[Editors’ note: This essay is one in a series—the Good Governance Papers—organized by Just Security. In these essays, leading experts explore actionable legislative and administrative proposals to promote non-partisan principles of good government, public integrity, and the rule of law. For more information, you can read the Introduction by the series’ editors.]

The federal government runs an extensive, sprawling, and disorganized system of prepublication review that imposes extraordinary constraints on the capacity of former government employees to write and speak freely. The government has valid reasons to protect national security information. But the current system is a wildly overbroad and thus unconstitutional prior restraint on the legitimate speech of former officials. This constraint is especially pernicious because those whose speech is unconstitutionally constrained are best placed to analyze and express views on national security and related policies about which the public has little knowledge or lacks necessary context to hold officials to account.

Every U.S. intelligence agency and many other federal agencies impose a lifetime prepublication review requirement on at least some subset of their former employees. For example, a partially redacted memorandum for the Central Intelligence Agency specifies that its prepublication requirements apply to “all current and former Agency employees and contractors, and others who are obligated by CIA secrecy agreement.”

In 2014, over five million people held security clearances—or 1.5 percent of the entire U.S. population, and more than the entire population of Norway. Even this far understates the reach of the modern prepublication review process, because all former government employees who have held classified access but who are no longer employed by the federal government (including the authors of this piece) are bound to comply with prepublication review for life. Hence today a significant portion of the U.S. population—almost certainly much more than 1.5%—are formally bound to submit their writings to the federal government for prior review.

How could this possibly comport with the First Amendment, which at its core was designed to prevent prior government restraints on citizen speech? The answer, the government maintains, can be found in a 1980 Supreme Court  per curiam decision in United States v. SneppSnepp was a former CIA officer who had signed “an agreement promising that he would ‘not … publish … any information or material relating to the Agency, its activities or intelligence activities generally, either during or after the term of [his] employment … without specific prior approval by the Agency.’” Despite this agreement, he published a “book about CIA activities on the basis of this background and exposure” after he left the CIA. The district court found that he “deliberately and surreptitiously violated his obligation to submit all material for prepublication review.” The government sued to enforce its agreement, and the district court enjoined Snepp from violating his agreement and imposed a constructive trust on the book’s proceeds.

When the issue made it to the Supreme Court, the brief per curiam opinion focused almost exclusively on the constructive trust. It addressed the First Amendment issues in a cursory footnote that rested on Snepp’s distinctive role within the CIA.  The sum total of the Court’s analysis was this: “The Government has a 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 agreement that Snepp signed is a reasonable means for protecting this vital interest.”

At the time, the prepublication process was quite limited and the number of affected authors and works was tiny. In 1978, for example, only 62 draft works were submitted for review. As we detail in our amicus brief in the pending case Edgar v. Ratcliffe (from which this article draws), in the aftermath of Snepp, the prepublication review process grew from the limited program the Court approved to the multi-agency expansive system of prepublication review we have today.

The Problematic Prepublication Review Process

Several features of today’s prepublication review system work to deny free speech to millions of Americans. The duty to submit manuscripts for prepublication review is broad and vague. It also subjects former government employees’ writings to review for reasons that have nothing to do with protecting “the secrecy of information important to our national security” – the basis on which the more limited system in effect in 1980 was upheld in Snepp.

Our own experiences, as detailed further in the aforementioned amicus brief, illustrate some of the problems. Several years ago, out of an abundance of caution, we submitted for prepublication review a draft Washington Post op-ed, entitled The government’s prepublication review process is broken. It clearly included no classified information. Nonetheless, the prepublication reviewers at the Department of Defense cleared it on the condition that we include the following disclaimer statement: “The views expressed in this article are those of the author and do not reflect the official policy or position of the Department of Defense or the U.S. Government.” By clearing the piece with a condition that had nothing to do with classified information, the reviewers leveraged the threat of criminal penalties to impact the message of the op-ed and asserted jurisdiction over a manuscript with no classified information in a way that discourages future writings that also contain no classified information. Others have had similar experiences. Mark Fallon, for example, wrote an op ed evocatively entitled, The Government Had to Approve This Op-Ed.

Former National Security Advisor John Bolton’s much more prominent recent experience with the process, however, reveals two central problems that others have repeatedly encountered.

First, because the process is plagued by an absence of binding rules, it permits the government to suppress speech in politically self-serving ways. The expert career official in charge of Bolton’s prepublication review, Ellen Knight, determined through the deliberative prepublication process that the book contained no classified information. At that point, Knight maintains through her lawyer, the prepublication process was “commandeered by political appointees for a seemingly political purpose.” A new National Security Council (NSC) staff political appointee determined that the book was replete with classified information even though he made this judgment before he was even trained for the task. Political officials in the White House also urged Knight to delay clearing a chapter on Ukraine during impeachment. This is an extreme, high-stakes example, but many others have experienced political opportunism in the process due to a lack of clear standards and process, a very vague test for what counts as “classified information,” and extreme secrecy. Though those subject to review are sometimes reluctant to complain for fear of antagonizing reviewers, the complaint in a lawsuit filed by the ACLU and the Knight First Amendment Institute at Columbia University demonstrates that the plaintiffs in that case believed their work had been subject to heightened scrutiny when it was critical of or unfavorable to the government.

Second, the process is subject to unnecessary and seemingly opportunistic delays. In Bolton’s case the government appeared to drag out the review process based on the book’s possible impact on impeachment and then the 2020 election. This reflects the fact that each intelligence agency has guidelines about how long review of different kinds of manuscripts will take, but none is binding, and they frequently are not followed. Where speakers have deadlines in their publishing contracts, delays can prevent the publication or impose significant costs on the speaker and publisher. Yet the vast majority of authors abide by the delays since the government threatens criminal prosecution or to recoup royalties.

Bolton was one of the few authors to proceed to publication in the face of these threats. Most authors simply abide by the extended prior restraint. As a revealing example, former CIA analyst Nada Bakos submitted a book manuscript for review in October 2015 and did not receive a response until August 2017, nearly two years later. Multiple intelligence agencies sought to require redactions of Bakos’s manuscript and all but one of the agencies refused to meet with Bakos to discuss the redactions. These agencies agreed to meet only after Bakos sued—an expensive step that most cannot afford. Brad Moss, a lawyer who has represented officials going through the process, explained that he has tried to battle the problem in the past by bringing an “Undue Delay Claim” as a “mechanism to get them to finish the review,” explaining that he’s “had it take years in the past.” Faced with “long delays,” a former employee writing under the pen name “Rebecca H.” wrote, some authors may decide “to ditch their projects altogether.”

The Costs of the Current System

The restrictions not only hurt the former government employees, they also lead to an impoverished public discourse on matters on which they are uniquely well informed. In a system where every former government employee with access to classified information must obtain government approval to speak about “military matters” or “national security issues,” public discourse on these matters is bound to be poorer. To the extent that political considerations slow or block publication of works critical of a current president, senior advisors, or administration policy, the system can operate to slant public discussion of national security matters in a partisan or pro-government direction by current and former clearance holders. These developments, in turn, make it more difficult for the American public to understand the decisions the government is making on its behalf and thus to exercise effective democratic oversight over those decisions.

The current cumbersome, time-consuming, and seemingly arbitrary prepublication review process, by chilling free speech, also drives talent away from government service. As Stephen Aftergood has put it, that talent has “declined to accept a security clearance and the employment possibilities it provided precisely because it entailed what they considered an intolerable limitation of their personal freedom to speak, write, and publish.” This is particularly true for those whose professions revolve around writing, teaching, or speaking about national security.

Last, the current prepublication process places former employees in the untenable position of deciding whether to abide by unreasonable prior restraints or replace the government’s submission standard with a more reasonable one of their own making. To cite one high-profile example, former CIA director and Secretary of Defense Leon Panetta grew so frustrated with the agency’s prepublication review of his memoir that he “allowed his publisher to begin editing and making copies of the book before he had received final approval from the CIA.” He is far from alone. Former government officials often decide to go ahead and publish without seeking prepublication review because the process is so cumbersome, slow, and often unfair.

The Solution

A number of principles should guide reforms to the system of prepublication review to protect the speech rights of former government officials—and their ability to inform the public—while also protecting the government’s valid concerns about protecting information relevant to U.S. national security. These elements could be codified in an Executive Order, or a Presidential Memorandum could direct department and agency heads to promulgate specific rules in accordance with these principles. Legislation constraining and regulating prepublication review would be more difficult to enact but it would be more effective, because its limits on prepublication review would not be subject to unilateral change by a future president. A few basic principles should guide reform:

Provide narrow, objective, and definite standards for what must be submitted. The grounds for prepublication review must be clear and uniform across agencies. The current system fails to provide narrow, objective, and definite standards with respect to what materials former employees must submit for review and what content agency officials may censor. This leaves former government employees unsure of what must be submitted, which has the effect of chilling speech. In particular, only written material directly related to classified matters on which a former government employee worked should be subject to review. Moreover, only writings that might reasonably contain or be derived from classified information should be subject to review.

Limit who is subject to prepublication review. Jameel Jaffer, of the Knight Institute, argues for limiting the scope of those subject to prepublication review to those who have had Top Secret/Special Compartmented Information clearance within the last 10 years. We agree that this reform would allow the pre-publication review process to focus on those who most likely possess information that could do real harm to U.S. national security, thus justifying the prior restraint on speech entailed in the prepublication review system.

Limit prepublication review to review for classified information. Inspection for classified information should be the only basis for review. The government has no compelling interest in viewpoint discrimination or in reviewing unclassified or publicly available information. When reviewers scrutinize submissions for other reasons (like criticism of government policy, accuracy, or privilege), the government restricts speech in ways unrelated to protecting national security and thus not justified by its legitimate interests.

Meaningful Deadlines. Prepublication review should have binding deadlines for completion—ideally no longer than 30 days. In Freedman v. Maryland, the Supreme Court held that prior restraints must yield a determination “within a specified brief period.” Right now, the prepublication review process has no binding deadlines for review. That leaves former employees at the mercy of a process that can drag on long past what is reasonable. If a former employee receives no response within thirty days, he or she should be permitted to proceed with publication.

Reason-giving and appeal. When an agency censors materials under the prepublication review process, it must be required to give clear reasons and permit swift appeals. At present, when a publication is held or rejected, it is not always clear why. The process of appeal is slow and cumbersome and can create such a hurdle to publication that many simply give up.


To be sure, the government has a legitimate interest in protecting “the secrecy of information important to our national security.” But the government has a wide array of mechanisms for protecting classified information, including administrative and criminal penalties for former government employees who actually disclose classified information. A properly designed system of prepublication review can be part of the government’s toolkit, but any such regime of prior restraint requires careful scrutiny to avoid the abuses and vulnerabilities that currently characterize the system. The current system for prepublication review has grown so sprawling, cumbersome, and overbroad that it simply goes beyond what the First Amendment allows. In contrast, a more narrowly tailored system of the kind described here would achieve the government’s legitimate aim of protecting classified information from unauthorized disclosure without impermissibly imposing on the free speech rights on former government employees.