While the presidential office has been transformed into a representative institution, it lacks proper organs for the exercise of that function. . . . [N]o constitutional means are provided whereby he may carry out his pledges.
—Henry Jones Ford, The Rise and Growth of American Politics: A Sketch of Constitutional Development, 1898
Introduction
The idea that the presidency is a representative institution—that the president represents the people—developed over time. The framers of the Constitution made the popular connection indirect in all institutions except the House of Representatives. Representation in the Anglo-American political tradition had developed as a means to present local grievances to the nation, formulate legislation, and check executive power.Footnote 1 As originally conceived, executive authority had a stronger constitutional basis than a popular one. Selected by the Electoral College, the president—the Chief Magistrate—was to be an independent officer who would execute the laws, serving also as a check on potential legislative tyranny.Footnote 2 It follows that the idea of presidential representation is, constitutionally speaking, a stretch, and that the more fully it is expressed institutionally, the more problematic its intrusions upon the government’s intended formal arrangements are likely to be. The “proper organs” for presidential representation, to use Henry Jones Ford’s phrase, will, by degrees, scramble envisioned constitutional roles. This article takes a fresh look at the Budget and Accounting Act of 1921 [BAA] in this light. The BAA was the first instance in which Congress passed a law that relied upon the idea of presidential representation as its core design assumption.
Linking Presidential Representation to Institutional Change
On its face, the idea of presidential representation is simple: the president is the only officer of government (besides the vice president) selected by a national constituency.Footnote 3 The literature on this subject takes that observation as a point of departure, testing whether and how well the office represents the whole citizenry. With representation operationalized as either centrism or universalism, the literature finds, by and large, that the presidency falls short on both counts.
Focusing on a standard of centrism, which posits that presidents should primarily respond to the national median voter as opposed to their party’s median voter, one group of scholars has found that presidents have primarily behaved as partisans.Footnote 4 Emphasizing a standard of universalism, which asserts that presidents should represent the national interest as opposed to any particularistic interests, a second group has found that presidents have instead prioritized key states to build an Electoral College majority.Footnote 5
These contributions have provided valuable insights. However, casting representation as a normative standard by which to judge the office’s performance misses the idea’s more profound impact on government and politics. Whatever its limitations, this is an idea that spurred institutional development, rearranging political authority, especially between Congress and the presidency. Indeed, one political theorist advises that we should consider not just what representation is, but also “the effects of its invocation.”Footnote 6 I propose to study the effects of a particular claim of representation, to examine the institutional forms that attached themselves to it, and to consider their inherent limitations as constitutional adaptations.Footnote 7
A recent proposal put forth by two political scientists brings these issues into sharp relief. Like Henry Jones Ford a century earlier, William Howell and Terry Moe point directly to the Constitution as the source of governing problems in modern America and offer greater presidential power as the solution. Decrying the parochialism of Congress and the separation of powers, they propose a constitutional amendment providing greater presidential control over agenda setting: the president would propose bills to Congress for a straight up-or-down majority vote within a limited period of time.Footnote 8 This envisioned change explicitly assumes that the president will represent the national interest relatively more than members of Congress, and therefore will submit cohesive, nationally oriented legislation for Congress’s consideration.Footnote 9 Their proposal illustrates two developmental implications. First, it exemplifies how the idea of presidential representation can influence proposed institutional reforms. Second, whatever may be the feasibility of such a proposal, it vividly illustrates the mismatch between the constitutional structure of authority and the idea of presidential representation. The demands of this idea are likely to prove insatiable, for they can only be met by discarding basic principles of this government’s organization.
Indeed, the proposal indicates the kinds of remedies this idea portends.Footnote 10 Most obviously, it would require a presidential agenda setting power to ensure that a national perspective is formally put before Congress. The specifics of this power can vary along a continuum, ranging at least from the weaker power to recommend measures found in Article II, to a stronger power of regularly submitting detailed bills, to a more sweeping power, along the lines of the Howell and Moe proposal, of a virtual monopoly on agenda setting by prohibiting congressional amendments to presidential initiatives and requiring a congressional up-or-down vote. The idea of presidential representation is likely to also require that the president’s perspective be presented to Congress, rather than the views of cabinet officers or other executive officials. Therefore, presidents should have some measure of unitary control over the executive branch, attempting to ensure that other executive officers cannot submit or advocate for alternative proposals.Footnote 11
Some scholars have considered the importance of the idea of presidential representation in examining political developments in the nineteenth century. Thomas Jefferson used it to fashion and promote a national political party; Andrew Jackson used it to claim an electoral mandate for constraining congressional action with the removal of federal deposits from the National Bank. Jackson’s idea of an electoral mandate for presidential action on a campaign commitment was contested between supportive Democrats and suspicious Whigs.Footnote 12 It was legitimized when Abraham Lincoln—previously a Whig opposed to Jackson’s view—claimed a mandate from the 1864 election.Footnote 13 The institutional purposes attached to the idea in the nineteenth century—presidential vetoes on policy grounds, use of the removal power, and claims of election mandates—were, however, different from those pursued in the twentieth.
In the twentieth century, presidency-oriented reformers increasingly associated the concept with two other entailments that combined to form the basis of the institutional presidency—a formal license for agenda setting and greater executive organizational capacity. Both entailments were debated and provided for in the BAA of 1921. Making the president responsible for initiating the budget process and creating the Bureau of the Budget, the statute reveals the terms upon which the idea of presidential representation had become legitimized by 1921 and the extent to which constitutional relationships were reconfigured. Just as a new “rhetorical presidency” was layered upon the old Constitution, a new institutional presidency—promising greater presidential independence—would be as well.Footnote 14
Explaining the BAA of 1921
At issue in this article is a critical departure, a pivotal elaboration of presidential representation in which Congress did not simply tolerate a presidential pretension but candidly acknowledged its own institutional incapacity and promoted the institutionalization of presidential remedies. Indeed, as Sean Gailmard and John Patty point out, the institutionalization of the presidency required a congressional “supply side,” rather than just a presidential “demand side.”Footnote 15
I argue that the idea of presidential representation was not just a vehicle used instrumentally to pass the BAA, but instead, that the act was utilized to advance a new conception of American government. The idea animated reformers prior to settling on budget reform as their primary vehicle, and it was taken up by key actors at all stages of the reform process—proposing reforms, considering legislation, and implementing the new law. Notably, the BAA shows the institutional presidency emerging first on what might otherwise be considered its most unlikely constitutional ground.Footnote 16 The intimate connection that had developed between finance and representation in the Anglo-American political tradition dates back to the time of the Magna Carta.Footnote 17 In the United States, this connection was formalized in Article I, as revenue bills would originate in the House, giving the power of the purse to the chamber closest to the people.Footnote 18 The question thus arises as to what circumstances and arguments could have led the House to concede this vital advantage. Indeed, given this history, the importance of conceiving of the president as a representative of the whole people to reforms involving presidential budgeting becomes more apparent.
There is a notable paradox as well in the specific timing of the law’s passage. While the initial proposal from President Taft’s Commission on Economy and Efficiency in 1912 came in the midst of Progressive enthusiasm for executive power, the law was not passed until 1921—after World War I, after congressional-executive relations had deteriorated, after the repudiation of Woodrow Wilson, and after public reaction against Progressivism had set in. In that context, one might have expected the president’s role to recede. Yes, World War I had brought about a severe debt problem, but solutions that did not require presidential involvement were conceivable and considered (see Table 1). Instead, presidential initiative was formalized in the budget process and the president’s organizational capacities were augmented. In effect, Congress accepted that institutionalizing presidential leadership was required even for a return to “normalcy.”Footnote 19
Table 1. Examples of Possible National Budget Solutions
Alternative explanations for the BAA’s passage that do not point to the importance of presidential representation cannot adequately account for the type of reform enacted. Some have pointed to Congress simply seeking to solve its collective-action problem in budgeting.Footnote 20 While Congress clearly acknowledged this as a problem, other solutions that would not involve the president, such as centralizing the committee process or giving more authority to the Treasury Secretary, were feasible and had precedent (see Table 1). Thus, the collective-action problem alone was not sufficient for Congress to decide on a presidential solution. A more recent account put forth by Gailmard and Patty emphasizes Congress wanting to provide the president with information in using powers already possessed.Footnote 21 Their explanation correctly points to Congress wanting the president to be well informed by placing the Bureau of the Budget under his control, but it is insufficient to explain why Congress would grant the president a new and formalized responsibility that he had generally been excluded from previously.Footnote 22 An explanation solely focused on the idea of efficiency is also inadequate.Footnote 23 Simply seeking a more efficient budget would not in itself require a presidential solution. Rather, an assumption of presidential representation was the core logic for reformers who claimed the president would behave more efficiently than legislators in budgeting. As I will show, beneath the amorphous consensus on the goal of efficiency, the idea of presidential representation was a contested claim in debates over budget reform. A purely partisan explanation of unified government also cannot account for the BAA’s passage in 1921 because budget reform also passed during divided government in 1920, and the margin of passage in 1921 was overwhelmingly bipartisan. Finally, while rising debt from World War I was the proximate cause of the law, this alone cannot explain the particular design chosen in the law.Footnote 24 An intellectual program of reform based upon the purported merits of presidential representation was needed to supply the law’s design.
Method and Evidence
To demonstrate that the idea of presidential representation was the core assumption behind the design of the BAA, I recount the central events of the national budget movement by process tracing, tracking the reforms that successfully made it into the passed law and explaining the significance of these departures from previous practice.Footnote 25 Because the BAA involved compromises, I examine the resistance reformers encountered and the limitations of their success as registered in the final law, comparing the innovations proposed and the resistance aroused for key proposals and legislation. Altogether, I show that while the timing of the act was caused by World War I debt, the design of the law was crucially influenced by a popularized idea of the president’s role as national representative, leading to a presidency-oriented rather than a Congress-oriented solution.
The use of the language of presidential versus congressional representation and the raising of constitutional issues, as opposed to simply focusing on partisanship or regional interests, signals that the debate over budget reform centered upon questions of institutional design. Two types of statements are evidence for reference to the idea of presidential representation. First, statements criticizing members of Congress or other executive branch officials as parochial or beholden to special interests in budgeting, while arguing that the president should have greater responsibility instead, are implicit arguments for presidential representation. By promoting a presidential solution to a perceived problem of congressional representation, as opposed to alternative options for institutional design, such statements implicitly assume that the president will be more likely to represent the national interest. Second, statements can also explicitly spell out the claim of presidential representation, laying out the purported logic that the president is assumed to act only in the interest of the people as a whole because he represents a national constituency. Because presidents would naturally claim to represent the nation as a whole, I look primarily to other actors who might have different interests than the president—including legislators, executive branch officers, prominent reformers, academics, and journalists—for such statements, which indicate the influence of the idea on the design of a new budget process.
The article proceeds as follows. First, I describe the reform context in which the idea of presidential representation was popularized in elite discourse, and I show how the idea became connected to the budget movement. Second, I review the two principal proposals that formed the eventual basis of the law’s design. Third, I examine the proposed legislation and its passage. Fourth, I consider how the law was implemented. Finally, I conclude by considering the BAA as a provisional political development.
Ideas and Institutional Change
Questions involving the relationship of political ideas and institutional change are central to American political development.Footnote 26 Do the ideas or institutions come first?Footnote 27 How well are ideas translated into institutional forms? To show that the idea of presidential representation influenced the budget reform movement, rather than the movement inventing the idea post hoc, I briefly review the broader context of political thought and proposals for institutional change during this period before turning to how this idea was applied to budgeting.
The Reform Context: Progressive Faith in the Presidency
Progressives sought to release the energies of government for the common good by overcoming parochial and partisan interests.Footnote 28 Reformers believed executives were the means for gaining this focus on the whole. Governors increasingly demonstrated the potential of leadership based on an executive electoral connection.Footnote 29 Changes in presidential behavior vis-à-vis Congress—including more attempts to direct policy and rally public support for legislation—pointed to presidential leadership.Footnote 30
Praise for presidential representation and criticisms of congressional representation were prominent in elite discourse. A disgruntled congressman lamented, “The claim of President Jackson that the President was the direct representative of the whole people is to-day very often heard.”Footnote 31 “No one else represents the people as a whole,” lectured Woodrow Wilson at Columbia.Footnote 32 The president, wrote Theodore Roosevelt, was a steward “bound . . . to do all he could for the people.”Footnote 33 The president could “capture the public imagination” and check the perceived localism of Congress.Footnote 34 Congress, characterized by “hide-and-seek vagaries of authority,” was the stronghold of “special and local interests”; legislators valued “equity between the different parts” more than “the general interests” of the nation.Footnote 35
Accepting the premise developed in the nineteenth century that presidents uniquely represented the entire nation, Progressives sought to institutionalize it, thinking the office’s capacities were insufficient.Footnote 36 Though not agreeing on all proposals, most reformers focused on two institutional remedies, borrowing from the British parliamentary idea of cabinet government and what modern legal parlance refers to as the unitary executive theory.Footnote 37 The first was an enhanced presidential agenda-setting power. Bridging the separation of powers through a formal connection to Congress, the president, selected by a national constituency, would propose bills considering the needs of the whole country.Footnote 38 Some even sought to have department heads take the floor in Congress to introduce and advocate for the president’s legislation.Footnote 39 The second solution involved placing the president at the head of a more unitary executive branch. To overcome the perceived lack of coordination among separate departments, which could communicate with Congress without consideration of the president’s views, reformers sought to create new organizational capacity to manage the executive branch and ensure that only presidential proposals went before Congress.
Though these reforms were connected, they had potentially different constitutional implications. Compared with presidential agenda setting, making the president more clearly head of the executive branch might plausibly claim greater support in the executive power invested in the president by Article II. Nonetheless, despite contemporary originalist claims about the unitary executive theory, Progressives largely viewed granting the president new organizational capacity as a new development that would be at the discretion of Congress to bring about, as it would significantly change the operations of government.Footnote 40 Furthermore, while proponents of presidential agenda setting could point to the Article II power of recommending measures to Congress, stronger powers to set Congress’s agenda in certain policy areas directly challenged legislative prerogatives. Some candidly acknowledged that the president’s legislative powers from the Constitution were insufficient. According to Henry Jones Ford, Andrew Jackson had established the veto power as a representative tool, but “the correlative function, the legislative initiative, still dependent as it is upon congressional acquiescence, has shown no access of strength.”Footnote 41 “Without such an initiation,” wrote John Burgess, “the veto power does not give the President an equal part in the legislation power.”Footnote 42 Gamaliel Bradford called for an inversion. Rather than Congress passing laws and the president reacting by signing or vetoing, the president “should himself” submit legislation “for acceptance or rejection” by Congress; “the veto should be applied the other way.”Footnote 43
Under the Guise of Efficiency: Applying Presidential Representation to Budgeting
In pursuing budget reform, Progressives used the language of “efficiency.”Footnote 44 A cadre of presidency-oriented reformers took advantage of the efficiency consensus to push their own program of institutionalizing presidential representation, seeking to achieve the institutional remedies of presidential agenda setting and greater organizational capacity. Given the constitutional basis of revenue bills originating in the House, this would bring constitutional issues to the fore.
Efficiency—“the idiom of [that] generation,” as F. Scott Fitzgerald wrote—was an amorphous concept encapsulating many reform impulses.Footnote 45 Some viewed the concept as promising economical spending, while many Progressives viewed it in terms of broader social goals.Footnote 46 But efficiency alone was not a rationale for a specifically presidential cure for the nation’s alleged budgetary sickness. Rather, questions and claims of representation were raised and contested.Footnote 47 Claims that the president was the most likely political actor to propose efficient budgets relied upon an assumed logic of presidential representation. Congress’s perceived inefficiency in budgeting was chalked up to the localistic incentives of individual legislators to seek money for their respective districts and states without consideration of a national perspective.Footnote 48 Moreover, agencies routinely overspent their budgets and requested further appropriations to make up the deficiency.Footnote 49 These critiques of Congress as parochial or individual executive departments as myopic, when accompanied by claims that the president should propose a budget, relied on an implicit assumption of presidential representation. Indeed, as one contemporary critic argued, while many reformers “frequently urged” economy as a rationale, in fact they assumed “an inevitable connection” between “responsible government and the so-called executive budget.”Footnote 50
The existing congressional budget process did exclude the president. The original act establishing the Treasury Department had been unique in clearly specifying the Secretary’s responsibilities to Congress. In the early years of American government, Treasury Secretaries Alexander Hamilton and Albert Gallatin had wielded influence over the estimates.Footnote 51 Despite Andrew Jackson’s subordination of the Treasury in the Bank War, there subsequently continued to be no stipulation that the president would have a role in budgeting.Footnote 52 Presidential influence over executive branch estimates was irregular and not formalized.Footnote 53 Indeed, after the Civil War, rather than involving the president, Congress made the new Appropriations Committee responsible for reviewing estimates, but the process was soon decentralized again.Footnote 54 By the early twentieth century, budgeting involved individual departments and agencies submitting nonrevisable estimates to the Treasury Secretary to be sent to the House. Neither the president nor the Treasury Secretary was responsible for individual estimates or coordinating among departments. Congress reigned supreme. Various committees, especially Ways and Means and Appropriations, considered and revised the estimates for particular departments. But no part of government holistically oversaw financial matters.Footnote 55
But beyond criticizing congressional inefficiency, key proponents of budget reform had more idealistic expectations of presidential representation, explicitly touting the president as best able to embody the nation, serve as its overall spokesman, focus on national priorities, and alert the whole country to the importance of seemingly mundane aspects of budgeting. Far from being just a sideshow to arguments about efficiency, this broader conception of the idea of presidential representation was present at the genesis of the reform program that led to the BAA, exercising a crucial influence on the law’s design.
Moreover, institutionalizing presidential representation in budgeting raised constitutional issues. Pressing the implications, reformers identified the separation of powers as an obstacle to be overcome.Footnote 56 Henry Jones Ford described budgetary issues as “symptomatic of [a] general constitutional disease” in which “our national representative assembly fails to discharge this constitutional function successfully.” The House, the primary locus of budget initiation, was not up to the task.Footnote 57 Explicitly connecting his proposed reform to the idea of presidential representation, Ford called for a presidential agenda-setting power in budgeting “to subordinate particular interests to the general interest.”Footnote 58 Democracy was at stake: if the president’s “power of initiative is abridged, the sovereignty of the people is impaired.”Footnote 59
Frederick Cleveland, the leading protagonist of the budget reform movement, also admitted purposes beyond efficiency. Since Congress confused responsibility for policy, democracy would be better served by being able to hold an executive accountable for governance.Footnote 60 An executive budget—accounting for past policy decisions, assessing present financial conditions, and allowing for long-term planning—would be a means to this end, allowing the public to properly evaluate its leaders. As Cleveland stated, “The only person or persons who can formulate and submit for consideration a plan or program for the government as a whole is the President and his advisors.”Footnote 61
Implicitly and explicitly, reformers argued for the importance of presidential representation to a new budget process, seeking to grant the presidency an initiatory responsibility and organizational capacity commensurate with its allegedly superior representative role. In the politics of reform, the language of efficiency would fail to submerge the larger structural stakes.
Proposals: Piercing the Efficiency Veil
Two proposals especially influenced the BAA of 1921: the President’s Commission on Economy and Efficiency [PCEE] report of 1912, spearheaded by Frederick Cleveland, and the Institute for Government Research’s [IGR] proposal of 1919, authored by William Willoughby. In this section, I show how the proposals drew directly from the idea of presidential representation, as both contained the idea’s two main institutional referents, a presidential budget and new executive organizational capacity. As indicators of constitutional resistance, I also examine the response to the PCEE and how it influenced adjustments in the IGR proposal.
A Proposal Stillborn: The President’s Commission on Economy and Efficiency
While Republican President William Howard Taft sought to prioritize economy and efficiency in government, he considered presidential representation to be the core assumption of budget reform.Footnote 62 Since the president “is the one whose method of choice and whose range of duties have direct relation to the people as a whole and the government as a whole,” Taft later wrote, he would likely “feel the necessity for economy in total expenditures.”Footnote 63 Seeking to achieve a presidency-oriented reform, Taft sought discretion to run a study; in 1910, Congress relented and appropriated funds. Revealing the scope of his ambition, Taft chose Frederick Cleveland to lead the inquiry. Cleveland did not disappoint, suggesting the need to centralize executive branch authority. Pleased, Taft expanded the inquiry to a full commission in 1911.Footnote 64 The two principal innovations proposed by the commission—an executive budget and a Bureau of Administrative Control—were justified in terms of presidential representation.
First, the executive budget would place a national perspective before Congress and improve the president’s connection to public opinion. Taft explained that the existing budget process failed to properly inform the public of government business because it lacked a publicized presidential plan.Footnote 65 The president’s budget proposal would include a summary message and summaries of general finances, expenditures, estimates, and suggestions for changes in law to facilitate “greater economy and efficiency.”Footnote 66 But a central problem of not having an executive budget process, the commission argued, was that the president lacked a tool to keep in touch with popular feeling: “Without a definite method of getting his concrete proposals before the country the Executive, as the one officer of the Government who represents the people as a whole, lacks the means for keeping in touch with public opinion with respect to administrative proposals—both the Congress and the Executive are handicapped in thinking about the country’s needs.”Footnote 67 Instead, the executive budget would “enable the President, as Chief Executive and representative of the people at large, to get before the country a definite proposal.” Additionally, the president would be responsible for deciding whether to approve “action taken by the Congress on its own initiative,” and he would determine whether measures promoted the “public welfare” and should be executed. Congress would be responsible to take action on “definite proposals” submitted by the administration.Footnote 68 The president’s national perspective was the linchpin of the process.
Significantly, the commission described how its vision of presidential budgetary agenda setting would depart from American precedents. The report noted that “executive authority” in other nations possessed “powers of initiation and leadership,” while “legislative authority” possessed “merely powers of final determination and control.” Because the United States had this relationship backwards, the “use of a budget would require that there be a complete reversal of procedure by the Government.”Footnote 69 Furthermore, the Article II provision for the president to recommend measures was deemed insufficient.Footnote 70
Second, the proposed Bureau of Administrative Control would enhance the president’s organizational capacity. As “the central information plant for the Government,” it would help the president and cabinet know “what are the current problems and conditions that require immediate consideration.” Among the functions performed by the bureau would be auditing and budget preparation. Crucially, to ensure that the president’s own views went before Congress rather than agency perspectives, the budget would “be formulated in a central office which is responsible directly to the President and not under any one department.”Footnote 71
Despite its name, the commission’s scope of inquiry revealed a broader scope than economy and efficiency, provoking backlash even before it completed its work. Both executive branch agencies and Congress resisted the commission. Conscious of its traditional relationship to Congress, the Treasury Department only grudgingly cooperated with Cleveland when Taft personally intervened. In 1912, Congress pushed back against the commission by rejecting a $250,000 funding request, giving instead $75,000 and stipulating that the appropriation pay only three salaries instead of five.Footnote 72
After hampering the study, Congress rejected its grandiose plans. The presidential budgeting proposal, involving mainly an Article I power, portended a significant change to constitutional relationships. The proposed Bureau of Administrative Control, which would be used to implement a new budget system, likewise would alter the relationship of the departments to Congress. But the budgeting recommendations provoked more congressional resistance.Footnote 73 Democratic committee chairmen “humanely disposed of” its report.Footnote 74 Furthermore, in July 1912, when Taft tried to impose the new budget process on a recalcitrant Congress by directing departments to submit their estimates in the manner prescribed by the commission’s recommendations, Congress countered by requiring submission of the estimates according to procedures under existing law. Taft responded by submitting estimates to Congress himself to complement the standard budget procedure, but Congress ignored his message. Finally, though Taft sought to extend the commission even after losing in the 1912 election, Congress refused to extend the appropriation past June 1913.Footnote 75
Political calculations did help fuel this defense of congressional prerogatives. Between the authorization of the PCEE and its report, Democrats had ended Republican unified party control with their victory in the 1910 midterm elections. Moreover, House rules had been decentralized after the revolt against Speaker Joseph Cannon (R-Ill.) in 1910, giving back more authority to individual legislators and making it even less likely that they would surrender prerogatives to the president.Footnote 76 But revealingly, the commission faced resistance because it had clarified the ambition behind institutionalizing its vision of presidential representation. The report had assailed Congress as the enemy of efficiency and praised the presidency as its champion. The potential scope of the president’s proposed initiative power was also ambiguous. By citing Britain’s budget process as a model, the report seemed to leave open the possibility that the president would present a plan allowing for no congressional amendments, a reform that went farther than Congress was willing to consider. House members recognized that their constitutionally granted powers were at stake. Indeed, in response to “many of their influential constituents” approving of Taft’s efforts, House Democrats proposed legislative budget alternatives instead, including either a committee that would report estimates of available revenue or a centralized single appropriations committee in each chamber.Footnote 77 Proposing his legislative plan, Representative J. Swagar Sherley (D-Ky.) called Congress “the only logical representative of a free people,” and Representative Victor Murdock (R-Kans.) praised Sherley’s plan for its “virtue” of giving the legislature, not the president, the initiative.Footnote 78 Finally, though deficits were rising, they had not yet reached post–World War I crisis levels.Footnote 79
With Democrat Woodrow Wilson assuming office in 1913, the prospects for an executive budget briefly appeared brighter. But despite his keen interest in budget reform and public administration, Wilson deferred to his congressional supporters, preferring the strictly legislative solution of a single committee controlling all appropriations.Footnote 80 Because his vision of executive-led government required strong party discipline to accomplish a legislative program, “Wilson traded administrative leadership for congressional leadership.”Footnote 81
Though disappointed, Taft correctly predicted that the commission’s recommendations would influence future reforms.Footnote 82 States increasingly gave budget responsibility to governors—the “popularly elected chief.”Footnote 83 After neither party had committed to the cause in the 1912 election, Democrats and Republicans both embraced budget reform in their 1916 platforms, though a partisan division emerged: Democrats favored a legislative solution, while Republicans now embraced Taft’s proposals.Footnote 84
For the moment, however, the limits on the potential for institutionalizing presidential representation in the American constitutional system had been exposed. It would take a greater crisis for Congress to reconsider reform.
A Strategic Adjustment: The Institute for Government Research
In 1916, the Institute for Government Research, a new organization formed to promote efficiency and effective governance, prioritized national budget reform. However, its political strategy differed from that of the PCEE. IGR trustees chose PCEE member William Willoughby as director rather than Frederick Cleveland, believing Willoughby would be the more effective political operator with Congress. Suspecting support for reform might surge after the war, he immediately set to work.Footnote 85 Despite feeling the PCEE had strategically bungled by prompting Congress to recoil at threats to its prerogatives, Willoughby nonetheless took the report as the starting point for his proposal.Footnote 86 His plan also contained the two institutional entailments of presidential representation—a presidential budget and new executive organizational capacity.
First, Willoughby recommended making the president responsible for formulating an annual budget, which would make Congress more effectively consider “the general interests of the government as a whole.” Though he focused more on the president’s responsibility to provide information to Congress than to the public, he crucially argued that the executive budget would allow citizens to exercise “a real popular control” upon “their representatives, legislative and executive.”Footnote 87 However, seeking to head off the resistance faced by the PCEE, Willoughby rejected a monopoly presidential agenda-setting power in favor of a revisable presidential budget that Congress could alter through the proposed new budget committees in each chamber.Footnote 88
Second, in proposing new executive organizational capacity, Willoughby shifted to a more managerial emphasis for the president’s role rather than bearing down on Congress. He focused attention on making the president more clearly the head of the executive branch, which might appear less constitutionally threatening given the president’s executive power under Article II. Nevertheless, Willoughby viewed this as a departure, noting that the Treasury Secretary had been the original officer directed by Congress to prepare estimates in 1789: “Congress had no intention of establishing the President in the position of head of the administration.”Footnote 89 Indeed, “the fundamental basis for effecting this reform” was “the new conception now entertained regarding the President as the responsible head of the administration.”Footnote 90 Unlike the British, who recognized the Treasury as the superior authority in financial matters, Willoughby wanted the president to be “established by law” as “the sole authority by whom requests for the grant of funds for the executive and administrative branches of the government shall be made of Congress.”Footnote 91 To ensure budget proposals reflected only the president’s views, a new budget bureau would be placed “under the immediate authority and direction of the Chief Executive.”Footnote 92
These proposals included the institutional solutions attached to the idea of presidential representation, but in a reconfigured, less confrontational form. Willoughby preferred the strong executive powers of the British system, but recognized that Congress was unlikely “to make such a radical change.” Thus, he expressed a strategic preference: “The system proposed goes as far as it is believed that Congress is prepared to go at this time.”Footnote 93
Legislation: Congress Recognizes Its Incompetence
The IGR proposal soon made its mark, as a crisis precipitated legislative action. Rising debt from World War I—the total debt skyrocketed from $1 billion in 1916 to over $25 billion by 1919—exposed Congress’s inability to halt the growth in expenditures and gave the movement to enact a presidential budget momentum. Citizens also had an increasingly direct stake in federal finances under the new system of income, corporate, and inheritance taxes. Despite divided government, the House Select Committee on the Budget, recognizing Congress’s failure to control the deficits, urged the adoption of an executive budget in 1919. Woodrow Wilson, who previously had supported budgeting through a single appropriations committee, later announced his concurrence.Footnote 94
Proposals favoring presidential participation in budgeting now preempted alternatives that excluded the president, such as the earlier proposals to have a single committee review departmental estimates or to centralize the appropriations process. Indeed, the implicit endorsement of presidential representation is apparent when considering the total range of possibilities Congress did or could have considered (see Table 1). In addition to addressing the issue solely through changes to the committee process, Congress could have given authority to prepare budgets to the Treasury Secretary alone or could have created its own resource for budgeting like the later Congressional Budget Office.Footnote 95 Furthermore, some states and cities had commissions to propose budgets.Footnote 96 However, though Congress chose a solution involving the president formally for the first time, it rejected granting the president a supermajority agenda-setting power, revealing the limit to which it was willing to empower the president.
Though increased debt and the new tax system were proximate causes of the BAA’s passage, explaining the law’s design requires accounting for the persistent influence of the idea of presidential representation. In addition to the implicit assumption of presidential representation in promoting a presidential solution for a perceived problem of congressional parochialism, I show that this idea was explicitly discussed as the basis for legislation. To determine the extent to which Congress provided for an institutionalization of presidential representation, I examine the innovations adopted in both the 1920 and 1921 versions of the passed bill and what limitations were imposed. After Wilson vetoed the first act in 1920 over removal power concerns, Warren Harding signed the second version in 1921. The timing of the law’s passage is interesting because it came amid the public and congressional reaction against Progressivism, against Wilson’s presidency, and against executive overreach in the League of Nations debacle.Footnote 97 While Harding and Congress sought to curb budget deficits as part of the return to “normalcy,” they advocated greater presidential responsibility. As I show in this section, the fact that a new budget process formally involving the president advanced in this political environment marked the emergence of the idea of presidential representation as a new common sense.Footnote 98
Hearings: The House Select Committee of 1919
The hearings of the House Select Committee on the Budget, chaired by Representative James Good (R-Iowa), who also was the Appropriations Committee chairman, reveal that Congress was aware of how proposed budget reforms were justified based upon the idea of presidential representation. Despite warnings from critics that a presidential solution would be a significant departure—“I think we had better stick pretty close to the Constitution with its division of powers well defined and the taxing power close to the people,” wrote Joseph Cannon—Congress increasingly recognized its failure to control the debt, and calls for presidential responsibility increased.Footnote 99
For example, former congressman J. Swagar Sherley—who had previously proposed a legislative budget involving a committee on estimates—testified in favor of giving the president a budget bureau and the sole power to appoint its director. While also advocating committee centralization, Sherley wanted the president to set the agenda: “The legislative body should not undertake the forming of a budget until after action by the executive branch.”Footnote 100 Likewise, Representative John Nance Garner (D-Tex.) noted the difficulty of encouraging individual legislators to reduce expenditures, arguing that Congress should only see the president’s budget rather than individual department and bureau proposals.Footnote 101
Both principal proponents of budget reform testified. Explaining that the first institutional entailment of the idea—a presidential budget—relied upon the purported logic of presidential representation, Frederick Cleveland told the committee that “the assumption that lies back of the suggestion that the Executive should be held responsible is this: . . . the Executive is the one man that is elected by the people at large and represents the whole country . . . the viewpoint of his vision must be countrywide . . . he must be in a position of coming to have some definite program or plan that is comprehensive.Footnote 102 William Willoughby, describing the second entailment of new organizational capacity, argued that the purpose of reform was “definitely locating responsibility with the President.” Giving budget responsibility to the Treasury Secretary would fail to subordinate the rest of the cabinet, so the president needed more direct control over budget preparation.Footnote 103 However, Willoughby also framed his proposal as a way to help Congress fulfill its responsibilities.Footnote 104
Others echoed their testimony. Former Secretary of War Henry Stimson argued for presidential budgetary initiative because “the Executive brings to bear . . . the viewpoint of the Nation as a whole as against the [legislature’s] view of an aggregate of disputants.”Footnote 105 Former PCEE member Frank Goodnow sought a new officer “who would . . . be able to stand up under the demands of the spending departments.”Footnote 106 Assistant Secretary of the Navy Franklin Roosevelt also called for an officer “directly under the President himself” to prepare a budget. Foreshadowing his own administrative reform efforts, he hoped a budget system would augment presidential control over the administration.Footnote 107
Testimony skirting the president’s new budget role was challenged. When Samuel Lindsay downplayed changes to Congress’s role in budgeting, Representative Joseph Byrns (D-Tenn.) confronted him: “I do not know of anything that would tend more to put Congress under the domination of the Executive.” Justifying it as a necessity, Lindsay admitted the scope of the proposed change: “It is true you are centralizing the power of initiation in the Executive that does not vest there now, and you are limiting the power of initiation that now vests in individual Members of Congress.”Footnote 108
The hearings clarified the growing consensus for the president to be granted a formal role in the budget process, while hinting at what limits would be imposed. The president, emphasized Good, would initiate the budget process, but Congress would retain the ability to amend the budget and would have an independent audit.Footnote 109 Still, the hearings augured a significant reform.Footnote 110
Veto in 1920
The first version of the BAA passed in 1920. Solutions excluding the president from budgeting found fewer advocates.Footnote 111 Willoughby assisted James Good in drafting a bill, which passed the House 285–3 on October 21, 1919. The Senate passed reform without opposition on May 1, 1920, but differences between the House and Senate versions had to be reconciled.Footnote 112 While the passed law contained notable innovations, some of the limitations it imposed revealed continued congressional ambivalence over presidential budgeting.
First, the bill provided for presidential agenda setting: the president would submit a yearly budget to Congress. Joseph Byrns, who had worried about enhancing presidential authority in the hearings, now emphasized that “the President . . . an elective officer of the United States, is made responsible to Congress and to the country.”Footnote 113 No one but the president, asserted Representative Edward Taylor (D-Colo.), “ought to be responsible absolutely for the fiscal and economic policy and system of our Government.”Footnote 114 The lack of a stronger supermajority agenda-setting power for the president underscored the boundary of what Congress was willing to accept.Footnote 115 Nevertheless, Congress’s passage of an enhanced presidential role in an Article I power indicated the influence of the budget reform movement.Footnote 116
Second, the bill augmented the president’s organizational capacity by creating the Bureau of the Budget [BOB]. However, this provision was central to the dispute between the House and Senate bills. In the conference compromise, BOB was placed in the Treasury Department and the Treasury Secretary was made BOB’s director, responsible for preparing the president’s budget.Footnote 117 Senator Medill McCormick (R-Ill.) explained that the Senate agreed with the House on “fixing upon the President the ultimate responsibility” for “the annual budget,” but preferred the Treasury Secretary to draft it. A separate budget officer might threaten the cabinet.Footnote 118 Since many reformers had advocated for a different officer directly under the president to prepare the budget, this was a shortcoming.
Finally, the 1920 bill placed a new General Accounting Office under congressional control, and even the strongest presidency-oriented bills considered had given Congress control of the audit.Footnote 119 The Comptroller General would be removable only by concurrent resolution, not requiring presidential approval.Footnote 120 Wilson thus vetoed the bill, asserting this infringed upon the president’s removal power.Footnote 121 The veto provided an opportunity for an even more presidency-oriented reform to emerge.
Passage in 1921
Despite Wilson’s veto, it was clear budget reform would soon pass in some form. Both party platforms embraced the presidential cure in 1920.Footnote 122 Entering office alongside a Republican Congress for unified government in 1921, President Warren Harding sought to deliver a promised deficit reduction. Ironically, though he had pledged to seek “normalcy,” Harding realized he needed new tools for presidential leadership, confessing that he sought a fundamental departure.Footnote 123 From a presidential perspective, the bill passed during the early part of Harding’s administration was stronger.
Both the House and Senate overwhelmingly passed budget reform, but they continued to have differences. First, the House placed an independent BOB solely under presidential control, while the Senate placed it again in the Treasury Department. Second, the House bill continued to allow for removal of the Comptroller General without the president’s approval, while the Senate bill provided for a joint resolution requiring presidential approval.Footnote 124 In the conference compromise, BOB was placed in the Treasury Department, but the director and assistant director would be appointed by the president without Senate consent and placed under his direct authority. Furthermore, though Congress retained an independent audit, it agreed to provide for removal of the Comptroller (who would hold office for fifteen years during good behavior) through a joint resolution requiring the president’s signature.Footnote 125
Thus, the law contained substantial innovations that had been inspired by the idea of presidential representation, marking, in essence, a first recognition in statute that the president was the nation’s chief representative.Footnote 126 It provided for presidential agenda setting with an executive budget and new executive organizational capacity through BOB. The law also prohibited any other members of the administration from submitting appropriations requests without a congressional demand to do so.Footnote 127 Reflecting on the act’s significance decades later, one scholar called it “probably the greatest landmark in our administrative history except for the Constitution itself.”Footnote 128 However, limits were imposed as well. The president did not simply gain power at Congress’s expense. Instead, “the independent audit was Congress’s quid pro quo for the President’s budget bureau.”Footnote 129 While the House had wanted BOB to be solely a presidential agency, it was placed in the Treasury Department. Congress retained the ability to amend the budget, and later acts consolidated appropriations authority in single committees in the House and Senate.Footnote 130
Floor debates revealed the significance that Congress attributed to the law. “The responsibility is laid on the President to outline a policy,” said Senator McCormick.Footnote 131 Discussing the compromise, James Good asserted that the House had mostly achieved its wishes. Importantly, this House bias corresponded to an enhanced institutional capacity for the president to fulfill his representative role:
[The House bill] assumed that the President, being the only official of the United States that is elected by all the people, and the only official who is designated by the Constitution to give Congress, from time to time, information on the state of the Union, the President must lay out a work program for the Government, and the appropriations that would necessarily follow would only be to supply the money to do the work in accordance with that work program.Footnote 132
Good emphasized giving the president the most direct control over BOB that the Senate would accept. BOB’s location “mattered very little” to the House given that it would not be subjected to the Treasury Secretary’s control.Footnote 133 Instead, “the real meat in the section is the power granted,” which was “only” to “the President.” Senate confirmation for the director and assistant director had been avoided because those positions would be “peculiarly the President’s staff,” ensuring that the budget “reflected [the president’s] sentiment.”Footnote 134 Finally, the director could be changed at any point, especially with new administrations.Footnote 135
The compromise received bipartisan praise.Footnote 136 John Nance Garner was pleased that the budget director was placed under presidential control. Believing the director could become “the second largest man in the executive department,” he relished a scenario in which the director and the Treasury Secretary disagreed: “He will be able to look at the Secretary of the Treasury and say, ‘You will cut out this expenditure. This is what I am going to abolish.’ ‘Who is this that is speaking to me?’ ‘It is the representative of the President of the United States himself.’”Footnote 137 This fervor was all the more remarkable since Garner had previously lambasted budget reformers for attacking Congress too much.Footnote 138
The compromise shows that the act was a tentative institutionalization of presidential representation. It would have been better for presidency proponents if BOB had been placed solely under the president outside the departmental structure. But when that was not possible for the time being, they prioritized placing the director under presidential control.Footnote 139 Given its Article I prerogatives, the House’s eagerness to ensure the president’s views would prevail in budget proposals is somewhat remarkable. However, rather than granting a presidential agenda-setting power that would limit congressional amendments, the House focused on the president’s control of BOB, touching on an Article II power. By knowingly passing a law with institutional innovations based on an assumption that the president’s national constituency would make him seek greater economy in expenditures, Congress had institutionalized presidential representation.
Implementation: Presidentialism in the Service of Normalcy
With the BAA’s passage in 1921, a paradox arose—the prospect of using the presidency to return to normalcy. The Progressive Era had been characterized by bold attempts at presidential leadership.Footnote 140 Yet despite the reaction against Progressivism, Congress and President Harding discovered that fiscal retrenchment required its own increment of presidentialism. Satirically describing this new presidential responsibility, F. Scott Fitzgerald wrote that “a good President ought to be able to tell just how much we could afford.”Footnote 141
Though the law was supposed to apply to the 1923 fiscal year, the Harding administration boldly decided, without formal authority, to implement it early and devise a budget for 1922.Footnote 142 The conservative Harding admitted to aggressively using the presidency to “restore sane and normal ways again.”Footnote 143 Nudging Harding to adopt a stronger view of presidential authority was Charles Dawes, his budget director. Dawes underscored the significance of using the new budget process: it “marks the passing (and is intended so to do) of the old system.”Footnote 144 The law ensured that departments would be “made to better accord with the plan which the President had established.”Footnote 145 Though Dawes was criticized by some for focusing BOB too much on reducing costs and not on broader policy ends, he unquestionably viewed the budget process through the lens of presidential representation: “Nothing should be allowed to withdraw the attention of the public from the duty and powers of the President.”Footnote 146
Even after Dawes left his post, enthusiasm for using this new budgetary authority continued unabated. Harding put his own business-oriented spin on presidential representation: “What we are doing is not for ourselves . . . not for the President . . . but for the people—the stockholders of this great business.”Footnote 147 Seeking to ensure that only the presidential perspective would be put before Congress in budget proposals, Harding warned executive branch employees that testifying with estimates “in excess of the Executive recommendation” would be “sufficient reason” for being fired.Footnote 148 His successor, Calvin Coolidge, issued the same warning.Footnote 149
Harding, Dawes, and Coolidge implemented the law in a manner that augmented presidential authority. Their focus on fiscal retrenchment aligned with the Republican Congress’s wishes, but caused consternation among some of the more ambitious presidency-oriented reformers. It would remain for future efforts in the coming decades to seek to institutionalize the presidency in more policy areas.
Conclusion: A Provisional Achievement
The idea of presidential representation is more than a standard by which to judge presidents; it is a significant prod to the development of a presidency-centered government. The case of the Budget and Accounting Act of 1921 shows both the achievements and limits of this institutionalization of presidential representation. I have demonstrated that reformers were influenced by presidential representation, that their proposals and laws incorporated some of its key institutional entailments, and that the reforms departed from previously established constitutional arrangements. The president gained substantial agenda-setting power by initiating the budget process and was given new organizational capacity to exercise this responsibility. Marking the boundary of the achievement, Congress retained the ability to amend the president’s proposals, created an independent audit, and, until 1939, located the BOB in the Treasury Department. While the immediate motivation for the law was the challenge of post–World War I deficits, the law’s design cannot be explained without accounting for the idea of presidential representation. Furthermore, Congress continued the institutionalization of the presidency in other policy areas—tariffs and trade (1934), executive reorganization (1939), employment and economic management (1946), and national security (1947)—further providing presidents with new licenses for agenda setting and new executive organizational capacities.Footnote 150 While enacted in response to various challenges of the day, these statutes collectively marked recognition of the president as the nation’s chief representative.Footnote 151
A fundamental problem, however, is indicated when the influence of presidential representation on the development of the institutional presidency is reflected back upon the constitutional structure. It is revealing that Howell and Moe now propose a constitutional amendment for presidential agenda setting. They endorse the claim of presidential representation, but they recognize the continued boundaries Congress and the Constitution have presented.Footnote 152 The institutionalization of presidential representation in statutes has always been unstable, pushing against the boundaries of the written frame. This is not to say that the acts are unconstitutional, but rather that they often reflect attempts to alter constitutional relationships.Footnote 153
Congress retains the capacity to adjust this authority if it thinks presidents have diverged too far from its own political purposes. The two institutional entailments contained in the 1921 law were not equally durable. In the wake of conflict with the Nixon administration over its impoundments of funds, Congress sought to reclaim its Article I budget prerogatives in 1974, creating its own alternative budget process and establishing the Congressional Budget Office.Footnote 154 But the president’s enhanced managerial capacity—arguably more in line with Article II—endured in the form of the Office of Management and Budget (created in 1970) and in the enhanced priority of regulatory review in the 1980s (though Congress did in 1974 begin requiring Senate confirmation of the OMB Director and Deputy Director).Footnote 155 This suggests that powers granted to the president may be more durable to the extent that they are more easily related to Article II and more vulnerable when touching upon Article I.
Notwithstanding Congress cooling on its enthusiasm for presidential budgetary initiatives, the purported promise of presidential representation has endured, becoming the animating force for another reform. The line-item veto, allowing presidents to veto specific items in appropriations bills, would allegedly “permit Presidents to better represent the public interest” and “throw a spotlight of public scrutiny onto the darkest corners of the Federal budget.”Footnote 156 Demonstrating again how presidential representation anticipates reforms that can stretch from the Constitution, this reform’s passage in 1996 was soon invalidated by the Supreme Court. Even so, the House has subsequently passed other versions.Footnote 157
Though the institutional demands of presidential representation remain open to negotiation, congressional pushback also has raised problems. Donald Trump is only the latest president to have his proposed budget mostly ignored, and Congress has taken back some of its constitutional prerogatives. But the record of congressional performance is still poor: failures to pass budgets on time, reliance on continuing resolutions, government shutdowns, and even an ad hoc supercommittee.Footnote 158 Presidents bear responsibility along with Congress for the situation, but nonetheless, criticisms leveled against congressional incompetence are magnified by the fact that alternative institutional arrangements based on the alleged benefits of presidential representation are at the ready.Footnote 159 In effect, Congress is now indicted politically for asserting its constitutional prerogatives and refusing the president tools commensurate with his recognized status as the nation’s leading representative.Footnote 160 Thus, the idea of presidential representation in budgeting now persists precisely on its force as an idea. It is used to badger Congress, expose its weaknesses, and agitate for changes that anticipate a different kind of government altogether.
