Summary
In Hernandez, we did not look for any of the extra indicia of public importance that today's majority now piles onto the test.
Summary of this case from Adams v. Cnty. of SacramentoOpinion
No. 21-16007
08-05-2022
Steven J. Serbalik (argued), Steven J. Serbalik PLC, Scottsdale, Arizona, for Plaintiffs-Appellants. Stephen B. Coleman (argued), Pierce Coleman PLLC, Scottsdale, Arizona, for Defendants-Appellees.
Steven J. Serbalik (argued), Steven J. Serbalik PLC, Scottsdale, Arizona, for Plaintiffs-Appellants.
Stephen B. Coleman (argued), Pierce Coleman PLLC, Scottsdale, Arizona, for Defendants-Appellees.
Before: Richard A. Paez, Richard R. Clifton, and Paul J. Watford, Circuit Judges.
WATFORD, Circuit Judge:
In 2013, the City of Phoenix's Police Department adopted a new policy governing its employees' use of social media. Among other things, the policy prohibits employees from engaging in speech on social media that would be "detrimental to the mission and functions of the Department," "undermine respect or public confidence in the Department," or "impair working relationships" of the Department. In 2019, the Department concluded that Sergeant Juan Hernandez violated the policy by posting content to his personal Facebook profile that denigrated Muslims and Islam. When the Department took steps to discipline Hernandez, he sued, alleging that the Department was retaliating against him for exercising his First Amendment right to freedom of speech.
The district court rejected Hernandez's First Amendment retaliation claim on the ground that his speech did not address matters of public concern and was therefore not entitled to constitutional protection under the balancing test established in Pickering v. Board of Education of Township High School District 205, Will County , 391 U.S. 563, 88 S.Ct. 1731, 20 L.Ed.2d 811 (1968). The court also rejected Hernandez's claim that certain provisions of the Department's social media policy are facially invalid. We reverse the dismissal of Hernandez's First Amendment retaliation claim but largely affirm the district court's rejection of his facial challenge to the Department's policy.
I
With the rise in popularity of social media platforms like Facebook and Twitter, many employers have adopted policies regulating their employees' use of social media. In some instances, those policies govern not only what employees may say when speaking while on the job, but also what they may say when speaking in their capacity as private citizens outside the workplace. Such policies are founded on the recognition that what employees say on their own time can negatively affect both an employee's ability to perform his or her job duties and the employer's ability to effectively carry out its mission.
The Phoenix Police Department promulgated a policy governing its employees' use of social media in August 2013 by adopting Operations Order 3.27. Titled "Social Media Use Policy," the order established a comprehensive set of regulations and guidelines that apply to the use of social media both on and off the job. The policy explicitly covers an employee's "personal use" of social media and, under that heading, provides two general admonitions: (1) "Department personnel are cautioned their speech and related activity on social media sites may be considered a reflection upon their position, and, in some instances, this Department," and (2) "Personal social media activity must not interfere with work duties or the operation of the Department."
After the Department adopted the policy, Hernandez, while off-duty, posted various news articles and memes created by others to his personal Facebook profile. Although he posted this content primarily to generate discussion among his friends and family, any member of the general public could view it. Hernandez's Facebook profile did not explicitly state that he worked as a Phoenix police officer, but he posted other content, such as photos of himself in uniform, from which that fact could be determined.
Hernandez made the four posts at issue here between September 2013 and January 2014. Because the content of the posts is relevant to whether they receive First Amendment protection, we describe them briefly below and have included copies in the appendix to this opinion.
The first post is a meme depicting a series of mugshot-like photos of men above a caption that reads, "The most common name for a convicted gang rapist in England is ... Muhammad." Below that in smaller type is additional text stating, "Note to the British media – these gangs are not comprised of ‘Asians’; they are Muslims."
The second post is a meme depicting a photo of what appears to be a British cab driver opening the door to his cab. The text accompanying the photo states, "You just got to love the Brits," followed by two paragraphs of text describing a supposed encounter between a "devout Muslim" and a cab driver in London:
A devout Muslim entered a black cab in London. He curtly asked the cabbie to turn off the radio because, as decreed by his religious teaching, he must not listen to music because, in the time of the prophet, there was no music, especially Western music which is the music of the infidel.
The cab driver politely switched off the radio, stopped the cab and opened the door. The Arab Muslim asked him, "What are you doing?" The cabbie answered,
"In the time of the prophet, there were no taxis, so piss-off [sic ] and wait for a camel!"
The third post is another meme titled, "Recent Contributions to Science by Islam." It depicts photos of four men to whom quotes are attributed in different fields of science. Each of the quotes attempts to mock the supposed contributions to science made by Islamic scholars or scientists.
The last post is an article published by the Minority Report Blog under the headline, "Military Pensions Cut, Muslim Mortgages Paid By US!" Beneath the headline is a blurb that states, "The Obama Administration cut Military pensions but found 300 million to send to Muslim's [sic ] overseas to help pay for their mortgages." The record does not contain the full text of the article, which could no longer be accessed by the time the Department conducted its investigation.
For more than five years, none of these posts came to the attention of the public or caused any turmoil within the Phoenix Police Department. That changed in June 2019 when the Plain View Project, which maintains a database of Facebook posts from law enforcement officers across the country, published a collection of posts from various officers of the Phoenix Police Department. Many of the posts reflected bias against racial or religious minorities or contained content that would be offensive to members of such groups. The Plain View Project's publication of these posts generated a firestorm of public criticism of the Department and considerable negative media attention.
The Plain View Project's database included 11 posts from Hernandez. The Phoenix Police Department's Professional Standards Bureau conducted an internal investigation to determine whether any of Hernandez's posts violated the Department's social media policy. Its investigation focused on the four posts described above. During an interview with investigators, Hernandez explained that he posted the items in question to drive discussion about issues that were in the news at the time, such as cultural assimilation and veterans' benefits.
The Bureau concluded that, taken together, the four posts violated section 3.27.9.B.(6) of the Department's social media policy. That provision states:
Department personnel are free to express themselves as private citizens on social media sites to the degree that their speech does not impair working relationships of this Department, are detrimental to the mission and functions of the Department, that undermine respect or public confidence in the Department, cause embarrassment to the Department or City, discredit the Department or City, or undermine the goals and mission of the Department or City.
The Bureau found that Hernandez's Facebook posts "potentially reduced or contributed to the erosion of public trust, were inflammatory to certain groups, and/or created dissention in the community by promoting hate, violence, racism, bias, or beliefs inconsistent with the Phoenix Police Department's Purpose Statement and Guiding Values." The Bureau also found that Hernandez's posts "do not align with the distinguishing features, essential functions and required knowledge as outlined in the City of Phoenix classification for a Police Sergeant." Based on the Bureau's findings, Hernandez faced discipline ranging from a suspension of 40 hours without pay up to termination.
Before the Department's Disciplinary Review Board or the Chief of Police decided what disciplinary sanction to impose, Hernandez and the Arizona Conference of Police and Sheriffs (known by the acronym AZCOPS) filed this lawsuit. Hernandez and AZCOPS sought a preliminary injunction barring the defendants—the City of Phoenix, the Chief of Police, and the Commander of the Professional Standards Bureau—from disciplining Hernandez for his Facebook posts. Any such disciplinary action, they argued, would violate his First Amendment free speech rights. The district court denied the request for a preliminary injunction and neither plaintiff appealed.
According to the complaint, Hernandez and a majority of the Phoenix Police Department sergeants and lieutenants are members of AZCOPS, and AZCOPS has provided legal representation to Hernandez throughout these proceedings.
Hernandez and AZCOPS, now joined by a third plaintiff, Phoenix Police Department Lieutenant Mark Schweikert, filed an amended complaint asserting three causes of action. The first cause of action alleges a claim under 42 U.S.C. § 1983 on two distinct theories. Plaintiffs contend, first, that the Department violated the First Amendment by retaliating against Hernandez for exercising his right to freedom of speech, and second, that certain provisions of the Department's social media policy are facially invalid under the First Amendment because they are vague and overbroad. Plaintiffs' second cause of action reasserts Hernandez's retaliation claim under the Arizona Constitution's First Amendment analogue. The third cause of action appears to allege a claim for municipal liability under a failure-to-train theory. Plaintiffs seek damages incurred during the Department's investigation of Hernandez's posts, such as attorney's fees; an injunction barring the Department from imposing any form of discipline against Hernandez for his posts; and an order enjoining further enforcement of the provisions of the social media policy that plaintiffs contend are facially invalid.
Defendants moved to dismiss the action under Federal Rule of Civil Procedure 12(b)(6) for failure to state a claim. The district court granted the motion as to Hernandez's First Amendment retaliation claim, plaintiffs' facial overbreadth challenge, and the related state law claim. The court concluded that Hernandez's retaliation claim failed because none of his posts addressed a matter of public concern, which meant they were not entitled to constitutional protection. The court rejected plaintiffs' facial overbreadth challenge on the ground that the Department's social media policy prohibited only those categories of speech that could reasonably be expected to disrupt the Department's mission and operations—ends that the Supreme Court has held are constitutionally permissible. The court did not dismiss plaintiffs' facial vagueness challenge given defendants' failure to adequately brief the issue. Because that portion of plaintiffs' claims survived, the court also declined to dismiss the municipal liability claim.
Discovery proceeded as to the facial vagueness challenge alone, but the court ultimately granted defendants' motion for summary judgment as to that claim. The court concluded that the challenged provisions are sufficiently clear to inform officers of "what the Policy proscribes in the vast majority of its intended applications." Having determined that plaintiffs could not establish an underlying constitutional violation, the court entered summary judgment for defendants on plaintiffs' remaining claim alleging municipal liability. Following entry of final judgment, plaintiffs timely appealed. II
We begin by addressing the district court's dismissal of Hernandez's First Amendment retaliation claim, which is predicated on the theory that the Department violated his right to freedom of speech by seeking to discipline him for his Facebook posts. Our court has developed a series of five sequential steps to analyze First Amendment retaliation claims brought by government employees, drawn from the tests established in Pickering and Mt. Healthy City Board of Education v. Doyle , 429 U.S. 274, 97 S.Ct. 568, 50 L.Ed.2d 471 (1977) : "(1) whether the plaintiff spoke on a matter of public concern; (2) whether the plaintiff spoke as a private citizen or public employee; (3) whether the plaintiff's protected speech was a substantial or motivating factor in the adverse employment action; (4) whether the state had an adequate justification for treating the employee differently from other members of the general public; and (5) whether the state would have taken the adverse employment action even absent the protected speech." Eng v. Cooley , 552 F.3d 1062, 1070 (9th Cir. 2009). Only the steps drawn from Pickering —the first, second, and fourth—are at issue here.
Under what has become known as the Pickering balancing test, a government employee bears the initial burden of showing that he spoke on a matter of public concern and that he did so in his capacity as a private citizen, rather than as an employee. Connick v. Myers , 461 U.S. 138, 147, 103 S.Ct. 1684, 75 L.Ed.2d 708 (1983) ; Eng , 552 F.3d at 1070–71. If the employee succeeds in making that threshold showing, his speech is entitled to constitutional protection and "the possibility of a First Amendment claim arises." Garcetti v. Ceballos , 547 U.S. 410, 418, 126 S.Ct. 1951, 164 L.Ed.2d 689 (2006). The burden then shifts to the government employer to show that it had an adequate justification for punishing the employee for his speech. To sustain its burden, the employer must show that "its own legitimate interests in performing its mission" outweigh the employee's right to speak freely. City of San Diego v. Roe , 543 U.S. 77, 82, 125 S.Ct. 521, 160 L.Ed.2d 410 (2004) (per curiam). The objective of this framework is to strike "a balance between the interests of the [employee], as a citizen, in commenting upon matters of public concern and the interest of the State, as an employer, in promoting the efficiency of the public services it performs through its employees." Pickering , 391 U.S. at 568, 88 S.Ct. 1731.
As with any balancing test requiring the weighing of competing interests, no hard-and-fast rules dictate where the balance is to be struck in a particular case. But a few general principles guide us. On one side of the balance, the Supreme Court has held that government employers have a strong interest in prohibiting speech by their employees that impairs close working relationships among co-workers, impedes performance of the speaker's job duties, interferes with the effective functioning of the employer's operations, or undermines the employer's mission. Rankin v. McPherson , 483 U.S. 378, 388, 390, 107 S.Ct. 2891, 97 L.Ed.2d 315 (1987) ; Connick , 461 U.S. at 151–52, 103 S.Ct. 1684 ; Pickering , 391 U.S. at 570, 572–73, 88 S.Ct. 1731. Government agencies are, after all, in the business of providing public services, so when an employee "begins to do or say things that detract from the agency's effective operation, the government employer must have some power to restrain her." Waters v. Churchill , 511 U.S. 661, 675, 114 S.Ct. 1878, 128 L.Ed.2d 686 (1994) (plurality opinion).
On the other side of the balance, government employees have an interest in speaking out "to bring to light actual or potential wrongdoing or breach of public trust" within their agencies, Connick , 461 U.S. at 148, 103 S.Ct. 1684, since they are often uniquely situated to inform the public about "government corruption and abuse." Dahlia v. Rodriguez , 735 F.3d 1060, 1066–67 (9th Cir. 2013) (en banc). The First Amendment interests at stake in this context have as much to do with the public's right to hear what an employee has to say about government operations as with the employee's right to speak freely. Roe , 543 U.S. at 82, 125 S.Ct. 521. The Supreme Court has held that the relative value of the employee's speech in advancing First Amendment interests factors into the balancing calculus, such that "the State's burden in justifying a particular discharge varies depending upon the nature of the employee's expression." Connick , 461 U.S. at 150, 103 S.Ct. 1684. The more substantially an employee's speech involves matters of public concern, the weightier the government employer's interests must be in preventing disruption of the workplace or impairment of the employer's mission. Id. at 152, 103 S.Ct. 1684.
In this case, the district court ruled that Hernandez's First Amendment retaliation claim did not survive the first step of the analysis, as his Facebook posts did not address matters of public concern. (No one contests that Hernandez spoke in his capacity as a private citizen rather than as an employee of the Phoenix Police Department when he made the posts in question.) Whether speech addresses a matter of public concern is an issue of law subject to de novo review. Id. at 148 n.7, 103 S.Ct. 1684 ; Cochran v. City of Los Angeles , 222 F.3d 1195, 1200 (9th Cir. 2000).
What constitutes speech on a "matter of public concern" remains somewhat hazy, despite the decades that have passed since the concept was first employed. Over the years, the Supreme Court has defined the concept in only the most general of terms. See, e.g., Connick , 461 U.S. at 146, 103 S.Ct. 1684 ("any matter of political, social, or other concern to the community"). The concept is useful primarily to draw a contrast with speech that is not entitled to constitutional protection in this context—namely, speech on "matters only of personal interest," such as speech addressing "a personal employment dispute" or "complaints over internal office affairs." Id. at 147, 148 n.8, 149, 103 S.Ct. 1684. Most speech falling outside that purely private realm will warrant at least some First Amendment protection and thus will qualify as speech on a matter of public concern for purposes of the Pickering balancing test. Tucker v. California Department of Education , 97 F.3d 1204, 1210 (9th Cir. 1996).
A notable exception has been speech involving pornographic material featuring police officers, which both the Supreme Court and our court have held does not involve speech on a matter of public concern despite the lack of connection to any sort of internal workplace dispute. See City of San Diego v. Roe , 543 U.S. 77, 84, 125 S.Ct. 521, 160 L.Ed.2d 410 (2004) (per curiam); Dible v. City of Chandler , 515 F.3d 918, 927 (9th Cir. 2008).
To determine whether an employee's speech addressed a matter of public concern, we consider the content of the statements, the form (time, place, and manner) of the statements, and the context in which the statements were made. Connick , 461 U.S. at 147–48, 103 S.Ct. 1684. Analyzing those elements here leads us to conclude that Hernandez's Facebook posts qualify as speech on matters of public concern.
As for content, Hernandez's posts assuredly did not address an internal workplace grievance or complaints about internal office affairs. They instead addressed matters of social or political concern that would be of interest to others outside the Phoenix Police Department. Even if only "a relatively small segment of the general public" might have been interested in the subject of Hernandez's posts, that is sufficient. Roe v. City and County of San Francisco , 109 F.3d 578, 585 (9th Cir. 1997).
Take the post with the article headlined "Military Pensions Cut, Muslim Mortgages Paid By US!" That article addressed, at least in part, the subject of government spending priorities, which has long been regarded as a matter of public concern. See Ulrich v. City and County of San Francisco , 308 F.3d 968, 978–79 (9th Cir. 2002). The post with the meme containing mugshot photos criticized the British media for the way it was identifying men convicted of rape in England. Subjects that receive media coverage "almost by definition involve[ ] matters ‘of public concern,’ " Roe , 109 F.3d at 585, so it follows that speech criticizing the media's coverage of a particular subject qualifies as a matter of public concern as well. The post with the meme about the British cab driver at least tangentially touched on matters of cultural assimilation and intolerance of religious differences in British society, which again are topics of social or political concern to some segments of the general public. And the last post with the meme about "Recent Contributions to Science by Islam" addressed subjects that, according to exhibits attached to Hernandez's complaint, received media attention at the time and sparked heated public debate.
Both form and context also weigh in Hernandez's favor. Hernandez posted each of the items at issue on his own time, outside the workplace, using his personal Facebook profile. The intended audience of his posts was not limited to Hernandez's fellow employees, and the posts could be viewed by any member of the general public. We agree with the Fourth Circuit that "publicly posting on social media suggests an intent to ‘communicate to the public or to advance a political or social point of view beyond the employment context.’ " Liverman v. City of Petersburg , 844 F.3d 400, 410 (4th Cir. 2016) (quoting Borough of Duryea v. Guarnieri , 564 U.S. 379, 398, 131 S.Ct. 2488, 180 L.Ed.2d 408 (2011) ). The context in which Hernandez's posts were made also supports the conclusion that the posts were not tied to any workplace dispute or grievance. The exhibits attached to Hernandez's complaint suggest that issues relating to immigration and cultural assimilation were topics of media attention at the time, and Hernandez alleges that he posted the items in question to foster discussion on those topics.
It is true that each of Hernandez's posts expressed hostility toward, and sought to denigrate or mock, a major religious faith and its adherents. The Supreme Court has made clear, however, that "[t]he inappropriate or controversial character of a statement is irrelevant to the question whether it deals with a matter of public concern." Rankin , 483 U.S. at 387, 107 S.Ct. 2891 (holding that speech expressing hope that a second attempt at assassinating the president would prove successful qualified as speech on a matter of public concern). Indeed, the Court has held that even overtly hateful speech denigrating gay men as a means of protesting the United States' tolerance of "homosexuality in the military" qualified as speech on a matter of public concern. Snyder v. Phelps , 562 U.S. 443, 454–55, 131 S.Ct. 1207, 179 L.Ed.2d 172 (2011). Speech that expresses hostility toward racial or religious minorities may be of particularly low First Amendment value at the next step of the Pickering balancing test (as we note below), but its distasteful character alone does not strip it of all First Amendment protection.
Having concluded that Hernandez's Facebook posts constitute speech on matters of public concern at the first step of the Pickering balancing test, we would ordinarily proceed to the next step and assess whether the Phoenix Police Department has shown an adequate justification for punishing Hernandez's otherwise protected speech. We cannot do so here, however, because the district court dismissed Hernandez's First Amendment retaliation claim at the motion-to-dismiss stage. The factual record before us is therefore limited to the allegations in the amended complaint and documents incorporated by reference. See Stoyas v. Toshiba Corp. , 896 F.3d 933, 938 (9th Cir. 2018). (In ruling on defendants' motion to dismiss, the district court properly declined to consider evidence introduced during the preliminary injunction hearing. See Fed. R. Civ. Proc. 12(d) ; Courthouse News Service v. Planet , 750 F.3d 776, 780 (9th Cir. 2014).) Although it seems likely that Hernandez's posts could impede the performance of his job duties and interfere with the Phoenix Police Department's ability to effectively carry out its mission, no evidence of the actual or potential disruptive impact caused by Hernandez's posts is properly before us at this stage of the proceedings. We therefore reverse the district court's dismissal of Hernandez's First Amendment retaliation claim and remand for further development of the factual record.
Because we reverse the district court's dismissal of Hernandez's retaliation claim, we also reverse the district court's dismissal of his related claim under the Arizona Constitution (Count 2). We do not disturb the district court's dismissal of plaintiffs' municipal liability claim (Count 3), as plaintiffs have not challenged dismissal of that claim on appeal, and the claim was not adequately pleaded in any event.
In remanding the case, we do not mean to suggest that the Department will face a particularly onerous burden to justify disciplining Hernandez for his posts, given the comparatively low value of his speech. Government employee speech that exposes wrongdoing or corruption within the employee's own agency lies at "the apex of the First Amendment" in this context. Moser v. Las Vegas Metropolitan Police Department , 984 F.3d 900, 906 (9th Cir. 2021). Needless to say, Hernandez's Facebook posts occupy a much lower rung on the First Amendment hierarchy, and indeed they touched on matters of public concern "in only a most limited sense." Connick , 461 U.S. at 154, 103 S.Ct. 1684. On the other side of the scale, a police department's determination that an officer's speech warrants discipline is afforded considerable deference, see Cochran , 222 F.3d at 1201, and police departments may permissibly consider the special status officers occupy in the community when deciding what limitations to place on officers' off-duty speech. Rankin , 483 U.S. at 390, 107 S.Ct. 2891 ("The burden of caution employees bear with respect to the words they speak will vary with the extent of authority and public accountability the employee's role entails."). Speech by a police officer that suggests bias against racial or religious minorities can hinder that officer's ability to effectively perform his or her job duties and undermine the department's ability to effectively carry out its mission. Locurto v. Giuliani , 447 F.3d 159, 182–83 (2d Cir. 2006).
III
We turn next to plaintiffs' facial challenge to certain provisions of the Department's social media policy. Plaintiffs focus much of their attack on a provision prohibiting social media posts "that are detrimental to the mission and functions of the Department, that undermine respect or public confidence in the Department, could cause embarrassment to the Department or City, discredit the Department or City, or undermine the goals and mission of the Department or City." Plaintiffs contend that this provision and provisions employing similar language should be facially invalidated because they are unconstitutionally overbroad and vague.
A
We largely agree with the district court's rejection of plaintiffs' facial overbreadth challenge. In analyzing such challenges in the public employment context, we apply a modified Pickering balancing analysis that closely tracks the test used for First Amendment retaliation claims. See United States v. Treasury Employees , 513 U.S. 454, 466–68, 115 S.Ct. 1003, 130 L.Ed.2d 964 (1995) ( NTEU ). We first ask whether the challenged restriction applies to employees' speech in their capacity as private citizens on matters of public concern. If it does, we then ask whether the government has an adequate justification for treating its employees differently from other members of the general public. Barone v. City of Springfield , 902 F.3d 1091, 1102 (9th Cir. 2018).
Because the challenge in this context targets a prospective restriction on a broad category of expression, rather than punishment imposed after-the-fact for a specific instance of speech, the government bears a heavier burden to justify the scope of the restriction. Id. at 1105. The government must show that the combined First Amendment interests of the public and current and future employees are outweighed by the speech's " ‘necessary impact on the actual operation’ of the Government." NTEU , 513 U.S. at 468, 115 S.Ct. 1003 (quoting Pickering , 391 U.S. at 571, 88 S.Ct. 1731 ). The harms on which the government relies must be "real, not merely conjectural," and the proposed restriction must "in fact alleviate these harms in a direct and material way." Id. at 475, 115 S.Ct. 1003 (quoting Turner Broadcasting System, Inc. v. FCC , 512 U.S. 622, 664, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994) ). A restriction on speech is facially overbroad if, applying this standard, "a substantial number of its applications are unconstitutional, judged in relation to the [provision's] plainly legitimate sweep." United States v. Stevens , 559 U.S. 460, 473, 130 S.Ct. 1577, 176 L.Ed.2d 435 (2010) (quoting Washington State Grange v. Washington State Republican Party , 552 U.S. 442, 449 n.6, 128 S.Ct. 1184, 170 L.Ed.2d 151 (2008) ).
At the first step of the analysis, defendants do not dispute that the challenged provisions apply to employee speech on matters of public concern made in an employee's capacity as a private citizen, outside the scope of official duties. That fact is plain from the text of the policy, which applies to all social media use by the Department's employees, even when off-duty. The challenged provisions are found in a section of the policy that explicitly regulates "personal use" of social media, and because the policy contains no subject-matter carve outs, speech on matters of public concern will necessarily be covered.
We next ask whether the Department has advanced an adequate justification for the restrictions. Because the district court dismissed plaintiffs' facial overbreadth challenge on the pleadings, we are again left without a developed factual record as to the harms on which the Department relies to justify imposition of its social media policy. But most of the restrictions challenged here directly promote the same interests that the Supreme Court has already held to be valid bases for imposing restrictions on public employee speech. As noted earlier, government employers have a strong interest in prohibiting speech by their employees that undermines the employer's mission or hampers the effective functioning of the employer's operations. Rankin , 483 U.S. at 388, 390, 107 S.Ct. 2891 ; Connick , 461 U.S. at 151–52, 103 S.Ct. 1684. That interest justifies the policy's restrictions on social media posts that are "detrimental to the mission and functions of the Department" or which "undermine the goals and mission of the Department or City." Police departments also have a strong interest in maintaining a relationship of trust and confidence with the communities they serve, see Roe , 543 U.S. at 81, 125 S.Ct. 521, which justifies the policy's restriction on speech that would "undermine respect or public confidence in the Department." Given how closely these clauses of the policy track interests that the Department may constitutionally pursue, we cannot say that a "substantial number" of the policy's applications are unconstitutional, judged in relation to the policy's "plainly legitimate sweep." Stevens , 559 U.S. at 473, 130 S.Ct. 1577.
We reach a different conclusion with respect to the clauses of the policy prohibiting speech that would "cause embarrassment to" or "discredit" the Department—most notably the provision that states: "Employees are prohibited from using social media in a manner that would cause embarrassment to or discredit the Department in any way."
We do not think plaintiffs' facial overbreadth challenge to these clauses may be rejected at the motion-to-dismiss stage. The Department does not have a legitimate interest in prohibiting speech merely because the Department might find that speech embarrassing or discrediting, just as it does not have a legitimate interest in prohibiting all negative or disparaging speech about the Department. See Barone , 902 F.3d at 1105–06. The Department has a legitimate interest in prohibiting embarrassing or discrediting speech to the extent such speech could reasonably be expected to disrupt the workplace, hinder the Department's mission, or undermine the public's confidence in and respect for the Department. But the social media policy already prohibits speech generating those detrimental effects. And unlike those more targeted prohibitions, the "embarrass" and "discredit" clauses are entirely self-regarding and not constrained by any demonstrable impact on the Department or its ability to function. It is thus far from clear what additional work the "embarrass" and "discredit" clauses could be doing here, beyond broadening the scope of the policy to authorize discipline for social media activity that the Department may not have a sufficiently strong interest in prohibiting. That is particularly concerning from an overbreadth standpoint because virtually all speech that lies at the core of First Amendment protection in this area—for example, speech exposing police misconduct or corruption—could be expected to embarrass or discredit the Department in some way.
We do not foreclose the possibility that the Department may be able to produce evidence at the summary judgment stage justifying the clauses prohibiting speech that would embarrass or discredit the Department. But in the absence of a developed factual record, we cannot conclude that plaintiffs' facial overbreadth challenge to those clauses fails as a matter of law.
Plaintiffs separately challenge section 3.27.9.B.(7) of the social media policy, which provides that "Department personnel may not divulge information gained while in the performance of their official duties." We do not think plaintiffs' facial overbreadth challenge to this provision can be rejected at the motion-to-dismiss stage either. Although the Department has a strong interest in prohibiting the disclosure of confidential information, such as information that could jeopardize ongoing investigations, the challenged provision sweeps much more broadly. See Moonin v. Tice , 868 F.3d 853, 865, 867 (9th Cir. 2017). It prohibits the disclosure of any information gained while on the job. Yet public employees are uniquely positioned to expose wrongdoing or corruption within their agencies precisely because they acquire information while on the job to which the public otherwise lacks access. A policy that prohibits public employees from divulging any information acquired while on the job would silence speech that warrants the strongest First Amendment protection in this context. For that reason, the provision challenged here "must be tailored to protect information the government has a legitimate interest in keeping confidential." Id. at 873. Because plaintiffs' challenge was resolved on the pleadings, the Department has not yet had an opportunity to produce evidence attempting to establish that this provision is appropriately tailored.
In short, we hold that the district court properly rejected plaintiffs' facial overbreadth challenge except as to the clauses prohibiting social media activity that (1) would cause embarrassment to or discredit the Department or (2) divulges any information gained while in the performance of official duties.
B
We affirm the district court's rejection of plaintiffs' facial vagueness challenge to the same provisions discussed above. To prevail on this challenge, plaintiffs must show either that the challenged provisions fail to afford employees "a reasonable opportunity to understand what conduct [they] prohibit[ ]," or that the provisions permit "arbitrary and discriminatory enforcement." Hill v. Colorado , 530 U.S. 703, 732, 120 S.Ct. 2480, 147 L.Ed.2d 597 (2000). If it is clear what the challenged provisions proscribe "in the vast majority of [their] intended applications," they cannot be deemed unconstitutionally vague on their face. Id. at 733, 120 S.Ct. 2480 (citation omitted).
Plaintiffs argue that the Department's social media policy is unconstitutionally vague because the challenged provisions offer no discernible standard that allows officers to determine what types of social media posts would subject them to discipline. We disagree. As discussed earlier, most of the challenged provisions simply track the language that the Supreme Court has used to describe the circumstances in which government employers have a strong interest in restricting employee speech. The provisions draw from Supreme Court precedent to prohibit speech that undermines, interferes with, or is detrimental to the Department's goals and mission and its relationship with the public—concepts that it would be infeasible for the Department to describe with exhaustive specificity. See Arnett v. Kennedy , 416 U.S. 134, 161, 94 S.Ct. 1633, 40 L.Ed.2d 15 (1974) (plurality opinion). Like many employment policies, the challenged provisions are framed in broad and general terms that nonetheless provide sufficient guidance to employees as to the types of social media posts that are prohibited. See Waters , 511 U.S. at 673, 114 S.Ct. 1878 (plurality opinion) (noting that policies governing public employee speech may be framed in language that might be deemed impermissibly vague if applied to the public at large). Further, the Department already expects employees to be familiar with and promote its goals and mission. Plaintiffs have not demonstrated that officers tasked with adhering to the social media policy will, in most cases, be unable to determine whether their speech will have a deleterious impact on the Department's operations.
In advancing their facial vagueness challenge, plaintiffs rely heavily on the deposition testimony of Lieutenant Eric Pagone, a former member of the Professional Standards Bureau who was charged with supervising investigations into alleged violations of the Department's social media policy. Pagone was asked to opine on whether hypothetical social media posts addressing hot-button political issues, such as abortion, would run afoul of the policy. He could not give definitive opinions in response to the hypotheticals posed to him, explaining that in most instances he would need additional contextual information to make an informed judgment. Rather than demonstrating the facial invalidity of the Department's policy, Pagone's answers merely reflect the fact that deciding whether any given social media post violates the policy involves a heavily fact-and context-dependent exercise that often cannot be performed with only the bare content of a hypothetical post. In any event, as alluded to earlier, uncertainty about the correct resolution of edge cases "will not warrant facial invalidation if it is clear what the [policy] proscribes ‘in the vast majority of its intended applications.’ " California Teachers Association v. State Board of Education , 271 F.3d 1141, 1151 (9th Cir. 2001) (quoting Hill , 530 U.S. at 733, 120 S.Ct. 2480 ). In our view, the Department's social media policy meets that standard.
* * *
We affirm the district court's dismissal of plaintiffs' facial overbreadth challenge, with the two caveats noted above. We also affirm the district court's entry of summary judgment for defendants on plaintiffs' facial vagueness challenge and their municipal liability claim. We reverse the district court's dismissal of Hernandez's First Amendment retaliation claim and his related claim under the Arizona Constitution.
AFFIRMED in part, REVERSED in part, and REMANDED.
The parties shall bear their own costs on appeal.
Appendix ?
?
?
?