THE FIRST great object of a constitution, believed many of the Framers of the United States Constitution, is to enable the government to protect the people from external attack. Relative to the track records of other countries, the U.S. government’s success rate in that regard has been little short of astonishing. Through 225 years of threats from air, sea and land that have claimed hundreds of millions of lives in other countries, only a tiny handful of Americans have fallen victim to such attacks. So remarkable has been the government’s record that in the aftermath of the tragedy of September 11, 2001, Americans were at a loss to recall the last major deadly attack within North America (in June 1876, when Sioux Indians wiped out Custer’s cavalry at the Battle of the Little Big Horn, killing more than 250). Few today are surprised to learn that lightning strikes have killed more Americans in the last twenty years than terrorist attacks.

Enabling the government to protect the people, however, is hardly the Constitution’s only purpose. Its second great object, the Framers believed, is to protect the people from the government. In this respect, the risk has risen considerably, for the greater the government’s capacity to protect against external threats, the greater the internal threat from the government itself. The Framers sought to meet that internal threat in part by setting up a system in which the three branches of the federal government, in competing for power, would produce an equilibrium that would guard against autocracy. But today that equilibrium has largely broken down. In July, the CIA acknowledged that it had spied on its Senate oversight committee and then lied about it. Given the emblematic significance of that event, a brief recap is in order.

In 2009, the Senate Select Committee on Intelligence initiated a study of the CIA’s rendition, detention and interrogation activities. The CIA itself installed computers at a CIA facility for use by the committee staff in reviewing relevant documents. Some of the information on the computers was understood to consist of Senate documents and as such its availability was restricted to the committee staff. Nonetheless, five CIA officials—two attorneys and three information-technology staff members—surreptitiously accessed the documents. Following a related dispute with the committee staff, a CIA official accused that staff of crimes and filed a report with the Justice Department (which itself turned out to be based on inaccurate information).

In response, Senator Dianne Feinstein, chairman of the committee, took to the Senate floor in March 2014 to say that “CIA personnel had conducted a ‘search’—that was [CIA director] John Brennan’s word—of the committee computers . . . of the ‘stand-alone’ and ‘walled-off’ committee network drive containing the committee’s own internal work product and communications.” In addition, she said, the agency had removed files from the committee’s computers, read its staff members’ e-mail messages and tried to intimidate them. Brennan, however, was quick to deny any wrongdoing by the agency. “Nothing could be further from the truth,” he said. “I mean, we wouldn’t do that. I mean, that’s just beyond the scope of reason in terms of what we would do.”

On July 31, four months later, the CIA inspector general issued a one-page statement confirming the spying. Brennan apologized to Feinstein and the committee’s vice chairman, Senator Saxby Chambliss (though not to the committee, the Senate, the president or the public). President Barack Obama proceeded to assert his “full confidence” in Brennan. The CIA announced that a panel would be set up to look into the matter, including, presumably, Brennan’s own role. Its members would be selected by Brennan. Though the Justice Department declined to investigate, major questions remained unanswered. Who ordered the search? How many intrusions occurred? Who within the CIA was given the purloined documents? Were they transmitted beyond the CIA? Who within the White House, if anyone, was informed of the CIA’s searches? When did the White House learn of them? And what action did it take then?

This was not the first time that Obama had failed to take disciplinary action in response to a senior intelligence official’s public falsehood. Director of National Intelligence James Clapper, the official Obama designated to oversee the declassification of the torture report, testified on behalf of the Obama administra­tion before Feinstein’s committee on March 12, 2013. He was asked directly about the National Security Agency’s (NSA) surveillance by Senator Ron Wyden. “Does the NSA collect any type of data at all on mil­lions or hundreds of millions of Americans?” Wyden asked. Clapper responded, “No, sir.” Wyden followed up: “It does not?” Clapper replied, “Not wittingly.” Following the Edward Snowden disclosures, Clapper admitted that his testimony was false. On June 9, 2013, he described his response to NBC’s Andrea Mitchell as the “least untruthful” statement he could give, suggesting that he had understood the question and deliberated on how it should be answered. (Unlike Clapper’s, Brennan’s statement was not made to a congressional committee and therefore was not subject to potential criminal penalties.)

EVEN WITH the little that was publicly known, it was clear that the legal implications of the agency’s spying on its Senate oversight committee were nonetheless significant. The Constitution’s separation-of-powers doctrine supposes three separate and independent branches of government that do not encroach upon the constitutionally assigned functions of each other, including the oversight responsibilities of legislative committees. The Fourth Amendment prohibits unreasonable searches and seizures. The National Security Act of 1947, which set up the CIA, prohibits the agency from performing any “police, subpoena, law-enforcement powers, or internal-security functions.” Executive Order 12333 prohibits the CIA from conducting domestic searches or surveillance. In addition, various statutes impose criminal penalties. The Wiretap Act, for example, prohibits the intentional, unauthorized interception of electronic communications. The Computer Fraud and Abuse Act prohibits intentional, unauthorized access to computers. In a March 5, 2014, letter to Wyden, Brennan acknowledged that this act applies to the CIA.

The CIA’s spying was thus no trivial staff quarrel requiring merely a personal apology. Willfully deceiving a governmental fact-finding body, whether a court or a congressional committee, undermines the integrity of the American legal system. In the constitutional design, these organs were intended to be the government’s portals to truth. To carry out their duties, they depend upon an accurate assessment of the facts. When they are misled, their work product is suspect; judicial opinions and legislative findings then rest upon falsehood. The body politic casts votes based upon misinformation, electing candidates who would not otherwise hold office. The entire system of constitutional and electoral checks on abusive power is thereby corrupted.

Those who mislead no doubt believe that they do so for a greater good, the protection of the nation’s security. They are mistaken. The CIA, the NSA and other elements of the military/intelligence community do not exist merely to prevent airplanes from flying into buildings. Their larger mission is to protect the nation’s democratic institutions and the rule of law established by the Constitution. When elements of the national-security apparatus deceive Congress or the courts, they feed the perception that the whole system is rigged and undermine the very institutions that it is their mission to protect.

Distrust of government tends to become generalized. People who doubt government officials’ assertions on national-security threats are inclined to extend their skepticism to other arenas. Governmental assur­ances concerning everything from vaccine and food safety to the fairness of stock-market regulation and IRC investigations (not without reason) become widely suspect. The protection of legitimate national-security interests itself suffers if the public is unable to distinguish between measures vital to its protection and those assumed to be undertaken for reasons of doubtful validity.

Further, it does not strengthen the United States in its relations with other nations to engage in deception. It weakens our government when its institutions are seen around the world as hollow or its officials as duplicitous. The United States’ historic advantage in its international relations has been not merely military or economic. It has been reputational. Legislative and judicial monitors that operate independent of the executive branch, that are able to call the military and intelligence agencies to task when they run amok, lie at the core of America’s reputation for a robust rule of law. Whether the United States thrives or declines in this century will rest in large part on its ability or inability to maintain democratic accountability by safeguarding the integrity of its institutions.

President Obama thus said more than he intended when he stated, referring only to torture, that “we did some things that were contrary to our values.” As Senator John McCain said, in some ways the spying incident was “worse than criminal.” Had Obama acted consistently with American values—had the system worked—the president would have dismissed Brennan the moment his mistruth became evident. The Justice Department would immediately have initiated an investigation to determine whether CIA officials had violated the law by spying on the committee. The committee leadership would have subpoenaed Brennan at once and called him to testify, under oath, about what he knew and when he knew it. Congress would have been incentivized to do so by an outraged public, informed and galvanized by a record of judicial opinions from cases in which the courts had heard public testimony about the duplicity, kidnapping and brutality that every knowledgeable observer knew had stained American counterterrorism policy.

But the system did not work. Instead, Obama, more presider than decider, sat mutely for months while Clapper’s earlier dishonesty festered, even though Obama knew, or should have known, that the intelligence chief’s testimony was false. Obama’s silence signaled that official misstatement of the facts would now go unpunished, a premise that Brennan readily embraced. Indeed, the Justice Department, ever solicitous of maintaining friendships in Langley and Fort Meade, promptly dismissed Feinstein’s request for a criminal investigation of the CIA’s breach of trust, with the result that whether the CIA broke the law remained a matter of conjecture. The committee, thitherto led by cheerleaders for the CIA and the NSA, itself did nothing to fill the void. It had failed earlier to learn that the CIA ran secret prisons, waterboarded prisoners, made videotapes of the waterboarding or—after it found out—destroyed the videotapes. It had failed to learn how the administration used the phone records of American citizens that the NSA collected, or that Angela Merkel’s cell phone was being tapped—and a host of other embarrassments (many publicly revealed by Edward Snowden) that a competent oversight committee would have caught. The committee’s leadership had little to gain by focusing further public scrutiny on its own omissions and indifference to Clapper’s and Brennan’s deceit. Even some defenders of NSA surveil­lance acknowledged that the oversight committees could not be trusted. “Clearly, they’ve been co-opted,” said McCain. “There’s no doubt about that.”

The courts joined the committee in behaving as an annex of the military/intelligence community. The rubber-stamp record of the Foreign Intelligence Surveillance Court—the closest thing the nation has to a national-security court—in approving warrant requests has made it the butt of jokes. But its lamentable history is not unique. At the time of Clapper’s statement, it was well nigh impossible to find a single case in which anyone claiming to have suffered even the gravest injury as the result of the U.S. government’s counterterrorism policies had recovered a dime in damages. In fact, it is still hard to find any case in which any plaintiff has even been allowed to litigate any counterterrorism claim on the merits. Challenges have been regularly dismissed before any plaintiff has had a chance to describe what happened either before the courts or, often more important, the court of public opinion.

The system’s failure, then, has been far more than a failure of the truth-finding process, or even a failure to prevent torture; its failure has been nothing less than a collapse of the equilibrium of power, the balance expected to result from ambition set against ambition, the resistance to encroachment that was supposed to keep the three branches of the federal government in a state of equilibrium and to protect the people from the government.

How could this have happened?

MUCH OF the answer can be found in Walter Bagehot’s theory of the British government. He presented it in the 1860s to explain the evolution of the country’s political system. While not without critics, his theory has been widely acclaimed and has generated significant com­mentary. Indeed, it is something of a classic on the subject of institutional change, and it foreshadowed modern organi­zational theory. Bagehot’s view went something like this:

Power in Britain reposed initially in the monarch alone. Over the decades, however, a dual set of institutions emerged. One set comprises the monarchy and the House of Lords. These Bagehot called the “dignified” institutions—dignified in the sense that they provide a link to the past and excite the public imagination. Through theatrical show, pomp and historical symbolism, they exercise an emotional hold on the public mind by evoking the grandeur of ages past. They embody memo­ries of greatness. Yet it is a second, newer set of institutions—Britain’s “efficient” institutions—that do the real work of governing. These are the House of Commons, the cabinet and the prime minister. As Bagehot put it:

Its dignified parts are very complicated and somewhat imposing, very old and rather venerable; while its efficient part . . . is decidedly simple and rather modern. . . . Its essence is strong with the strength of modern simplicity; its exterior is august with the Gothic gran­deur of a more imposing age.

Together these institutions make up a “disguised republic” that obscures the massive shift in power that has occurred, which, if widely understood, would create a crisis of public confidence. This crisis has been averted because the efficient institutions have been careful to hide where they begin and where the dignified institutions end. They do this, Bagehot suggested, by ensuring that the dignified institutions continue to partake in at least some real governance and that the efficient institutions partake in at least some inspiring public ceremony and ritual. This promotes continued public deference to the efficient institutions’ decisions and continued belief that the dignified institutions retain real power. These dual institu­tions, one for show and the other for real, afford Britain exper­tise and experience in the actual art of governing while at the same time providing a façade that generates public acceptance of the experts’ decisions. Bagehot called this Britain’s “double government.” The structural duality, some have suggested, is a modern reification of the “Noble Lie” that, two millennia before, Plato had thought necessary to insulate a state from the fatal excesses of democracy and to ensure deference to a class of efficient guardians.

Bagehot’s theory may have overstated the naïveté of Britain’s citizenry. When he wrote, probably few Britons believed that Queen Victoria actually governed. Nor is it likely that the country’s prime ministers, let alone 658 members of the House of Commons, could or did consciously and intentionally conceal from the British public that it was really they who governed. Big groups keep big secrets poorly. Nonetheless, Bagehot’s endur­ing insight—that dual institutions of governance, one public and the other concealed, work side by side to maximize both legitimacy and efficiency—is worth pondering as one possible explanation. There is no reason in prin­ciple why the institutions of Britain’s juridical offspring, the United States, ought to be immune from the broader bifurcat­ing forces that have driven British institutional evolution.

As it did in the early days of Britain’s monarchy, power in the United States lay initially in one set of institutions—the presi­dency, Congress and the courts. These are America’s “dignified” institutions. Later, however, a second institution emerged to safeguard the nation’s security. This, America’s “efficient” insti­tution (actually, more a network than an institu­tion), consists of the several hundred executive officials who sit atop the military, intelligence, diplomatic and law-enforcement departments and agencies that have as their mission the protection of America’s security. Large segments of the public continue to believe that America’s con­stitutionally established, dignified institutions are the locus of governmental power. That belief allows both sets of institutions to maintain public support and legitimacy. Enough exceptions exist to sustain that illusion. But when it comes to defining and protecting national security, the public’s impres­sion is mistaken. America’s efficient institution makes most of the key decisions concerning national security, removed from public view and from the electoral and constitutional restrictions that check America’s dignified institutions. The United States has, in short, moved beyond a mere imperial presidency to a bifurcated system—a structure of double government—in which even the president now exercises little substantive control over the general direction of U.S. national-security policy. Whereas Britain’s dual institutions evolved toward a concealed republic, America’s have evolved in the opposite direction, toward greater central­ization, less accountability and emergent autocracy.

The birth date of Britain’s efficient institution is difficult to determine, having evolved over time. America’s did not. President Harry S Truman, more than any other president, is responsible for creating the nation’s “efficient” national-security apparatus. Under him, Congress enacted the National Security Act of 1947, which unified the military under a new secretary of defense, set up the CIA, created the modern Joint Chiefs of Staff and established the National Security Council (NSC). Truman also set up the NSA, which was intended at the time to monitor communications abroad. Friends as well as detractors viewed Truman’s role as decisive. Honoring Truman’s founding role, let us substitute “Trumanite” for “efficient,” referring to the network of sev­eral hundred high-level military, intelligence, diplomatic and law-enforcement officials within the executive branch who are responsible for making national-security policy.

TRUMAN’S NATIONAL-SECURITY initiatives were contro­versial, with liberal and conservative positions in the debate curiously inverted from those prevalent in current times. In the late 1940s and early 1950s, congressional liberals generally supported Truman’s efforts to create more centralized national-security institutions on the theory, held by many and summa­rized by Michael Hogan, that “peace and freedom were indivisible, that American power had to be mobilized on behalf of democracy ‘everywhere,’ and that tradition had to give some ground to this new responsibility.” Senator Hubert Humphrey of Minnesota, for example, dismissed objections to the consti­tutionality of the new arrangements: “It is one thing to have legalistic arguments about where the power rests,” he said, but another to straitjacket a president in trying to deal with a totalitarian state capable of swift action. Stalin could strike a deathblow at any time, he argued; as a result, “those days of all the niceties and formalities of declarations of war are past.” Under these conditions, “it is hard to tell . . . where war begins or where it ends.” Senator Paul Douglas of Illinois insisted that U.S. military power should support democracy “everywhere.” Unanswered aggression would lead only to further aggression, he suggested, requiring the United States to move to a posture of permanent military preparedness.

Congressional conservatives, by contrast, feared that Truman’s ballooning national-security payrolls, reliance upon military solutions to tackle international problems and efforts to centralize decision making posed a threat to democratic institutions and the principle of civilian leader­ship. Republican senator Edward V. Robertson of Wyoming, for example, worried that Truman’s military consolidations could amount to the creation of an “embryonic” general staff simi­lar to that of Germany’s Wehrmacht. A new national intelli­gence agency, he said, could grow into an American “gestapo.” Republican senator William Langer of North Dakota and his allies believed that the Soviet threat was exaggerated; in their eyes, the real enemy was the Pentagon, where “military lead­ers had an insatiable appetite for more money, more men, and more power, whatever the cost to democracy.” The con­servatives invoked the specter of a “garrison state,” a “police state” and a “slave state” run by “power-grabbing bureau­crats.” They saw peacetime military conscription as “aping the military clique of Hitler” and leading to a “complete mili­tarization of the country,” creating a “permanent military caste.” Republican congresswoman Katherine St. George of New York, recalling George Washington’s Farewell Address, foresaw the possibility of military domination of the nation’s civilian leadership. Republican senators John Bricker and Robert Taft of Ohio and Homer Capehart of Indiana voted to cap the size of active U.S. military forces in part to halt what they regarded as “a drift from ‘congressional responsibility’ to ‘administra­tive policymaking’ . . . which would destroy the ‘liberty of the people.’” “The truth is that we are slowly losing our freedoms as we move toward the garrison state,” said the Republican leader of the House of Representatives, Joseph W. Martin of Massachusetts.

Truman himself appeared to share these concerns, at least to an extent. He was “very strongly anti-FBI,” according to his aide Clark Clifford. Truman was “afraid of a ‘Gestapo’” and wanted to “hold [the] FBI down.” Although a military officer would be permitted to head the CIA, Truman accepted the proviso in the National Security Act described above, under which the agency would be prohibited from performing any police or law-enforcement functions. As for the military, while waste­ful duplication had to be eliminated and better coordination established, Truman feared that collective deliberation could force the president to share responsibility and decision-making power, resulting in a diminution in presidential authority and a weakening of civilian control over the military. With half of the members of the new NSC coming from the military, Truman believed it would be difficult for the president to ignore their recommendations, even though their counsel was only advisory. Truman was particularly annoyed by interser­vice rivalries and pressure from military lobbyists to increase their services’ budgets. “We must be very careful that the mili­tary does not overstep the bounds from an economic standpoint domestically,” he wrote. He also believed that “most of them would like to go back to a war footing.” But he considered the new national-security apparatus necessary to rein in the mili­tary as well as to improve the United States’ ability to respond to the looming (though exaggerated) Soviet threat. The Hoover Commission warned in 1949 that the Joint Chiefs had come to act as “virtually a law unto themselves” and that “central­ized civilian control scarcely exists” in certain military depart­ments. Internecine warfare among the services had come to undermine the nation’s defense. Truman believed that his new national-security architecture was the best bet to bolster the capacity of the nation to meet security threats while safeguard­ing the democratic institutions that the newly empowered mili­tary and intelligence organizations were expected to protect.

SIXTY YEARS later, sitting atop its national-security institutions, an intragovernmental network that has descended from what Truman created now manages the real work of protecting the nation’s security. Its members are smart, hard-working, public-spirited officials, careerists as well as in-and-outers. They exercise their authority not because of some vast, nefarious conspiracy, but rather as the result of structural incentives embedded deeply within the American political system. They define security primarily in military terms and tend to consider military options before political, diplomatic or law-enforcement alternatives for an understandable reason: relative to other governmental agencies, the American military is extremely proficient and widely respected. They share the premise of Madeleine Albright’s famous question to Colin Powell: “What’s the point of having this superb military . . . if we can’t use it?” They also favor existing policies over new, different ones, in part because senior officials—their bosses—were their authors. In economic terms, their programs are “sticky down”—much more difficult to end than to expand or to continue.

This basic dynamic, well known to organizational behaviorists, represents the principal reason that U.S. national-security policy has changed so little from the George W. Bush to the Obama administration. As a candidate for president, Obama repeatedly, forcefully and eloquently promised fundamental change in that policy. It never happened. U.S. policies on rendition, covert operations, cyberwar, military detention without trial or counsel, drone strikes, NSA surveillance, whistle-blower prosecutions, nonprosecution of waterboarders, reliance on the state-secrets privilege and a variety of other national-security issues all have remained largely the same. The explanation lies not simply in the huge number of holdovers in high-level policy-making positions; the reality is that structural incentives have given these policies a life of their own—allowing them to run “on autopilot,” as Secretary of State John Kerry described one NSA program, largely immune from constitutional and electoral restraints.

A variety of legislative and judicial reforms have been suggested, aimed generally at restoring a semblance of institutional balance. Given the prevailing incentive structure, however, none are likely to succeed. The first difficulty with the proposed reforms is circularity. All rely upon the Madisonian institutions—Congress, the courts and the presidency—to restore power to the Madisonian institutions by exercising the very power that the Madisonian institutions lack. All assume that the Madisonian institutions, in which all reform proposals must necessarily originate, can somehow magically impose those reforms upon the Trumanite network or that the network will somehow merrily acquiesce. All sup­pose that the forces that gave rise to the Trumanite network can simply be ignored. All assume, at bottom, that Madison’s scheme can be made to work—that an equilibrium of power can be restored—without regard to the root cause of the disequilibrium.

That root cause is difficult to discuss in a democracy, for it lies in the electorate’s own deficiencies. This is the second great obstacle the reform proposals confront; on this point Bagehot’s and Madison’s theories converge. Bagehot argued that when the public becomes too sophisticated to be misled any longer about who holds governmental power but not informed enough to play a genuine role in governance, the whole structure will “fall to the earth,” in his phrase. Madison, contrary to popular belief, did not suggest that the system that he and his colleagues designed was self-correcting. The Framers did not believe that merely setting “ambition against ambition” within the government would by itself save the people from autocracy. They believed that this competition for power would not occur absent an informed and engaged public—what Robert Dahl has called the “adequate citizen,” the citizen able and willing to undertake the responsibilities required to make democracy work. Thomas Jefferson spoke for many of the Framers. He said: “If a nation expects to be ignorant and free, in a state of civilization, it expects what never was and never will be.” Competition between institutions was thus written into the constitutional architecture not as a substitute for civic virtue—there is none—but as a backstop, as an additional safeguard to forestall the rise of autocracy. But that backstop was not freestanding: it, too, depended upon an electorate possessed of civic virtue.

If anything, the essentiality of civic virtue has grown over the years. In the early days of the Republic, public-policy issues were less intricate, and the franchise was de jure or de facto more restricted. A smaller electorate was more capable of mastering the more straightforward issues it faced. As Louis Henkin pointed out, however, the United States has since changed gradually from a republic to a democracy—an “ultra-democracy,” Bagehot believed. The problems government has faced over the years have become more complex, and a greater base of civic knowledge has thus become indispensable for responsible participation in the process of governance. Yet a cursory glance at consistent survey results confirms what former Supreme Court justice David Souter has described today as the public’s “pervasive civic ignorance.”

The numbers are sobering. A 2011 Newsweek survey showed that 80 percent of Americans did not know who was president during World War I; 40 percent did not know whom the United States fought in World War II; and 29 percent could not identify the current vice president of the United States. Far more Americans can name the Three Stooges than any member of the Supreme Court. One poll has found that 71 percent of Americans believe that Iran already has nuclear weapons. In 2006, at the height of U.S. military involvement in the region, 88 percent of Americans aged eighteen to twenty-four could not find Afghanistan on a map of Asia, and 63 percent could not find Iraq or Saudi Arabia on a map of the Middle East. Ilya Somin’s fine book Democracy and Political Ignorance analyzes the problem in depth. The great conundrum is that the public’s ignorance does not derive from “stupidity”—average raw IQ scores actually have increased in recent decades—so much as it derives from simple rationality: Why spend time and energy learning about national-security policies that cannot be changed?

That is the nub of the negative feedback loop in which the United States is now locked. Resuscitating the Madisonian institutions requires an informed, engaged electorate, but vot­ers have little incentive to be informed or engaged if they believe that their efforts would be for naught—and as they become more uninformed and unengaged, they have all the more reason to continue on that path. The Madisonian institutions thus continue to atrophy, the power of the Trumanite network continues to grow and the public continues to disengage.

Should this trend continue, and there is scant reason to believe it will not, it takes no great prescience to see what lies ahead: outward symbols and rituals of national-security governance that appear largely the same, concealing a Trumanite network that takes on the role of a silent directorate, and Madisonian institutions that, like the British monarchy and House of Lords, quietly and gradually are transformed into museum pieces.

Michael J. Glennon is professor of international law at Tufts University’s Fletcher School of Law and Diplomacy. This essay is adapted from his book National Security and Double Government (Oxford University Press, 2014) and a January 2014 article of the same title in the Harvard National Security Journal.