On the evening of April 8, 1952, Harry Truman went on national radio and told the country that the federal government had taken possession of its steel mills. A labor dispute threatened to halt production during the Korean War, and rather than wait for Congress to act, Truman signed a directive numbered 10340 and seized the mills outright. No statute authorized the move. He acted on the claim that the office itself carried an inherent authority to meet emergencies, and that the claim was enough. Seven weeks later the Supreme Court told him he was wrong, in a decision that historians still treat as the clearest constitutional boundary ever drawn around a president acting alone.
That single episode holds the entire story of the instrument this article follows across two centuries. A pen. A claim of authority drawn from nowhere in particular in the constitutional text. And, eventually, a court willing to say no. The directive Truman signed belonged to a family of presidential actions that the framers never named, never authorized in so many words, and never imagined would grow into a quasi-legislative tool capable of integrating an army, interning a population, creating federal agencies, and reshaping the entire regulatory machinery of the government. George Washington signed eight such directives across two terms. Franklin Roosevelt signed 3,721 across twelve years. Bill Clinton, operating in what looked from the raw numbers like a season of restraint, signed 364. Reading those three figures in sequence tells you that something fundamental changed, but it does not tell you what, and the gap between the count and the meaning is the puzzle this institutional biography exists to solve.
The Constitutional Silence at the Center of the Story
Begin with the text, because the text is where the strangeness lives. The Constitution does not contain the phrase “executive order.” It does not authorize the president to issue binding directives with the force of law. What it does contain are three spare provisions that every modern defense of the practice is built upon. Article II opens by vesting “the executive Power” in a president, a phrase whose breadth has been argued about since the first administration. It instructs the president to “take Care that the Laws be faithfully executed,” a duty that implies some discretion about how execution happens. And it names the president commander in chief, a title that in wartime has carried the broadest claims of all. From those fragments, and from the simple practical necessity of running an executive branch, the directive grew.
Kenneth Mayer, whose 2001 book gave the field its most influential title, argues that this constitutional foundation is real rather than improvised, that presidents issuing directives are exercising genuine authority rooted in Article II and in the statutes Congress passes delegating power to the executive. The opposing reading, associated with William Howell, treats the directive less as a constitutional instrument than as a move in an ongoing bargaining game between the president and Congress, a way of changing the status quo and forcing everyone else to react. Phillip Cooper, in his comprehensive study of the various tools a president wields by signature alone, documents how the practice hardened over time from informal note into formal legal artifact. Andrew Rudalevige situates the whole arc inside a larger account of how the presidency accumulated power across the twentieth century and then surged again after 2001. And the broader theoretical frame, developed by Terry Moe and Howell, casts unilateral action of every kind as a structural feature of the office, something presidents reach for not because of personality but because the institution rewards it.
These scholars do not agree, and their disagreement is not cosmetic. Mayer reads a constitutional instrument; Howell reads a bargaining chip; the distinction shapes whether you see the growth charted in this article as legitimate institutional development or as a slow erosion of the separation of powers. The argument advanced here is that both readings capture something true at different moments in the arc, and that the instrument changed character so completely between 1789 and the late twentieth century that a single theory cannot hold it. The directive Washington signed and the directive Roosevelt signed share a name and almost nothing else.
The Cabinet Argument That Set the Pattern
Before the directive had acquired any settled form, Washington’s cabinet staged the argument that would shadow every later dispute about presidential power exercised alone. The occasion was the 1793 Neutrality Proclamation, in which Washington declared that the United States would stay out of the war between Britain and revolutionary France. The proclamation was not a numbered directive, but it raised the identical question: could the president, acting without Congress, bind the nation through an instrument the Constitution did not explicitly authorize? Alexander Hamilton, writing as Pacificus, defended the action with a sweeping reading of the Article II vesting clause, arguing that the grant of “the executive Power” carried with it a reservoir of authority broad enough to cover any action not specifically denied to the president or assigned to another branch. James Madison, writing as Helvidius at Jefferson’s urging, answered with the narrow reading, insisting that the vesting clause conferred no general grant of power and that authority over war and foreign relations belonged chiefly to Congress. The full reconstruction of how Washington came to issue the proclamation and split his cabinet over it appears in the examination of the 1793 neutrality decision.
The Pacificus-Helvidius exchange matters to this institutional biography because it is the original version of the Mayer-Howell disagreement, conducted by the founders themselves in the first administration. Hamilton’s expansive theory of inherent executive power is the intellectual ancestor of every later defense of unilateral action, including Truman’s doomed claim in the steel seizure. Madison’s narrow theory is the ancestor of every later challenge, including the Youngstown majority. The argument was never resolved, because the Constitution genuinely underdetermines the answer, and the indeterminacy is precisely what allowed the directive to grow. An instrument resting on a contested reading of an ambiguous clause could expand whenever a president had the political room to claim the broad reading and shrink only when a court or a Congress had the will to enforce the narrow one. The two-century arc that follows is, in one sense, simply the record of which reading prevailed at each moment, and the broad reading prevailed far more often than the narrow one.
What Washington Actually Signed
The earliest directives were not policy. They were housekeeping. When Washington issued what scholars later counted as the first of his eight orders, on June 8, 1789, he was asking the holdover department heads from the Confederation government to prepare reports on the state of their offices so the new administration could understand what it had inherited. This is the genetic origin of the entire institution: a chief executive communicating instructions to the officers who worked beneath him. There was nothing quasi-legislative about it. No private citizen was bound. No law was made. A manager was managing.
The number eight is itself a historian’s artifact rather than a fact Washington would have recognized. Pre-twentieth-century directives were not numbered, not centrally filed, and not consistently distinguished from proclamations, letters of instruction, or routine administrative correspondence. When researchers in the modern era went back to assemble retrospective counts, they had to decide what qualified, and different decisions produce different totals. Some scholars count more than eight for Washington; some count fewer. The instability of the early numbers is not a flaw in the record so much as a clue about the instrument’s nature in its first century. It was so informal that nobody bothered to keep score.
The founding-era totals reflect that informality with almost comic clarity. Jefferson is credited with four. Madison with one. Monroe with one. John Quincy Adams with three. These are the numbers of men who governed a small federal establishment and saw the directive as a minor administrative convenience. Even Andrew Jackson, who expanded presidential power in nearly every other dimension and whose 1832 bank veto rewired the relationship between the office and Congress, signed only twelve. The instrument that would later become a weapon of unilateral policy was, in the hands of the founders and their immediate heirs, barely a tool at all. The reader who wants to understand how dramatically Jackson reconceived presidential authority should look not at his directive count but at his 1832 destruction of the Second Bank of the United States, where the real expansion happened. The directive was simply not yet the vehicle for that kind of ambition.
The Antebellum Drift
Between Jackson and Lincoln the numbers crept upward without changing in kind. Van Buren signed ten, Tyler seventeen, Polk eighteen, Taylor five in a presidency cut short by death, Fillmore twelve, Buchanan sixteen. The one outlier is Franklin Pierce at thirty-five, a figure that looks dramatic against his neighbors but dissolves on inspection into the same administrative routine that defined the era, with a heavier hand on patronage and customs-house management rather than any new conception of the instrument. William Henry Harrison signed none, having died after a month in office, a reminder that the count is hostage to longevity as much as to philosophy.
What matters about the antebellum drift is what did not happen. Across four decades of intensifying sectional crisis, mounting tension over slavery, and the slow collapse of the party system, no president reached for the directive as a way to make policy that Congress would not. The instrument stayed inside the executive branch, addressed to officials, concerned with the mechanics of administration. The reason is partly philosophical, a Whig and Jacksonian inheritance that located the lawmaking power firmly in the legislature, and partly practical, because the federal government was small enough that a president could run it without inventing new tools. The directive would not become a policy weapon until a crisis large enough to justify the leap arrived. It arrived in 1861.
The Whig theory of the presidency deserves emphasis because it explains the restraint that the numbers reflect. The Whigs, who produced four presidents between 1841 and 1853, held a constitutional philosophy that deliberately subordinated the executive to Congress, regarding the veto as an instrument to be used sparingly and the directive as nothing more than administrative housekeeping. Their presidents governed accordingly, treating the office as an executor of the legislative will rather than an independent source of policy. Even the Democrats of the era, heirs to a more expansive Jacksonian conception of presidential power, used the directive almost exclusively for patronage and the management of the federal workforce, the customs houses, and the land offices that constituted the bulk of the federal establishment. Pierce’s thirty-five orders, the antebellum high, consisted largely of patronage and customs administration rather than any assertion of policy authority. The restraint was not an accident of small numbers; it was a considered position about where lawmaking power belonged, and it held until the war made restraint impossible. The contrast with the twentieth century is the whole point: the instrument did not grow because presidents discovered a power that had always been latent in it, but because crises overrode the philosophy of restraint that had kept it small.
Lincoln and the War-Power Leap
The Civil War transformed the directive the way it transformed nearly everything about the presidency. Lincoln is credited with forty-eight, a modest count by later standards but a qualitative break from everything before. The war forced him to act at a speed and scale that the ordinary legislative process could not match, and he reached repeatedly for instruments that let him act alone. The most consequential of these is not even conventionally counted as an executive order, which itself reveals how unstable the category was. The Emancipation Proclamation, issued on January 1, 1863, was formally a war-power proclamation rather than a numbered directive, grounded in Lincoln’s authority as commander in chief to seize enemy resources, which in the logic of the document included the enslaved labor sustaining the Confederate war effort. Yet in substance it did what later generations would recognize as the work of a sweeping executive order: it changed the legal status of millions of people by presidential signature, justified entirely by the emergency of war.
The Emancipation Proclamation matters to this institutional biography precisely because of the ambiguity surrounding its classification. It demonstrates that by 1863 the boundary between proclamation, war measure, and directive had become a matter of label rather than substance, and that a president in a sufficient crisis could reshape the country by signature alone. The constitutional theory was narrow, tied tightly to the war power and to property held by belligerents, and Lincoln himself worried that a peacetime court might find it wanting. But the precedent it set was broad. It showed what the instrument could do when a president decided the stakes justified acting without Congress, and every wartime president who followed studied that example.
The Emancipation Proclamation was not the only war measure Lincoln took by signature, and the others fill out the picture of how completely the war remade the instrument. In the weeks after Fort Sumter, with Congress out of session, Lincoln called up the militia, expanded the regular army and navy beyond their authorized strength, ordered a blockade of southern ports, and authorized the suspension of the writ of habeas corpus along the military line between Washington and Philadelphia, all by executive action and without prior legislative sanction. The blockade was an act of war ordinarily requiring a congressional declaration; the army expansion exceeded the limits Congress had set; and the suspension of habeas corpus touched one of the most fundamental protections in the Constitution. The full reconstruction of the legal reasoning behind that suspension, and the constitutional crisis it provoked, appears in the analysis of Lincoln’s 1861 habeas decision. Taken together these measures show that the war did not merely increase the number of directives Lincoln signed; it expanded the conceivable scope of what a directive could do, pushing the instrument into the gravest questions of war powers and civil liberties. When Congress reconvened, it ratified most of what Lincoln had done, but the precedent that a president could act first and seek ratification later, in an emergency, was the lasting institutional residue.
The numbering itself carries a Lincoln fingerprint. When the government finally began assigning numbers to directives, the order designated as Executive Order 1 was a Lincoln directive from 1862, establishing a provisional court in Louisiana. The assignment was retroactive and somewhat arbitrary, made decades later by archivists imposing order on a chaotic record, but it is fitting that the numbering begins in the war that first made the instrument matter. The reader tracking the broader pattern of how war reshaped the office will find the directive count is only one measure among many, and the aggregate per-president comparison across the whole sweep of American history makes the wartime spikes visible at a glance.
The Reconstruction and Gilded Age Institutionalization
After Lincoln the numbers climbed in a way that finally reflected a change in the instrument’s role. Andrew Johnson signed seventy-nine, a sharp rise driven by the administrative demands of occupying and governing the defeated South. Grant signed 217, a figure that marks the directive’s transition from occasional tool to routine instrument of governance. The leap from Johnson to Grant is the first moment in the story where the count genuinely signals institutionalization rather than the accident of a crisis or the quirk of a record-keeper. Grant was running an enormously expanded federal apparatus, charged with enforcing Reconstruction, managing a growing civil service, and administering federal authority across a continent, and the directive had become the natural medium for that work.
The Gilded Age presidents pushed the numbers higher still. Hayes signed ninety-two, Arthur ninety-six, Cleveland 113 in his first term, Benjamin Harrison 143, Cleveland 140 in his second, McKinley 185. Garfield signed only six, having been shot four months into his term, another casualty count rather than a philosophical statement. The steady upward march across these administrations reflects the growth of the federal establishment after the war, the expansion of the civil service, the increasing complexity of managing customs, public lands, Indian affairs, and a federal workforce that the directive was the convenient tool for directing. None of these orders rewrote American life the way Lincoln’s wartime measures had. They were administrative, and they were numerous, and their accumulation built the habit of governing by directive that the twentieth century would weaponize.
A significant share of the Gilded Age output concerned the public lands, and this strand of the record deserves attention because it became the seedbed of the conservation power that Roosevelt would later wield so dramatically. Under the Forest Reserve Act of 1891, Benjamin Harrison began withdrawing millions of acres of public land from settlement by directive, setting aside forest reserves that would eventually become the national forest system. Cleveland continued the practice in his second term, and the cumulative effect was to establish that a president could lock up enormous tracts of the public domain through a stroke of the pen, under authority Congress had delegated. The land-withdrawal directives are the missing link between the administrative orders of the nineteenth century and the policy orders of the twentieth. They were technically administrative, exercises of delegated authority over the public domain, yet their consequences were vast and permanent, reshaping the map of the American West. By the time Roosevelt arrived with his stewardship theory, the precedent that a president could govern the land by directive was already firmly established, and he had only to expand a practice his predecessors had built.
Theodore Roosevelt and the Progressive Surge
The first genuinely modern surge belongs to Theodore Roosevelt, who signed 1,081, more than every president before him combined many times over. The jump from McKinley’s 185 to Roosevelt’s 1,081 is the steepest proportional break in the entire chart, and it announces the arrival of the modern conception of the office. Roosevelt understood the presidency as a stewardship that authorized him to do anything the Constitution and the laws did not forbid, the inverse of the older view that he could do only what they affirmatively permitted. The directive was the perfect instrument for a man who held that theory, because it let him act first and dare Congress and the courts to stop him.
The signature Roosevelt achievement that ran through the instrument was conservation. Using the authority Congress granted in the 1906 Antiquities Act, Roosevelt designated national monuments by directive at a pace and scale that astonished his contemporaries and infuriated his critics, protecting the Grand Canyon and dozens of other sites by signature. The Antiquities Act is a crucial case for the institutional argument because it shows the directive at its most legitimate, exercised under explicit statutory authority that Congress had granted precisely so the president could act. This is Mayer’s case in its strongest form: a directive grounded in delegated power, doing exactly what the legislature authorized, achieving public goods the ordinary process was too slow to deliver. Yet it is also the beginning of the pattern that would later trouble observers, because once a president learns to use delegated authority expansively, the line between executing the law and making it begins to blur.
Taft, Roosevelt’s chosen successor and then his bitter rival, signed 724, sustaining the new baseline even as he governed in a more legalistic and restrained spirit. The numbers, having leapt under Roosevelt, did not retreat. This is the first appearance of what this article will call the institutional ratchet, the tendency of each expansion to set a new floor rather than a temporary peak. Once Roosevelt established that a president could govern at the scale of a thousand directives across two terms, the instrument did not shrink back to McKinley’s dimensions. The ratchet would prove to be the single most important dynamic in the entire story, and it is worth naming it now because every subsequent administration confirms it.
Wilson, the World War, and the Administrative State
Woodrow Wilson signed 1,803, and the bulk of that staggering total came from the mobilization for the First World War. Between 1917 and 1919 Wilson created wartime agencies, commandeered industries, set prices, allocated resources, and built an emergency administrative apparatus almost entirely by directive, operating under broad delegations of authority that a Congress in wartime was willing to grant and that Wilson interpreted generously. The War Industries Board, the Fuel Administration, the Railroad Administration, and a sprawling network of emergency bodies came into being through the instrument. For roughly two years the United States was governed to an unprecedented degree by presidential signature, and the experience left a permanent mark on the conception of what the office could do in a crisis.
The legal scaffolding for the Wilson surge deserves a closer look, because it reveals how a wartime Congress can hand a president the keys to the entire executive branch. The Overman Act of 1918 authorized Wilson to reorganize executive agencies and redistribute their functions by directive for the duration of the war, an extraordinary delegation that effectively let him redraw the administrative map of the government at will. Under that authority and others like it, Wilson took control of the railroads through the Railroad Administration, seized the telegraph and telephone networks, set wages and prices through the War Industries Board under Bernard Baruch, and managed fuel distribution down to the level of ordering nonessential factories closed on particular days to conserve coal. The federal government Wilson built by directive in 1917 and 1918 was, for its brief life, the most powerful peacetime-into-wartime administrative apparatus the country had ever seen, and it was assembled almost entirely through instruments the Constitution never named. The structure was dismantled after the armistice, but the demonstration could not be undone, and when the next great emergency arrived in 1933 the Wilson precedent was waiting.
The Wilson surge is the second great confirmation of the ratchet and the first clear instance of the house thesis this series carries throughout. The argument running through every article in this collection holds that the modern presidency was forged in four crises, the Civil War, the Great Depression, the Second World War, and the Cold War, and that every emergency power created in those crises outlived the emergency that justified it. Wilson’s wartime apparatus was supposed to be temporary. Much of it was dismantled after the armistice. But the demonstration that a president could govern an industrial economy by directive in an emergency could not be undismantled. It entered the institutional memory of the office as a precedent available for the next crisis, and the next crisis, when it came in 1933, was domestic and even larger.
The 1920s presidents kept the numbers high while draining them of Wilson’s significance. Harding signed 522 in less than three years before his death, Coolidge 1,203 across the rest of the decade, Hoover 995 in a single term. These are enormous totals, larger in Coolidge’s case than Roosevelt’s, and they are the strongest evidence in the entire dataset that raw counts mislead. Coolidge, the most temperamentally passive president of the modern era, a man whose governing philosophy amounted to a conviction that the government should do as little as possible, signed more directives than Theodore Roosevelt, the most activist president of his time. The explanation is that the directive had become the routine medium of an expanded administrative state, and that much of the 1920s output was procedural minutiae, civil-service classifications, land withdrawals, and bureaucratic housekeeping, rather than the bold policy strokes the high numbers might suggest. Any reader who takes the counts at face value will conclude that Coolidge was a more aggressive executive than Roosevelt, which is precisely backward, and recognizing why is the key to reading the entire chart correctly.
Franklin Roosevelt and the Peak of the Instrument
Then comes the figure that dwarfs everything: 3,721, signed by Franklin Roosevelt across twelve years and one month, the longest tenure in American history and the most consequential use of the directive ever recorded. The Roosevelt total is so large that it distorts the visual scale of any chart it appears on, and it represents the convergence of two crises, the Great Depression and the Second World War, each of which alone would have produced a surge and which together produced the peak.
The Depression mobilization began on Roosevelt’s fourth day in office. Facing a banking collapse, he declared a national bank holiday by proclamation on March 6, 1933, closing every bank in the country and freezing financial transactions while Congress assembled emergency legislation. The action rested on a strained reading of a 1917 wartime statute, and it worked, restoring enough confidence that deposits flowed back when the banks reopened. The bank holiday set the tone for the entire New Deal, which proceeded through a torrent of directives establishing agencies, setting industrial codes, allocating relief, and reorganizing the executive branch faster than Congress could legislate. The reader who wants to understand how the New Deal recast the federal government in its first hundred days will find the legislative side of that story in the fifteen bills Roosevelt pushed through in the spring of 1933, but the directive was the constant companion of every statute, filling in the operational detail and extending the reach of the new state.
The war produced the second surge and the darkest single use of the instrument in American history. On February 19, 1942, Roosevelt signed Executive Order 9066, authorizing the military to designate exclusion zones from which any person could be removed, the legal foundation for the forced relocation and incarceration of roughly 120,000 people of Japanese ancestry, most of them American citizens. The order made no explicit reference to race or ancestry, delegating to military commanders the authority that was then used to empty the West Coast of its Japanese American population. The Supreme Court upheld the resulting program in 1944, a decision later repudiated by history and eventually by the Court itself. Executive Order 9066 stands as the definitive demonstration of the instrument’s danger, a single signature that suspended the constitutional rights of an entire population on the basis of a wartime emergency, and the full reconstruction of how Roosevelt came to sign it belongs to the dedicated examination of the internment decision. For the purposes of this institutional biography, 9066 is the moment when the directive’s capacity to do grave and irreversible harm became undeniable. The instrument that began as Washington’s request for departmental reports had become, in eighty years, a tool capable of imprisoning citizens by signature.
Roosevelt’s 3,721 is the apex of the raw count, and it is also the moment the count begins to mislead in a new way. After Roosevelt, the numbers fall, and they fall steeply, even as presidential power continues to grow. Understanding why requires holding two facts together: that Roosevelt’s total reflects a genuine peak of governing by directive, and that the decline after him does not reflect a retreat of presidential power but a change in its plumbing.
The Post-War Decline That Was Not a Decline
Truman signed 907, less than a quarter of Roosevelt’s total, yet his presidency includes two of the most consequential directives in the entire chart and one of the most important constitutional defeats. On July 26, 1948, Truman signed Executive Order 9981, ordering the desegregation of the armed forces and establishing the principle of equal treatment and opportunity regardless of race. The order was an act of moral and political courage, issued in an election year against the resistance of the military establishment and the southern wing of his own party, and it accomplished by signature a transformation that Congress would not have passed. It is the clearest case in the dataset of the directive used for civil rights, a precedent that Kennedy and Johnson would build upon, and it shows the instrument at its most admirable, a president acting alone to advance justice that the legislative process was blocking.
The steel seizure of 1952, with which this article opened, shows the opposite face. When Truman signed Executive Order 10340 and took possession of the steel mills, he relied not on any statute but on a claim of inherent executive authority to act in an emergency. The steel companies sued, and in Youngstown Sheet and Tube Company versus Sawyer, decided in June 1952 and reported at 343 U.S. 579, the Supreme Court ruled the seizure unconstitutional. The decision is the single most important legal limit ever placed on the directive, and its lasting influence comes from a concurring opinion by Justice Robert Jackson, who proposed a three-part framework for judging presidential action. A president acts at the height of his power when Congress has authorized the action, in a zone of uncertainty when Congress is silent, and at the lowest ebb when acting against the expressed or implied will of Congress. Jackson’s framework became the controlling logic for evaluating unilateral action, and it established the principle that a directive cannot manufacture authority the president does not otherwise possess. The instrument has limits, and Youngstown drew the clearest one.
Eisenhower signed 484, the lowest total of any two-term president of the modern era, a figure that fits his reputation for restraint and his preference for working through institutional channels. Yet his presidency contains one of the most dramatic deployments of the instrument. In September 1957, when the governor of Arkansas used the National Guard to block the court-ordered integration of Central High School in Little Rock, Eisenhower signed Executive Order 10730, federalizing the Arkansas National Guard and removing it from the governor’s control, and ordered regular Army troops to enforce the desegregation order. The directive turned the soldiers who had been blocking nine Black students into the soldiers who escorted them into the building. It was a stark assertion of federal authority over a defiant state, accomplished by signature, and the full account of how Eisenhower reached the decision to send the 101st Airborne into an American city is told in the reconstruction of the Little Rock deployment. For this institutional biography, 10730 demonstrates that the significance of a directive bears no relationship to its administration’s raw count. Eisenhower signed fewer orders than almost any modern president, and one of them changed the trajectory of the civil rights struggle.
The post-war decline in numbers, then, conceals rather than reveals the trajectory of presidential power. The counts fell because the mechanism of presidential influence shifted. As the federal government built out a vast apparatus of agencies, each with its own delegated rulemaking authority, presidents increasingly achieved their goals by directing those agencies to issue regulations rather than by signing directives themselves. A modern president who wants to change environmental policy does not issue a thousand directives; he appoints the head of an agency and signals the direction, and the agency produces the rules. The raw executive-order count captures less and less of what a president actually does, which means the decline from Roosevelt’s 3,721 to Eisenhower’s 484 measures a change in technique, not a reduction in power.
The Cold War and the National-Security Directive
The fourth of the crises that this series identifies as the forge of the modern presidency, the Cold War, expanded the instrument along a track that the raw counts barely register. Truman’s 907 directives include not only the desegregation order and the steel seizure but the founding instruments of the national-security state. In March 1947, the same month he announced the doctrine that would define containment, Truman signed Executive Order 9835 establishing the federal employee loyalty program, which subjected millions of government workers to investigation of their political associations and set the template for the loyalty and security apparatus of the early Cold War. The loyalty order is a civil-liberties episode of considerable gravity, accomplished entirely by directive, and it shows the instrument used to build the domestic machinery of the Cold War in the same way Wilson had used it to build the machinery of the First World War.
The deeper Cold War development was the emergence of an entire parallel category of secret directives that never appeared in the public count at all. The national-security directive, issued under various names across administrations, governed covert operations, intelligence priorities, and military planning, and by its nature it was classified, unnumbered in the public register, and invisible to the kind of count this article tabulates. The growth of this secret track is the clearest evidence that the visible directive count understates presidential unilateral action, because the most consequential national-security decisions of the Cold War were made through instruments deliberately kept off the public ledger. Eisenhower, whose public count of 484 marks him as a restrained user of the ordinary directive, governed the covert side of the Cold War through a dense structure of national-security directives that authorized interventions and intelligence operations around the world. The instrument had bifurcated: a public track, subject to the Federal Register and the courts, and a secret track, subject to neither, and the count captures only the first. Any honest assessment of the Cold War presidency has to hold both tracks together, and doing so confirms that the apparent post-war restraint in the visible numbers conceals a vast expansion of unilateral authority in the invisible ones.
The Two Faces of the Same Signature
Two directives, signed six years apart, mark the moral extremes of the instrument and explain why a count can never capture its meaning. Roosevelt’s Executive Order 9066 of 1942 and Truman’s Executive Order 9981 of 1948 were produced by the same office, through the same constitutional mechanism, with the same legal force, and they point in opposite moral directions. The first authorized the incarceration of an entire population on the basis of ancestry. The second dismantled racial segregation in the very armed forces that had carried out the war during which the incarceration happened. Reading the two together is the most efficient way to understand why the directive resists single-metric evaluation.
The structural similarity is the unsettling part. Both orders delegated broad authority to subordinates rather than spelling out every consequence. Executive Order 9066 never named Japanese Americans; it authorized military commanders to designate exclusion zones, and the commanders did the rest. Executive Order 9981 declared a policy of equality and established a committee to implement it, leaving the operational work of desegregation to the services and the implementing structure. In both cases a president set a direction by signature and let the machinery of the executive branch carry it out, which is exactly how the modern instrument works. The mechanism is morally neutral. It can deliver justice or injustice with equal facility, because it is a tool for converting presidential will into administrative action, and the character of the action depends entirely on the will behind it.
This is the scope problem made vivid. A count treats these two orders as two units, interchangeable with each other and with a routine reclassification of a civil-service grade. The count cannot see that one of them ranks among the gravest civil-liberties violations in American history while the other ranks among the most important advances for racial equality. Any evaluation of the instrument that rests on numbers will therefore miss everything that matters about it, because what matters is concentrated in a handful of consequential orders whose significance the count flattens to one. The lesson is that the directive must always be judged order by order, weighing consequence rather than counting frequency, and that the aggregate numbers are useful only for tracking the formal behavior of signing, never for assessing the instrument’s effect on the country.
The Modern Instrument: Kennedy Through Clinton
Kennedy signed 214 in less than three years, a rate consistent with the post-war pattern, and used the instrument for one of the signature creations of his presidency. On March 1, 1961, Executive Order 10924 established the Peace Corps on a pilot basis, before Congress had passed the legislation that would later authorize it permanently. The order let Kennedy launch a program he had promised in the campaign without waiting for the legislative process, a classic instance of the directive used to seize initiative and create facts on the ground that Congress would then ratify. It is a small order in its immediate legal scope but a revealing one, showing how a president uses the instrument to lead rather than follow.
Lyndon Johnson signed 325 and used the directive to advance the cause that defined his domestic legacy. Executive Order 11246, signed in September 1965, required federal contractors to take affirmative action to ensure equal employment opportunity regardless of race, color, religion, sex, or national origin, building on the principle Truman had established with the desegregation of the military and extending it into the economy. The order created an enforcement structure that has shaped employment law for decades and remains one of the most consequential civil-rights actions ever taken by directive. Johnson’s mastery of the legislative process was unmatched, yet on this front he chose the instrument, because it let him reach into the practices of every company that did business with the government in a way no single statute could.
Nixon signed 346 and used the instrument for a reorganization that reshaped the executive branch itself. Executive Order 11541, signed in 1970, implemented the creation of the Office of Management and Budget, transforming the old Bureau of the Budget into a far more powerful instrument of presidential control over the federal government. The new office gave the president centralized authority over budgeting and management across the entire executive branch, a concentration of control that fit Nixon’s broader project of strengthening the presidency against the bureaucracy and Congress alike. The directive that created the modern OMB is a quiet entry in the chart, easy to overlook beside the dramatic civil-rights orders, but its long-term effect on the balance of power inside the government was enormous, because it gave every subsequent president a tool for imposing his priorities on the sprawling agencies that nominally worked for him.
Ford signed 169, Carter 320, and the numbers settled into the range that would define the late twentieth century. The two presidencies are worth pausing on because they show the instrument absorbed into ordinary governance, used for economic management, energy policy, and intelligence reform rather than for dramatic single strokes. Ford, governing in the aftermath of Watergate and amid stagflation, used directives heavily for economic stabilization and for the reorganization of the intelligence community, issuing an order in 1976 that imposed the first significant restrictions on intelligence-agency activities after the abuses the congressional investigations had exposed. Carter, facing the energy crisis, used the instrument to build out the apparatus of energy policy and to advance civil-service reform, and he issued the directives that structured the federal response to the oil shocks of the period. Neither president treated the directive as a tool of confrontation with Congress in the manner that would later become routine; both used it as a normal instrument of administration in difficult times. The character of the instrument in the Ford and Carter years is the calm before the regulatory-control revolution that Reagan would launch, the moment when the directive was simply a standard tool of governance rather than the contested weapon it would become in the polarized decades that followed.
Then came two directives that revealed the instrument’s final modern form, as a tool for controlling the regulatory state that had grown up around the presidency. Reagan signed 381, and the most institutionally significant was Executive Order 12291, issued in February 1981, which required federal agencies to submit major regulations to the Office of Management and Budget for cost-benefit analysis before they could take effect. The order centralized control over the regulatory output of the entire executive branch in the White House, giving the president a mechanism to slow, shape, or stop the rules that agencies produced. It was a directive about directives, an instrument for governing the way the government governed, and it represented the maturation of the technique that had been quietly replacing the raw directive since Roosevelt. Reagan did not need to sign thousands of orders to control policy; he needed to sign one that put the regulatory apparatus under his review.
Clinton signed 364, and his use of the instrument confirmed both its modern character and the partisan symmetry of the technique. Executive Order 12898, signed in February 1994, required federal agencies to identify and address the disproportionate environmental and health effects of their programs on minority and low-income populations, embedding environmental-justice analysis into the regulatory process. Like Reagan’s cost-benefit order, it was a directive that shaped how agencies did their work rather than a one-time policy stroke, and it showed that the regulatory-control technique Reagan had pioneered served a Democratic president pursuing different ends just as well as it had served a Republican one. The Clinton White House grasped the political appeal of the instrument with unusual frankness. The aide Paul Begala, describing the administration’s enthusiasm for acting by signature, reportedly summarized the mood in a phrase that became famous for its candor about the appeal of unilateral action, something close to “stroke of the pen, law of the land.” The line captured exactly why presidents of both parties had come to rely on the instrument: it let them make policy without the painful, uncertain, often futile work of moving legislation through Congress.
The Agency Workaround and the Vanishing Count
The most consequential development in the instrument’s modern history is one that the chart cannot show, because it consists precisely of activity that does not register as a numbered directive. As the federal government built out its vast structure of agencies across the middle of the twentieth century, Congress delegated rulemaking authority to those agencies on an enormous scale, authorizing them to issue regulations that carry the full force of law. A president who controls the agencies controls that rulemaking, and he controls the agencies through appointments, through budget authority, and through the kind of centralized review that Reagan’s Executive Order 12291 formalized. The result is that a modern president can transform vast areas of policy without ever signing a directive that would show up in the count, simply by directing the agencies that now do the work the directive once did.
This is why the post-Roosevelt decline in the numbers is the most misread feature of the entire dataset. A naive reading sees Roosevelt’s 3,721 falling to Eisenhower’s 484 and Reagan’s 381 and concludes that the presidency grew more restrained, that the era of governing by signature passed with the Second World War. The opposite is true. Presidential power over policy grew continuously across the post-war decades, but the mechanism migrated from the directive to the agency. When a modern administration wants to change environmental policy, labor policy, or financial regulation, it does not issue hundreds of directives; it installs the right administrators and signals the direction, and the agencies produce rules that fill thousands of pages of the Federal Register. None of that volume appears in the executive-order count, which captures only the formal directives the president signs personally, an ever-shrinking fraction of the unilateral action the executive branch actually undertakes.
Andrew Rudalevige’s account of the accumulating presidency is essential here, because it locates the directive within a larger ecosystem of unilateral tools that grew together. The directive is one instrument among several, alongside agency rulemaking, signing statements that announce how the president will interpret a law, and the broad national-security authorities that expanded after the Second World War and again in later decades. Reading the directive in isolation, by its count alone, produces a false picture of decline; reading it as one component of a larger and growing apparatus of unilateral action produces the accurate picture of continuous expansion. The count fell because the work moved, not because the work stopped. This migration is the single most important thing to understand about the modern instrument, and it is invisible to anyone who reads the numbers without it.
There is a deeper irony embedded in the workaround. The very instrument that came to control the agencies, the centralized regulatory review that Reagan established and every successor retained, is itself a directive, Executive Order 12291 and its successors. The most powerful modern use of the instrument is to govern the agencies that have absorbed the instrument’s old function. A president signs one directive to control the production of thousands of agency rules, achieving through a single signature a leverage over policy that Roosevelt could only have matched with hundreds of individual orders. By this measure the directive did not decline at all; it became more powerful even as it became less frequent, concentrating its force in a small number of structural orders that reach across the entire executive branch. The count went down; the power per order went up; and the net effect was an instrument more potent than its raw numbers will ever reveal.
The Orders the Courts and Congress Pushed Back
The directive is vast, but it is not lawless, and an honest account has to register the moments when the other branches drew lines. Youngstown is the most famous, but it is not the only one, and the pattern of pushback reveals the real boundaries of the instrument. The courts have repeatedly held that a directive cannot exceed the authority the president holds under the Constitution or a statute, and they have struck down or narrowed specific orders that reached too far. The 1952 steel seizure is the landmark because it involved a direct claim of inherent emergency power with no statutory basis, the purest test case, and the Court’s rejection of that claim established the principle that the instrument executes power rather than creating it.
Justice Robert Jackson’s three-tier framework from his Youngstown concurrence remains the governing analysis, and its durability comes from its realism about how presidential power actually works. Jackson observed that a president acts with maximum legitimacy when Congress has authorized the action, because then he wields his own power plus whatever Congress can delegate. He acts in a zone of twilight when Congress is silent, where the answer depends on the circumstances and the gloss that practice has put on the constitutional text. And he acts at his lowest ebb when he proceeds against the expressed or implied will of Congress, where his claim must rest on his own constitutional authority minus whatever Congress can constitutionally subtract. The steel seizure fell into the third category, because Congress had considered and declined to authorize the seizure of struck industries, and a directive in that posture could not stand. The framework matters because it explains why the ratchet, however far it has turned, has not produced an unlimited executive: the directive operates most freely in the first two zones and faces real constraint in the third.
Congress has its own tools, and it has used them. A statute generally overrides a directive, so Congress can simply legislate against an order it dislikes, subject to the president’s veto and the two-thirds majority required to override it. Congress can defund the implementation of an order, starving it of the appropriations needed to carry it out. And Congress can repeal the statutory authority on which a directive rests, pulling the legal foundation out from under it. These tools are blunt and politically costly, which is why Congress uses them rarely, but their existence is the reason the directive remains, in principle, a subordinate instrument rather than a sovereign one. The deepest check, though, is the simplest: a directive can be revoked by the next president as easily as it was issued. Policy made by signature can be unmade by signature, which is why so much of what presidents accomplish through the instrument proves impermanent, surviving only until an administration of the other party arrives with a pen of its own. The reversibility is the instrument’s great weakness and the structural reason that durable change still requires the slow, painful work of legislation that the directive was invented to avoid.
The Findable Artifact: The Full Count and the Rate That Corrects It
The single most useful thing this article can give a reader is the complete dataset, presented in a way that exposes the distortion in the raw numbers. The table below lists every president from Washington through Clinton with the documented count of directives, the approximate per-year rate that normalizes for the length of each tenure, the landmark order most associated with the administration, and a scope classification indicating the dominant character of that president’s use of the instrument. The per-year rate is the crucial column, because it strips out the longevity effect that makes Roosevelt’s twelve-year total incomparable to a four-year presidency and reveals the actual intensity of each administration’s reliance on the device.
| President | Directives | Approx. per year | Landmark order | Dominant scope |
|---|---|---|---|---|
| Washington | 8 | 1.0 | First departmental directive, 1789 | Administrative |
| Jefferson | 4 | 0.5 | Routine departmental instruction | Administrative |
| Madison | 1 | 0.1 | Administrative | Administrative |
| Monroe | 1 | 0.1 | Administrative | Administrative |
| J.Q. Adams | 3 | 0.8 | Administrative | Administrative |
| Jackson | 12 | 1.5 | Patronage and removal directives | Administrative |
| Van Buren | 10 | 2.5 | Administrative | Administrative |
| W.H. Harrison | 0 | 0 | Died in office | None |
| Tyler | 17 | 4.4 | Administrative | Administrative |
| Polk | 18 | 4.5 | Wartime administration, Mexican War | Administrative |
| Taylor | 5 | 3.8 | Administrative | Administrative |
| Fillmore | 12 | 4.6 | Administrative | Administrative |
| Pierce | 35 | 8.8 | Patronage and customs management | Administrative |
| Buchanan | 16 | 4.0 | Administrative | Administrative |
| Lincoln | 48 | 11.7 | Emancipation Proclamation, 1863 | Emergency, national security |
| A. Johnson | 79 | 20.3 | Reconstruction administration | Administrative |
| Grant | 217 | 27.1 | Civil-service and Reconstruction orders | Administrative |
| Hayes | 92 | 23.0 | Civil-service reform directives | Administrative |
| Garfield | 6 | 12.0 | Assassinated early | Administrative |
| Arthur | 96 | 27.4 | Civil-service classification | Administrative |
| Cleveland (1st) | 113 | 28.3 | Land and pension administration | Administrative |
| B. Harrison | 143 | 35.8 | Land and forest reserve withdrawals | Administrative |
| Cleveland (2nd) | 140 | 35.0 | Pullman Strike enforcement context | Administrative |
| McKinley | 185 | 41.1 | Spanish-American War administration | National security |
| T. Roosevelt | 1,081 | 144 | Antiquities Act monument designations | Regulatory, conservation |
| Taft | 724 | 181 | Conservation and land orders | Administrative, regulatory |
| Wilson | 1,803 | 225 | WWI mobilization agencies | Emergency, national security |
| Harding | 522 | 218 | Land and administrative orders | Administrative |
| Coolidge | 1,203 | 215 | Civil-service and land classification | Administrative |
| Hoover | 995 | 249 | Land withdrawals, administrative orders | Administrative |
| F. Roosevelt | 3,721 | 308 | EO 9066 internment; 1933 bank holiday | Emergency, national security |
| Truman | 907 | 117 | EO 9981 armed-forces integration | Civil rights, national security |
| Eisenhower | 484 | 61 | EO 10730 Little Rock federalization | Civil rights, national security |
| Kennedy | 214 | 76 | EO 10924 Peace Corps | Administrative, foreign policy |
| L. Johnson | 325 | 63 | EO 11246 affirmative action | Civil rights, regulatory |
| Nixon | 346 | 62 | EO 11541 Office of Management and Budget | Administrative, structural |
| Ford | 169 | 70 | Administrative and economic orders | Administrative |
| Carter | 320 | 80 | Civil-service and energy orders | Regulatory, administrative |
| Reagan | 381 | 48 | EO 12291 regulatory cost-benefit review | Regulatory |
| Bush Sr. | 166 | 42 | Administrative and Gulf War orders | National security, administrative |
| Clinton | 364 | 46 | EO 12898 environmental justice | Regulatory |
The per-year column tells a story the raw totals hide. The intensity of governing by directive peaked not with the largest absolute total but across the late Progressive and 1920s presidencies and into Roosevelt, where rates climbed above two hundred orders annually. After Roosevelt the rate collapses, with every president from Truman forward governing at well under half the Roosevelt-era pace and most of them under a quarter of it. A reader looking only at the totals would think the instrument grew steadily across the twentieth century and then plateaued. The rate reveals instead that the directive surged, peaked around the Second World War, and then declined sharply as a measured behavior, even as presidential power continued to expand through other channels. This is the central quantitative finding of the article, and it is the foundation of the argument that raw counts cannot be read as a power index.
This inversion deserves a name, because it is the single most counterintuitive fact in the dataset and the one most likely to mislead a casual reader. Call it the count-power inversion: the period of the highest directive counts and rates, roughly Roosevelt through the early 1950s, preceded a long era of falling counts during which presidential power over policy continued to grow. The numbers peak and then fall; the power rises and keeps rising. The two curves move in opposite directions across the second half of the twentieth century, which means that anyone who treats the directive count as a proxy for presidential assertiveness will read the trend exactly backward, concluding that the office grew more restrained precisely when it was growing more dominant. The count-power inversion is the reason this article insists on pairing every number with a landmark order and a scope classification, and it is the analytical contribution a reader cannot get from a bare list of totals.
A second artifact makes visible the institutional infrastructure that turned the informal directive into a formal legal instrument. The directive did not simply grow in number; it grew in legal definition, acquiring a numbering system, a publication requirement, and a body of constraining case law over the course of a century.
| Year | Event | Significance |
|---|---|---|
| 1862 | Lincoln directive later designated Executive Order 1 | Retroactive starting point of the numbering system |
| 1907 | State Department begins systematic numbering | First consistent identification of directives |
| 1935 | Federal Register Act | Established modern numbering and mandatory publication |
| 1946 | Administrative Procedure Act | Subjected agency action to procedural requirements |
| 1952 | Youngstown Sheet and Tube v. Sawyer, 343 U.S. 579 | Established that directives cannot exceed statutory authority |
| 1976 | Government in the Sunshine and transparency era | Heightened disclosure norms around executive action |
| 1981 | Executive Order 12291 | Centralized regulatory review in the White House |
The 1935 Federal Register Act is the hinge in this timeline. Before it, directives were not systematically published, and a citizen could be bound by a presidential order he had no reliable way to read. The Act required that directives be numbered and printed, transforming the instrument from a private communication within the executive branch into a public legal document subject to the ordinary apparatus of legal notice. The numbering that scholars use to count orders is itself a product of this infrastructure, which is why pre-1907 counts are so unstable and why Executive Order 1 is a retroactive designation rather than a contemporaneous fact. The instrument acquired its modern legal identity in the 1930s, exactly as Roosevelt was using it more intensively than any president before or since.
The Complication: Why the Counts Mislead
The argument cannot end with the chart, because the chart is treacherous, and an honest institutional biography has to disarm it before drawing conclusions. There are three distinct ways the raw counts mislead, and each one matters.
The first distortion is the pre-1907 numbering problem. Before the State Department began systematically numbering directives, the record is incomplete and inconsistent. Many informal directives, letters of instruction, and administrative orders that would now be counted simply were not recorded as such, which means the early totals understate the true volume of presidential directive-issuing. Washington’s eight, Jefferson’s four, Madison’s one: these are not full inventories of every binding instruction those presidents issued to their officers, but artifacts of a retrospective counting exercise applied to an era that did not keep score. The early numbers are real, but they are floors rather than complete totals, and the apparent explosion of activity in the late nineteenth century partly reflects improved record-keeping rather than a pure increase in behavior.
The second distortion is the delegation problem, and it is the most important. After the build-out of the administrative state, presidents increasingly achieved policy through agency rulemaking rather than direct directives. A modern president who wants to transform a policy area appoints the right agency head and signals the direction, and the agency issues regulations that carry the force of law without the president ever signing a directive. The raw executive-order count captures none of this, which means the post-Roosevelt decline measures a shift in technique rather than a reduction in unilateral authority. The modern president governs through the regulatory state at least as forcefully as Roosevelt governed through directives, but the activity shows up in the Federal Register as agency rules rather than as numbered executive orders. Reagan’s Executive Order 12291 makes this concrete: a single directive that gave the White House control over the entire regulatory output of the government accomplished more, in terms of policy leverage, than hundreds of routine orders, precisely because it operated at the level of the agencies that now did the heavy lifting.
The third distortion is the scope problem. Individual directives vary enormously in consequence, from a routine reclassification of a civil-service position to the incarceration of an entire population. A count treats Executive Order 9066 and an order adjusting the boundary of a federal land reserve as equal units, which is absurd. The dataset that produces Coolidge’s 1,203 is dominated by procedural minutiae, while the forty-eight Lincoln orders include the legal architecture of emancipation. Any reading that weights the instrument by count rather than by consequence will reach conclusions exactly opposite to the truth, ranking the passive Coolidge above the activist Roosevelt and treating the restrained Eisenhower, whose 10730 sent the Army into an American city, as a minor user of the instrument. The scope problem is why this article pairs every count with a landmark order and a scope classification, because the qualitative dimension is the one that matters and the one the numbers erase.
Holding these three distortions together produces the correct way to read the chart. The raw totals are a record of a particular formal behavior, the signing of numbered directives, that surged in the early twentieth century, peaked under Roosevelt, and declined afterward as the technique of governing shifted toward the agencies. They are not a measure of presidential power, which grew continuously even as the directive count fell. And they are not a measure of significance, which is concentrated in a small number of consequential orders scattered across administrations with wildly varying totals. The reader who internalizes these three corrections can use the chart correctly; the reader who takes the numbers at face value will get the whole story backward.
The Verdict
The executive order, followed across two centuries, is the clearest single instrument of the imperial-presidency expansion that this series tracks, but its clarity comes from its qualitative transformation rather than its quantitative growth. The directive that Washington signed to request departmental reports and the directive that Roosevelt signed to authorize internment share a name and a constitutional silence and almost nothing else. Between them lies a change of kind, not merely of degree, in which an informal mechanism for communicating instructions inside the executive branch became a quasi-legislative instrument capable of reshaping American life by signature.
The transformation happened in the crises. Each of the four crises that this series identifies as the forge of the modern presidency produced a surge in the instrument and a permanent expansion of its conceivable scope. The Civil War made Lincoln’s war-power directives the precedent for emergency action. The First World War, though it falls outside the four-crisis frame, demonstrated that a president could govern an industrial economy by directive. The Depression and the Second World War together produced Roosevelt’s peak and the internment order that revealed the instrument’s full capacity for harm. The Cold War sustained the apparatus and added the steel seizure that produced Youngstown, the one durable limit. After each surge the instrument did not return to its prior dimensions. The ratchet held. This is the InsightCrunch executive-order ratchet, the principle that each crisis-era expansion of the instrument established a new floor of conceivable presidential action rather than a temporary peak, so that every president inherits a wider range of unilateral possibility than his predecessor enjoyed.
The verdict on the historiographic dispute is that Mayer and Howell are each right about a different era. Mayer’s reading, that the directive is a genuine constitutional instrument grounded in Article II and statutory delegation, describes the instrument at its best, in Roosevelt’s conservation orders, in Truman’s integration of the armed forces, in the exercises of clearly delegated authority that the framers’ design contemplated. Howell’s reading, that the directive is a bargaining move in an ongoing struggle with Congress, a way of changing the status quo and daring the other branches to respond, describes the modern instrument in its political reality, the Begala “stroke of the pen” mentality that treats the directive as a shortcut around a Congress too divided to act. Both are true. The instrument is simultaneously a legitimate constitutional tool and a weapon of unilateral politics, and which face it shows depends on the order and the moment. Cooper’s account of the formalization, the slow acquisition of numbering and publication and legal definition, explains how the instrument acquired the legitimacy that makes Mayer’s reading possible, while Rudalevige’s account of accumulating presidential power explains the structural pressure that makes Howell’s reading inevitable. The evidence does not force a choice between them; it forces the recognition that the instrument is two things at once.
The one firm boundary is the one Truman discovered in 1952. Youngstown established that the directive cannot manufacture authority the president does not otherwise possess, that the instrument executes power but does not create it, and that a court will enforce the limit when a directive reaches beyond statute and constitution into pure assertion of inherent emergency authority. Jackson’s three-tier framework remains the controlling logic, and it is the reason the ratchet, however far it has turned, has not produced an unlimited executive. The directive is vast, but it is not infinite, and the line is real.
The Legacy and the House Thesis
The executive order’s two-century arc is the strongest single confirmation of this series’ governing argument that the modern presidency was built in four crises and that every president since inherits an office designed for emergencies that have passed. The instrument grew because the crises demanded fast action, and it never shrank because the precedents the crises established could not be unestablished. A modern president, facing an ordinary political situation in ordinary times, commands an instrument forged for total war and economic collapse, and the temptation to use the crisis tool for routine politics is structural, built into the office by the accumulated precedents of the four emergencies. Begala’s candid phrase about the appeal of governing by signature is the voice of that temptation, and it is bipartisan, because the office produces it regardless of who holds it.
The directive’s evolution also illuminates the broader institutional dynamics that this series examines through other mechanisms. The pardon power, traced from Washington’s clemency for the Whiskey Rebellion participants through Clinton’s controversial final-day grants, exhibits a similar ratchet, in which each expansion of the norm-range of what a president can do by clemency widens the field for his successors, and the parallel between the two unilateral instruments is examined in the institutional biography of the presidential pardon. The directive and the pardon are the two great unilateral powers, one constructed from constitutional silence and statutory delegation, the other granted explicitly in a single clause, and both have grown through the same dynamic of crisis-driven expansion followed by normalization. Studying them together reveals that the imperial presidency is not a single seizure of power but an accumulation of ratchets, each instrument widening separately, the whole adding up to an office vastly more capable of acting alone than the framers designed or intended.
The deepest implication is that the instrument’s growth was not chiefly a story of presidential ambition, though ambition played its part, but of institutional logic. Moe and Howell’s framework, which casts unilateral action as a structural feature of the office rather than a function of personality, is vindicated by the arc. The passive Coolidge signed more directives than the activist Theodore Roosevelt. The restrained Eisenhower federalized a state’s National Guard. The numbers do not track temperament because the instrument is not a matter of temperament; it is a matter of the situation the office places its occupant in and the tools the accumulated precedents make available. Every president reaches for the directive because the office rewards reaching for it, and the office rewards reaching for it because four crises taught it to. That is the lesson of the two-century arc, and it is the reason the executive order will keep growing, ratchet by ratchet, crisis by crisis, until something more durable than a single court decision constrains it.
Frequently Asked Questions
Q: Does the Constitution actually authorize executive orders?
No, not explicitly. The Constitution never uses the phrase “executive order” and contains no provision granting the president authority to issue binding directives with the force of law. The practice rests on three implied sources: the Article II vesting clause placing “the executive Power” in the president, the take-care clause requiring faithful execution of the laws, and the commander-in-chief designation that has supported the broadest claims in wartime. Beyond these constitutional fragments, most directives rest on statutory delegation, where Congress passes a law granting the president authority to act and the president exercises it by directive. Kenneth Mayer argues this foundation is genuine constitutional authority; William Howell treats the directive more as a political bargaining tool. The 1952 Youngstown decision established the key limit: a directive cannot create authority the president does not otherwise hold from the Constitution or a statute.
Q: How many executive orders did each president sign from Washington to Clinton?
The documented counts run from a single order for Madison and Monroe up to 3,721 for Franklin Roosevelt. Washington signed 8, Jefferson 4, Lincoln 48, Grant 217, Theodore Roosevelt 1,081, Wilson 1,803, Coolidge 1,203, Hoover 995, and Roosevelt the record 3,721. After Roosevelt the numbers fall sharply: Truman 907, Eisenhower 484, Kennedy 214, Johnson 325, Nixon 346, Ford 169, Carter 320, Reagan 381, Bush Sr. 166, and Clinton 364. The early counts are unstable because directives were not systematically numbered before 1907, so figures for the founding-era presidents are retrospective reconstructions rather than complete inventories. The full table in this article pairs each count with a per-year rate and a landmark order, which is the only honest way to read the data.
Q: Why did Franklin Roosevelt sign so many more executive orders than anyone else?
Roosevelt’s 3,721 reflects the convergence of the longest tenure in American history with two simultaneous crises. He served twelve years and one month, far longer than any other president, which alone inflates the total. More importantly, he governed through two emergencies that each demanded rapid unilateral action: the Great Depression, which produced the torrent of New Deal directives establishing agencies and reorganizing the executive branch, and the Second World War, which produced mobilization orders and the internment authorization of Executive Order 9066. Either crisis alone would have generated a surge; together they produced the peak of the entire chart. Yet even Roosevelt’s total can mislead, because many of his directives were routine administrative measures rather than consequential policy strokes, and the raw number conflates the bank holiday and the internment order with bureaucratic housekeeping that carried no comparable weight.
Q: What was Executive Order 9066 and why does it matter so much?
Executive Order 9066, signed by Roosevelt on February 19, 1942, authorized the military to designate exclusion zones from which any person could be removed. It became the legal foundation for the forced relocation and incarceration of roughly 120,000 people of Japanese ancestry, the majority of them American citizens. The order never mentioned race or ancestry explicitly, instead delegating broad authority to military commanders who then used it to empty the West Coast of its Japanese American population. The Supreme Court upheld the program in 1944, a ruling later repudiated by history and eventually by the Court itself. The order matters to the institutional history because it is the definitive demonstration of the directive’s capacity for grave, irreversible harm, showing that a single signature could suspend the constitutional rights of an entire population on the basis of a claimed wartime emergency.
Q: What did the Youngstown steel seizure case decide about executive orders?
In Youngstown Sheet and Tube Company versus Sawyer, decided in June 1952 at 343 U.S. 579, the Supreme Court ruled that Truman’s seizure of the steel mills by Executive Order 10340 was unconstitutional. Truman had relied not on any statute but on a claim of inherent presidential authority to act in an emergency, and the Court rejected that claim. The decision established the most important limit ever placed on the directive: an executive order cannot exceed the authority the president holds under the Constitution or a statute, and it cannot manufacture power from a bare assertion of emergency. Justice Robert Jackson’s concurring opinion proposed the lasting three-tier framework: presidential power is highest when Congress authorizes the action, uncertain when Congress is silent, and lowest when the president acts against Congress’s expressed will. Jackson’s framework remains the controlling logic for evaluating unilateral presidential action.
Q: Why do executive order counts drop after Roosevelt if presidential power kept growing?
The decline in raw counts measures a change in technique rather than a reduction in power. After the build-out of the administrative state, presidents increasingly achieved their policy goals through agency rulemaking rather than direct directives. A modern president who wants to transform a policy area appoints the right agency head, signals the direction, and the agency issues regulations carrying the force of law, all without the president signing a single directive. This activity appears in the Federal Register as agency rules, not as numbered executive orders, so the raw count captures less and less of what a president actually does. Reagan’s Executive Order 12291, which centralized White House control over the entire regulatory output of the government, accomplished more in policy leverage than hundreds of routine orders precisely because it operated at the level of the agencies that now did the heavy lifting. The count fell; the power did not.
Q: Was the Emancipation Proclamation an executive order?
Not formally. The Emancipation Proclamation, issued January 1, 1863, was technically a war-power proclamation rather than a numbered executive order, grounded in Lincoln’s authority as commander in chief to seize the resources of an enemy at war, which in the document’s logic included the enslaved labor sustaining the Confederacy. The distinction matters because it reveals how unstable the category was in the nineteenth century. In substance, the Proclamation did what later generations would recognize as the work of a sweeping directive: it changed the legal status of millions of people by presidential signature, justified by emergency. The narrow war-power theory left Lincoln worried that a peacetime court might reject it, which is part of why he pushed for the Thirteenth Amendment. But as precedent it demonstrated that a president in sufficient crisis could reshape the nation by signature alone, a lesson every wartime successor absorbed.
Q: Which president used executive orders most intensely relative to time in office?
Measured by per-year rate rather than raw total, the most intense users cluster in the late Progressive and 1920s era and peak with Roosevelt. Hoover signed at roughly 249 orders per year, Wilson at about 225, Coolidge at about 215, Harding at about 218, and Roosevelt at about 308. These rates dwarf the modern presidents, who almost all governed at well under half the Roosevelt pace. Eisenhower’s rate of about 61 per year and Reagan’s of about 48 reflect the post-war shift toward governing through agencies. The per-year measure corrects the longevity distortion that makes Roosevelt’s twelve-year total incomparable to a four-year presidency, and it reveals that the intensity of governing by directive surged in the early twentieth century and then declined sharply, even as overall presidential power continued to grow through other channels.
Q: What was Executive Order 9981 and what did it accomplish?
Executive Order 9981, signed by Truman on July 26, 1948, ordered the desegregation of the United States armed forces and established the principle of equal treatment and opportunity for all service members regardless of race. Truman issued it in an election year, against the resistance of much of the military establishment and the southern wing of his own Democratic Party, accomplishing by signature a transformation that Congress would not have passed through legislation. The order is the clearest case in the entire dataset of the directive used to advance civil rights, and it established a precedent that Kennedy and Johnson would build upon with later orders on equal employment and affirmative action. It demonstrates the instrument at its most admirable: a president acting alone to advance justice that the legislative process was actively blocking, using the unilateral character of the directive as a feature rather than a flaw.
Q: How did the Federal Register Act change executive orders?
The 1935 Federal Register Act is the hinge in the institutional history of the directive. Before it, executive orders were not systematically published, which meant a citizen could be legally bound by a presidential directive he had no reliable way to read. The Act required that directives be numbered and printed in a public register, transforming the instrument from a private communication within the executive branch into a public legal document subject to the ordinary apparatus of legal notice. The numbering system that scholars now use to count orders is itself a product of this infrastructure, which is why pre-1907 counts are so unstable and why Executive Order 1 is a retroactive designation applied to an 1862 Lincoln directive rather than a contemporaneous fact. The instrument acquired its modern legal identity in the 1930s, precisely as Roosevelt was using it more intensively than any president before or since.
Q: Why did Coolidge sign more executive orders than Theodore Roosevelt?
This is the single best illustration of why raw counts mislead. Coolidge, the most temperamentally passive president of the modern era, whose governing philosophy amounted to a belief that the government should do as little as possible, signed 1,203 directives, more than the 1,081 signed by Theodore Roosevelt, the most activist president of his time. The explanation is that by the 1920s the directive had become the routine medium of an expanded administrative state, and much of Coolidge’s output consisted of procedural minutiae: civil-service classifications, land withdrawals, and bureaucratic housekeeping. Roosevelt’s lower total included far more consequential strokes, particularly the conservation orders that protected millions of acres under the 1906 Antiquities Act. Anyone who reads the counts as a measure of activism will conclude Coolidge was more aggressive than Roosevelt, which is exactly backward, and recognizing why is the key to reading the whole chart correctly.
Q: What executive orders shaped civil rights the most?
Three directives form the civil-rights spine of the chart. Truman’s Executive Order 9981 of 1948 desegregated the armed forces and established the principle of equal treatment regardless of race. Eisenhower’s Executive Order 10730 of 1957 federalized the Arkansas National Guard and sent regular Army troops to enforce the court-ordered integration of Central High School in Little Rock, turning the soldiers who had blocked Black students into the soldiers who escorted them inside. Johnson’s Executive Order 11246 of 1965 required federal contractors to take affirmative action to ensure equal employment opportunity, extending the equal-treatment principle into the economy and creating an enforcement structure that shaped employment law for decades. Together these three show the directive used at its most consequential for justice, each one accomplishing by signature a change the legislative process was too slow or too blocked to deliver, and each building on the precedent of the one before.
Q: How did Reagan’s Executive Order 12291 change the way government works?
Executive Order 12291, issued in February 1981, required federal agencies to submit major proposed regulations to the Office of Management and Budget for cost-benefit analysis before the rules could take effect. The order centralized control over the entire regulatory output of the executive branch in the White House, giving the president a mechanism to slow, shape, or stop the rules that agencies produced. It represented the maturation of the technique that had been quietly replacing the raw directive since Roosevelt: rather than signing thousands of orders, a president could sign one that put the regulatory apparatus under his review. In effect it was a directive about directives, an instrument for governing how the government governed. Its long-term significance lies in establishing centralized regulatory review as a permanent feature of the presidency, a tool that every subsequent president of both parties has retained and used to impose priorities on the agencies.
Q: What is the difference between an executive order and a proclamation?
The distinction is often more about form and tradition than about legal force. Executive orders are directives to officers and agencies of the executive branch, instructing them how to carry out their duties or implementing presidential policy. Proclamations are typically addressed to the public at large and have historically been used for ceremonial declarations, such as holidays, and for substantive actions like Lincoln’s Emancipation Proclamation or Roosevelt’s bank holiday. The line between them has always been blurry, and the same substantive action could often have been cast as either. The Emancipation Proclamation is the prime example: formally a proclamation grounded in the war power, it accomplished what later observers would recognize as the work of a sweeping executive order. Both instruments rest on the same constitutional and statutory foundations, and both can carry the force of law when properly grounded in presidential authority, which is why scholars often count consequential proclamations alongside numbered orders.
Q: Can Congress or the courts overturn an executive order?
Yes, through several mechanisms. Congress can override a directive by passing legislation that contradicts it, since a statute generally trumps an executive order, though the president may veto such legislation and force Congress to muster a two-thirds majority. Congress can also defund the implementation of an order or repeal the statutory authority on which it rests. The courts can strike down a directive that exceeds the president’s constitutional or statutory authority, as the Supreme Court did in the 1952 Youngstown decision, which established that an order cannot manufacture power the president does not otherwise hold. A subsequent president can also revoke or modify a predecessor’s order by issuing a new directive, which is why much policy made by executive order proves less durable than policy enacted by statute. The reversibility of directives is both their appeal, since they can be issued quickly, and their weakness, since they can be undone just as quickly.
Q: Did the executive order make the presidency more powerful than the framers intended?
The evidence strongly supports that conclusion, though the mechanism was institutional rather than purely the result of presidential ambition. The framers designed an office with carefully limited powers, vesting lawmaking in Congress, and they never contemplated an instrument that would let presidents make quasi-legislative policy by signature. The directive grew through four crises, the Civil War, the Depression, the Second World War, and the Cold War, each of which demanded rapid unilateral action and each of which left behind precedents that outlived the emergency. This article calls the dynamic the executive-order ratchet: each crisis-era expansion set a new floor of conceivable presidential action rather than a temporary peak. Terry Moe and William Howell’s framework, treating unilateral action as a structural feature of the office, explains why the growth happened regardless of who held the office. The passive Coolidge and the restrained Eisenhower used the instrument alongside the activist Roosevelt because the situation, not the temperament, drove the behavior.
Q: Why did presidents of both parties embrace executive orders by the late twentieth century?
Because the instrument solved a problem common to presidents regardless of ideology: the difficulty of moving policy through a divided and slow-moving Congress. Reagan’s Executive Order 12291 establishing regulatory cost-benefit review and Clinton’s Executive Order 12898 requiring environmental-justice analysis pursued opposite political ends, yet both used the same technique of shaping how agencies did their work. The bipartisan appeal was captured candidly by the Clinton aide Paul Begala, who reportedly summarized the administration’s enthusiasm for acting by signature in a phrase close to “stroke of the pen, law of the land.” The line expressed exactly why the instrument crossed party lines: it let a president make policy without the painful, uncertain work of legislation. By the late twentieth century the directive had become a standard tool of governance for both parties, with each new administration using it to advance its own priorities and to undo the orders of its predecessor.
Q: What does Executive Order 1 actually contain and why is it numbered that way?
Executive Order 1 is a Lincoln directive from 1862 establishing a provisional court in Louisiana during the Civil War. Its designation as the first order is retroactive and somewhat arbitrary, made decades later by archivists who imposed a numbering system on a chaotic historical record. When the State Department began systematically numbering directives in 1907, it worked backward through the existing record and assigned number one to that Lincoln order, even though presidents had been issuing directives since Washington and many earlier orders existed. The arbitrary starting point is itself a clue about the instrument’s nature: the directive was so informal in its first seventy years that no one had bothered to number or systematically preserve the orders, and the modern count begins only where the record becomes legible. It is fitting, though accidental, that the numbering begins in the Civil War, the crisis that first made the instrument genuinely consequential.
Q: How do scholars disagree about what executive orders really are?
The central disagreement runs between Kenneth Mayer and William Howell. Mayer, whose influential book gave the field its defining title, argues that executive orders are genuine policy instruments with real constitutional foundations in Article II and statutory delegation, exercised legitimately within the framework of the office. Howell treats them instead as bargaining tools within the ongoing politics between president and Congress, a way of changing the status quo and forcing the other branches to react. Phillip Cooper documents the slow formalization of the instrument from informal note to legal artifact, while Andrew Rudalevige situates the directive inside a broader account of accumulating presidential power across the twentieth century and beyond. Terry Moe and Howell together developed the structural framework casting all unilateral action as a built-in feature of the office. The disagreement is substantive because it determines whether the growth charted across two centuries represents legitimate institutional development or a steady erosion of the separation of powers.
Q: What is the executive-order ratchet?
The executive-order ratchet is the principle, named in this article, that each crisis-era expansion of the directive established a new floor of conceivable presidential action rather than a temporary peak that would later recede. After Theodore Roosevelt demonstrated that a president could govern at the scale of a thousand directives, the instrument never shrank back to the dimensions of his predecessors. After Wilson showed that a president could run an industrial economy by directive in wartime, the demonstration entered the institutional memory of the office as an available precedent. After Roosevelt reached the peak of 3,721, the technique shifted toward agency rulemaking but the underlying scope of unilateral authority did not contract. The ratchet explains why presidential power grew continuously across the twentieth century even as raw directive counts fell after Roosevelt: the floor kept rising, crisis by crisis, so that every president inherits a wider range of unilateral possibility than his predecessor enjoyed, constrained durably only by the limit the Youngstown decision drew in 1952.