Speech First, Inc. v. Timothy Sands

69 F.4th 184

Opinion

PUBLISHED

UNITED STATES COURT OF APPEALS FOR THE FOURTH CIRCUIT

No. 21-2061

SPEECH FIRST, INC., Plaintiff – Appellant,

v.

TIMOTHY SANDS, in his individual capacity and official capacity as President of Virginia Polytechnic Institute and State University,

Defendant – Appellee.

------------------------------

SOUTHEASTERN LEGAL FOUNDATION; LIBERTY JUSTICE CENTER; CATO INSTITUTE; AMERICAN COUNCIL OF TRUSTEES AND ALUMNI; FOUNDATION FOR INDIVIDUAL RIGHTS IN EDUCATION; ALLIANCE DEFENDING FREEDOM,

Amici Supporting Appellant.

Appeal from the United States District Court for the Western District of Virginia, at Roanoke. Michael F. Urbanski, Chief District Judge. (7:21-cv-00203-MFU)

Argued: October 25, 2022 Decided: May 31, 2023

Before WILKINSON and DIAZ, Circuit Judges, and MOTZ, Senior Circuit Judge.

Affirmed by published opinion. Senior Judge Motz wrote the opinion, in which Judge Diaz joined. Judge Wilkinson wrote a dissenting opinion.

2

ARGUED: James Hasson, CONSOVOY MCCARTHY PLLC, Arlington, Virginia, for Appellant. William H. Hurd, ECKERT SEAMANS CHERIN & MELLOTT, LLC, Richmond, Virginia, for Appellee. ON BRIEF: J. Michael Connolly, Cameron T. Norris, CONSOVOY MCCARTHY PLLC, Arlington, Virginia, for Appellant. Matthew B. Kirsner, Richmond, Virginia, Michael McAuliffe Miller, Renee Mattei Myers, Harrisburg, Pennsylvania, Jessica A. Glajch, ECKERT SEAMANS CHERIN & MELLOTT, LLC, Washington, D.C., for Appellee. Celia Howard O’Leary, Kimberly S. Hermann, SOUTHEASTERN LEGAL FOUNDATION, Roswell, Georgia, for Amicus Southeastern Legal Foundation. Ilya Shapiro, Thomas A. Berry, CATO INSTITUTE, Washington, D.C.; Reilly Stephens, Daniel Suhr, Jeffrey Jennings, LIBERTY JUSTICE CENTER, Chicago, Illinois, for Amici Liberty Justice Center, Cato Institute, and American Council of Trustees and Alumni. Darpana M. Sheth, Ronald London, Jeffrey D. Zeman, FOUNDATION FOR INDIVIDUAL RIGHTS IN EDUCATION, Philadelphia, Pennsylvania, for Amicus Foundation for Individual Rights in Education. Gordon D. Todd, Mackenzie J. Siebert, Robert M. Smith, SIDLEY AUSTIN LLP, Washington, D.C.; John J. Bursch, Cody S. Barnett, ALLIANCE DEFENDING FREEDOM, Washington, D.C., for Amicus Alliance Defending Freedom.

3

DIANA GRIBBON MOTZ, Senior Circuit Judge:

Speech First, Inc., which identifies itself as a national organization committed to protecting the rights of college students, initiated this action against Timothy Sands, the President of the Virginia Polytechnic Institute and State University (Virginia Tech or the University). Speech First asserts that two Virginia Tech policies — the Bias Intervention and Response Team Policy (the Bias Policy) and the Informational Activities Policy — violate the First Amendment rights of its student members who attend Virginia Tech. Just four days after filing suit, Speech First asked the district court to preliminarily enjoin both policies. In a comprehensive, 53-page opinion, based on numerous findings of fact, the court refused to do so. The district court held that Speech First (1) lacked standing to challenge the Bias Policy because its members had suffered no injury in fact, and (2) failed to demonstrate a likelihood of success on the merits as to the Informational Activities Policy because the record was, at that time, inadequate as to that policy. Speech First now appeals. For the reasons that follow, we affirm.

I.

The district court’s express findings, and uncontradicted record evidence, establish the facts recounted here. Speech First does not challenge any of these facts.

The Bias Policy defines bias incidents as “expressions against a person or group because of the person’s or group’s age, color, disability, gender (including pregnancy), gender identity, gender expression, genetic information, national origin, political affiliation, race, religion, sexual orientation, veteran status, or any other basis protected by law.” Speech First, Inc. v. Sands, No. 21-00203, 2021 WL 4315459, at *8 (W.D. Va. Sept.

4

22, 2021). Virginia Tech concluded that such incidents are detrimental to the University community and accordingly established a Bias Policy that allows members of the University community to report incidents of bias that occur at Virginia Tech.

Complaining persons may choose to identify themselves by name, or they can file a report anonymously. Id. at *9. To make students feel comfortable reporting bias incidents, Virginia Tech has publicized the Bias Policy through a “See something? Say something!” campaign. The University accepts bias reports filed through various online platforms.

Once submitted, all reported bias incidents go to a panel of university administrators known as the Bias Intervention and Response Team (the BIRT). Id. at *8. The BIRT includes representatives from the Fraternity and Sorority Life Office, the Services for Students with Disabilities Office, the Virginia Tech Police Department, and the Housing and Residence Life Office. The district court expressly found that the “BIRT lacks any authority to discipline or otherwise punish students for anything.” Id. at *10. None of the BIRT’s interactions with students — whether a complaining student or a responding student — ever appear on a student’s academic transcript or disciplinary record. 1 Rather, as Dean Hughes explained in his written declaration, all BIRT-related records and

1

The BIRT is overseen by Virginia Tech’s Dean of Students Office. The Dean of Students, Dr. Byron Hughes, declared in writing that his office has “no role in student discipline,” but exists instead to “offer advising, care, and support to students.” It is the Student Conduct Office that “administers the Student Code of Conduct, which is ‘the exclusive process by which students are adjudicated to be in violation of University policy and sanctioned for any such violations.’” Sands, 2021 WL 4315459, at *9. Speech First does not seek to enjoin any portion of the Student Code of Conduct or the operations of the Student Conduct Office.

5

correspondence are kept separately within the Dean of Students Office’s case management system.

The BIRT meets weekly and reviews any newly submitted bias incident reports. At the outset of its review, the BIRT considers whether each bias incident report involves First Amendment-protected speech. Some complaints are dismissed outright because they do. In fact, the district court identified only two occasions when the BIRT referred protected speech to the Student Conduct Office and, in both cases, the Student Conduct Office concluded that the speech was constitutionally protected and so did not sanction the accused students. Id.

Virginia Tech believes that some complaints that do not violate the law or the Student Code of Conduct may nonetheless present an educational opportunity. In such cases, the BIRT sends a letter to both the complaining student and the responding student that invites them to take part in a voluntary conversation facilitated by an administrator in the Dean of Students Office. 2 Attendance at these meetings is entirely voluntary; if a student ignores the message or refuses to meet with the Dean of Students Office, the district

2

The BIRT also refers some “localized” bias incidents to non-punitive campus entities. For instance, “a complaint alleging an incident of bias between two roommates may be referred to Housing and Residence Life to address.” The BIRT takes a different approach to incidents involving speech that independently may violate the law or the Student Code of Conduct. The record reflects that the BIRT refers complaints “alleging criminal activity . . . to the Virginia Tech Police Department” and complaints “describing an incident that may violate the Student Code . . . to the Student Conduct Office.” But the record also demonstrates that the BIRT does not have any “special authority to refer cases.” Indeed, the district court found that any member of the university community (for example, students, faculty, or administrators) can report any incident directly to the Student Conduct Office or the Virginia Tech Police Department without ever involving the BIRT. See Sands, 2021 WL 4315459, at *12.

6

court found that “no further action is taken, and the student faces no consequences of any kind.” Id.

The second policy Speech First seeks to enjoin, the Informational Activities Policy, provides:

Informational activity is defined as the distribution of literature and/or petitioning for signatures where no fee is involved nor donations or contributions sought. Informational activities may be permitted if they are sponsored by a university-affiliated organization. Such activities require prior approval by the designated university scheduling office and are subject to university policies and the reasonable guidelines of the authorizing official.

In simple terms, the Informational Activities Policy regulates leafletting and the solicitation of signatures on campus.

The district court found that this policy requires students to make reservations with the Student Engagement and Campus Life Office (the Campus Life Office) to leaflet and solicit signatures at designated locations throughout campus. Id. at *22. The Campus Life Office awards such reservations for free, in the order they are received. Id. Students seeking to engage in informational activities must be “sponsored” by one of the university’s 750+ registered student organizations (RSOs). A student is not required to belong to an RSO as a member to leaflet on campus. Virginia Tech Br. at 14 & n.8. The student must simply secure the sponsorship of any existing RSO. Id. “RSOs may sponsor informational activities unrelated to their official purpose.” Id. at 14 n.8. For instance, “[t]he Chess Club could sponsor leaflets criticizing the government’s immigration policies; the Cave Club could sponsor leaflets supporting those policies.” Id.

7

Speech First, however, is not an RSO at Virginia Tech. Its members do not want to become an RSO (or seek an existing RSO’s sponsorship) but “want to independently distribute literature about conservative ideas and collect signatures for petitions that support conservative causes.” 3 Sands, 2021 WL 4315459, at *23. They assertedly refrain from doing so out of fear that they “will be punished for engaging in ‘informational activities’ without [prior approval and] the sponsorship of a ‘university-affiliated organization.’” Id.; Speech First Opening Br. at 13.

As noted above, Speech First has not and does not challenge any of these facts. But Speech First asserts that these facts establish that the Bias Policy caused its members to “censor their speech” and that the Informational Activities Policy is an impermissible prior restraint and speaker-based restriction, and therefore both policies violate the First Amendment. Speech First Opening Br. at 11–13. For this reason, Speech First maintains that the district court abused its discretion in refusing to preliminarily enjoin both policies.

II.

We therefore must determine whether the court did, in fact, abuse its discretion in denying Speech First’s motion for a preliminary injunction. See Di Biase v. SPX Corp.,

3

In the district court, Speech First submitted anonymous declarations so stating on behalf of three of its student members (identified as Students A, B, and C). Because none of those students still attend Virginia Tech, the University moved to dismiss this action as moot and Speech First moved to supplement the appellate record with affidavits from four additional students (identified as Students D, E, F, and G) who assert identical injuries and are currently enrolled at Virginia Tech. We denied the University’s motion to dismiss and granted Speech First’s motion to supplement. We thus consider the declarations of Students D–G as part of the record on appeal. See ECF No. 76; Fed. R. App. P. 10(e); Loc. R. 10(d).

8

872 F.3d 224, 229 (4th Cir. 2017). In making this determination, we examine the district court’s factual findings for clear error and consider its legal conclusions de novo. Id.; see also dmarcian, Inc. v. dmarcian Eur. BV, 60 F.4th 119, 138 (4th Cir. 2023). When a party moves for a preliminary injunction, as Speech First did, it invites the district court to act as the finder of fact on a limited record. See Fed. R. Civ. P. 52(a); Scotts Co. v. United Indus. Corp., 315 F.3d 264, 274 (4th Cir. 2002). The moving party also bears the burden of demonstrating that it is likely to succeed on the merits. Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008).

When reviewing the denial of a preliminary injunction, an appellate court must credit the district court’s factual findings unless clearly erroneous. 4 Scotts Co., 315 F.3d at 274; see also dmarcian, Inc., 60 F.4th at 138. Such deference is appropriate because district courts enjoy a comparative “advantage in hearing and ‘weighing’ evidence.” United States v. Shea, 989 F.3d 271, 277 (4th Cir. 2021). Thus, unless, in denying a preliminary injunction request, the district court “rest[ed] its decision on a clearly erroneous finding of a material fact, or misapprehend[ed] the law with respect to underlying issues in litigation,” we defer to its judgment. Leaders of a Beautiful Struggle

4

Of course, if instead the district court had granted a motion by Virginia Tech to dismiss Speech First’s complaint, we would have a very different obligation — namely, we would be bound to construe the complaint’s factual allegations most favorably to plaintiff Speech First. As Judge Quattlebaum recently explained in Menders v. Loudoun County School Board, 65 F.4th 157, 160 (4th Cir. 2023), “in reviewing an order granting a motion to dismiss, we accept the . . . facts from the amended complaint and the incorporated exhibits as true and draw all reasonable inferences from them in favor of the [plaintiffs].” See also Int’l Refugee Assistance Project v. Trump, 961 F.3d 635, 648 (4th Cir. 2020) (noting that “important differences exist between the two standards”).

9

v. Balt. Police Dep’t, 2 F.4th 330, 339 (4th Cir. 2021) (en banc) (quoting In re Search Warrant Issued June 13, 2019, 942 F.3d 159, 171 (4th Cir. 2019), as amended (Oct. 31, 2019)). 5 The record before us shows that the district court took seriously its factfinding responsibility. The court considered hundreds of pages of exhibits, sworn declarations from Virginia Tech students and administrators, various campus policies and internal campus documents, and screenshots taken from Virginia Tech’s website. Speech First itself submitted much of this evidence. The district court also provided the parties an opportunity for a lengthy oral argument on the preliminary injunction motion. After considering the totality of the record before it, the court issued a thorough opinion setting forth its factual findings and legal conclusions. We turn now to review of that opinion and the substance of Speech First’s challenge.

5

As the dissent recognizes, Speech First mounts facial (rather than as-applied)

challenges to the policies here. See Dis. Op. at 36; see also Speech First Opening Br. at 17, 33. The dissent concedes, as it must, that “facial challenges are disfavored.” Dis. Op. at 36 (citing Wash. State Grange v. Wash. State Republican Party, 552 U.S. 442, 450 (2008)). Indeed, in United States v. Chappell, 691 F.3d 388, 392 (4th Cir. 2012) (Wilkinson, J.), this court rejected a First Amendment facial challenge, cautioning that such challenges rely on speculation, risk premature judicial interpretation of government actions, and run counter to principles of judicial restraint. So it is here.

10

III.

Speech First initially contends that the district court erred in concluding that it lacks standing to challenge the Bias Policy on behalf of its student members. 6 See Lujan v. Defs. of Wildlife, 504 U.S. 555, 560 (1992) (explaining that “standing is an essential and unchanging part of the case-or-controversy requirement of Article III”). Without standing to sue, Speech First cannot show that it is likely to succeed on the merits. “To establish standing, a plaintiff must show: (1) an injury in fact; (2) a sufficient causal connection between the injury and the conduct complained of; and (3) a likelihood that the injury will be redressed by a favorable decision.” Wikimedia Found. v. NSA, 857 F.3d 193, 207 (4th Cir. 2017). The district court held that Speech First’s members had suffered no injury in fact, and thus Speech First lacked standing, because the Bias Policy does not proscribe any constitutionally protected activity. 7 Indeed, the court concluded the Bias Policy does not proscribe “anything at all.” Sands, 2021 WL 4315459, at *10.

A.

Speech First grounds its challenge in the First Amendment, which protects free speech and expression. See U.S. Const. amend. I (“Congress shall make no

As a membership organization, Speech First has standing as a representative of its

6

members if its members would have standing to sue in their own right. See S. Walk at Broadlands Homeowner’s Ass’n, Inc. v. OpenBand at Broadlands, LLC, 713 F.3d 175, 183–84 (4th Cir. 2013).

7

We recently explained in another school speech case that “[t]he limitation of federal court jurisdiction to actual cases or controversies is a bedrock principle fundamental to our judiciary’s role in our system of government . . . [and] ‘[o]ne element of the case- or-controversy requirement is that [plaintiffs] . . . must establish that they have standing to sue.’” Menders, 65 F.4th at 162 (quoting Raines v. Byrd, 521 U.S. 811, 818 (1997)).

11

law . . . abridging the freedom of speech . . . .”). It “prohibits the state from interfering with the expression of unpopular, indeed offensive, views.” Nat’l Socialist White People’s Party v. Ringers, 473 F.2d 1010, 1016 (4th Cir. 1973) (en banc). And of course, “state colleges and universities are not enclaves immune from the sweep of the First Amendment.” Healy v. James, 408 U.S. 169, 180 (1972).

Direct prohibitions of speech and expression naturally invite constitutional scrutiny.

For instance, as we explained in Benham v. City of Charlotte, 635 F.3d 129, 135 (4th Cir. 2011), “[a]t the most basic level, the denial of a permit to engage in expressive activity altogether constitutes a First Amendment injury.” Here, Speech First does not contend that the Bias Policy established an outright ban on speech. But speech need not be banned outright to trigger First Amendment protections. Individuals suffer a concrete injury even when the state has simply “chilled” the right to engage in free speech and expression. See, e.g., Cooksey v. Futrell, 721 F.3d 226, 235 (4th Cir. 2013). That is what Speech First maintains the University did here in instituting the Bias Policy.

A First Amendment claim premised on chilling speech, like this one, is cognizable only when the asserted chill “would likely deter ‘a person of ordinary firmness’ from the exercise of First Amendment rights.” Constantine v. Rectors & Visitors of George Mason Univ., 411 F.3d 474, 500 (4th Cir. 2005); see also Laird v. Tatum, 408 U.S. 1, 13–14 (1972). “Allegations of a subjective ‘chill’ are not an adequate substitute for a claim of specific present objective harm or a threat of specific future harm.” Laird, 408 U.S. at 13– 14. Such allegations, standing alone, cannot demonstrate an injury in fact. See Benham, 635 F.3d at 135. Thus, Speech First has standing to lodge this challenge to the Bias Policy

12

only if its members’ asserted self-censorship is objectively reasonable. See id.; Laird, 408 U.S. at 13–14.

Speech First advances two interrelated arguments as to why its members acted in an objectively reasonable manner in assertedly self-censoring. First, it maintains that the University has used implicit threats to deter unfavored speech. Speech First Opening Br. at 24. Second, Speech First argues that Virginia Tech has imposed a burdensome administrative regime that would cause an objectively reasonable student to refrain from engaging in politically charged speech. Id. We consider each argument in turn.

1.

As to its initial contention, Speech First asserts that “the entire point of the BIRT is to implicitly threaten students with discipline if they say something ‘biased.’” Id. The district court rejected this contention, expressly finding that the “BIRT lacks any authority to discipline or otherwise punish students for anything.” Sands, 2021 WL 4315459, at *10.

We recently explained why this factual finding is critical in considering another student speech challenge to a university policy. See Abbott v. Pastides, 900 F.3d 160, 176 (4th Cir. 2018). There, we pointed out that a “credible threat of enforcement” is the sine qua non of a speech chilling claim. Id. Accordingly, a putative plaintiff suffers no cognizable injury if she lacks an “objectively good reason for refraining from speaking and ‘self-censoring’ instead.” Id.

The district court found that the BIRT’s interaction with students is limited to offering an optional “educational opportunity for better understanding protected speech and the role of tolerance in the campus community.” Sands, 2021 WL 4315459, at *8, *10.

13

The court further found that the Bias Policy does “not proscribe anything at all,” or require anything of anyone. Id. at *10. In fact, the Bias Policy espouses tolerance for competing views, assuring that “[a]ll community members,” including those accused of perpetrating bias, “will be treated with respect, consideration, concern, and care.” Id. at *9. Some bias incidents do not involve protected speech and “may be adjudicated through the student conduct process.” Id. But the district court found that Virginia Tech does not and cannot adjudicate matters involving protected speech. Id. at *9, *12.

Speech First attempts to liken the Bias Policy to the coercive tactics found unconstitutional in Bantam Books, Inc. v. Sullivan, 372 U.S. 58 (1963). There, the Court held that “informal sanctions,” including “the threat of invoking legal sanctions and other means of coercion, persuasion, and intimidation,” were sufficient to confer a First Amendment injury. Id. at 67, 71–72. In Bantam Books, a state entity known as the “Commission to Encourage Morality in Youth” sent to book and magazine publishers “notices, phrased virtually as orders,” urging them to stop publishing disfavored materials. Id. at 59–63, 68. Those notices included language thanking each publisher “for his ‘cooperation’ with the Commission” and reminding each that the Commission was mandated “to recommend to the Attorney General prosecution of purveyors of obscenity.” Id. at 62. The Commission also circulated to local police departments lists of objectionable publications, and local police officers commonly paid personal visits to offending publishers “to learn what action [the publisher] had taken” in response to the Commission’s notice. Id. at 63. The Court held the Commission’s activities unconstitutional because their purpose and effect were to deter disfavored speech. Id. at 67, 71–72.

14

The BIRT shares little common ground with the Bantam Books Commission and the Commission’s attempt to suppress publications it deemed objectionable. The Bantam Books Commission, unlike the BIRT, wielded great coercive authority. The Commission sent police officers to pay personal visits to disfavored publishers. Id. at 63. And it issued written orders to those publishers it sought to threaten, reminding them that the Commission had an obligation to recommend criminal prosecution for “purveyors of obscenity.” Id. at 62. One publisher had received “at least 35 such notices at the time [the] suit was brought.” 8 Id. at 61. And critically, the trial court in Bantam Books “found as a fact” that “compliance with the Commission’s directives was not voluntary.” Id. at 68.

Here, the district court found as a fact that the BIRT does not mandate involuntary compliance or anything of the sort. Rather, the court found as a fact that “[a]ll that [the] BIRT has the authority to do is to ‘invite’ students to participate in a ‘voluntary conversation’ about the alleged bias or refer reports elsewhere.” Sands, 2021 WL 4315459, at *10. The record establishes that the BIRT does not even extend an invitation for a voluntary conversation in response to every complaint it receives. Rather, the BIRT often dismisses complaints because they involve constitutionally protected activity.

8

Speech First seems to suggest that the University’s “See something? Say something!” posters serve the same purpose as the orders and circulated lists in Bantam Books. The comparison is unpersuasive. In its posters, the University does not threaten or forbid anything. Indeed, Virginia Tech’s posters, unlike the Bantam Books Commission’s threats, are directed not to those who utter constitutional speech but to those who may complain about it. Thus, the University promotes its bias reporting mechanism not to chill its students’ speech but instead to encourage civility. The University’s posters do not state, or even suggest, that disciplinary sanctions will be imposed on anyone because of their speech.

15

The district court expressly found that even when the BIRT does extend an invitation to meet, there is “no evidence that students feel obligated to come to these voluntary meetings” with the Dean of Students. Id. at *12. None of Speech First’s student members have offered any evidence that they (or their peers) feel pressured to attend the meetings. 9 For these reasons, the district court found that the BIRT neither imposes discipline nor suggests in any way that it can impose discipline. Thus, Speech First’s reliance on Bantam Books fails. 10 Speech First further argues that Virginia Tech students are threatened by the BIRT’s special “referral power.” But in truth, the BIRT’s ability to refer matters is neither special nor much of a power. As to it being “special,” the district court found that the “BIRT may report a Student Code violation just like any other member of the Virginia Tech

9

The dissent conceals this lack of evidence with fiery rhetoric and more than a dozen hypothetical questions. Neither provide a sound basis for rejecting the district court’s explicit findings of fact. Indeed, our friend in dissent has explained that the disfavored nature of facial challenges (like the one at hand) cannot be overcome by “hypotheticals . . . conjure[d] up” for support. See Chappell, 691 F.3d at 393.

10

The remaining cases on which Speech First relies offer it even less support. The Second Circuit’s opinions in Levin v. Harleston, 966 F.2d 85 (2d Cir. 1992) and Okwedy v. Molinari, 333 F.3d 339 (2d Cir. 2003), concerned high-ranking government officials who threatened to use their expansive authority to punish the speakers. By contrast, here the district court found that the BIRT is administered by a university department — the Dean of Students Office — that, unlike the government officials in those cases or even the Student Conduct Office at Virginia Tech, has no enforcement authority and “no role in student discipline.” Sands, 2021 WL 4315459, at *9. For this reason, the district court found that the bias incident reporting process is entirely separate from a student’s academic and disciplinary records. See id. at *10. The asserted threat here is also distinct from the one presented in Backpage.com, LLC v. Dart, 807 F.3d 229, 231 (7th Cir. 2015). Unlike the local sheriff in that case, Speech First does not contend that the BIRT accuses anyone of facilitating odious criminal acts, or that it sends students anything like “cease and desist” letters printed on official law enforcement letterhead.

16

community.” Id. (emphasis added). As to the “power” of the BIRT referral, unlike any other member of the university community, who could report any conduct, the Bias Policy permits the BIRT only to refer conduct that is not constitutionally protected and that independently may violate the Student Code of Conduct or the law. In short, because the Bias Policy has specifically disclaimed the ability to adjudicate matters involving protected speech, the policy is far from a “thinly veiled threat[] to institute criminal [or other disciplinary] proceedings,” like that in Bantam Books. 372 U.S. at 68.

Speech First also did not offer any evidence that BIRT referrals occur with any frequency, or that they are more likely to result in discipline than referrals from other members of the University community. In fact, the only example of a BIRT referral to which Speech First points ended with the referred student receiving no sanctions. Compare Speech First Opening Br. at 26 with Sands, 2021 WL 4315459, at *10. Moreover, when the BIRT refers an incident to another University entity, that other entity also must determine whether the respondent’s conduct is constitutionally protected. The district court found that this process “shows that if one part of the university makes a mistake, the university as a whole will ensure that protected speech remains protected.” Sands, 2021 WL 4315459, at *12.

In light of the above facts, the district court did not clearly err in finding that the “BIRT lacks any authority to discipline or otherwise punish students for anything.” Id. at *10. We recently held in Abbott that a plaintiff asserting First Amendment chill cannot prevail unless it demonstrates that a challenged policy conveys a “credible threat of enforcement.” 900 F.3d at 176. We must follow that holding here, and it requires rejection

17

of Speech First’s argument. Accordingly, as the district court found, Speech First’s student members have failed to allege that they have sustained an injury in fact.

2.

We consider next Speech First’s alternative argument that Virginia Tech, through the BIRT, has created an elaborate bureaucratic regime that burdens the exercise of free speech. Speech First essentially maintains that the BIRT process itself is a form of punishment.

Again guided by our recent decision in Abbott, we cannot agree. There, Ross Abbott, a student at the University of South Carolina, received a university official’s approval to host a “Free Speech Event” involving displays of offensive words and symbols, evidently in an attempt to “draw attention to the various threats to free speech on [college] campuses.” Id. at 164–65. Immediately thereafter, a different university official informed Abbott that the university had received complaints about Abbott’s event and therefore ordered Abbott to “arrange an appointment to fully discuss the charges as alleged.” Id. at 165. Abbott was also told “not [to] contact the named complainant, or discuss the complaints with any member of the University community.” Id. The university warned him that “[i]f the matter could not be resolved otherwise,” the university would investigate him formally. Id. Abbott, like Speech First, asserted that the university’s actions would cause a reasonable student to self-censor. Id. at 169. Abbott mounted both an as-applied challenge and a facial challenge; the discussion in Abbott of the latter is instructive here.

Speech First’s legal theory in support of its facial challenge to Virginia Tech’s Bias Policy is largely the same as the theory advanced by Abbott to challenge the University of

18

South Carolina’s policy. See id. at 178. Like Abbott, Speech First contends that the challenged policy creates a process that is so administratively burdensome in and of itself that it creates a First Amendment injury. We rejected Abbott’s argument and, for the same reasons, we reject Speech First’s similar argument.

In Abbott, echoing the Supreme Court’s discussion in Susan B. Anthony List v.

Driehaus, 573 U.S. 149 (2014), we acknowledged that some administrative processes may be so onerous as to amount to a sanction, and thus may be sufficient to confer First Amendment standing. 900 F.3d at 178. We cited as examples a “‘significantly intrusive’ FBI field investigation into a plaintiff’s political beliefs and personal life,” as well as a federal agency’s “extraordinarily intrusive and chilling” eight-month-long investigation “into certain plaintiffs’ protected speech and beliefs.” Id. at 179 (citing Clark v. Library of Congress, 750 F.2d 89, 92–95 (D.C. Cir. 1984); White v. Lee, 227 F.3d 1214, 1228, 1237–38, 1242 (9th Cir. 2000)). But we emphasized that a “threatened administrative inquiry will not be treated as an ongoing First Amendment injury sufficient to confer standing unless the administrative process itself imposes some significant burden, independent of any ultimate sanction.” Id. at 179.

Applying this standard in Abbott, we explained:

Even an objectively reasonable “threat” that the plaintiffs might someday have to meet briefly with a University official in a non-adversarial format, to provide their own version of events in response to student complaints, cannot be characterized as the equivalent of a credible threat of “enforcement” or as the kind of “extraordinarily intrusive” process that might make self-

censorship an objectively reasonable response.

19

Id. The Seventh Circuit, in rejecting another challenge by Speech First to a university bias policy, expressly relied on this “useful guidance” in Abbott to hold that “[i]t follows that if a mandatory meeting does not demonstrate a credible threat of enforcement, neither does an invitation to an optional one.” Speech First v. Killeen, 968 F.3d 628, 640–41 (7th Cir. 2020), as amended on denial of reh’g and reh’g en banc (Sept. 4, 2020). Such is the case here. Indeed, although Abbott provided only “guidance” for the Seventh Circuit, here Abbott controls.

For these reasons, we cannot accept Speech First’s argument that the BIRT process is so burdensome that an objectively reasonable student would self-censor to avoid encountering it.

3.

Both of Speech First’s theories as to why the Bias Policy violates the First Amendment — i.e., that the Bias Policy chills speech through implicit threats and that the BIRT resolution process is burdensome enough to itself chill speech — suffer from the same fundamental problem: Speech First has not shown that the Bias Policy credibly threatens injury to the organization’s members. And for this reason, Speech First’s members have not demonstrated the injury in fact necessary to establish standing.

To bring a speech chilling claim, a plaintiff must at least show that it has “suffered an injury or threat of injury that is ‘credible,’ not ‘imaginary or speculative.’” Cooksey, 721 F.3d at 235 (quoting Babbitt v. United Farm Workers Nat’l Union, 442 U.S. 289, 298 (1979)); cf. Clapper v. Amnesty Int’l USA, 568 U.S. 398, 414 (2013) (holding that respondents could not establish injury by relying on a “speculative chain of possibilities”).

20

It is undisputed that the Bias Policy itself does not set forth or contemplate sanctions and that the BIRT has no power to impose any sanctions. Nor is there any evidence that the BIRT makes threats suggesting that it can punish students.

Speech First offers only speculation in support of its argument that it has suffered an injury in fact. Because the district court’s factual findings make clear that no record evidence establishes any such injury, the organization has failed to establish an injury in fact and so lacks standing to challenge the Bias Policy.

B.

We recognize that Speech First has brought similar challenges against universities located in the Fifth, Sixth, Seventh, and Eleventh Circuits. See Speech First, Inc. v. Fenves, 979 F.3d 319 (5th Cir. 2020), as revised (Oct. 30, 2020); Speech First, Inc. v. Schlissel, 939 F.3d 756 (6th Cir. 2019); Killeen, 968 F.3d at 628; Speech First, Inc. v. Cartwright, 32 F.4th 1110 (11th Cir. 2022).

In each of these cases, Speech First sought a preliminary injunction. And, in each, the district court denied Speech First’s motion for lack of standing. See Speech First, Inc. v. Fenves, 384 F. Supp. 3d 732, 743 (W.D. Tex. 2019) (finding that the record was “without any evidence of a credible threat of enforcement”); Speech First, Inc. v. Schlissel, 333 F. Supp. 3d 700, 713 (E.D. Mich. 2018) (finding that “[t]he evidence in the present matter . . . reflects no threats—direct, subtle, or implied”); Speech First, Inc. v. Killeen, No. 19-03142, slip op. at 31 (C.D. Ill. Sept. 17, 2019) (finding that “being reported [to the bias response team] . . . results in essentially no consequences”); Speech First, Inc. v. Cartwright, No. 21-313, 2021 WL 3399829, at *5 (M.D. Fla. July 29, 2021) (finding that

21

Speech First presented no evidence that the university policy “compels student involvement”). Speech First subsequently noted an appeal from each judgment.

The Seventh Circuit affirmed. That court followed what we believe to be the only appropriate approach: it based “its standing analysis around a series of factual findings by the district court.” See Killeen, 968 F.3d at 649 (Brennan, J., concurring in part and dissenting in part). After examining those findings and concluding that none were clearly erroneous, the Seventh Circuit upheld the district court’s judgment denying a preliminary injunction to Speech First, as we do here. Id. at 639, 647 (majority opinion).

The other appellate courts vacated and remanded, but only after seemingly ignoring the factual findings of the respective district courts. The Fifth and Sixth Circuits never even mentioned the applicable standard for reviewing facts found by a district court. Neither court seemed to acknowledge the deference due those findings. Accordingly, Judge White, dissenting from the Sixth Circuit’s decision, noted that the majority had reversed the district court’s ruling on standing despite “not identify[ing] any clear error in that district court’s factual findings.” Schlissel, 939 F.3d at 773 (White, J., dissenting). The Eleventh Circuit, for its part, apparently relied on its own factual finding: “that the average college student would be intimidated, and quite possibly silenced, by the policy.” Cartwright, 32 F.4th at 1124. Not only did the district court in Cartwright make no such finding, its decision suggests the opposite is true. See Cartwright, 2021 WL 3399829, at *5.

Like our colleagues in the Seventh Circuit, and Judge White on the Sixth Circuit, we believe that the district court’s factual findings must serve as the starting point for our

22

analysis. Here, Speech First has failed to demonstrate that a single one of the district court’s numerous findings of fact is clearly erroneous. And those findings more than adequately support the court’s legal conclusion that Speech First’s student members have not demonstrated an injury in fact. Therefore, Speech First is without standing to challenge the Bias Policy.

C.

Our country rightfully places great value on the freedom of speech, and any statute or regulation implicating speech receives close scrutiny. Freedom of speech, after all, is expressly protected in the very first of the original amendments to our Constitution. But the First Amendment does not stand in the way of modest efforts to encourage civility on college campuses.

The Supreme Court has explained that governmental entities may engage in speech in order to “‘promote a program’ or ‘espouse a policy.’” Shurtleff v. City of Boston, 142 S. Ct. 1583, 1587 (2022) (quoting Walker v. Texas Div., Sons of Confederate Veterans, Inc., 576 U.S. 200, 208 (2015)). As the Second Circuit explained in Okwedy v. Molinari, 333 F.3d 339, 344 (2d Cir. 2003), “[w]hat matters is the distinction between attempts to convince and attempts to coerce.” See also Backpage.com, LLC v. Dart, 807 F.3d 229, 230 (7th Cir. 2015) (explaining that the First Amendment permits “government expression” but forbids government “intimidation”).

Although Virginia Tech clearly would prefer that its students engage in respectful discourse, the district court found no evidence that the University sought to pursue this goal through intimidation or threats. In fact, the court found no evidence that the Bias Policy

23

has imposed or threatened to impose any discipline on anyone. Virginia Tech’s Bias Policy therefore falls within the bounds of acceptable government speech.

Moreover, in considering Speech First’s challenge, we are also mindful of the fact that colleges and universities occupy a special place in our society. They expose students to a wide range of topics that they might otherwise overlook, as well as to persons whose backgrounds and life experiences are far distant from their own. And, for many students, college life is all-encompassing; college campuses are where students reside, work, form bonds with classmates, and refine their worldviews.

It is vital, therefore, that “[t]eachers and students . . . remain free to inquire, to study and to evaluate, to gain new maturity and understanding,” and to engage in robust debate about current events. Whitehill v. Elkins, 389 U.S. 54, 60 (1967) (quoting Sweezy v. New Hampshire, 354 U.S. 234, 250 (1957)). The Constitution, however, does not require us to ignore that universities have not always been places where such open dialogue is accepted from everyone. With this history in mind, many universities — Virginia Tech among them — find it equally vital to communicate that their campuses are places welcoming to all students, whatever those students’ backgrounds and whatever their political, social, or religious views. 11 Just as universities may legitimately strive to promote intellectual curiosity, so too they may legitimately strive to promote civility and a sense of belonging among the student body.

11

Indeed, federal antidiscrimination statutes obligate universities to ensure that student-on-student harassment does not bar students’ access to educational opportunities and benefits. See, e.g., Davis ex rel. LaShonda D. v. Monroe Cnty. Bd. of Educ., 526 U.S. 629, 633 (1999).

24

That is what Virginia Tech’s Bias Policy seeks to achieve. Rather than show indifference for the role that harmful stereotypes and discriminatory tropes play in all facets of society, see, e.g., McIver v. Bridgestone Americas, Inc., 42 F.4th 398, 413 (4th Cir. 2022) (Motz, J., concurring in the judgment), the University here has devised a way to educate its student body about both “protected speech and the role of tolerance in the campus community.” Sands, 2021 WL 4315459, at *8. This is precisely the type of government speech that the First Amendment permits. See Walker, 576 U.S. at 207–08.

IV.

We turn to Speech First’s challenge to Virginia Tech’s Informational Activities Policy. The district court held that Speech First did have “standing to pursue its challenge” to this policy because its student members’ “intended conduct [was] arguably proscribed by the policy.” 12 Sands, 2021 WL 4315459, at *22–23. But the court further held that the record, as to this policy, was too incomplete for it to forecast whether Speech First was likely to prevail on the merits. Id. at *24.

The court reasoned that even if the Informational Activities Policy does restrict Speech First’s members’ ability to leaflet and solicit signatures on campus, the policy may ultimately be a “reasonable time, place, and manner restriction[]” that is permitted under

12

We are unconvinced by Virginia Tech’s argument that Speech First lacks standing to challenge the Informational Activities Policy. It is uncontroverted that the policy bars Speech First’s members from distributing literature unless they first place a reservation through the Campus Life Office and secure the sponsorship of an RSO. And, unlike Gilles v. Torgersen, 71 F.3d 497, 500–01 (4th Cir. 1995), on which the University relies, the University has not effectively exempted Speech First from these requirements. Accordingly, the policy’s limitation on speech and expression, even if ultimately constitutionally permissible, is at least sufficient to confer standing.

25

the First Amendment. See id. at *23. Speech First contends that the district court committed reversible error in concluding that it needed “a more developed record” to determine whether the Informational Activities Policy is constitutional. Speech First Opening Br. at 32–33; see also Sands, 2021 WL 4315459, at *24. Thus, notwithstanding the district court’s holding, Speech First asks us to enjoin the Informational Activities Policy at this early stage in the litigation.

Speech First maintains that the present record requires a finding that the Informational Activities Policy is unlawful both as an impermissible prior restraint on speech and as an impermissible speaker-based regulation.

Speech First rests much of its prior restraint argument on our opinion in Child Evangelism Fellowship of Maryland, Inc. v. Montgomery County Public Schools, 457 F.3d 376 (4th Cir. 2006), which Speech First asserts is “directly on point” and required the district court to enjoin the Informational Activities Policy as facially unconstitutional — thus precluding the need for further factual development. See Speech First Opening Br. at 33–34; Speech First Reply Br. at 13–14. Speech First claims that the Informational Activities Policy does exactly what Child Evangelism forbids, i.e., it confers “unbridled discretion” upon school administrators to limit First Amendment activity. Speech First Opening Br. at 32–33. Given Speech First’s heavy reliance on Child Evangelism, we set forth the particulars of the case in some detail.

That case concerned a public school district that permitted outside organizations to use the schools’ take-home flyer forum to distribute flyers for elementary school students to take home to their parents. Child Evangelism, 757 F.3d at 378–79. The Child

26

Evangelism Fellowship (CEF), which hosted faith-centered “Good News Club” meetings, sought to advertise its activities through that take-home flyer forum. Id. The school district’s “policy governing access to the take-home flyer forum” conferred on the school district “broad discretion over flyer distribution.” Id. CEF, invoking the First Amendment, sought an injunction that would require the school district to grant CEF access to the take- home flyer forum. Id. at 380. When the district court refused to do so, we reversed, holding that the school district policy did “not provide adequate protection for viewpoint neutrality” because it left the school district with “unfettered discretion . . . to control access to the take-home flyer forum.” Id. at 378.

According to Speech First, like the take-home flyer policy in Child Evangelism, Virginia Tech’s Informational Activities Policy confers “unbridled discretion” on the University that can be used to disallow speech. Speech First Opening Br. at 32. But the district court made explicit findings to the contrary. 13 The court explained that, upon receipt of an informational activities request, the Campus Life Office “merely confirms the location is available and then approves [the request].” Sands, 2021 WL 4315459, at *22. Thus, the court found that the Informational Activities Policy is nothing more than a “reservation system to which all student organizations have access.” Id. Accordingly, the

13

Speech First maintains that the district court’s findings are irrelevant because Child Evangelism focused only on “the plain language of the [take-home flyer] policy.” Speech First Opening Br. at 33 (quoting Child Evangelism, 457 F.3d at 387). This contention eliminates the context of the quoted phrase and misreads our holding in Child Evangelism. When we discussed “the policy” in Child Evangelism, we considered both the written policy and the factual record as to the policy’s implementation. Id. at 379, 387 n.7, 388. The district court here did not err in engaging in the same analysis.

27

district court found, as a matter of fact, that the Informational Activities Policy does not confer any discretion on the University. Id.

Alternatively, Speech First argues that the Informational Activities Policy is still an unconstitutional prior restraint even if it is only a time, place, and manner restriction. See Speech First Opening Br. at 34–35. This is so, according to Speech First, because the University has not adequately explained the need for such a policy in the first place. See id.

But we have repeatedly stressed that injunctive relief is an extraordinary remedy and that it is the plaintiff’s burden to demonstrate entitlement to such relief. See, e.g., Leaders of a Beautiful Struggle, 2 F.4th at 339; Roe v. Dep’t of Def., 947 F.3d 207, 219 (4th Cir. 2020), as amended (Jan. 14, 2020). At this stage of the proceedings, it is Speech First that must demonstrate that it is likely to succeed in its challenge to the Informational Activities Policy — a tough hurdle to overcome, to be sure. See Cantley v. W. Virginia Reg’l Jail & Corr. Facility Auth., 771 F.3d 201, 207 (4th Cir. 2014); see also Leaders of a Beautiful Struggle, 2 F.4th at 355 (Wilkinson, J., dissenting) (“[P]reliminary injunctions should not be casually awarded . . . .”).

This burden is made more difficult here because, as the district court found, the University has offered evidence that it has limited physical space, and that implementing a reservation system “ensures fair and equitable access” to its finite resources. Sands, 2021 WL 4315459, at *22; see also Widmar v. Vincent, 454 U.S. 263, 267 n.5 (1981) (observing that universities possess “authority to impose reasonable regulations” compatible with their educational mission). The district court found that, in contrast, Speech First failed to

28

produce any evidence suggesting that Virginia Tech’s stated rationale was unsupported, unreasonable, or a subterfuge. See Sands, 2021 WL 4315459, at *24. Given the state of the record, we can hardly fault the district court for refusing to issue an injunction.

Speech First separately maintains that the Informational Activities Policy is an unconstitutional speaker-based regulation, and that no further factual development is necessary to determine the policy’s unconstitutionality. See Speech First Opening Br. at 35–37. But Speech First’s speaker-based challenge is plagued by the same misunderstanding of the preliminary injunction burden as its prior restraint challenge. Again, Speech First bears the burden of demonstrating it is likely to prevail on the merits, and once more falls short.

The Informational Activities Policy requires students to secure the sponsorship of an RSO before they can distribute leaflets on campus. According to Speech First, “[t]his is a classic speaker-based restriction,” and so constitutionally forbidden. See id. at 36. Speech First suggests that the RSO preference embedded in Virginia Tech’s Informational Activities Policy is no different than the one found unconstitutional by the Eighth Circuit in Turning Point USA at Arkansas State University v. Rhodes, 973 F.3d 868 (8th Cir. 2020). See Speech First Opening Br. at 36. Not so.

In Turning Point, the Eighth Circuit considered an Arkansas State University policy that permitted tabling in the patio area outside of the student union, but only for “registered student organizations and University departments.” 973 F.3d at 873. The university’s sole rationale for imposing this restriction was that it wanted the union patio to exist as a “comfortable, living-room atmosphere.” Id. at 879. Because there was “no rational

29

relationship” between the university’s nonspecific justification and the restriction it imposed, the Eighth Circuit held that the policy was unconstitutional. Id.

Virginia Tech, by contrast, has not relied on an amorphous desire to make its campus more “comfortable.” Rather, the University maintains, and the district court expressly found, that the University relies on physical capacity concerns and only restricts informational activities “to official organizations so that Virginia Tech’s limited physical resources can be used for the benefit of the most students.” Sands, 2021 WL 4315459, at *22. This explanation may, or may not, be a “good reason[] for distinguishing between registered student organizations and other members of the university community.” Turning Point USA, 973 F.3d at 879. It is simply too early to tell because, as the district court also found, the record lacks “information about the demands on reservable spaces by RSOs and the availability of alternatives for students who are not members of RSOs.” Sands, 2021 WL 4315459, at *24. Such information would appear vital to the fact-intensive inquiry inherent in determining whether the Informational Activities Policy fits within the contours of the First Amendment. Given the inadequacy of the record evidence pertinent to the Informal Activities Policy, we cannot second guess the district court’s considered judgment.

“[A]n undeveloped record not only makes it harder for a plaintiff to meet [its]

burden of proof, it also cautions against an appellate court setting aside the district court’s exercise of its discretion.” Siegel v. LePore, 234 F.3d 1163, 1175 (11th Cir. 2000) (en banc); see also Bowe v. Bd. of Election Comm’rs of City of Chicago, 614 F.2d 1147, 1153 (7th Cir. 1980). Once this case is returned to the district court, and after further factual

30

development has taken place, it will be for that court to determine in the first instance whether the Informational Activities Policy complies with the First Amendment. Without a developed record, the district court did not err, let alone clearly err, in determining that Speech First has not yet shown that it is likely to succeed on the merits. See Di Biase, 872 F.3d at 230. We accordingly affirm its present refusal to enjoin the Informational Activities Policy.

V.

“[S]ecuring reversal of a denial of preliminary relief is an uphill battle for any movant.” League of Women Voters of N. Carolina v. North Carolina, 769 F.3d 224, 250 (4th Cir. 2014) (Motz, J., dissenting); see also Leaders of a Beautiful Struggle, 2 F.4th at 357 (Wilkinson, J., dissenting); Sierra Club v. Georgia Power Co., 180 F.3d 1309, 1310 (11th Cir. 1999); Associated Gen. Contractors of California, Inc. v. Coal. for Econ. Equity, 950 F.2d 1401, 1405 (9th Cir. 1991). This is so because “the ultimate decision to grant the preliminary injunction” is confined to the district court’s discretion. See League of Women Voters, 769 F.3d at 250 (Motz, J., dissenting) (citing Miller v. Mitchell, 598 F.3d 139, 145 (3d Cir. 2010)). “Indeed, because the law places a thumb on the scale against the issuance of preliminary injunctions, our deference should be even greater when the district court denies a preliminary injunction.” Leaders of a Beautiful Struggle, 2 F.4th at 357 (Wilkinson, J., dissenting) (emphasis in original).

In this case, though, none of these considerations seem to matter to the dissent.

Throughout its lengthy exposition, the dissent disregards the district court’s findings of fact

31

and replaces them with its own conjecture. 14 Though there are many examples, a few suffice to make the point.

While the dissent proclaims that “[c]ollege students hoping to stay on the administration’s good side will not view the [BIRT’s] ‘invitation’ as voluntary,” the district court found as a fact that there was “no evidence that students feel obligated to come to these voluntary meetings.” Compare Dis. Op. at 48 with Sands, 2021 WL 4315459, at *12; see also Killeen, 968 F.3d at 643 (observing that the majority of students contacted by the University of Illinois’ bias response team declined the invitation to meet). While the dissent contends that Virginia Tech students have good reason to fear “significant reputational harm” from the BIRT, the district court found as a fact that nothing about the BIRT process would ever appear on a student’s academic transcript or disciplinary record. Compare Dis. Op. at 51 with Sands, 2021 WL 4315459, at *10. While the dissent discounts Virginia Tech’s affirmative efforts to protect free speech as mere “promises and assurances,” the district court found as a fact that even “if one part of the university makes a mistake [about protected speech], the university as a whole will ensure that protected speech remains protected.” Compare Dis. Op. at 45 with Sands, 2021 WL 4315459, at *12. And while the dissent complains that the Informational Activities Policy provides “no discernable standards by which to evaluate tabling requests,” the district court found

14

Sometimes, when the district court’s findings of fact prove too inconvenient, the dissent resorts to mischaracterizing them as “claims” by the majority. See, e.g., Dis. Op. at 47, 53.

32

as a fact that the policy is nothing more than “a reservation system.” Compare Dis. Op. at 61 with Sands, 2021 WL 4315459, at *22.

Moreover, rather than deferring to the district court’s prudent decision to wait for a fuller record before deciding whether to enjoin the Informational Activities Policy, the dissent shrugs off the court’s entirely proper decision to delay ruling on the question pending development of the factual record as a trivial focus on “narrow details.” See Dis. Op. at 54.

Most telling is what the dissent does not do: it does not meaningfully recognize the procedural posture of this case, it does not apply the proper deferential standard of review to the district court’s many findings of fact, and it does not even attempt to adhere to our holdings in Abbott. The dissent instead concocts a novel standard of review for preliminary injunctions in which an appellate court can conjure up subjective fear and then treat it as objective fact, view the record in the light least favorable to the non-movant, and strip district courts of their discretion to determine whether a factual record is sufficiently developed. The dissent’s misguided journey produces a dramatic read, but it comes nowhere close to offering a basis for upending the district court’s careful exercise of its discretion.

For all these reasons, the judgment of the district court is AFFIRMED.

33

WILKINSON, Circuit Judge, dissenting:

Consider a 19-year-old sophomore at Virginia Tech sitting in a favorite class, one involving the role that race, ethnicity, and gender play in contemporary American politics. During a lively class discussion, an interesting but controversial topic comes up. She considers raising her hand to add her thoughts to this fascinating debate, but she hesitates.

She remembers hearing about the University’s Bias Intervention and Response Team, which Virginia Tech established to “eliminate acts of bias” through “immediate direct or indirect responses to bias-related incidents.” She cannot recall how “bias incident” was defined but thinks it was something about “expressions against a person” in a protected class. She knows that biased speech can be reported anonymously online. In fact, Virginia Tech “encourages” students “to make a report” if they “hear or see something that feels like a bias incident” even if they are “unsure.” She vaguely remembers that those reported for bias will be invited to a meeting with the Dean of Students or referred to another University office. Students are told the meetings are voluntary, but word travels quickly on college campuses, and she does not want to be “that girl who got reported.” She cannot recollect whether those who get accused of bias get in trouble with the University, but she knows the Dean of Students keeps a file of all complaints.

She thought she had an insightful comment to add to the discussion, but it might not be worth risking an encounter with the bias response team, especially because the team comprises representatives from the offices of Inclusion and Diversity, Student Conduct, the Dean of Students, and the Virginia Tech Police Department.

34

Faced with these circumstances, what would a reasonable student do? Speak up and risk an anonymous report? Or keep her head down, sit silently, and avoid the potential fallout? A student in this situation will almost always choose the latter. And this is how Virginia Tech objectively chills speech.

The First Amendment should prevent this danger. The Supreme Court has underscored “in a number of cases that constitutional violations may arise from the deterrent, or ‘chilling,’ effect of governmental regulations that fall short of a direct prohibition.” Laird v. Tatum, 408 U.S. 1, 11 (1972). When pressure and intimidation lurk behind the state’s policy such that a reasonable individual “is chilled from exercising her right to free expression” and instead chooses “self-censorship,” the steadfast protections of the First Amendment should be summoned into action. Benham v. City of Charlotte, 635 F.3d 129, 135 (4th Cir. 2011).

My friends in the majority claim any fears about chilling are the product of the “dissent’s misguided journey,” Majority Op. at 32, which I suspect was not intended as a compliment. Odysseys aside, the stark reality of the record is that programs like the Bias Intervention and Response Team (BIRT) are being used in the service of discouraging that open inquiry from which education draws its very meaning and sustenance.

Virginia Tech’s Bias Intervention and Response Team is not the benign little system the majority imagines. It is intimidating to the extent that a student of “ordinary firmness” would be deterred from exercising her First Amendment rights. Benham, 635 F.3d at 135. Chilling effects of even well-intended government policies present “an evil of constitutional proportions” for the government has an obligation to promote free speech,

35

and at the very least, not to discourage it. Leslie Kendrick, Speech, Intent, and the Chilling Effect, 54 Wm. & Mary L. Rev. 1633, 1655 (2013).

The majority today ignores the plight of the college sophomore, declaring instead that Speech First fails to show that BIRT objectively chills speech. Due to Virginia Tech’s perfunctory promise to uphold the First Amendment and its proffered assurance that BIRT cannot directly punish students, the majority concludes that no reasonable student would be dissuaded from exercising her free speech rights. This conclusion neglects real-world consequences. The reality is that Virginia Tech has constructed a complex apparatus for policing and reporting whatever administrators may deem “biased speech.” The intricate program has a straightforward effect: students self-censor, fearing the consequences of a report to BIRT and thinking that speech is no longer worth the trouble.

How did it ever come to this—that such a fine and distinguished university would institute a policy with such incipient inquisitional overtones, one that turns its campus into a surveillance state? The First Amendment guarantees to everyone not just passive access to but active participation in the marketplace of ideas. Today, the majority breaks that promise to a segment of society who needs it most—college students.

The damage done to the First Amendment is not confined to BIRT. It is compounded by the University’s Informational Activities Policy. The whole is even worse than the sum of its parts. The University has somehow managed to offend virtually every cardinal principle of First Amendment law. The First Amendment comes in dead last by its reckoning. The Amendment exists at the sufferance of a bureaucracy dedicated to eliminating “bias.” To say that campus life has lost its way severely understates the distance

36

from productive dialogue that once nurtured the capacity of young minds for critical thought. Slowly the critical now succumbs to the conventional. How sad. I would remand this case with directions to the district court to enjoin this ill-conceived experiment in its entirety, thereby allowing the University a new start, one which returns the fresh air of free speech to its rightful place in campus life.

I.

The majority claims that this litigation comes to us in a premature posture. It argues that the record is devoid of the injury needed to possess standing and is undeveloped as to the application of the Informational Activities Policy. See Majority Op. at 32. The majority asks that we kick this whole case down the road. What we have before us, however, is wholly sufficient to resolve this matter here and now. The grant or denial of a preliminary injunction is an appealable order. It is so for a reason. Where the appellants as here have a strong case of ultimate success on the merits, they deserve preliminary injunctive relief and will suffer irreparable harm if it is not awarded. See Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008). This is especially so in connection with the First Amendment. That Amendment and those invoking its protections should not be placed at the sufferance of extended rounds of litigation. In this sense, the Amendment functions like an immunity. The longer its beneficiaries languish in litigation, the more its value and meaning is lost. See Mitchell v. Forsyth, 472 U.S. 511, 525–29 (1985).

I am aware that this is a facial challenge to BIRT. I am further aware that facial challenges are disfavored. Wash. State Grange v. Wash. State Republican Party, 552 U.S. 442, 450 (2008); United States v. Chappell, 691 F.3d 388, 392 (4th Cir. 2012). The majority

37

says the plaintiffs lack standing. Virginia Tech suggests that, even if Speech First has standing, we ought to hold off any ruling until we receive evidence as to BIRT’s application.

I would reject this course of action. In the First Amendment context, facial challenges are appropriate where a “substantial number” of the applications of an “impermissibly overbroad” policy are “unconstitutional[] judged in relation to the statute’s plainly legitimate sweep.” Wash. State Grange, 552 U.S. at 449 n.6 (internal quotation marks omitted). The depressive effect of this policy is so evident from its face that we do not need to wait for one deprivation of protected speech after another to occur. I doubt we shall ever know just how many deprivations there are, for the simple reason that self- censorship seldom comes into view. The prohibitive effect on speech will follow from this policy as surely as the night follows the day. See City Council of City of Los Angeles v. Taxpayers for Vincent, 466 U.S. 789, 797 (1984) (explaining that policies are “unconstitutional on their face” when “any attempt to enforce” such policies “would create an unacceptable risk of the suppression of ideas”); see also Sec’y of State of Md. v. Joseph H. Munson Co., 467 U.S. 947, 965 n.13 (1984) (collecting cases).

Reviewing the very terms of this scheme should persuade anyone that a regime this inimical to campus free speech cannot be permitted to stand. The educative effect that colleges should ideally have upon the exercise of basic civil liberty is set at naught by the University’s view that speech freedom no longer belongs first and foremost to individual Americans but to the collective and authoritative hand of the state that would suppress them.

38

The majority would insulate this flawed system from challenge by holding that Speech First lacks standing to bring suit. Plaintiffs must typically suffer an injury in fact to establish standing. In the First Amendment context, however, plaintiffs may “refrain from exposing themselves to sanctions under the policy, instead making a sufficient showing of self-censorship,” thus establishing a “chilling effect on their free expression that is objectively reasonable.” Abbott v. Pastides, 900 F.3d 160, 176 (4th Cir. 2018) (internal quotation marks omitted). The majority makes much of the statement in Abbott that it is critical that there be “a credible threat of enforcement” for a policy to objectively chill speech. Id. Because BIRT does not directly punish students for speech, the majority concludes there is no credible threat and therefore no chilling effect.

If we confine “credible threat of enforcement” to direct punishment by BIRT, we leave out the policies that stop short of formal penalties but nonetheless exact a heavy toll. What formal penalties is the majority waiting for? A reprimand? Probation? A suspension? An expulsion? If BIRT does not formally impose the particular penalty, its referrals to other offices nonetheless set the process in motion. Of course Speech First has standing. It has shown that the policy at issue causes students to self-censor for fear of being reported, thus effecting an objective chill on speech. The majority talks about such consequences as being wholly fanciful and theoretical. How can it know? Thoughts that are formulated in the mind but never escape the lips still detract from the store of First Amendment expression.

When a challenged policy “risks chilling the exercise of First Amendment rights, the Supreme Court has dispensed with rigid standing requirements.” Cooksey v. Futtrell, 721 F.3d 226, 235 (4th Cir. 2013) (internal quotation marks omitted). Should

39

constitutionally questionable policies escape judicial scrutiny due to overly stringent standing analysis, “[s]ociety as a whole then would be the loser.” Joseph H. Munson, 467 U.S. at 956. Speech First has shown a variety of ways in which BIRT will take action against protected speech, thus demonstrating a credible threat of enforcement.

Let’s examine some of these.

A.

BIRT’s own self-description lays bare Virginia Tech’s persistent efforts to impose a bureaucratic superstructure that dampens speech. Because the majority glosses over the policy’s practical consequences, it is important to lay it out from beginning to end exactly as it is presented to students. Once the full policy is exposed, stripped of fig-leaf assurances, its oppressive nature has nowhere to hide.

1.

The very terms of BIRT betray its chilling effect. Virginia Tech proclaims BIRT’s purpose in the most practiced bureaucratese, saying a lot while saying nothing. Unpacking its meaning is no small task. BIRT was created to “eliminate acts of bias” through “immediate direct or indirect responses to bias-related incidents.” J.A. 369. The University claims to need this policy because “[b]ias-related incidents often represent a conflict of competing and opposing values,” and conventional solutions “will rarely go far enough to do the adaptive work needed to resolve[] gaps between values, beliefs, and behavior.” Id. The “Bias Intervention and Response Team” is thus meant to be “both proactive and responsive” in addressing the “challenges presented in a community where inclusion and

40

dissent exist in a way that often results in marginalization, isolation, and loneliness.” J.A. 368.

How dense will all this get? The composition of BIRT seems further designed to intimidate any student brave enough to utter controversial thoughts. To achieve its goals, BIRT consists of “a representative from” the office of the “Dean of Students,” “Office for Equity and Accessibility, Office for Inclusion and Diversity, Student Conduct, [and the] Virginia Tech Police Department.” J.A. 370. Other “representatives from offices may be called upon to participate with BIRT due to the nature of an incident.” Id. These representatives are supposed to “compile a report for the community that offers a narrative about[] . . . shift[ing] our culture towards a more positive and pluralistic society.” Id.

A reasonable student reading BIRT’s purpose is left with the ominous impression that BIRT was created to “proactive[ly]” address problems caused by “dissent[ing]” “values” and “beliefs,” whatever they may happen to be. BIRT will “resolve[] gaps between values, beliefs, and behavior” whenever the supposed gaps appear because typical solutions “will rarely go far enough.” To achieve its goal of eliminating bias, BIRT enlists a cast of unnamed but high-ranking University officials and police.

This garbled declaration reads like a mission statement of a committee dedicated to rooting out dissent. A student whose views fail to align with campus orthodoxy would think twice before speaking under such a menacing regime. The majority says that “what matters is the distinction between attempts to convince and attempts to coerce.” Majority Op. at 22 (quoting Okwedy v. Molinari, 333 F.3d 339, 344 (2d Cir. 2003)). When the stated goal of the bias response team is to “eliminate” bias, we are faced not with a gentle effort to

41

convince students to be unbiased but with a systematic effort to coercively drive out views that strike administrators the wrong way. Telling students “You should stop saying biased things or else we might need to reeducate you” is no benevolent attempt at persuasion. Moreover, as soon as students see that the bias response team includes members of the Office of Student Conduct and Virginia Tech Police Department, they will adjust their behavior to avoid getting reported. Their parents might be displeased and their future prospects compromised if they had a run-in with disciplinary officials and local law enforcement. Better to just keep quiet.

2.

The plain language of its policy makes BIRT even more troubling. Virginia Tech provides a vague but expansive definition of “bias.” Students are told that “bias incidents” are “expressions against a person or group because of the person’s or group’s age, color, disability, gender (including pregnancy), gender identity, gender expression, genetic information, national origin, political affiliation, race, religion, sexual orientation, veteran status, or any other basis protected by law.” J.A. 333.

With a definition this loose and rambling, almost anything could be framed as a bias incident. Indeed, the number of bias reports nearly doubled from 2017 to 2018. Speech First, Inc. v. Sands, No. 7:21-CV-00203, 2021 WL 4315459, at *10 (W.D. Va. Sept. 22, 2021). This trend either shows Virginia Tech’s community is growing more biased—an odd result given the University’s efforts to “eliminate” such problems—or it reveals that members of the community are increasingly exploiting the policy’s uncertain language. This trend will give rise to another—as reports increase, speech decreases.

42

Worse yet, the explicit language of the policy shows that BIRT invades the realm of protected speech. It defines a “bias incident” as “expressions,” and lists examples such as “words or actions that contradict the spirit of the Principles of Community,” “jokes that are demeaning to a particular group of people,” and “posting flyers that contain demeaning language or images.” Id. at *8. Additionally, BIRT’s preset list of nineteen potential bias offenses includes “Comment in Class or Assignment,” “Comment in Person,” “Comment in Writing or on Internet,” “Comment via Email/Text,” “Comment via Phone/Voicemail,” and “Written Slur.” J.A. 149.

The students in this case, through their submitted declarations, explained that they “enrolled in the University because [they] wanted to learn in a challenging environment where students and faculty are free to engage in lively, fearless debate and deliberation.” Pl.’s Mot. to Supplement R. (Student Decl.) 8, 12, 16, 20, ECF No. 67. Yet they are told that words contradicting the spirit of the community’s principles can be reported as biased speech. Such an “overly broad” policy “creates a ‘danger zone’ within which protected expression may be inhibited.” Dombrowski v. Pfister, 380 U.S. 479, 494 (1965). The plain language alerts students that even their comments in class and online are constantly and subjectively parsed for bias.

In effect, the prolix text of the policy permits students to report speech for no other reason than they were offended by what was said. In fact, the district court recounted a few such reports of bias, including the following:

[1] a report that the words “Saudi Arabia” were written on a whiteboard outside of a student’s dorm room, alleging bias based on “national or ethnic origin;” [2] a report that a student in a University residence hall overheard

43

several male students privately “talking crap about the women who were ‘playing’ in a snowball fight,” calling them not “athletic,” which the complainant reported as discrimination based on “gender;” and [3] a report that a student told a joke that included “Caitlyn Jenner’s deadname” during a classroom lecture, which was reported as discrimination on the basis of gender identity.

Sands, 2021 WL 4315459, at *10 (internal citations omitted).

These examples illustrate how students can report speech as biased based on nothing more than its obvious poor taste or the fact that some listener took offense. But offensive speech can seldom be isolated with precision. Speech exists in an environment in which the useful and insightful are often commingled with the outrageous and profane. But the “fact that society may find speech offensive is not a sufficient reason for suppressing it.” FCC v. Pacifica Found., 438 U.S. 726, 745 (1978). To the contrary, “if it is the speaker’s opinion that gives offense, that consequence is a reason for according it constitutional protection.” Id.

Despite the “bedrock principle underlying the First Amendment” that the state may not silence “the expression of an idea simply because society finds the idea itself offensive or disagreeable,” Texas v. Johnson, 491 U.S. 397, 414 (1989), BIRT’s sweeping and vague language enables students to report anything the most sensitive are offended by. The students in this case fear that because the “definition of ‘bias’ is so broad and vague,” they are “confident that someone will find [their] speech to be ‘biased.’” Student Decl. 9, 13, 17, 21. Those offended can take steps to avoid future exposure or, better still, engage the speaker with their well-founded disapproval. The cure for distasteful speech is tasteful dialogue, not the conversion of an illustrious campus into a First Amendment dead zone.

44

3.

The BIRT policy goes from bad to worse. Virginia Tech establishes a regime of comprehensive surveillance. The University enthusiastically encourages its students to report a bias incident even if they are “unsure” that an incident qualifies as biased. J.A. 200. As a “student, if you hear or see something that feels like a bias incident, statement, or expression, we encourage you to make a report. In short, if you see something, say something!” Id. Where students are urged to report on one another, mutual suspicions fester, as any society bereft of basic freedoms can attest. Anyone at Virginia Tech can submit reports of bias anonymously on Virginia Tech’s website through an online reporting tool, called the Bias Incident Reporting Form. Sands, 2021 WL 4315459, at *9. Surveillance is total and constant: Undergraduate and professional students “can be referred for bias-related behavior” from “admission to commencement.” J.A. 372. And “bias-related incidents can occur on or off campus, including on social media and other digital platforms.” Sands, 2021 WL 4315459, at *9.

BIRT’s eye is nothing short of ubiquitous. In the classroom, on social media, and off campus, Virginia Tech encourages its students to keep watch for any biased speech. By extending its sphere of surveillance off campus, Virginia Tech stretches well beyond the boundaries of acceptable regulation. Courts “must be more skeptical of a school’s efforts to regulate off-campus speech, for doing so may mean the student cannot engage in that kind of speech at all.” Mahanoy Area Sch. Dist. v. B.L. ex rel. Levy, 141 S. Ct. 2038, 2046 (2021). Speech First’s student members wish to “engage in open and robust intellectual debate with [their] fellow students” about controversial topics “in the classroom, in other

45

areas of campus, [and] online.” See, e.g., Student Decl. 9. But they are “reluctant to openly express” their opinions because they know that their classmates can anonymously report them on campus or off, creating an oppressive atmosphere of scrutiny from which there is no reprieve. Id. In short, students stay silent because they can be reported anytime, anywhere. It is “clear that the average college student would be intimidated, and quite possibly silenced, by the [University’s] policy.” Speech First, Inc. v. Cartwright, 32 F.4th 1110, 1124 (11th Cir. 2022).

B.

The majority believes all this evidence cannot show a chilling effect because Virginia Tech soothes its students’ fears through numerous promises and assurances. First, Virginia Tech emphasizes that BIRT itself does not have any authority to punish students and only refers reports to other offices. Second, when someone is accused of bias, Virginia Tech claims the most likely course of action is that the Dean of Students will invite the accused to a voluntary meeting. Third, students have nothing to fear because how could anyone even suggest that the bias response team would ever penalize constitutionally protected speech.

All these placations miss the mark. Virginia Tech’s claims do not change the fact that a reasonable student, considering the policy and accompanying repercussions as a whole, would hardly regard the regime in a kindly, avuncular light.

1.

Let’s take the matter of BIRT referrals first. Even if BIRT cannot formally punish students and only refers cases to other offices, it gets the ball rolling. A referral to other

46

offices by definition carries its own administrative imprimatur. BIRT admits that it refers cases to offices such as the Office of the Dean of Students, Fraternity and Sorority Life, Student Conduct, Equity and Accessibility, Housing and Residence Life, and Title IX. J.A. 360. Indeed, after a student reported that inappropriate words had been written on a hallway whiteboard, BIRT reviewed the incident and “notified VTPD” and Title IX. J.A. 172. The district court found that “BIRT has referred protected speech to” the Office of Student Conduct “at least twice since 2017.” Sands, 2021 WL 4315459, at *10.

So once a complaint is shipped to one of these other entities, reasonable students could fear that they will be subject to that office’s disciplinary authority. Even if students believe that BIRT itself cannot punish them (a big “if”), that hardly dispels their fears of getting in trouble with another office. Indeed, the students below said that they are “afraid that Dean of Students Office will keep a record on [them], share the allegations with others at the university, call [them] in for meetings, or refer the allegations to the Office of Student Conduct, the Office of Equity and Accessibility, or the Virginia Tech Police Department.” Student Decl. 9, 13, 17, 21.

Consider the scenario in which an eavesdropper anonymously reports a student for gender bias. Even if the accused student thinks that BIRT cannot punish him, he may fear that BIRT will refer him to the office of Title IX which may well be hostile to the student’s point of view. Then will he not worry that the office of Title IX may discipline him for his alleged bias? It matters not that BIRT cannot dole out punishment if it can simply refer the case to another office that can. Being reported to BIRT may be only the first step. Who knows what awaits? Uncertainty itself can cause students to fall silent rather than speak up.

47

2.

Virginia Tech also claims that, where BIRT or the Dean of Students directly responds, all that will happen is the Dean will “invite” the accused student to a “voluntary conversation.” Sands, 2021 WL 4315459, at *10. An invitation is commonly thought of as something one looks forward to. So what sort of invitation is this?

The majority claims that Speech First’s members have not offered “evidence that they (or their peers) feel pressured to attend the meetings.” Majority Op. at 15. It emphasizes that “the district court found as a fact that there was ‘no evidence that students feel obligated to come to these voluntary meetings.’” Id. at 31 (quoting Sands, 2021 WL 4315459, at *12).

Does the majority really believe this invite is no different from students inviting one another to drop by down the hall for a Friday night pizza? No! This is an invitation from the Dean to the student to come to the Dean’s office, not for tea or coffee, but for the express purpose of discussing the student’s speech. There is an imbalance here. Dean versus student. State versus the individual. This is precisely the sort of imbalance that the First Amendment has historically refused to tolerate. Why should we ever do so here?

Speech First students not surprisingly explained that they are “afraid” that the Dean will keep a record on them or share the allegations with others at the University. See, e.g., Student Decl. 17, 21. While the majority claims that no trace of the meeting will “appear on a student’s academic transcript or disciplinary record,” Majority Op. at 31, there is no dispute that a record of the meeting will be kept on permanent file within the Dean of Students Office’s case management system, see id. at 4–5; J.A. 372. College students

48

hoping to stay on the administration’s good side will not view the “invitation” as voluntary or as something to which one may simply send “regrets.” With the invitation comes pressure to attend the meeting; better to avoid the whole darn thing by keeping one’s mouth shut.

As the Eleventh Circuit recognized, “the students targeted here are . . . teenagers and young adults who, it stands to reason, are more likely to be cowed by subtle coercion.” Cartwright, 32 F.4th at 1123. In multiple areas of law, the Supreme Court has expressed concern with state coercion of young people, as they are more susceptible to threats and intimidation. See id. at 1123–24. Simply telling students the meeting is “voluntary” will do little or nothing to put anyone at ease.

Additionally, “the very name” Bias Intervention and Response Team “suggests that the accused student’s actions have been prejudged to be biased.” Speech First, Inc. v. Schlissel, 939 F.3d 756, 765 (6th Cir. 2019). A reasonable student could certainly reach this conclusion given the policy’s use of terminology such as “victim,” “bystander,” “perpetrator,” “targeted,” and “accused.” J.A. 369–70. As such, the name and language of BIRT “intimates that failure to” attend the meeting with the Dean may result in a type of default judgment against the student. Schlissel, 939 F.3d at 765. If a student fears the accusations will be taken as true unless he contests them, he would not skip the meeting and risk exposure to “far-reaching consequences, including reputational harm or administrative action.” Id. If he does skip the meeting, it may well be out of apprehension as to what even worse possibilities await him there.

49

What’s more, although Virginia Tech claims a voluntary meeting on the carpet with the Dean is the only response from the school, the policy communicated to students warns otherwise. Virginia Tech touts numerous “interventions for addressing” bias, such as “[a]dvocacy for community members impacted by mistreatment,” reporting “a narrative about acts of bias-related behavior” to the community, and further educating the “campus to ensure that bias-incident reporting systems are publicized.” J.A. 372. The University also cautions students that “BIRT can consider an array of responses to include: temporary or permanent changes to on-campus housing” or “a community alert” about the incident “from the police department.” J.A. 368. Reasonable students could thus conclude that a report to BIRT will result in more than a “voluntary meeting.” Despite the Dean’s assurances, the reasonable student would likely keep quiet rather than gamble with the repercussions.

3.

When accused of chilling constitutionally protected speech, Virginia Tech’s response is, “trust us, BIRT will uphold free speech.” Despite devoting much time and many resources to BIRT and proclaiming a goal of eliminating bias, Virginia Tech declares it will administer this policy with a dispassionate hand.

Why should this be comforting? What is protected speech to one may seem unprotected to another. The First Amendment is not to be so casually consigned to the eye of the beholder. It is true that Virginia Tech’s written policy affirms its commitment to free speech by saying, “BIRT will examine and review each complaint through the lens of free and protected speech.” J.A. 370. But this nebulous assurance offers cold comfort. The very

50

next sentence warns that “[s]ome bias-related incidents may violate the Student Code of Conduct and may be adjudicated through the student conduct process.” Id. This is reinforced by the fact that BIRT has referred protected speech regarding undocumented immigration to the Office of Student Conduct before. See Sands, 2021 WL 4315459, at *10. Furthermore, although “Virginia Tech cannot adjudicate matters that are deemed protected speech,” it tells students that “[b]ehavior that is discriminatory or otherwise hurtful to members of the community is addressed through educational interventions.” J.A. 370. The next paragraph reminds students that “[r]egardless of whether incidents violate policy or are insensitive, it is crucial that response occurs in a timely and consistent manner.” Id.

Students are also warned that if BIRT believes bias exists, “[i]nterventions of either an educational or restorative nature will . . . be conducted by the office closest to the students,” and the “case will be logged in the [Dean of Students Office’s] case management system.” J.A. 372. If BIRT finds no bias, however, the case is still logged, indicating that even reports of non-biased, protected speech are kept on file. Id.

Through these terms, the University communicates to students that it will intervene even in instances of protected speech. On the one hand, the administration promises that BIRT does not punish protected speech, but then on the other, it tells students that protected, but controversial speech will be subject to the University’s reeducation efforts. The reasonable student hears a vague promise to protect free speech followed by BIRT doubling down on the “crucial” need for responding “in a timely and consistent manner”

51

to bias with “educational interventions,” a creative euphemism for disruptive measures meant to stop the student from engaging in heterodox speech again.

The fact that the University keeps a file of all complaints makes dealing with BIRT all the more discomforting. There is no point in keeping something on file if it will never be used for anything. A young college student would be concerned that a filed report may affect her standing with the University in some way. Even assuming the University keeps the student’s name confidential, that protection is not ironclad. If word gets out—either through an announcement about the bias incident or through the old-fashioned rumor mill—the student could face significant reputational harm. Such news being spread to the school could have “dramatic effects such as currying disfavor with a professor, or impacting future job prospects.” Schlissel, 939 F.3d at 765.

Similarly, it damages a student’s reputation if her classmates know that she had to undergo an “educational or restorative” intervention, implying that the student is either ignorant or in need of moral restoration. “Because reputational damage can impair a student’s prospects for academic and professional success, objectively reasonable students may be expected to behave in ways that mitigate their exposure to any allegation that might trigger a bias investigation.” Speech First, Inc. v. Killeen, 968 F.3d 628, 652 (7th Cir. 2020) (Brennan, J., concurring). The briefest study of the McCarthy era teaches that even baseless accusations can be enough to ruin someone’s good name.

Though stopping short of formal discipline, the collateral consequences of being reported to BIRT create a system where “process is punishment.” Keeping complaints on record, subjecting students to “educational interventions,” and possibly referring students

52

to another office is punishment enough. If a city were to announce that it will respond to accusations of unpopular speech with an “educational intervention” or “moral restoration” team, we would not let its constitutional violations off the hook simply because the team disclaimed its obvious authority to punish and promised always to respect the First Amendment. It does not save a policy from constitutional scrutiny for the state to say that its interventions are benign and nonpunitive. The “possibility the Government could have imposed more draconian limitations on speech never has justified a lesser abridgment.” Denver Area Educ. Telecommunications Consortium, Inc. v. FCC, 518 U.S. 727, 809 (1996) (Kennedy, J., concurring in part).

Virginia Tech cannot paper over the prohibitive tenor of BIRT with promises to be on its best behavior. The First Amendment “does not leave us at the mercy of noblesse oblige” and does not condone an unconstitutional system “merely because the Government promised to use it responsibly.” United States v. Stevens, 559 U.S. 460, 480 (2010). In other words, the First Amendment entertains no “pinky promise” exception to chilling speech.

Given the admission that the University will respond to, record, and attempt to correct constitutionally protected speech, the undisputed facts show that a reasonable student is “chilled from exercising her right to free expression” and would instead choose “self-censorship.” Benham, 635 F.3d at 135. BIRT presents a credible threat of enforcement because even assuming a credulous student believes that BIRT cannot punish, he is still likely to self-censor to avoid at least the procedural consequences and at most the reputational harms that inhere in every sinew of this oppressive system.

53

In sum, Virginia Tech’s bias response policy chills speech by fostering an atmosphere of anonymous, constant surveillance from, as we have noted, “admission to commencement.” With its vague definition of “bias,” any joke or off-hand remark or controversial comment can be contorted into a “bias incident.” This environment objectively targets the freedom of speech. The only question is whether we give Virginia Tech a license to censor because its policies come with gilded assurances. The majority claims that “even ‘if one part of the university makes a mistake about protected speech, the university as a whole will ensure that protected speech remains protected.’” Majority Op. at 31 (quoting Sands, 2021 WL 4315459, at *12). What scant solace to know that the bureaucracy may correct its infractions at some point or the other down the road. Why not design a system that heads off First Amendment violations before they occur? The majority believes the University deserves a pass. But state authority is often suppressive before it is wielded. Every day these policies stand, someone somewhere is driven to silence.

II.

The problems free speech faces on Virginia Tech’s campus unfortunately do not end there. The majority also affirms the district court’s decision declining to enjoin Virginia Tech’s Informational Activities Policy. This policy governs the “distribution of literature and/or petitioning for signatures,” J.A. 225, and limits these activities to “specified tabling locations,” J.A. 420. Per the policy, students may engage in these forms of speech “if they are sponsored by a university affiliated organization” and receive “prior approval by the designated university scheduling office.” J.A. 225. Despite acknowledging that the policy “clearly proscribe[s]” speech, the district court concluded that it did not have enough

54

information to determine whether the challenged policy is an acceptable time, place, and manner restriction on speech. Sands, 2021 WL 4315459, at *23–24.

In assessing the Informational Activities Policy, both the district court and the majority focused on narrow details, asking questions such as how many tables are available and how often are they reserved. See id. at *24; Majority Op. at 29. This attention to particulars neglects the policy’s broader consequences. To be clear, the policy forbids students from passing out flyers or collecting signatures anywhere on campus except at designated tabling locations that one must reserve ahead of time. To do so, the student must belong to a registered student organization or be sponsored by one. There is plenty in the record as it now stands to enjoin this policy as an unreasonable restraint on speech.

A.

The district court committed reversible error by failing to conduct the appropriate legal analysis of the Informational Activities Policy. The majority acknowledges that we review the district court’s “legal conclusions de novo.” Majority Op. at 8. Yet the majority does no such thing, deferring wholly to the district court’s views of the law. Looking anew at the legal standard and undisputed facts, a few things are clear. First, the tabling locations identified by the policy are limited public forums and thus restrictions on their use must be reasonable. Second, because the policy proscribes certain types of fundamental speech anywhere on campus except for where allowed by the University and only when given prior approval, the policy is an unreasonable constraint on speech.

55

1.

Determining what kind of forum is at issue is essential to our First Amendment analysis “because the extent to which the Government may limit access” depends on “the nature of the forum.” Cornelius v. NAACP Legal Def. & Educ. Fund, Inc., 473 U.S. 788, 797 (1985). The district court, for its part, did not even attempt to analyze the forum.

Not deciding which type of forum was at issue was error. This is a legal question that our circuit has already answered. A public university “campus is a limited public forum.” American Civil Liberties Union v. Mote, 423 F.3d 438, 444 (4th Cir. 2005). This makes sense. A university campus, “at least for its students, possesses many of the characteristics of a public forum.” Widmar v. Vincent, 454 U.S. 263, 267 n.5 (1981). For example, universities are replete with walkways and thoroughfares, features that are similar to traditional public forums. These areas are generally available to students even if they are inaccessible to the public. It is therefore evident that Virginia Tech’s campus is at least a limited public forum.

Virginia Tech’s tabling locations under the policy are also limited public forums.

The “government creates a designated [or limited] public forum when it purposefully makes property generally available to a class of speakers.” Warren v. Fairfax Cnty., 196 F.3d 186, 193 (4th Cir. 1999) (internal quotation marks omitted). Here the tables have been made generally available to students through the school’s official policy.

The policy at issue delineates a group of speakers who may use the forum—i.e., sponsored individuals—from a group of speakers who may not—i.e., non-sponsored individuals. Such a policy that allows certain people to speak while disallowing others must

56

be “viewpoint neutral and reasonable in light of the objective purposes served by the forum.” Id. at 194. Put differently, the “touchstone for evaluating” the policy’s regulations “is whether they are reasonable in light of the purpose which the forum at issue serves.” Perry Educ. Ass’n v. Perry Loc. Educators’ Ass’n, 460 U.S. 37, 49 (1983).

Courts assess reasonableness by looking at, inter alia, whether “substantial alternative channels . . . remain open” despite the policy. Perry, 460 U.S. at 54; see also Ball v. City of Lincoln, Nebraska, 870 F.3d 722, 737 (8th Cir. 2017); Uptown Pawn & Jewelry, Inc. v. City of Hollywood, 337 F.3d 1275, 1281 (11th Cir. 2003). Courts also ask whether “the nature of the property is inconsistent with expressive activity.” Cornelius, 473 U.S. at 803; see also Turning Point USA at Arkansas State University v. Rhodes, 973 F.3d 868, 875–77 (8th Cir. 2020). Last, courts should account for university administrators’ expertise in creating educational policies. Christian Legal Soc. Chapter of the Univ. of California, Hastings Coll. Of the L. v. Martinez, 561 U.S. 661, 686 (2010); see also Rhodes, 973 F.3d at 877. Even the most cursory consideration of these factors reveals Virginia Tech’s policy is unreasonable.

2.

First, Virginia Tech’s policy does not keep open substantial alternative channels for the distribution of pamphlets and the collection of signatures. Students who are not part of a registered student organization or sponsored by one are completely banned from engaging in signature collection or pamphlet distribution anywhere on campus. Moreover, even students who are members of a student organization are banned from distributing pamphlets or collecting signatures unless they have reserved a table through the University.

57

It is not an exaggeration to say that Virginia Tech’s policy leaves no alternative channels for students who wish to exercise their rights to collect signatures and distribute literature.

Virginia Tech argues that a plethora of alternative avenues for speech remain open to students, ranging from the wide array of electronic media to old-fashioned bulletin boards. This does not suffice. The framers of our Constitution would be shocked to think that petitioning and leafletting would be subject to prior state approval. We should not accept the University’s invitation to downplay the importance of these two core mediums of free speech. Imagine Samuel Adams or Thomas Paine beseeching the state at some table for permission to pamphleteer. They would hardly abide such a prior restraint.

Handing out leaflets “is the essence of First Amendment expression” and no “form of speech is entitled to greater constitutional protection.” McIntyre v. Ohio Elections Comm’n, 514 U.S. 334, 347 (1995). Leafleting and pamphleteering often allow communication face to face. And the pamphleteer may well feel his speech has greater resonance away from a state-sanctioned table. “For the Revolutionary generation . . . the pamphlet had peculiar virtues as a medium of communication,” for it “allowed one to do things that were not possible in any other form.” McCullen v. Coakley, 573 U.S. 464, 489 n.5 (2014) (quoting Bernard Bailyn, The Ideological Origins of the American Revolution 2 (1967)). The importance of petitioning likewise cannot be understated as it is expressly protected by the First Amendment alongside the freedom of speech.

Second, when considering the purpose served by the forum, we must remember, as the majority notes, “colleges and universities occupy a special place in our society.” Majority Op. at 23. These educational institutions have historically been places for the free

58

exchange of speech in the “marketplace of ideas.” Healy v. James, 408 U.S. 169, 180 (1972). Thus, when examining a restriction on informational activities like the one at issue here, we are not dealing with a restriction on the distribution of literature on a busy downtown street or even in a public park; we are dealing with a ban on advocacy and information in one of America’s most important free speech communities. Virginia Tech’s restriction on speech does not reasonably match the purposes served by a college campus. The proposed activity at issue involves students passing out leaflets or collecting signatures. And the nature of the forum here is overwhelmingly consistent with promoting a free flow of information. The distribution of pamphlets, the collection of signatures, and the advocacy and debate that accompany those two are perfectly aligned with a college’s mission of intellectual growth.

Finally, Virginia Tech argues that we should defer to its expertise. Yet none of the University’s proffered justifications for the policy have anything to do with education; they all deal with the orderly administration of campus property. The University’s expertise is limited to education, not constitutional law. Courts are “the final arbiter of the question whether a public university has exceeded constitutional constraints, and we owe no deference to universities when we consider that question.” Martinez, 561 U.S. at 686. Moreover, even if we agreed with Virginia Tech on this point, deference to administrators alone would not overcome the lack of alternative channels of speech and the inconsistency between the policy and the forum.

59

3.

Considering these factors illuminates the unreasonableness of Virginia Tech’s Informational Activities Policy. The University attempts to justify the policy with a handful of feeble reasons. I find none of them compelling.

Virginia Tech claims that this policy is necessary to ensure that access to the limited tabling locations is “fair and equitable.” J.A. 420. We should not reward such artificial scarcity. If Virginia Tech did not completely ban informational activities from its entire campus except at designated tables, then there would be no scarcity of possible locations. We should not shift the focus away from the sheer sweep of the policy to mundane questions about table reservations and locations. To do so would impose artificial limits on our review. Our job is to assess the constitutionality of the policy and its effects on campus as a whole, not simply to diagnose the procedures of the table-reservation policy asserted by the state. A university cannot severely restrict speech—and here Virginia Tech enforces a broad ban on leafleting and signature gathering on campus—and then hide behind the ostensible fairness of its limited reservation system. Virginia Tech cannot manufacture scarcity and then use that as a justification for banning speech.

The University also claims its system is necessary to ensure that student groups have access to “high-traffic areas of Campus without impeding student movement (for example, by blocking the exits to busy classrooms) or invading student living spaces (for example, soliciting door to door in residence halls).” J.A. 420. Again, the policy makes little sense in addressing this concern. A policy against pamphleteering in front of classroom exits or

60

soliciting in residence halls would accomplish the intended goal. Such concerns do not reasonably justify a ban on every walkway and open-air venue on the entire campus.

Similarly, Virginia Tech claims that it is “reasonable to prefer those students who, by forming an organization, have shown some likelihood of structure and continuity in their contribution to the marketplace of ideas.” Response Br. at 46. But this reasoning flips the First Amendment on its head. The First Amendment is not needed to protect speech that is popular enough to attract a group; it exists to safeguard an individual’s speech that is so “unpopular or distasteful” that other groups will not sponsor it. Denver Area Educ. Telecommunications Consortium, 518 U.S. at 785 (Kennedy, J., concurring).

B.

Speech First’s likelihood of success on the merits is further bolstered because the Informational Activities Policy constitutes an unconstitutional prior restraint on speech. A prior restraint exists when a regulation gives “public officials the power to deny use of a forum in advance of actual expression.” Se. Promotions, Ltd. V. Conrad, 420 U.S. 546, 553 (1975). “Any prior restraint on expression comes” with a “‘heavy presumption’ against its constitutional validity.” Org. for a Better Austin v. Keefe, 402 U.S. 415, 419 (1971).

In the context of public schools and universities, we have acknowledged that prior restraints are only permissible if they are viewpoint neutral, reasonable, and limit government discretion. See Mote, 423 F.3d at 446; see also Taylor v. Roswell Indep. Sch. Dist., 713 F.3d 25, 46 (10th Cir. 2013). Thus, to “pass constitutional muster,” a viewpoint- neutral preapproval system “must contain adequate standards to guide the official’s decision.” Wag More Dogs Liab. Corp. v. Cozart, 680 F.3d 359, 372 (4th Cir. 2012)

61

(internal quotation marks omitted). In other words, a policy cannot give officials unbridled discretion. See, e.g., Atlanta Journal & Constitution v. City of Atlanta Dep’t of Aviation, 322 F.3d 1298, 1310–11 (11th Cir. 2003); DeBoer v. Village of Oak Park, 267 F.3d 558, 572–73 (7th Cir. 2001); Lewis v. Wilson, 253 F.3d 1077, 1079 (8th Cir. 2001); Summum v. Callaghan, 130 F.3d 906, 919 (10th Cir. 1997). A policy that grants too much ex-ante discretion to a university official cannot survive constitutional scrutiny.

Virginia Tech’s policy states that requests to distribute pamphlets and collect signatures “require prior approval by the designated university scheduling office and are subject to university policies and the reasonable guidelines of the authorizing official.” J.A. 225. What’s more, requests are reviewed by considering “overall campus safety and security, any special circumstances relating to university activities, and the impact such activity may have on the university.” J.A. 223. If there are fuzzier or more opaque standards than these, they do not come readily to mind.

The majority denies there is any discretion involved in this policy by saying that it is “nothing more” than a reservation system. Majority Op. at 26. On its face, however, the policy places unreviewable discretion in the hands of university administrators with no discernable standards by which to evaluate tabling requests. The only limitations placed on the University come in the form of vague language referencing unspecified “reasonable guidelines” and the equally unspecified “impact such activity may have on the university.” Rather than provide clarity and limit discretion, these additional factors only give administrators more subjective and unverifiable ways to deny a request.

62

For example, if a student requests to hand out pamphlets promoting a disfavored cause, the school could say his request has been denied due to “university policies.” Such a terse rejection would prevent the student from engaging in speech and would be unreviewable. The “absence of express standards renders it difficult to differentiate between a legitimate denial of access and an illegitimate abuse of censorial power.” Child Evangelism Fellowship of MD, Inc. v. Montgomery Cnty. Pub. Sch., 457 F.3d 376, 386 (4th Cir. 2006) (internal quotation marks omitted). Indeed, among “the dangers posed by unbridled discretion” is the censor’s “ability to hide unconstitutional viewpoint discrimination.” Id. The policy’s express terms establish a prior restraint that gives an administrator unchecked discretion, funneling all requests to distribute literature or collect signatures through some black box of government preapproval.

It is important to view the BIRT and Informational Activities policies in tandem.

The Informational Activities Policy intensifies the chilling effect of BIRT. The two policies in one sense are disparate but in another sense are of the same piece, for both angle toward an identical end goal: You cannot speak on campus without Virginia Tech’s approval. I regret the majority’s failure to consider the effect of the BIRT policy and the leafletting prior restraint together. The policies chill speech by channeling student expression into a process that must be approved by the state—either on the front end when trying to distribute pamphlets, or on the back end when a student is reported for bias. Virginia Tech effectively creates “a University-controlled clearinghouse for speech [that] can deter students from speaking out.” Killeen, 968 F.3d at 652 (Brennan, J., concurring).

63

III.

Making matters worse, today’s decision splits from three of our sister circuits. Given the dangers that bias response bureaucracies pose to free expression, it is no wonder that the Fifth, Sixth, and Eleventh Circuits have held that these policies objectively chill speech.

The majority’s divergence from our colleagues on the Sixth Circuit is especially troubling because the record shows that Virginia Tech’s policy is even worse than the one at issue at the University of Michigan. The bias response policy there defined a “bias incident” as “conduct that discriminates, stereotypes, excludes, harasses or harms anyone in our community based on their identity.” Schlissel, 939 F.3d at 762 (emphasis added). Virginia Tech’s policy here does not even feign such a limitation; it explicitly defines a “bias incident” as “expressions.” J.A. 333 (emphasis added). Even though Michigan was attempting to proscribe conduct and not speech, the Sixth Circuit nonetheless held that Michigan’s policy objectively chilled student expression. Schlissel, 939 F.3d at 765.

The Eleventh Circuit considered a university bias response team that ostensibly could not punish students but could only refer them to other university actors. The court held that “the district court erred in focusing so singularly on the . . . power to punish.” Cartwright, 32 F.4th at 1122. While “[p]unishment is no doubt relevant to the objective- chill analysis, and may well be sufficient to prove the requisite chill . . . analogous precedent makes clear that it is not decisive and, in any event, is not uniformly necessary.” Id. The question, our colleagues determined, was a simple one with a simple answer: Would “the average college-aged student . . . be intimidated—and thereby chilled from exercising her free-speech rights—by subjection to the bias-related-incidents policy?” The

64

answer was yes. Id. at 1124. Given the option of a narrow focus on the power of punishment or a broader examination of the policy’s consequences, our sister circuit wisely chose the latter. I cannot say the same about the decision today.

Likewise, the Fifth Circuit held that a bias-response team that did not engage in investigations or punishment still objectively chilled speech given the team’s ability to refer students to other offices. Speech First, Inc. v. Fenves, 979 F.3d 319, 333 (5th Cir. 2020). And like Virginia Tech here, the University of Texas there also stressed its commitment to the freedom of speech, stating that it “expressly protect[s] and encourage[s]” free speech. Id. The Fifth Circuit nonetheless found that these “paeans” to the First Amendment did not “detract[] from the likelihood that the University’s policies” chill protected speech. Id. at 333–34. All three cases closely resemble ours, yet their legal analysis strives to protect First Amendment rights where ours stumbles. 1 The majority wishes away the circuit conflict by declaring that three of our sister circuits either “ignor[ed] the factual findings of the respective district courts” or made up their “own factual finding.” Majority Op. at 21. I would give our fine colleagues on the Fifth, Sixth, and Eleventh Circuits more credit. If the majority wishes to disagree with them, it at least owes it to them and the law more generally to engage with their reasoning at a higher level. Our sister circuits did not ignore pertinent facts or make stuff up. They

1

The majority is wrong to say the Seventh Circuit reached a different result. The students in that case provided only a “bareboned declaration,” which did not indicate whether they intended to engage in speech that would invoke the consequences of the bias response policy. Killeen, 968 F.3d at 644. As we have made clear, that evidentiary deficiency is emphatically absent here given the plainly professed expressive intentions of the students. See, e.g., Student Decl. 8, 12, 20.

65

sized up an important issue of constitutional law and put forth a thoughtful analysis of it. One need only check the above opinions themselves to confirm this conclusion.

The majority’s decision today is all the more unfortunate because this is no mere theoretical disagreement. This circuit split creates a patchwork of First Amendment jurisprudence for schools across the country. On the vitally important issue of free speech on college campuses, we should avoid the needless creation of circuit splits, which results in students in Michigan, Florida, and Texas being protected from unconstitutional policies while students in Virginia remain exposed.

IV.

Viewpoint discrimination is arguably the cardinal First Amendment sin, but that is exactly where the BIRT policy leads. When anyone can be anonymously reported for an offense vaguely defined as a “bias incident,” one must abandon common sense to think that all students will feel an equal effect. The truth is that those who believe their views are the least popular will be the first to clam up. No student who knows his speech will rouse applause will hesitate to make his voice heard. But the student who wishes to dissent will shrink from the stage. Because of this lopsided effect, BIRT commits the First Amendment’s original sin, an “egregious” transgression from which the “government must abstain.” Rosenberger v. Rector & Visitors of Univ. of Va, 515 U.S. 819, 829 (1995).

A.

It is a “core postulate of free speech law” that the “government may not discriminate against speech based on the ideas or opinions it conveys.” Iancu v. Brunetti, 139 S. Ct. 2294, 2299 (2019). When a government policy, even inadvertently, amplifies the speech

66

of one viewpoint while chilling the speech of another, that policy destabilizes the foundation of the First Amendment, which “forbids the government to regulate speech in ways that favor some viewpoints or ideas at the expense of others.” Taxpayers for Vincent, 466 U.S. at 804.

The Supreme Court has admonished us to beware of policies that may become viewpoint discriminatory in application. The Court found that a law forbidding registration of “immoral or scandalous” trademarks had criteria so broad and vague that the government was rejecting trademarks merely because they were “offensive to many Americans.” Iancu, 139 S. Ct. at 2301. The Court concluded that “a law disfavoring ‘ideas that offend’ discriminates based on viewpoint, in violation of the First Amendment.” Id. Although the government insisted that it would only interpret “immoral or scandalous” in a non- viewpoint-discriminatory way, the Court held that “we cannot accept the Government’s proposal, because the statute says something markedly different.” Id. The Court thus struck down the provision because it resulted in “viewpoint-discriminatory application.” Id. at 2300.

B.

Virginia Tech provides a similarly broad definition of “bias” that will predictably lead to students reporting “ideas that offend” while permitting ideas that please. We know the students involved in this case wish to “speak passionately” about topics such as affirmative action, Black Lives Matter, and abortion, but they refrain from doing so out of fear of being reported. Student Decl. 7–20.

67

One need not be a mystic to anticipate BIRT’s foreseeable effects: Students holding conservative or otherwise heterodox views on controversial topics are more likely to be reported for bias. Indeed, “anonymous reports carr[y] particular overtones of intimidation to students whose views are ‘outside the mainstream.’” Fenves, 979 F.3d at 338. It takes courage to express unpopular or “incorrect” views, even under the very best of circumstances. Pile BIRT on top of all this, and students with pro-life, anti-affirmative action, restrictive immigration views and traditional religious beliefs have been made conspicuous targets. Sometimes, people at the University do not even try to hide their hostility toward conservative students who hold dissenting values. When students filed this current lawsuit, a Virginia Tech tenured professor took to Twitter to publicly call the students “conservative shitbags” who were “suing the school because they’re bigots.” J.A. 274.

It is beyond wrong to place these students in the crosshairs. It was beyond wrong in the civil rights era to make those courageous voices for racial equality subject to vilification or worse. It was beyond wrong to make American pacifists in times of war feel beyond the pale of civil discourse. The First Amendment does not permit the fevers of majority passions to deny the minority its say.

Per Virginia Tech’s own data, only one in five students feel comfortable expressing ideas in class that are “probably only held by a minority of people.” Sands, 2021 WL 4315459, at *2. Furthermore, the Foundation for Individual Rights and Expression (FIRE) observed in its 2017 report that bias response policies “risk becoming tools . . . for imposing some form of political or intellectual orthodoxy” on campus. J.A. 246. Professors

68

themselves have remarked that bias response policies “result in a troubling silence” where students are “afraid to speak their minds” because their peers can “leverage bias reporting policies to shut down unpopular or minority viewpoints.” J.A. 264.

The more distasteful public authority finds someone’s views, the more it should seek to protect the speaker’s right to express them. BIRT runs in exactly the opposite direction. BIRT thus blesses state-sanctioned silencing of minority views, allowing “the majority [to] draw[] a formidable circle around thought,” where inside “those limits, the writer is free; but unhappiness awaits him if he dares to leave them” for he is then exposed to “persecutions every day.” Alexis de Tocqueville, Democracy in America 244 (Harvey C. Mansfield & Delba Whitney ed. & trans., U. Chi. Press 2000) (1835). Like in Iancu, the policy here favors one group over another based solely on viewpoint, distinguishing between “those inducing societal nods of approval and those provoking offense and condemnation.” 139 S. Ct. at 2300.

Sometimes, of course, the majority view is right. Sometimes the conventional wisdom is sound. Sometimes the “politically correct” is indeed correct. But it must prove its worth through testing, not complacency. Through dialogue, not suppression. The “[f]reedom to think as you will and to speak as you think are means indispensable to the discovery and spread of political truth.” Whitney v. California, 274 U.S. 357, 375 (1927) (Brandeis, J., concurring). Yet in “both concept and design,” BIRT’s “efforts to encourage students to anonymously” report their classmates for an endless litany of ill-defined offenses “subvert free and open inquiry and invite fears of political favoritism.” Keith E. Whittington, Free Speech and the Diverse University, 87 Fordham L. Rev. 2453, 2466

69

(2019). Viewpoint discrimination merits clear eyes, not blind ones, for in situations like this and Iancu, it can be de facto as well as de jure.

V.

The majority says the “First Amendment does not stand in the way of modest efforts to encourage civility on college campuses,” Majority Op. at 22, and that Virginia Tech “has devised a way to educate its student body about” the role that “harmful stereotypes and discriminatory tropes play in all facets of society,” id. at 24. All well and good. But while everyone agrees that promoting civility and discouraging discrimination is a good thing, the majority’s vague invocation of civility has no limiting principle.

If, in the name of fostering civility and eliminating stereotypes, the state is allowed to take action against individuals that includes keeping a record of one’s wrongspeak, then the stalwart protections of the First Amendment will slowly but inevitably recede. Any speech that is controversial can reflexively be labeled uncivil or discriminatory. Where would the majority draw the line between tolerable speech and the intolerable? The civil and the uncivil? Or the harmful and the helpful? Why should the state be in the business of drawing such lines at all? A public university cannot unfurl some omnipresent banner of civility to silence its students.

The sweet, innocent little system the majority envisions wishes merrily away the suppressive cards in the state’s exclusive hand. This has it all backwards. Universities should be the first to embrace the “well-known aphorism, ‘I disapprove of what you say, but I will defend to the death your right to say it.’” Mahanoy Area Sch. Dist., 141 S. Ct. at

70

2046. Virginia Tech flips the adage on its head, encouraging students to respond, “I disapprove of what you say, and I’m going to report you for it.”

Speech at universities is distinct. It attempts to probe conventional thought rather than confirm it. At its worst, it can be petty and banal. But at its best, academic discourse can rise above the sort of pablum one might hear at a political convention or the repetition of pat, dogmatic lines that often pass for serious discussion. At a time when Americans increasingly sort themselves through soundbites and talking points, speech at a university should counter these trends and embrace the cross-pollination of ideas from diverse viewpoints. It is no coincidence that ancient Athens introduced both dialogue to theater and democracy to the world around the same time—dialogue gives rise to democratic engagement because political truth emerges most fully from the conflict of different points of view. Unfortunately, Virginia Tech’s policies restrict the role that speech can play in our democracy at the very time the unique capacities of university speech are needed most. Its bias response team sounds less like a catalyst for dialogue and more like some Ministry of Truth.

Of course there are reasonable time, place, and manner restrictions on speech that administrators can establish, see Perry, 460 U.S. at 45–46, such as protecting the privacy of dorms and the orderly progression of classes. Encouraging dialogue does not mean countenancing disruption. The heckler has no right to veto the speech that others come to hear. And the University may obviously take protective measures to stem violent misconduct against students, as occurred so tragically at Virginia Tech itself in the depraved killing of so many good and fine people on April 16, 2007. When speech

71

threatens to verge into violent misconduct, University officials remain free to act in the best interest of those for whom any community would naturally seek to care.

It is not too much, however, to expect University administrators to appreciate the difference between expressive speech and violent action. This distinction is basic not only to the First Amendment, but to the way in which our society is run and governed. That basic separation must prevail in the campus environment. It is, I recognize, difficult to understand when campus dialogue will lead to mutual respect and appreciation on the one hand or when it may result in heated tempers on the other. It is not, however, for government to steer the consequences of pure speech, lest the line between free and repressive societies be lost.

For all its faults, higher education is not some expensive bauble that has outlived its usefulness. It remains essential to the transmission of the upper reaches of learning to upcoming generations, to the nation’s innovation and growth, and to the ability of students to appreciate the past even as they anticipate the future. Universities and monasteries kept alive the fickle light of learning in the Middle Ages. It is not naïve to think universities today can illuminate a better way. America depends mightily on its great universities, of which Virginia Tech is surely one. What we have before us in this case is unworthy of it.

I respectfully dissent. I would reverse and remand this case with directions that the district court enjoin the BIRT and Informational Activities policies at issue. If they are the future of our nation’s universities, we shall all share in the loss.