What once was nearly unthinkable – vesting the President with the power to designate wide swaths of information that cannot lawfully be discussed – has become status quo. The Obama-Biden administration played an important role in getting us to this point, and the Trump-Pence administration sped things along further still. The Biden-Harris administration should muster the political courage and imagination to seek a reversal of course, a task that they can achieve by encouraging a rethinking of the Justice Department’s prosecution priorities regarding leaks of national security information to the media and by introducing or supporting legislation to curtail the sweeping powers currently available under the Espionage Act of 1917 to prosecute such media leaks. Although the early signs from the Biden-Harris Administration are not terribly encouraging, the respect that the administration has voiced for speech and press freedoms provides some basis for hope.
One key, early step on the path to our current state of affairs came in January of 1917. As war engulfed Europe, Woodrow Wilson introduced draft legislation to Congress. That June, following the United States’ entry into the war, Wilson signed the bill into law as the 1917 Espionage Act.1 The Act criminalized classic espionage, meaning the passing of closely held national security information to foreign governments or to enemies of the United States for nefarious purposes. Yet Wilson had wanted the Act to do more. Fearing that journalists and their sources could also harm national security by publishing sensitive information, Wilson’s draft included a provision authorizing the President, “in a time of war,” to designate categories of information that it would be a crime to convey to unauthorized recipients, even if the speaker’s intentions were benign.2 The provision sparked a media outcry, and Congress did not include it in the final legislation. The debate in Congress at the time reflected a consensus among the provision’s proponents and opponents alike that, in its absence, the Espionage Act only punished the conveyance of national security information with the intent of aiding an enemy of the United States.3
Despite the 65th Congress’ apparent conviction that the Espionage Act was not an official secrets act, the Act’s plain language left the door open to just such an interpretation. In provisions now codified at 18 U.S.C. §§ 793(d) and (e), the Act makes it a crime for anyone who possesses information “relating to the national defense” to “willfully” communicate it to an unauthorized person or to fail to return it on demand, if the possessor “has reason to believe” the information “could be used to the injury of the United States or to the advantage of any foreign nation.”4 When the communicated items are tangible––such as documents or photographs, rather than orally conveyed information––the Act does not require a “reason to believe” in the materials’ potential injuriousness.
The breadth of these provisions did not fully escape the attention of the 81st Congress when it considered amendments to the Act in 1950. For example, Senator Kilgore expressed concern that “the bill could “theoretically . . . make practically every newspaper in the United States . . . into criminals without their doing any wrongful act.”5 The 81st Congress’ attention, however, was focused mostly on higher-profile parts of the Act that were directly responsive to the Cold War, including provisions “that made it unlawful to conspire to establish a totalitarian dictatorship in the United States, the broad registration requirements, and the powers of the Subversive Activities Control Board.”6 In comparison, the prospect of an official secrets act seemed too far-fetched at the time to command much attention. Indeed, Attorney General Tom C. Clark submitted a letter for the congressional record in response to Senator Kilgore’s concerns. In it, he wrote that the Act’s “language [and] history,” and “the integrity of the three branches of the Government . . . would indicate that nobody other than a spy, saboteur, or other person who would weaken the internal security of the Nation need have any fear of prosecution.”7
Although it took quite a while, the Act’s potential to operate as an official secrets act –turning virtually every unauthorized conveyance of classified information into a prosecutable crime — increasingly has been realized. One important step toward that end occurred in 1951, when President Truman signed an Executive Order establishing the nation’s first peacetime classification system.8 Just eleven years prior to that, President Roosevelt had signed the first Executive Order on classification, shortly after World War II broke out in Europe.9 Courts have since used the classification system to clarify some of the Act’s terminology and, having done so, to help stave off vagueness and overbreadth challenges to the Act.10 Perhaps as importantly, the existence of an increasingly large classification system – indeed, the system has expanded substantially since 1951 – has normalized the notion that ubiquitous secret-keeping and far-reaching measures to enforce it are inevitable, even healthy parts of government. It may surprise some readers to learn, for example, that Truman’s 1951 classification order was “decried by the press, members of Congress, and others who considered it “‘unwarranted peacetime censorship.’”11 Today, of course, the classification regime “towers over the one that struck Americans as frighteningly radical in the 1950s.”12
It was perhaps inevitable that as the classification system grew to cover ever-larger swaths of newsworthy information, and as courts interpreted the Act to cover the unauthorized conveyance or retention of virtually any classified information, prosecutions for media leaks would eventually follow. The first such prosecution was brought in 1957 against Army Colonel Jack Nickerson. The second was brought in 1971 against Daniel Ellsberg and Anthony Russo for leaking the Pentagon Papers; and the third was brought against naval intelligence analyst Samuel Morison in 1984. The fourth, brought against Lawrence Franklin in 2005, involved Franklin’s disclosure of information to two lobbyists whom themselves were accused of leaking the information “to members of the media, foreign policy analysts, and officials of a foreign government.”13
The Espionage Act’s transformation into a tool for prosecuting leaks of information to the media accelerated dramatically during the Obama-Biden administration. That administration prosecuted eight government employees under the Espionage Act for leaking information to the media or retaining information in connection with suspected media leaks.14 Not to be outdone by its predecessor, the Trump-Pence administration initiated five such prosecutions in its single term, opened substantially more leak investigations than had the Obama-Biden administration, and created a new unit within the FBI to investigate media leaks. The Trump-Pence Administration also obtained an indictment against Julian Assange for violating the Espionage Act by publishing classified information via Wikileaks. The Assange indictment goes where the Obama-Biden administration had deemed it too dangerous to tread; the latter reportedly had concluded that indicting Assange would open the door to prosecuting mainstream U.S. news organizations and journalists.15
It remains to be seen whether the Biden-Harris administration will proceed along the trajectory set by its predecessors – including the administration in which Biden served as Vice President – by wielding the Espionage Act with increasing frequency against media leakers and keeping the door open to prosecuting publishers. Two early signs point in that direction. First, the administration chose to appeal a British court’s ruling blocking Julian Assange’s extradition to the United States. A number of press freedom and transparency organizations had urged the administration to drop the appeal and dismiss the underling indictment, citing concerns “about the way that a precedent created by prosecuting Assange could be leveraged – perhaps by a future administration – against publishers and journalists of all stripes.” Second, the administration has already obtained its first conviction in an Espionage Act case, albeit one that began in the Trump-Pence Administration. Specifically, former intelligence analyst Daniel Hale pleaded guilty on March 31st to a single count under the Espionage Act for retaining and transmitting classified information to a journalist. Prosecutors did not drop four remaining counts against Hale, but rather placed them in abeyance, “giving the government at least the theoretical opportunity to bring the other counts to trial.” The Justice Department’s press release announcing the plea suggested a continued categorical intolerance for media leaks. It stated that “Daniel Hale knowingly took classified documents and disclosed them without authorization, thereby violating his solemn obligations to our country. We are firmly committed to seeking equal justice under the law and holding accountable those who betray their oath to safeguard national security information.”
Despite these beginnings, it is plausible that the Biden-Harris administration still could chart its own path, one more mindful of the dangers that media leak (and publisher) prosecutions pose to speech and press freedoms. For one thing, the Assange and Hale cases were initiated in the Trump-Pence Administration, and new administrations often find it easier “to stay the course” in pending cases.16 Furthermore, both Joe Biden and Kamala Harris consistently criticized Donald Trump’s anti-press threats and rhetoric and have extolled the value of robust investigative reporting “to keep government honest and the public informed.” It is true that the Biden-Harris administration’s actions thus far are more mixed than its rhetoric would suggest. Apart from the Assange and Hale cases, for example, the administration has “been denying members of the news media visibility and access to immigration facilities at the U.S. – Mexico border” and has resisted calls to publish attendee details from its virtual meetings, despite pledging to publish its in-person visitor logs. Nonetheless, the administration’s apparent eagerness to be viewed as an ally of a free and active press suggests that they might be amenable to growing pressure from civil liberties and transparency groups to rethink Espionage Act leak and publication prosecutions going forward.
Thus far, I have only gestured in this essay at reasons why the Biden-Harris Administration should chart a new course with respect to media leak prosecutions. Elsewhere, I have argued at length that such prosecutions raise serious First Amendment and policy concerns that call for new legislative and judicial approaches to liability and sentencing alike.17 As David Schulz and I put it in a forthcoming article,
[C]lassified information is, after all, information; to convey it is to speak. Insofar as such communications concern government, foreign affairs, or public policy, they are in a realm that scholars and jurists routinely place at the very core of the First Amendment. Suppressing media leaks also raises a worry at the heart of much free speech theory and doctrine: government actors may single out that speech (i.e., those media leaks) that cast them in a bad light.18
Indeed, courts routinely subject prosecutions or other government punishments based on the content of speech – including punishments for speech that may incite violence or that embarrasses public officials — to highly searching review. As a matter of theory, nothing about the classified information context should change this protective posture. To the contrary, “[t]he notion that the executive branch – or even the political branches acting in tandem – can erase or substantially diminish the robust First Amendment protections that would otherwise apply, simply by declaring swaths of information ‘classified,’ flies in the face of core free speech principles.”19 And the realities of the bloated classification system bear out these theoretical concerns. “Endemic overclassification . . . is a real-life manifestation of the notion that the government will abuse its powers to stifle debate about itself;” it “betrays the folly of classification exceptionalism – that is, of the notion that . . . doctrinal protections should shrink away at the wielding of a classification stamp.”20
Should the Biden-Harris administration prove amenable to such arguments, what ought it to do? First, Attorney General Garland could incorporate case-sensitive assessment factors – such as public interest considerations and harm evaluations – into the Department’s investigative and prosecutorial priorities.21 This is warranted, at minimum, as a replacement for the careful, multi-factor assessments that the Justice Department’s internal guidelines long demanded its prosecutors make before subpoenaing journalists. The necessity for such subpoenas has been drastically reduced in media leak cases, given the relative ease today of tracking down sources through electronic clues and third-party records.22 Second, the President should support long-needed legislative reform in the area of media and publication leaks. There are many possibilities for legislative reform, as I and others have suggested elsewhere. A non-exhaustive list includes: amending the Espionage Act to limit its application solely to cases of classic espionage; incorporating into any new, media-leaks-focused legislation either a balancing test to weigh public interest and harm factors or a subjective intent requirement; and adding a statutory public interest defense to the Espionage Act or to new, media-leaks-focused legislation.23
Before considering specific reforms, of course, the Biden-Harris Administration must be willing to acknowledge that the trajectories for media leak and publication prosecutions pose intolerable dangers to speech and press freedoms and to the public’s ability to hold its governors to account. To acknowledge and address this reality is not to surrender to chaos and anarchy, leaving government officials helpless to protect legitimate national security secrets. It is, however, a necessary first step to righting the terrible imbalance wrought by the status quo, one in which little more than executive say-so is needed to punish speech that poses little security threat and that conveys information that the public is better off for knowing.
a. Robins Kaplan Professor of Law, University of Minnesota Law School; Newton N. Minow Visiting Professor of Law, Northwestern University Pritzker School of Law.
1. The definitive history of the Act was written by Harold Edgar and Benno C. Schmidt, Jr., and published in 1973: Harold Edgar & Benno C. Schmidt, Jr., The Espionage Statutes and Publication of National Defense Information, 73 Colum. L. Rev. 929 (1973).
2. Id. at 950–65. Similarly, the 65th Congress rejected a provision that would have given content to the Act’s reference to information that unauthorized persons are “not entitled to receive.” The rejected provision would have empowered the President “to designate any matter, thing, or information belonging to the Government, or contained in the records or files of any of the executive departments, or of other Government offices, as information relating to the national defense, to which no person [other than those duly authorized] shall be lawfully entitled.” Id. at 1006–09.
3. Id. at 953–58.
4. 18 U.S.C. § 793(d), (e). 1950 amendments to the Act split the restriction in Section 793 into to two provisions, (d) and (e), dealing separately with individuals having authorized possession of information and those with no authorization; restated the scope of the restriction to include “information relating to the national defense” (which remains undefined); and, added a scienter requirement (“[T]he possessor has reason to believe [the information] could be used to the injury of the United States or the advantage of any foreign nation . . . .”). See Edgar & Schmitt, supra note 1, at 998–1000.
5. Edgar & Schmitt, supra note 1, at 1025.
6. Id. at 1028.
7. Id. at 1026.
8. Arvin S. Quist, Security Classification of Information 50–51 (Vol. 1, 2002).
9. Id. at 9; Sam Lebovic, From Censorship to Classification, The Evolution of the Espionage Act, in Whistleblowing Nation: The History of National Security Disclosures And The Cult Of State Secrecy 54 (Kaeten Mistry & Hannah Gurman eds., 2020).
10. For example, the Act prohibits communicating information “relating to the national defense” to persons “not entitled to receive it.” Courts have held that one’s entitlement to receive information can be determined by their clearance (or lack thereof) under the classification system. Courts have also interpreted “relating to the national defense” to refer, in part, to information’s status as closely held, which can be satisfied by its classification status. See, e.g., Heidi Kitrosser & David Schulz, A House Built on Sand: The Constitutional Infirmity of Espionage Act Prosecutions for Leaking to the Press, 19 First Am. L. Rev. (forthcoming 2021) (manuscript at 65–67), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3825222 [https://perma.cc/U43R-XBX7].
11. Luther A. Huston, Brownell Praises Information Plan, N.Y. Times, Oct. 29, 1953, at 20; see also U.S. Adds Controls on Security Data, N.Y. Times, Sept. 26, 1951, at 17; Quist, supra note 8, at 50–51.
12. Kitrosser & Schulz, supra note 10, at 28–29.
13. Id. at 32–39; United States v. Rosen, 445 F. Supp. 2d 602, 607–11 (2006).
14. See Gabe Rottman, A Typology of Federal News Media “Leak” Cases, 93 Tul. L. Rev. 1147, 1182–85, Table 1 (2019) (counting only the prosecutions brought under Section 793).
15. See, e.g., Michael M. Grynbaum & Marc Tracy, ‘That’s Called News Gathering’: Charges Alarm Advocates of Press Freedom, Wash. Post, May 24, 2019, at A8; Sari Horwitz, Julian Assange Unlikely to Face U.S. Charges Over Publishing Classified Documents, Wash. Post (Nov. 25, 2013), https://www.washingtonpost.com/world/national-security/julian-assange-unlikely-to-face-us-charges-over-publishing-classified-documents/2013/11/25/dd27decc-55f1-11e3-8304-caf30787c0a9_story.html [https://perma.cc/686F-GWGL].
16. Charlie Savage explains, for example, that “litigation calendars can force early decisions about whether to proceed or shift direction in some cases. It is often easier to stay the course, based on an argument that the issue can be revisited later when there is more time. But once the new administration has started down that path, it owns the policy as a matter of political and bureaucratic reality and so can effectively get locked in.” Charlie Savage, Civil-Liberties Groups Ask U.S. to Drop Assange Case, N.Y. Times, Feb. 9, 2021, p. A18.
17. See, e.g., Kitrosser & Schulz, supra note 9; Heidi Kitrosser, Leak Prosecutions and the First Amendment: New Developments and a Closer Look at the Feasibility of Protecting Leakers, 56 Wm. & Mary L. Rev. 1221 (2015); Heidi Kitrosser, Free Speech Aboard the Leaky Ship of State: Calibrating First Amendment Protections for Leakers of Classified Information, 6 J. Nat’l. Sec. L. & Pol’y 409 (2013); Heidi Kitrosser, Classified Information Leaks and Free Speech, 2008 U. Ill. L. Rev. 881 (2008).
18. Kitrosser & Schulz, supra note 10, at 7.
19. Id. at 11.
20. Id. at 12.
21. See Principles of Federal Prosecution, U.S. Dep’t Just., at 9-27.230, https://www.justice.gov/jm/jm-9-27000-principles-federal-prosecution (last visited Apr. 26, 2021) [https://perma.cc/EAS8-AYKV] (“[I]n the interest of allocating [the Justice Department’s] limited resources . . . from time to time the Attorney General may establish national investigative and prosecutorial priorities.”).
22. See Kitrosser & Schulz, supra note 10, at 41–49.
23. Id. at 78–89 (citing some suggestions for reform made by the authors and others).
The full text of this is available to download as a PDF.