Tommia Dean v. Neil Warren

12 F.4th 1248

Opinion

[PUBLISH]

IN THE UNITED STATES COURT OF APPEALS

FOR THE ELEVENTH CIRCUIT

________________________

No. 19-14674

________________________

D.C. Docket No. 1:18-cv-04224-TCB

TOMMIA DEAN, Plaintiff - Appellant,

versus NEIL WARREN,

Defendant - Appellee.

________________________

Appeal from the United States District Court for the Northern District of Georgia

________________________

(September 2, 2021)

Before WILLIAM PRYOR, Chief Judge, JILL PRYOR and ED CARNES, Circuit Judges.

JILL PRYOR, Circuit Judge:* Joining a national movement intended to curb police brutality against African Americans and advance the cause of racial justice, Tommia Dean, a cheerleader at one of Georgia’s public universities, and a handful of her teammates kneeled during the pre-game national anthem at one of the university’s football games. What followed, according to Dean, was a public and private conspiracy— orchestrated by the university’s leadership, the county sheriff, and a Georgia legislator—to deprive Dean and her teammates of their First Amendment rights. Seeking redress, Dean filed this action alleging violations of 42 U.S.C. § 1983 and § 1985(3).

The only issue on appeal is whether the district court erred by dismissing Dean’s § 1985(3) claim against the sheriff, Neil Warren. The district court concluded that Dean failed to state a § 1985(3) claim because she failed to plausibly allege that Warren possessed the requisite class-based animus. After careful review, and with the benefit of oral argument, we agree with the district court that Dean failed to surmount § 1985(3)’s class-based animus bar under the

*

There are two majority opinions for the Court in this case. One is by Judge Jill Pryor, which all three judges join. The other majority opinion, which adds an alternative holding, is the concurring opinion of Chief Judge William Pryor, joined by Judge Ed Carnes.

3

standard established by Supreme Court precedent. We therefore affirm the district court’s order dismissing Dean’s claim against Warren.

I. BACKGROUND

A. Factual Background 1

During the National Football League’s 2016 season, Colin Kaepernick, a San Francisco 49ers quarterback, refused to stand for the national anthem. Instead, he kneeled in protest of police brutality against African Americans.2 His kneeling ignited a national controversy concerning police brutality, the appropriateness of kneeling in protest during presentations of the national anthem, and racial injustice in America.

Kaepernick’s kneeling encouraged other athletes to kneel, including a group of African American cheerleaders at Kennesaw State University, a public university in Georgia. Dean, the plaintiff, was one of those cheerleaders. After she and her teammates kneeled in protest for the first time in September 2017, several public and private individuals conspired to prevent Dean and her teammates from kneeling during the national anthem as a form of protest.

1 Because this case is an appeal from an order granting a motion to dismiss, we accept as true all well-pled allegations in the complaint and construe them in the light most favorable to Dean. Blevins v. Aksut, 849 F.3d 1016, 1018–19 (11th Cir. 2017). We therefore recite the facts as Dean has alleged them.

2 “African Americans” is the term Dean used in her complaint and briefing on appeal.

4

When the President of KSU, Samuel Olens, learned of the kneeling, he emailed the university’s Vice President of Student Affairs to express concern about the protest, directing the Vice President to schedule a meeting with the cheerleaders the following week. The day after the cheerleaders first kneeled, Earl Ehrhart, then a Georgia state legislator and chair of the state House committee with control over the budget for Georgia’s public universities, called Scott Whitlock, KSU’s Senior Assistant Athletic Director, instructing him that the cheerleaders should be prohibited from kneeling during the national anthem. 3 The next day, the presidents of the University System of Georgia gathered to discuss the kneeling incident. During that gathering, the University System’s staff relayed to the presidents, including Olens, legal advice from the office of Georgia’s Attorney General. That advice instructed the presidents that the First Amendment protected the students’ right to kneel in protest during the national anthem, so long as the expression was not disruptive. The presidents were also informed that no action should be taken on the protests without first discussing it with the University System.

That same day, Ehrhart called Olens to tell him that cheerleaders must not be permitted to kneel during the national anthem and those who continued to kneel

3 According to the complaint, Ehrhart has, in the past, described himself as “the funding source” for the University System and has had a history of wielding his power of the purse to pressure administrators of Georgia’s state universities to comply with his personal agenda.

5

should be removed from the team. During this call, or by a subsequent communication, Olens assured Ehrhart that the cheerleaders would not kneel again. Ehrhart was not the only individual to relay the message that the kneeling must cease to Olens that day. Neil Warren, the county sheriff, did too.

Two days later, Whitlock and Matt Griffin, then-interim Athletic Director at KSU, announced to an assembled group of KSU officials that the cheerleaders would not be allowed on the field during the anthem but would instead remain in the stadium’s tunnel. Assistant Athletic Director Michael DeGeorge questioned the timing of the policy change that came days after the protest. That same day, Griffin met with Olens to obtain permission to implement the “tunnel rule”: the new policy requiring cheerleaders to remain in the tunnel during the national anthem. Olens gave Griffin permission to implement the tunnel rule to appease Ehrhart and Warren.

Two days later, Warren again called Olens about the protests. Olens assured Warren that the protests would not happen again, informing him that the newly adopted tunnel rule prevented the cheerleaders from taking the field during the national anthem. That day, the Marietta Daily Journal published an article stating, “Ehrhart said Attorney General Chris Carr and Olens have been helpful in the

6

situation and he [Ehrhart] believes the behavior [the anthem kneeling] will not occur at KSU again.” Doc. 1 ¶ 26. 4 The next day, one week after the first kneeling protest, the tunnel rule took effect. For the first time in KSU’s history, the cheerleaders were not on the field during the national anthem. Instead, they were held in the tunnel by KSU Athletic Department representatives until the anthem concluded.

Warren and Ehrhart celebrated the development, taking credit for the tunnel rule. In a text message, Warren boasted, “[n]ot letting the cheerleaders come out on the field until after national anthem [sic] was one of the recommendations that Earl [Ehrhart] and I gave him [Olens]!” Id. ¶ 29. In a text to Warren, Ehrhart commented, “He [Olens] had to be dragged there but with you and I pushing he had no choice. Thanks for you[r] patriotism my friend.” Id.

Two days after the tunnel rule was implemented, an opinion writer for the Marietta Daily Journal warned the cheerleaders that they had better not insist on kneeling during the national anthem. He informed Dean and her teammates that “the sheriff, any sheriff, is about the last person you want to mess with in any county” and warned them that their “antics” had “riled [up] . . . something fierce [in Warren].” Id. ¶ 34. He explained, “Again these guys [Warren and Ehrhart] are very upset with you. Don’t say I didn’t warn you.” Id.

4 “Doc.” numbers refer to the district court’s docket entries.

7

Some KSU students were dismayed by the administration’s apparent attempt to prevent the cheerleaders from kneeling during the anthem. They protested in support of the cheerleaders. The Chancellor of the University System of Georgia directed the Board of Regents’ Office of Legal Affairs to conduct a review of the actions KSU took following the cheerleaders’ protest.5 While that review was ongoing, the cheerleaders were again held off the field in the tunnel during the national anthem, where four members, including Dean, again kneeled in protest.

Less than a month later, facing pressure from students, faculty, the press, and the Board of Regents, Olens wrote a public letter to the KSU community explaining that, at the next home game, the tunnel rule would be abolished and the cheerleaders would again be permitted to take the field before presentation of the national anthem. In the letter, Olens recognized that the central message of the protest concerned political issues of national concern.

The Regents’ report was issued days later. It found that Olens was aware of the tunnel rule’s implementation and did not prevent it. It also found that Olens acted contrary to the University System Office’s instruction to provide it with notice of any policy changes made in response to the cheerleaders’ kneeling.

5 The University System of Georgia is a state agency composed of 26 of Georgia’s higher education institutions. The Board of Regents oversees the institutions that compose the University System of Georgia. The Chancellor is elected by the Board’s members and is the Board’s chief executive officer and the University System’s chief administrative officer.

8

Finally, it questioned a justification KSU officials gave for the tunnel rule, that it was designed to eliminate a two-minute gap in the pregame music, implying that the justification was likely pretextual given the timing of the rule’s implementation and the fact that the two-minute gap was remedied by other means. B. Procedural History In a single action, Dean sued two groups of defendants. In Count 1, she alleged that Olens, Whitlock, and Griffin deprived her of her expressive speech rights under the First and Fourteenth Amendments. See 42 U.S.C. § 1983. Count 1 is not relevant to this appeal.6 In Count 2, Dean alleged that Ehrhart and Warren conspired to deprive her of her constitutional rights.7 See 42 U.S.C. § 1985(3). Dean alleged that Warren engaged in the conspiracy “because of her race,” that is, because Dean and her protesting teammates are African American. Id. ¶ 50. She also alleged that Warren engaged in the conspiracy “because . . . she was protesting police brutality against African Americans.” Id.

6 After the district court entered the order giving rise to this appeal, the parties filed a joint motion for voluntary dismissal of the § 1983 defendants pursuant to a settlement agreement. The district court granted the motion, dismissing the § 1983 defendants. Thus, Dean’s only remaining count is the one at issue in this appeal.

7 Dean’s § 1985(3) claim against Ehrhart is also irrelevant because Dean appealed the district court’s dismissal order only as to Warren. So, we recount the procedural history as it relates to Warren only.

9

Warren moved to dismiss for failure to state a claim. In the motion, Warren argued, among other things, that Dean’s claim failed as a matter of law because Dean had not stated a plausible § 1985(3) claim that he had acted with a racial or otherwise class-based discriminatory motive.

The district court agreed. It explained that § 1985(3) “requires a showing of some racial, or perhaps otherwise class-based, invidiously discriminatory animus behind the conspirators’ action.” Doc. 36 at 9 (internal quotation marks omitted). Dean’s allegations, the court concluded, failed to make this showing. In assessing whether Dean could state a claim under § 1985(3), the district court conceptualized three theories by which she might survive the class-based animus requirement.

First, Dean alleged a direct race-based theory of animus—that Warren undertook the conspiracy “because of her race.” Doc. 1 ¶ 50. The district court concluded that she failed to state a § 1985(3) claim under this theory because she failed to allege sufficient facts showing that Warren was motivated by her race.

Second, Dean alleged an indirect race-based theory of animus—that Warren undertook the conspiracy because the content of her protest concerned police brutality against African Americans, which is a political issue implicating race. The district court concluded that she failed to state a claim under this theory because the complaint failed to allege sufficient facts linking Warren’s conduct to the content of the cheerleaders’ protest. Alternatively, the court concluded that,

10

even if Warren’s conduct was aimed at the protest’s content, Dean’s indirect race- based theory was not cognizable under binding precedent. See Bray v. Alexandria Women’s Health Clinic, 506 U.S. 263 (1993).

Third, Dean alleged a political class-based theory of animus—that Warren undertook the conspiracy because of her membership in a political class, namely, the class of people protesting police brutality against African Americans. The district court acknowledged that it is unsettled in this circuit whether a § 1985(3) claim may be predicated upon political class-based animus and, relying on nonbinding authority, rejected Dean’s political class-based theory.

After reasoning that none of Dean’s theories surmounted the class-based animus bar, the district court concluded that she failed to state a § 1985(3) claim and dismissed Warren from the action. This is Dean’s appeal.

II. STANDARD OF REVIEW We review de novo the grant of a defendant’s motion to dismiss, accepting allegations in the complaint as true and construing them in the light most favorable to the plaintiff. See Blevins v. Aksut, 849 F.3d 1016, 1018–19 (11th Cir. 2017).

11

III. ANALYSIS

A § 1985(3) plaintiff must satisfy four requirements, one of which disposes of this appeal.8 This case turns on the requirement that the plaintiff allege the defendant conspired against her “for the purpose of depriving . . . any person or class of persons of the equal protection of the laws, or of equal privileges and immunities under the laws.” Griffin v. Breckenridge, 403 U.S. 88, 102–03 (1971) (quoting 42 U.S.C. § 1985(3)). The Supreme Court has described this requirement as the plaintiff’s obligation to allege, among other things, “some racial, or perhaps otherwise class-based, invidiously discriminatory animus” behind the defendant’s action taken in furtherance of the conspiracy. Bray, 506 U.S. at 267–68 (quoting Griffin, 403 U.S. at 102). That animus standard requires that the defendant proceeded on his course of conduct “because of, not merely in spite of, its adverse effects upon an identifiable group.” Id. at 271–72 (internal quotation marks omitted).

Like the district court, we read Dean’s complaint to have alleged three theories of § 1985(3) liability. The district court concluded that each theory failed

8 The § 1985(3) defendant must have (1) conspired or gone in disguise on the highway or on the premises of another, (2) for the purpose of depriving, either directly or indirectly, any person or class of persons of the equal protection of the laws, or of equal privileges and immunities under the laws, (3) taken or caused an action to be taken in furtherance of the conspiracy’s object, and (4) injured an individual’s person or property or deprived her of exercising any right or privilege of a United States citizen. Griffin v. Breckenridge, 403 U.S. 88, 102–03 (1971).

12

either because the facts alleged in the complaint failed to plausibly state a claim or because the theory, as a matter of law, could not support § 1985(3) liability. We discuss each of these theories in turn. After careful review, and with the benefit of oral argument, we affirm the district court’s ruling that Dean failed to state a claim under § 1985(3).9 A. Dean’s Direct Race-Based Theory We agree with the district court that Dean’s direct race-based theory cannot succeed because she failed to plead sufficient facts supporting it. The theory is that Warren undertook the conspiracy to prevent Dean and her protesting teammates from exercising their First Amendment rights because they are African American. Warren does not dispute—and we have no doubt—that, as a matter of law, a plaintiff could satisfy the class-based animus requirement with such a theory. It is well-settled that a conspiracy aimed at depriving African Americans of constitutional rights because they are African American falls within § 1985(3)’s

9 Warren argues that Dean failed, in her opening brief, to challenge an alternative ground for dismissal and therefore abandoned the issue. The district court ruled that “[a]n independent ground for granting [Warren’s] motion to dismiss” was that Dean failed to allege that the § 1985(3) defendants’ conspiracy violated a law other than § 1985(3). Doc. 36 at 26 n.2. According to Warren, because Dean failed to attack that alternative ground, the district court’s order must be affirmed. See Sapuppo v. Allstate Floridian Ins. Co., 739 F.3d 678, 680 (11th Cir. 2014). We disagree. Although Dean did not mention the district court’s alternative ruling as such, she addressed in her opening brief the underlying question, arguing that she alleged a violation of law other than § 1985(3)—the First Amendment. Because Dean does not lose this appeal for failure to challenge the district court’s alternative ground, and we resolve the appeal on the class-based animus issue, we do not address whether the district court erred by dismissing Dean’s claim for failure to allege an independent legal violation.

13

heartland. See Griffin, 403 U.S. at 103 (holding that targeting plaintiffs because of their race falls within the core of § 1985(3)). Dean’s problem is that her complaint fails to plead sufficient facts to plausibly support the theory.

Under Federal Rule of Civil Procedure 8(a)(2), when determining whether a complaint survives a motion to dismiss, we begin by identifying allegations that, because they are merely conclusory or state legal conclusions, are entitled to no presumption of truth. Ashcroft v. Iqbal, 556 U.S. 662, 678–79 (2009). Then, turning to the remaining allegations, we consider whether the complaint “contain[s] sufficient factual matter, accepted as true, to ‘state a claim to relief that is plausible on its face.’” Id. at 678 (quoting Bell Atl. Corp. v. Twombly, 550 U.S. 544, 570 (2007)). Plausibility rests between “possibility” and “probability,” and we determine whether a complaint satisfies that standard by “draw[ing] on [our] judicial experience and common sense.” Id. at 678–79.

Dean’s complaint contains two allegations that arguably support her theory that Warren was motivated by Dean’s and her teammates’ race. First, the complaint alleged that “[a]ll of the KSU cheerleaders who kne[eled] during the national anthem are African American.” Doc. 1 at ¶ 50. Second, Dean alleged that Warren “engaged in the conspiracy against [her] because of her race.” Id. Dean concedes that the second allegation does not help her clear the plausibility bar because, as her brief puts it, the allegation was intended to “provide[] the legal

14

framework for the factual allegations of the [c]omplaint.” Appellant’s Br. at 11; see Iqbal, 556 U.S. at 680–82 (holding that plaintiff’s allegation that defendants subjected him to harsh conditions “solely on account of his religion, race, and/or national origin” was a “formulaic recitation of the elements” of plaintiff’s claim and “not entitled to be assumed true” (alteration adopted) (internal quotation marks omitted)). And the first allegation, standing alone, does not nudge Dean’s direct race-based theory of § 1985(3) liability across the line from possible to plausible. Iqbal, 556 U.S. at 682 (holding that complaint did not “contain facts plausibly showing that [the defendants] purposefully adopted” a discriminatory policy even though the complaint alleged that the challenged policy had a disparate impact). That Warren’s targets are African American, without more, does not make it plausible that he targeted the cheerleaders because they are African American.

Dean resists this conclusion by pointing to other allegations in her complaint that she contends support her direct race-based theory. These include allegations that: the content of the cheerleaders’ protest concerned police brutality against African Americans specifically; the cheerleaders engaged in the protest in solidarity with professional athletes who themselves were protesting racial injustice; similar protests were occurring nationwide; the protests were initiated by Kaepernick, who is African American; the protests were praised by some as patriotic and condemned by others; Warren pressured Olens into implementing the

15

tunnel rule; and Olens eventually recognized that the cheerleaders were protesting issues of national concern.

These allegations do not change our view of Dean’s direct race-based theory. Certainly they “relat[e] to race and racial issues,” Appellant’s Br. at 12, and they support the theory that Warren discriminated against Dean because of the content of her protests (a subject to which we will return). But discrimination based on the protesting of racial issues, no matter how compelling those issues are, is simply not the same as discrimination based on the race of the protestors. Thus, the additional allegations do not support Dean’s theory that Warren discriminated against her because she is African American. Our “experience and common sense,” Iqbal, 556 U.S. at 679, tell us that Warren’s conduct was likely motivated by the nature and content of the kneeling cheerleaders’ message and that his conduct therefore would have been the same had some or even most of the cheerleaders been white. Therefore, Dean failed to allege that Warren undertook the conspiracy because Dean and her teammates are African American. We agree with the district court that Dean’s direct race-based theory was insufficiently pled. B. Dean’s Indirect Race-Based Theory By contrast, Dean’s indirect race-based theory fails on legal, rather than factual, grounds. This theory maintains that Warren undertook the conspiracy to prevent Dean and her protesting teammates from exercising their First Amendment

16

rights because they were “protesting police brutality against African Americans.” Doc. 1 at ¶ 50. Put differently, based on her allegations that Warren undertook the conspiracy to thwart the protests, which were intended and widely understood to concern racial issues, Dean argues that Warren undertook the conspiracy to stop the cheerleaders’ kneeling because of its racial message and thereby violated § 1985(3). In our view, this theory of § 1985(3) liability is precluded by Supreme Court precedent.

To explain why Dean’s indirect race-based theory is legally inviable we must delve into § 1985(3)’s background and review the controlling Supreme Court caselaw. Section 1985(3) is kin to the better known § 1983. See John C. Jeffries et al., Civil Rights Actions: Enforcing the Constitution 798 (4th ed. 2018) (explaining that both sections are derived from the Civil Rights Act of 1871, also known as the Ku Klux Klan Act, and therefore that the sections “share much of the same legislative history and many of the same objectives”). Section 1983 and § 1985(3) differ, however, in significant respects. Unlike § 1983, which created a cause of action against state actors who deprive individuals of rights secured by federal law, § 1985(3) created a cause of action against private individuals who, among other things, undertake conspiracies “for the purpose of depriving, either directly or indirectly, any person or class of persons of the equal protection of the laws, or of equal privileges and immunities under the laws.” 42 U.S.C. § 1985(3); see Griffin,

17

403 U.S. at 96–101. That § 1985(3) “reach[es] private action does not, however, mean that it was intended to apply to all tortious, conspiratorial interferences with the rights of others.” Griffin, 403 U.S. at 101. As the Supreme Court explained, “[t]he language requiring intent to deprive of equal protection, or equal privileges and immunities, means that there must be some racial, or perhaps otherwise class- based, invidiously discriminatory animus behind the conspirators’ action.” Id. at 102.

Griffin held that § 1985(3)’s animus requirement was satisfied when the complaint alleged that two white defendants, motivated by their mistaken belief that the African American plaintiffs were traveling in a vehicle operated by a civil rights worker, conspired to block the vehicle’s passage and violently terrorized the plaintiffs. Id. at 89–91, 102–03. Since Griffin, the Court has elaborated on the animus requirement in two cases that control the outcome of this appeal. See United Bhd. of Carpenters & Joiners of Am., Loc. 610, AFL-CIO v. Scott, 463 U.S. 825 (1983); Bray, 506 U.S. 263. We turn to them now.

In Scott, the Court considered whether “conspiratorial discrimination against employees of a non-unionized entity” satisfied § 1985(3)’s animus requirement. 463 U.S. at 829. The lower courts agreed that it did. Id. at 829–30. The Supreme Court reversed. Id. at 830. Taking no position on the question whether § 1985(3) “reaches conspiracies other than those motivated by racial bias,” the Court rejected

18

the proposition that § 1985(3) “forbids conspiracies against workers who refuse to join a union.” Id. at 835. The Court acknowledged that § 1985(3) plainly forbids “class-based animus . . . against [African Americans] and those who championed their cause.” See id. at 836. It went on to hold, though, that the statute’s legislative history suggested that Congress did not intend for it to extend to “conspiracies motivated by economic or commercial animus” and that such an expansive construction of § 1985(3) was “not compelled” by its plain language. Id. at 837–39.

In Bray, the Court offered its most thorough discussion of the animus requirement to date. It held that a conspiracy to obstruct access to an abortion clinic was not actionable under § 1985(3) because the abortion-clinic plaintiffs failed to satisfy the animus requirement. 506 U.S. at 266–68. The Court explained that the animus requirement is comprised of two analytically distinct inquiries: (1) whether there is a “qualifying class” and (2) whether the defendant was motivated by discriminatory animus against the class. Id. at 269.

Beginning with the first inquiry, the Court rejected the possibility that the “‘class’ of ‘women seeking abortion’” was a qualifying class. Id. It reasoned that, “[w]hatever may be the precise meaning of a ‘class’ for purposes of Griffin’s speculative extension of § 1985(3) beyond race, the term unquestionably connotes something more than a group of individuals who share a desire to engage in

19

conduct that the § 1985(3) defendant disfavors.” Id. (explaining that a § 1985(3) plaintiff cannot “simply defin[e] the aggrieved class as those seeking to engage in the activity the defendant has interfered with”). The Court concluded that the class of women seeking abortion was too gerrymandered to count as a “qualifying class” that could satisfy § 1985(3)’s animus requirement. Id.

The Court then turned to whether the abortion clinics satisfied the qualifying class inquiry by “alleg[ing] [that the] class-based discrimination [was] directed not at ‘women seeking abortion’ but at women in general.” Id. The Court concluded that it was “unnecessary to decide” whether the class of women was “a qualifying class under § 1985(3)” because the abortion clinics could not demonstrate that the anti-abortion activists’ “opposition to abortion reflect[ed] an animus against women.” Id. In other words, the Court assumed “women” was a qualifying class and held that the plaintiffs’ claim nevertheless failed under the second inquiry, that is, whether the defendants were motivated by discriminatory animus against women as a class. Id. at 269–70.

The second inquiry requires a defendant to have been “motivated by a purpose (malevolent or benign) directed specifically at [the qualifying class]” by reason of the essential characteristic of that class. Id. at 270. Because the record did not contain direct evidence suggesting that the defendants targeted women as a class, the Bray plaintiffs could satisfy this requirement “only if one of two

20

suggested propositions [was] true: (1) that opposition to abortion can reasonably be presumed to reflect a sex-based intent, or (2) that intent is irrelevant, and a class-based animus can be determined solely by effect.” Id. The Court concluded that “[n]either proposition [was] supportable.” Id.

As to the first proposition, the Court explained that “[s]ome activities may be such an irrational object of disfavor that, if they are targeted, and if they also happen to be engaged in exclusively or predominantly by a particular class of people, an intent to disfavor that class can readily be presumed.” Id. For this reason, a “tax on wearing yarmulkes is a tax on Jews.” Id. But, the Court explained, the anti-abortion activists’ conduct in Bray was not tantamount to a tax on yarmulkes. According to the Court, “opposition to voluntary abortion cannot possibly be considered such an irrational surrogate for opposition to (or paternalism towards) women.” Id. “Whatever one thinks of abortion,” the Court reasoned, “it cannot be denied that there are common and respectable reasons for opposing it, other than hatred of, or condescension toward (or indeed any view at all concerning), women as a class—as is evident from the fact that [both men and women are on both sides of the abortion issue].” Id.

As further support for its conclusion, the Court pointed to Griffin’s use of the term “invidious,” which means “tending to excite odium, ill will, or envy; likely to give offense; esp., unjustly and irritatingly discriminating.” Id. at 274

21

(alteration adopted) (quoting Webster’s Second International Dictionary 1306 (1954)). In using this term, Griffin announced a high bar for plaintiffs. See id. The “goal of preventing abortion,” the Court explained in Bray, “does not remotely qualify for such harsh description [as invidious].” Id. The Bray plaintiffs, therefore, failed to show that opposition to abortion could reasonably be presumed to reflect a sex-based intent that cleared the animus bar.

Turning to the second proposition, the Court concluded that the abortion clinics failed to satisfy the discriminatory animus element of § 1985(3)’s class- based animus requirement with the theory that animus could be presumed because the conduct targeted by the § 1985(3) conspiracy was conduct that only members of the qualifying class engaged in. Id. at 271. The Court explained that even though “voluntary abortion [was] an activity engaged in only by women, to disfavor it [was not] ipso facto to discriminate invidiously against women as a class.” Id. (footnote omitted). As in the constitutional equal protection context, discriminatory animus under § 1985(3) requires that the defendant acted “at least in part ‘because of,’ not merely ‘in spite of,’ [his actions’] adverse effects upon [members of the qualifying class].” Id. at 272. That conclusion follows, the Court declared, “not because . . . Equal Protection Clause jurisprudence is automatically incorporated into § 1985(3), but rather because it is inherent in the requirement of a class-based animus, i.e., an animus based on class.” Id. at 272 n.4. Because

22

§ 1985(3) requires “because of” not merely “in spite of” targeting, the abortion clinics could not state a § 1985(3) claim based on the theory that the defendants’ conduct had merely a discriminatory effect on women. Id. at 273–74.

Bray’s discussion of § 1985(3)’s class-based animus requirement forecloses Dean’s indirect race-based theory. To explain why, we must look closer at the allegations supporting that theory. Dean alleged that Warren engaged in the conspiracy because she was protesting police brutality against African Americans. We will assume Dean sufficiently alleged that Warren was motivated by the content of the protest, which was inextricably bound up with race. Taking her allegations as true, we can conceive of two possible variants of Dean’s indirect race-based theory. The first variant supposes that the class Warren’s conspiracy targeted was African Americans protesting police brutality against African Americans. The second supposes that the targeted class was African Americans. We next turn to the two variants of Dean’s indirect race-based theory and evaluate them under Bray.

First, based on Bray we must reject the possibility that a class of African Americans protesting police brutality against African Americans can support a § 1985(3) claim. Bray squarely held that the class of women seeking abortion was too gerrymandered to support a § 1985(3) claim because the class was defined merely by the conduct the defendants disfavored. Dean has given us no rationale

23

for distinguishing, at the qualifying class stage of the analysis, the class of African Americans protesting police brutality from the class of women seeking abortion, which the Supreme Court rejected as a qualifying § 1985(3) class. And we do not think such a line could persuasively be drawn.10 Second, we must reject the possibility that Dean can survive the class-based animus requirement if the qualifying class is African Americans. To be sure, African Americans is a qualifying class. The problem here, though, is that this version of Dean’s indirect race-based theory fails on a different ground: it does not clear the high bar established in Bray for alleging invidious discriminatory animus against that class.

Dean did not sufficiently allege that Warren was “motivated by a purpose . . . directed specifically at [African Americans] as a class,” that is, a “purpose that focuses upon [African Americans] by reason of their [race].” Id. at 270.11 So, she,

10 Indeed, Dean appears to acknowledge that the class of African Americans protesting police brutality is too gerrymandered to survive Bray. See Appellant’s Br. at 18 (characterizing the “protected class” underlying her race-based theories as “African American[s]”).

11 To understand why this is so, it is helpful to consider the Bray Court’s analysis of when a conspirator’s conduct is “directed specifically” at a class “as a class.” Bray, 506 U.S at 270. The Court explained that conduct preventing women from practicing law, justified by the stereotype that women are fragile, targets women “because they are women” and therefore satisfies the invidious discriminatory animus requirement. See id. By contrast, the demonstrations at issue in Bray were not “motivated by a purpose . . . directed specifically at women as a class” because the defendants “define[d] their [actions] not with reference to women” but with reference to the goal of preventing abortion. Id. In our view, Dean’s claim is closer to the abortion clinics’ claim in Bray than it is to the Court’s hypothetical concerning female lawyers prevented from practicing law because of their fragility. Like in Bray, the conduct Dean complains of is arguably a proxy for targeting a class protected by § 1985(3) and,

24

like the Bray plaintiffs, can survive the class-based animus requirement only indirectly, that is, if (1) opposition to cheerleaders kneeling during the national anthem to protest police brutality can reasonably be presumed to reflect a race- based intent or (2) Warren’s class-based animus can be inferred by discriminatory effect. Dean’s complaint fails on both counts.

In assessing whether the first proposition is true, we must follow Bray’s approach. Bray requires us to ask whether kneeling to protest police brutality against African Americans during the anthem is “such an irrational object of disfavor” and a practice “engaged in exclusively or predominantly by [members of the qualifying class—African Americans]” that an intent to disfavor African Americans “can readily be presumed” by opposition to that conduct. Id. In considering that question, we must assess whether, according to the standard outlined in Bray, there are “common and respectable reasons” for opposing the practice (kneeling during the national anthem to protest police brutality against African Americans)—reasons other than “hatred of, or condescension toward (or indeed any view at all concerning), [African Americans] as a class.” Id. We must

like in Bray, neither allegations nor evidence suggests the defendants were self-consciously intending to target a qualifying class because of the defendants’ beliefs about the class’s defining features. In the Court’s hypothetical, by contrast, there is direct targeting of women based on a stereotype rooted in their womanhood.

25

also assess whether class members and non-members “are on both sides of the issue.” Id.

In our view, this analysis dooms Dean’s claim just as it doomed the abortion clinics’ claim in Bray. Our experience informs us, at least as much as the Bray Court’s experience informed it, that there are members of all racial groups on both sides of the anthem-kneeling issue and that there are reasons to disfavor anthem kneeling that have nothing to do with hatred of, or condescension toward, African Americans. We readily acknowledge that some Americans oppose anthem kneeling out of racial animus. But we cannot deny that some believe anthem kneeling is unpatriotic and disrespectful and that no one should make any political statement, by any means, during a presentation of the national anthem. For others, their conception of patriotism may be entangled with their political beliefs, leading them to disfavor anthem-kneeling protests concerning race more than anthem- kneeling protests with different political messages. These individuals may, for example, believe that African Americans do not disproportionately suffer at the hands of police on account of race (but rather for some other reason) or that the policy changes advocated by groups supportive of kneeling would harm—rather than help—African American communities. The question is not whether we share these beliefs or whether we think they withstand scrutiny, but rather whether a sizable percentage of the American population shares them and whether, from

26

them, we “can readily . . . presume[]” a “hatred of, or condescension toward (or indeed any view at all concerning)” African Americans as a class. Id. We think a sizable percentage of Americans likely do hold these views, and from these views we cannot presume a hatred of or condescension toward African Americans as a class. Applying Bray, as we are bound to do, we conclude that the relationship between opposing anthem kneeling as a form of protesting police brutality against African Americans and hatred of, or condescension toward, African Americans is insufficient to demonstrate the animus that § 1985(3) requires.12 Dean’s indirect race-based theory, then, “comes down [to]” whether she can surmount the class-based animus requirement with a discriminatory effect theory. Id. at 271. We understand Bray to have foreclosed that possibility for Dean. See id. at 270 (holding that the proposition that the defendants’ “class-based animus [could] be determined solely by effect” was “[not] supportable”). For even if some plaintiff might prevail with such a theory, which is a question we have no occasion to answer, we cannot say Dean’s discriminatory effect argument is stronger than

12 We recognize that the inquiry into whether there are “common and respectable reasons” for opposing a practice may appear conjectural and dependent upon the moral judgments of the inquirer. And we recognize that the inquiry into whether members and non- members of the qualifying class are on both sides of the issue is both an empirical question that Bray asks us to undertake without record evidence and ill-defined. For instance, how many members of the qualifying class (or is it a percentage?) must be on the defendant’s side of the issue before a plaintiff becomes unable to establish animus? And how are we to determine that? In deciding this case, however, our duty is not to question Supreme Court precedent but to apply it.

27

the Bray plaintiffs’ discriminatory effect argument because we cannot say that anthem kneeling is engaged in more predominantly by African Americans than voluntary abortion is engaged in by women. The strongest version of Dean’s indirect race-based theory—that opposition to anthem kneeling creates a presumption of hatred toward African Americans or that the discriminatory effect of that opposition on African Americans can support a § 1985(3) claim—therefore fails.

Dean’s arguments to the contrary are unavailing. Rather than grapple with Bray, she attempts to sidestep it, arguing that it is distinguishable because it did not implicate race, which is the central focus of § 1985(3). Dean is correct that Bray did not implicate race and that “the central concern of Congress in enacting § 1985(3)” was class-based invidious animus “against [African Americans] and those who supported them.” Scott, 463 U.S. at 835. Even so, she has failed to persuade us either that (1) Bray’s analytical approach to § 1985(3) claims does not control when race is implicated or (2) Bray’s analytical approach applies, but is modified, in this context.

First, we think Bray’s discussion of § 1985(3)’s class-based, invidiously discriminatory animus requirement, which is comprised of the qualifying class element and the animus element, applies to § 1985(3) cases generally. See Bray, 506 U.S. at 271–72 & n.4 (concluding the “same principle” that applies in the

28

equal protection context—that discriminatory purpose requires “because of” not merely “in spite of” targeting—also “applies to the ‘class-based, invidiously discriminatory animus’ requirement of § 1985(3)” because that principle is “inherent in the requirement of a class-based animus”).

Second, assuming Bray’s analytical approach controls, we do not think the animus requirement—upon which Dean’s claim founders—is less demanding when the § 1985(3) plaintiff alleges a class implicating race. Dean has assuredly asserted a better qualifying class (African Americans) than the one grounding the § 1985(3) claim in Bray (women). Indeed, the class of African Americans is the strongest qualifying class imaginable because protecting African Americans and their advocates from the “massive, organized lawlessness that infected our Southern States during the post-Civil War era” was the chief purpose of § 1985(3)’s passage. Bray, 506 U.S. at 307 (Stevens, J., dissenting). We cannot say, however, that Bray offers us the latitude to apply a less demanding animus requirement to Dean’s claim on account of the strength of her qualifying class. Rather, Bray seems to have articulated the animus requirement for § 1985(3) claims across the board. See Bray, 506 U.S. at 274 (holding that, because opposition to abortion was not discrimination against women, the abortion clinics’

29

claim was “not the stuff out of which § 1985(3) ‘invidiously discriminatory animus’ is created”). 13 Under Bray, once we conclude or assume that the plaintiff has alleged a qualifying class, we turn to whether the animus element is satisfied. See Bray, 506 U.S. at 269–70 (assuming women are a qualifying class before turning to whether the animus requirement was satisfied). As we have discussed, Dean’s claim, construed as alleging the indirect race-based theory of liability, failed to allege animus under Bray. Her indirect race-based theory, therefore, fails to state a claim. C. Dean’s Political Class-Based Theory Dean’s political class-based § 1985(3) theory is also precluded by Bray. The theory is that Warren undertook the conspiracy to prevent Dean and her protesting teammates from exercising their First Amendment rights because Dean and her protesting teammates were members of a political class: protestors of police brutality against African Americans.

13 Although we do not parse judicial decisions as if they are statutes, see Bryan A. Garner et al., The Law of Judicial Precedent 2 & n.4 (2016) (explaining that not every grammatical detail of a precedential opinion is sacred), we find it informative that the Supreme Court’s formulations of the animus requirement consistently suggest that all § 1985(3) plaintiffs must allege invidiously discriminatory animus, see, e.g., Griffin, 403 U.S. at 102 (“racial, or perhaps otherwise class-based, invidiously discriminatory animus”); Scott, 463 U.S. at 834 (quoting Griffin, 403 U.S. at 102). Not only do the commas in Griffin’s oft-recited formulation suggest that a plaintiff alleging a racial class must also allege invidiously discriminatory animus, the language “otherwise class-based” suggests that a racial class is just one example of the set of classes § 1985(3) envisions and therefore that the invidious discriminatory animus requirement applies equally to other classes cognizable by § 1985(3) as it does to racial classes.

30

The Supreme Court has long reserved the question whether § 1985(3)

prohibits conspiracies targeting political classes. See Scott, 463 U.S. at 835–37 (“follow[ing] [Griffin’s] course” by withholding judgment on the question whether § 1985(3) covers conspiracies targeting political classes). And the courts of appeals are divided on this question. See Farber v. City of Paterson, 440 F.3d 131, 139 & n.7, 143 (3d Cir. 2006) (collecting cases discussing, and taking a position in, the split). We have never answered the question with binding precedent, and this case provides us with no opportunity to do so, for it can be resolved on narrower grounds. See PDK Lab’ys Inc. v. U.S. Drug Enf’t Admin., 362 F.3d 786, 799 (D.C. Cir. 2004) (Roberts, J., concurring) (“[Where] it is not necessary to decide more, it is necessary not to decide more . . . .”). 14 We need not decide whether political classes can ever serve as a basis for a § 1985(3) claim because the particular political class Dean alleged does not satisfy Bray. As we have explained, Bray forbids § 1985(3) plaintiffs from grounding

14 A published, binding case from our predecessor court suggested that conspiracies targeting political classes were actionable under § 1985(3). See Kimble v. D.J. McDuffy, Inc., 648 F.2d 340, 347 (5th Cir. June 1981) (en banc) (explaining that the class at issue in that case, persons who had filed worker’s compensation claims or personal injury suits against employers, was not—like classes “based on political beliefs or associations”—“envisioned by the framers of [§ 1985(3)]”); see also Bonner v. City of Prichard, 661 F.2d 1206, 1207 (11th Cir. 1981) (en banc) (adopting as binding all decisions of the former Fifth Circuit handed down prior to the close of business on September 30, 1981). We nevertheless regard the issue as undecided because this language in Kimble was dicta—the Fifth Circuit could have reached its conclusion that a class of persons who had filed worker’s compensation claims or personal injury suits against their employers was not cognizable under § 1985(3) without concluding that political classes are cognizable.

31

their claims on classes defined by the conduct the defendants oppose. 506 U.S. at 269 (rejecting the possibility that “women seeking abortion” constitutes an actionable class because “[w]hatever may be the precise meaning of a ‘class’ for purposes of Griffin’s speculative extension of § 1985(3) beyond race, the term unquestionably connotes something more than a group of individuals who share a desire to engage in conduct that the § 1985(3) defendant disfavors”). Dean has offered us no argument that, although women seeking abortion is not an actionable class, protestors of police brutality is—and we can think of none. See Farber, 440 F.3d at 136 (explaining that although “‘women,’ or ‘registered Republicans,’ may constitute an identifiable ‘class’ . . . a more amorphous group . . . such as ‘women seeking abortion’ or ‘persons who support political candidates’ [does not]” (alteration adopted) (citations omitted)); Aulson v. Blanchard, 83 F.3d 1, 5–6 (1st Cir. 1996) (holding that a § 1985(3) class “[must be] comprised of a distinctive and identifiable group” and therefore rejecting a class defined as “persons who support other persons opposed to the politics of the old guard” (internal quotation marks omitted)). We therefore reject Dean’s political class-based theory of § 1985(3) liability without opining on whether a more identifiable and discrete political class may ground a § 1985(3) claim.15

15 Rather than attempt to distinguish Bray’s holding that women seeking abortion is not an actionable § 1985(3) class, Dean seeks to strengthen her political class-based theory by citing

32

IV. CONCLUSION

For these reasons, we conclude that each theory by which Dean may have overcome § 1985(3)’s class-based animus requirement fails. We therefore affirm the district court’s grant of Warren’s motion to dismiss.

AFFIRMED.

two out-of-circuit, pre-Bray cases for the proposition that political classes are especially worthy of § 1985(3) protection when those classes—like the one at issue in this case—are intertwined with race. See Hobson v. Wilson, 737 F.2d 1 (D.C. Cir. 1984); Hampton v. Hanrahan, 600 F.2d 600 (7th Cir. 1979). We do not find Hobson and Hampton persuasive in light of Bray.

33

WILLIAM PRYOR, Chief Judge, joined by ED CARNES, Circuit Judge:

I join Judge Jill Pryor’s majority opinion in full. I write separately to explain why, even if Tommia Dean had properly stated a claim under section 1985(3), I would still affirm the dismissal of her complaint. Dean engaged in government speech when she cheered in a team uniform at a football game on behalf of her public university. Because the Free Speech Clause does not restrict government speech, Kennesaw State University did not violate her First Amendment rights when it prevented her from kneeling on the field. Without a First Amendment violation, Dean’s section 1985(3) claim fails. See 42 U.S.C. § 1985(3); Great Am. Fed. Sav. & Loan Ass’n v. Novotny, 442 U.S. 366, 372 (1979) (“Section [1985(3)] provides no substantive rights itself; it merely provides a remedy for violation of the rights it designates.”).

Government speech is not regulated by the Free Speech Clause. Pleasant Grove City v. Summum, 555 U.S. 460, 467 (2009). “When the government exercises the right to speak for itself, it can freely select the views that it wants to express. This freedom includes choosing not to speak and speaking through the removal of speech that the government disapproves.” Mech v. Sch. Bd. of Palm Beach Cnty., 806 F.3d 1070, 1074 (11th Cir. 2015) (alteration adopted) (citations and internal quotation marks omitted).

34

Although “we lack a precise test for separating government speech from private speech,” we are required to consider three factors: “history, endorsement, and control.” Cambridge Christian Sch., Inc. v. Fla. High Sch. Athletic Ass’n, 942 F.3d 1215, 1230 (11th Cir. 2019) (internal quotation marks omitted). The first factor, history, “ask[s] whether the type of speech under scrutiny has traditionally communicated messages on behalf of the government.” Id. at 1232 (internal quotation marks omitted). The second factor, endorsement, “asks whether the kind of speech at issue is often closely identified in the public mind with the government, or put somewhat differently, whether observers reasonably believe that the government has endorsed the message.” Id. at 1232–33 (citation and internal quotation marks omitted). And the third factor, control, “asks whether the relevant government unit maintains direct control over the messages conveyed through the speech in question.” Id. at 1234 (internal quotation marks omitted).

All three factors strongly suggest that cheerleading at Kennesaw State University is government speech. First, as a matter of history, cheerleaders have traditionally been understood to communicate an important message on behalf of their schools: cheer for our team to win the game. Cheerleading began in the second half of the nineteenth century “as an informal, sporadic activity” by enthusiastic spectators, but it soon “evolved” into an organized, school- administered program “as a way to intensify spectator involvement within the

35

increasingly structured context of formal athletic competition.” Mary Ellen Hanson, Go! Fight! Win!: Cheerleading in American Culture 10–12 (1995). “Just as elite athletes were singled out to compete for the college, cheerleaders representing the college were . . . selected for their tumbling skills and extroverted personalities.” Id. at 13. Today, “sideline cheerleaders generally strive to entertain audiences or solicit crowd reaction at sport or school functions.” Biediger v. Quinnipiac Univ., 691 F.3d 85, 103 (2d Cir. 2012).

Second, there is no doubt that Kennesaw State University endorses the message conveyed by its cheerleading team. On the website of its athletics department, the university says that its cheerleaders “are part of the spirit and tradition of [Kennesaw State University] Athletics. As student-athletes and ambassadors for Kennesaw State University, they promote school spirit at all home football . . . games, . . . school events, and community functions throughout the season.” Spirit Squad, Kennesaw State University Owls, https://ksuowls.com/sports/2019/1/3/spirit2019.aspx (all Internet materials as visited August 23, 2021, and available in Clerk of Court’s case file). They do so “on government property at government-sponsored school-related events.” Sante Fe Indep. Sch. Dist. v. Doe, 530 U.S. 290, 302 (2000). And as “ambassadors” of the university, they are expected to convey a message of which the university approves. The university’s endorsement of that message is even more apparent in

36

the context of the national anthem and other pre-game rituals, which are “inseparably associated with ideas of government.” Cambridge Christian, 942 F.3d at 1233. “In this context[,] the members of the . . . audience must perceive [that the message of the cheerleaders is] delivered with the approval of the school administration.” Sante Fe, 530 U.S. at 308.

Finally, Dean’s complaint does not allege that Kennesaw State University has relinquished control over the cheerleading squad or the message it conveys. Ordinarily, cheerleading is a school-sponsored activity managed by school- employed coaches and limited to students who successfully try out for the squad and commit to attending practices and games. Cheerleaders wear school uniforms and colors, receive school funding, train to perform choreographed acts, and cheer from the sidelines of school property where members of the public are ordinarily not allowed. It would be highly unusual for a public university to allow its cheerleaders—while they are in uniform on the field at a football game—to say or do whatever they please. For example, Auburn University does not allow its cheerleaders to root for its rival—and reigning national champion in football—the Crimson Tide of the University of Alabama, as much as they might want to do so.

Dean argues that the issue of government speech should not be resolved at the motion-to-dismiss stage, but she does not identify any facts left to be developed that could salvage her claim. “To survive a motion to dismiss, a complaint must

37

contain sufficient factual matter, accepted as true, to state a claim to relief that is plausible on its face.” Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009) (internal quotation marks omitted). Dean provided no facts in her complaint that would allow us “to draw the reasonable inference” that cheerleaders at Kennesaw State University are engaging in private speech when they wear school uniforms and cheer from the sidelines of a football game. Id. She did not allege a plausible violation of the First Amendment.

Absurd results would follow if the First Amendment protected Dean’s right to express herself while she is in uniform on the field. By this logic, Dean would have a right to perform her own unapproved, self-choreographed cheer, to cheer for the opposing team, or to refrain from cheering at all. She would also have a right to stage a hunger strike, to hold up campaign posters for a political candidate, to entertain the crowd with expressive dance, to cut up the American flag, or to wear a leather jacket over her cheerleader uniform with the words “f*** the draft” stitched onto the back. No one doubts that Dean has these rights as a citizen. But when she puts on a school uniform and joins her teammates on the sidelines of a game, she is expected to cheer for her school team. Although students do not “shed their constitutional rights to freedom of speech or expression at the schoolhouse gate,” Tinker v. Des Moines Indep. Cmty. Sch. Dist., 393 U.S. 503, 506 (1969), they do “voluntarily subject themselves” to some limitations when they try out for

38

the cheerleading squad, Vernonia Sch. Dist. 47J v. Acton, 515 U.S. 646, 657 (1995).

39

JILL PRYOR, Circuit Judge, concurring, Naturally, I agree with the reasoning and conclusions in the majority opinion I authored. I write separately to voice my discomfort with applying the test for class-based discriminatory animus from Bray v. Alexandria Women’s Health Clinic, 506 U.S. 263 (1993). In my view, the test requires federal courts to engage in an inquiry unsuited to the judiciary, and I question whether that inquiry is justified by 42 U.S.C. § 1985(3)’s text or any independent legal principle.

The majority opinion I authored describes in detail the inquiry Bray demands. See J. Pryor Maj. Op. at 18–29. I will review it briefly and explain why I believe it is an inquiry unfit for the judiciary to undertake.

A § 1985(3) plaintiff must allege “some racial, or perhaps otherwise class-

based, invidiously discriminatory animus behind the conspirators’ action.” Griffin v. Breckenridge, 403 U.S. 88, 102 (1971). Bray articulated the controlling analytical framework for evaluating whether a § 1985(3) plaintiff satisfied that requirement. See J. Pryor Maj. Op. at 18, 27–29. Under that framework, a plaintiff must first allege that a “qualifying class” was the target of the § 1985(3) defendant’s conspiracy. Bray, 506 U.S. at 269. Second, she must allege that invidiously discriminatory animus was “behind” the conspiracy’s actions against the class. Id. (internal quotation marks omitted). The Supreme Court and lower courts alike have struggled with the first inquiry, for it is difficult to discern what

40

types of classes qualify for § 1985(3) protection. 1 That inquiry is not the subject of my concern here, for there is no dispute that African Americans is the quintessential qualifying class under § 1985(3). But, as this case demonstrates, even when Bray’s qualifying class requirement is clearly met, applying Bray’s framework is vexing nonetheless because of Bray’s second inquiry.

Bray’s second inquiry requires a plaintiff to demonstrate animus either directly or indirectly. Id. at 269–70. A plaintiff demonstrates animus directly when she demonstrates that the defendant’s conduct is “motivated by a purpose (malevolent or benign) directed specifically at [the qualifying class] as a class,” that is, “a purpose that focuses upon [the qualifying class] by reason of [the characteristics defining that class].” Id. The Bray plaintiffs failed to allege direct animus because the anti-abortion activists targeted abortion, which the Court regarded as arguably a proxy (at best) for targeting women, and the record

1 See, e.g., United Bhd. of Carpenters & Joiners of Am., Local 610, AFL-CIO v. Scott, 463 U.S. 825 (1983) (describing it as a “difficult” and “close question” whether § 1985(3) extends beyond racially motivated conspiracies to politically motivated conspiracies); Bray, 506 U.S. at 295–96 (Souter, J., concurring in part and dissenting in part) (arguing that because equal protection principles apply to even “routine” legislative classifications, § 1985(3) “ought to cover discriminations that would be impermissible under rational-basis scrutiny”); Farber v. City of Paterson, 440 F.3d 131, 139 & n.7 (3d Cir. 2006) (describing the circuit split on the question whether animus against political classes qualifies for § 1985(3) relief); Compare Lyes v. City of Riviera Beach, 166 F.3d 1332, 1336–39 (11th Cir. 1999) (en banc) (holding that women are a protected class under § 1985(3)) with id. at 1351 (Tjoflat, J., concurring in part and dissenting in part) (concluding that § 1985(3) does not “cover[] conspiracies motivated by sex-based animus”).

41

suggested that the anti-abortion activists thought of themselves as motivated by an anti-abortion, rather than anti-women, sentiment. Id. at 270. In this case, Dean failed to allege direct animus because Warren targeted anthem kneeling, which is arguably a proxy (at best) for targeting African Americans, and Dean’s complaint alleged that Warren was motivated by anti-anthem kneeling, rather than anti- African American, sentiment. J. Pryor Maj. Op. at 23 n.11.

As Bray recognized, however, just because a defendant who targets an activity closely associated with a qualifying class does not acknowledge that her actions target the class (rather than merely the activity associated with the class) does not mean that the defendant should necessarily escape § 1985(3) liability. Bray, 506 U.S. at 270. This is because a defendant with invidious discriminatory animus, the Court explained, may have targeted a qualifying class indirectly. Id. If the circumstances lead to an inference of animus, the plaintiff should therefore be permitted to use § 1985(3) to hold the defendant accountable. Id. (explaining that even though the plaintiffs failed to demonstrate direct animus, § 1985(3) liability was still possible). So, the question becomes, under what circumstances might invidious discriminatory animus be inferred?

Here is where the analysis goes sideways. After Bray, the primary (and perhaps exclusive) means by which a plaintiff may demonstrate animus indirectly is if “opposition to [the targeted activity] can reasonably be presumed to reflect a

42

[class]-based intent.” Id. at 270–73 (rejecting the argument that § 1985(3) liability could be premised upon the discriminatory effect that targeting abortion providers had on women). The contours of that analysis are puzzling. Bray instructs us to consider whether the targeted activity is “such an irrational object of disfavor that, if [it is] targeted, and [if it] also happen[s] to be engaged in exclusively or predominantly by a particular class of people, an intent to disfavor that class can readily be presumed.” Id. at 270 (explaining that a tax on yarmulkes could be presumed to reflect animus toward Jews).

In my view, whether the behavior the defendant targeted is “engaged in exclusively or predominantly,” id., by the qualifying class’s members calls for an unusual degree of speculation and empirical guesswork from an appellate court tasked with answering a question of law. Although it is easy to conclude in the Bray Court’s example of the yarmulke tax that the predominant wearers of yarmulkes are Jews, most § 1985(3) cases are not so easy. This case is a compelling example: do African Americans predominate among anthem kneelers? 2

2 The predominance inquiry also raises a problem of scope. At what level of generality do we characterize the conduct the defendant targeted? In this case, instead of anthem kneeling, is it “anthem kneeling by athletes at the collegiate and professional level?” Or is it “kneeling to protest police brutality,” more generally? Depending on which version of the question we ask, the answer to whether class members predominantly engage in the targeted conduct may change. See Laura Douglas-Brown and Jen King, Emory Community Unites Against Racist Violence, Emory News Center (June 5, 2020), https://news.emory.edu/stories/2020/06/er_vigils_coverage/campus.html (reporting that, in an event unrelated to the national anthem, thousands of members of the Emory University

43

The predominance inquiry is not merely difficult to answer; it is also, I think, misguided. Asking whether the targeted activity is predominantly or exclusively engaged in by members of the protected class does not adequately capture whether the defendant is motivated by class-based animus. A defendant may be motivated by class-based animus even if her most immediate victims are not members of that class. “[A] racial conspiracy against [Black people] does not lose that character when it targets in addition white supporters of [B]lack rights.” Id. at 271 n.2. It should therefore make little difference to the question whether the anti-abortion activists in Bray were motivated by anti-women sentiment that some of the anti-abortion activists’ direct victims were male medical and support personnel. See id. By the same token, it should make little difference in this case to the question whether Warren was motivated by anti-Black sentiment that many individuals who engage in anthem-kneeling are white. Thus I fail to grasp why a targeted activity must be “exclusively or predominantly” practiced by class members for a court to infer the defendant was motivated by a “[class]-based intent.” Id. at 270.

These are not my only objections to Bray’s articulation of the animus element of the “racial, or perhaps otherwise class-based, invidious discriminatory

community gathered to kneel in silence for eight minutes and forty-six seconds to protest a notorious case of police brutality, the killing of George Floyd, a Black man who suffered for that period of time under the knee of a law enforcement officer later convicted of his murder).

44

animus” requirement. The predominance inquiry is only the iceberg’s tip. From there the test gets harder to apply and less well-suited to judicial decisionmaking. According to Bray, to determine whether animus can be presumed from the circumstances, we must also consider (to an unquantifiable degree) whether there “are common and respectable reasons for opposing [the targeted conduct],” reasons “other than hatred of, or condescension towards (or indeed any view at all concerning, [the class members].” Id. We consider, that is, whether the defendant’s conduct is a “surrogate for opposition” to the class members that is comparably “irrational” to the defendant’s motivation in the yarmulke hypothetical. Id.

And, from there, we add to the adjudicatory mix (again, to an unquantifiable degree) whether class members and non-members are “on both sides of the issue,” meaning both sides of the moral or political position that presumably motivates the defendant’s conduct. Id. In Bray that position was “opposition to abortion,” id., and, in this case, it is (as best I can tell) opposition to anthem kneeling as a political statement. See J. Pryor Maj. Op. at 24.

So, after undertaking the empirical (and in my view, misguided)

predominance inquiry, to faithfully apply Bray we must ask the “common and respectable reasons” question and the “both sides” question. The former is a mixed descriptive/normative exercise, for we must estimate whether a sufficient

45

(undefined) number of Americans hold the views held by the defendant (that is, whether the defendant’s hypothetical reasons are “common”) and whether the defendant’s hypothetical reasons for opposing the conduct are worthy of respect (that is, whether the defendant’s reasons are “respectable”). 3 Quite an inquiry— but we are not finished yet.

Before concluding the inquiry, we must consider yet another empirical matter—whether class members and non-members are on “both sides of the issue.” Id. at 25. But how many class members must be on the defendant’s “side” (or how many non-class members must be on the plaintiff’s “side”) before the “both sides” inquiry points to an answer? Bray gives us no guidance.

It seems to me that the inquiries Bray calls for are better suited to pollsters and perhaps political philosophers than judges. That should be particularly troubling in this context, where cases have concerned, and will inevitably concern, profound and controversial social issues. See Griffin, 403 U.S. at 90 (racial terrorism); Scott, 463 U.S. at 827–28 (labor unions); Bray, 506 U.S. at 266 (abortion); Farber v. City of Paterson, 440 F.3d 131, 133 (3d Cir. 2006) (partisan politics); see also Bray, 506 U.S. at 325 (Stevens, J., dissenting) (“The [Bray] Court’s view requires a subjective judicial interpretation inappropriate in the civil

3 Respectable, Merriam-Webster’s Unabridged Dictionary, https://unabridged.merriamwebster .com/unabridged/respectable (last visited Sept. 1, 2021) (defining “respectable” as “worthy of esteem or deference”).

46

rights context, where what seems rational to an oppressor seems equally irrational to a victim.”).

Our obligation to engage in Bray’s inquiry would be lamentable, though palatable, if § 1985(3)’s text or some independent legal principle required it. But I have grown concerned that neither does.

Griffin’s class-based animus requirement was “derived from the statute’s legislative history,” as Justice O’Connor observed in her dissent in Bray, and Bray’s “fine[]” “pars[ing]” of Griffin made the class-based animus requirement an “‘element’ of the § 1985(3) cause of action that does not appear on the face of the statue.” See Bray, 506 U.S. at 347 (O’Connor, J., dissenting). As far as independent legal principles, the Court’s § 1985(3) jurisprudence has been dominated by constitutional concerns. See Griffin, 403 U.S. at 95–96, 101–02 (holding that the class-based animus requirement was necessary to avoid perceived “constitutional shoals”); see also Bray, 506 U.S. at 299 (Souter, J. concurring in part and dissenting in part) (explaining that the Court’s “concerns with constitutionality . . . most probably left a lesser [§ 1985(3)] than Congress intended”). The constitutional concerns necessitating the class-based animus requirement have been mostly expressed as the platitude that § 1985(3) must not be permitted to become a “general federal tort law.” Griffin, 403 U.S. at 102. I wonder whether there is truly a real danger of turning the statute into a general

47

federal tort law, however, given two narrowing principles: that § 1985(3) “creates no rights” but is rather a vehicle for enforcing a not-yet-defined set of federal rights, Great Am. Fed. Sav. & Loan Ass’n v. Novotny, 442 U.S. 366, 376 (1979), and that it requires the plaintiff to demonstrate that the defendant “aimed at” the right, i.e., that “impairment [of the right was] a conscious objective of the enterprise,” Bray, 506 U.S. at 275.4 At minimum, if there are sound constitutional reasons for giving § 1985(3) a constricted construction, they should be articulated—not merely gestured at. See Brett M. Kavanaugh, Fixing Statutory Interpretation, 129 Harv. L. Rev. 2118, 2146 (2016) (book review) (discussing the dangers inherent in invoking constitutional avoidance when there are “mere questions of unconstitutionality rather than actual unconstitutionality”).

I understand Bray to demand our panel’s result. But I worry that § 1985(3)

jurisprudence has gone awry and deserves another look.

4 The Supreme Court has not described the full set of rights that may be vindicated by § 1985(3), but it has suggested that a § 1985(3) action aimed at vindicating First Amendment rights is viable if state action is implicated. See Scott, 463 U.S. at 833 (explaining that because “the right claimed to have been infringed ha[d] its source in the First Amendment . . . [which] restrains only official conduct, to make out their § 1985(3) case, it was necessary for [plaintiffs] to prove that the state was somehow involved in or affected by the conspiracy”).