If nothing else, the presidency of George W. Bush will be recorded by historians as a heightening of the “imperial presidency,” particularly in the realm of national security. Over the past seven years, the White House has been remarkably successful in seizing near-complete control of the military, security, and intelligence apparatus of the federal government, a turn that has raised significant concern among civil libertarians of all political stripes. As Al Gore said in a 2006 speech, “The American values we hold most dear have been placed at serious risk by the unprecedented claims of the Administration to a truly breathtaking expansion of executive power.”

But Bush and Vice President Dick Cheney cannot claim all the credit, or be tagged with all the blame, for the executive branch’s current ascendance. In fact, its rise has a long and complex history an inconvenient fact that neither supporters nor detractors of the Administration like to dwell upon. After all, if Bush and Cheney are the problem, then removing them from office is the cure–not to mention a powerful electoral motivator. Yet the notion that the imperial presidency will vanish on January 20, 2009, is both unfounded and hazardously naïve.

To be sure, Bush and Cheney have aggressively pushed the envelope on all elements of executive power. Going back to his time in the Ford Administration, Cheney has nurtured a belief in the supremacy of the Oval Office–recall his vigorous and ultimately successful defense of absolute secrecy for his energy task force, well before September 11. After the attacks, however, opportunities for the exercise of the “monarchical prerogatives” (as Cheney’s staff wrote in the Iran-Contra minority report, published while he was in Congress) multiplied, with the Administration adopting policies of torture, warrantless surveillance, and indefinite offshore detention. Even as legislators were negotiating carefully defined new statutory surveillance and detention authorities in the USA Patriot Act, a team of lawyers in the White House and the Justice Department were assembling justifications for open-ended, warrantless surveillance and unlimited detention at Guantánamo Bay. Both the process that yielded these policies and their results have elicited harsh condemnation from a range of critics, from Nadine Strossen of the American Civil Liberties Union (ACLU) and Bob Herbert of the New York Times on the left to Nixon White House counsel John Dean and former Republican congressman Bob Barr on the right.

Yet missing from many of these critiques has been an exploration and appreciation of the broader trends that created the conditions for a heightening of the imperial presidency–most notably, changes in the size, managerial complexity, and legal instrumentation of the executive branch’s national security powers. None of these trends is wholly new, and noticing them is hardly an insight. Political scientists and scholars from Arthur Schlesinger Jr.–who popularized the term “imperial presidency” in his 1973 book of the same name–onward have charted the executive branch’s growth over the past 70 years. Historians like Andrew Bacevich in The New American Militarism and James Carroll in House of War have eloquently highlighted, in particular, the growth of a national security bureaucracy. Equally, scholars such as Andrew Rudalevige in The New Imperial Presidency and John Burke in The Institutional Presidency have mapped the increasing managerial and organizational capacity of the White House. And most recently, Philip J. Cooper in By Order of the President and Kenneth Mayer in With the Stroke of a Pen have highlighted the growing salience of the president’s unilateralist law-making tools. Nevertheless, what is still missing is an account of how these different developments combined to fuel the unprecedented growth of the national security presidency in terms of sheer scale, the centralization of its powers, and the proliferation of executive orders and other unchecked executive tools. Along the way, each of these developments has met little or no resistance from Congress.

No matter what the next president thinks about the war on terror, we should not expect the underlying balance of power between the branches to change overnight. Nor should we expect presidential incentives and opportunities to shift radically: Power is tough to give up, especially when, from an executive’s risk-averse perspective, it makes for very efficient policymaking in our interdependent, terror-ridden world. Yet there are measures that Congress can take to reassert its power and its rightful role in our constitutional system. And no matter who becomes the next president, progressives must be committed to reining in the virtually unchecked power of the national security presidency.

A Different Beast

During the twentieth century, the executive branch of the federal government grew in ways the Framers could not have predicted. In 1830, the federal government had roughly 11,000 employees. In 1930, it had 608,915 employees, and by 2004, it had 2,649,319. Growth in the national security state has been even more marked. During World War II, General William “Wild Bill” Donovan directed the Office of Strategic Services (OSS), the nation’s wartime intelligence service, which even at its acme never had more than 13,000 members. In November 1944, almost six months before the fall of Berlin, Donovan proposed to President Franklin D. Roosevelt that the United States create a permanent “Central Intelligence Service.” Roosevelt turned him down, fearing public animosity (the plan, noted the press, had all “the earmarks of a Gestapo”).

But by 2005, less than 60 years later, the federal government was spending $44 billion on its 16 permanent intelligence agencies and their combined staff of more than 100,000. Six intelligence components produce intelligence; six collect it; three collect and build it; and nine components, agencies, and sets of officials use it. At its founding, the CIA was so starved for cash that it resorted to skimming money from the Marshall Plan. Now it has a secret budget running into the billions, used not just to fund operations but also to suborn other countries’ intelligence services into evasions of federal law. Whatever the merits of its policy justifications, this transformation’s constitutional consequence–a shift in power toward the executive branch, which oversees the combined apparatus–is clear.

This growth in intelligence spending since 1946 has been largely driven by technological change, which in turn has vested the executive branch with a new kind of power. The emergence of large, private databanks holding individual financial and transactional histories, for example, has provided a new, easily accessed tool for national security agencies to pry into individuals’ lives, and an incentive to develop computational resources and data-mining algorithms to leverage that possibility. Telecommunication’s reliance on economies of scale and network effects–you couldn’t build the Internet for one person alone–gives the government unprecedented opportunities to piggy-back on private initiatives. For example, it plugs into the massive telecommunications routers that carry most of our email and telephone calls in order to vacuum out huge swathes of our correspondence. Technology, in short, adds a qualitative dimension to the quantitative change in executive branch scale.

Congress has failed to respond adequately to these monumental shifts in executive power. As Fritz Schwarz and I have discussed at length elsewhere, a Senate investigation in the mid-1970s, led by Senator Frank Church, yielded a comprehensive accounting of intelligence abuses in the Cold War era and an equally capacious list of reform recommendations. But limited political capital meant that while congressional oversight committees were strengthened and a federal statute against warrantless wiretapping was enacted, too many of the Church Committee’s necessary reforms were left on the table. Moreover, such periodic bouts of attention could not compensate for Congress’s longer, and more significant, failures with respect to national security powers. Three, in particular, merit special scrutiny.

First, Congress has never provided clear ex ante limits on intelligence authorities, as it did with the rest of the federal bureaucracy. In the 1930s, recognizing the impossibility of direct management of the vast post–New Deal regulatory state, Congress enacted a range of framework statutes to set agencies’ goals and establish checks and balances within them, essentially fire alarms and watchdogs to signal to Congress when something went awry. These laws set the ground for new federal bureaucracies, but also carefully defined and limited bureaucratic mandates for agencies such as the Department of Labor, the Federal Communications Commission, and the Food and Drug Administration. By contrast, the 1947 National Security Act conjured up the CIA in six terse and uninformative paragraphs. From its inception, the CIA pushed the envelope of its legal authority, plunging early on, for example, into covert, psychological operations against leftist political parties in democratic Western Europe.

Second, Congress failed to create intradepartmental oversight mechanisms, as it did in other parts of the postwar regulatory state. The Administrative Procedure Act of 1946, for example, not only acknowledged administrative agencies’ large discretion, but it also imposed ongoing oversight constraints: Most importantly, it allowed private parties to intervene in agency rulemaking and sue to prevent “arbitrary or capricious” agency action. In subsequent decades, Congress added open government laws, such as the 1966 Freedom of Information Act, the 1974 Privacy Act, the 1976 Government in the Sunshine Act, and the 1972 Federal Advisory Committee Act.

But this infrastructure of legislative constraint largely bypassed the national security agencies. Private parties cannot easily challenge the rules and decisions of national security agencies, as they can when other agencies act. Because Congress has allowed exceptions to disclosure and privacy laws for security agencies, it is hard even to know what to challenge. Exceptions from information- and procedure-forcing laws, moreover, are compounded by judicial quiescence, rendering existing oversight mechanisms toothless. Under the Supreme Court’s scalpel, for example, parts of the Freedom of Information Act that might have forced some public oversight of CIA activity have been nullified. And groups that monitor intelligence agencies, such as ACLU and the Center for National Security Studies, too often lack the statutory levers available to groups that track, say, the vagaries of the Environmental Protection Agency or the Food and Drug Administration.

Finally, oversight of security agencies by Congress itself has been weak to nonexistent, with the exception of moments such as the Church Committee and the 9/11 Commission. Reflecting on Cold War intelligence practices in 1976, former CIA Director William Colby commented, “The old tradition was that you don’t ask. It was a consensus that intelligence was apart from the rules.” Congress neither knew about nor cared to know about assaults on elected governments in Guatemala and Iran. At home, the CIA started a letter-opening program targeting hundreds of people, from the members of Student Nonviolent Coordinating Committee to Richard Nixon. At the behest of the Johnson White House, it began a domestic spying program, “Operation Chaos,” to track antiwar groups and root out foreign influences. Even on these domestic matters, historian David Barrett writes, “Congress deferred to presidents and leaders of the CIA” without knowing what the Agency was doing.

Oversight failure has been fairly constant regardless of whether we have divided government or the same party in the White House and in command on Capitol Hill. The last five years have been instructive on this count. From 2002 to 2006, Congress enacted one measure with truly significant civil liberties consequences in the national security arena–the September 2006 Military Commissions Act, which complemented a December 2005 effort to deny detainees in Guantánamo habeas corpus, the traditional judicial remedy for unlawful detention. Yet even after the return of divided government in January 2007, neither house of Congress has been able to pass habeas restoration legislation. In fact, in August 2007, both (Democratically controlled) houses rushed to enact the Protect America Act, which created a categorical exemption to the surveillance laws and guaranteed the continuance of open-ended surveillance. Legislative oversight and constraint, in short, are not much more in evidence today than they were in 2004–or even 1954.

The President in His Labyrinth

As the executive branch grew in size and power, it also became more centralized, not only in national security areas, but in all its endeavors. This is the second significant trend that laid the groundwork for the contemporary imperial presidency. And, like the growth in sheer executive power, it too has been abetted by congressional errors and apathy.

Through statutes and post–New Deal administrative devices, the White House over the decades has fashioned tools to exercise direct and immediate control over a far-flung federal bureaucracy. What began as a worthy effort at efficiency, however, has evolved into a means to steal a march on Congress by formulating and pushing into action public policies without legislative sanction. Under the current president, these tools have been superseded by an even more centralized form of decision-making: a small cadre of vice-presidential intimates making critical decisions about national security, often without the knowledge or participation of cabinet members. Early on, for example, they developed and implemented the idea of military commissions standing outside the purview of American justice. Nevertheless, this coterie did not achieve this status overnight; the “Cheney cabinet” has been a long time coming, and the centralization of power in the Executive Office of the President has been a consistent, bipartisan development.

Presidents have not always had labyrinthine groves of policy and communications staff. It was not until 1939–well into the New Deal–that the Executive Office of the President was created and the president had the authority from Congress to hire new staff. Subsequent presidencies enlarged the White House’s capacity. Nixon, for example, saw the larger federal bureaucracy as his enemy and fashioned a “counter-presidency” inside the White House with more than double the staff of his predecessor. Reagan built on this legacy by centralizing the process of agency rulemaking, using the Office of Information and Regulatory Affairs within the White House to ratchet up the Oval Office’s influence.

President Clinton circumvented an unresponsive Congress by issuing formal, published memoranda providing specific direction about the goals, means of implementation, and enforcement strategies for agency action. During his first term Clinton promulgated detailed instructions to the Treasury Department stopping the import of foreign-made assault pistols and improving enforcement of licensing laws. In his second term, he imposed new notification rules and a moratorium on certain assault weapons. All of this was done without congressional action. In 2001, Harvard Law School Dean Elena Kagan could describe, without contradiction, the ascendance of “presidential administration,” whereby presidents had “personal ownership” of the activities of federal agencies and could use this ownership to impose policy change across the federal policy spectrum, from health care and welfare reform to gun control and civil rights. Today, personal presidential ownership of policy-making extends to the sanction of counterterrorism policies such as extraordinary rendition and warrantless surveillance. From one angle, the Bush Administration’s free-wheeling unilateralism when it comes to interrogation and detention is merely the dark side of Clinton’s exuberant, and often celebrated, unilateral use of executive agencies.

As the president’s toolkit grew, Congress struggled and failed to find effective devices for the oversight and containment of executive power. Until 1983, it used “legislative vetoes” to disapprove specific executive actions through bicameral votes which did not require presidential presentment. Legislative vetoes were never terribly effective tools; they were mostly used to reverse decisions to suspend deportation of non-citizens. But after 1983, the Supreme Court denied Congress even this power by holding the legislative veto unconstitutional. In 1996, a legislative effort to craft a new means for congressional review of agency action foundered. At the same time, Congress gradually ceded control of the budgeting process to the White House through statutes and sheer disorganization, losing yet another oversight tool. The present Democratic majority’s inability to use Congress’s unequivocal power of the purse to constrain the Iraq adventure, in spite of public support, merely illustrates this larger collapse of legislative control of the federal fiscal agenda.

On national security matters, the decline of legislative authority, and the extent of congressional complicity, is even starker. The National Security Council (NSC) has not only made foreign-policy making more efficient, it also has been used, along with other White House devices, to wrest authority from Congress. The current president has taken this centralization of power even further. Statutory mechanisms of presidential control of national security policy have been supplanted by an even more centralized process of informal decision-making that lacks legislative imprimatur. This new system bypasses the NSC and the other mechanisms designed to bring control into the White House, and instead concentrates power in the vice president’s office. Hence, when Senator Bob Graham, as the new head of the Intelligence Oversight Committee, traveled to the White House in October 2001 to meet with the President, he was told by Bush that Cheney “has the portfolio for intelligence activities.” Indeed, many of the most important national security decisions of the past six years emerged from the vice president’s office. Mere months after September 11, for example, the vice president secured presidential sign-off on an executive order establishing military commissions, leaving both the national security advisor and the secretary of state out of the loop. White House debates on the Iraq war, former Secretary of State Colin Powell has said, were characterized by the same dynamic. Essential facts and figures were ignored, and the views of key actors were routinely misrepresented, willfully or not, by the small group of people most often alone with the president–i.e., Cheney and Secretary of Defense Donald Rumsfeld.

Cheney, in effect, boiled down the national security policy process to a mere handful of White House decisionmakers. But without the changes wrought by from FDR to Clinton’s innovative presidential administration, he could not have assumed this degree of concentrated control. Sixty years of executive centralization created the foundation for this, the non plus ultra of executive unilateralism.

Executive Tools

A growing and centralizing executive has been further empowered by an expanded set of executive law-making tools. National security policy today is almost always achieved via unilateral executive orders from the White House. And when Congress wants to push back against the executive, devices in the Constitution initially designed to disperse political power, such as the veto, now perversely serve to concentrate power in the executive branch.

Presidents initially used executive orders to convey instructions inside the executive branch. However, as the New Deal progressed and the federal government swelled, presidents increasingly came to rely on them for a broad array of matters. Between 1920 and 1998, presidents issued more than 10,000 unilateral directives. Across the board, the use of unilateral executive orders picked up under Clinton. By one count, Reagan issued nine directives on domestic policy in his eight years in office, George H.W. Bush issued four in his single term–and Clinton issued 107.

But the executive-order presidency is most noteworthy, and troublesome, in the national security arena. In fact, the history of post–World War II security policy can be told in executive orders, beginning with the landmark 1950 NSC-68 that defined the Cold War paradigm. According to the Government Accountability Office, White Houses between 1961 and 1988 issued 1,042 presidential directives on “national security,” from telecommunications security to Soviet immigration, from sealifts to the structure of the NSC. Less than a quarter are published, although every one is binding on the entire executive branch. Each White House also has its own name for these security directives Presidential Directives, National Security Directives, National Security Policy Directives, or the like. Such orders are not only difficult to find (they are classified), they also remain opaque to public comment and criticism. It is difficult to get worked up about something that does not even have a consistent name.

Secret, unilateral executive directives issued without congressional approval are in the background of the most controversial of the Bush Administration’s post-9/11 security policies. On September 17, 2001, Bush signed a secret presidential finding that allowed the CIA to hold and interrogate suspects; the finding did not require the CIA, when detaining and transferring suspects, to seek case-by-case approval from the White House, the State Department, or the Justice Department. It also released vast new funds to coax foreign intelligence services into cooperation with the CIA. The order led to the establishment of a global network of secret CIA prisons (known as “black sites”) and “outsourced” detentions by unscrupulous secret services across the Middle East, North Africa, and Asia. This global detention system, which still persists alongside more notorious facilities such as the Guantánamo Bay Naval Base, violates numerous human rights conventions and customary international law prohibitions that bind the United States, not to mention domestic criminal prohibitions concerning torture and conspiracy to abet torture. While it costs taxpayers millions (or billions) per year, there is no practical means to ascertain whether it is any more effective than the Bush Administration’s other reckless uses of military might overseas.

Simply put, secret national security orders dramatically change the balance of constitutional power. The Constitution envisages a scheme whereby Congress, acting first, sets the agenda, and the president can respond via exercise of the veto. When the president acts first with an executive order, however, Congress no longer sets the agenda. Rather, it responds to the status quo preference of the executive branch, and must muster supermajorities in both houses to overcome an inevitable veto.

Congress has accordingly failed to find the political resources to respond. Bicameral, veto-proof supermajorities are rare in an era in which, as Thomas Mann and Norman Ornstein have noted, “The majority party, including the leaders of Congress, see themselves as field lieutenants in the president’s army far more than they do as members of a separate and independent branch of government.” This is especially true on national security issues, where a president can use legislative opposition as an opening for partisan critique come election time. Today, it is the executive that makes the law–especially when it comes to national security–and not Congress.

Rebalancing the Branches

Realizing that the damage to the Separation of Powers doctrine is not the handiwork of Bush or Cheney alone is the first step to wisdom. Like the causes, the solutions to imbalanced government are structural, rather than a matter of partisan politics. The national security state is not going to wither any time soon, and the systemic advantages of the executive are unlikely to change. Nor, given the very real national security threats, is it wholly a bad thing. Rather, resolving the imbalance in the Separation of Powers doctrine requires the installation of new accountability mechanisms, not wholesale reconsideration of the national security state. It means taking the principle of checks and balances and applying it assiduously to the executive, instead of recklessly chopping away at its powers.

While this task is a large one, it has a few straightforward principles. First, there is a need to craft new information-forcing mechanisms that address informational asymmetries between the branches and prevent the executive from leveraging unsubstantiated successes to consolidate its authority. As a result of technological changes, increasing administrative sophistication, and the use of secret executive orders, the White House maintains critical information advantages over the other branches. In the past few years, it has moved to consolidate that advantage through aggressive innovations in the use of “executive privilege,” with officials claiming immunity not only with respect to specific information, but also with respect to the duty to testify at all before Congress. Congress can, however, respond through measures that force information disclosures and impose penalties both for refusing to disclose and also for failing to disclose in a timely fashion. It could reinforce statutory disclosure obligations with respect to intelligence activities, with heightened rules for detention and interrogation practices that have recently produced the lion’s share of American human rights abuses. By providing a clear and enforceable framework for classification rules in a statute, Congress could eliminate implausible, but frequently used, claims to secrecy that hinder public revelation of executive wrongdoing. And through a statute on executive privilege, Congress could preempt the White House’s inclination to make every dispute over access to information into a constitutional issue.

Second, restoration of the Separation of Powers doctrine demands a gamut of new devices to ensure accountability within the executive branch–what Georgetown Law Professor Neal Katyal has called an “internal checks and balances.” This would range from a stronger system of inspectors general, the statutory office responsible for internal auditing of executive branch activity; to better protection for whistleblowers; to clarification of the ethical rules that cover government lawyers, who have an obligation to the Constitution as well as to the sitting president (whatever Alberto Gonzales might have thought). Legislation currently pending on the Hill to strengthen the independence and investigative reach of inspector generals is a start. That bill should be followed by new statutes for special prosecutors who focus on the abuse and self dealing in national security powers. A new framework for special prosecutors could be modeled on existing internal Justice Department guidelines for the Independent Counsel statute, which have been largely successful in avoiding the incentive structures that earlier enabled abuse of the statute. Congress should also deepen whistleblower protections that have been eroded during the current administration.

Finally, if Gonzales’s tenure does indeed have something to teach the nation, it is the importance of transparency in the operation of the executive branch legal institutions. Executive orders and internal legal opinions have been instrumental in entrenching executive power at the expense of the other branches and the ability of the public to assess meaningfully what is done in its name. These need to be brought to public light–which is one measure Congress could push for even in the absence of legislation. Legal opinions or executive orders that invoke Article II powers to curtail or negate the effect of a federal statute, in particular, should be presumptively public, even if they concern national security matters. This entails, at a minimum, releasing currently classified opinions on enhanced interrogation tactics, NSA surveillance, and extraordinary rendition. And in the future, if an attorney general adopted the position implied in confirmation hearings by Judge Michael Mukasey–i.e., that torture could not be constitutionally prohibited–it would mean releasing that opinion and airing that decision to the plenary public and congressional scrutiny and criticism.

We should be under no illusion that the political will to make these changes will be easy to muster, or that an increased flow in information would necessarily improve the quality of oversight and the balance between the branches. As recent debates on the Protect America Act and RESTORE Act of 2007 make clear, there is scant legislative appetite for dealing with the Administration’s often-dubious claims to act in the name of national security, let alone for accounting for the recent redistribution of constitutional power. But without an honest look at the state of constitutional play today, without an understanding of the historical connections between growing executive power, presidential administration, and executive law-making, there is little hope that reform will even be considered–and that the very essence of our Constitution can be redeemed.