In a post yesterday, Steve concluded that the DNI’s new Directive 119 — which, broadly speaking, prohibits employees of the Intelligence Community from unauthorized “contacts” with the media about intelligence “sources, methods, activities and judgments” — does not violate the First Amendment. I don’t think the matter is so clear-cut, however.
Steve suggests that courts may view the constitutionality of Directive 119 as controlled by the Supreme Court’s decision in Garcetti v. Ceballos, in which a 5-4 majority held that the First Amendment does not offer any protection at all to government employees’ speech made “pursuant to their official duties.” Steve contends that “the majority opinion [in Garcetti] was fairly clear that its holding reached any and all speech that a public employee could not have engaged in solely as a private citizen–i.e., any and all speech that he could only undertake because of his professional duties.” In other words, Steve reads Garcetti to have held that an employee has no First Amendment right at all to speak about something she learned only by virtue of her public employment.
In fairness to Steve, some lower courts have adopted this understanding of Garcetti. I think it’s a considerable overreading, however. Garcetti involved speech that not only was about the public workplace, but was itself undertaken as part of the employee’s duties. The Court did not have any occasion to address whether there is any constitutional protection for speech that an employee makes outside her duties about information she learned by virtue of her job. As it happens, this very question was at issue in a case argued before the Court yesterday morning, Lane v. Franks . . . and from all that appeared at oral argument, not a single Justice was buying the defendant’s argument that employee speech is unprotected merely because the speaker learned of the information in question by virtue of his government employment (or pursuant to his official duties).
I think the better view (one the Court will confirm in Lane) is that the private, unauthorized speech of a government employee to the media is entitled to some First Amendment solicitude, even if that speech is about matters the employee has learned by virtue of her employment. Indeed, most importantly for purposes of Directive 119, that’s the view of the United States.
So what is the relevant First Amendment standard? It’s a form of Pickering balancing, but with a somewhat greater burden on the government. A broad, categorical pre-approval requirement, of the kind Directive 119 prescribes, “chills potential speech before it happens,” in contrast to a discrete “adverse [employment] action taken in response to actual speech.” And as the Supreme Court explained in the National Treasury Employees Union case (from which those quotes are taken), the Court is especially wary of approving such wholesale, ex ante regulation of government employee speech, which “deters an enormous quantity of speech before it is uttered,” by “a massive number of potential speakers,” based “only on speculation that the speech might threaten the Government’s interests.” Because of these concerns, in such cases “the Government’s burden is greater . . . than with respect to an isolated disciplinary action. The 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 outweighed by that expression’s ‘necessary impact on the actual operation’ of the Government.”
So how would the Pickering/NTEU balance be struck as to the DNI’s Directive 119? On the one hand, the regulation is not as severe as the one at issue in NTEU, since it does not by its terms prohibit any substantive category of employee speech to the media; instead, it “merely” requires employees to obtain approval before speaking to the press about “intelligence-related information.” See Justice Sotomayor’s comment from the oral argument yesterday: “I certainly understand the government being able to tell people ‘don’t go out and talk to the newspapers about this; don’t go out and talk about X publicly, because we should be able to control those kinds of disclosures.'” (In this respect, Directive 119 might be viewed as analogous to traditional “Touhy” regulations (see United States ex rel. Touhy v. Ragen (1951)), pursuant to which many agencies establish centralized review and approval processes for handling subpoenas that seek information from government employees.)
On the other hand, Directive 119 might yet be vulnerable to substantial constitutional challenge, because of what appears to be considerable overbreadth in its drafting.
The Supreme Court has already held, in the Snepp case, that CIA agents can be obliged, as a condition of employment, not to publish any information relating to the Agency without first submitting the information for clearance, in order to protect the government’s “compelling interest in protecting both the secrecy of information important to our national security and the appearance of confidentiality so essential to the effective operation of our foreign intelligence service.”
The preclearance requirement in Snepp, however, did not itself impose any substantive limitations on the employee’s right to publish CIA-related information. Indeed, the government did “not deny–as a general principle–Snepp’s right to publish unclassified information.” The preclearance process merely gave the CIA “an opportunity to determine whether the material he proposed to publish would compromise classified information or sources,” and separate laws properly prohibited the employee’s publication of such classified information.
Subsequent to Snepp and NTEU, two leading court of appeals cases considered the constitutionality of other forms of agency preclearance requirements and prepublication restrictions.
In Weaver v. USIA (1996), a divided panel of the U.S. Court of Appeals for the D.C. Circuit upheld a regulation that required all employees of the State Department, USIA and U.S. AID to submit speaking, writing, and teaching materials on matters of “official concern” to their employers for review prior to publication. Essential to the court’s analysis, however, was the government’s reading of the regulation not to authorize any form of punishment for publication of material disapproved by the agency, so long as it was submitted for review beforehand. (The preclearance process permitted the government “to take corrective action” before the unauthorized disclosure of classified information occurred or before publication of statements that would insult or embarrass foreign governments or foreign leaders.) The court was careful to explain that
if, contrary to the government’s proposed interpretation, the regulation were read to authorize punishment for publication of material disapproved under the criteria of . . . inaccuracy, inconsistency with current foreign policy, or significant potential to affect U.S. foreign relations in an adverse manner—then the regulation would raise serious constitutional issues. For example, it is doubtful that the agency could, consistent with [Pickering], penalize publications devoid of non-public information, by employees with non-sensitive responsibilities (e.g., a driver, a payroll accountant), writing in a context where their statements could not possibly be viewed as representing the agency or the United States, simply because the publication took a view “inconsistent with current foreign policy.”
Two years later, in Harman v. City of New York (1998), the U.S. Court of Appeals for the Second Circuit held that the First Amendment prohibited policies forbidding employees in New York City agencies from speaking with the media regarding any policies or activities of the agency without first obtaining permission from the agency’s media relations department. The court disagreed with the D.C. Circuit’s rationale for upholding the policy in Weaver, but also stressed that the policies in New York would have been invalid even under Weaver, since they “allow[ed] the agencies to determine in advance what kind of speech will harm agency operations instead of punishing disruptive remarks after their effect has been felt. For this reason, the regulations run afoul of the general presumption against prior restraints on speech.”
How would Directive 119 fare under these precedents? It appears to be more similar to the invalid policies in Harman than to policy the court upheld in Weaver. It would be one thing if Directive 119 merely required preclearance for all media contacts by IC employees, so that the agencies could identify for the employees any classified and otherwise privileged information that the employees could not disclose under other laws and regulations. But the central provision of the Directive provides categorically that “[c]ontact by IC employees with the media on covered matters must be authorized by their IC element”–lest the employee be subject to possible loss of a security clearance or employment termination. And, most significantly, the Directive does not explain when, if ever, the IC element must “authorize[]” employees to contact the media. It does not, in particular, provide that such authorization must be forthcoming except in cases where the employee might reveal classified or otherwise privileged information. From all that appears on the face of the Directive, administrative officials might have unbridled discretion to withhold authorization based upon whatever criteria they choose. They could, for example, refuse to “authorize” employee speech to the media even about nonclassified, unprivileged information that is already the subject of public debate, and do so even in a case where it is clear that the employee is speaking in her personal capacity, and not as a representative of the government. (Nor does Directive 119 establish a time limit by which the authorization decision must be made. And, as Steve Aftergood has emphasized, Directive 119 is also oddly underinclusive, in a sense that it is limited to contacts with the “media,” thereby discriminating against speech to the press, in particular: “The new Directive creates an anomalous situation in which . . . an intelligence community employee is at liberty to discuss unclassified ‘intelligence-related information’ with his or her next-door neighbor. But if the neighbor happened to be a member of the media, then the contact would be prohibited altogether without prior authorization.”)
I’m not aware of any other preclearance policies that are so restrictive (or that discriminate against speech to the media). (My knowledge of such practices is limited, however. I would welcome information from Just Security readers about similar policies and precedents.) And therefore, at least absent further clarification and tailoring, Directive 119 would appear to be vulnerable to constitutional challenge under NTEU, Weaver and Harman. On the other hand, it’s been more than 15 years since the last of those decisions, so it’s difficult to predict with confidence how the courts would resolve the question today.