Illustration by Steve Brodner
For the past eight years, George W. Bush has treated the White House much as Kenneth Grahame’s Mr. Toad treated a new automobile—like a shiny toy to be wrecked by racing the motor, spinning smoke from the tires, and smashing through farmyards until the wheels come off. Bush got to the Oval Office despite having lost the popular vote, and he governed with a fine disdain for democratic and legal norms—stonewalling congressional oversight; detaining foreigners and U.S. citizens on his “inherent authority”; using the Justice Department as a political cudgel; ordering officials to ignore statutes and treaties that he found inconvenient; and persisting in actions, such as the Iraq War, that had come to be deeply unpopular in Congress and on Main Street.
Time capsule: The Bush Files
Readers are invited to submit mementos—from quotes to videos to books to movies—that best capture the essence of the Bush years.
Understandably, most Americans today are primarily concerned with whether Barack Obama can clean up Bush’s mess. But as Bush leaves the White House, it’s worth asking why he was able to behave so badly for so long without being stopped by the Constitution’s famous “checks and balances.” Some of the problems with the Bush administration, in fact, have their source not in Bush’s leadership style but in the constitutional design of the presidency. Unless these problems are fixed, it will only be a matter of time before another hot-rodder gets hold of the keys and damages the country further.
The historian Jack N. Rakove has written, “The creation of the presidency was [the Framers’] most creative act.” That may be true, but it wasn’t their best work. The Framers were designing something the modern world had never seen—a republican chief executive who would owe his power to the people rather than to heredity or brute force. The wonder is not that they got so much wrong, but that they got anything right at all.
According to James Madison’s Notes of Debates in the Federal Convention of 1787, the executive received surprisingly little attention at the Constitutional Convention in Philadelphia. Debate over the creation and workings of the new Congress was long and lively; the presidency, by contrast, was fashioned relatively quickly, after considerably less discussion. One important reason for the delegates’ reticence was that George Washington, the most admired man in the world at that time, was the convention’s president. Every delegate knew that Washington would, if he chose, be the first president of the new federal government—and that the new government itself would likely fail without Washington at the helm. To express too much fear of executive authority might have seemed disrespectful to the man for whom the office was being tailored.
Washington’s force of personality terrified almost all of his contemporaries, and although he said little as presiding officer, he was not always quiet. Once, when an unknown delegate left a copy of some proposed provisions lying around, Washington scolded the delegates like a headmaster reproving careless prep-schoolers, and then left the document on a table, saying, “Let him who owns it take it.” No one did.
Even when Washington remained silent, his presence shaped the debate. When, on June 1, James Wilson suggested that the executive power be lodged in a single person, no one spoke up in response. The silence went on until Benjamin Franklin finally suggested a debate; the debate itself proceeded awkwardly for a little while, and was then put off for another day.
Many of the conversations about presidential authority were similarly awkward, and tended to be indirect. Later interpreters have found the original debates on the presidency, in the words of former Supreme Court Justice Robert H. Jackson, “almost as enigmatic as the dreams Joseph was called upon to interpret for Pharaoh.”
In the end, the Framers were artfully vague about the extent and limits of the president’s powers. Article I, Section 8 of the Constitution, which empowers Congress, runs 429 words; Article II, Section 2, the presidential equivalent, is about half as long. The powers assigned to the president alone are few: he can require Cabinet members to give him their opinions in writing; he can convene a special session of Congress “on extraordinary occasions,” and may set a date for adjournment if the two houses cannot agree on one; he receives ambassadors and is commander in chief of the armed forces; he has a veto on legislation (which Congress can override); and he has the power to pardon.
The president also shares two powers with the Senate—to make treaties, and to appoint federal judges and other “officers of the United States,” including Cabinet members. And, finally, the president has two specific duties—to give regular reports on the state of the union, and to “take care that the laws be faithfully executed.”
All in all, the text of Article II, while somewhat ambiguous—a flaw that would be quickly exploited—provided little warning that the office of president would become uniquely powerful. Even at the convention, Madison mused that it “would rarely if ever happen that the executive constituted as ours is proposed to be would have firmness enough to resist the legislature.” In fact, when citizens considered the draft Constitution during the ratification debates in 1787 and 1788, many of their concerns centered on the possibility that the Senate would make the president its cat’s-paw. Few people foresaw the modern presidency, largely because the office as we know it today bears so little relation to that prescribed by the Constitution.
The modern presidency is primarily the intellectual handiwork not of “the Framers” but of one Framer—Alexander Hamilton. Hamilton’s idea of the presidency can be found in a remarkable speech he gave to the convention, on June 18, 1787. In it, Hamilton argued that the president should serve for life, name Cabinet members without Senate approval, have an absolute veto on legislation, and have “the direction of war” once “authorized or begun.” The president would be a monarch, Hamilton admitted, but an “elective monarch.”
Hamilton’s plan was so far from the mainstream of thought at the convention that none of its provisions was ever seriously discussed. Nonetheless, Hamilton was and remains the chief theorist of the presidency, first in writing his essays for The Federalist and then in serving as George Washington’s secretary of the Treasury. In this latter role, acting as Washington’s de facto prime minister, Hamilton took full advantage of the vagueness and brevity of Article II, laying the groundwork for an outsize presidency while the war-hero Washington was still in office.
In The Federalist, Hamilton had famously proclaimed that “energy in the executive is a leading character in the definition of good government.” Just how much energy he favored became clear during America’s first foreign crisis, the Neutrality Proclamation controversy of 1793. When Britain and France went to war, many Americans wanted to aid their Revolutionary ally. But Washington and the Federalists were rightly terrified of war with the powerful British Empire. Washington unilaterally proclaimed that the United States would be neutral.
France’s American supporters, covertly aided by Thomas Jefferson, fiercely attacked Washington for exceeding his constitutional authority. The power to make treaties, they said, was jointly lodged in the president and the Senate; how could Washington unilaterally interpret or change the terms of the treaty of alliance with France?
Under the pen name “Pacificus,” Hamilton wrote a defense of Washington’s power to act without congressional sanction. The first Pacificus essay is the mother document of the “unitary executive” theory that Bush’s apologists have pushed to its limits since 2001. Hamilton seized on the first words of Article II: “The executive power shall be vested in a President of the United States of America.” He contrasted this wording with Article I, which governs Congress and which begins, “All legislative powers herein granted shall be vested in a Congress of the United States.” What this meant, Hamilton argued, was that Article II was “a general grant of … power” to the president. Although Congress was limited to its enumerated powers, the executive could do literally anything that the Constitution did not expressly forbid. Hamilton’s president existed, in effect, outside the Constitution.
That’s the Bush conception, too. In 2005, John Yoo, the author of most of the administration’s controversial “torture memos,” drew on Hamilton’s essay when he wrote, “The Constitution provides a general grant of executive power to the president.” Since Article I vests in Congress “only those legislative powers ‘herein granted,’” Yoo argued, the more broadly stated Article II must grant the president “an unenumerated executive authority.”
Hamilton’s interpretation has proved durable even though there is little in the record of constitutional framing and ratification to suggest that anyone else shared his view. In times of crisis, power flows to the executive; too rarely does it flow back. And while Washington himself used his power wisely (Jeffersonians found out in 1812 that pulling the British lion’s tail was poor policy), it was during his administration that the seeds of the “national-security state” were planted.
The system that the Framers developed for electing the president was, unfortunately, as flawed as their design of the office itself. When Madison opened discussion on presidential election in Philadelphia, he opined that “the people at large” were the “fittest” electorate. But he immediately conceded that popular election would hurt the South, which had many slaves and few voters relative to the North. To get around this “difficulty,” he proposed using state electors. Electoral-vote strength was based on a state’s total population, not on its number of voters—and the South received representation for three-fifths of its slaves both in the House of Representatives and in the Electoral College.
Scholars still debate whether the Framers foresaw the prospect of a contested presidential election, followed by a peaceful shift of power. (Remember that, as Shakespeare pointed out in Richard II, kings left office feet first.) Some members of the founding generation believed that a duly elected president would simply be reelected until his death, at which point the vice president would take his place, much like the Prince of Wales ascending to the throne.
Perhaps as a result, the mechanics of presidential election laid out in the Constitution quickly showed themselves to be utterly unworkable. The text of Article II contained no provision for a presidential ticket—with one candidate for president and one for vice president. Instead, each elector was supposed to vote for any two presidential candidates; the candidate who received the largest majority of votes would be president; the runner-up would be vice president. In 1800, this ungainly system nearly brought the country to civil war. Thomas Jefferson and Aaron Burr ran as a team; their electors were expected to vote for both of them. Jefferson assumed that one or two would drop Burr’s name from the ballot. That would have given Jefferson the larger majority, with Burr winning the vice presidency. But due to a still-mysterious misunderstanding, all the electors voted for both candidates, producing a tie in the electoral vote and throwing the election to a House vote.
The ensuing drama lasted six days and 36 ballots before Hamilton threw Federalist support to Jefferson (as much as he despised Jefferson, he regarded Burr as “an embryo-Caesar”). This choice began the chain of events that led to Hamilton’s death at Burr’s hands three years later. More important, the imbroglio exposed the fragility of the election procedure.
In 1804, the Electoral College was “repaired” by the Twelfth Amendment; now the electors would vote for one candidate for president and another for vice president. This was the first patch on Article II, but far from the last—the procedures for presidential election and succession were changed by constitutional amendment in 1933, 1951, 1961, and 1967. None of this fine-tuning has been able to fix the system. In 1824, 1876, 1888, and 2000, the Electoral College produced winners who received fewer popular votes than the losers, and it came startlingly close to doing so again in 2004; in 1824, 1876, and 2000, it also produced prolonged uncertainty and the prospect of civil unrest—or the fact of it.
Even when the election system works passably, a president-elect must endure another indefensible feature of the succession process. In England, a new prime minister takes office the day after parliamentary elections; in France, a newly elected president is inaugurated within a week or two. But when Americans choose a new leader, the victor waits 11 weeks—nearly a quarter-year—to assume office. The presidential interregnum is a recurrent period of danger.
Originally, a new president didn’t take office until March 4. This long delay nearly destroyed the nation after the 1860 election. During the disastrous “secession winter,” Abraham Lincoln waited in Illinois while his feckless predecessor, James Buchanan, permitted secessionists to seize federal arsenals and forts. By March 1861, when Lincoln took office, the Civil War was nearly lost, though officially it had not even begun.
In 1932, Franklin Roosevelt crushed the incumbent, Herbert Hoover, but had to wait four months to take office. During that period, Hoover attempted to force the president-elect to abandon his proposals for economic reform. Roosevelt refused to commit himself, but the resulting uncertainty led the financial system to the brink of collapse.
The Twentieth Amendment, ratified in 1933, cut the interregnum nearly in half, but 11 weeks is still too long. After his defeat in 1992, President George H. W. Bush committed U.S. troops to a military mission in Somalia. The mission turned toxic, and Bill Clinton withdrew the troops the following year. Clinton was criticized for his military leadership, perhaps rightly—but the Constitution should not have permitted a repudiated president to commit his successor to an international conflict that neither the new president nor Congress had approved.
As the elder Bush did, an interregnum president retains the power of life or death over the nation. As Clinton did, an interregnum president may issue controversial or corrupt pardons. In either case, the voters have no means of holding their leader accountable.
The most dangerous presidential malfunction might be called the “runaway presidency.” The Framers were fearful of making the president too dependent on Congress; short of impeachment—the atomic bomb of domestic politics—there are no means by which a president can be reined in politically during his term. Taking advantage of this deficiency, runaway presidents have at times committed the country to courses of action that the voters never approved—or ones they even rejected.
John Tyler, who was never elected president, was the first runaway, in 1841. William Henry Harrison had served only a few weeks; after his death, the obscure Tyler governed in open defiance of the Whig Party that had put him on the ticket, pressing unpopular proslavery policies that helped set the stage for the Civil War.
Andrew Johnson was the next unelected runaway. Politically, he had been an afterthought. But after Lincoln’s assassination, Johnson adopted a pro-Southern Reconstruction policy. He treated the party that had nominated him with such scorn that many contemporaries came to believe he was preparing to use the Army to break up Congress by force. After Johnson rebuffed any attempt at compromise, the Republican House impeached him, but the Senate, by one vote, refused to remove him from office. His obduracy crippled Reconstruction; in fact, we still haven’t fully recovered from that crisis.
American political commentators tend to think loosely about exertions of presidential authority. The paradigm cases are Lincoln rallying the nation after Fort Sumter, and Roosevelt, about a year before Pearl Harbor, using pure executive power to transfer American destroyers to embattled Britain in exchange for use of certain British bases. Because these great leaders used their authority broadly, the thinking goes, assertions of executive prerogative are valid and desirable.
Certainly there are times when presidential firmness is better than rapid changes in policy to suit public opinion. Executive theorists in the United States often pose the choice that way—steady, independent executive leadership or feckless, inconstant pursuit of what Hamilton called “the temporary delusion” of public opinion. But not all shifts in public opinion are delusive or temporary. An executive should have some independence, but a presidency that treats the people as irrelevant is not democratic. It is authoritarian.
Lincoln and Roosevelt asserted emergency powers while holding popular mandates. Lincoln had just won an election that also provided him with a handy majority in Congress; Roosevelt was enormously popular, and in 1940 his party outnumbered the opposition 3-to-1 in the Senate and by nearly 100 seats in the House.
But sometimes a president with little or no political mandate uses the office to further a surprising, obscure, or discredited political agenda. Under these circumstances, what poses as bold leadership is in fact usurpation. The most egregious case arises when a president’s policy and leadership have been repudiated by the voters, either by a defeat for reelection or by a sweeping rejection of his congressional allies in a midterm election. When that happens, presidents too often do what George Bush did in 2006—simply persist in the conduct that has alienated the country. Intoxicated by the image of the hero-president, unencumbered by any direct political check, stubborn presidents in this situation have no incentive to change course.
When the voters turn sharply against a president mid-term, his leadership loses some or all of its legitimacy, and the result can be disastrous. Clinton was decisively repudiated in November 1994. After the election, the administration and the new Republican Congress remained so far apart on funding decisions that the government had to shut down for 26 days in 1995 and 1996. This episode is now remembered for Clinton’s political mastery, but it was actually a dangerous structural failure. (Imagine that the al-Qaeda attacks of September 11, 2001, had happened instead on December 20, 1995, when the stalemate had forced the executive branch to send most of its “nonessential” employees home.)
To sum up, while George W. Bush may have been a particularly bad driver, the presidency itself is, and always has been, an unreliable vehicle—with a cranky starter, an engine too big for the chassis, erratic steering, and virtually no brakes. It needs an overhaul, a comprehensive redo of Article II.
Constitutional change is a daunting prospect. But consider how often we have already changed the presidency; it is the Constitution’s most-amended feature. And this is the moment to think of reform—the public’s attention is focused on the Bush disaster, and ordinary people might be willing to look at the flaws in the office that allowed Bush to do what he did.
So how should the presidency be changed?
First, voters should elect presidents directly. And once the vote is counted, the president-elect (and the new Congress) should take office within a week. Americans accustomed to the current system will object that this would not allow enough time to assemble a Cabinet—but in England and France, the new chief executive considers ministerial nominations before the election. A shorter interregnum would force the creation of something like the British shadow cabinet, in which a candidate makes public the names of his key advisers. That would give voters important information, and provide the president with a running start.
Next, Article II should include a specific and limited set of presidential powers. The “unitary executive” theorists should no longer be allowed to spin a quasi-dictatorship out of the bare phrase executive power; like the responsibilities of Congress, those of the president should be clearly enumerated.
It should be made clear, for example, that the president’s powers as commander in chief do not crowd out the power of Congress to start—and stop—armed conflict. Likewise, the duty to “take care that the laws be faithfully executed” needs to be clarified: it is not the power to decide which laws the president wants to follow, or to rewrite new statutes in “signing statements” after Congress has passed them; it is a duty to uphold the Constitution, valid treaties, and congressional statutes (which together, according to the Constitution, form “the supreme law of the land”).
After a transformative midterm election like that of 1994 or 2006, the nation should require a compromise between the rejected president and the new Congress. A president whose party has lost some minimum number of seats in Congress should be forced to form the equivalent of a national-unity government. This could be done by requiring the president to present a new Cabinet that includes members of both parties, which the new Congress would approve or disapprove as a whole—no drawn-out confirmation hearings on each nominee. If the president were unwilling to assemble such a government or unable to get congressional approval after, say, three tries, he would have to resign.
This would not give Congress control of the executive branch. A resigning president would be replaced by the vice president, who would not be subject to the new-Cabinet requirement. This new president might succeed politically where the previous one had failed (imagine Al Gore becoming president in 1995, and running in 1996—and perhaps in 2000—as an incumbent). And that possibility would discourage the new congressional majority from simply rejecting the compromise Cabinet. Resignation might be worse for them than approval.
As a final reform, we should reconsider the entire Hamiltonian concept of the “unitary executive.” When George Washington became president, he left a large organization (the Mount Vernon plantation) to head a smaller one (the federal government). But today, the executive branch is a behemoth, with control over law enforcement, the military, economic policy, education, the environment, and most other aspects of national life. That behemoth is responsible to one person, and that one person, as we have seen, is only loosely accountable to the electorate.
In other areas, the Framers solved this problem neatly: they divided power in order to protect against its abuse. Congress was split into the House and the Senate to ensure that the legislative process would not be so efficient as to absorb powers properly belonging to the other branches. The problem now is not an overweening Congress but an aggrandized executive branch; still, the remedy is the same. We should divide the executive branch between two elected officials—a president, and an attorney general who would be voted in during midterm elections.
As we are learning from the ongoing scandal of the torture memos, one of the drawbacks of a single executive is that Justice Department lawyers may consider it their job to twist the law to suit the White House. But the president is not their client; the United States is. Justice Department lawyers appointed by an elected attorney general would have no motive to distort law and logic to empower the president, while the White House counsel’s office, which does represent the president, would have every incentive to monitor the Justice Department to ensure that it did not tilt too strongly against the executive branch. The watchmen would watch each other.
This arrangement would hardly be unprecedented: most state governments elect an attorney general. The new Article II could make clear that the president has the responsibility for setting overall legal policy, just as governors do today.
None of these changes would erode the “separation of powers.” That happens only when a change gives one branch’s prerogatives to another branch. These changes refer in each instance back to the people, who are the proper source of all power. The changes would still leave plenty of room for “energy in the executive” but would afford far less opportunity for high-handedness, secrecy, and simple rigidity. They would allow presidential firmness, but not at the expense of democratic self-governance.
It’s not surprising that the Framers did not understand the perils of the office they designed. They were working in the dark, and they got a lot of things right. But we should not let our admiration for the Framers deter us from fixing their mistakes.
Our government is badly out of balance. There is a difference between executive energy and autocratic license; between leadership and authoritarianism; between the democratic firmness of a Lincoln and the authoritarian rigidity of a Bush. The challenge we face today is to find some advantage in Bush’s sorry legacy. Reform of the executive branch would be a good place to start.