The following is an excerpt from a memo I submitted to the House Permanent Select Committee on Intelligence for its pending consideration of restrictions on the press's publication of unauthorized disclosures of classified information:
A central question before the Committee is this: Should the United States criminally punish the press for publishing classified information? This inquiry poses a prospect unprecedented in American history. For more than 215 years, the United Stateshas managed to flourish in the absence of any federal legislation directly prohibiting the press from publishing government secrets. The absence of such legislation is no accident. It clearly fulfills the promise of the First Amendment: “Congress shall make no law . . . abridging the freedom . . . of the press.”
The First Amendment is not an absolute. The press may be held accountable for publishing libel, obscenity, false advertising, and the like. As the Supreme Court observed more than sixty years ago, “such utterances are no essential part of any exposition of ideas, and are of such slight value as a step to truth that any benefit that may be derived from them is clearly outweighed by the social interest in order and morality.”
But government secrets are something else entirely. There is nothing inherent about government secrets that would make their publication of only “slight value as a step to truth.” To the contrary, the publication of government secrets may be extraordinarily valuable to the proper functioning of a self-governing society. Indeed, the very notion that the United Stateswould punish the press for publishing government secrets seems incompatible with the most fundamental tenets of public accountability.
But, of course, there are secrets and there are secrets, and in exploring this matter it may be helpful to distinguish three different types of secrets. First, there are what we might call “illegitimate” government secrets. In this category of secrets, government officials are attempting to shield from public scrutiny their own misjudgments, incompetence, misconduct, venality, cupidity, corruption, or criminality. In a self-governing society, it is vital that such secrets must be exposed. What makes this difficult is that government officials attempting to maintain such secrets may invoke the claim of national security as a cover. We know from historical experience that this happens all-too-often.
Second, there are “legitimate but newsworthy” government secrets. The publication of such a secret may harm the national security and have substantial “value as a step to truth.” For example, the publication of secret information that Army rifles routinely misfire might be both harmful and beneficial to the national interest. Or the publication of secret information that the security of our nuclear power plants is inadequate might both endanger and further the national interest. In such situations, it is often difficult to know which effect predominates.
Third, there are “legitimate and non-newsworthy” government secrets. The public disclosure of such secrets may harm the national security and have only “slight value as a step to truth.” An example would be a publication disclosing that the United States has broken the enemy’s code, in circumstances in which this disclosure furthers no legitimate public interest. Of course, whether any particular publication furthers a legitimate public interest is commonly a matter of dispute, so it may be easier to state this category in the abstract than to apply it in practice.
In principle, the government should never be able to punish the publication of “illegitimate” secrets and should be able to punish the publication of “legitimate and non-newsworthy” secrets. The middle category, which is no doubt the largest, is the most difficult to assess because there are both real costs and real benefits to disclosure. A central challenge to a free society is to distinguish wisely among these three types of secrets. Particularly in the context of criminal prosecutions of the press, the problems of complexity and vagueness can be daunting.
To provide reasonable guidance to the press, avoid chilling the publication of information that is important to the public interest, and limit the dangers of unchecked prosecutorial discretion, we need clear, simple, straightforward rules. Such rules, by definition, will be imperfect. They will inevitably protect either too much or too little expression, and they will inevitably protect either too much or too little secrecy. This is a dilemma.
To resolve this dilemma we should look to the lessons of history. As noted earlier, for more than two centuries the United States has opted not to prohibit the press to publish government secrets. Indeed, in the entire history of the United States the federal government has never criminally prosecuted the press for publishing government secrets. Perhaps surprisingly, this has been an extraordinarily successful solution.
At one point in our history Congress seriously entertained the idea of enacting legislation that would have prohibited such publications. It is instructive to recall how Congress addressed the question. Only three weeks after it voted a formal declaration of war under Article I, section 8 of the Constitution, Congress began debate on what would become the Espionage Act of 1917. Although the Act was directed primarily at espionage, the original bill included what we can call the “press” provision. This provision would have made it unlawful in time of war for the press to publish any information that the president declared to be “of such character that it is or might be useful to the enemy.” The proposed provision added that “nothing in this section shall be construed to limit or restrict any discussion, comment, or criticism of the acts or policies of the Government.” Not surprisingly, the press provision provoked heated debate.
When the provision was first presented to the House of Representatives on April 30, 1917, Representative Edwin Webb of North Carolina defended it on the ground that, “in time of war, while men are giving up their sons and while people are giving up their money,” the press should be willing to give up its right to publish what the president “thinks would be hurtful to the United States and helpful to the enemy.” Representative Andrew J. Volstead of Minnesota asked pointedly how the nation would feel if American troops were “sent to the bottom of the sea as a result of information” published by the press because Congress had failed to enact the provision.
Opposition to the provision was fierce. Representative Simeon Fess of Ohio warned that “in time of war we are apt to do things” we should not do. Republican Senator Henry Cabot Lodge of Massachusetts expressed concern that the government officials who would administer this provision would use their authority to stifle legitimate criticism of the government. Representative Medill McCormick of Illinois added that he was appalled to think that if an epidemic were to break out in the Army the proposed provision might empower the president to prohibit the press from “drawing public attention to the condition of the troops.”
In response, proponents of the provision invoked the clause guaranteeing that “nothing in this section shall be construed to limit or restrict any discussion, comment, or criticism of the acts or policies of the Government.” Opponents replied that it was impossible effectively to criticize the policies of the government without discussing the information on which the criticism was based.
When it began to appear that the press provision would go down to defeat, President Wilson made a personal appeal to Congress, stating that the provision was “absolutely necessary to the public safety.” Members of Congress were unpersuaded. On May 31, 1917, the House of Representatives defeated the provision by a vote of 184 to 144, with 36 Democrats joining the Republican opposition. This ended consideration of the press provision for the duration of the war.
With the benefit of hindsight, it is clear that this was a remarkable and pivotal victory for American freedom. As it turned out, there was not a single instance during World War I, and not a single instance thereafter, in which the press’s publication of a “legitimate but newsworthy” government secret seriously harmed the national interest. The lesson of this experience is that the best course for the United States is to refrain from threatening to criminally punish the press for publishing “legitimate but newsworthy” government secrets. Although one can imagine hypothetical circumstances in which such a publication might seriously harm the national security, 215 years of experience has demonstrated that such legislation is unnecessary, and would do more harm than good.
As the Members of Congress understood in 1917, for the United States government to wield the power to prosecute the press for such disclosures would give government officials a dangerous lever with which to intimidate and threaten the press. To grant government officials such power would seriously jeopardize the ability and willingness of the press to expose to public scrutiny what should be exposed and would undermine the press’s vital role in our constitutional system. In 1917, Congress made a wise and courageous judgment. Nothing that has happened in the intervening nine decades warrants a different judgment today.
But that still leaves the third category of government secrets – those that are “legitimate and non-newsworthy.” The publication of these secrets could harm the national interest without contributing meaningfully to informed public debate. In principle, then, the government should be able to prohibit the publication of such secrets. The problem, though, is that it is not easy even to “know such secrets when we see them.” The very concept of “non-newsworthy” is elusive. This is a serious difficulty, but it is not necessarily insurmountable. It should be possible reasonably to limit the uncertainty by clearly and narrowly defining what is prohibited.
It might be useful to work backwards from the paradigm example of the government secret that should not be published. Suppose a newspaper publishes the fact that the United States has broken the al Qaeda code, and as a consequence the terrorists change their cipher. Suppose also that there is no legitimate public interest in the publication of this information. That is, the publication of this information does not reveal any plausible illegality, incompetence, venality, or misjudgment by government officials. In such circumstances, it hardly seems unreasonable to punish the newspaper for its action.
This example suggests two factors that may help define the scope of a constitutionally permissible criminal prohibition. First, the newspaper knew or was reckless in not knowing that the publication would create a clear and imminent danger of a grave harm to the national security. Second, the newspaper knew or was reckless in not knowing that the publication of this information served no legitimate public interest.
With these two elements in place, it is possible to craft a narrowly drawn law that addresses the most serious dangers to the national security, while at the same time protecting the freedom of the press and the compelling national interest in free and robust discussion of matters of public concern.
Would it be good public policy to enact such a law? On balance, I think not. Once again, I return to the lessons of history. Even if such a law is constitutional, it is neither necessary nor wise. In more than two centuries of experience, the problem addressed by this “law” has never actually arisen. This would be a law in search of a problem. This is never a sound basis for legislation, and certainly not when dealing with a freedom as precious as the freedom of the press. As a matter of wise public policy, Congress had it right in 1917. Even a law drawn as carefully as the one I have defined would cause more mischief than it is worth. Some things are simply best left alone.
I do not mean to suggest, by the way, that the government has no legitimate interest in keeping military secrets. Certainly, it does. But the way to protect this interest is not by prosecuting the press. It is, rather, by refining the government classification system to focus on matters that seriously threaten the national security and then preserving confidentiality by punishing (in a constitutionally permissible manner) government employees who unlawfully leak such information.
As the Yale constitutional scholar Alexander Bickel once observed, this is surely a “disorderly situation,” but it is the best we can do. If we give the government too much power to punish the press, we risk too great a sacrifice of public deliberation; if we give the government too little power to control secrecy “at the source,” we risk too great a sacrifice of secrecy. The American solution has been to reconcile the irreconcilable values of secrecy and freedom by guaranteeing an expansive right of the press to publish and a strong power of the government to prohibit leaks. The American solution is imperfect and unruly, but it has served our nation well for more than two hundred years.
 Chaplinsky v. New Hampshire, 315 U.S. 568, 572 (1942).
 On the debate over the press provision, see Geoffrey R. Stone, Perilous Times: Free Speech in Wartime from the Sedition Act of 1798 to the War on Terrorism 146-149 (W. W. Norton 2004). It is sometimes suggested that another provision of the 1917 Espionage did prohibit some press publications, but this interpretation of that provision is erroneous, particularly when the act as a whole is read in light of the debate over the press provision. See Geoffrey R. Stone, Scared of Scoops, New York Times (May 8, 2006).
 Alexander Bickel, The Morality of Consent 79-82 (Yale University Press 1975).