The safety of our country depends, in no small part, on our government’s ability to keep national security secrets. Yet at a moment when threats are proliferating, and the need for secret intelligence about our adversaries is at a premium, our system for protecting national security secrets is in shreds. It began to fall apart during the Bush administration, when a series of sensitive counterterrorism secrets were leaked to the New York Times and published on its front page. The disarray became more acute when Bradley Manning indiscriminately leaked thousands of classified diplomatic cables and military reports to Julian Assange, who published them on his electronic bulletin board, WikiLeaks. The last thread of the fabric was pulled out by Edward Snowden when he stole and then distributed a huge collection of secret documents, possibly numbering in the millions, from the National Security Agency and other US intelligence bodies.
At the same time that our government demonstrates a staggering inability to protect its secrets, there is near-universal acknowledgment of rampant mis- and over-classification within our national security apparatus. A system of irresistible and seemingly intractable incentives works to create pressure for over-classification in every government agency with foreign-policy responsibility. General Michael Hayden, former director of both the CIA and the NSA, has referred to the “routine” by which the intelligence community “elevates information to a higher level” of classification than is warranted. As of last year, there were some 58,000,000 pages of documents 25 years or older in the National Archives awaiting declassification, the overwhelming share of them made innocuous by the passage of time. Countless contemporary documents, also innocuous, are marked with the secrecy stamp. General Hayden’s testimony, that of numerous other officials, and a wealth of documentary evidence suggest that a great many of the secrets that our government does manage to keep need not be secret at all.
With necessary secrets leaking out and unnecessary secrets kept tight in massive vaults, we are in the midst of a quiet crisis that is certain to have profound repercussions for our national defense in the years and decades ahead. The breakdown is also certain to have no less far-reaching implications for preserving the wide-open flow of information on which our democracy rests. We have arrived at a juncture where it is necessary to think seriously about reform of a far-reaching and highly complex system that has resisted reform for decades. We have also arrived at a moment when serious reform may be possible. But reform in which direction?
One place to begin is with the famous observation by the late New York Times reporter and columnist James Reston that the ship of state is the only ship that leaks from the top. That leaking from the top, from the White House and the upper echelons of the various national-security bureaucracies, is a phenomenon that has been visible in all recent administrations. Even as presidents complain bitterly about leaks and vilify leakers, the fact is that in all recent administrations, White House aides and cabinet officials have consistently engaged in rampant leaking themselves. One only has to look at the wealth of highly classified information contained in any of Washington Post reporter Bob Woodward’s books about the presidency, or in New York Times reporter David Sanger’s recent Confront and Conceal, to get a sense of the scale and scope of the phenomenon.
In a brilliant Harvard Law Review article, “The Leaky Leviathan,” Columbia Law School professor David Pozen argues persuasively that most of this upper-level government leaking is useful, necessary, and virtually ineradicable. “Leakiness”—Pozen’s term for the porosity of our government—lubricates the policy process and serves a great many different purposes for a great many powerful players. It has helped the executive branch, writes Pozen, “to secure the necessary leeway and legitimacy for governance.” At the same time, “curious citizens are derivative beneficiaries. Even though particular leaks may cause real damage, an accommodating approach to enforcement has in the aggregate supported, rather than subverted, the government’s general policymaking capacity.” The corollary that follows from Pozen’s analysis is that even if it were feasible to track down and prosecute White House, cabinet, and sub-cabinet leakers, doing so would go some distance toward crippling the informal mechanisms by which our government works and it would deprive the public of valuable information about the direction of the country.
Nonetheless, this kind of upper-level leaking, whatever benefits it bestows, is also deeply corrosive. The most obvious deleterious consequence is, of course, the dissemination of truly sensitive information of the kind that can compromise central national security goals. It is indisputable that given their instrumental nature most upper-level leaks—daily on display in our major news outlets—do no identifiable harm at all. Yet some small but significant fraction does cause serious damage even if it is difficult in many instances for those outside government to specify its nature.
The high-level leaking also has indirect effects that are harmful in a more global and systemic fashion. Disseminating classified information without authorization is a violation of the law. When a law is violated regularly and brazenly, we have a classic “broken window” problem. Like the low-level vandalism in inner cities that fosters an environment that encourages more serious crime, upper-level leaking breeds disrespect for the classification system and the rule of law throughout the government and society at large. It sends a signal into the bowels of the various national security bureaucracies that leaking is not only acceptable but can even be virtuous. There is perhaps no other area in American life, apart from illegal immigration, where there is such a wide gap between laws and regulations and actual behavior.
Even as high-level leaking encourages leaking down below, it undermines law enforcement efforts to stanch leaks. In almost every one of the eight leak prosecutions that have gone forward over the past few years, the defendants have argued that they are being unfairly targeted by prosecutors for engaging in behavior that is the Washington norm. In the AIPAC case, the lobbyists Steven Rosen and Keith Weissman stood accused of receiving classified information from a Pentagon employee and distributing it to journalists and foreign officials. In contesting the charges, the two defendants attempted to subpoena a parade of high-ranking government officials with the goal of demonstrating that the acts for which they were indicted were nothing more, as one of their briefs asserted, than “the well-established official Washington practice of engaging in ‘back channel’ communication with various non-governmental entities and persons for the purpose of advancing US foreign policy goals.”
Stephen Jin-Woo Kim, the State Department contract employee who earlier this year pleaded guilty to charges under the Espionage Act for leaking classified information to Fox News, pressed the same point. As his attorney argued, “lower-level employees . . . are prosecuted because they are easier targets or often lack the resources or political connections to fight back. High-level employees leak classified information to forward their agenda or to make an administration look good with impunity.” Convicted CIA leaker John Kiriakou argued in his defense that “when the government chooses among similarly situated people and charges only those who have publicly spoken out against the government’s position, the government engages in selective prosecution.”
Claims of selective prosecution like these have gone nowhere in the courtroom. But they resonate in the court of public opinion and are employed to exculpate, at least morally if not legally, those accused of the crime of leaking. Thus, in the view of Jack Shafer of Reuters, Edward Snowden merely did “in the macro what the national security establishment does in the micro every day of the week to manage, manipulate and influence ongoing policy debates.” The investigative reporter Michael Isikoff of NBC News wonders how the Obama administration can “credibly prosecute mid-level bureaucrats and junior military officers for leaking classified information to the press when so many high-level officials have dished far more sensitive secrets to Woodward?” The same complaint against the double standard is regularly put forward by the defenders of leakers like Snowden and Manning. The blogger Glenn Greenwald—an ardent partisan of Snowden and one of the three individuals to whom Snowden passed portions of his trove of stolen documents—notes that the Obama administration, even as it cracks down on leakers, has used leaks to “glorify the President, or manipulate public opinion, or even to help produce a pre-election propaganda film about the Osama bin Laden raid.” They have been leaking “continuously,” says Greenwald, and are “more responsible for such leaks than anyone.”
Whatever one thinks of Greenwald’s extremist politics, and whether or not one accepts his contention about the quantity of leaks flowing from the Obama administration, the larger argument embedded in his remarks and in those of Shafer and Isikoff is impossible to refute. Our laws governing secrecy are enforced in a way that invites charges of double standards and caprice. In a system in which millions of people hold security clearances, the hypocrisy on display has the effect, at least on some quotient of that sizable workforce, of breeding disrespect for the classification system. The nation is now paying a price for that disrespect in the appearance of figures like Manning and Snowden, who justify leaking secrets wholesale. That same disrespect is reflected in the way those two leakers—one now a convicted felon serving time in the Fort Leavenworth military prison, and the other a fugitive from justice in Vladimir Putin’s Russia—are lauded by many as “whistleblower” heroes. In the words of the New York Times editorial board, much of the material released by Manning is of “public value,” while Snowden, for his part, “has done his country a great service.”
What can be done to repair this broken part of our secrecy system?
Two directions are logically possible. The first is to try to make the behavior of government officials better conform to the rules. The second is to try to make the rules better conform to the behavior of government officials. The former is what the Obama administration has been trying to accomplish, albeit in a lopsided and ineffective manner. Even before Snowden appeared, Attorney General Eric Holder’s Justice Department had taken extraordinary steps to apprehend leakers, going so far as to use subpoena powers to rummage through the personal emails and telephone records of journalists. It has prosecuted a record number of leakers. It named a respected journalist seeking classified information from a government official as an unindicted “co-conspirator” in a plot to violate the Espionage Act. Continuing where the Bush administration left off, it is taking the exceptional measure of seeking to compel a journalist—James Risen of the New York Times —to testify about his source in a leak prosecution. The intelligence community, for its part, has set in motion “insider threat” programs to identify future Mannings and Snowdens by means of more intensive screening and polygraphing of employees. It is putting in place dual key safeguards of computer networks, greater compartmentalization of secrets, and tighter restrictions on what kinds of interactions its employees can have with the press and the public.
But none of this deals at all with high-level leaking. Although President Obama has pledged to “root out” leakers, he emphatically denies that his own national-security staff ever leaks. “The notion that my White House would purposely release classified national security information is offensive. It’s wrong,” the president has flatly stated, even as a body of evidence contradicts him. With the president in denial, a crackdown on high-level leakers is almost certainly not going to occur under his watch, just as it did not occur under the watch of any of his predecessors for all the reasons Pozen indicates; high-level leaking is simply too integral to the smooth functioning of the system. We are not going to see a determined effort to put an end to practices that, though often blatantly in violation of the classification rules and sometimes the espionage statutes, frequently serve the White House well.
This brings us to the second approach: making the rules better conform to the behavior of officials.
A starting point for analysis here is Pozen’s observation that upper-level leakers have a very different moral and political standing from the low-level civil servants who have been prosecuted by the DOJ. The nature of their position inevitably affords them a broader view of the consequences of leaking information than a civil servant typically enjoys. Unlike low level bureaucrats, they are in a good position to evaluate the potential damage from any given disclosure and balance it against the value of the objectives being sought. More significantly, they are also in a good position to be held responsible for their actions outside of the criminal justice system. They are appointees of the president, serve at his pleasure, and, as such, are politically accountable in a way that low-ranking civil servants are not. Both their blunders and their accomplishments reflect on the elected leader at the top. When a mistake is made in other realms of national security—say, for example, a failure to provide adequate security to an American embassy that is then overrun by terrorists—that is not treated as a criminal violation, but as a lapse in planning and judgment that deserves both a full airing of facts and a political response.
The treatment or mistreatment of national security secrets by high level officials is little different and should be handled accordingly. Exempting such officials from the criminal strictures and/or the administrative rules that currently bind their words and actions regarding classified information would eliminate the double-standard with some strokes of the pen. With such a change in place, prosecutions of low level civil servants who violate their oaths to protect secrets would be put on a more solid moral footing. For low-level leakers are not in any way politically representative or accountable to the public. Yet their illicit and anonymous disclosures can have profound effects on the course of American defense and foreign policy. Their actions therefore constitute an assault on democratic procedures and principles for which they deserve to be sanctioned.
High-ranking officials have a completely different standing and our security system should be altered to reflect that critical fact. The press and the public at large could be led to grasp the principle under which a leaker down in the depths of the bureaucracy is punished for an unauthorized disclosure while the law leaves a White House or other top official engaging in seemingly similar conduct unmolested. By decriminalizing upper-level leaking, the claim of selective prosecution could be put to rest and a measure of integrity restored to the classification system as a whole.
There are no doubt many ways to accomplish such a shift, and my purpose here is not to set them out in any detail, but simply to show some possible pathways to a destination, with full awareness that the complexities of the system would make any shift an extraordinary challenge of draftsmanship.
One such pathway would be to amend the espionage statutes to exempt a designated group of high-level officials from the restrictions prohibiting unauthorized disclosure of national defense information. As they currently stand, those statutes apply to “whoever” violates the specified terms. This could be altered so that some class of presidential subordinates, whether detailed in the legislation itself or left to the discretion of the president, are exempted from some or all restrictions of the law. This would formalize unwritten rules and ratify practices that are already customary. As Zbigniew Brzezinski, Jimmy Carter’s national security adviser, once put it, “I don’t leak, I brief.” That remark encapsulated the fact that, given his rank in government and proximity to the president, he did not need to concern himself with violating the terms of the espionage statutes, no matter what information he disclosed. The power to brief rather than to leak could be extended by law to a broader circle of presidential appointees. Exactly how broad is a question of obvious and critical importance—and also one of detail and implementation—that need not detain us here.
An alternative path in the same direction would avoid the need for congressional action. Our current system of classification and declassification was created and designed entirely by executive order, not by an act of Congress. Since the establishment of the system by President Roosevelt in 1940, presidents have issued revisions to suit their own purposes. The most recent one is Executive Order 13526, issued by President Obama in 2009. Like its predecessors, it makes the originator of any secret the party responsible for declassifying it. Thus, if, say, the CIA generates a secret, the CIA then has final word on whether it can be declassified (although the president, as the “supervisory” official of the CIA director, has the ultimate authority to decide the matter).
This system of agency ownership makes a good deal of sense because the originating agency has the best picture of why and how the secret was created and what equities are at risk if it is disclosed. Nonetheless, it could be modified to address the issue at hand. The circle of decision makers could be enlarged. Decisions about removing a secrecy stamp are often a judgment call. The agencies should of course retain their power to declassify their own secrets as they see fit. But those who wield the most power and are most accountable should also be treated as responsible parties. It must be noted, of course, that an official declassifying a secret inside a bureaucracy, on the one hand, and an official providing a secret to a reporter, on the other hand, are very different things. The former involves a process that relies on extensive formal procedures, while the latter skips over those procedures. But these critical differences notwithstanding, the power to disclose secrets is one that flows from the president and should be shared with those subordinates who are closest to him.
One objection to such a proposal, whether accomplished by legislation or executive order, is that, in response to a broken window problem, it merely defines deviancy downward. Decriminalizing leaking and/or broadening the circle of officials who possess the power to declassify information would accelerate rather than slow the flow of genuinely sensitive material into the public domain. Like the decriminalization of drugs, it would legalize behavior that, if anything, needs to be discouraged. After all, the objection here would go, not every high-level leak is set in motion to advance a legitimate governmental purpose. Some are motivated by naked political considerations. Others result from even less defensible behavior, like the quest for personal advancement or the desire to be a Washington player and feel important.
Undoubtedly, this kind of undesirable leaking would continue under any regime. Even with our current criminal laws on the books, it is commonplace. Yet it is also relatively rare for this kind of high-level leaking to lead to releases of information that endanger national security or human life. We are talking here, after all, not about Bradley Mannings and Edward Snowdens, but well-regarded men and women with known track records, vetted by the president before being appointed by him, and then vetted again before being confirmed by the Senate. There is no reason to expect that these officials will recklessly compromise extremely sensitive national security secrets about intelligence sources and methods or cryptologic systems or war plans and the like. Even as matters stand today, it is not the law and the prospect of criminal punishment that keeps high-ranking officials from disclosing such sensitive things. It is patriotism and common sense—which in the terminology of political scientists are called “norms.” In any event, the key point is that disclosure will not be free of risk even for high-ranking officials. Criminal liability may be reduced or eliminated but a powerful system of restraints would still remain in place: not in the criminal but in the political domain.
There is a significant precedent for this already built into our system. Some actors in our constitutional order have long enjoyed the discretion to reveal secrets free from fear of criminal prosecution and subject only to political consequences. The Speech or Debate clause of the Constitution offers a laboratory in which we can observe how such a system works.
According to Article I, Section 6 of the Constitution, “for any Speech or Debate in either House, [Senators and Representatives] shall not be questioned in any other Place.” This empowers members of Congress to release classified information if they so wish, as long as it is done in the course of their legislative business. For such an action, they might be sanctioned or even expelled by their own body, or punished by the voters at the next election, but they would not be in legal jeopardy. There are several cases in point.
In March 1987, when Rep. George E. Brown gave a speech on the floor of the House offering highly classified details of the KH-11 and KH-12 reconnaissance satellites, intelligence officials were outraged. Yet Brown was immune from prosecution. He was, however, sanctioned by his colleagues, who removed him from his seat on the House Intelligence Committee. In another instance, also in 1987, Senator Patrick Leahy was removed from his position as vice-chair of the Senate intelligence committee and resigned from the committee itself after it came to light that he had passed a report about the Iran-Contra affair to NBC News. In the same year, Senator David Durenberger, chair of the intelligence committee, was reprimanded by the Senate Ethics Committee for his unauthorized disclosures regarding American spying on Israel.
No criminal prosecution was contemplated in any of these cases. Nor could it have been thanks to the Speech or Debate clause and the Supreme Court‘s interpretation of that clause in the case of Gravel v. United States. In 1971 Senator Maurice “Mike” Gravel of Alaska attempted to place the highly classified trove of documents—the Pentagon Papers—given to him by Daniel Ellsberg into the Congressional Record, but failed because of the absence of the required quorum. He then entered the full text of the thousands of pages of purloined documents into the record of the Senate subcommittee that he chaired. A grand jury attempted to investigate Gravel’s actions and issued a subpoena to gain his testimony. Invoking the Speech or Debate clause, Gravel filed a motion to quash the subpoena and the matter rose to the Supreme Court.
The Court held that the Senator could not be drawn into the legal proceedings: “we have no doubt that [Sen. Gravel] may not be made to answer either in terms of questions or in terms of defending himself from prosecution—for the events that occurred at the subcommittee meeting.” Exceedingly relevant to the issue at hand, the Court also found that the privilege invoked by Gravel extended to one of his subordinates. An aide who had helped him with the Pentagon Papers project had also come under scrutiny by the grand jury. The Court ruled that the aide enjoyed a qualified exemption from criminal sanction: “the privilege available to the aide is confined to those services that would be immune legislative conduct if performed by the Senator himself.”
Although Congress is assumed by many in the executive branch to be a significant source of leaks, congressional leaking seldom draws much attention from either prosecutors or the public. The immunity of its members makes it a hard target—under most circumstances, an impossible target—for prosecutors. The representative nature of its members gives their actions, and even their sometimes foolish loose lips, a measure of legitimacy. Presumably, it is thanks to the Speech or Debate clause that disclosures emanating from Congress rarely if ever elicit charges of double standards and hypocrisy when they go unpunished.
This is not to say that extending the principles underpinning the Speech or Debate clause to some small fraction of the executive branch would leave the covered officials free from all fear of punishment for mishandling classified information. As noted earlier, they would remain vulnerable to other forms of punishment short of criminal prosecution. For one thing, they could be administratively sanctioned or publicly reprimanded by the president or simply dismissed. For another thing, if the president declines to punish them or fails to punish them sufficiently, he himself and his party could be held to account at the ballot box by the voters or—in extreme cases—he could be impeached.
Leakers themselves could, theoretically at least, also be impeached. Section 4 of Article II of the Constitution provides that “__all civil Officers__ of the United States, shall be removed from Office on Impeachment for, and Conviction of, Treason, Bribery, or other High crimes and Misdemeanors” (emphasis added). Two questions immediately arise from this: Do presidential appointees qualify as “civil Officers? And does leaking for illicit purposes constitute a “High crime” or “Misdemeanor”?
The “civil Officers” clause has primarily been directed at federal judges who misuse their position. But presidential appointees also fall under its purview. To be sure, impeachment of executive branch civil officers is even rarer than impeachment of presidents: there has only been one such case in our history. In 1876, William W. Belknap, Secretary of War under President Grant, was impeached by the House of Representatives for bribery (and then acquitted by the Senate). But no matter how rare, the precedent stands: presidential appointees are civil officers and subject to removal from office by the Congress.
As for whether leaking constitutes a “high Crime” or “Misdemeanor,” the question can be reformulated to make it more general: must a transgression (like a security leak) rise to the level of criminality to warrant impeachment? In other words, could a leak that was not a violation of criminal law nevertheless be grounds for impeachment? Precedent suggests the answer is yes. During the impeachment proceedings against Associate Supreme Court Justice William O. Douglas in 1970, the subcommittee of the Judiciary Committee charged with formulating procedures concluded that impeachment could occur for acts “that involved serious dereliction from public duty, but not necessarily in violation of positive statutory law or
forbidden by the common law.” It went on to find that “When such misbehavior occurs in connection with the federal office, actual criminal conduct should not be a requisite to impeachment of a judge or any other federal official. While such conduct need not be criminal, it nonetheless must be sufficiently serious to be offenses against good morals and injurious to the social body.”
Congress has arrived at similar conclusions in other impeachment proceedings and “has repeatedly defined ‘other high Crimes and Misdemeanors’ to be serious violations of the public trust, not necessarily indictable offenses under criminal laws.” Leaking national security secrets for illicit purposes would in many instances qualify as a serious violation of the public trust. As a Congressional Research Service report observed in an overview of impeachment, “the American system seems more designed to protect the public interest than to punish the person impeached.” That is precisely how we should treat illicit disclosures of classified information when they emanate from the highest levels of our government.
Robert Litt, the General Counsel of the Office of the Director of National Intelligence, remarked recently that a primary lesson that he and his colleagues had learned from the Snowden fiasco “is that we would likely have suffered less damage from the leaks had we been more forthcoming about some of our activities, and particularly about the policies and decisions behind those activities.” Being forthcoming, he continued, means reassessing “how we strike the balance between the need to keep secret the sensitive sources, methods and targets of our intelligence activities, and the goal of transparency with the American people about the rules and policies governing those activities.” Contending with the problem of high-level leaking and the double standard it engenders should be part of that reassessment.
Of course it is only one piece—and by no means the biggest piece—of a highly complex puzzle. Along with the much needed (but remarkably ham-fisted) tightening up that the Obama administration has been conducting, we also need better protection for genuine whistleblowers, faster and more comprehensive declassification, and a careful look at the incentive structure that has led to the promiscuous use of the secrecy stamp on innocuous documents. The shift I have proposed here is certain to elicit fierce criticism and/or resistance from some quarters. No doubt much of that criticism and resistance will be based upon valid concerns about how best to protect America’s secrets. Even if the reform I have suggested turns out to be problematic in ways that prove insuperable, my hope, at the very least, is that it will provoke a discussion of how best to repair some broken windows that have been costing the country dearly.