`
`
`
`In the
`United States Court of Appeals
`For the Eleventh Circuit
`
`____________________
`
`No. 19-10604
`
`____________________
`
`
`
`
`
`
`
`ROBERT W. OTTO,
`
`JULIE H. HAMILTON,
`
`versus
`
`CITY OF BOCA RATON, FLORIDA,
`
`COUNTY OF PALM BEACH, FL,
`
`
`
` Plaintiffs-Appellants,
`
` Defendants-Appellees.
`
`
`
`____________________
`
`Appeal from the United States District Court
`
`for the Southern District of Florida
`
`D.C. Docket No. 9:18-cv-80771-RLR
`
`____________________
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 2 of 110
`
`2
`
`
`
`19-10604
`
`
`Before WILLIAM PRYOR, Chief
`
`Judge, WILSON,
`
`JORDAN,
`
`ROSENBAUM, JILL PRYOR, NEWSOM, BRANCH, GRANT, LUCK, LAGOA,
`
`and BRASHER, Circuit Judges.
`
`BY THE COURT:
`
`A petition for rehearing having been filed and a member of
`
`this Court in active service having requested a poll on whether this
`
`case should be reheard by the Court sitting en banc, and a majority
`
`of the judges in active service on this Court having voted against
`
`granting rehearing en banc, it is ORDERED that this case will not
`
`be reheard en banc.
`
`
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 3 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`1
`
`GRANT, Circuit Judge, joined by BRANCH and LAGOA, Circuit
`
`Judges, concurring in the denial of rehearing en banc:
`
`First Amendment jurisprudence is straightforward in at least
`
`one respect: it “requires that content-based speech restrictions
`
`satisfy strict scrutiny. And unless restrictions meet that demanding
`
`standard, whether the speech they target should be tolerated is
`
`simply not a question that we are allowed to consider, or a choice
`
`that we are allowed to make.” Otto v. City of Boca Raton, 981 F.3d
`
`854, 870 (11th Cir. 2020) (quotation and citations omitted). The
`
`city and county ordinances in this case, which prohibit talk therapy
`
`on a particular—and particularly controversial—subject, are no
`
`exception to this rule.
`
`The challenged ordinances “prohibit therapists
`
`from
`
`engaging in counseling or any therapy with a goal of changing a
`
`minor’s sexual orientation, reducing a minor’s sexual or romantic
`
`attractions (at least to others of the same gender or sex), or
`
`changing a minor’s gender identity or expression—though support
`
`and assistance to a person undergoing gender transition is
`
`specifically permitted.” Id. at 859. The perspective enforced by
`
`these local policies is extremely popular in many communities.
`
`And the speech barred by these ordinances is rejected by many as
`
`wrong, and even dangerous. But the First Amendment applies
`
`even to—especially to—speech that is widely unpopular.
`
`The panel opinion thoroughly explains why a fair-minded
`
`and neutral application of longstanding First Amendment law
`
`dooms the ordinances. We write separately here to address our
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 4 of 110
`
`2
`
` GRANT, J., Concurring
`
`19-10604
`
`colleagues’ dissenting opinions and to reiterate the importance of
`
`the First Amendment protections at stake. Today’s dissenters
`
`decry the result of the panel decision—namely, that speech they
`
`consider harmful is (or may be) constitutionally protected. But to
`
`reach their preferred outcomes, they ask us to ignore settled First
`
`Amendment law.
`
`Consider our well-established standard of review for First
`
`Amendment cases.
`
` When reviewing constitutional
`
`facts
`
`underlying possible violations of the freedom of speech, we apply
`
`de novo, or plenary, review. ACLU of Florida, Inc. v. Miami-Dade
`
`Cnty. Sch. Bd., 557 F.3d 1177, 1203 (11th Cir. 2009); see also Bose
`
`Corp. v. Consumers Union of U.S., Inc., 466 U.S. 485, 501 n.17, 505-
`
`–06 & 506 n.24 (1984). Judge Jordan correctly applied this standard
`
`when writing for this Court in Wollschlaeger v. Governor of
`
`Florida, an en banc case in which we held that the government
`
`could not block doctors from speaking to their patients about guns.
`
`See 848 F.3d 1293, 1301 (11th Cir. 2017) (en banc). Remarkably, he
`
`now attacks that standard, emphasizing that we ordinarily review
`
`a district court’s “factual findings for clear error” in an appeal from
`
`the grant or denial of a preliminary injunction. Indep. Party of
`
`Florida. v. Sec’y, Florida, 967 F.3d 1277, 1280 (11th Cir. 2020).
`
`Jordan Dissent at 1.
`
`That
`
`is true—but “First Amendment
`
`issues are not
`
`ordinary.” ACLU of Florida, 557 F.3d at 1203. It has long been the
`
`rule that when we consider a preliminary injunction implicating
`
`the freedom of speech, “our review of the district court’s findings
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 5 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`3
`
`of ‘constitutional facts,’ as distinguished from ordinary historical
`
`facts, is de novo.” Id. (quoting CAMP Legal Def. Fund, Inc. v. City
`
`of Atlanta, 451 F.3d 1257, 1268 (11th Cir.2006)). Historical facts are
`
`the straightforward findings of the circumstances surrounding a
`
`case—here, for example, the dates on which the ordinances were
`
`passed. Constitutional facts, in contrast, are the “core facts” that
`
`determine whether a First Amendment violation has occurred. Id.
`
`at 1205.
`
`Because “the reaches of the First Amendment are ultimately
`
`defined by the facts it is held to embrace,” appellate courts must
`
`ourselves decide “whether a given course of conduct falls on the
`
`near or far side of the line of constitutional protection.” Hurley v.
`
`Irish-Am. Gay, Lesbian & Bisexual Grp. of Boston, Inc., 515 U.S.
`
`557, 567 (1995). Here, the question of whether the ordinances
`
`regulate speech or conduct—as Judge Jordan puts it, whether the
`
`therapy is “just talk”—goes well beyond historical fact. See Jordan
`
`Dissent at 13. To defer on a factual issue so intertwined with the
`
`legal questions at stake would be to implicitly delegate legal
`
`judgment to the district court as well.
`
`We cannot duck controversial issues by evading the
`
`standard of review for constitutional facts. The panel, as our
`
`precedents require, applied the proper standard: “plenary review.”
`
`Wollschlaeger, 848 F.3d at 1301. And we are puzzled that Judge
`
`Jordan objects to applying the same standard he used in
`
`Wollschlaeger.
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 6 of 110
`
`4
`
` GRANT, J., Concurring
`
`19-10604
`
`The next dissent also ignores the law of this Circuit and the
`
`Supreme Court. Citing dozens of interest group publications—
`
`none of which are in the record—Judge Rosenbaum criticizes the
`
`panel majority’s “uninformed take on talk therapy.” Rosenbaum
`
`Dissent at 2; see id. at 3–7, 71–75 (citing publications). But we are
`
`not charged with performing our own internet investigation on the
`
`questions that come before us. In fact, doing so is out of bounds.
`
`See, e.g., Turner v. Burnside, 541 F.3d 1077, 1086 (11th Cir. 2008)
`
`(“We do not consider facts outside the record.”). Our role is to
`
`independently review the record, not to develop it further.
`
`Our role is also to apply the precedents that bind us, and
`
`Judge Rosenbaum’s attempts to justify the ordinances only reveal
`
`that it is impossible to do so under existing law. To start, the
`
`dissent recognizes that ordinances like these are “necessarily
`
`content-based and would not survive the general presumption
`
`against content-based regulations and strict scrutiny.” Rosenbaum
`
`Dissent at 24. Exactly. As the panel opinion explains, the studies
`
`offered to the district court in support of the regulations contained
`
`“ambiguous proof” and “equivocal conclusions.” Otto, 981 F.3d at
`
`868–69 (quoting Brown v. Ent. Merchs. Ass’n, 564 U.S. 786, 800
`
`(2011)). That is not enough to meet the “demanding standard” that
`
`strict scrutiny requires. Id. at 868 (quoting Brown, 564 U.S. at 799).
`
`Indeed, the dissent also concedes that—even considering the
`
`dramatic number of interest group publications and press releases
`
`that it identifies—these specific regulations cannot survive strict
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 7 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`5
`
`scrutiny. See Rosenbaum Dissent at 24, 3–7, 71–75 (interest group
`
`publications).
`
`Because ordinary First Amendment law will displace these
`
`speech bans, creative thinking is required to save them. In its
`
`attempt to persuade the reader otherwise, the dissent misreads
`
`First Amendment precedents. Take National Institute of Family &
`
`Life Advocates v. Becerra (NIFLA). Judge Rosenbaum cites that
`
`case as showing that the Supreme Court “permit[s] governments
`
`to impose content-based restrictions on speech with[] persuasive
`
`evidence . . . of a long (if heretofore unrecognized) tradition to that
`
`effect.” Rosenbaum Dissent at 11; NIFLA, 138 S. Ct. 2361, 2372
`
`(2018) (quotations omitted). Those brackets do a lot of work. Here
`
`is the unaltered quotation: “This Court’s precedents do not permit
`
`governments to impose content-based restrictions on speech
`
`without ‘“persuasive evidence . . . of a long (if heretofore
`
`unrecognized) tradition”’ to that effect.” Id. (quoting United States
`
`v. Alvarez, 567 U.S. 709, 722 (2012) (plurality opinion) (quoting
`
`Brown, 564 U.S. at 792)) (ellipsis in original). Again—the Supreme
`
`Court’s precedents “do not” permit content-based speech
`
`restrictions without persuasive evidence that a long tradition of
`
`such restrictions exists.
`
`Read correctly, NIFLA emphasizes that content-based
`
`regulation is heavily disfavored and that there is no tradition of
`
`regulating professional speech. Id. “As with other kinds of
`
`speech,” it explains, “regulating the content of professionals’
`
`speech poses the inherent risk that the Government seeks not to
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 8 of 110
`
`6
`
` GRANT, J., Concurring
`
`19-10604
`
`advance a legitimate regulatory goal, but to suppress unpopular
`
`ideas or information.” Id. at 2374 (quotation and brackets omitted).
`
`That is why speech does not lose First Amendment protection
`
`“merely because it is uttered by ‘professionals’”—including doctors
`
`or therapists. Id. at 2371–72. It is impossible to rewrite NIFLA to
`
`make a loophole for this one category of speech bans, no matter
`
`how popular they may be.
`
`Make no mistake: these regulations are content-based
`
`restrictions of speech, not conduct. Talk therapy is certainly a form
`
`of treatment. But it “consists—entirely—of words.” Otto, 981 F.3d
`
`at 865. If this speech is conduct, “the same could be said of teaching
`
`or protesting,” of “[d]ebating” and “[b]ook clubs.” Id. The
`
`professional setting of this speech does not transform it into
`
`conduct. Nor does characterizing it as a “scientifically based
`
`healthcare treatment technique” governed by a standard of care.
`
`Rosenbaum Dissent at 24–25. And NIFLA’s refusal to recognize a
`
`lesser-protected category of “professional speech” only confirmed
`
`what this Court already understood in Wollschlaeger: “Speech is
`
`speech, and it must be analyzed as such for purposes of the First
`
`Amendment.” 848 F.3d at 1307 (alteration and quotation omitted);
`
`see NIFLA, 138 S. Ct. at 2371–75.
`
`Having fully exhausted existing free speech doctrine, the
`
`dissent attempts to trailblaze its own. Yet again, that move is
`
`barred by precedent. The Supreme Court has admonished that the
`
`Constitution bars “any freewheeling authority to declare new
`
`categories of speech outside the scope of the First Amendment.”
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 9 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`7
`
`Alvarez, 567 U.S. at 722 (quotation omitted). And it reiterated that
`
`warning in NIFLA, reminding us that courts must be “reluctant to
`
`mark off new categories of speech for diminished constitutional
`
`protection”—especially when such categories would be exempt
`
`from “the normal prohibition on content-based restrictions.” 138
`
`S. Ct. at 2372 (quotations omitted).
`
`Those rebukes should always be enough to induce caution.
`
`But they carry even more force here because in NIFLA the
`
`Supreme Court was specifically criticizing other circuit courts’
`
`approval of “professional speech” bans just like the ones we now
`
`consider. Id. at 2371–72 (citing King v. Governor of New Jersey,
`
`767 F.3d 216, 220, 232–33 (3d Cir. 2014) (upholding a therapist
`
`speech ban virtually identical to the ones here after concluding that
`
`“a licensed professional does not enjoy the full protection of the
`
`First Amendment when speaking as part of the practice of her
`
`profession”), and Pickup v. Brown, 740 F.3d 1208, 1222, 1227–1229
`
`(9th Cir. 2014) (upholding a similar ban, again on the rationale that
`
`it regulates conduct, not speech)); see also Wollschlaeger, 848 F.3d
`
`at 1309. Nor can we forget that the Court specifically “stressed the
`
`danger of content-based regulations in the fields of medicine and
`
`public health.” NIFLA, 138 S. Ct. at 2374 (quotation omitted).
`
`The Supreme Court’s warnings, like so much else from
`
`NIFLA, find no place in the dissent. Judge Rosenbaum proposes a
`
`brand-new category of speech regulation exempt from strict
`
`scrutiny—one that not only rejects our well-established aversion to
`
`viewpoint-based speech restrictions, but actually builds viewpoint
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 10 of 110
`
`8
`
` GRANT, J., Concurring
`
`19-10604
`
`into the analysis. The dissent suggests that we give special
`
`treatment to speech restrictions prohibiting “licensed professionals
`
`from practicing, on populations from whom informed consent
`
`cannot reliably be obtained, treatment techniques that (1) do not
`
`meet the prevailing standard of care, (2) are not shown to be
`
`efficacious, and (3) are associated with a significant increase in the
`
`risk of death”—in short, restrictions that apply only to what the
`
`dissent
`
`calls
`
`“Life-threatening Treatment Techniques.”
`
`Rosenbaum Dissent at 46. This is not a category at all. It is a
`
`description of disfavored speech that bears no resemblance to the
`
`other analytical brackets set out by the Supreme Court. It
`
`privileges the current views of certain professional organizations.
`
`And it requires significant work to even decipher. As a “category,”
`
`this misses the constitutional mark by a mile.
`
`The innovation does not stop there. Although Judge
`
`Rosenbaum “concede[s]” that the talk therapy banned in this case
`
`is “speech, not conduct,” one would not know it from the analysis
`
`that follows. Rosenbaum Dissent at 34. The dissent rejects the
`
`existing frameworks for evaluating laws that burden free speech,
`
`turning instead to caselaw relating to substantive due process and
`
`fundamental rights, concepts that are unrelated to this case and
`
`invoked by none of the parties. Using Washington v. Glucksberg
`
`to support a speech restriction is a novel approach. 521 U.S. 702
`
`(1997). Glucksberg, after all, did not involve a First Amendment
`
`challenge; it outlined limits on substantive due process. Id. at 727–
`
`28. Yet the dissent insists that Glucksberg erects “three guardrails”:
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 11 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`9
`
`it “focuses on the informed opinion of the healthcare community”;
`
`“suggests that the standard of care in question must be supported
`
`by research on the matter” (requiring, of course, that the research
`
`be “acceptable”); and “suggests that informed consent must be
`
`unable to mitigate the dangers of the Life-threatening Treatment
`
`Technique within the universe of clients on whom the law
`
`prohibits
`
`the practice of
`
`the Life-threatening Treatment
`
`Technique.” Rosenbaum Dissent at 60, 62, 64. That is a
`
`remarkable set of takeaways from Glucksberg.
`
`Equally remarkable,
`
`the dissent pivots
`
`to Planned
`
`Parenthood v. Casey in search of a fresh standard of review for its
`
`new category of speech. Rosenbaum Dissent at 67–71. In the
`
`dissent’s view, the plurality opinion in Casey invites us to apply a
`
`“reasonableness” inquiry when testing the constitutionality of
`
`speech restrictions
`
`justified under
`
`the dissent’s
`
`tripartite
`
`Glucksburg analysis. See Planned Parenthood of Se. Pennsylvania
`
`v. Casey, 505 U.S. 833, 883 (1992) (plurality opinion), overruled by
`
`Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228, 2242
`
`(2022).
`
`This approach is a house of cards. To start, NIFLA was clear
`
`that the Casey standard does not apply to regulations of “speech as
`
`speech.” NIFLA, 138 S. Ct. at 2373–74. The rational basis
`
`“reasonableness” standard applies only to regulations of conduct
`
`that incidentally burden speech. Id. And as the dissent (at least
`
`ostensibly) concedes, that category does not fit the speech at issue
`
`here. Rosenbaum Dissent at 34. But despite that concession, the
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 12 of 110
`
`10
`
` GRANT, J., Concurring
`
`19-10604
`
`dissent excises any traditional levels of scrutiny for speech
`
`restrictions; all that remains is judgment of “reasonableness.” Is it
`
`even plausible that a judge who has already concluded that a
`
`particular kind of speech
`
`is a “Life-threatening Treatment
`
`Technique” will then conclude that it would be unreasonable to
`
`ban it? The question answers itself. If there is a standard better
`
`designed to allow speech that judges like and disallow speech that
`
`judges dislike, we do not know what it is.
`
`Indeed, the dissent’s “Glucksburg guideposts,” apart from
`
`their creativity, are designed with one audience in mind. Who
`
`decides which professional bodies qualify as “leading” when
`
`considering the “informed opinion of the healthcare community”?
`
`Who defines the “jurisdiction” of those “leading professional
`
`bodies”? “Acceptable research” by whose standards? “Unable to
`
`mitigate the dangers” according to whom? The answer, of course,
`
`is judges. This category of speech and its circular test would
`
`replace all existing First Amendment doctrines with one
`
`question—whether a judge approves of the speech.
`
`But whether speech is protected does not depend on
`
`whether judges, or communities, like it. The Constitution gives
`
`the government “no power to restrict expression because of its
`
`message, its ideas, its subject matter, or its content.” Reed v. Town
`
`of Gilbert, 576 U.S. 155, 163 (2015) (quotation omitted). The
`
`government cannot be trusted to prohibit only bad speech. And
`
`our role as an independent judiciary is to enforce the First
`
`Amendment, not to decide which ideas are worthy of immunity
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 13 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`11
`
`from government regulation—or which professional groups can
`
`make that decision for us.
`
`Truthfully, the dissent’s unfailing trust in professional
`
`groups is surprising given their track records on the very subject at
`
`issue. Well-intentioned professional associations “may hit the right
`
`mark—but they may also miss it.” Otto, 981 F.3d at 869. As the
`
`panel opinion points out, only a few decades ago the exact set of
`
`“leading professional bodies” that the dissent trusts to regulate
`
`speech—and the research they relied on—endorsed treating
`
`homosexuality as a mental disorder. See id. at 869–70; American
`
`Psychiatric Association, DSM-I (1952); DSM-II (1968); DSM-II 6th
`
`printing change (1973); DSM-III (1980). Under the dissent’s
`
`proposed test, this Court would have been required to uphold
`
`government bans on talk therapy that encouraged ideas about
`
`gender identity and sexual orientation that fell outside the social
`
`orthodoxy of that era. But that defies the First Amendment’s
`
`“fundamental principle that governments have no power to restrict
`
`expression because of its message, its ideas, its subject matter, or its
`
`content.” NIFLA, 138 S. Ct. at 2371 (quotations omitted). This
`
`country’s guarantee of free expression has fostered many political,
`
`social, and religious debates, with our citizens encouraging one
`
`another to consider and reconsider the consensus position. It has
`
`never been the judiciary’s role to moderate those debates, and we
`
`should not start now.
`
`
`
`Even less convincing is the claim made by today’s dissenters
`
`that our decision in Wollschlaeger has no bearing on this case. See
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 14 of 110
`
`12
`
` GRANT, J., Concurring
`
`19-10604
`
`Rosenbaum Dissent at 41–45; Jordan Dissent at 10–11. Judge
`
`Jordan argues that a different procedural posture and a lack of
`
`disputed facts render Wollschlaeger so inapplicable that it can
`
`provide “no support” for the panel’s decision in Otto. Jordan
`
`Dissent at 11. And Judge Rosenbaum simply draws lines between
`
`the substantive content prohibited in Wollschlaeger and those
`
`prohibited here, concluding that our earlier decision “does not in
`
`any way conflict with” her proposed approach because the statute
`
`there “could be understood to require” doctors to “violate the
`
`standard of care” rather than follow it. Rosenbaum Dissent at 43,
`
`45.
`
`These attempts to distinguish our most relevant recent
`
`precedent are not persuasive. Wollschlaeger explicitly held that
`
`“content-based restrictions on speech by those engaged in a certain
`
`profession” deserve heightened review. 848 F.3d at 1311 (rejecting
`
`both a comparison to Casey and application of rational basis
`
`review). Indeed, it expressed “serious doubts” about the Ninth
`
`Circuit’s characterization of the same kind of therapy as conduct
`
`rather than speech. Id. at 1309. And it emphasized that “the
`
`enterprise of labeling certain verbal or written communications
`
`‘speech’ and others ‘conduct’ is unprincipled and susceptible to
`
`manipulation.” Id. at 1308 (quotation omitted). It is no wonder
`
`that the district court found itself “stymied by the Eleventh
`
`Circuit’s analysis in Wollschlaeger” when it considered the
`
`possibility that it could uphold the ordinances as regulating some
`
`form of conduct. Wollschlaeger squarely precludes that argument.
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 15 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`13
`
`It also precludes pulling a lax conduct-based standard of review out
`
`of Casey to perform an end-run on free speech doctrine in the
`
`professional context: “state officials cannot successfully rely on a
`
`single paragraph in the plurality opinion of three Justices . . . to
`
`support the use of rational basis review here.” Id. at 1311. The
`
`dissenting opinion’s attempt to convert a case striking down a
`
`speech ban for doctors into a case supporting a speech ban for
`
`therapists is spirited, but it fails to get the job done.
`
`One final point. States need not shutter their licensing
`
`boards in light of this Court’s decision in Otto. Regulatory
`
`authority is alive and well—just as robust as it was before the
`
`opinion. Indeed, though Otto was published nearly two years ago,
`
`we have no indication that therapy has become “a Wild West of
`
`anything goes—no matter how detrimental to clients’ health.”
`
`Rosenbaum Dissent at 25. Nor was there any such result in the
`
`years following our decision in Wollschlaeger, which also refused
`
`to allow content-based restrictions on professional speech. And
`
`that’s no surprise, because “[t]his case, like Wollschlaeger, is not
`
`about licensure requirements. It is about speech.” Otto, 981 F.3d
`
`at 866–67 (footnote and citation omitted). The State did not lose
`
`its ability to regulate the medical profession simply because it was
`
`compelled to respect constitutional boundaries. Nor, we add, have
`
`the parties raised the specter of thwarted health and safety
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 16 of 110
`
`14
`
` GRANT, J., Concurring
`
`19-10604
`
`regulation so vividly imagined by Judge Rosenbaum. That concern
`
`is of the dissent’s own making.1
`
`We take some comfort in the fact that NIFLA’s dissenters
`
`also lobbed charges that the majority there imperiled health and
`
`safety regulations. See NIFLA, 138 S. Ct. at 2380–81 (Breyer, J.,
`
`dissenting); id. at 2376 (majority opinion) (responding). And we
`
`are confident that the fears of regulatory impotence expressed here
`
`will be similarly relieved in good time. The panel opinion itself
`
`explains that states can penalize harmful speech and hold
`
`accountable those who hurt children. Otto, 981 F.3d at 870.
`
`License revocations, professional suspensions, malpractice suits,
`
`even criminal charges—all are on the table for professionals who
`
`violate the public trust. But “broad prophylactic rules in the area
`
`of free expression” remain suspect, no matter how much a judge
`
`may wish to engineer an exception for speech that seems
`
`particularly risky. Id. (quoting NAACP v. Button, 371 U.S. 415, 438
`
`(1963)) (brackets omitted).
`
`
`1 We add that the panel opinion does not directly affect Florida’s regulatory
`authority at all. The ordinances here are the legislative products of local
`governments, but Florida law commits regulatory authority to the State. See
`Fla. Stat. ch. 491 (regulatory authority over therapists); id. ch. 456 (regulatory
`authority over health professionals); see also Vazzo v. City of Tampa, 415 F.
`Supp. 3d 1087, 1107 (M.D. Fla. 2019).
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 17 of 110
`
`19-10604
`
` GRANT, J., Concurring
`
`15
`
`*
`
`*
`
`*
`
`Neither the panel opinion nor this Circuit’s decision against
`
`en banc review express any view on the efficacy or desirability of
`
`the speech at issue in this case. Nor do they condone or ignore the
`
`struggles faced by many LGBTQ youth. But “we cannot react to
`
`that pain by punishing the speaker. As a Nation we have chosen a
`
`different course.” Snyder v. Phelps, 562 U.S. 443, 461 (2011). What
`
`this Circuit has done—indeed, all it has done—is uphold the
`
`protections of the First Amendment for unpopular speech. That
`
`can be hard to do. But if the First Amendment only protected
`
`speech that judges and politicians approved of, it would not be of
`
`much use. We concur in the Court’s decision not to rehear this
`
`case en banc.
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 18 of 110
`
`19-10604
`
` JORDAN, J., Dissenting
`
`1
`
`JORDAN, Circuit Judge, joined by WILSON, Circuit Judge, and by
`
`ROSENBAUM and JILL PRYOR, Circuit Judges, as to Parts I-IV, dis-
`
`senting from the denial of rehearing en banc:
`
`Judge Rosenbaum makes a number of salient points in her
`
`dissent as to why, insofar as the First Amendment is concerned,
`
`SOCE therapy should be evaluated as a form of medical treatment.
`
`As the author of Wollschlaeger v. Governor of Florida, 848 F.3d
`
`1293 (11th Cir. 2017) (en banc), I think the characterization of
`
`SOCE therapy presents a difficult question. And although I am not
`
`sure who is right—Judge Rosenbaum or the panel majority—with
`
`respect to the First Amendment analysis, the issue is sufficiently
`
`important to merit en banc review.
`
`I also believe en banc consideration is warranted for a less
`
`complex but no less important reason. As I hope to explain, the
`
`panel majority in this preliminary injunction appeal ignored the
`
`clear error standard of review—never acknowledging or applying
`
`it —and substituted its own factual findings for those of the district
`
`court on important issues.
`
`I
`
`When we hear an appeal from the denial or grant of a pre-
`
`liminary injunction, we review the district court’s “factual findings
`
`for clear error.” Indep. Party of Fla. v. Sec’y, 967 F.3d 1277, 1280
`
`(11th Cir. 2020). That standard of review is so long-standing and
`
`unremarkable that it is by now gospel. Here is the way Judge Mar-
`
`cus put the matter some 20 years ago: “Preliminary injunctions are,
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 19 of 110
`
`2
`
` JORDAN, J., Dissenting
`
`19-10604
`
`by their nature, products of an expedited process often based on an
`
`underdeveloped and incomplete evidentiary record. As is usually
`
`the case, the [district] court is in a far better position than this Court
`
`to evaluate the evidence, and we will not disturb its factual findings
`
`unless they are clearly erroneous.” Cumulus Media, Inc. v. Clear
`
`Channel Commc’ns, Inc., 304 F.3d 1067, 1171 (11th Cir. 2002) (ci-
`
`tations omitted). It’s hard to improve on that explanation.
`
`The district court in this case took evidence from the parties
`
`and received proposed findings of fact and conclusions of law from
`
`them following oral argument. Then, in its order denying a pre-
`
`liminary injunction, the district court evaluated the evidence and
`
`made a number of important factual determinations. See Otto v.
`
`City of Boca Raton, 353 F. Supp. 3d 1237, 1241, 1258–70 (S.D. Fla.
`
`2019) (Otto I).
`
`The panel majority acknowledged the general abuse of dis-
`
`cretion standard for preliminary injunction appeals, but it did not
`
`mention, much less apply, the subsidiary clear error standard for
`
`underlying factual findings. See Otto v. City of Boca Raton, 981
`
`F.3d 854, 862 (11th Cir. 2020) (Otto II). Indeed, the phrases “clear
`
`error” or “clearly erroneous” are nowhere to be found in the panel
`
`majority’s opinion.
`
`Maybe the panel majority thought that the clear error stand-
`
`ard was inapplicable because the district court did not base its fac-
`
`tual findings on credibility determinations. But if that was the un-
`
`stated reason for its failure to acknowledge and apply the clear er-
`
`ror standard, the panel majority was mistaken. “Findings of fact,
`
`
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 20 of 110
`
`19-10604
`
` JORDAN, J., Dissenting
`
`3
`
`whether based on oral or other evidence, must not be set aside un-
`
`less clearly erroneous.” Fed. R. Civ. P. 52(a)(6). Not surprisingly,
`
`the Supreme Court has held that under Rule 52(a) the clear error
`
`standard applies not only to factual findings based on credibility de-
`
`terminations but also to findings based on “physical or documen-
`
`tary evidence or inferences from other facts.” Anderson v. Besse-
`
`mer City, 470 U.S. 564, 573 (1985).
`
`II
`
`A factual finding “is clearly erroneous when[,] although
`
`there is evidence to support it, the reviewing court on the entire
`
`evidence is left with the definite and firm conviction that a mistake
`
`has been committed.” Anderson, 470 U.S. at 573 (internal quota-
`
`tion marks and citation omitted). But a finding is not clearly erro-
`
`neous simply because the reviewing court would have weighed the
`
`evidence differently or reached a different outcome. See id. at 574.
`
`If there are two permissible views of the evidence or the district
`
`court’s account of the evidence is “plausible in light of the record
`
`viewed in its entirety,” then the district court’s finding is not clearly
`
`erroneous. Id. The clear error standard is therefore “highly defer-
`
`ential.” Bellitto v. Snipes, 935 F.3d 1192, 1197 (11th Cir. 2019) (in-
`
`ternal quotation marks and citation omitted).
`
`Here the district court made several important findings of
`
`fact that the panel majority ignored, mischaracterized, or revised.
`
`In the interest of brevity, I will highlight two of the important find-
`
`ings by the district court and the findings that the panel majority
`
`substituted in their place.
`
`
`
`
`
`USCA11 Case: 19-10604 Date Filed: 07/20/2022 Page: 21 of 110
`
`4
`
` JORDAN, J., Dissenting
`
`19-10604
`
`A
`
`The district court found that there is a distinction between
`
`the plaintiffs’ “expression of their views about SOCE, their advo-
`
`cacy of SOCE, . . . their discussions with minor clients about
`
`SOCE,” and their “efforts, through a medical intervention, by a li-
`
`censed provider, to therapeutically change a minor’s sexual orien-
`
`tation.” Otto I, 353 F. Supp. 3d at 1244, 1264, 1269. In other words,
`
`the district court found that the “practice” or “perform[ance]” of
`
`SOCE therapy is different from “a dialogue between patient and
`
`provider” about that treatment, even one in which a plaintiff “com-
`
`mend[ed] and recommend[ed]” it. Id. at 1256, 1269 (emphasis
`
`omitted).
`
`The distinction is highlighted in the district court’s factual
`
`determination that the s