Affirmed in part, reversed in part, and remanded by published opinion. Judge WILKINSON wrote the opinion, in which Judge TRAXLER and Judge HENDRICKS joined.
WILKINSON, Circuit Judge:
Two police officers challenge disciplinary actions for violations of their Department’s social networking policy. The district court denied relief on most of their claims. While we are sensitive to the Department’s need for discipline throughout the chain of command, the policy here and the disciplinary actions taken pursuant to it would, if upheld, lead to an utter lack of transparency in law enforcement operations that the First Amendment cannot countenance. For the reasons that follow, we affirm in part, reverse in part and remand for further proceedings.
I.
The pertinent facts in this case are not in dispute. Plaintiffs Herbert Liverman and Vance Richards were veteran police officers in the City of Petersburg’s Police Department. Both served as field officers under Chief John Dixon, who led the Department. Dixon in turn served under the general direction of the City Manager.
In April 2013, Chief Dixon issued a general order revising the Department’s social networking policy. That policy governs officers’ use of social media platforms. The preface to the revised policy prohibits in sweeping terms the dissemination of any information “that would tend to discredit or reflect unfavorably upon the [Department] or any other City of Petersburg Department or its employees.” J.A. 161. The central provision of the policy, which we will refer to as the Negative Comments Provision, states:
Negative comments on the internal operations of the Bureau, or specific conduct of supervisors or peers that impacts the public’s perception of the department is not protected by the First Amendment free speech clause, in accordance with established case law.
J.A. 162. Another provision, which we label the Public Concern Provision, specifies:
Officers may comment on issues of general or public concern (as opposed to personal grievances) so long as the comments do not disrupt the workforce, interfere with important working relationships or efficient work flow, or undermine public confidence in the officer. The instances 'must be judged on a case-by-case basis.
Id.
The policy nonetheless “strongly discourages employees from posting information regarding off-duty activities” and provides that violations will be forwarded to .the Chief of Police for “appropriate disciplinary action.” J.A. 163.
This case concerns the Department’s application of the social networking policy to the following conversation between Liver-man and Richards. While off-duty on June
17, 2013, Liverman posted a message to his Facebook page:
Sitting- here reading posts referencing rookie cops becoming instructors. Give me a freaking break, over 15 years of data collected by the FBI in reference to assaults on officers and officer deaths shows that on average it takes at least 5 years for ah officer to acquire the necessary skill set to know the job and perhaps even longer to acquire the knowledge to teach other officers. But in todays world of instant gratification and political correctness we have rookies in specialty units, working as field training officer’s and even as instructors. Becoming a master of your trade is essential, not only does your life depend on it but more importantly the fives of others. Leadership is first learning, knowing and then doing.
J.A. 398. More than thirty people “liked” or commented on this post! Richards, also off-duty at the time, commented as follows:
Well said bro, I agree 110%... Not to mention you are seeing more and more younger Officers being promoted in a Supervisor/ or roll. It’s disgusting and makes me sick to my stomach DAILY. LEO Supervisors should be promoted by experience... And what comes with experience are “experiences” that “they” can pass around to the Rookies and younger less experienced Officers. Perfect example, and you know who I’m talking about.How can ANYONE look up, or give respect to a SGT in Patrol with ONLY 1 l/2yrs experience in the street? Or less as a matter of fact. It’s a Law Suit waiting to happen. And you know who will be responsible for that Law Suit? A Police Vet, who knew tried telling and warn the admin for promoting the young Rookie who was too inexperienced for that roll to begin with. Im with ya bro_smh
J.A. 399. Later that day, Liverman responded to Richards with a comment of his own:
There used to be a time when you had to earn a promotion or a spot in a specialty unit.. .but now it seems as though anything goes and beyond officer safety and questions of liability, these positions have been “devalued”-... and when something has no value, well it is worthless.
Id. Richards then replied:
Your right.The next 4yrs can’t get here fast enough... From what I’ve been seeing I don’t think I can last though. You know the old “but true” saying is.... Your Agency is only as good as it’s Leader(s)... It’s hard to “lead by example” when there isn’t one... .smh
J.A. 400.
Among those who liked or commented on the Facebook postings, most were current or former Department officers. Two sergeants, Liverman’s and Richards’s supervisors, learned of the exchange and notified Chief Dixon of the issue. Dixon determined that the statements violated the Department’s social networking policy and instructed the sergeants to discipline the officers. In the disciplinary action forms, the Department stated that Liver-man’s follow-up comment and both of Richards’s comments violated the Negative Comments Provision. They each received an oral reprimand and six months’ probation, but were advised that such discipline would not affect their eligibility for promotion. Both the City Manager and Human Resources Director signed the personnel action forms indicating their probationary status.
Several weeks later, however, Chief Dixon altered the- qualifications for promotion. The new protocol expressly excluded any officers on probation from participating in the promotion process. Accordingly, when Liverman and Richards applied for open sergeant positions, the Department notified them that they were ineligible to sit for the promotional exam.
On October 1, 2013, the two officers sent a letter informing the City that they intended to challenge the disciplinary actions. Shortly thereafter, Liverman and Richards were the subject of several complaints and investigations within the Department. Based on the findings, Chief Dixon decided to fire Liverman, but Liyer-man resigned before receiving notice of his termination.
On March 5, 2014, Liverman and Richards filed a six-count complaint in federal district court under 42 U.S.C. § 1983, seeking damages and other relief against Chief Dixon and the City for various violations of the First Amendment. The two officers claimed that the social networking policy infringed their free speech rights in Counts One (Liverman) and Two (Richards). Liverman and Richards also challenged the adverse disciplinary actions taken pursuant to the policy in Counts Three and Four, respectively. Finally, they alleged in Counts Five and Six that the Department instituted subsequent investigations against them in retaliation for proceeding with the instant suit.
The district court granted Liverman summary judgment on his claim that the social networking policy infringed his right to free speech, but nonetheless found that Chief Dixon was entitled to qualified. immunity because the policy fell within a gray zone. On Liverman’s challenge to the disciplinary action, the court found that qualified immunity again shielded Dixon’s decision because the contours of protected speech in this area were not clearly established. The district court next denied relief on Richards’s challenges to the policy and the discipline, holding that Richards’s speech was purely personal and thus not protected by the First Amendment. For both of their retaliation claims, the court concluded that the subsequent internal investigations were not retaliatory. This appeal followed.
n.
The legal framework governing public employee speech claims is well known. Public employees may not “be compelled to relinquish the First.Amendment rights they would otherwise enjoy as citizens to comment on matters of public interest.”
Pickering v. Bd. of
Educ., 391 U.S. 563, 568, 88 S.Ct. 1731, 20 L.Ed.2d 811 (1968). Underlying this principle is the recognition that “public employees are often the members of the community who are likely to have informed opinions as to the operations of their public employers.”
City of San Diego v. Roe,
543 U.S. 77, 82, 125 S.Ct. 521, 160 L.Ed.2d 410 (2004) (per curiam). Nonetheless, a citizen who accepts public employment “must accept certain limitations on his or her freedom.”
Garcetti v. Ceballos,
547 U.S. 410, 418, 126 S.Ct. 1951, 164 L.Ed.2d 689 (2006). Government employers enjoy considerable discretion to manage their operations, and the First Amendment “does not require a public office tó be run as a roundtable for employee complaints over internal office affairs.”
Connick v. Myers,
461 U.S. 138, 149, 103 S.Ct. 1684, 75 L.Ed.2d 708 (1983).
Courts begin the First Amendment inquiry by assessing whether the speech at issue relates to a matter of public, concern.
See Pickering,
391 U.S. at 568, 88 S.Ct. 1731. If speech is purely personal, it is not protected and the inquiry is at an end. If, however, the speech is
of public concern, courts must balance “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.”
Id.; see also Connick,
461 U.S. at 142, 103 S.Ct. 1684.
Against this backdrop, we turn to the officers’ First Amendment challenges to the Department’s social networking policy and the subsequent disciplinary actions taken against them.
A.
The district court granted summary judgment to Liverman on his challenge to the social networking policy, but denied Richards’s parallel claim on the grounds that his speech was not protected by the First Amendment. We hold that the Department’s social networking policy is unconstitutionally overbroad and, for the following reasons, award judgment to Richards on his claim as well.
Although regulations on social media use may appear to present novel issues, we agree with the district court that such questions are amenable to the traditional analysis set forth in
Connick
and
Pickering.
Indeed, the particular attributes of social media fit comfortably within the existing balancing inquiry: A social media platform amplifies the distribution of the speaker’s message—which favors the employee’s free speech interests—but also increases the potential, in some cases exponentially, for departmental disruption, thereby favoring the employer’s interest in efficiency. What matters to the First Amendment analysis is not only the medium of the speech, but the scope and content of the restriction.
Here we deal with a broad social networking policy setting forth the parameters of public employee speech. In
United States v. Nat’l Treasury Employees Union (NTEU),
513 U.S. 454, 115 S.Ct. 1003, 130 L.Ed.2d 964 (1995), the Supreme Court addressed how courts should apply
Pickering
when a generally applicable statute or regulation (as opposed to a post-hoc disciplinary action) operates, as a prior restraint on speech.
NTEU
involved a statute that prohibited federal employees from accepting any compensation for giving speeches or writing articles, even when the topic was unrelated to the employee’s official duties.
See id.
at 457, 115 S.Ct. 1003. Emphasizing that the honoraria ban impeded a “broad category of expression” and “chills potential speech before it happens,” the Court held that “the Government’s burden is greater with respect to this statutory restriction on expression than with respect to [the] isolated disciplinary action[s]” in
Pickering
and its progeny.
Id.
at 467, 468, 115 S.Ct. 1003. Accordingly, “[t]he Government must show that the interests of both potential audiences and a vast group of present and future employees in a broad range of present and future expression are outwéighed by that expression’s ‘necessary impact on the actual operation' of the Government.”
Id.
at 468, 115 S.Ct. 1003 (quoting
Pickering,
391 U.S. at 571, 88 S.Ct. 1731). Further, the government “must demonstrate that the recited harms are real, not merely conjectural, and that the regulation will in fact alleviate these harms in a direct and material way.”
Id.
at 475, 115 S.Ct. 1003.
The threshold question in this case is whether the Department’s policy regulates officers’ rights to speak on matters of public concern. There can be no doubt that it does: the restraint is a virtual blanket prohibition on all speech critical of.the government employer. The explicit terms of the Negative Comments Provision prevent plaintiffs and any other officer from
making unfavorable comments on the operations and policies of the Department, arguably the “paradigmatic” matter of public concern.
Sanjour v. EPA,
56 F.3d 85, 91 (D.C. Cir. 1995);
see also Roe,
543 U.S. at 80, 125 S.Ct. 521.
Weighing the competing interests on either side of the
Pickering/NTEU
balance, we begin by noting the astonishing breadth of the social networking policy’s language. The policy seeks to prohibit the dissemination of any information on social media “that would tend to discredit or reflect unfavorably upon the [Department].” J.A. 161. In particular, the Negative Comments Provision proscribes “Megative comments on the internal operations of the Bureau”—which could be just about anything—or on the “specific conduct of supervisors or peers”—which, again, could be just about anything. J.A. 162.
The interests of “present and future employees” and their “potential audiences” in such speech is manifestly significant.
See NTEU,
513 U.S. at 468, 115 S.Ct. 1003. We do not, of course, discount the capacity of social media to amplify expressions of rancor and vitriol, with all its potential disruption of workplace relationships that
Con-nick
condemned. But social networking sites like Facebook have also emerged as a hub for sharing information and opinions with one’s larger community. And the speech prohibited by the policy might affect the public interest in any number of ways, including whether the Department is enforcing the law in an effective and diligent manner, or whether it is doing so in a way that is just and evenhanded to all concerned. The Department’s law enforcement policies could well become a matter of constructive public debate and dialogue between law enforcement officers and those whose safety they are sworn to protect. After all, “government employees are often in the best position to know what ails the agencies for which they work.”
Waters v. Churchill,
511 U.S. 661, 674, 114 S.Ct. 1878, 128 L.Ed.2d 686 (1994) (plurality opinion). But this policy will cut short all of that. To repeat, it squashes speech on matters of public import at the very outset.
Because the Department’s social - networking policy unmistakably imposes a significant burden on expressive activity, we next consider whether the Department has adequately established “real, not merely conjectural” harms to its operations.
See NTEU,
513 U.S. at 475, 115 S.Ct. 1003. Chief Dixon’s primary contention is that divisive social media use undermines the Department’s interests in maintaining camaraderie among patrol officers and building community trust. These are, to be sure, legitimate interests. “When close working relationships are essential to fulfilling public responsibilities, a wide degree of deference to the employer’s judgment is appropriate.”
Connick,
461 U.S. at 151-52, 103 S.Ct. 1684. And such deference applies with special force to police departments because they are “paramilitary—discipline is demanded, and freedom must be correspondingly denied.”
Maciariello v. Sumner,
973 F.2d 295, 300 (4th Cir. 1992).
Here, however, the Department fails to satisfy its burden of demonstrating actual disruption to its mission. Apart from generalized allegations of budding “divisiveness” and claims that some “patrol officers sought [shift] transfers,” J.A. 502, Chief Dixon presented no evidence of any material disruption arising from plaintiffs’—or any other officer’s—comments on social media. We do not deny that officers’ social media use might present some potential for division within the ranks, particularly given the broad audience on Facebook. But the speculative ills targeted by the social networking policy are not sufficient
to justify such sweeping restrictions on officers’ freedom to debate matters of public concern.
See Connick,
461 U.S. at 152, 103 S.Ct. 1684;
McVey v. Stacy,
157 F.3d 271, 279 (4th Cir. 1998) (Murnaghan, J., concurring in part and concurring in the judgment) (“A stronger showing of public interest in the speech requires a concomitantly stronger showing of government-employer interest to overcome it.”).
Defendants’.fallback argument is that, even if the Negative Comments Provision itself is overbroad, the Public Concern Provision significantly narrows the reach of the social networking policy. This second provision, which permits comments on “issues of general or public concern ... so long as the comments do not disrupt the workforce,” J.A. 162, is ostensibly more aligned with the case-by-case analysis of
Connick
and
Pickering.
But the milder language in a single provision does not salvage the unacceptable over-breadth of the social networking policy taken as a whole. Indeed, the Public Concern Provision does not purport to nullify or otherwise supersede the blanket censorship endorsed by the Negative Comments Provision. If the Department wishes to pursue a narrower social media policy, then it can craft a regulation that does not have the chilling effects on speech that the Supreme Court deplored. We cannot, however, allow the current policy to survive as a management- and disciplinary mechanism.
B.
Plaintiffs next assert that the district court erred in dismissing their challenges to the Department’s disciplinary actions. We agree..In fact, the facial overbreadth of the social networking policy is borne out by the disciplinary actions taken pursuant to it.
When evaluating an ex post disciplinary action, rather than an ex ante restraint on speech, the nature of our review is narrower than the analysis under
NTEU.
In this context, our court has adopted the traditional
Connick/Pickering
three-part test to determine whether a public employee has sustained a First Amendment challenge to an adverse employment action. First, we determine whether the employee spoke as a citizen on a matter of public concern. Second, we evaluate whether the employee’s interest in First Amendment expression outweighs the employer’s interest'in the efficient operation of the workplace. And finally, we decide whether the protected speech was a substantial factor in the employer’s decision to take adverse employment action.
McVey,
157 F.3d at 277-78.
The first inquiry, once again, is whether Liverman and Richards spoke on matters of public concern. “Speech involves a matter of public concern when it involves an issue of social, political, or other interest to a community.”
Kirby v. City of Elizabeth City,
388 F.3d 440, 446 (4th Cir. 2004). In resolving this question, the Supreme Court has directed courts to examine the “content, form, and context of a given statement.”
Connick,
461 U.S. at 147-48, 103 S.Ct. 1684. Although defendants are certainly correct that “personal complaints and grievances about conditions of employment” are not matters of public concern,
Campbell v. Galloway,
483 F.3d 258, 267 (4th Cir. 2007), they misconstrue the thrust of Liverman’s and Richards’s comments.
The form and context of the comments indicate that plaintiffs did -in fact speak on an issue of public concern. Regarding the form of speech, we find it significant that the officers chose Facebook as the forum for their communication. As our colleague Judge Traxler has recognized, Facebook is a dynamic medium through which users
can interact and share news stories or opinions with members of their community.
See Bland v. Roberts,
730 F.3d 368, 385 (4th Cir. 2013). Similar to writing a letter to a local newspaper,
see Pickering,
391 U.S. at 569-70, 88 S.Ct. 1731, 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,”
Borough of Duryea v. Guarnieri,
564 U.S. 379, 131 S. Ct. 2488, 2501, 180 L.Ed.2d 408 (2011). Further, the officers’ Facebook comments were prompted by other “posts referencing rookie cops becoming instructors.” J.A. 398. Accordingly, the context of the speech buttresses our conclusion that Liverman and Richards were not simply airing personal grievances but rather were joining an ongoing public debate about the propriety of elevating inexperienced police officers to supervisory roles.
The content of the Facebook comments further confirms that they dealt with issues of public import. Defendants seek to carve up the Facebook colloquy and assert that Liverman’s and Richards’s comments should be considered separately. Yet this court has previously rejected attempts to “divide[ ] [speech] into discrete components to conduct a constitutional analysis on each.”
Stroman v. Colleton Cty. Sch. Dist.,
981 F.2d 152, 157 (4th Cir. 1992). Because we do not have “license to ignore the. portions” of the communication that touch on a matter of public concern, we must view the statements “as a single expression of speech to be considered in its entirety.”
Campbell,
483 F.3d at 267. This approach is consistent with the typical experience on social media, where users engage .in interactive discussions through a series of posts and comments. Liverman’s initial post invited others to pick up on his observations; Richards' responded, and they began a public dialogue about the Department’s promotion policies. Their comments, therefore, should be read in conjunction as part of a single conversation on the qualifications of instructors and the increasing number of rookies thrust into teaching roles.
Taken together, plaintiffs’ statements stand in stark contrast to the sort of “individualized concerns” this court has characterized as personal grievances.
See Brooks v. Arthur,
685 F.3d 367, 374 (4th Cir. 2012). Each veteran officer grounded his statements in specialized knowledge and expressed a general “concern about the inability of the [Department] to carry out its vital public mission effectively.”
Cromer v. Brown,
88 F.3d 1315, 1325-26 (4th Cir. 1996). Liverman’s initial post cited an FBI study that underscored the danger of promoting green officers, and his subsequent comment noted the implications for “officer safety and questions of liability.” J.A. 398-99. Notwithstanding his more colloquial tone, Richards touched on the same issues of public import in his responses. First, he agreed with Liverman’s observations and echoed the concerns about “more and more younger Officers being promoted.” J.A. 399. Then he turned to the issue of skill development raised by the FBI study and concluded that “LEO .Supervisors should be promoted by experience” and the “Agency is only as good as it’s Leader(s).” J.A. 399-400.
Whether plaintiffs were correct or not in their views is not the issue. The matter they addressed was of more than personal import. We thus have no trouble finding that plaintiffs’ Facebook comments, which addressed risks posed by the Department’s inexperienced supervisors, raised issues of public concern.
See, e.g., Brooks,
685 F.3d at 375 (explaining that when an employer’s practice “crosses a line to the point that imperils the public welfare ... then the public would rightly be concerned about the matter”);
Goldstein v. Chestnut
Ridge Volunteer Fire Co.,
218 F.3d 337, 353 (4th Cir. 2000) (holding that firefighter’s complaints about inadequate training and unsafe procedures during emergency calls were matters of public concern).
The second and third prongs of the
Connick/Pickering
inquiry are not in genuine dispute. Serious concerns regarding officer training and supervision are weighty matters that must be offset by an equally substantial workplace disruption. Chief Dixon failed to establish a reasonable apprehension that plaintiffs’ social media comments would meaningfully impair the efficiency of the workplace.
See Maciariello,
973 F.2d at 300. Finally, defendants do not seriously dispute that plaintiffs’ Facebook comments were a substantial factor in the decision to discipline them—indeed, both disciplinary action forms cited violations of the Negative Comments Provision as the sole basis for the oral reprimand and probation. J.A. 427-28.
In light of the First Amendment protection accorded to the officers’ posts, we conclude that the discipline they received pursuant to the social networking policy was unconstitutional.,
C.
In the alternativej Dixon contends that the Department’s decisions to adopt the social. networking policy and take disciplinary action pursuant to the Negative Comments Provision are entitled to qualified immunity. The doctrine of qualified immunity shields government officials “who commit constitutional violations but who, in light of clearly established law, could reasonably believe that their actions were lawful.”
Henry v. Purnell,
652 F.3d 524, 531 (4th Cir. 2011) (en banc).
Having found that Dixon violated the officers’ First Amendment rights, we must consider whether such rights were “clearly established”- at the time of the events at issue. “We do not require a case directly on point” in order to conclude that the law was clearly established, “but existing precedent must have placed the statutory or constitutional question beyond debate.”
Ashcroft v. al-Kidd,
563 U.S. 731, 131 S.Ct. 2074, 2083, 179 L.Ed.2d 1149 (2011).
Dixon first argues that he acted reasonably in adopting the social networking policy because the policy purported to track the subtle balancing calculus in
Pickering.
We agree that officials “are not liable for bad guesses in gray areas,”
Maciariello,
973 F.2d at 298, and “do not expect [police chiefs] to be judges and to have the training to sort through every intricacy of case law.”
Bland,
730 F.3d at 393. But this case does not involve gray areas: the right against such a sweeping prior restraint on speech was clearly established and then some. Indeed, it is axiomatic that the government may not ban speech on the ground that it expresses an objecting viewpoint.
See R.A.V. v. City of St. Paul,
505 U.S. 377, 112 S.Ct. 2538, 120 L.Ed.2d 305 (1992). Accordingly, there can be no doubt that prohibiting any “[n]ega-tive comments on the internal operations of the Bureau, or specific conduct of supervisors or peers”—even comments of great public concern—violates the First Amendment. J.A. 162.
Dixon also asserts that the disciplinary actions taken pursuant to the policy were reasonable in light of the vague boundaries distinguishing public and private speech. Given the patent unconstitutionality of the social networking policy,, however, efforts to enforce the policy are similarly suspect. After all, the core of the policy was a prohibition on legitimate speech and, as
detailed above, we have little difficulty locating the officers’ speech within this protected sphere. Plaintiffs raised serious concerns regarding the Department’s training programs and the promotion of inexperienced supervisors, both of which are matters of public concern. As this court has held time and again, it was clearly established law that such speech is protected by .the First Amendment.
See, e.g., Brooks,
685 F.3d at 375;
Goldstein,
218 F.3d at 353.
We appreciate the need for order and discipline in the ranks.
See Maciariello,
973 F.2d at 300 (recognizing that “greater latitude is afforded to police department officials in dealing with dissension”). At the same time, we cannot countenance an arm of government with such enormous powers being removed to this extent from public scrutiny. This is not an all-or-nothing matter; there is a balance to be struck. But the Department’s social networking policy, and the disciplinary actions taken to enforce it, lean too far to one side. We therefore hold that Chief Dixon is not entitled to qualified immunity.
III.
Finally, plaintiffs argue that Dixon retaliated against them for filing their First Amendment suit. We agree with the district court that their retaliation claims are without merit.
Plaintiffs argue that the retaliation took the form of investigating their conduct on the force. We cannot conclude, however, that plaintiffs have raised an issue of triable fact that the investigations were pretextual.
See McVey,
157 F.3d at 277-78. There were independent bases for each investigation. Liverman was investigated twice. In notifying the City of his First Amendment claims, Liverman requested a wide range of personnel records. While searching for responsive documents, the Department discovered that Liverman had sent sexually explicit emails to a female officer. The Department launched an investigation for sexual harassment, during which Liverman admitted to engaging in sexual misconduct on Department, property and while on duty. Additionally, Liv-erman was investigated for ah incident in which he ignored Chief Dixon’s orders and failed to maintain his duty post as directed.
Richards was also investigated twice. Both inquiries were opened as a result of complaints initiated not by Chief Dixon but by his fellow officers. The first complaint related to a report Richards allegedly made to the media about another officer’s spouse. The Department concluded the investigation within one week, after Richards demonstrated his innocence. The second complaint arose from his involvement with the Department's Shop with a Cop program. Once again, the Department determined that the allegations were unfounded.
Apart from generalized assertions regarding the existence of the investigations, plaintiffs fail to offer any evidence that the investigations were retaliatory. Far from groundless “fishing expeditions,” Appellants’ Br. at 34, each arose from discrete allegations of misconduct. Without more, we see no reason to question the legitimacy of the Department’s investigations. After all, simply filing a
Pickering
claim does not confer indefinite immunity on employees or insulate them from subsequent investigation and discipline for unrelated misconduct. Granting relief on plaintiffs’ retaliation- claims would handcuff the Department by forcing inaction even where there is police behavior that warrants close review. Speech is one thing; misconduct something else. There are countless unobjectionable reasons why a police department might want to investigate an officer’s performance, including absence from work,
tardiness, insubordination, illegal activity, and basic failure to carry out one’s duties in a competent and impartial fashion. The garden-variety investigations into Liver-man’s and Richards’s conduct were no different, and we therefore reject their claims of retaliation.
IV.
The City argues that Liverman and Richards have failed to establish municipal liability. The district court agreed. We remand on this question to give the district court a chance to assess under the appropriate standard municipal liability for establishing the policy under which plaintiffs were disciplined.
Under Section 1983, a local government may be held liable for injuries suffered due to the “execution of a government’s policy or custom, whether made by its lawmakers or by those whose edicts or acts may fairly be said to represent official policy.”
Monell v. Dep’t of Soc. Servs.,
436 U.S. 658, 694, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978). Municipal liability “attaches only where the decisionmaker possesses final authority to establish municipal policy with respect to the action ordered.”
Pembaur v. City of Cincinnati,
475 U.S. 469, 481, 106 S.Ct. 1292, 89 L.Ed.2d 452 (1986). Whether an entity possesses the requisite authority is a question of state law.
Id.
at 483, 106 S.Ct. 1292.
Here, the district court concluded that the City was not liable with respect to any of plaintiffs’ claims. The court’s holding rested on a city ordinance providing that the Chief of Police “serve[s] at the pleasure of the city manager” and is “under the direction and control of the city manager.”
Liverman v. City of Petersburg,
106 F.Supp.3d 744, 769 (E.D. Va. 2015). Because Dixon does not have the final say over Department matters, the court indicated, plaintiffs failed to show that Chief Dixon “possesses the final authority required to establish municipal liability.”
Id.
This analysis misapprehends the, nature of the requisite authority. We deal here not merely with an individual employment decision,
see Crowley v. Prince George’s Cty.,
890 F.2d 683, 687 (4th Cir. 1989), but a broad policy setting forth all the ground rules for employee speech. An entity has “final” authority to set this sort of policy when no further action is needed for the policy to take effect. The Supreme Court has expressly noted that “[a]uthority' to make municipal policy ... may be delegated by an official who possesses such authority” to another official.
Pembaur,
475 U.S. at 483, 106 S.Ct. 1292. Here the fact that Dixon serves “under the direction and control of the city manager” does not necessarily establish that he lacked final authority to promulgate the policy whose validity has been successfully challenged herein. We must therefore remand to the district court to .undertake a more particularized inquiry into whether Chief Dixon possessed final authority to set policies on the parameters of speech on the part of those law enforcement officers under his command. If so, the City may also be held liable for the injuries that were caused by the applications of that policy.
V.
Running a police department is hard work. Its mission requires capable top-down leadership and a cohesion and esprit on the part of the officers under the chiefs command. And yet the difficulty of the task and the need for appropriate disciplinary measures to perform it still does not allow police departments to wall themselves off from public scrutiny and debate. That is what happened here. The sensitivity of all the well-known issues that surround every police department make such
lack of transparency an unhealthy state of affairs. The advent of social media does not provide cover for the airing of purely personal grievances, but neither can it provide a pretext for shutting off meaningful discussion of larger public issues in this new public sphere.
To recapitulate: We hold that the Department’s social networking policy was unconstitutional and that the disciplinary measures taken against plaintiffs pursuant to .that policy were likewise impermissible. The patent overbreadth of the policy negates Chief Dixon’s qualified immunity defense. We find no merit, however, in plaintiffs’ retaliation claims, which involved investigations for alleged police misconduct independent of any issues of free speech. As to municipal liability, we remand for further proceedings in accordance with the foregoing directions. Remedial issues are also best left for remand, although in light of all that has transpired, reinstatement is not an equitable option. The calculation of attorneys’ fees must of course await the conclusion of proceedings on remand.
The judgment of the district court is accordingly affirmed in part, reversed in part, and remanded for further proceedings consistent with this decision,
AFFIRMED IN PART, REVERSED IN PART, AND REMANDED.