. Id. at 585, 122 S.Ct. 1700. But five Justices recognized to some degree how “the national variation in community standards constitutes a particular burden on Internet speech.” See id. at 597, 122 S.Ct. 1700 (Kennedy, J., concurring in the judgment, with Souter and Ginsberg, JJ.); id. at 589–91, 122 S.Ct. 1700 (Breyer, J., concurring in part); id. at 586, 122 S.Ct. 1700 (O'Connor, J., concurring in part). If the law in Ashcroft burdened national speech by threatening to impose local standards, the Act burdens speech by doing the opposite. “Live performances,” in contrast to the online speech in Ashcroft and the mailed brochures in Miller, are an inherently local medium. (Even if the Act's coverage of “presentation[s] in front of a live audience” includes online performances, FDBPR's enforcement actions are tied to brick-and-mortar establishments.) The audience for a live performance is disproportionately local—perhaps the only people likely to drive from Tallahassee to see a drag show in Miami are FDBPR inspectors. Yet the Act requires people to calibrate their speech, no matter how locally targeted, according to the tastes of those living hundreds of miles away.In Ashcroft, even the three Justices who felt the least compelled to apply a national standard for Internet speech said so because they thought Internet speakers could just limit access from particular jurisdictions. 535 U.S. at 583, 122 S.Ct. 1700 (opinion of Thomas, J.). “If a publisher wishes for its material to be judged only by the standards of particular communities, then it need only take the simple step of utilizing a medium that enables it to target the release of its material into those communities.” Id. But here, the Act regulates a quintessentially local medium—performance before a “live audience”—yet would judge it by a geographically broad standard. This is especially puzzling when compared with Florida's more general law against giving children material “harmful to minors” (in a variety of mediums). See Fla. Stat. § 847.012. That law does not specify that content is to be evaluated by statewide community standards but just “standards in the adult community as a whole.” Id. § 847.001(7)(b). The potential mismatch of using statewide standards for local speech while using local standards for broadly distributed speech only adds to our concerns with the Act.

11.   Florida's pre-existing “harmful to minors” statute, like the Act, does not adjust the “appeals to a prurient, shameful, or morbid interest” prong based on age. See Fla. Stat. §§ 827.11(1)(a)1; 847.001(7)(a).

12.   The age-variable standard distinguishes Florida's law from two recently challenged statutes elsewhere. The Sixth Circuit addressed a Tennessee law regulating “adult-oriented performances that are harmful to minors.” See Friends of George's, Inc. v. Mulroy, 108 F.4th 431, 433 (6th Cir. 2024) (quoting Tenn. Code. Ann. § 7-51-1401(3)(A)). That statute used a general “harmful to minors” standard, which Tennessee's high court had previously interpreted the same way we did Webb: obscene even for seventeen-year-olds. Id. at 434 (citing Davis-Kidd, 866 S.W.2d at 522–23, 528). Because the Friends of George's plaintiffs did not allege that their performances (which included drag shows) lacked serious value even for seventeen-year-olds, they were denied standing. Id. at 439. In contrast to the Tennessee law, the Act's age-variable standard broadens its scope and accordingly gives standing to a wider range of parties.The Act also differs from Texas's law requiring online purveyors of “sexual material harmful to minors” to verify their users are over eighteen. Tex. Civ. Prac. & Rem. Code § 129B. The Supreme Court recently heard arguments in a case challenging the law's constitutionality. See generally Free Speech Coalition, Inc. v. Paxton, No. 23-1122 (docketed Apr. 16, 2024). But Texas's law, like Tennessee's, uses a general “harmful to minors” standard. So the Court's decision about the Texas statute will not necessarily affect our decision here.

13.   Indeed, the statute Ginsberg upheld barred sales of “harmful to minors” materials to children “under 17 years of age.” Ginsberg, 390 U.S. at 631, 88 S.Ct. 1274.

14.   While states have different ages of majority, each state of course has just one.

15.   We need not resolve today all the complexities of analyzing the overbreadth of statutes whose scope partially overlaps with other law. But we note that counting already-prohibited speech within the “legitimate sweep” of a statute for overbreadth purposes would create obvious problems. Doing so would incentivize legislators to draft statutes redundantly to inflate the statute's “legitimate sweep” and thus dilute the law's unconstitutional applications in the overbreadth analysis. See R. George White, The Problems of Overbreadth and What to Do About Them, 60 Hous. L. Rev. 1115, 1141 (2023).

16.   Of course, nothing we say today eliminates the substantive limits on what speech may be deemed obscene as to minors. It remains true that “[s]peech that is neither obscene as to youths nor subject to some other legitimate proscription cannot be suppressed solely to protect the young from ideas or images that a legislative body thinks unsuitable for them.” Erznoznik, 422 U.S. at 213–14, 95 S.Ct. 2268. Neither statutory specificity nor the other prongs of the Miller test are a blank check for suppression of disfavored speech. See Stevens, 559 U.S. at 479–80, 130 S.Ct. 1577.

17.   The dissent argues that the Supreme Court's decision in Moody counsels against consideration of a facial challenge here. Dissenting Op. at ––––. But our analysis shows that “a substantial number of [the Act's] applications are unconstitutional, judged in relation to the statute's plainly legitimate sweep.” Moody, 603 U.S. at 723, 144 S.Ct. 2383. We think the “lewd conduct” provision has in essence no “legitimate sweep”: the rest of the Act's prohibitions cover the bulk of content that can be deemed obscene under Miller, if not all of it, and in any event they cover everything that escapes the specific-definition requirement for obscenity regulation. As for the age-variable requirement, our precedent in Webb, Florida's recent history of enforcing similar statutes, the “harmful to minors” statutes already on Florida's books, Hamburger Mary's expressed fears of enforcement, and the state's inability to articulate any substantive guideposts for interpreting the age-variable standard convince us that the Act has little, if any, legitimate sweep, while the Act in practice would be likely to directly and indirectly chill much more speech. Cf. Webb, 919 F.2d at 1502–09 (accounting for direct and indirect effects of speech regulation in conducting facial overbreadth analysis).This Act also simply does not present the kind of uncertainty Moody confronted. The Court there noted that “[c]laims of facial invalidity often rest on speculation about the law's coverage and its future enforcement.” Moody, 603 U.S. at 723, 144 S.Ct. 2383 (cleaned up). (Of course, this speculation just poses risks to facial adjudication—not an absolute bar.) The parties in Moody focused on how the challenged laws would apply in certain contexts: primarily, in social-media platforms’ content moderation of public or semi-public posts, like “Facebook's News Feed.” Id. at 724, 144 S.Ct. 2383. “But argument ․ revealed that the laws might apply to, and differently affect, other kinds of websites and apps” with different technological mechanisms and First Amendment implications. Id. at 718, 144 S.Ct. 2383. And especially “when confronted with the application of a constitutional requirement to new technology,” the Court wished to “proceed with caution.” Id. at 796, 144 S.Ct. 2383 (Alito, J., concurring in the judgment).This case could not be more different. The Act deals with one of the oldest technologies: “live performance.” Fla. Stat. § 827.11(1)(a). No one has suggested that the Act might apply in some context significantly different from the one presented as the “heartland” of the law, let alone one that raises meaningfully different First Amendment questions. Moody, 603 U.S. at 724, 144 S.Ct. 2383. Some “live performance[s]” may take place at restaurants or bars, others in more traditional theaters or even outdoors, but the basic speech dynamics are the same. In contrast, the laws in Moody potentially reached speech settings nearly as disparate as “direct messaging,” “events management,” “email,” “online marketplace ․ customer reviews,” and “payment service[s].” Id. Moody warns against considering only apples when reviewing a law that regulates oranges and pears, too, but the Act regulates only apples.

1.   Judges Jordan, Rosenbaum, and Brasher comprised the panel. The majority order was issued “By the Court.”

2.   The Supreme Court denied the application for a stay on November 16, 2023. One month earlier, on September 29, 2023, it had granted certiorari in Moody v. NetChoice. As discussed below, the Court's decision in Moody, issued on July 1, 2024, clarified the standards for evaluating facial challenges. The Court may have declined to take up the procedural question because, as explained below, Moody’s framework compels reversal of the broad injunction here. See infra Part V.

3.   The Court further noted that “ ‘[l]ewd’ and ‘lascivious’ are words in common use, and the definitions indicate with reasonable certainty the character of acts and conduct which the Legislature intended to prohibit and punish, so that a person of ordinary understanding may know what conduct on his part is condemned.” Chesebrough, 255 So. 2d at 677.

4.   The Supreme Court explicitly rejected the idea “that the canon [of constitutional avoidance] has less force in the context of an overbreadth challenge.” Hansen, 599 U.S. at 781, 143 S. Ct. at 1946 n.3.

5.   Our own federal statutes are rife with examples of catchall phrases that are influenced by their preceding terms. Consider obstruction statutes such as 18 U.S.C. § 1519. That statute penalizes anyone who “knowingly alters, destroys, mutilates, conceals, covers up, falsifies, or makes a false entry in any record, document, or tangible object” to obstruct a federal investigation. 18 U.S.C. § 1519. In Yates v. United States, the Supreme Court was asked whether a fish was a “tangible object” under the statute. Yates, 574 U.S. at 531–32, 135 S. Ct. at 1078–79. Answering that question in the negative, a majority of the Court relied on canons such as noscitur a sociis and ejusdem generis to confine the term's meaning to “objects one can use to record or preserve information, not all objects in the physical world.” See id. at 536, 135 S. Ct. at 1081; id. at 549–51, 135 S. Ct. at 1089 (Alito, J., concurring). The lesson from this and similar cases is that we should err on the side of interpreting statutes narrowly, rather than more broadly, based on the statutory context. See also Fischer, 603 U.S. at 488, 144 S. Ct. at 2184 (“The idea is simply that a general phrase can be given a more focused meaning by the terms linked to it.”).

6.   The case also involved some uncertainty about the standing of various parties to litigate the dispute on appeal. Although the district court held that Article XXVIII was unconstitutional, it denied the plaintiff's request for an injunction because “she ha[d] not established an enforcement threat sufficient to warrant [such] relief.” Arizonans for Off. Eng., 520 U.S. at 55, 117 S. Ct. at 1063. After judgment, two parties, the Arizonans for Official English Committee (“AOE”) and its chairman, moved to intervene as defendants under Federal Rule of Civil Procedure 24. Id. at 56, 117 S. Ct. at 1063. As proponents of the amendment, they wanted to defend its constitutionality on appeal. Id. Meanwhile, the original plaintiff, Maria-Kelly Yniguez, expressed reluctance about any further appeal, believing that she had effectively won the suit and needed no other relief. Id. The district court denied the intervention motion for lack of standing, but the Ninth Circuit disagreed. Id. at 57, 117 S. Ct. at 1064. It held that “AOE, as principal sponsor of the ballot initiative, qualified to defend Article XXVIII on appeal.” Id. at 58, 117 S. Ct. at 1064. And it held that the case was not moot, even though Yniguez resigned from state employment during the appeal, because she may be entitled to nominal damages. Id. at 59–60, 117 S. Ct. at 1065. On certiorari review, the Supreme Court expressed “grave doubts” about whether AOE had standing to appeal. Id. at 66, 117 S. Ct. at 1068. And the Court viewed Yniguez's case as having been mooted by her resignation. Id. at 72, 117 S. Ct. at 1071. Although the mooting event occurred after the district court's judgment, the complex posture of the case—“and the federalism concern [the Court] next consider[ed]”—led it to “conclude that vacatur down the line” was appropriate. Id. at 74–75, 117 S. Ct. at 1072.

7.   Williams was a case from the Eleventh Circuit. See United States v. Williams, 444 F.3d 1286 (11th Cir. 2006). The full quote from the Supreme Court was an admonishment to our Circuit: “[T]he Eleventh Circuit's error is more fundamental than merely its selection of unproblematic hypotheticals. Its basic mistake lies in the belief that the mere fact that close cases can be envisioned renders a statute vague. That is not so. Close cases can be imagined under virtually any statute.” Williams, 553 U.S. at 305–06, 128 S. Ct. at 1846.

8.   Judge Brasher has already flagged these concerns—albeit without the benefit of Moody, which now makes the point unmistakable. See HM Florida-Orl, 2023 WL 6785071, at *4–6 (Brasher, J., dissenting from the order denying motion for a partial stay).

9.   Dealing with a facial challenge under Moody in this context “strikes me as a daunting, if not impossible, task.” See Moody, 603 U.S. at 745, 144 S. Ct. at 2409 (Barrett, J., concurring). It “likely forces [the] court to bite off more than it can chew,” having to dream up all applications of a law. See id. at 747, 144 S. Ct. at 2411. But that is a feature, not a bug. As I have explained, “facial challenges are [strongly] disfavored.” See id. at 744, 144 S. Ct. at 2409. Requiring courts to examine every imaginable application—many of them involving parties not before the court—ensures that we do not strike down statutes merely because they might be problematic in a few hypothetical scenarios. That makes universal invalidation hard—painfully hard—because it bears repeating that the judiciary's role is to decide “Cases” and “Controversies,” not to serve as a roving constitutional council.

ROSENBAUM, Circuit Judge: