Take Up Your Pen
Unilateral Presidential Directives in American Politics
Graham G. Dodds
2013 | 304 pages | Cloth $69.95
View main book page
Table of Contents
Chapter 1. Unilateral Directives and the Presidency
Chapter 2. The Constitutional Executive
Chapter 3. Judicial Sanction
Chapter 4. Early Unilateral Presidential Directives
Chapter 5. Theodore Roosevelt and the Rise of Unilateral Presidential Directives
Chapter 6. Unilateral Presidential Directives from Roosevelt to Roosevelt: Taft through
Chapter 7. Unilateral Presidential Directives from the Postwar Era to the Present Day
Chapter 8. Conclusions
Excerpt [uncorrected, not for citation]
Unilateral Directives and the Presidency
Stroke of the pen, law of the land. Kind of Cool.
—Paul Begala, Advisor to President Clinton (1998)
The history of executive orders is, to a great extent, a narrative of the evolution of presidential power.
—Robert B. Cash, "Presidential Power" (1963)
Outrageous, or Ordinary?
On December, 15, 2005, Americans were shocked to learn that President George W. Bush had issued an executive order directing the National Security Agency (NSA) to engage in domestic spying on U.S. citizens. Bush secretly issued the order in 2002 as part of the government's effort to prevent another terrorist attack on the scale of those committed on September 11, 2001. But Bush's policy appeared to contradict the legal processes established by the Foreign Intelligence Surveillance Act (FISA) of 1978 (Pub. L. 95-511, 92 Stat. 1783), which required the government to obtain a warrant before engaging in domestic spying.
Public indignation at Bush's unilateral directive for domestic spying was immediate and widespread. A poll conducted a month after the public revelation showed that 51 percent of Americans opposed Bush's action and 58 percent supported the appointment of a special prosecutor to investigate the legality of the program. The ACLU complained that the president was "very willing to sacrifice civil liberties" and might have authorized "criminal activity."
Major newspapers also criticized Bush's action. In an editorial three days after its initial story, the New York Times described the program as "illegal government spying" and charged that "Mr. Bush's team cannot be trusted to find the boundaries of the law, much less respect them." The Chicago Tribune said that the Bush administration evidently "thinks it can scorn the law." The Philadelphia Inquirer complained that "no president unilaterally should be able to declare that a part of the Bill of Rights is null and void." And the Dallas Morning News said, "It won't do to tell a free people that their government is spying on them in possible violation of the law for their own good," and it encouraged Congress to "hold the president accountable."
In the halls of Congress, senators and representatives were outraged about Bush's spying directive. The day after the story broke, the Senate refused to reauthorize the Patriot Act, several senators demanded a joint inquiry by the Judiciary and Intelligence Committees, Senator Arlen Specter (then R-PA) threatened hearings, and Senator John Kerry (D-MA) said the president seemed to think he was "above the law." Senator Russ Feingold (D-WI) sought to pass a resolution censuring the president for his action, and some members of Congress even spoke of possible impeachment.
Criticism of Bush's unilaterally imposed spying program even came from members of the Bush administration itself. On New Year's Day, 2006, the Times reported that in March 2004, the acting attorney general, James Comey, had determined that the domestic spying program was likely unconstitutional (even though it had been previously reauthorized by the Department of Justice), so he decided not to renew approval for it. The chief of staff, Andrew Card, and the White House counsel, Alberto Gonzales, then made a dramatic visit to the hospital bed of the attorney general, John Ashcroft, who was recovering from gallbladder surgery, in a fruitless attempt to gain his approval. For some critics, it was ironic that Ashcroft, the father of the Patriot Act, was belatedly assuming the mantle of protector of civil liberties and constitutional niceties, but it suggested that Bush's unilateral action had indeed gone too far.
Five months later, the judiciary also condemned Bush's order. In ACLU v. NSA, 438 F. Supp. 2d 754, E.D. Mich. (2006), Judge Anna Diggs Taylor ruled that Bush's action was unconstitutional. Her opinion admonished, "There are no hereditary kings in America and no powers not created by the Constitution."
In short, Bush's policy was widely condemned by the public, the press, members of Congress from both political parties, the judiciary, and even officials within the administration. So what happened? Nothing. The public outcry dimmed, and Bush's executive order and policy remained in place.
Even when yet more damning details emerged, nothing changed. On May 11, 2006, USA Today revealed that the government was using information gleaned from domestic spying to compile a massive database, including tens of millions of phone calls made by customers of the nation's three largest telecommunications companies, in apparent violation not only of FISA and the Fourth Amendment but also of various statements and promises issued by officials in the Bush administration after the domestic spying program was exposed. But despite this revelation, the program continued. And some polls suggested that the American people had even come to embrace Bush's policy: a Washington Post poll conducted shortly after the database revelation found that 63 percent of respondents supported the NSA program and only 35 percent opposed it. Moreover, Congress acted to legitimate Bush's policy in 2007 (Pub. L. 110-55, 121 Stat. 552) and again in 2008 (Pub. L. 110-261, 122 Stat. 2436), giving legislative sanction to what had been the president's unilateral policy and granting the telecommunications companies immunity from lawsuits for violations of privacy.
Thus, even though the president's executive order initially elicited so much outrage, and even after more and more controversial details emerged, nothing changed, and it was quickly back to business as usual. It was almost as if controversial unilateral presidential policymaking itself were just business as usual, which is exactly what this book argues it is.
The Argument of the Book
The above episode is emblematic of a much broader puzzle in U.S. constitutional politics and interbranch relations, as presidents have often used unilateral directives such as executive orders and proclamations to impose controversial policies, and Congress and the courts have at times complained but have seldom offered much in the way of real resistance. This book seeks to explain how we got to this point and why it matters. The basic argument of the book may be summarized as follows.
In unilateral presidential directives, American political development has seen a major expansion of presidential power that rests on vague justifications and has been relatively unchecked. This book documents and explains this development. The explanation has roots in the Constitution's ambiguity and the character of executive power, which is inherently resistant to strict limits. But the rise of more activist central governance from the Progressive era on—advanced first by Theodore Roosevelt and then by a variety of successors, with congressional, judicial, and popular acquiescence—has led to more extensive governance by unilateral presidential directives. This state of affairs has strengths as well as weaknesses, but it is hard to reconcile with any but very thin conceptions of democracy. Whether Americans really want or need governance by unilateral presidential directives is a central question for American politics today and will be in the years ahead.
The ABCs of UPDs
Some basic background information about unilateral presidential directives (UPDs) may be useful as a foundation for the above argument. Public awareness of unilateral presidential directives has been growing over the last dozen years or so, sparked in part by Barack Obama's use of executive orders and presidential memoranda to reverse many of his predecessor's policies, by several controversial unilateral directives that George W. Bush promulgated for the "war on terror" and other matters, and by what critics regarded as an unprecedented rash of late-term executive orders by Bill Clinton. But the historical record is much richer than recent public attention: presidents have long relied on unilateral directives to enact their preferences across a wide range of policy areas. Every president has issued executive orders or similar directives. Even William Henry Harrison, who was president for only thirty-one days, issued a proclamation, calling Congress into a special session.
Unilateral directives are documents that the president issues to direct the activities of the executive branch. As such, they afford presidents an independent means of controlling a wide range of governmental actions. More controversially, they may enable presidents to unilaterally enact their own policy preferences by a mere stroke of a pen, as they can serve to prompt congressional action, to preclude it, or to circumvent a recalcitrant Congress. As the Clinton White House adviser Rahm Emanuel explained in 1998, "Sometimes we use it in reaction to legislative delay or setbacks. Sometimes we do it to lead by example and force the legislative hand. Obviously, you'd rather pass legislation that can do X, but you're willing to make whatever progress you can on an agenda item." Or as Clinton White House Communications Counsel Paul Begala described unilateral presidential directives, "Stroke of the pen, law of the land. Kind of Cool."
Indeed, unilateral presidential directives can be an attractive resource for presidential action. And presidents generally either want to act or have to act. Most presidents enter the office with an ambitious agenda; and at times the nation requires action, and executive action may be the best or only option. Given various institutional constraints on other types of presidential action, unilateral presidential directives can be a convenient means to many ends.
Sometimes presidents use these directives for minor or noncontroversial matters, but other times they employ them for major or highly controversial policies. Better-known examples of unilateral presidential directives include Abraham Lincoln's Emancipation Proclamation, FDR's wartime internment of Japanese Americans, Harry Truman's desegregation of the military, JFK's creation of the Peace Corps, Bill Clinton's proclamations of new national monuments, and George W. Bush's executive orders for faith-based initiatives, the detention of suspected terrorists, and domestic spying. But there have been thousands of other unilateral presidential directives, and they have figured in many of the most important and controversial episodes in American political history. Unilateral presidential directives can therefore tell us a lot about American politics in general and the presidency in particular. As one of the earliest studies of executive orders put it, "The history of executive orders is, to a great extent, a narrative of the evolution of presidential power."
As the controversy about Bush's executive order for domestic spying and the other points noted above suggest, unilateral presidential directives are an important political phenomenon. And some basic background information about them suggests that they are also an intriguing political phenomenon, as their definition, justification, limits, format, numbering, and cataloging are surprisingly loose and inexact.
There are over two dozen different types of unilateral presidential directives. A study by the Congressional Research Service (CRS) in 2007 identified twenty-seven distinct types: administrative orders, certificates, designations of officials, executive orders, general licenses, homeland security presidential directives, interpretations, letters on tariffs and international trade, military orders, thirteen different types of national security instruments, presidential announcements, presidential findings, presidential reorganization plans, proclamations, and regulations. That list fails to mention presidential determinations and memoranda, so the total number of types of unilateral presidential directives may be twenty-nine. But the different names do not always designate different "tools": many directives are very similar in terms of their substance and authority, regardless of what the president decides to call them. And the definitions of many of these devices are ambiguous or even nonexistent.
Executive orders are the best known and most common type of unilateral presidential directive. They date to the earliest days of the republic, but the term "executive order" was not regularly applied to unilateral presidential directives until the late nineteenth century, and the first directives to be officially designated as executive orders were military orders by Abraham Lincoln. There is no official definition of what constitutes an executive order; there is no law—or even an executive order—that defines what an executive order is. A report by the U.S. House of Representatives in 1957 provisionally defined them as follows: "Executive orders are written documents denominated as such. . . . Executive orders are generally directed to, and govern actions by, government officials and agencies." This definition certainly captures part of what executive orders are, but it is inadequate, as its authors realized: "Essentially an Executive order is a written document issued by the President and titled as such by him or at his discretion." That circular reformulation indicates the difficulty of a strict definition. The following informal definition by a legal librarian is perhaps somewhat more helpful: "Executive orders are the formal means through which the President of the United States prescribes the conduct of business in the executive branch. Executive orders are presidential directives issued to federal government agencies or officials. An executive order is basically a document the President issues and designates as such." More to the point, executive orders are a primary means by which presidents impose their will by directing the activities of the U.S. government.
Proclamations are another main type of unilateral presidential directive. Prominent examples of proclamations include George Washington's Neutrality Proclamation and Abraham Lincoln's Emancipation Proclamation. Like executive orders, proclamations are generally written documents that the president issues to direct governmental action, and they lack a strict definition. Accounts disagree somewhat about whether or how these two devices differ. Legally and constitutionally, there is no difference between executive orders and proclamations. In Wolsey v. Chapman, 101 U.S. 755 (1880), the Supreme Court found no material distinction between them.
Congress has also treated executive orders and proclamations as being very similar, if not interchangeable. Both are subject to the publication requirements of the Federal Register Act of 1935, but presidents have often not distinguished between them or consistently used one label or the other. Therefore, the House report of 1957 concluded, "The difference between executive orders and proclamations is more one of form than substance." Similarly, in 1974, when the authors of a U.S. Senate report on executive orders asked officials at the Justice Department and the Federal Register how to distinguish between executive orders and proclamations, the officials acknowledged that the devices were interchangeable. According to the Senate report, "The arbitrariness of this system is illustrated by two instances of very similar situations of federally-enforced school integration in the South. In the first instance, President Eisenhower relied on an executive order, while, only a few years later, President Kennedy issued a proclamation."
Nevertheless, some scholars have suggested that proclamations and executive orders differ in terms of the substance of their use. For example, proclamations tend to be the tool used for presidential directions about international trade, and indeed many proclamations have addressed trade. But there are also instances of executive orders addressing international trade. Alternatively, some claim that proclamations often concern foreign affairs, while executive orders are usually used for domestic policy purposes. However, executive orders have at times been used for foreign affairs, as have proclamations for domestic affairs. Yet another possible distinction is that proclamations are the means by which presidents grant pardons. But there are some instances of executive orders being used for this purpose, too.
A fourth possible distinction is that proclamations are often used for hortatory or symbolic purposes, while executive orders are generally used for more substantive matters. George Washington issued the first proclamation on October 3, 1789, designating November 26 a day of public thanksgiving, and many subsequent proclamations have been used for similar symbolic purposes. The Senate report of 1974 noted the hortatory nature of many proclamations as a possible means of distinguishing them from executive orders: "In general, it appears that Proclamations are issued when it is felt that the decree is addressed to the public at large. They tend to be hortatory in nature, proclaiming national days of celebration or ceremonial events." However, this distinction is inexact, or at least anachronistic. According to Glendon Schubert, "It is generally true that before the Civil War, proclamations of the President were used most usually for hortatory and ceremonial purposes; but there has been an increasing tendency since that time for proclamations to serve as vehicles for direct substantive lawmaking." Furthermore, executive orders have at times been used for merely symbolic purposes.
Beyond these several problematic distinctions, there is one respect in which proclamations do appear to differ from executive orders and other unilateral presidential directives in their usage. Proclamations are the means by which presidents enact treaties. Strictly speaking, the Senate does not ratify treaties but rather gives its "advice and consent" for treaties that the president has negotiated. After the Senate approves a treaty, the president then ratifies it by formally notifying the other parties to the treaty, and this notification usually takes the form of a proclamation.
In addition to executive orders and proclamations, presidential memoranda constitute a third, important type of unilateral presidential directives. Like executive orders and proclamations, memoranda are written documents via which the president directs governmental actions. In one notable memorandum, Bill Clinton directed the secretary of health and human services to end the moratorium on funding research involving fetal tissue. Memoranda are very similar to executive orders, and hence also to proclamations. Phillip Cooper, a professor of public administration, calls memoranda "executive orders by another name," and in Lower Brule Sioux Tribe v. Deer, 911 F. Supp. 395, D.S.D. (1995), a circuit court suggested that memoranda are legally interchangeable with executive orders.
However, there is a key difference among these three types of unilateral presidential directives, as memoranda are not required be published, while executive orders and proclamations are. According to the Senate report of 1974: "If a document is not specifically designated as an 'Executive Order' or 'Presidential Proclamation,' the decision of whether or not it will be published as a part of the public record is left to the discretion of the President and his advisers. If he wishes a document to have 'general applicability and legal effect,' he will presumably have it published. If, however, the order is directed only to an official or an agency and does not purport to regulate the conduct of private citizens, there is no legal necessity for its publication. Most executive directives fall into this category." The Obama administration published most of its memoranda and also listed them on the White House Web site, but earlier administrations have been less forthcoming. The lack of publication and publicity mean that memoranda may be more difficult for Congress or courts or future presidents to reverse. Presidents are aware of this, and the number of memoranda has been increasing over the past few presidencies.
Beyond executive orders, proclamations, and memoranda, there are several other types of unilateral presidential directives that have at times been significant. For example, presidential determinations can be a means of exercising executive discretion in subjecting entities to regulations, as Bill Clinton's "Presidential Directive" 95-45 exempted the Air Force's secret "Area 51" military base in Nevada from environmental disclosure laws. So-called administrative orders have been used to create and organize the Federal Emergency Management Agency (FEMA), among other purposes. And the various types of national security directives are certainly important and afford presidents a powerful means of independent policymaking, but they are closely tied to the president's capacity as commander in chief and are usually veiled in secrecy. Most national security directives are not made public. According to the General Accounting Office (GAO), even the relevant congressional committees often do not see these directives.
The ambiguous number and nature of unilateral presidential directives, and the unclear relations among them, can easily lead to confusion. For example, George H. W. Bush's Executive Order 12,807 of 1992 was intended to direct the Coast Guard to return Haitian refugees to Haiti. However, the order did not specifically mention Haitian refugees per se. That detail was contained in a press release, which stated: "President Bush has issued an executive order which will permit the U.S. Coast Guard to begin returning Haitians picked up at sea directly to Haiti." In Sale v. Haitian Centers Council, 509 U.S. 155 (1993), the Supreme Court ruled that the press release was a sufficient articulation of the policy, as if inexactitude in unilateral presidential directives were to be expected and tolerated. Such confusion can be compounded by the fact that presidents are generally free to decide what to call a particular directive and can even create entirely new types of directives if they want.
Given the ambiguity among the different types of unilateral presidential directives, it makes sense to construe the topic broadly, rather than to focus narrowly on one particular type. This book focuses mostly on executive orders, proclamations, and memoranda. They are arguably the most common, most important, and most accessible types of unilateral presidential directives, and they are similar in their justifications and usage. Furthermore, I focus on nonmilitary directives, since military orders are a fairly discrete set of unilateral presidential directives, and insofar as they are rooted in the president's constitutional position as commander in chief, they may be less constitutionally controversial. I also exclude presidential signing statements from my analysis, since they also constitute a distinctly different type of unilateral presidential tool, and they differ from other directives in that they are less clearly legally binding.
If executive orders, proclamations, memoranda, and other unilateral presidential directives merely expressed the president's view, then they would be important but not necessarily determinative. However, these directives are not mere statements of presidential preferences; rather, they establish binding policies and have the force of law, ultimately backed by the full coercive power of the state. In Armstrong v. United States, 80 U.S. (13 Wall.) 154 (1871), the Supreme Court considered the legal status of a proclamation and decided that such directives are public acts to which courts must "give effect." In other words, in the eyes of the judiciary, unilateral presidential directives are just as binding as laws. In 1960, Senator Robert Byrd (D-WV) advised his colleagues, "Keep in mind that an executive order is not statutory law." Politically, that may be true, as unilateral presidential directives represent the will only of the chief executive and lack the direct endorsement of congressional majorities. But constitutionally and legally, a unilateral presidential directive is as authoritative and compulsory as a regular law, at least until such time as it is done away with by Congress, courts, or by a future unilateral presidential directive.
The justifications for unilateral presidential directives and the limits on them are essentially the same, namely Congress and the Constitution. Bill Clinton's chief of staff, John Podesta, alluded to both of these justifications in defending Clinton's extensive use of unilateral directives after the Republican congressional victories of 1994: "The president has had a sense that he has authority that comes from statutes as well as from the Constitution that he can exercise to make progress for the American people."
Many unilateral presidential directives are authorized by the legislative branch. Congress often delegates power to the president or leaves it to the president to determine when a certain state of affairs exists or when a certain preauthorized action is warranted. Many unilateral presidential directives are issued pursuant to this sort of authority. Additionally, presidents have at times creatively invoked long forgotten statutory grants of power to justify unilateral directives that Congress never anticipated. In addition to congressional authorization, unilateral presidential directives can also be justified by the Constitution. Some presidential actions are clearly provided for in the Constitution, but most unilateral presidential directives are not. Instead, they are justified by various broad grants of power in Article II and other general aspects of the Constitution's treatment of the executive.
Just as Congress and the Constitution justify unilateral presidential directives, they also limit them. Unilateral presidential directives are lawlike, but laws have priority over unilateral presidential directives. If Congress passes a law, then a president cannot override it by a unilateral directive. Furthermore, Congress is free to override a particular unilateral presidential directive, or even to limit or eliminate certain types of unilateral presidential directives. For example, the Federal Pollution Control Act Amendments of 1972 (Pub. L. 92-500, 96 Stat. 816, now commonly referred to as the "Clean Water Act") effectively nullified the system of water pollution permits that Richard Nixon established via Executive Order No. 11,574 of 1970. Congress may also vote to withhold the funds that would be necessary to implement a unilateral presidential directive, which is tantamount to nullifying the directive. For example, on his second full day in office, Barack Obama signed Executive Order No. 13,492 to close the detention center at the U.S. naval base in Guantanamo Bay, Cuba. However, four months later, the Senate voted 90-6 to withhold $80 million that the Obama administration had requested to pay for the closure, thereby effectively blocking it.
Aside from Congress, the Constitution also limits what presidents can do via unilateral directives. The Supreme Court and other courts have at times struck down unilateral presidential directives as unconstitutional, and various limits on unilateral presidential directives have evolved in case law. As circuit judge Diana Gribbon Motz wrote in Al-Marri v. Wright, 433 F. Supp. 2d 774 (2006), about Bush's unilateral directive for the detention of suspected terrorists without recourse to the writ of habeas corpus, "The President cannot eliminate constitutional protections with a stroke of a pen."
While Congress and the Constitution are the main limits on unilateral presidential directives, two other limits should be noted, namely politics and presidents. The use of unilateral presidential directives is, of course, subject to various political considerations and limits. Even if a potential unilateral presidential directive might be constitutional and even authorized by Congress, it might nevertheless be so problematic politically as to be impracticable or impossible. And future presidents are free to issue unilateral directives to overturn earlier unilateral presidential directives. Calvin Coolidge alluded to this in his second State of the Union address, when he asked Congress for a law to make revisions to civil service employment, rather than have to rely on a mere unilateral presidential directive: "Otherwise the Executive order of one administration is changed by the Executive order of another administration, and little real progress is made."
However, the limits on unilateral presidential directives are less confining that one might think. First, there is a great deal of leeway between the limits of the Constitution and the will of Congress. As long as a unilateral presidential directive does not clearly violate the Constitution or a law, it is potentially legitimate. Second, while the legislative and judicial branches can overturn unilateral presidential directives, they seldom do so.
In terms of the judiciary's curtailment of unilateral presidential directives, Kenneth Mayer reports that "between 1789 and 1956, state and federal courts overturned only 16 executive orders." The number overturned in more recent years has been correspondingly small: according to Terry Moe and William Howell, of the roughly 4,000 executive orders issued between 1942 and 1996, only 86 were challenged in court, and presidents won in 86 percent of those few cases. The rare occasions when courts do overturn executive orders may be dramatic, but they are very much the exception rather than the rule. By some accounts, even the Court's decision in Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579 (1952), against Harry Truman's seizure of the steel industry was more of a fluke than a broad, principled curtailment of unilateral presidential action. It is not clear whether several court decisions against George W. Bush's directives for suspected terrorists detained in Guantanamo constitute a major rebuke or are of more limited or narrow significance.
Similarly, congressional attempts to reverse particular executive orders are rare and seldom successful. According to Neil Kinkopf, a special assistant to the Office of Legal Counsel at the Department of Justice under Clinton, "In the 25 years from January 1973 through the end of 1997, legislation to overturn an executive order was introduced on 37 occasions." Moe and Howell report that only 3 of those 37 attempts were successful. According to Mayer, "Congress [has] explicitly invalidated an executive order of any substance" only twice since 1970. And Adam Warber finds that while Congress terminated 206 executive orders between 1936 and 2001, those amounted to only 4.7 percent of all significant executive orders issued during that time.
Indeed, unilateral presidential directives are fairly safe and durable means of policymaking. According to one study, 40 percent of executive orders issued between 1959 and 1998 were still in effect as of 2008, and the average executive order in that time remained unchanged for twelve years. Altogether, the above figures suggest that executive orders and other directives give presidents a means of unilateral policymaking that may be largely immune from the checks of the other two branches. And that points to one of the ways in which unilateral presidential directives can be constitutionally troubling.
Unilateral presidential directives are arguably in tension with two of the most fundamental aspects of the U.S. constitutional architecture: the separation of powers and checks and balances. Insofar as unilateral presidential directives enable the executive to legislate unilaterally, they violate a strict separation of powers. And insofar as the other two branches have often been unable or unwilling to resist or reverse these executive edicts, unilateral presidential directives also call into question the efficacy of traditional checks and balances. But despite these constitutional concerns—and because courts have accepted and even affirmed them, and Congress seldom overrides them and usually passively acquiesces to them—unilateral presidential directives are a legitimate and binding means of presidential policymaking.
Even though unilateral presidential directives are important for policymaking and can be controversial both politically and constitutionally, they have often not been subject to systematic processing or careful accounting and bookkeeping. The procedures by which presidents have promulgated unilateral directives have varied greatly. Sometimes unilateral presidential directives are years in the making and are debated at great length before being signed, but sometimes they are issued with no advance notice. Sometimes they are initiated by the president himself, generally after consultation with top advisors or officials, but other times they emerge from deep within the vast executive bureaucracy. For example, consider Lyndon Johnson's Executive Order No. 11,377, which authorized the U.S. Tariff Commission "to monitor the annual consumption of whisk brooms in America, noting the types, numbers, and uses made of these brooms." While Johnson signed the order, he did not initiate it or give it much consideration or effort.
Presidents have periodically tried to impose a regular process for issuing executive orders and other unilateral directives, and they have done so by issuing unilateral directives. Herbert Hoover issued an executive order for this purpose in 1929, as did Truman in 1948 and JFK in 1962. Truman's and Kennedy's directives were both entitled "Preparation, Presentation, Filing, and Publication of Executive Orders and Proclamations" and both outlined a process for issuing executive orders that went from the Bureau of the Budget to the attorney general to the Federal Register and finally to the president. Later, executive orders from Johnson, Carter, and George W. Bush made further changes to these procedures. The attorney general's office subsequently delegated its review authority to the Office of Legal Council, which now reviews most proposed unilateral presidential directives. Additionally, the White House counsel now routinely reviews executive orders and at times helps to draft them. Despite these measures, there is still periodic confusion about how exactly presidents issue them. For example, Kennedy's and Carter's first executive orders were criticized as not following the usual procedures.
Another quirk about unilateral presidential directives is their format. Early executive orders, for example, were often rather casual. In 1851, Millard Fillmore wrote notes on parts of a map of proposed public grounds in Washington, D.C., in order to approve the plan. Those notes could well be construed as executive orders. Many other early executive orders consisted only of presidential endorsements of legal briefs or memos. According to the Senate report of 1974, "Successive Presidents wrote, 'Approved,' or 'Let it be done,' or other short comments and these jottings sufficed to stamp a proposal with the authority of the Presidential imprimatur." While these informal orders were once the norm, there were also more formal early ones, "vested with the full trappings and dignity associated with official national documents—the use of highly formulaic language, and the impression of the Great Seal of the United States executed by the Secretary of State."
Ulysses Grant established an official format and style for executive orders in 1873, but as late as 1906, "orders were treated most informally—for example Executive order 396 for that year is not even dated." More format changes were enacted, via executive order, in the 1920s, and a 1931 executive order called for further changes in form and style. In 1936, FDR's Executive Order No. 7,298 established more regulations for the format of executive orders, and these regulations were revised by other executive orders in 1948 and 1962. The later guidelines stipulated that orders should have a title and should cite the authority under which they are issued. But these norms were not always followed then, nor are they now.
In more recent years, executive orders and proclamations have tended to be phrased like laws, usually with a preamble that cites a problem to be addressed and the sources of the president's authority to act, via a series of clauses that precede the directive saying "Whereas. . . ." And recent executive orders often have concluded with a legalistic disclaimer, saying that nothing in the order shall be construed to impair the statutory authority of a department or agency and denying that the order establishes any enforceable right or benefit.
Beyond their processing and format, there is also significant uncertainty as to the number of unilateral presidential directives. By some estimates, presidents have issued over 17,000 executive orders and 8,300 proclamations since 1789. But the precise numbers of these directives are simply not known. This is because for most of the nation's history nobody bothered to keep track of them. Routinized numbering of executive orders did not begin until 1907, when the State Department began assigning chronological numbers to the executive orders that it had collected. However, by one account, less than half the presidential orders were filed with the State Department. Furthermore, the numbering began with an executive order from Lincoln dated October 20, 1862: all executive orders from the seventy-three years before then (even those that were in the State Department records) were omitted. Earlier executive orders were later inserted into the list, using numbers that fell between existing numbers, such as "1A" or "28-1." Some executive orders were and still are secret, but many classified executive orders were inserted into the regular numbering with a letter indicating their existence and their classified nature.
Despite these efforts to fill in the gaps, "some unnumbered orders exist[ed] as late as 1941." A project in 1942 by the Works Projects Administration (WPA) sought to rectify some of these problems and successfully collected over 1,500 unnumbered executive orders dating back to 1789, but there are likely thousands more. According to one scholar, estimates "of the number of pre-1907 orders which were never serialized and deposited for the record range from 15,000 to 50,000." In terms of proclamations, 4,500 or more may have been lost. For these reasons, even cautious estimates of the numbers of unilateral presidential directives must be regarded with some skepticism.
Aside from their number, there is no comprehensive, authoritative compendium of unilateral presidential directives. As their incomplete numbering indicates, many executive orders were not originally published, and many old executive orders are still unpublished. In August 1869, Ulysses Grant issued a proclamation recognizing insurgents in Cuba, but this proclamation was never published, apparently because it was pigeonholed by Secretary of State Hamilton Fish. One might think that lack of publication might pose a problem for the effectiveness of unilateral presidential directives, but in Lapeyre v. United States (1873) the Court found that executive orders are valid even if they are not published. Obviously, this is true of executive orders that are classified or secret. According to the General Accounting Office, of the 1,042 directives on national security matters that presidents issued between 1961 and 1988, less than one-quarter were published.
In 1895, Congress passed the General Printing Act, under which the "Documentary Catalog" began. This act sought to catalog every executive order ever printed by the Government Printing Office in "slip form" (that is, as a single sheet of paper), whether the orders were included in the numbered series or not. Also in 1895, Congress authorized Rep. James Richardson (D-TN) to compile a collection of major presidential documents starting with George Washington. Richardson omitted executive orders from his first volume of documents in 1895, but he included them in his second volume in 1897. Altogether, Richardson's Compilation of the Messages and Papers of the Presidents contains eight large volumes and runs through the presidency of William McKinley.
In 1927 Felix Frankfurter wrote, "The formation and publication of executive orders and rules and regulations are in this country still in a primitive stage, which only serves to render more portentous the operation of these forms of law." The inadequate bookkeeping of executive orders became acute after FDR issued so many, and in 1935 "the government discovered that it had brought an indictment and taken an appeal to the Supreme Court without realizing that the portion of the regulation on which the proceeding was based had been eliminated by an executive order." To prevent this sort of confusion, Congress passed the Federal Register Act (44 U.S.C. Chapter 15) in 1935 requiring that executive orders and proclamations be published in the Federal Register. This greatly helped to facilitate the basic bookkeeping and cataloging of unilateral presidential directives, but problems have persisted. Again, anything not designated as an executive order or a proclamation need not be placed in the Federal Register. Some memoranda are published there, but some are not, and directives dealing with national security are generally not published anywhere.
There are several online compilations of unilateral presidential directives, which are good if incomplete resources. Executive orders from the present day back through FDR's presidency are now available online via the National Archives and Records Administration and the Federal Register. The American Presidency Project—a collection of over one hundred thousand documents relating to the presidency maintained by the University of California at Santa Barbara—has an extensive online database of many executive orders and proclamations. The Avalon Project at Yale Law School contains many early directives. And even Wikipedia has many executive orders and proclamations.
In sum, the lack of a clear definition of most unilateral presidential directives; the variable nature of their processing and format; and their flawed numbering and incomplete cataloging and publication all suggest that they are an unusual and intriguing political phenomenon. The casual accounting and bookkeeping of unilateral presidential directives may be surprising given their importance. And it may be the result of administrative sloppiness, legitimate needs for discretion and secrecy, and even intentional executive obfuscation, but lax congressional oversight and the haphazard nature of the evolution of unilateral presidential directives have also played roles in this state of affairs.
Presidents can and do issue unilateral directives at any time during their administrations and for a wide variety of purposes. However, some general patterns have emerged over the years. For example, presidents tend to issue a lot of unilateral directives early in their terms, particularly in recent decades. Of the forty-one executive orders that Obama issued in his first year as president, he issued fourteen of them during his first three weeks in office (and also issued eleven memoranda and two proclamations during that time), often to reverse policies that George W. Bush had enacted via unilateral directives. Presidents commonly issue these orders in their first day in office, and many candidates for the presidency often promise that if elected they will issue a particular directive as their first official act as president. Scholars and journalists are increasingly aware of this tendency. Just before Barack Obama was inaugurated, Slate magazine noted that "newly minted presidents have often used executive orders to set the tone for their administration right out of the gate," and it identified "the top ten Bush executive orders that Obama should scrap immediately." Presidents also tend to issue a lot of unilateral directives late in their terms. In part, this practice derives from presidents' desires to cement their legacies before leaving office. In some cases, it may also be a means of scoring partisan political points by enacting policies that could be politically embarrassing for a successor of the other party to overturn.
Aside from regularities in the timing of unilateral presidential directives, there are also other patterns in their usage. For example, political factors like the levels of a president's popularity, the president's level of support in Congress, and the presence of divided government may be associated with the number of directives issued, though not always in the ways one might expect. Party affiliation also seems to matter, as there are somewhat different patterns of unilateral directive usage for Democratic and Republican presidents. Additionally, there have been different substantive patterns of usage concerning unilateral presidential directives over the years, with different issues dominant in different eras, as I will show.
Despite being politically important, constitutionally controversial, and procedurally intriguing, unilateral presidential directives have received surprisingly little scholarly attention, at least until recently. Indeed, the few scholarly accounts that do exist invariably note the paucity of such scholarship. For example, the political scientist Kenneth Mayer wrote in 1999 that "despite the importance of executive orders, only a handful of studies have considered their use and significance." Similarly, a law review article from 2000 stated that "to date, there is very little scholarship on executive orders." And writing about unilateral presidential directives in 2002, Phillip Cooper claimed, "Curiously, the literature on the presidency largely ignores the tools of presidential direct action."
This neglect may be due to the tendency of presidency scholarship to focus on character, personality, and leadership style instead of the constitutional and legal bases of presidential power. The dominance of methodological approaches derived from behavioralism may have caused academics largely to ignore the constitutional and legal bases of the presidency. With a few exceptions, scholars of the presidency simply have not devoted sufficient attention to the constitutional and legal bases of presidential power in general, let alone to the use of unilateral presidential directives in particular.
By most accounts, mass media have also paid little attention to unilateral presidential directives. For example, Adam Warber contends that "media reports on these directives appear to be minimal." But at times, the press has noted particular unilateral presidential directives. William Howell reports that between 1945 and 2001, the New York Times "granted page-one coverage to a total of 290 executive orders, or 7.7 percent of all orders issued" in those 56 years.
Some of the more substantial treatments of unilateral presidential directives are in the form of governmental publications, which provide basic introductions to the topic. For example, the WPA undertook a study of executive orders in 1942 and published two lists of them (numbered and unnumbered) with a brief introduction. In 1957, the House Committee on Government Operations issued a report titled "Executive Orders and Proclamations: A Study of a Use of Presidential Powers," and in 1974 the Senate Special Committee on National Emergencies and Delegated Emergency Powers issued a report titled "Executive Orders in Times of War and National Emergencies," which included another report titled "Executive Orders: A Brief History of Their Use and the President's Power to Issue Them." Since then, the Congressional Research Service (CRS) has produced several brief overviews on executive orders and other unilateral presidential directives.
Much of the academic literature on unilateral presidential directives is not authored by political scientists. For example, conservative think tanks have produced a couple of general overviews. And legal librarians have written a couple of pieces to help researchers find unilateral presidential directives. Most of the academic literature on unilateral presidential directives appears in law reviews, and it can be divided into four categories. First, there are several overviews of the general jurisprudence on the topic. A second branch of the small law review literature more narrowly addresses the justiciability (for example, concerns with standing and enforcement) of executive orders. A third type addresses the use of unilateral directives in a particular policy area. And a fourth type consists of in-depth case studies of a specific directive.
In political science, the literature on unilateral presidential directives is quite small. Some American politics textbooks and basic reference books on the presidency mention executive orders and proclamations, but few treat the topic at any length. Over the last fifteen years or so, there have been just over two dozen journal articles about unilateral presidential directives. Half of these are quantitative or employ regression analysis or formal models, though some also contain accessible overviews of the topic. There are just over one dozen articles on unilateral presidential directives in political science and related journals that are qualitative. Some of these concern a particular unilateral presidential directive. Others concern a particular president. One addresses a particular policy area. And several are authored or co-authored by Phillip Cooper and concern the differences among various types of unilateral directives and their relation to public administration and administrative law. In addition to these qualitative journal articles, there have also been several book chapters on the general topic. There is a little bit of scholarship that considers unilateral executive decrees in comparative perspective. There is also some literature that briefly notes executive orders in the course of examining the overall role of the executive in policymaking. And there have been a number of dissertations and academic conference papers on executive orders and other unilateral presidential directives.
There are presently only a half-dozen books on unilateral presidential directives, and they vary considerably in their focus and their utility. The first of these was Ruth Morgan's The President and Civil Rights (1970), a qualitative study of how presidents from Franklin Roosevelt to Lyndon Johnson used executive orders for civil rights. Morgan found that these executive orders were largely successful in enacting desired change and also served to prompt later congressional action on civil rights. Generalizing from this one policy area, she concluded that executive orders can provide an effective means of presidential policymaking in the face of congressional inaction, provided the president has sufficient public support.
Kenneth Mayer's With the Stroke of a Pen: Executive Orders and Presidential Power (2001) examines executive orders in multiple policy areas. Mayer's book surveys the use of executive orders for institutional organization, foreign affairs, and civil rights and assesses its implications for conceptions of presidential power. Mayer purports to utilize the perspective of "new institutional economics," but his discussion actually contains little rational choice institutionalism or formal theory and is an accessible and informative general introduction to executive orders. Phillip Cooper's By Order of the President: The Use and Abuse of Executive Direct Action (2002) examines several different tools of "executive direct action," by which he means unilateral presidential directives. Cooper's book builds on his earlier articles, as he compares executive orders and proclamations with other types of unilateral presidential directives in terms of their usage by recent presidents and their relative strengths and weaknesses.
William Howell's Power without Persuasion: The Politics of Direct Presidential Action (2003) draws on aspects of game theory and new institutional economics and develops parts of a couple of earlier articles to present a spatial model of unilateral presidential action, complete with algebraic proofs. Adam Warber's Executive Orders and the Modern Presidency: Legislating from the Oval Office (2006) is primarily a quantitative study based on an evaluation of the policy content of every executive order listed in the Federal Register from 1936 through the Clinton presidency, which he employs to ascertain the extent to which presidents have been strategic in using executive orders. And Ricardo Jose Pereira Rodrigues's The Preeminence of Politics: Executive Orders from Eisenhower to Clinton (2007) examines executive orders from Eisenhower through Clinton in three policy areas: equal employment opportunities, regulatory review, and environmental policy.
In sum, the existing literature on unilateral presidential directives is small. The gap in the literature is no longer the gaping hole that it was several years ago, but the presidential use of unilateral directives remains an understudied topic. The existing treatments tend to be overviews, topical, brief, or quantitative, and many of them fail to consider unilateral presidential directives other than executive orders. Furthermore, they seldom display much historical sensitivity; apart from how and why recent presidents have used unilateral directives, there is the issue of how they came to be able to do so. There is very little written about the development of unilateral presidential directives—where they came from, how they were used before recent times, and what their overall impact has been. To their credit, the authors of a leading formal model of unilateral presidential action acknowledge that historical institutionalism could be a good approach to the topic, but as yet there is no such account. This book seeks to begin to remedy those shortcomings.
The Plan of the Book
Having suggested that unilateral presidential directives are an important, intriguing, timely, and understudied topic, it now remains to explain the purpose and plan of this book in some detail. My aim is to explore where these directives came from; how they were legitimated, accepted, and routinized; and what their impact has been. I seek to explain how presidents came to be able to make law by a mere stroke of the pen, how they have used this power, and what its implications are for understanding the presidency, American politics, and political development. If the entire project were to be described in the traditional social science terms of independent and dependent variables, then one formulation would have the dependent variable as executive power, with the independent variable as personal agency, and with unilateral presidential directives as an intervening variable, but the account presented here is far more nuanced and complex than that. The historical development described in the following chapters begins with constitutional ambiguity and presidential initiative, followed by judicial acceptance and legislative acquiescence, leading to greater presidential use of unilateral directives and only sporadic judicial or legislative resistance, something that may be difficult to square with the norms of constitutional democracy.
More specifically, my analysis of unilateral presidential directives consists of two main parts, each of which is multifaceted: I argue that the presidential use of unilateral directives can best be understood in terms of two developmental preconditions and four developmental stages. I devote a chapter to each of these six points.
The developmental preconditions may be regarded as enabling conditions that made possible but did not cause the development of unilateral presidential directives. In other words, they effectively created or cleared the constitutional space for significant usage of unilateral presidential directives. The two preconditions are constitutional ambiguity and early judicial sanction, which correspond roughly to constitutional theory and constitutional law respectively, and they are the subjects of the second and third chapters.
Chapter 2 considers the place of unilateral presidential directives in the American constitutional order. Since the U.S. Constitution does not specifically mention unilateral presidential directives, their constitutional status is largely bound up with the broader question of the nature and scope of executive power. I examine the Constitution's treatment of the executive and find that it is equivocal. Some clauses may support wide presidential powers, perhaps including the power to issue various unilateral directives, but other clauses suggest a highly limited executive, and unilateral presidential directives seem to be in tension with the overall architecture of separation of powers and also checks and balances. The Constitution neither clearly provides for unilateral presidential directives nor clearly precludes them, and its overall treatment of executive power is highly ambiguous.
In order to elucidate the ambiguous nature of executive power, I then turn to related resources in political theory and also English and early American history. I find that these considerations show the possibility of broad executive powers like unilateral directives being compatible with a limited, constitutional executive office, but they fail to resolve the ambiguity. Faced with a persistent plurality of plausible competing conceptions of executive power, I make two claims, one theoretical and one methodological. First, partly following Harvey Mansfield, I argue that the constitutional ambiguity of executive power is ineliminable, as it reflects ambiguities inherent in the nature of constitutionalism and also executive power. Second, I argue that therefore the best way to understand the constitutional propriety of unilateral presidential directives is to examine how the matter has developed in the political realm, specifically in terms of the interbranch struggle over constitutional politics. This chapter thus justifies my methodology, which combines the traditional public law approach with more recent scholarship on historical institutionalism and American political development. The small scholarly literature that exists on unilateral presidential directives largely ignores historical considerations, and the topic of how these directives developed seems well suited for an American political development approach.
Given the contingent nature of executive power in general, and therefore of unilateral presidential directives in particular, I proceed to examine the evolution of case law on the subject in Chapter 3. I find that the judiciary first endorsed the constitutional legitimacy of unilateral presidential directives in two maritime cases in the early nineteenth century: Little v. Barreme (1804) and The Orono (1812). In the first case, Chief Justice John Marshall acknowledged the propriety of executive orders in general but struck down the particular order because it contradicted a policy that Congress had previously set. In the second case, the judiciary again affirmed the propriety of executive orders in general but struck down the particular order for violating a specific constitutional clause. Thus, courts had established the legitimacy of executive orders, subject to constitutional and congressional restrictions, by the time the nation was only 23 years old, fully 140 years before the Court famously reiterated those same principles in striking down Harry Truman's executive order seizing the steel industry in Youngstown Sheet & Tube Co. v. Sawyer (1952). It may be that the Court accepted executive orders as constitutionally permissible because it thought they would be limited to similar naval issues or other military or national security matters. Yet, the judiciary's endorsement of executive orders continued after the early 1800s, such that by Lapeyre v. United States (1873) and Wolsey v. Chapman (1880), it had explicitly acknowledged that executive orders were being used to make binding law in a variety of contexts.
After exploring these two developmental preconditions in Chapters 2 and 3, I turn to the actual development of unilateral presidential directives in Chapters 4 through 7. Textual ambiguity and judicial acceptance permitted but did not ensure the development of unilateral presidential directives; that depended on circumstance, ideology, and especially individual presidential initiative. I begin by considering a measure of the number of executive orders issued, specifically the average number issued per year by each president from George Washington through Barack Obama. This is based on the best data available, and it compensates for the different lengths of different presidencies and thereby facilitates meaningful comparisons across different administrations (see Figure 1). Four areas of this picture of the development of executive orders seem to stand out: (1) a period of very low usage of executive orders from the founding of the country through the nineteenth century, (2) a sharp rise under Theodore Roosevelt, (3) a less dramatic but steady increase from Taft through FDR, and (4) a marked decline in usage after World War II.
For the reasons mentioned earlier, the numbers of executive orders are not entirely reliable. And of course executive orders are only one type of unilateral presidential directive. Moreover, as Phillip Cooper has noted, "mere counting tells us little," in that it fails to distinguish between important and unimportant unilateral presidential directives. Nevertheless, the statistical changes are fairly clear and may serve at least as the starting point for deeper, more careful analysis. Indeed, I find that apparent statistical breaks correspond to fairly discrete stages in the development of executive orders and other unilateral presidential directives, with different forces driving their development at different times. I then devote a chapter to examining the development of unilateral presidential directives in each of these four phases.
The internal organization of these four chapters varies, with divisions based on substance, significance, and ideology, but the discussion within each rubric is chronological. Some directives are mentioned only briefly or included in a list of similar directives, while others are examined at length. Much of this could be contested; the chapter divisions, the thematic divisions within each chapter, the inclusion or exclusion of specific directives, and the amount of space devoted to any particular directive or set of directives could all be different. But this organizational scheme is intended to enhance our understanding, to make sense of a long, strange history, and to provide an accessible narrative of each period and the overall development of unilateral presidential directives.
Chapter 4 considers the development of unilateral presidential directives from the founding of the country through the nineteenth century. Notwithstanding the fact of early judicial acceptance of unilateral presidential directives, presidents did not much use this new policymaking tool for quite some time. Few scholars have paid much attention to early unilateral presidential directives, and the few who have done so tend to discern little of note. But I argue that early unilateral presidential directives were more numerous and more important than conventional wisdom holds. I divide the discussion of early unilateral presidential directives into three groups, based on their notoriety: small, medium, and large. The "large" directives include those that are best known, such as George Washington's Neutrality Proclamation and Abraham Lincoln's Emancipation Proclamation. The "small" ones include various little known directives for public lands, Indians, and government administration. And the "medium" group includes directives for domestic unrest, Reconstruction, Mormon insurrection, participation in foreign conflicts, and labor. All together, these early unilateral presidential directives served to solidify politically the legitimacy of the devices, setting the stage for much wider usage later.
The nature of the use of unilateral presidential directives changed dramatically with Theodore Roosevelt, and Chapter 5 is devoted to TR and the rise of unilateral presidential directives. In executive orders and proclamations, TR found the perfect means by which to implement his "Stewardship" theory of presidential leadership. TR used executive orders in far greater numbers and for greater purposes than any previous president, often in the face of strong congressional opposition. Although TR did not turn to executive orders and proclamations to enact progressive reforms of trusts and tariffs, and his uses of the devices to advance new coinage and phonetic spelling were reversed, he successfully utilized them to create dozens of national monuments and wildlife refuges, and to change governmental regulations and structure, and he used the mere threat of an executive order to resolve the Pennsylvania coal strike in 1902. TR's conscious choice to use unilateral presidential directives for a variety of purposes served to establish a new understanding of the constitutional executive as a leading force in Progressive government, an understanding that was contested but finally triumphed under later presidents. Thus, TR plays the leading role in this drama, and as such he alone gets his own chapter.
In Chapter 6, I examine the development of unilateral presidential directives from the end of the first Roosevelt through the second Roosevelt, that is, from Taft through FDR. I argue that the practice of presidents regularly using unilateral directives for significant purposes was not just an aberration under TR but rather became further institutionalized and entrenched in this period. William Howard Taft articulated a more reserved approach to executive leadership than TR, yet Taft issued executive orders at roughly the same rate as his predecessor. Woodrow Wilson used unilateral directives in greater numbers and for more significant purposes during World War I, often with legislative sanction. The three Republican presidents of the 1920s continued to issue a great number of executive orders and proclamations, but the use of unilateral presidential directives rose to new heights with Franklin Roosevelt, who issued far more unilateral directives than any other president. FDR's barrage of executive orders was the impetus for the Federal Register Act and the regular publication of presidential executive orders and proclamations. I conclude that FDR fully and perhaps permanently institutionalized the Progressive view of the presidency and the practice of presidents regularly issuing many unilateral directives for significant purposes, norms that TR first introduced.
In Chapter 7, I examine unilateral presidential directives from the postwar period to the present day, focusing on five main substantive areas: war and national security, labor, race, administration, and environment. I also consider the dramatic decline in the number of executive orders issued in the postwar period, and I explore several possible reasons for it. Despite the decline in the number of executive orders, I find that presidents in this period continued to use unilateral presidential directives to further their political and policy goals. In many cases, these uses followed precedents and patterns first established by TR. Even with evolving issue areas, periodic congressional attempts to resist unilateral presidential directives, and the rare court case striking down a directive, there is every indication that presidents will continue to use unilateral directives for a variety of important and controversial purposes.
Chapter 8 explicitly discusses some of the analytical threads that are implicit in the preceding seven chapters. I offer some thoughts about the implications of my analysis for broader debates in both academic political science and real world politics.
Much of this book is devoted to little-known actions of presidents from long ago, but its goal is to shed light on an aspect of presidential power that has long had a great impact on the presidency and American politics and that continues to the present day. What follows is part constitutional analysis and part history lesson in the course of exploring how presidents have unilaterally enacted their policy preferences by a mere stroke of a pen, and why it matters.