Chapter 1

The Problem

How to Regulate State Secrecy?

In the introduction we noted that there is broad agreement that state secrecy is justified so long as it is used to protect national security and not to conceal wrongdoing. However, there is also broad agreement that in practice state secrecy has become the “all-purpose means” by which the presidency has “sought to dissemble its purposes, bury its mistakes, manipulate its citizens and maximize its power.”1 Assuming this complaint is at least partially correct, it raises the question of why it has proven so difficult to regulate the employment of state secrecy. Why, in other words, has practice been able to diverge so far from principle?

The leading explanation places the blame on fairly recent developments. The presidency, Arthur Schlesinger argues, is responsible for transforming the “legitimate system of restriction” established by the Framers “into an extravagant and indefensible system of denial.” Because those who receive “prestige and protection” from secrecy control the classification system, he writes, concealment has “overridden its legitimate objectives.”2 There is, however, something puzzling about this claim. How can the Constitution, otherwise revered for establishing checks and balances, have allowed the president a free hand when it comes to state secrecy? The culprits, according to Schlesinger, have been war and conflict, which have prompted deference from lawmakers and judges, and “unquestioning” loyalty from citizens, thereby ensuring that “the only control over the secrecy system” has come to be “exercised by the executive branch itself.”3

But this explanation only deepens the mystery. Did the Framers not foresee that war and conflict might open the door to extensive government secrecy? A number of scholars believe not. The Framers, they maintain, intended to construct a political system where secrecy would be a rare and short-lived exception rather than an enduring presence. This claim was first pressed by an earlier generation of scholars, including Schlesinger, Henry Steele Commager, Raoul Berger, and Daniel Hoffman, who took the view that the Framers supported the “principle of disclosure.”4 And it has since been reiterated by, among others, Geoffrey Stone, Robert Pallitto and William Weaver, David Pozen, and Heidi Kitrosser, who have declared that the Constitution is premised on a “philosophy of openness.”5

The claim that the Framers expected state secrecy to be the exception rather than the norm in American government has rightly been challenged in recent years.6 Abraham Sofaer, Mark Rozell, Stephen Knott, and Gabriel Schoenfeld, among others, have marshaled evidence showing that early administrations, which were led by prominent Framers, readily utilized covert action and withheld information from citizens and lawmakers. Even so, because these scholars have not examined in detail the examples and ideas that influenced the Framers, they have somewhat understated the case that can be made against the prevailing view. I cannot address this lacuna at length here, but even a brief overview of the broader intellectual and historical context will show that the Framers cannot be understood to have supported “the principle of disclosure” or “a philosophy of openness.”

There are, then, at least three reasons to be skeptical about the view that the Framers were leery of state secrecy. In the first place, we should bear in mind that the Framers undoubtedly knew that prior republics had readily employed secrecy. Though neither Athens nor Rome developed an infrastructure or corps dedicated to the collection and protection of secret intelligence, the Framers would have learned from the classical sources they knew well that the Athenians and Romans had utilized an array of sources and methods to obtain intelligence on an ad hoc basis, and that their leaders had made frequent, if often private, efforts to obtain secret intelligence when this was necessary and feasible.7

The Framers’ consideration of the history of modern republics would only have sharpened this understanding. The experience of the English Commonwealth would have been especially telling in this regard. In the century preceding the English Civil War, the business of intelligence in England had been transformed by the emergence of a new breed of officials who came to occupy the office of principal secretary.8 The men who filled this office—Thomas Cromwell, William Cecil, and Francis Walsingham—were renowned for their “peculiar knowledge” of “dangers both abroad and at home,” which they obtained by building an effective, far-reaching, and secretive “intelligence system” consisting of clerks, translators, archivists, code-breakers, forgers, and messengers, through whom the principal secretaries controlled a “shifting legion of freelancers” and dozens of foreign “correspondents.”9 Did a sense of republican propriety lead the officers of the English Commonwealth to spurn the methods and practices utilized by the principal secretaries? Far from it. These officials readily followed their Tudor and Stuart forerunners in cultivating a far-flung network of spies and correspondents.10 Indeed, they went further, secretly coordinating the publication of propaganda and institutionalizing the covert interception of mail.11 The same holds true after the Glorious Revolution. Over the course of the eighteenth century, the executive in England (increasingly the prime minister rather than the monarch) immersed itself ever deeper in the business of intelligence, as can be discerned from the steady growth of public funding on this front. Under James I and Charles I, the funds made available to the secretaries of state under this head had been seven hundred pounds each; by 1786, the budget for intelligence was twenty-five thousand pounds.12 These resources were used to cement the precedents established in the sixteenth and seventeenth centuries—as the Framers experienced firsthand during the Revolutionary War.13

A second reason to be skeptical about the claim that the Framers were wary of state secrecy is that many of the republican theorists they were familiar with endorsed it as valuable.14 Recall that the first major step toward the institutionalization of secret intelligence had actually been taken in the republican city-states of Renaissance Italy. Having discerned the value of information in navigating the treacherous world they inhabited, these city-states were among the first to post ambassadors abroad.15 As the primary mission of an ambassador was to communicate back to the chancery whatever intelligence he could obtain, the emergent enterprise of modern diplomacy naturally prompted calls for discretion, most notably in Francesco Guicciardini’s Dialogue on the Government of Florence; there it is asserted that informants are more likely to cooperate with a “closed regime” than an “open regime,” because no one “wants to reveal a hidden secret” in a place where “it will no sooner be said than publicized.”16

Even more striking are the claims put forward in the aftermath of the English Civil War. Confronted with the charge that representative assemblies (such as Parliament) were prone to indiscretion, the defenders of the English Commonwealth replied that these bodies could—and indeed should—maintain secrecy by delegating sensitive business to smaller bodies appointed by the people’s representatives. Thus we find James Harrington recommending in The Commonwealth of Oceana that those affairs that must be conducted in secrecy “for the good of the commonwealth” should be entrusted to a “Council of War.”17 Similarly, Marchamont Nedham argues in The Excellencie of a Free State that the “prudence, time, and experience” required to manage “secrets of state” are best obtained by entrusting such matters to an executive council.18 John Milton, too, recommends in The Ready and Easy Way to Establish a Free Commonwealth the establishment of a “Council of State” for the purposes of “carrying on some particular affairs with more secrecy and expedition.”19

The utility of state secrecy is appreciated in eighteenth-century republican literature as well. For instance, in A Short Introduction to Moral Philosophy Francis Hutcheson describes “secret and speedy execution” as one of the four points “to be aimed at” in constituting an ideal polity.20 Meanwhile, David Hume argues in his “Idea of a Perfect Commonwealth” that the executive ought to be made “absolute” in the sphere of foreign affairs because “otherwise there could be no secrecy or refined policy.”21 Richard Price, that influential friend of the American revolutionaries, writes in his Additional Observations on the Nature and Value of Civil Liberty that what distinguished a free government from an ideal one was that the latter joined to the liberty enjoyed by the former the “dispatch, secrecy, and vigour” required to successfully execute the “will of the community.”22 And citing “the disclosure of public counsels and designs” as one of the “evils” of democracy, William Paley warns in The Principles of Moral and Political Philosophy that a well-ordered regime ought to guard against the sort of “officious and inquisitive interference with the executive functions” that would lead to the disclosure of “what it is expedient to conceal.”23

Now, to be sure, there were, from the sixteenth century onward, a number of republican theorists who were troubled by state secrecy, especially after the Renaissance had shown it to be one of the foremost elements of arcana imperii (mysteries of state)—that body of stratagems by which rulers could maintain their rule over their unwitting subjects.24 This was a point that one of the Commonwealth’s fiercest critics, Clement Walker, stressed when he charged that Charles I had been replaced not by a republican regime, but by “forty tyrants” (the number of members in the Commonwealth’s Council of State), who were “dispatching all affairs privately and in the dark, whereas Justice delights in the light, and ought to be as public as the common air.”25 But the defenders of the English Commonwealth had a simple answer to this objection: secrecy, they argued, was necessary, and had been seen as such in Greece, Rome, and Florence too.26 Note also that later republican critiques of state secrecy, such as Algernon Sidney’s famous riposte against arcana imperii in his Discourses Concerning Government, written just prior to the Glorious Revolution, were directed at the exercise of secrecy by absolute monarchs rather than republican magistrates. Indeed it would be strange if Sidney thought a republic ought to eschew secrecy, since in his view “the best Government” is that “which best provides for War.”27

Certainly, not everyone was convinced by the argument from necessity. Over the long eighteenth century—a time of intricate diplomatic maneuvering and incessant war—the notion that state secrecy posed a standing threat to the public interest found subscribers. The most striking critique in this vein is Jeremy Bentham’s “A Plan for an Universal and Perpetual Peace,” wherein Bentham depicts the elimination of state secrecy as a prerequisite to the achievement of peace. “Between the interests of nations,” he writes, “there is nowhere any real conflict,” and where such conflicts appear, he argues, they can be traced to deliberate or inadvertent misunderstandings fostered by state secrecy.28 Therefore, “if we think peace better than war,” he writes, then secrecy “can not be too soon abolished.”29 Bentham was not the only one to think this way; this was a view shared by a diverse set of theorists, including the abbé de Saint-Pierre, Pierre-André Gargaz, the marquis de Condorcet, Immanuel Kant, and Jean-Jacques Rousseau. Though these writers did not discuss state secrecy at length, their various “peace plans,” which sought to replace diplomacy and intrigue with orderly international confederations, favored openness. Note, however, that prominent Framers explicitly rejected such proposals as impractical.30

The third reason to doubt that the Framers were leery of state secrecy is that there is evidence to show that they viewed it as an essential element of statecraft.31 Certainly, a number of important texts from the Revolutionary period underscored the importance of state secrecy. John Adams’s “Thoughts on Government,” for instance, described secrecy as one of the properties “essential” to the exercise of executive power.32 In “The Genuine Principles of the Ancient Saxon or English Constitution,” a radical Whig writing under the name Demophilus observed that even the best government would be short-lived unless it provided for its self-defense that “dangerous, but necessary engine of state, a standing army, whose operations must be conducted with all possible secrecy and dispatch.”33 And Theophilus Parsons’s remarkable report, “The Essex Result,” warned that “want of secrecy may prevent the successful execution of any measures, however excellently formed and digested.”34

That the leaders of the Revolution appreciated state secrecy is evidenced most clearly, though, by the measures they introduced to preserve the confidentiality of proceedings in the Second Continental Congress, and by their use of covert action under the auspices of the Committee of Secret Correspondence, which was made responsible for an array of intelligence functions, including espionage, covert operations, and postal interception—activities that, in the words of the participants, called for “great circumspection and impenetrable secrecy.”35 Notably, these measures remained in place even after the Revolutionary War had drawn to a close.36 Another important piece of evidence in favor of a wider political consensus on the value of secrecy is provided by the adoption of Article IX of the Articles of Confederation, authorizing Congress to withhold from publication those proceedings “relating to treaties, alliances or military operations, as in their judgement require secrecy.”37 Though the states proposed a number of changes to the Articles of Confederation prior to the document’s eventual ratification, they did not challenge this provision. Nor, for that matter, did they try to amend this provision once their delegates to Congress pleaded an inability to share information that Congress had recorded in its Secret Journal.

That the Framers appreciated the value of state secrecy becomes only clearer as we consider the period leading up to the Constitutional Convention. Far from viewing Congress’s extensive use of secrecy with alarm, the Framers appear to have drawn from the Revolution and its aftermath the lesson that the country needed a political system capable of ensuring rather than reducing secrecy. This is evident in George Washington’s correspondence with James Madison in March 1787. Welcoming the decision to meet in Philadelphia, Washington averred that “a thorough reform of the present system is indispensable” because it lacked, among other things, the “secrecy and dispatch … characteristic of good government.”38 This was not the first time Washington had expressed such a view. In a letter to Henry Knox, written a month earlier, he emphasized that a central failing of the Continental Congress was that it was “defective in that secrecy, which, for the accomplishment of many of the most important national objects, is indispensably necessary.”39

Why did the men who were to draft the Constitution consider secrecy “characteristic of good government”? One must appreciate that these men saw international politics as a ruthless business. They accepted, as John Jay would write in The Federalist No. 4, that “nations in general will make war whenever they have a prospect of getting anything by it.”40 At the same time, they were unwilling to rely on leagues or alliances as a principal means of securing peace. The fragility of the “triple and quadruple alliances” that had recently been formed in Europe had taught them, Alexander Hamilton would write in The Federalist No. 15, “how little dependence is to be placed on treaties which have no other sanction than the obligations of good faith.”41 This belief was one of the principal reasons why the leading men of the day were votaries of a powerful national government. “If a federal constitution could chain the ambition, or set the bounds to the exertions of all other nations,” Madison would argue in The Federalist No. 41, “then indeed might it prudently chain the discretion of its own government, and set bounds to the exertions for its own safety.” But, in the absence of such sureties, said Madison, “with what color of propriety could the force necessary for defence, be limited by those who cannot limit the force of offence?”42 At the same time, the Framers noted that power, though necessary, was not sufficient for the end that they had in mind. The security and stability of the nation, they observed, depended in no small measure on how power was exercised. In particular, they agreed that a government ought to display “energy”—that is, “decision, activity, secrecy, and dispatch.”43

None of what has been said so far should be seen as implying that late eighteenth-century Americans were unquestioning proponents of state secrecy. On the contrary, they were well aware of the accompanying danger. For example, having acknowledged that the preservation of society required an executive capable of “secrecy and expedition” and that “expedition, secrecy, and dispatch” are present only when execution is undertaken by “one or a small number of persons,” Parsons warned in “The Essex Result” that “from a single person, or very small number, we are not to expect that political honesty, and upright regard to the interest of the body of the people, and the civil rights of each individual, which are essential to a good and free constitution.”44 “For these properties,” he wrote, “we are to go to the body of the people,” who “have always a disposition to promote their own happiness.”45

Consider as well what Zabdiel Adams, cousin to both John Adams and Samuel Adams, had to say in his “Election Sermon” before the General Court of Massachusetts in May 1782. Speaking in the wake of controversy over Congress’s recommendation that the states impose a 5 percent duty on imports in order to rescue the finances of the United States, Adams observed that “there is a maxim often mentioned of late, that there should be no mysteries in government.” But this maxim, Adams pointed out, could not always be followed because “necessities of the state sometimes require great secrecy”; “the most important expedition or negotiation might otherwise fail.” Nonetheless, “where secrecy is not essential,” he continued, “there the authority ought to make known the necessity of their measures.” For “if taxes are heavy, and people know not to what uses they are applied,” Adams warned, then “they conclude that they are swallowed up in a manner not beneficial to the public.”46

The concern that secrecy could obscure abuses of power was voiced most clearly, though, during the debates over the Constitution, particularly with respect to Article I, Section 5, Clause 3, which permits Congress to withhold the publication of those parts of their journals of proceedings that “may in their Judgment require Secrecy.”47 Every time such objections were raised, however, the supporters of the Constitution firmly responded that decisions about what information should be concealed ought to be left to officials. In Pennsylvania, for example, James Wilson acknowledged that “some gentlemen” were worried that the Senate and the president might commit the United States to treaties enacted in secret. Confessing that he, too, was “not an advocate for secrecy in transactions relating to the public; not generally even in forming treaties,” Wilson declared that upon reflection he had been compelled to accept that it would “be extremely improper” to publish treaties when “secrecy may be necessary.”48 A month later the same concern was voiced in North Carolina. John Steele admitted there that “he had heard objections” to the provision allowing Congress to exercise secrecy, but he joined with William Davie and James Iredell in emphasizing that “every principle of prudence and good policy pointed out the necessity of not publishing such transactions as related to military arrangements and war.”49 In Massachusetts, William Widgery, a county lawyer who had declared earlier in the proceedings that rulers “ought never to have a power they could abuse,” informed his fellow delegates that the words “excepting such parts as may in their judgment require secrecy” implied that “Congress might withhold the whole journals under this pretence, and thereby the people be kept in ignorance of their doings.” This assertion provoked a curt response from Nathaniel Gorham, who informed Widgery that “many things in great bodies are to be kept secret, and records must be brought to maturity before [being] published.” “Would it be policy,” he asked, “to inform the world of the extent of the powers to be vested in our ambassador, and thus give our enemies [an] opportunity to defeat our negotiations?”50

The most revealing debate of all occurred in Virginia, where Patrick Henry informed his fellow delegates that his “greatest objection” to the Constitution was that it undermined “true responsibility” because it allowed Congress “not to publish what parts they think require secrecy.”51 This opening salvo was easily deflected by Edmund Randolph, who reminded his audience that “without secrecy, no government can carry on its operations on great occasions.”52 Randolph’s reply prompted Henry to refine his position. “Such transactions as relate to military operations or affairs of great consequence, the immediate promulgation of which might defeat the interests of the community,” he declared, “I would not wish to be published, till the end which required their secrecy should have been effected.” But the impact of this more refined argument was blunted when Henry characteristically veered off course and urged Americans to imitate the “manly boldness” of the British, who “divulged to all the world their political disquisitions and operations,” a fanciful claim that attracted John Marshall’s ire.53 Do the British, Marshall asked, “deliberate in the open fields” when they determine “the propriety of declaring war, or on military arrangements?” “No,” he continued, “the British government affords secrecy when necessary, and so ought every government.”54

Henry responded to this challenge by retracing his steps. Though Marshall had indicated that “war was the case wherein secrecy was most necessary,” said Henry, the Constitution also permitted the Senate to withhold information in other contexts as well, most notably the formation of treaties. This meant, he argued, that “the federal veil of secrecy” would prevent citizens from “knowing or being able to punish” those responsible for the passage of unfavorable treaties. Henry challenged this “destruction of responsibility” by revisiting the nuanced position that he had carved out earlier. “I did not wish that transactions relative to treaties should, when unfinished, be exposed,” he argued, “but it should be known, after they were concluded, who had advised them to be made, in order to secure some degree of certainty that the public interest shall be consulted in their formation.”55 But his final opponent, Madison, was not convinced by the wisdom of this demand either. Henry, said Madison, desired that once a treaty had been made, “the public ought to be made acquainted with every circumstance relative to it.” But “the policy of not divulging the most important transactions and negotiations of nations,” Madison argued, “is universally admitted.”56

Constitutional Puzzles

I have suggested that the Framers did not subscribe to a “philosophy of openness” or promote the “principle of disclosure.” On the contrary, they anticipated and defended the use of state secrecy as an essential element of statecraft. This raises an obvious question: did they not worry about the possible misuse of secrecy? The Anti-Federalists, as we have seen, certainly did. Henry charged that the president and Congress could find it in their mutual interests to keep instances of maladministration out of the public’s sight. Hence he stressed the need for what we might today term “compulsory declassification.” Madison, as we have seen, rejected this idea on the grounds that the delicate nature of international affairs militated against compelled or scheduled disclosure. But Madison was not content simply to say that state secrecy was necessary. He also emphasized that the Constitution’s provisions would combat the abuse of secrecy—“there can be no real danger,” Madison said to Henry, “as long as the government is constructed on such principles.”57 What did Madison mean by this?

Recall that in the period leading up to the Constitutional Convention, the Framers had become increasingly vocal about the idea that energy was a vital prerequisite of national security and stability, and that the Continental Congress was incapable of exhibiting the requisite degree of energy. A major factor was Congress’s inability to maintain that “secrecy which is the life of execution and despatch.”58 According to those most familiar with the problem, the very “nature and construction” of Congress—its size and varied membership in particular—militated against the maintenance of internal discipline.59 As a result, no one at the convention appears to have doubted that one of the president’s key functions would be to employ secrecy in a more effective manner. According to Rufus King, when Wilson moved on June 1, 1787, that the “executive consist of a single person,” he explicitly stated that an executive “ought to possess the powers of secrecy, vigour and dispatch.”60 In the debate that followed, few, if any, of the delegates questioned this claim.61 Instead, they focused on what form the executive should take so as to strike an appropriate balance between energy and safety. The most notable objection came from George Mason. Mason observed that “unity in the executive” promised “the secrecy, the dispatch, the vigour and energy” that not only he but also “the ablest and most candid defenders of republican government” acknowledged as “great advantages.”62 But allowing the executive to consist “only of one person,” he argued, raised the specter of usurpation. Far better then, he concluded, to create an executive council, an arrangement that would sacrifice a degree of energy in return for greater security against the misuse of executive power.63

Mason was not alone in taking this view. Randolph and Elbridge Gerry offered much the same diagnosis and prescription.64 But these views did not find sufficient support in the convention, which ultimately rejected both Randolph’s proposal for an executive council and Mason’s subsequent call for an advisory council.65 The delegates, Oliver Ellsworth and David Ramsay later recounted, decided that not only secrecy, but also responsibility, and thereby accountability, were best served by vesting executive power in one person.66 The reasoning behind this view was made clearer in the state conventions and in The Federalist No. 70. In North Carolina, Davie explained that while “the superior energy and secrecy wherewith one person can act, was one of the principles on which the Convention went … a more predominant principle was the more obvious responsibility of one person.” “It was observed” by delegates, he said, “that if there were a plurality of persons, and a crime should be committed, when their conduct came to be examined, it would be impossible to fix the fact on any one of them, but that the public were never at a loss when there was but one man.”67 Similarly, having observed that the presidency was best designed to take the lead on national security matters, Hamilton reassured his audience in The Federalist No. 70 that the elective nature of the office and the singularity of the occupant provided citizens with the “two greatest securities” possible against the misuse of delegated power—namely, “the restraints of public opinion” and “the opportunity of discovering with facility and clearness the misconduct of the persons they trust.”68

But how credible are these claims? How can citizens “fix the facts” or ascertain misconduct with “facility and clearness” when the president has the authority to conceal the relevant information (or to disclose only selected bits)? Indeed, how might citizens even discover wrongdoing in the first place? This, as we have seen, was the concern that led Henry to demand the timely disclosure of military and diplomatic secrets. And Madison, as we have also seen, responded—completely reasonably—that the delicate nature of international affairs made it inadvisable to compel prompt disclosure. But, then again, if citizens can scrutinize the secret conduct of their representatives only after their representatives have judged that the relevant information can be safely shared with them, then what security do they have that their representatives will not simply decline to share incriminating or embarrassing information?

A reply might be sought in what Madison described as the “auxiliary precautions” contained in the Constitution, most notably the separation of powers, which, he argued, would oblige the government “to control itself.”69 But how precisely would this mechanism constrain the misuse of state secrecy? In particular, how could it overcome the objection voiced by a host of Anti-Federalists, including Henry, that the president and a cabal in Congress could jointly conceal wrongdoing? The Federalist No. 26 provided an answer. Hamilton wrote there that those who feared that Congress and the president might build up “vast augmentations of military force” had failed to appreciate that the Constitution required Congress to annually approve military appropriations. “Can it be supposed,” Hamilton asks, that when debates on these appropriations took place, “there would not be found one man, discerning enough to detect so atrocious a conspiracy, or bold or honest enough to apprise his constituents of the danger?” “If such presumptions can fairly be made,” Hamilton argues, “there ought at once to be an end of all delegated authority.”70 In other words, Hamilton’s argument was that the separation of powers provides a simple but effective safeguard against secret machinations—since any member of Congress can sound the alarm, ambition would have every chance to counteract ambition. The Anti-Federalists’ fears of cabals and combinations were unrealistic, he implied, because in practice it would be extremely difficult for conspirators to quietly secure the collective agreement required to cheat the public.

There are, however, some wrinkles in this story. The first is that neither Madison nor Hamilton seems to have thought that officials entrusted with state secrets have a right to disclose such information as they see fit. Consider, for example, Madison’s commentary on the events of December 1782 when David Howell, Rhode Island’s delegate to Congress, was found to have leaked to the Providence Gazette news of a friendly overture from the Swedish court.71 Howell had leaked the news, which had been recorded in the Secret Journal, because he believed it vindicated his stance that the United States would be able to raise new loans in Europe, and that Congress therefore did not have to impose a 5 percent import duty that Rhode Islanders opposed. Claiming to have informed his constituents of “such things as they have a right to know,” Howell subsequently defended his action as an exercise of “the freedom of speech.”72 But his colleagues were thoroughly unimpressed by this claim. Howell’s response, Madison observed in his Notes on Debates, provoked “universal indignation” in Congress, because Howell’s colleagues viewed his actions as having betrayed the Swedish court and presented the public with a distorted picture of the United States’ financial dealings that they could not correct without revealing “many delicate transactions.”73 Not surprisingly, Howell’s defense of his action was firmly repudiated—on Hamilton’s motion—as “highly derogatory to the honor and dignity of the United States in Congress.”74

This episode does not of course rule out the possibility that Madison and Hamilton believed that officials have a moral right to disclose secret information as and when this reveals a conspiracy against the people, and that Howell deserved to be castigated only because he had exposed secret information without due cause. But this quite plausible interpretation is complicated by the fact that neither Madison nor Hamilton elaborated on what might constitute due cause for making an unauthorized disclosure. When was a conspiracy truly a conspiracy (as opposed to the fevered imagination or self-serving claim of a partisan)? Howell, for his part, certainly seems to have thought his disclosure had defeated a conspiracy.75

The other wrinkle in the claim that the separation of powers would counter the misuse of state secrecy is that the Framers left Congress dependent on the president for access to secret information. This becomes evident when we contrast The Federalist No. 53 and No. 75. In the former, Madison discusses what a legislator must do to gain competence. “No man can be a competent legislator,” he writes, “who does not add to an upright intention and a sound judgement, a certain degree of knowledge of the subjects on which he is to legislate.” One such “branch of knowledge which belongs to the acquirements of a federal representative,” Madison says, “is that of foreign affairs”—a subject that will “sometimes demand particular legislative sanction and cooperation.” While “some portion” of this knowledge can be acquired from “a man’s closet” and some of it “can only be derived from the public sources of information,” Madison concludes that “all of it will be acquired to the best effect by a practical attention to the subject during the period of actual service in the legislature.”76 But how could legislators attain such familiarity with foreign affairs when, as Hamilton noted in The Federalist No. 75, the degree of secrecy required to safeguard intelligence was “incompatible with the genius of a body so variable and numerous” as the House?77

The conflict between The Federalist No. 53 and No. 75 was not unprecedented. The proceedings of the Constitutional Convention reveal that the confusion ran deep. On August 11, 1787, the delegates agreed to incorporate language that would allow the House and the Senate to withhold publication of such of their records “as may in their judgment require secrecy.”78 But consider what happened when, on September 7, the delegates turned to discuss Article II, Section 2, Clause 2, which authorized the president to make treaties “by and with the advice and consent of the Senate.” At this point, Wilson moved that the House be included in this clause alongside the Senate on the grounds that since “treaties are to have the operation of laws, they ought to have the sanction of laws.” The “only objection” to including the House, Wilson said, was the “circumstance of secrecy in the business of treaties.”79 This was a remarkable argument because, as we have seen, less than a month earlier, the delegates had authorized the House to exercise secrecy. Why grant the House the authority to exercise secrecy if it could not be trusted to maintain secrets? This question was not raised at the time. Instead, Roger Sherman challenged Wilson’s motion by reemphasizing the importance of secrecy, arguing that “the necessity of secrecy in the case of treaties forbade a reference of them to the whole Legislature.”80 His colleagues appear to have been persuaded, for they rejected Wilson’s motion by a margin of 10–1.

This was not to be the end of the story. On September 14, Gerry joined with Mason to propose that only the Senate be allowed to exercise secrecy. Their proposal was rejected 7–3, Madison reports, after the delegates were informed that “cases might arise where secrecy might be necessary in both Houses,” for instance with respect to “measures preparatory to a declaration of war.”81 Why did the delegates believe that the House would be able to keep preparations for war secret, but that it would not be able to keep diplomatic proceedings secret? Had the delegates changed their minds? This seems unlikely given what advocates of the Constitution had to say in the various state conventions. In Virginia, for instance, Francis Corbin affirmed—in Madison’s presence—that the House had been “excluded from interposing in making treaties, because large popular assemblies are very improper to transact such business, from the impossibility of their acting with sufficient secrecy, despatch, and decision, which can only be found in small bodies.”82 In South Carolina, Charles Cotesworth Pinckney put the point more directly still. “Can secrecy,” he asked, “be expected in sixty-five members?” “The idea,” he said, “is absurd.”83

What about the Senate? A number of the Constitution’s defenders seem to have believed that the Senate’s role in foreign affairs would help combat the abuse of secrecy. In Pennsylvania, for example, Wilson argued that since “neither the President nor the Senate, solely, can complete a treaty; they are checks upon each other, and are so balanced as to produce security to the people.”84 Similarly in South Carolina, Pinckney, having ridiculed the indiscreetness of the House, went on to argue that “the Senate, from the smallness of its numbers … joined with the President, who is the federal head of the United States, form together a body in whom can be best and most safely vested the diplomatic power of the Union.”85 But consider what Jay has to say in The Federalist No. 64. “There are cases,” he writes, “where the most useful intelligence may be obtained, if the persons possessing it can be relieved from apprehensions of discovery.” Such persons, he claims, “would rely on the secrecy of the President, but would not confide in that of the Senate, and still less in that of a large popular Assembly,” which is why the convention did well to require the president to act by the advice and consent of the Senate, but simultaneously vesting in him alone the management of the “business of intelligence in such a manner as prudence may suggest.”86 In other words, according to The Federalist No. 64, even the Senate would not have independent or unmediated access to secret intelligence.

Strikingly, Jay does not discuss here the possibility that the president’s monopoly of secret intelligence could make it harder for the Senate to exercise its coordinate powers in opposition to the president. On the contrary, he focuses solely on the advantage of this division of functions, saying that it shows how “the Constitution provides that our negotiations for treaties shall have every advantage which can be derived from talents, information, integrity, and deliberate investigations, on the one hand, and from secrecy and despatch on the other.”87 But there is something puzzling about this claim. If the president is required to share secret intelligence with the Senate, then this defeats the purpose of vesting control over the “business of intelligence” in his hands, namely, to take advantage of his greater capacity to maintain discretion. But if the president is not required to share secret intelligence, then what is to prevent him from sharing only information advantageous to his own cause? It is tempting to assume that the Framers expected the Senate to extract information by withholding cooperation on funds, appointments, and the approval of treaties. But then again, if the point of creating the office of president is to maintain discretion, then how can it be sensible for the Senate to compel the disclosure of information that the president judges ought to be kept secret? And if the reply is that the Senate should utilize compulsion only when this is truly necessary, then one must ask how Senators can ascertain when to utilize compulsion, since they will not know in advance whether or not the information held by the president is genuinely sensitive.

Insulated Circumstances

I have been arguing that though the Framers did not expect the executive to maintain only a few, short-lived secrets, the checks and balances they devised feature a number of silences about how to counter the abuse of state secrecy. I have identified three such silences. First, how can citizens hold the president and Congress accountable for the use of secrecy, since the president and Congress get to decide when, if ever, to disclose secret information to the public? Second, how can lawmakers check the executive’s use of state secrecy, since the president gets to decide when, if ever, to disclose secret information to Congress? And third, under what circumstances, if any, are officials entitled to disclose secret information that reveals wrongdoing? Since these lacunae are not insignificant, it is only reasonable to assume that succeeding generations ought to have noticed and then sought to fill these silences. As we shall now see, the effects of these lacunae certainly were felt in the nineteenth century, but not acutely enough or often enough to compel the fashioning of a remedy.

In the nineteenth century there appear to have been few challenges at the level of principle to the notion that the president and Congress were entitled to maintain secrecy in the national interest. Certainly, presidents did not display any sense of anxiety when they requested Congress to join with them in concealing information, be it in the matter of ransoms paid to Algeria, diplomatic initiatives with European powers, or negotiations with Mexico.88 This principle was also readily defended by the leading commentators of the day. In A View of the Constitution, for instance, William Rawle observes that calls for the president to disclose information relating to foreign affairs had been justified on the ground that “in republics there ought to be few or no secrets.” But this is, Rawle writes, “an illusory opinion, founded on ideal conceptions, and at variance with the useful practice of mankind,” since exposing the transactions of a cabinet to “the public eye” would merely serve to impede its operations and allow “improper advantages” to be taken.89 This point is reiterated in Justice Joseph Story’s Commentaries on the Constitution, which stresses the need for secrecy in foreign relations, especially with regard to the formation of treaties, on the grounds that “no man at all acquainted with diplomacy” could have failed to realize that “the success of negotiations as often depends upon their being unknown by the public, as upon their justice or their policy.”90

This is not to imply that no criticism of state secrecy was voiced during this time. Certainly there were members of Congress who spoke up quite strongly on behalf of publicity.91 But even these members typically conceded that the final say on how far this ideal could be realized lay with the president. For instance, during the epic debate on the Jay Treaty in 1796, Representative Abraham Baldwin argued that “the importance of having many Government secrets was diminishing,” since experience had shown that “the greater the publicity of measures the greater the success.” Moreover, in a “free Government,” said Baldwin, the “arguments for and against measures [ought] to be known to the people.” Nevertheless, he ultimately affirmed that the president was entitled to exercise secrecy if there was “any temporary impropriety” associated with making the relevant information public.92 Similarly, Article I, Section 5, Clause 3 of the Constitution continued to attract scattered criticism in the decades immediately after Ratification, most notably in St. George Tucker’s View of the Constitution.93 But such views were evidently in the minority, as the Senate and the House faced little opposition when they established rules formalizing the secrecy of their executive sessions.94

More controversial, though, were the attempts of presidents to withhold military and diplomatic information from Congress. This was a domain over which presidents asserted their control from quite early on, as the Senate learned when it asked President Washington in 1794 for details pertaining to the ongoing negotiation of the Jay Treaty. President Washington agreed to share some information, but he withheld “those particulars, which in my judgment, for public considerations, ought not to be communicated.”95 The House soon learned that it was even more disadvantaged on this front. In 1796, it demanded access to all the materials relating to formulation of the (now completed) Jay Treaty before it would make the relevant appropriations. But President Washington declined to share the information, arguing that “even when brought to a conclusion a full disclosure of all the measures, demands, or eventual concessions which may have been proposed or contemplated would be extremely impolitic.”96 The evidence indicates that these precedents hardened over the course of the nineteenth century. Sofaer and Rozell have very ably documented the relevant instances.97 I would add only that the leading commentaries of the day evidently did not view this practice as a travesty. James Kent, for instance, in his Commentaries on American Law notes forthrightly that though in republics “determinations respecting peace, as well as war” had hitherto been made in assemblies, the Constitution had been influenced by the realization that “secrecy and despatch” rendered it “expedient to place this power in the hands of the executive department.”98

Why did these claims not provoke a more sustained critique from Congress? It was not as if lawmakers were unaware of what this arrangement boded. On the contrary, they stood up at regular intervals to voice concern that the president’s control over the flow of secret information threatened to hollow out Congress’s coordinate powers. For instance, in 1796 (now Representative) Madison observed that the president’s claim that he had the right to withhold information relating to foreign relations (in the context of the Jay Treaty) could make it difficult for the House to deliberate “on the subjects submitted to them by the Constitution.”99 In 1826 when President John Quincy Adams proposed to send ministers to the Panama Congress, his critics in Congress, alarmed by what this implied for America’s policy of abstaining from foreign entanglements, demanded that he share further information about the mission. When he declined to comply in full, Senator Robert Hayne complained that while the president “leaves us free to act as we think proper, he refuses to furnish us with the information on which alone we could act, and for which we had respectfully called.”100 Then again in 1843 when the Senate was confronted with reports indicating that President John Tyler was secretly negotiating with the British on the future of Oregon, Senator William Allen complained that “the hitherto tolerated practice” of “concluding the most important treaties” without “even informing this body that negotiations are pending” was a “practical and dangerous departure from the letter and spirit of the constitution.”101 The problem, his colleague Senator Thomas Benton explained, was that the secret formulation of treaties “deprived the Senate of their free action upon the ratification of treaties” because it left them facing a stark choice: either “to confirm what was done, or to embarrass the President.”102

In the end, though, these complaints had no lasting effect because leading members of Congress also admitted the weight of the argument on the other side. As we have seen, one of the central justifications for creating the presidency was that the size and structure of the Continental Congress had made it difficult to maintain secrecy. Given this history, it should come as no surprise that when Madison stood up to respond to the Washington administration’s refusal to share documents relating to the Jay Treaty, he could not help but admit that while the House “must have a right, in all cases, to ask for information which might assist in their deliberations,” the president also “had a right” and “a due responsibility” to withhold information that was of a nature “that did not permit a disclosure of it at the time.”103 This posture soon crystallized into precedent as Congress began to qualify its requests for secret information with a fateful formula—that the executive should share relevant information “unless in the opinion of the President it is contrary to the public interest” to do so.104

Needless to say, this formula did little to resolve the question of how Congress might safeguard the independent exercise of its coordinate powers. In the absence of clarity, members of Congress occasionally attempted to force the issue. More often than not, though, these attempts failed to obtain the support of Congress itself, as was the case in 1826 when Representative Samuel Ingham tried to persuade his colleagues that the gravity of the question before them (whether to fund delegates to the Panama Congress) provided sufficient reason to demand information from the president “without limitation of any kind.” Shall we tell our constituents, said Ingham, that we voted for the measure because “the President knew all about it?” This would be to admit, he said, “that we, the Representatives of the People … had agreed to become a mere registering assembly for Executive edicts.”105 Quite a few of Ingham’s colleagues agreed with this characterization of the problem. Representative James Hamilton argued, for instance, that to fail to inquire deeply would be to assume that “the President can do no wrong.”106 However, the majority still sided with Representatives Daniel Cook and Daniel Webster, who outlined a much more modest role for Congress when they argued that, in the absence of clear evidence of wrongdoing, there was no reason to doubt that “the President may have the best reasons” for withholding information.107

Other lawmakers were more blunt in their assessment of Congress’s claims. In 1831 a bloc of senators charged President Andrew Jackson with having violated the Constitution by secretly dispatching ministers to negotiate a treaty with the Ottoman Empire. Led by Senator Littleton Tazewell, the chairman of the Senate Committee on Foreign Relations, they argued that the Senate ought to decline the president’s request for appropriations to reimburse these ministers since the Senate had not confirmed their appointments. But Senator John Forsyth, the future secretary of state, pushed back, arguing that prior indiscretions, such as Senator Stevens Mason’s 1795 disclosure of the Jay Treaty, had taught the executive that senators could not be counted upon to live up to the “sacred obligation” to preserve the secrecy required to conduct foreign relations.108 And more recent experience, he added, had served only to make it apparent that “if a desire was felt that any subject should … become a topic of universal conversation, nothing more was necessary than to close the doors of the Senate Chamber, and make it the object of a secret, confidential deliberation.”109

Not every attempt to rouse Congress failed, though. In 1848, for instance, the House asked President James Polk to communicate information about the activities of John Slidell, the United States’ minister in Mexico. The request, Polk noted in his reply to the House, was “unconditional,” as it omitted the “customary and usual reservation contained in calls of either House of Congress upon the Executive for information relating to our intercourse with foreign nations.” However, Polk was not to be cowed. Arguing that he could not “violate an important principle, always heretofore held sacred by my predecessors,” he transmitted only information that he believed could be communicated “without serious injury to the public interest.”110 Unwilling to push the matter further, Congress reverted to the old formula. By the late nineteenth century, this formula was so well rehearsed that Hermann von Holst could state in The Constitutional and Political History of the United States that control over the flow of information allowed the president to judge “when the constitutional cooperation of the Senate shall begin.”111

I have been arguing that the absence in the nineteenth century of a sustained challenge to the executive’s monopoly of secret information can be attributed to continuing faith in the idea that the president had the right to determine when secret information should be made available to the public, and to Congress’s deference to the claim that the president was entitled to determine when it could have access to secret information. This account may lead one to wonder why citizens and lawmakers did not chafe more at the fact that the executive’s control over information made it harder for them to enforce accountability. Arguably, the significance of this monopoly was obscured by the United States’ insulated circumstances, which limited the scope and scale of state secrecy, and correspondingly limited the points of conflict between lawmakers and citizens on one side and the president on the other.112 As Francis Lieber observes in On Civil Liberty and Self Government, while the secrecy associated with diplomacy militates against the publicity required for self-government, “a great change has been wrought in modern times, and comparatively a great degree of publicity now prevails in the foreign intercourse of nations.”113 As evidence Lieber cites “one of our first statesmen,” who had privately written to him that “I would not give a dime for all the secrets that people may imagine to be locked up in the United States’ archives.”114

It might be objected that the first half of the nineteenth century witnessed more than a few presidents who made extensive use of secrecy, particularly in the cause of expansion. This is of course true—the acquisitions of Louisiana, Florida, Texas, Oregon, and California were furtive. Note, however, that the furtiveness in these cases was bounded in key respects: the objective of the secrecy was tightly defined, its duration was limited, and the outcome was unavoidably public. In other words, covert activities during this period did not emerge from or stay within the recesses of a security apparatus shrouded in deep secrecy. Indeed the United States had an emaciated security apparatus during this period.115 The principal vehicles for intelligence initiatives were “executive agents” who acted as “pathfinders for American foreign policy.”116 These agents were primarily amateurs who “worked alone and without the support of any far-flung intelligence organization.”117 The “highly sporadic and individualistic” nature of these activities in turn helped limit conflict between the president and Congress.118 For instance, presidents were able to avoid confrontations with Congress by obfuscating the precise role played by these executive agents, often communicating with them via private letters, a practice that allowed presidents to unhesitatingly comply with Congress’s requests for official correspondence.119 The transient nature of these activities also helped calm Congress’s fears. Consider, for example, President Polk’s message to Congress in 1846 where, having declined to share the certificates (or secret payments) drawn up by his predecessor, he diplomatically added, “[F]or my own part … I have had no occasion rendering it necessary in my judgment to make such a certificate, and it would be an extreme case which would ever induce me to exercise this authority.”120

The other reason the executive’s monopoly of secret information did not arouse prolonged concern is that even when the president did undertake covert activity, unauthorized disclosures ensured that lawmakers and citizens were “seldom left … ignorant of executive aims.”121 At times the disclosures came from Congress. Perhaps the best-known example dates to April 1844 when, in the face of political divisions over the wisdom of acquiring Texas, President John Tyler sought to have the treaty of annexation ratified in closed session. As part of the ratification process, the Tyler administration submitted documents pertaining to the negotiations. These documents included diplomatic correspondence suggesting that one of the annexation’s purposes was to bolster the institution of slavery. Senator Benjamin Tappan, an abolitionist opposed to the treaty, secretly disclosed this correspondence, thereby sparking public debate that contributed to the Senate’s rejection of the treaty.122

More often, though, unauthorized disclosures came from within the executive branch. Consequently, throughout the nineteenth century, presidents struggled to keep covert activities out of the press, especially during times of war.123 A prominent episode from the end of the century nicely summarizes the challenge they faced. During the Philippine War, President William McKinley’s administration sought to censor American correspondents posted in Manila. The experiment backfired after the press corps eventually cabled home a widely publicized statement deploring the censorship as a means of preventing Americans from knowing about the military’s excesses and failures. Reflecting on the episode, the Nation warned the president that “if there is one thing which no system of military discipline or official terrorism can long force American officers to do, it is to suppress facts to conceal blunders.” “Leakages,” it said,” are sure to occur,” and so, “the harder the President squats on the safety valve, the higher will he be blown when the explosion takes place.”124

It is not a new observation that unauthorized disclosures helped nineteenth-century citizens and lawmakers challenge secretive executive action. Schlesinger was one of the first to discern this “recurrent pattern” in American history—namely, that “when the republic faced a hard decision in foreign policy and the executive branch had not revealed facts that would enable the people to reach their own judgment, aggrieved citizens felt themselves morally warranted in violating a system of secrecy exploited (as they earnestly believed) by government against the national interest.”125 What Schlesinger glosses over, though, is that these disclosures were highly controversial. Tappan, for instance, was censured by the Senate and only narrowly escaped expulsion.126 At other points, Congress subpoenaed, interrogated, and even imprisoned journalists in an effort to discover their sources.127 In 1890, for example, the Senate created a so-called smelling committee that spent more than five months investigating reporters, staffers, and even senators, in an effort to trace the source of a recent spate of leaks.128 These episodes might well be seen as a manifestation of one of the aforementioned silences in the Framers’ theory. The Framers had implied that secrecy might legitimately be breached in the public interest, but then did not properly explain what constituted good cause. The consequence was controversy: like Rhode Island’s Howell, Tappan believed himself to be justified in disclosing secret information, while Congress evidently thought otherwise. Which side was in the right? The question was not posed frequently or acutely enough in the nineteenth century to provoke a search for a systematic answer.

Turning Points

I have argued that though the Framers did not fully explain how citizens or lawmakers would be able to bring the president to account for the employment of state secrecy, the limited scope and scale of state secrecy in the nineteenth century meant that their silence did not attract sustained attention at the time. This picture, as we shall now see, began to alter once America immersed itself in international politics.

In The American Commonwealth, published toward the close of the nineteenth century, James Bryce observed that republics usually struggled to “define the respective spheres of the legislature and executive in foreign affairs, for while publicity and parliamentary control are needed to protect the people, promptitude and secrecy are the conditions of diplomatic success”; America, however, had been fortunate enough to sidestep the problem because “happy America” stood “apart in a world of her own.”129 A little more than a decade after Bryce wrote these words, America’s long-standing aversion to foreign entanglements began to fade. Woodrow Wilson was one of the first to discern what this meant for the nation. Writing after the Spanish-American War of 1898, Wilson observed that from a constitutional perspective, “interesting things” might come out of the country’s “plunge into international politics and the administration of distant dependencies” such as Hawaii and the Philippines.130 For “when foreign affairs play a prominent part in the politics and policy of a nation,” he wrote, “its Executive must of necessity be its guide: it must utter every initial judgment, take every first step of action, supply the information upon which it is to act, suggest and in large measure control its conduct.”131

Congress soon made the same discovery. The point was brought home in 1906 by a famous debate between Senators John Spooner and Augustus Bacon as to whether Congress was entitled to request from President Theodore Roosevelt information about his decision to send delegates to the Algeciras Conference, which had been called to settle a dispute between France and Germany over control of Morocco. Bacon, fearing that the United States’ decision to attend indicated that “the whole policy of noninterference and nonentanglement is being given away and abandoned,” demanded to know “the limitations of the instructions” given to the American delegates.132 When Spooner questioned the propriety of Bacon’s demand on the grounds that Congress ought not to interfere in the conduct of negotiations, the latter responded that since the Constitution had authorized the president to make treaties “by and with the Advice and Consent of the Senate,” how can “it be proper for the Senate to offer advice or counsel to the President as to the policy or impolicy of a proposed treaty, and at the same time improper to ask for information upon which to base such advice or counsel?”133 Nonetheless, when pressed by Spooner to clarify whether the president was obliged to comply with Congress’s request, Bacon conceded that it was for the president to judge whether such a disclosure was “compatible with the public interest.”134 And so the senators ultimately concluded by reiterating what Madison had stated more than a century earlier: they agreed that “the President has the same right to refuse to share the information as the Senate has to request it.”135

Although the exchange between Bacon and Spooner offered the clearest indication yet that the precedents established in the nineteenth century would make it difficult for Congress to rein in a secretive executive, the scale of the challenge was not immediately evident. This is because as America entered the world stage, the statesmen of the day calmed fears by appealing to liberal-democratic ideals that had taken shape in nineteenth-century Europe, where revolutionaries like Giuseppe Mazzini and Lajos Kossuth, and liberals like Benjamin Constant and François Guizot, had argued that democracies ought to eschew state secrecy, which they associated with the old monarchies of Europe.136 In short order it became commonplace to assert, as former Secretary of State Elihu Root did in 1917, that a democracy was “incapable” of pursuing “sinister policies of ambition,” since the “open and public avowal and discussion which must precede their adoption by a democracy is destructive of them.”137 Following the United States’ entry into World War I, this view took center stage. The belief that the use of “secret counsels” had been responsible for allowing “so stupendous a contest” to be unleashed “without warning to the world,” led (now President) Wilson to urge that the lesson to be learned from the experience was that “the peace of the world must henceforth depend upon a new and more wholesome diplomacy” that would “proceed always frankly and in the public view.”138 Soon, however, the realization dawned that state secrecy had actually increased rather than decreased since Wilson’s famous “Fourteen Points” speech. As the political scientist and former envoy to China Paul Reinsch lamented in 1922, whereas “it was substantially true that the United States had no diplomatic secrets” immediately prior to World War I, now “even in the American government, particularly during and since the war, foreign affairs have been handled with what would ordinarily seem insufficient information to the public.”139

To the scholars who reflected on these developments, it was quite clear that the path ahead lay not in eschewing state secrecy but rather in ensuring that it was used responsibly. “It is all very well to say that an open and straightforward policy best befits a free and high-minded people,” Bryce wrote in Modern Democracies, “but if such a people should stand alone in a naughty world, it will have to suffer for its virtues.”140 At the same time, Bryce acknowledged that, as recent experience had shown, there was a need to address the possibility that state secrecy could obscure “unwise action” and the use of “dishonorable methods.”141 His response to this conundrum presaged future developments. “The risk that secrecy and discretion will be abused will be gradually lessened,” Bryce wrote, “the more public opinion becomes better instructed on foreign affairs, and the more that legislatures learn to give unremitting attention to foreign policy.”142

Bryce’s proposal that Congress involve itself more thoroughly in foreign affairs was echoed by other influential commentators including Edward Corwin, Quincy Wright, DeWitt Poole, Carl Friedrich, and Harold Laski.143 But how could Congress take on this responsibility when it relied on the president for information, as the debate between Senators Bacon and Spooner had underscored? The leading observers of the day were aware of the problem. “Parliamentary action is becoming notoriously ineffective,” Walter Lippmann wrote in Liberty and the News, because Congress’s “sources of information are hardly better than that of any other reader of the newspaper,” whereas the president has “an elaborate hierarchy reaching to every part of the nation and to all parts of the world.”144 This divergence, he noted, was steadily weakening the checks and balances established by the Framers because when the “legislature is haphazardly informed … the people themselves prefer to trust the executive which knows, rather than the Congress which is vainly trying to know.”145

The scholars who counseled faith in the separation of powers accepted that Congress was becoming ever more dependent on the president for information, but they had little to offer by way of concrete solutions. Wright’s suggestion that the president and Congress alleviate conflict over information sharing by cultivating informal “constitutional understandings” was typical of the period.146 But what if such “comity” were to break down? There would still be no cause for concern, Poole argued, because Congress could obtain whatever it needed through the judicious use of the powers at its disposal, including its control over finances, appointments, and the approval of treaties.147 What Poole did not address, though, was the question that Spooner had used to checkmate Bacon—that is, ought Congress to force the president to divulge information when the Framers had created the latter to keep secrets that they thought the former could not?

Bryce’s other proposal—the instruction of public opinion—accommodated the demand for great public oversight. But this proposal raised the question of who was to do the instructing. It was Lippmann, once again, who discerned the problem. If the task of instructing the public was left to officials, he warned, then propaganda would win the day. At a time when individual opinion relied more than ever on public opinion, there could be, he wrote, “no liberty for a community which lacks the information by which to detect lies.”148 Consequently, the need of the hour, Lippmann argued, was to increase the participation of disinterested experts, who could investigate policy questions without fear or favor and then guide public opinion in the appropriate direction.149 But how could such experts form “more comprehensive and distant views” when the relevant information might be secreted within the recesses of the state? “It is difficult to see,” Lippmann wrote, why, save for “a few diplomatic and military secrets,” all the information at the disposal of the State Department “should not be open to the scholars of the country.”150 The problem with this line of reasoning, though, was that, as Reinsch had correctly discerned, diplomatic and military secrets were not so few any more.

Few felt the expansion of secrecy more keenly than the press. Confronted with the Wilson administration’s concerted effort to introduce formal censorship during World War I, the press defended its freedom to cast the “pitiless light of publicity” on “every scandal” and “every blunder,” including those in the diplomatic arena, by arguing that in view of the propensity of state secrecy to foster misunderstanding and incompetence, “even reckless speech may be a moderating influence, whereas drastic censorship chokes the safety valve.”151 However, Congress’s debate on the Wilson administration’s proposal shows only limited support for this point of view. Certainly, a whole host of congressmen, led by Senator William Borah, objected to censorship on the grounds that a vigilant press was needed more than ever during war, a time “when sordidness and greed are always active,” thus emphasizing what Vincent Blasi would later term the “checking value” of the First Amendment.152 But the slender majority in Congress who supported this view conceded that the executive was entitled to prevent the disclosure of secret information in the first place, and that Congress retained the right to penalize the publication of information harmful to national security.153

If the questions raised by the proposals put forward in the wake of World War I were initially left unexplored, this was because even though state secrecy had now officially appeared on the scene—in the form of classification guidelines and the passage of the Espionage Act in 1917—the scope and scale of the United States’ national security apparatus was still limited at this point.154 Indeed, even as late as the onset of World War II, President Franklin Roosevelt was still relying on executive agents, who were, as one important observer noted, “amateurs without special qualifications and without training,” who lacked “special means of communication or other facilities,” and thus constituted a “small and uncoordinated force.”155 By the end of World War II, though, the picture was very different: whereas in 1916 the Department of State had had only four officials officially charged with managing intelligence matters, by 1945 upwards of fifteen thousand had worked for the Office of Strategic Service (OSS), which was only one of a plethora of intelligence organizations created during the war.156 Soon a new threshold was breached. With Pearl Harbor still fresh in the minds of decision-makers, and with the Soviet Union looming large on the horizon, support grew for the notion that America required a permanent peacetime intelligence establishment, and, in short order, a host of powerful new organizations including the Central Intelligence Agency (CIA) and the National Security Agency (NSA) had been established.157

Accompanying these institutional advances were constitutional developments formalizing the president’s right to control the flow of information relating to this emergent national security apparatus. The first such development came in the form of Executive Order 10290, which cited the president’s “implied powers” as the basis for establishing, in 1951, a thoroughgoing “system for the safeguarding of official information the unauthorized disclosure of which would or could harm, tend to impair, or otherwise threaten the security of the nation.”158 Another was the formal enunciation in 1954 of an “executive privilege” to withhold information from Congress. Drawing on new research detailing the instances in which presidents had previously withheld information from Congress, Attorney General Herbert Brownell now argued that the president had a right to withhold information from Congress. These precedents showed, he wrote, that “throughout our history the President has withheld information whenever he found that what was sought was confidential or its disclosure would be incompatible with the public interest or jeopardize the safety of the Nation.”159

A third development was the recognition in United States v. Reynolds (1953) of what has since come to be termed the “state secrets privilege”—that is, the right of the state to refuse to produce evidence sought by plaintiffs on the grounds that its disclosure would gravely harm national security. Although this privilege had already been recognized in the nineteenth century, it had rarely been invoked. However, once the United States began to undertake covert activities, the courts encountered plaintiffs seeking access to secret information to prove violation of their rights.160 Reynolds revealed that the judicial process could do little to lift the veil on such activities because “even the most compelling necessity,” the Supreme Court ruled, “cannot overcome the claim of privilege if the court is ultimately satisfied that military secrets are at stake.”161

As these events unfolded, observers now began to recognize for the first time how the checks and balances inherited from the Framers could be short-circuited by state secrecy. The heightened use of secrecy, Harold Lasswell warned, meant that “the member of Congress who takes up a critical attitude toward the executive on national defense issues will find himself in a less and less tenable position,” because he will not be able to share with the public information that challenges the president’s statements and policies.162 Wallace Parks pushed further. Congress, he noted, typically relies on experts from the “non-governmental community” for the information needed to check the executive. How can Congress receive such guidance in military and diplomatic matters, he asked, when “the necessary information and communications have been severely restricted?”163 Robert Dahl and Lewis Coser drew attention to the problem of democratic control in the face of increasing state secrecy. “Given a monopoly on knowledge,” Coser observed, “power holders can exert dominion over all those who are incapable of acting rationally because they do not know the real situation.” “Hence an increase in the secrecy of governmental action,” he concluded, “may be taken as an index of the drift toward the garrison state in America.”164 Meanwhile, Zechariah Chafee fretted that the “enormous recent expansion of the subjects which officials are seeking to hide from publication” meant that “official encroachments on the freedom of the press will be probable unless the boundary line between secrecy and publicity is very carefully demarcated.”165

These grim pronouncements prompted calls for a “proper balance” to be struck “between secrecy and disclosure.”166 But could officials be trusted to strike the right balance? Chafee sounded an early note of skepticism, warning that “officials must not do the demarcating” because they would be tempted to use this power to “hoist public safety as an umbrella to cover their own mistakes.”167 In a little over a decade, such skepticism became mainstream. One reason for the change was the U-2 incident of May 1960, when Americans discovered that President Eisenhower had misled them about provocative surveillance flights over the Soviet Union. The incident, Francis Rourke observed at the time, provided “a clear indication of the power that has come to rest in the hands of government officials to influence public attitudes in foreign affairs through their control over the release of information.”168 President Eisenhower’s subsequent claim that the administration had lied in order to deceive the Soviets only made matters worse, Rourke pointed out, not only because the explanation seemed implausible (as the Soviets already knew about the surveillance flights), but also because it implied that officials might deceive the American public again, if they deemed it necessary to do so.

The other reason for growing public skepticism was newfound public awareness of officials’ tendency to classify information whose disclosure could not realistically be said to threaten national security. This problem of “overclassification” was brought to the public’s attention by high-profile congressional hearings and news reports citing instances where the classification stamp had been used to conceal information ranging from the mundane to the embarrassing. As they searched for a theoretical explanation for this phenomenon, commentators latched on to Max Weber’s essay “Bureaucracy,” which had recently been translated into English. From Weber they took the idea that bureaucracies had self-interested reasons to extend secrecy “far beyond those areas where purely functional interests make for secrecy.” In particular, they became convinced that Weber was right to argue that “every bureaucracy seeks to increase the superiority of the professionally informed by keeping their knowledge and intentions secret.”169

If the bureaucracy could not be trusted to deal fairly, then how was secrecy to be checked? Crucially, prominent bipartisan investigations emphasized that unauthorized disclosures were not the answer. In 1956 the Defense Department Committee on Classified Information (the Coolidge Committee) declared that though “reasonable men may differ” over national security decisions, it was a “deplorable thing” for officials to “carry their ardor to the point of undermining the system on which the nation relies for the protection of its defense secrets.” Such action, the committee concluded, ought to be checked by punishment—“however high the motives of an individual might seem to himself, he is guilty of a serious offense and should be dealt with accordingly.” And should it prove difficult to identify the source, the committee added, the reporter ought to “be summoned to testify in a grand jury investigation in order to discover the source of the leak.”170 The Commission on Government Security (the Wright Commission), which reported in 1957, was even more blunt, calling for the “unequivocal prohibition” of unauthorized disclosures and the “vigorous prosecution of every offender.” “The final responsibility for the difficult decisions of what shall be secret,” Chairman Lloyd Wright declared, “must be confided in those loyal and devoted public servants who are qualified to make the judgment.” “No citizen,” he emphasized, “is entitled to take the law, and the safety of the nation, into his own hands.”171

The recommendations of the Coolidge Committee and the Wright Commission were not adopted. But the skepticism these recommendations embodied did reflect wider political concern about uncontrolled disclosures. Chafee, for instance, cautioned that allowing secret information to slip out in the form of gossip “is more dangerous than frank discussion in the general press.” It would be far preferable, he concluded, for Congress to determine what matters should be kept secret, thereby allowing for “frank official disclosures.”172 Rourke made much the same point. “As the situation now stands,” he observed in 1961, “the leak often serves as something of a safety valve.” But there was the distinct chance, he added, that the officials responsible for disclosing information could make erroneous judgments about the need for secrecy, thereby keeping secret either too much or too little.173

It was in this context that proposals to legislate a “right to know” gained traction. A series of publications by members of the American Society of Newspaper Editors, most prominently, Harold Cross’s The People’s Right to Know and James Wiggins’s Freedom or Secrecy, struck some of the first blows.174 Crucially, these writers did not oppose state secrecy itself. Rather, they warned that the unchecked use of the classification stamp meant that citizens were increasingly being deprived of the knowledge required to determine the “adequacy of policy and the fidelity of individual public servants.”175 Hence they proposed that classification decisions take due account of the public’s interest in the availability of information—and that the courts play the role of arbiters.

This argument found a receptive audience in Congress, where Representative John Moss and Senator Thomas Hennings championed it to great effect. “No one denies the necessity for true security measures in the interest of national defense,” Moss argued, but in recent years secrecy claims had been “so broadly stated as to constitute a real and present danger to our system of representative government.”176 The need of the hour, his colleague Hennings argued, was for guidelines ensuring that information would be withheld “only to the extent that the effective and proper exercise of the President’s power to conduct the foreign affairs of the nation requires it.”177 Such guidelines were soon put in place. In 1966 Congress passed the Freedom of Information Act (FOIA), which placed on the executive the burden of proving that the imposition of secrecy in a given case was “specifically authorized under criteria established by an Executive Order to be kept secret in the interest of national defense or foreign policy.”178 As the conference report accompanying the legislation explained, the objective of FOIA was to establish “a general philosophy of full agency disclosure unless information is exempted under clearly delineated statutory language and to provide a court procedure by which citizens and the press may obtain information wrongly withheld.”179

Around this time Congress also began attending to its own needs by contesting the president’s claim that he had a right to withhold information from it. Initially, it collated a rival set of precedents that evidenced “the power of Congress and its committees to obtain information deemed necessary to the legislative process.”180 Subsequently, it took on Attorney General Rogers’s claim that the executive privilege was founded on the separation of powers by making the case that the very same doctrine also supported its right to information, since “a legislative body cannot legislate wisely or effectively in the absence of information respecting the conditions which the legislation is intended to affect or change.”181 For the most part, though, both of these challenges were directed at invocations of an executive privilege to withhold unclassified materials pertaining to internal deliberations. As far as classified information was concerned, Congress continued to show deference. During this period the subcommittees responsible for overseeing the intelligence community were composed of a few senior members who, according to Frank Smist, “chose not to be involved and preferred to be uninformed.”182 This deferential attitude was soon to be cast aside, though.

The turning point came in the first half of the 1970s when a sequence of events exposed the extent to which state secrecy could be used to conceal questionable, and in some cases blatantly illegal, activities. In June 1971, the New York Times published the Pentagon Papers, a classified study of decision making leading up to and during the Vietnam War. The study revealed that successive administrations had “systematically lied, not only to the public but also to Congress” about the reasons for, and prospects of, the war.183 Then in July 1973, a whistleblower, Major Hal Knight, informed Congress that the Nixon administration had secretly ordered the bombing of Cambodia, even as it publicly claimed to be respecting that country’s neutrality.184 The ensuing investigation revealed that the administration had been systematically falsifying records, including those submitted to the Senate Armed Services Committee, in order to hide the secret bombing missions from public view. Not long after this revelation came Watergate. Though this scandal did not directly involve national security matters, the Supreme Court–ordered disclosure in July 1974 of tape recordings from the White House revealed that President Richard Nixon and his aides had attempted to ward off FBI investigators by hinting that the Watergate break-in was part of a secret national security operation.185 Finally, in December 1974 Seymour Hersh of the New York Times published the so-called Family Jewels, an internal CIA document listing the activities that the organization had undertaken in contravention of domestic and international law, including assassination plots against foreign leaders, the overthrow of foreign regimes, the infiltration of domestic political groups, and warrantless domestic surveillance.186

These revelations sparked outrage and transformed what had previously been a general uneasiness about state secrecy into outright hostility. It now began to be argued, most prominently in Schlesinger’s The Imperial Presidency, that Americans needed to combat “the rise of the religion of secrecy” by depriving the executive of its monopoly of the classification system.187 But how was this to be done? First Amendment scholars argued that recent events provided a clue. The seminal contribution came from Alexander Bickel. In Bickel’s view, the Pentagon Papers episode had demonstrated that the press provided the “countervailing power” against undue secrecy, because the First Amendment’s disapproval of prior restraint meant that while the government was allowed to “guard mightily” against leaks, it had little choice but to “suffer them if they occur.” Though the resulting cat-and-mouse “game” might be “disorderly,” Bickel observed, it was nonetheless “effective.”188

But this argument received a lukewarm reception. Louis Henkin, for instance, argued that “trial by battle and cleverness between the three estates and the fourth hardly seems the way best to further the various aims of a democratic society.” This “unhappy game,” he argued, “does not ensure that what should be concealed will not be uncovered,” and “the rare, haphazard, fortuitous, journalistic uncovering will hardly achieve effective public knowledge of all that should be known.”189 It would be far better, he concluded, for Congress and president to tackle the problem of overclassification. This point was echoed by Schlesinger, who also expressed hesitation about the wisdom of relying on a “rebellious collaboration between anonymous and disgusted officials and the press.” Since the rightness or wrongness of an unauthorized disclosure could be worked out only after such an action had already been undertaken, there was the danger, he warned, of too little or too much disclosure, depending on the proclivities of the officials and reporters involved. “Might it not be better,” he asked, “to maintain the balance between secrecy and disclosure in a less nerve-racking way?”190 To this end, he proposed two paths along which reform could proceed. The first was to establish “some form of appellate procedure” to help ensure that “classification decisions met standards of reason.”191 The second was to compel the president to “supply Congress the information necessary to responsible debate,” for instance, by establishing “as a matter of law that CIA intelligence analyses be made available to the relevant committees.”192

Schlesinger’s analysis proved prescient. Emboldened by the widespread distrust of the presidency, Congress enacted two major changes in the institutional framework regulating the employment of state secrecy. In 1974, it amended FOIA to authorize the courts to examine classified records in camera in order to determine whether they legitimately qualified to be withheld under various national security exemptions. These amendments, adopted over President Gerald Ford’s veto, effectively invited the courts to oversee the classification system. The other change was the creation of an intelligence oversight system in the form of the House Permanent Select Committee on Intelligence in 1977 and the Senate Select Committee on Intelligence in 1976.193 Congress also passed the Hughes-Ryan Act in 1974 and the Intelligence Oversight Act in 1980, making it compulsory for the president to keep select members of Congress “fully and currently informed” of “significant anticipated intelligence activity,” including covert operations.194 Congress did not, however, bolster protection for the officials, reporters, and publishers responsible for transmitting unauthorized disclosures, declining, for instance, to establish a reporter’s privilege to protect the identity of confidential sources or to revise the Espionage Act, which had been used to prosecute Daniel Ellsberg in the Pentagon Papers case.

In short order, however, it became clear that the enacted reforms had failed to challenge the president’s control over the flow of secret information. During President Ronald Reagan’s second term, citizens and lawmakers learned—once again via unauthorized disclosures—that the administration had secretly facilitated the sale of arms to Iran, covertly provided support to the Contras in Nicaragua in violation of the law, and utilized American media outlets to undertake a disinformation campaign targeted at Libya. Representative Norman Mineta, a member of the House intelligence committee during this period, described Congress’s position in memorable terms: “we are like mushrooms. They keep us in the dark and feed us a lot of manure.”195 Meanwhile, civil society activists discovered that judicial deference to the executive’s claims about the harm likely to be caused by the disclosure of classified information meant that FOIA could not help them obtain access to seemingly basic information about the executive’s activities—for example, the size of the intelligence budget. As Robert Deyling glumly reported in 1992 after surveying the empirical evidence, since the enactment of FOIA the courts had “ruled on hundreds of cases involving classified information, affirming the government’s decision to withhold the requested information in nearly every case.”196

These setbacks prompted further calls for reform. For instance, Harold Koh argued that Congress ought to delegate the oversight of national security matters to a “core group of members” comprising a handful of its highest-ranked officials: limiting the number of overseers would, he felt, make it harder for the president to refuse to share information on the grounds that Congress was prone to indiscretion.197 Sissela Bok, meanwhile, argued that the level of concealment in American government had become “pathological” owing to a deficiency at the heart of FOIA, whose proponents had failed to see that allowing information to be withheld on national security grounds would enable conniving officials to defeat the realization of publicity. Citing Weber, Bok warned that laws such as FOIA “can serve the public well only if the exceptions to them are kept to a minimum and are prevented from expanding.”198 This point was reiterated by the Commission on Protecting and Reducing Government Secrecy (the Moynihan Commission), which concluded in 1997 that there was a pressing need for “some check on the unrestrained discretion to create secrets” and for an “effective mode of declassification.” To this end, the commission recommended the establishment of an independent National Declassification Center to oversee “systematic declassification.”199 What the commission explicitly rejected, though, was the idea that unauthorized disclosures might serve as a means of countering overclassification. “There must be,” Senator Daniel Moynihan declared, “zero tolerance for permitting such information to be released through unauthorized means.”200

Barely had the ink dried on the Moynihan Commission’s report than the onset of the so-called war on terror prompted the administrations of Presidents George Bush and Barack Obama to employ an array of covert capabilities. As events unfolded, it quickly became clear that the executive continued to maintain a stranglehold over the flow of information relating to the use of these capabilities. For instance, in 2002 Congress authorized the use of military force in the wake of assertions by officials that secret intelligence revealed Iraq to be developing weapons of mass destruction and aiding terrorist organizations hostile to the United States. When these assertions eventually proved to be unfounded, members of Congress drew the conclusion that there had been “an exaggeration” of the threat.201 According to Senator Dianne Feinstein, the episode underscored how vital it is for Congress to have “fairly presented, timely and accurate intelligence when they consider whether to invest in the President the authority as Commander-in-Chief to put American lives, as well as those of innocent civilians, at risk.”202 However, in spite of promises of closer oversight in the future, in 2006 a majority of the Senate intelligence committee found out about the NSA’s warrantless wiretapping program only after the New York Times published a story on the program. Peeved, Senator Ron Wyden, a committee member, complained that he and his fellow senators had been forced to hire a news-clipping service to bring such reports to their notice. “My line,” he is reported to have said, is “What do I know? I’m only on the Intelligence Committee.”203

Congress is not the only branch to have struggled to oversee the president’s secret activities during the war on terror. Over the past decade, the courts too have been hard-pressed to help citizens and lawmakers lift the veil on covert operations that have apparently violated the dignity and rights of individuals targeted in counterterrorism operations. For instance, FOIA has proven ineffective as a means of compelling the disclosure of documents detailing the treatment of suspected terrorists because judges continue to defer to the executive’s assessment of the harm likely to be caused by disclosure of such information. This record has led critics such as Pallitto and Weaver to declare that judges have “abdicated” the role that FOIA intended for them to play—namely, to serve as independent assessors of classification decisions.204 The courts have also proven unwilling to closely scrutinize the government’s invocations of the state secrets privilege. As a result, complainants who have been subjected to extraordinary rendition and warrantless wiretapping have found themselves denied a forum in which to establish their claims and seek judicial remedy—a “harsh result” that has also attracted criticism from legal scholars.205

Not every regulatory mechanism has proven ineffective, though. To the extent that citizens and lawmakers have become aware of potential wrongdoing in the past decade—the establishment of secret prisons, the practice of extraordinary rendition, and the existence of warrantless surveillance programs—this has been due to unauthorized disclosures. The executive’s response to this development has been unambiguous: the Bush and Obama administrations have together prosecuted more officials than all their predecessors combined.206 Notably, neither Congress nor the courts have intervened strongly on behalf of either officials or the press. On the contrary, lawmakers have routinely condemned such disclosures, while the courts have permitted legal action against officials and reporters to proceed. Not surprisingly, these developments have drawn strong criticism from First Amendment scholars such as Stone and Kitrosser, who have called for enhanced protection for officials and reporters on the grounds that the law currently “gives inordinate weight to secrecy at the expense of informed public opinion.”207 However, proposals of this variety have been fiercely opposed, most recently by Lillian BeVier and Schoenfeld, who argue that unauthorized disclosures are unacceptable because of the “injury to democratic rule when unelected individuals act to override the public will” as expressed by elected representatives.208

The Dilemma

To recapitulate: I began by asking why state secrecy is approved in principle and censured in practice. The prevailing explanation, as we have seen, blames the presidency for having exploited “popular fear” and “popular faith” during the Cold War to establish a secrecy system that it has since used to its advantage. The problem with this explanation, I have argued, is that the Framers clearly expected the executive to employ state secrecy. Hence the executive’s real and imagined transgressions cannot simply be a product of war hysteria; they must derive from something more deep-seated than that. The real cause, we have seen, is a silence in the Framers’ theory. The Framers authorized the president to employ secrecy in the public interest, but did not fully explain how citizens and lawmakers could know whether the president is in fact exercising this power responsibly. This silence did not produce lasting crises of confidence in the nineteenth century because the dearth of foreign entanglements afforded presidents little reason or opportunity to employ state secrecy extensively. However, once the United States immersed itself in international politics at the turn of the twentieth century, the concomitant increase in the scope and scale of secrecy magnified the impact of the Framers’ silence.

This conclusion raises an obvious question. For more than half a century now, scholars have addressed the Framers’ silence by pushing for reforms intended to loosen the president’s stranglehold over the flow of secret information. Why, then, does American public life continue to be roiled by controversies over the employment of state secrecy? The problem, as we have seen, is that it is not easy to fill the Framers’ silence. Contemporary efforts have arrived at an impasse. The regulatory mechanisms that have been championed in recent decades—the Freedom of Information Act and the establishment of congressional oversight committees in particular—have proven ineffective at exposing wrongdoing. Meanwhile, the regulatory mechanisms that have proven effective at exposing wrongdoing—the practices of whistleblowing and leaking—are condemned as unlawful and therefore illegitimate. It turns out, in other words, that the available safeguards are either ineffective or undesirable.

Can this dilemma be solved? That is, is it possible to transform legislative oversight and judicial review into more effective checks on the employment of state secrecy? And if not, are there conditions under which the making of unauthorized disclosures by officials, reporters, and publishers can be defended as legitimate? These are the questions we shall examine going forward. What we shall find is that this dilemma is far harder to solve than commentators have hitherto recognized.