I'm not a big fan of the law review article as a form. But every now and then, one comes along that is genuinely important, that sheds new and interesting light on an important issue, that cuts through the unilluminating manner in which the subject had been previously discussed, and that thus changes the way we think about a subject that had grown stale. So it is, at least in my opinion, with a fascinating new article (forthcoming in the Harvard Law Review) by David Pozen of Columbia Law School about leaks. Pozen has just put a draft up on SSRN of his article, "The Leaky Leviathan: Why the Government Condemns and Condones Unlawful Disclosures of Information."
Full disclosure: I read this article in draft and gave David comments on it.
Full disclosure number II: I also strongly recommended to the Harvard Law Review, when asked, that it publish it.
I also strongly recommend that Lawfare readers actually read it---in particular, that they read Section II, which deals with the systemic advantages to the executive branch of a leaky system. It's a genuinely important contribution. I have been on both sides of this. I have spent many years in the press and have, in that capacity, solicited my share of leaks. I also do work on national security law, however, and thus see the destructive effects of leaks too. This is the single best, most thorough, and most thoughtful treatment of the subject I have read.
Here's the introduction:
Ours is a polity saturated with, vexed by, and dependent upon leaks. The Bay of Pigs, the Pentagon Papers, warrantless wiretapping by the National Security Agency at home, targeted killings by the Central Intelligence Agency abroad: The contours of these and countless other government activities have emerged over the years through anonymous disclosures of confidential information to the press. Across the ideological spectrum, many Americans believe both that leaking “is a problem of major proportions” and that “our particular form of government wouldn’t work without it.” Episodically, leaks generate political frenzy. The country is in such a period at this writing. Mass releases of classified defense documents and diplomatic cables through WikiLeaks, followed by a series of news stories about some of the government’s most closely held national security programs, have unleashed a torrent of legislative and media responses, of recriminations and justifications. This “latest outbreak of leak panic” will soon fade; a new iteration will arrive in due course.
Our comprehension of leaking has not kept pace with our fascination. Even accounting for the secrecy that obscures its workings, the ratio of heat to light in commentary on the subject is extreme. Some valuable progress has been made. Journalists and ex-officials have chronicled the role of leaks in their work. Students of government and the press have limned leaks’ different forms and motivations. Legal theorists have considered their First Amendment implications. Yet for a variety of reasons, the literature reflects only a rudimentary understanding of leaks’ consequences, inside and outside government. More surprising, because the questions are more tractable, scholars have devoted scant attention to the constitutive elements of the leak, as a legal and bureaucratic concept, or to the policies the executive branch has developed to enforce relevant prohibitions. We know something about the phenomenology and constitutionality of leaks but next to nothing about how the government deals with them.
This Article begins to reveal that world. Drawing on a range of theoretical perspectives and original sources—interviews with journalists and executive branch officials, plus records requested through the Freedom of Information Act (FOIA)—it offers the first sustained account of the regulatory regime applicable to leaking. Superficially straightforward, this regime turns out to be an intricate ecosystem. At the most general level, the Article demonstrates that the story behind the U.S. government’s longstanding failure to enforce the laws against leaking is far more complicated, and far more interesting, than has been appreciated. More specifically, the Article argues that the executive branch has never prioritized criminal, civil, or administrative enforcement against leakers; that a nuanced set of informal social controls has come to supplement, and nearly supplant, the formal disciplinary scheme; that much of what we call leaking occurs in a gray area between full authorization and no authorization, so that it is neither “leaks” nor “plants” but what I will term pleaks that dominate this discursive space; that the executive’s toleration of these disclosures is a rational, power-enhancing strategy and not simply a product of prosecutorial limitations, a feature not a bug of the system; and that to appreciate these dynamics is to illuminate important facets of presidential power, bureaucratic governance, and the national security state in America today.
These claims require extensive elaboration. As a way into them, consider two features that mark the United States’ legal approach to unsanctioned disclosures of protected information. First, and most significantly, even though the Espionage Act of 1917 and other statutes broadly criminalize the gathering, receipt, and dissemination of national defense-related information and even though every modern president has decried the practice, an enormous amount of leaking to the press appears to occur and to go unpunished. The federal government has brought fewer than a dozen media leak prosecutions in the 96 years since the Espionage Act was enacted, six of them under the current administration. Available evidence suggests that civil and administrative sanctions are only marginally more common. Let us call this the punitive/permissive divide: The statutes on the books concerning leaks, and the political rhetoric associated with them, are so harsh, and yet the government’s actual treatment of the activity seems to have been so mild. There is a dramatic disconnect between the way our laws and our leaders purport to condemn leaking and the way they condone it—a rampant, pervasive culture of it—in practice.
Second, the courts have indicated that while the government has expansive legal authority to prosecute employees who leak, it has minimal authority to stop reporters who receive leaks from broadcasting what they learn, either through ex post penalties or prior restraints. In all cases, it seems, the government would have to prove a reporter’s actions threatened grave, immediate harm to national security interests. (It is hard to be sure because the government has never indicted one.) Let us call this the source/distributor divide: The First Amendment has been understood to provide so little protection for the leaker and yet so much protection for the journalist who knowingly publishes the fruits of the leaker’s illicit conduct and thereby enables the very harm—revelation of sensitive information to the public and to foreign adversaries—that the leak laws were designed to combat. In other areas of criminal law, downstream users of illegally obtained material are not similarly insulated from liability.
As a descriptive or diagnostic matter, the literature has pointed to several factors to explain the existence of these features. The leak laws are so rarely enforced, it is said, because the Justice Department finds it so difficult, at the investigatory stage, to identify culprits and so difficult, at the adjudicatory stage, to bring successful cases without divulging additional sensitive information. Courts and prosecutors have privileged journalists over leakers, it is said, because of the former’s special First Amendment status and the latter’s consent to nondisclosure as a condition of employment. Throughout these discussions, the comparison is often drawn to the United Kingdom’s notorious Official Secrets Act. Whatever else might be true of our legal and political regime for regulating leaks, virtually everyone agrees the United States would never abide such a sweeping criminal prohibition.
As a normative or justificatory matter, few have celebrated this “disorderly situation.” Many believe it to be regrettable if not outrageous. National security hawks and opposition members of Congress routinely call for legislative strengthening of the leak laws and more vigorous executive enforcement, including against members of the media. Civil libertarians have assailed the last two administrations’ “war” on leaking and sought federal shield legislation for journalists and enhanced doctrinal protections for their sources. The most thoughtful and influential defenses of the status quo—by Alexander Bickel, Jack Goldsmith, and Geoffrey Stone—are self-consciously ambivalent. They do not attempt to idealize or even rationalize the punitive/permissive and source/distributor divides, so much as to highlight the factors confounding any effort to strike an optimal balance in this area between national security needs and other democratic and constitutional goods. These analyses are also spare. They rely on discrete case studies and high-level constitutional theory to identify relevant values, precedents, and tradeoffs; they do not make any systematic inquiry into patterns and practices of leaking or enforcement or into the functional and strategic dimensions of the leak-law regime.
This Article aims to show why our “disorderly situation” cannot be understood without such inquiry and to explicate a more satisfying logic, a richer model, that better accounts for the seemingly incoherent law of leaks. The core claim is that the status quo, although ritualistically condemned by those in power, has served a wide variety of governmental ends at the same time as it has efficiently kept most disclosures within tolerable bounds. The leak laws are so rarely enforced not only because it is difficult to punish violators, but also because key institutional players share overlapping interests in vilifying leakers while maintaining a permissive culture of classified information disclosures.
The executive branch is where most of the action takes place. With respect to the punitive/permissive divide, commentators have widely assumed the executive would prefer to bring more cases, and then looked to constraints on realizing that preference. By concentrating on barriers to prosecution, these accounts overlook the ways in which powerful actors benefit from leak-law violations, as well as the ways in which leaking may be punished in the absence of a criminal proceeding, or indeed any formal sanction. The executive’s “leakiness” is often taken to be a sign of institutional failure. It may be better understood as an adaptive response to key external liabilities—such as the mistrust generated by presidential secret-keeping and media manipulation—and internal pathologies—such as overclassification and fragmentation across a sprawling bureaucracy—of the modern administrative state. Once these affirmative interests in leaking are identified, the source/distributor divide also begins to look a little less puzzling. To criminalize leaking at the publication stage, as well as the transmission stage, would not only raise difficult First Amendment concerns but also risk compromising the government’s instrumental use of the press.
Part I of the Article reviews the legal landscape and the available evidence on how leaks are used and punished. Part II first details the shortcomings of constraint-based rationales for the government’s permissiveness, and then advances an alternative explanatory theory grounded in the executive branch interests served by leakiness. These interests include preserving ambiguity as to the origins of unattributed disclosures and therefore the communicative flexibility of top officials, signaling trustworthiness, facilitating richer internal information flows, pacifying constituencies for transparency in Congress, the media, and civil society, and mitigating the classification system’s political and deliberative costs. Part III further fleshes out this account by uncovering and assessing the informal enforcement model that the executive appears, behind closed doors, to have applied to leaking. Part IV returns to the source/distributor divide and to the recent uptick in prosecutions and draws out additional evaluative and descriptive implications—including the surprising insight that the U.S. and U.K. legal regimes on leaks have, at least in their observable aspects, substantially converged.
Before proceeding further, several notes are in order. Throughout the pages that follow, the focus is on the U.S. federal executive branch and its information control practices relating to national security, broadly defined to include many matters of foreign policy. Some of the Article’s arguments may well carry over to other types of leaks, to other types of institutions, to subnational levels of government, maybe also to other mature democracies. At points, I will briefly discuss possible extrapolations. But national security leaks raise a number of distinct concerns and have always driven the legal conversation. Their treatment by the world’s dominant military power will, I trust, be of sufficient interest to many.
The heart of the Article is the positive analysis in Parts II and III and the window it provides into the regulation of leaking. While I hope in Part IV to draw some fruitful linkages to broader
questions in democratic, constitutional, and security theory, and while I hope more generally to facilitate normative projects of varied stripes, the Article is principally concerned with demonstrating how leaking works. The overriding aim is to provide an explanation (not a justification) for this regulatory regime in terms of the intersecting desires, beliefs, and constraints of presidents, political appointees, civil servants, legislators, journalists, and the institutions they populate. Although I cannot directly establish the intentions of many of these actors or rule out evolutionary factors—for instance, a natural tendency for the leakier components of government to gain in relative political power and thereby to propagate their disclosure norms—the Article tries to the extent possible to provide microfoundations for its theory, to specify mechanisms that reduce leakiness to the individual level.
Some pieces of the analysis were informed by roughly two-dozen interviews I conducted with current and former officials who have worked on top secret issues across the executive branch, as well as by my own experiences in government. Most of these officials served in a legal capacity. Almost all of the interviews were “on background,” which is to say the interviewees asked not to be identified and I agreed. The unsettling brand of methodological mimesis that results—this is an Article about anonymous government sources that itself relies on anonymous government sources—is not lost on me. I would have preferred to conduct interviews on the record and yet, like so many journalists and several scholars before me, I soon learned that few were willing to discuss anything interesting under those ground rules. Minimizing validity concerns, however, all points drawn from the interviews are consistent with the publicly available evidence, including the returns from my own FOIA requests, which I cite throughout. The interviews proved more confirmatory than revelatory. They deepened but did not transform my understanding of this evidence.
A final caveat: While I try in this Article to shine light on an opaque corner of the law, I certainly do not mean to suggest that I have fully uncovered, much less “solved,” the inner workings of the leak regime. The phenomenon of leaking is sufficiently heterogeneous and complex as to resist monolithic characterization. Perhaps the most generalizable contribution the Article aims to make is to point the way to new conceptual, theoretical, and empirical approaches to studying the subject.