A. Liberalism: Old

It is not an accident that the fathers of the new liberalism, such as John Dewey and Herbert Croly, and especially those of them who were most responsible for the role of administration today, such as Woodrow Wilson and Frank Goodnow,⁴

all found it necessary to focus attention on the American Declaration of Independence, the Constitution, and the theory of social compact represented in these documents.⁵ From the point of view of these men, the theory animating the founding documents was responsible for limiting government in the very ways that would make it impossible to realize their vision of the modern state. Particularly troubling was the transhistorical and limited account of the ends of government laid out in the Declaration, which the fathers of the new liberalism correctly identified as grounded in the language of John Locke's Second Treatise of Government.⁶ The purpose and role of government, according to the Declaration, is the same for all men by virtue of their common nature, and does not change from one generation to the next. The purpose of government, in the words of the Declaration, is to secure “certain unalienable Rights,” which the individual has according to “the Laws of Nature and of Nature's God,” and which are not granted by government but by his “Creator.” Government exists, therefore, for the very purpose of protecting these individual rights, as individuals would not otherwise consent to it.

The task of protecting individual natural rights in a democratic system of government was the primary challenge to the framers of the Constitution. This is why the authors of The Federalist focused on the problem of majority faction—on the possibility, or even likelihood, that a government empowered by a passionate majority would continually threaten the natural rights of the minority if its jurisdiction was not limited in extent and its powers were not channeled by a prudent design of national institutions. While the extent of national territory was cited by James Madison in Federalist No. 10 as a key means of controlling the effects of faction, federalism and the separation of powers served as the two primary, organizational features of the constitutional order. Both features were predicated, at least partly, on the notion that dividing power made it more likely that it would be exercised in a rational manner more conducive to the securing of rights. In the case of federalism, the national government would be “limited to certain enumerated objects,”⁷

thus sharing power with state and local authorities.

In the case of separation of powers, the legitimate authority of the union would be exercised by three coequal departments of government, each making sure that the others remained within the confines of their proper constitutional places.⁸

The fundamental aim of the separation of powers—which, as the fathers of progressive liberalism again correctly understood, came not only from Locke's Second Treatise but more directly from Montesquieu's Spirit of the Laws—was to safeguard rights against the possibility of arbitrary government.⁹ Indeed, Madison in Federalist No. 47, echoing Thomas Jefferson, redefined “tyranny” to mean the absence in government of the separation of powers.¹⁰ It is from this fundamental aim of separation of powers that we can discern three important tenets of American constitutionalism that bear directly on the administrative state, although this is by no means an exhaustive list. The first is the principle of nondelegation. If the separation of powers means anything at all, it means that one branch of government may not permit its powers to be substantially exercised by another branch.¹¹ The second tenet is a corollary of the first: there may be no combination of functions within a single branch. Madison, quoting Jefferson in the passage from Federalist No. 47 mentioned above, elaborates on this point: “The accumulation of all powers legislative, executive and judiciary in the same hands … may justly be pronounced the very definition of tyranny.”¹² Under this second tenet of the separation of powers, those making the law would also have to be subject to its being enforced upon them by an independent authority; those involved in execution could not make up the law as they went along, but would instead have to enforce laws that had been previously established by a separate authority; and those on whom the law was enforced could have their cases judged by an authority entirely separate from the one that had brought prosecution. The third tenet of the separation of powers is the responsibility of administration to the republican executive. The government remains “wholly popular,” in the words of Federalist No. 14, because those who carry out the law (administrators, under the traditional meaning of the term) are directly answerable to the president, who is elected.¹³ The Constitution grants all of “the executive power” to the president and requires him to “take care that the laws be faithfully executed.”¹⁴ Administration—as vigorous as some of the founders surely envisioned it—was thereby placed wholly within a single branch of government, and a clear line of political accountability for administrators was established.¹⁵

B. Liberalism: New

For the American pioneers in the new liberalism, this older, limited understanding of government stood in the way of the policy aims they believed the state ought to pursue in a world that had undergone significant evolution since the time of the founding. They believed that the role of government, contrary to the ahistorical notion of the earlier liberalism, ought to continually adjust to meet the new demands of a new age. As Wilson wrote in The State, “Government does now whatever experience permits or the times demand.”¹⁶

A carefully limited government may have been appropriate for the founding era, when the primary concern was throwing off central-government tyranny, but in order for government to handle the demands placed upon it by modern times, the founding-era restrictions on its powers and organization would have to be eased and the scope of government expanded. This is why Dewey criticized the founders for believing that their notions about government transcended their own age; they “lacked,” he explained, “historic sense and interest.”¹⁷ At the most fundamental level, therefore, the separation of powers was a deadly obstacle to the new liberalism, since it was an institutional system intended to keep the national government directed toward the relatively limited ends enumerated in the Constitution and the Declaration of Independence.

Beyond this fundamental difference on the very purpose of government, the three tenets of the separation of powers that I have described posed a particular problem for the new liberals' vision for national administration at the outset of the twentieth century. The range of activities they wanted the government to regulate was far too broad for Congress to handle under the original vision of legislative power. Instead, to varying degrees, the fathers of progressive liberalism envisioned a congressional delegation of regulatory power to an enlarged national administrative apparatus, which would be much more capable of managing the intricacies of a modern, complex economy because of its expertise and its ability to specialize. And because of the complexities involved with regulating a modern economy, it would be much more efficient for a single agency, with its expertise, to be made responsible within its area of competence for setting specific policies, investigating violations of those policies, and adjudicating disputes. The fulfillment of progressive liberalism's administrative vision, therefore, required the evisceration of the nondelegation doctrine and the adoption of combination of functions as an operating principle for administrative agencies. Furthermore, the new liberals believed that administrative agencies would never be up to the mission they had in mind if those agencies remained subservient to national political institutions. Since modern regulation was to be based on expertise—which was, its founders argued, objective and politically neutral—administrators should be freed from political influence. Thus, the constitutional placement of administration within the executive branch and under the control of the president was a problem, as the new liberals looked to insulate administrators not only from the chief executive but from politics altogether.

The achievement of these aims for national administration in the course of the twentieth century, at the expense of separation-of-powers constitutionalism, has been described in the legal literature and elsewhere, especially in Gary Lawson's seminal essay “The Rise and Rise of the Administrative State.” As Lawson explains, the Supreme Court ceased applying the nondelegation principle after 1935, and allowed to stand a whole body of statutes that enact the new vision of administrative power.¹⁸

These statutes, to varying degrees, lay out Congress's broad policy aims in vague and undefined terms, and delegate to administrative agencies the task of coming up with specific rules and regulations giving them real meaning. Lawson cites, for example, securities legislation giving to the Securities and Exchange Commission (SEC) the power to proscribe the use of “any manipulative or deceptive device or contrivance in contravention of such rules and regulations as the Commission may prescribe as necessary or appropriate in the public interest or for the protection of investors.” The agency, on the basis of its expertise, and not Congress, on the basis of its electoral connection, is charged with determining the policy that best serves the public interest. In another example, legislation on broadcast licenses directs that the Federal Communications Commission (FCC) shall grant licenses “if public convenience, interest, or necessity will be served thereby.”¹⁹ More recently, the Supreme Court, under Chief Justice William Rehnquist, made clear that there would be no revisiting the abandonment of nondelegation. In the case of Mistretta v. United States (1989), the Court upheld the statute which delegated to the U.S. Sentencing Commission the power to set sentences (or sentencing guidelines) for most federal crimes. If any case were going to constitute grounds for nondelegation review, it would have been this one. Congress created the Sentencing Commission as, essentially, a temporary legislature with no purpose other than to establish criminal penalties and then to go out of existence.²⁰ But Mistretta simply served as confirmation that the federal courts were not going to bring the legitimacy of the administrative state into question by resurrecting the separation of powers.

The second tenet of separation of powers—the prohibition on combining functions—has fared no better in modern constitutional and administrative law. As Lawson explains, “the destruction of this principle of separation of powers is perhaps the crowning jewel of the modern administrative revolution. Administrative agencies routinely combine all three governmental functions in the same body, and even in the same people within that body.” His example here is the Federal Trade Commission (FTC):

The FTC is a particularly apt example, since it was the “quasi legislative” and “quasi judicial” character of the FTC that was upheld in 1935, in the landmark Supreme Court case of Humphrey's Executor v. United States—the first time that the Court so clearly acknowledged that agencies technically within the executive branch could exercise substantially nonexecutive functions.²²

The new liberalism has also succeeded, at least partly, in defeating the third tenet of the separation-of-powers framework, by weakening the political accountability of administrators and shielding a large subset of agencies from most political controls. While the independence of so-called “independent regulatory commissions” and other “neutral” agencies is not as clearly established as delegation and combination of functions, the federal courts have certainly recognized the power of Congress to create agencies that are presumably part of the executive branch (where else, constitutionally, could they be?) but are nonetheless shielded from direct presidential control. Normally, this shielding is accomplished by limiting the president's freedom to remove agency personnel. The Humphrey's case, for example, overturned the president's removal of an FTC commissioner by reasoning that the Commission was more legislative and judicial than it was executive.²³

More recently, the Supreme Court upheld the independent counsel provisions of the Ethics in Government Act of 1978 (the provisions were subsequently repealed), concluding that even an office as obviously executive in nature as a prosecutor could be shielded from presidential control.²⁴ These rulings reflect the acceptance of a key tenet of the modern administrative state: that many areas of administration are based upon expertise and neutral principles, and must therefore be freed from the influence of politics. That such a notion has become ingrained in the American political mind-set was evidenced by the near universal outrage expressed over the Supreme Court's 2000 decision in FDA v. Brown and Williamson. In this surprising exception to its standard deference for agencies, the Court ruled that before the Food and Drug Administration (FDA) could promulgate and enforce regulations on tobacco, Congress first had to pass a law actually giving the agency the authority to do so.²⁵ The decision was denounced because it would subject tobacco regulation to the control of the people's elected representatives in Congress (where tobacco-state legislators might derail it), instead of giving carte blanche to FDA scientists to regulate in accord with their own expertise.

The acquiescence in the realms of law, politics, and culture to the concepts of delegation, combination of functions, and insulating administration from political control is explained by what legal scholars call the victory of “functionalism” over “formalism,” or what political theorists might loosely translate as “pragmatism” over “originalism.” Simply defined, a functionalist or pragmatic approach begins not with the forms of the Constitution, but with the necessities of the current age, thereby freeing government from the restraints of the Constitution so that the exigencies of today can be met. As Peter L. Strauss argues, “Respect for ‘framers' intent’ is only workable in the context of the actual present, and may require some selectivity in just what it is we choose to respect.”²⁶

This sentiment elevating expedience and efficiency over the separation of powers was expressed very clearly by Justice Harry Blackmun in his opinion for the Court in Mistretta: “Our jurisprudence has been driven by a practical understanding that in our increasingly complex society, replete with ever changing and more technical problems, Congress simply cannot do its job absent an ability to delegate power under broad general directives.”²⁷

The rise of the administrative state that is such an integral feature of the new liberalism thus required the defeat of the separation of powers as a governing principle, at least as it was originally understood, and its replacement by a system that allows delegations of power, combination of functions, and the insulation of administration from the full measure of political control. The remainder of this essay will examine the origins of modern administrative law and policymaking in the critique of the constitutional separation-of-powers system that was made by the fathers of progressive liberalism. In actual practice, the biggest steps toward defeating separation-of-powers constitutionalism took place during the New Deal, which was animated by a strong optimism about the capabilities, public-spiritedness, and objectivity of administrators. Nowhere was this optimism embodied more fully than in James Landis, the New Deal architect of the administrative state, who is the topic of the essay's next section. I proceed in reverse chronological order, beginning with Landis, in order first to identify the central premise of the modern administrative state before investigating where this premise might have originated. While Landis himself was not part of the Progressive era, his optimism reflected the Progressive political theory that had come at least a generation earlier; it was a manifestation of the idea of separating politics and administration, originated in America by Woodrow Wilson and Frank Goodnow. The final sections of the essay address the Progressive vision of replacing the separation of powers with the separation of politics and administration, and thus look to the role of Wilson and Goodnow as founders of administrative law and policymaking today.²⁸

When speaking of Wilson and Goodnow as founders of the administrative state, or of Landis as a New Deal facilitator of the growth of administration, I do not suggest that we see in the modern administrative state the complete fulfillment of the ideas of these men. The primary features of administration today—delegation, combination of functions, limited presidential control—are grounded in the notion of separating administration from politics. Like most political phenomena, however, this separation does not go quite as far in practice as it did in theory. In spite of the dramatic push toward establishing significant administrative discretion over policymaking, it still matters very much, for example, what happens in Congress and the presidency. The point of this essay is to explore the animating ideas behind the growth of the administrative state in the twentieth century, and to suggest the ways in which such ideas developed out of opposition to the animating ideas of founding-era liberalism. The principles of Wilson and Goodnow, later adopted by Landis, are in this way central to the very premise of the modern administrative state.

A. Background

When Franklin Roosevelt assumed the presidency in 1933, he embarked on a mission to enlarge vastly the scope of national government. This mission was not simply a response to the economic circumstances of the day, but was instead a conscious effort to put into practice the vision of a centralized, regulatory government that had been conceived by the Progressives a generation earlier. Such a mission was made clear by Roosevelt in the 1932 presidential campaign, where he identified Wilson as a key source of New Deal principles.²⁹

Roosevelt's advisor Felix Frankfurter looked to assemble a team to assist the new administration in crafting a plan to implement the goals on which Roosevelt had campaigned. Among others, Frankfurter called on his junior colleague at Harvard Law School, James Landis. Commissioned to work on securities legislation, Landis believed that the only way for the national government to manage capably the scope of affairs outlined in the New Deal program was to establish and empower a variety of regulatory agencies, whose expert staffs would be given significant discretion to implement Roosevelt's broad goals. It was through his work on securities legislation, and his subsequent service on the FTC and SEC, that Landis became the animating force behind the growth of modern administration as we know it today.³⁰

Landis wrote two substantial works of interest on the principles of the administrative state. The first was his seminal work The Administrative Process, which was based upon lectures Landis delivered after leaving the SEC in 1937 to become dean of Harvard Law School. As Harvard administrative law scholar Louis L. Jaffe has written, “his lectures in 1938 became inevitably a celebration, a defense, and a rationalization of the magnificent accomplishment in which he had played so brilliant a part.”³¹

Landis left Harvard in 1946 to return to public service as chairman of the Civil Aeronautics Board, but his experience there was disappointing, as President Truman did not reappoint him at the end of his term. Landis ended up working for the Kennedy family and thus was an advisor to John F. Kennedy during his presidential campaign and the early part of his presidency. It was in this capacity that Landis wrote his Report on Regulatory Agencies to the President-Elect in 1960, which contains some very useful reflections on the status of the administrative state that he had helped to form.

B. The evolutionary critique of constitutional formalism

In both of Landis's major works, the traditional forms of the Constitution are critiqued from an evolutionary point of view. Landis contended that there can be no fixed understanding of the ends and scope of government; rather, the role and even the structure of government must be constantly adjusted so that it can deal with the exigencies of the present. The growth of modern administration, which Landis fully conceded does not fit within the form of American constitutionalism, is a manifestation of this evolutionary critique. “In terms of political theory,” Landis explained, “the administrative process springs from the inadequacy of a simple tripartite form of government to deal with modern problems.”³²

In an entirely functionalist or pragmatic mode, the ultimate standard for Landis was efficiency. Government ought not do what the Constitution prescribes; rather, it ought to do that which enables it to deal most efficiently with the problems of the present. “There is no doubt,” Landis reasoned, “but that our age must tolerate much more lightly inefficiencies in the art of government.”³³

Landis was fully aware that with the advent of national administration as he envisioned it, the very understanding of the ends of government in the United States had changed. Under the old liberalism, he knew, government was erected for the sake of securing individual rights. The framers of the Constitution thus acted correctly when they implemented the separation-of-powers framework; this was a framework, Landis agreed, that would keep the national government focused on rights protection as its fundamental end. Nonetheless, modern government had to move beyond the separation of powers, since the end of government had changed from rights protection to what Landis called the “promotion of the welfare of the governed” or, more generally, “well-being.” For Landis, the establishment of the Interstate Commerce Commission (ICC) in 1887 marked the key turning point in America's self-understanding of the ends and scope of government. “What was important” about the ICC, Landis explained, “was the deliberate organization of a governmental unit whose single concern was the well-being, in a broad public sense, of a vital and national industry.” As for the understanding of the old government that the end of government was rights protection, Landis simply explained that “concessions to rectify social maladjustments thus had to be made.”³⁴

One of the major concessions to which Landis referred is that the new vision of national administration does not comport with the separation-of-powers structure of the Constitution. As I have noted, the original view of administration placed it wholly within the executive branch of government, subject to the superintendence of an elected president vested with “the executive power.”³⁵

For Landis, however, administration cannot be confined to exercising solely executive powers. In order for an agency to operate with maximum efficiency in its regulation of an industry for the purpose of promoting “well-being,” rulemaking, investigatory, prosecutorial, and adjudicatory powers need to be combined and placed at the agency's disposal.³⁶ This kind of combination is precisely what Landis liked about the ICC, where “a government had to be provided to direct and control an industry, and governance as a practical matter implied not merely legislative power or simply executive power, but whatever power might be required to achieve the desired results.”³⁷ “The resort to administrative process,” he subsequently explained, “is not, as some suppose, simply an extension of executive power.”³⁸ Thus, Landis reasoned that the Constitution cannot be read to require strict adherence to the separation of powers. Instead, the organization of the government's powers must be determined by the facts on the ground, and therefore he urged that interpreters look to “constitutional history”—that is, how the Constitution had been interpreted to accommodate historical changes—as opposed to the forms of the Constitution itself.³⁹ This new reading of the Constitution amounts to an admission that there exist legitimate powers of the national government that are not directly granted by the Constitution. Landis lamented that some critics insisted on finding a place for administration within the existing constitutional structure. This lament is what characterizes his response to the Brownlow Commission, which had been appointed by President Roosevelt in 1936 for the purpose of exerting executive supremacy in national government. Its membership consisted of Luther Gulick, Charles Merriam, and Louis Brownlow, who served as its chairman.⁴⁰ The Commission had been critical, to a degree, of administrative discretion. For Landis, the Commission had responded “hysterically” to the administrative process, and he mocked its labeling of administration as a “fourth branch.” Such criticism, he explained, came from too strong a fixation on the number “three” in thinking about the organization of government.⁴¹

C. The nature and extent of administrative power

The most significant reason that Landis's vision of administration could not be reconciled with original American constitutionalism was that he could not tolerate any notion of limiting the discretionary authority of administrators. Not only would administrators need to have at their disposal the combination of legislative, executive, and judicial powers, but Landis did not like talk of placing general limits on the extent to which administrators could employ these powers. Such talk, Landis reasoned, would violate the functionalist tenet of using necessity to determine the scope of state power. “Generalization as to the allowable limits of administrative discretion is dangerous,” he argued. He pointed to the Securities and Exchange Act of 1934, which he had helped to draft, as an example of how to create an agency with powers flexible enough to meet unforeseen exigencies. The discretionary language with which the act empowered the SEC was a vast improvement, Landis contended, over the 1933 Securities Act, which had placed authority for securities regulation in the FTC and limited the ways in which commissioners could operate.⁴²

Landis conceded that his vision of administration had to overcome a substantial objection: that it threatened rights by combining prosecutorial and adjudicatory powers in the same set of hands. Recognizing the need to answer this objection, Landis suggested that his lectures on the administrative process would offer a response to it. His response, however, simply falls back on the functionalist tenet of necessity or efficiency. He defended the combination of functions within agencies by describing the history of how such a combination came about—by explaining that necessity thrust such a change upon government. Furthermore, Landis believed that there are more than sufficient safeguards within agencies to guard against potential abuses; these safeguards, he contended, may not be “traditional,” but they are adequate.⁴³

Ultimately, the main requirement Landis seems to have had for ensuring competence in administration is the shielding of administrators from political influence. Since administration was politically neutral and based on the administrator's expertise in a particular area, political control could only corrupt the quality of administrative decision making. Administrators should, therefore, have the discretion to apply their expertise and to make their own determination of how best to serve the public interest in particular cases within their jurisdiction, free from the specific guidance of those elected by the people to serve their interests. In what has to be one of the clearest expressions of this sentiment that one is likely to see anywhere, Landis stated:

Landis's The Administrative Process catalogues the virtues that he believed are generated by securing maximum independence for administrators from political influence. In addition to the rule of expertise, Landis also emphasized consistency in administration as a primary virtue.⁴⁵

As a rule, administrators should have long tenure, which would both insulate them from political control and ensure consistency even across periods of political change.⁴⁶

This drive to separate administration from politics made Landis a strong admirer of the Supreme Court's Humphrey's decision in 1935. As I have noted, this decision limited the president's ability to remove commissioners of the FTC, thus creating an important precedent for shielding administrators from political influence. “The real significance of the Humphrey doctrine,” Landis observed, “lies in its endorsement of administrative freedom of movement.”⁴⁷

Landis himself was involved in this case in two ways. First, he was serving as a clerk for Justice Louis Brandeis when the Supreme Court, in the 1926 case of Myers v. United States,⁴⁸ handed down the precedent that Humphrey's subsequently overturned. Writing for the Court in Myers, Chief Justice William Howard Taft laid out a broad precedent requiring presidential control over executive branch officers (and thus the ability of the president to remove them at will). Both Landis and Brandeis recognized the threat that this ruling posed to the development of a modern administrative apparatus with significant discretionary powers and independence from the realm of politics, and thus Landis worked closely with Brandeis on his Myers dissent. Landis and Brandeis were particularly alarmed that Taft had gone out of his way to include independent commissions among the executive entities over which the Constitution required direct presidential control. In the second instance of Landis's involvement, the individual whose estate brought suit for back wages in the Humphrey's case was William E. Humphrey, who was a member of the Federal Trade Commission when Landis himself was appointed to the FTC by Franklin Roosevelt. Humphrey was a holdover on the FTC, and thus a continuous thorn in the side of New Dealers like Landis. Ironically, it was Landis himself who convinced the Roosevelt administration that the president had the authority to remove Humphrey at will, since Landis was so familiar with the Myers precedent. When Roosevelt eventually lost the case because he had followed Landis's advice, Landis was nonetheless quite pleased: the defeat for the president was meaningless in comparison with the great independence for administrators that the Humphrey's decision helped to secure.⁴⁹ This incident also helps to show that Landis was much more concerned with the freedom of administrators from politics than was Roosevelt. It was, after all, partly as a result of Roosevelt's belief that administrators should be tied more closely to his political agenda that he appointed the Brownlow Commission, the conclusions of which Landis ridiculed in The Administrative Process.⁵⁰

When Landis reflected on the status of administration in his 1960 Report on Regulatory Agencies, he continued to urge an even greater independence for administrators from political influence. To the extent that administrative agencies had failed to live up to the vision with which they were created, Landis made clear that the problem was not insufficient political control, but rather an insufficient separation of administration from politics. This is why Landis criticized ex parte contacts in administrative decision making⁵¹

—where administrators seek or receive input from other parties in the political institutions—and why he warned that politics was playing too great a role in the appointment of administrators. As he elaborated:

Landis readily acknowledged that keeping administrative decision making free from the influence of politics, and providing administrators with the latitude to exercise legislative, executive, and judicial powers, weakened the protection for individual rights contained within the traditional doctrine of the rule of law. The rule-of-law principle, after all, requires that administrators act only in accord with a preestablished rule. Absent such a principle, individuals and companies have no idea what they have to do in order to avoid the enforcement power of the state, thus putting them at the mercy of administrators who, literally, make up the rules as they go along. Landis complained, however, that the modern need for efficiency and discretion in administration simply made it impossible for administrators to be bound by such an inflexible system.⁵³

Furthermore, administrators would be hampered if their decisions were subject to review by an independent court. This bedrock protection for rights, Landis explained, interfered with the notion that administrators were the most expert in their field and thus were the most qualified to adjudicate disputes where individuals or companies were accused of wrongdoing.⁵⁴

He even lamented that the problem with allowing individuals or companies to have their cases heard in traditional courts is that such courts must be bound by the “limitations” of the law and judicial procedure put into place to protect rights. Traditional courts must rely only on evidence presented at trial, for example. When administrators adjudicate, by contrast, they do not have to be bound by the evidence presented, but can instead employ their own expertise and understanding of what is truly in the public interest. “For [the administrative] process to be successful in a particular field,” Landis argued, “it is imperative that controversies be decided as ‘rightly’ as possible, independently of the formal record the parties themselves produce.”⁵⁵ At the very least, Landis considered it essential that agency fact-finding not be second-guessed by an independent judiciary; otherwise, “the very efficiency of the system becomes threatened.”⁵⁶ These sentiments make perfect sense when one recalls that, for Landis, the ends of government had evolved. Getting a decision “right” in terms of the “well-being” of the general public interest was, to Landis, a far more important purpose of government than following specific rules and procedures designed to ensure the maximum protection for individual rights.

D. Why trust administrators?

It is not quite correct to say that Landis disregarded the importance of individual rights. It is the case, rather, that Landis so strongly trusted the objectivity and expertise of administrators that he could not conceive it likely that they would reach unjust decisions, so long as they were given sufficient freedom to do their jobs. It is this trust—or what one commentator has called his “exuberant optimism”⁵⁷

—that is really at the core of Landis's entire outlook on reforming government. As we will see in subsequent sections of this essay, Landis was a true Progressive at heart; his trust in the enlightenment and disinterestedness of administrators stands as an instructive contrast to the permanent self-interestedness that the framers of the U.S. Constitution saw in human nature.⁵⁸ Just as this sobriety about the potential for tyranny led the framers to circumscribe carefully the authority of the national government, Landis's passionate optimism fueled his call for maximum discretion for administrators. This is not to suggest that the framers denied discretionary power to the national government; no reader of Federalist No. 23, or many other papers of The Federalist for that matter, could draw such a conclusion. Rather, the framers understood that such discretion had to be channeled through the forms of the Constitution in order to be safe for liberty. Thus, as Alexander Hamilton explained in Federalist No. 23 and elsewhere, the vigorous discretion that the national government must have is made safe by the “most vigilant and careful attention of the people.”⁵⁹ For the people to exercise this kind of vigilance, the officers who exercise discretion must do so in a system of clear, electoral accountability. It is precisely this kind of accountability to the realm of politics from which Landis, by contrast, wanted to free administrators.

For Landis, there is something special about civil servants that somehow raises them above the ordinary self-interestedness of human nature. As we will see, for Landis's forebears in the Progressive era, such confidence in civil servants came from a faith that the progressive power of history had elevated public servants to a level of objectivity. They would, accordingly, be able to disregard their own private or particular inclinations in order to dedicate themselves to the objective good. Landis himself spoke glowingly of the selflessness of administrators, describing them as “men whose sole urge for public service is the opportunity that it affords for the satisfaction of achievement.”⁶⁰

Because of this disinterestedness, restraints on their discretion were unnecessary. Landis took this notion even further, suggesting that the selflessness and objectivity of administrators created a kind of moral imperative that the country allow them to rule:

For Landis, this dedication and the expertise of administrators together cure any potential problem that might come from removing the traditional separation-of-powers protections for individual rights. In the case of granting adjudicatory power to administrators, for example, Landis explained that this is a good thing both for individuals and for the common good, because the expertise of administrators makes them more competent than traditional courts and thus more likely to reach a just result. Therefore, he argued, “it is only delay that results from the insistence upon independent judicial examination of the administrative's [sic] conclusion.”⁶²

E. Reconsiderations?

Landis's faith in the competence and objectivity of administrators had not receded when he wrote his 1960 evaluation of the system he helped to create. It is true that Landis was disappointed with the record of what agencies had done with their discretion, and he had also had his own unsatisfactory experience with national administration. He even made what must have been a painful concession, admitting that “a deterioration in the quality of our administrative personnel has taken place, both at the top level and throughout the staff.”⁶³

But Landis did not draw from his disappointment over the actual experience of administration the conclusion that agencies might not have been so deserving of the trust he had placed in them or of the latitude they were given as a consequence of the New Deal. “Their continued existence is obviously essential for effective government,” Landis wrote. “The complexities of our modern society are increasing rather than decreasing…. [These complexities] all call for greater surveillance by government.”⁶⁴ Landis's 1960 evaluation is thus animated by a call for agencies to do even more, and for them to exercise even greater discretion in making national policy. If anything, in Landis's view, agencies had not abused their discretion but had, instead, been too deliberative and cautious.⁶⁵

The question for Landis in 1960, therefore, was how administration could be helped along to fulfill the great role that he still envisioned for it. First, Landis urged more money and more manpower. It was a frequent observation of Landis's report that agency inaction seemed to come, at least partly, from too great a workload. With greater budgets and staff, agencies would be able to regulate even more zealously.⁶⁶

Second, Landis saw a need to attract higher quality personnel into administrative positions. The problem was not that administrators had abused their trust, but rather that they had not been trusted enough. The way to attract better administrators, Landis argued, was to put even more power in administrative positions, thus demonstrating to the most talented individuals that national administration was a place where they could make a real difference. Thus, Landis advocated further insulating administration from politics, and contended that political interference had harmed the morale of agencies and had even kept good men away from administration entirely. The other way to attract higher quality administrators was to offer higher compensation and longer tenure. Longer tenure would, especially, give administrators “freedom from worry” about political influence.⁶⁷ Landis also worried about excessive procedural impediments to agency action. Perhaps more than anything else, regulatory activity had been slowed because agencies were being overly cautious in their procedures. Again, the problem was not too much discretion for agencies, but too little. As Landis observed, “The tendency here is again further to judicialize [sic] the administrative process and, in the opinion of many observers, to over-judicialize it to a point where stagnation is likely to set in.”⁶⁸

If Landis's faith in the competence and objectivity of administrators was at odds with the founders' sobriety over the self-interestedness of human nature, experience with the poor quality of administration did not lead him to change his mind. When Landis fretted in 1960 that administrators were making decisions out of concern for advancing their careers, or that administrators might be attracted by more money or increased job security, he did not draw the conclusion that these observations might just be evidence that the founders were correct about the self-interestedness of human nature. He wanted, instead, to see administrators take on an even greater role in national governance.

At first glance, Landis's faith in the objectivity of administrators, expressed in 1938 and affirmed in 1960, may seem at odds with a subsequent development in 1970s administrative law often referred to as “capture theory.” This development refers to a growing sentiment, especially among environmental interest groups, that administrators too frequently became “captured” by the industries and businesses they were supposed to regulate. These interest groups began to turn, in the 1970s, to increasingly receptive federal courts, seeking intervention either against agency inaction or against agency action deemed to be insufficiently vigorous.⁶⁹

In one sense, such action by interest groups and courts suggests a lack of confidence in the ability of administrators to do their jobs—not a sentiment one would easily associate with Landis. It is also the case, however, that Landis's views in 1960 prefigured the sentiments of activists in the 1970s. That is, Landis wanted agencies to do more, to be more active and zealous in regulating. And Landis, too, worried about the “capture” of regulators:

Landis and the 1970s activists thus shared a vision about the proper role of administration; both believed that administrators needed to be out front in acting on behalf of the public good. Both shared, in other words, in the original, New Deal vision for the role of agencies in national governance. Landis did not call, of course, for increasing judicial intervention in order to make administrators fulfill his vision for them; he thought that more money, power, and freedom of action would solve the problem. But the basic understanding of the broad role and mission of administration remained relatively consistent throughout the post–New Deal period. No one, in other words, sought to revive the nondelegation doctrine as a means of correcting any problems that were perceived with agencies or those who staffed them.

A. The evolving ends and scope of government

Like Wilson, Goodnow argued that government needed to adjust its very purpose and organization to accommodate modern necessities;⁸⁸

and, like Wilson, he believed that history had made obsolete the founders' dedication to protecting individual rights and their consequent design of a carefully limited form of national government. In Social Reform and the Constitution, Goodnow complained about “reverence” for the Constitution, which he regarded as “superstitious” and an obstacle to genuine political and administrative reform.⁸⁹ In Politics and Administration, Goodnow made clear that his push for administrative reform was not simply or even primarily aimed at correcting the corruption of the spoils system. Rather, administrative reform was, for Goodnow, instrumental to the end of achieving Progressive, big-government liberalism. Progressives had in mind a wide array of new activities in which they wanted national-government involvement; such involvement could not be achieved with the old system of placing administration under political direction:

Understanding administrative reform this way—as a means to securing the broader aims of Progressive liberalism—is what makes the work of Goodnow, and Wilson too, so much more significant to the development of modern American thought and politics than had been the case with the civil-service reformers.

Goodnow and his fellow Progressives envisioned an almost entirely new purpose for the national government. Government itself, therefore, had to be viewed through a historical lens. The principles of the original Constitution, Goodnow reasoned, may have been appropriate for the founding era, but now, “under present conditions, they are working harm rather than good.”⁹¹

The error that the founders made was not in constructing government as they did, but rather in thinking that their particular construction and manner of conceiving politics would transcend their own age, and would be appropriate for future ages as well. They did not realize the historical contingency of their principles.⁹² The modern situation, Goodnow argued, called for less focus on constitutional principle and form, and much greater focus on empowering and perfecting administration. He even repeated, using almost the same words, Wilson's proclamation from 1885 that the nation had to move from constitutional to administrative questions. Wrote Goodnow: “The great problems of modern public law are almost exclusively administrative in character. While the age that has passed was one of constitutional, the present age is one of administrative reform.”⁹³ In order to address the administrative questions that history was pressing upon the nation, Goodnow urged a focus not on the “formal” governing system (that is, the rules and procedures of the Constitution) but on the “real” governing system, which becomes whatever is demanded by the necessities of the time.⁹⁴

The focus of the old liberalism on government's permanent duty to protect individual rights was an impediment to the marked expansion of governmental power that Progressives desired; thus, the ideas that animated the old-liberal conception of government had to be discredited. Goodnow understood the political theory of the founding quite well. He knew that the notion that government's primary duty was to protect rights came from the theory of social compact—a theory which held that men are naturally endowed with rights prior to the formation of government, and thus consent to create government only insofar as it will protect their natural rights. The founders' system of government, Goodnow acknowledged, “was permeated by the theories of social compact and natural right.” He condemned these theories as “worse than useless,” since they “retard development”⁹⁵

—in other words, they prevent the expansion of government by their focus on individual liberty. The separation-of-powers limits on government, Goodnow realized, came from the founding-era concern for individual liberty: “It was the fear of political tyranny through which liberty might be lost which led to the adoption of the theories of checks and balances and of the separation of powers.”⁹⁶

Goodnow's critique of the founders' political theory came from the perspective of historical contingency. Their understanding of rights and the role of government, he argued, was based upon pure “speculation” and “had no historical justification.”⁹⁷

Here Goodnow employed the same critique as his fellow Hegelian Wilson, who had written in 1889 that the idea of social compact had “no historical foundation.”⁹⁸ Instead of an understanding of rights grounded in nature, where the individual possesses them prior to the formation of government, Goodnow urged an understanding of rights that are granted by government itself. He remarked favorably upon European trends in understanding rights as contingent upon government: “The rights which [an individual] possesses are, it is believed, conferred upon him, not by his Creator, but rather by the society to which he belongs. What they are is to be determined by the legislative authority in view of the needs of that society. Social expediency, rather than natural right, is thus to determine the sphere of individual freedom of action.”⁹⁹

Goodnow found it necessary to critique the theory of natural rights because he knew it was the foundation for the requirement of government based upon consent.¹⁰⁰

The principle of government by the consent of the governed was a problem for Goodnow and those with his vision of administrative power. Goodnow's vision required significant deference to expertise. The empowering of administrators, as he saw it, was justified not because the administrators had the consent of the people, but because they were experts in their fields. This is why Goodnow wanted to improve administration not by making it more accountable to politics and to the consent of the governed, but by making it less accountable. He observed and conceded that the doctrines of “sovereignty of the people and of popular participation in the operations of government” were an integral part of American political culture. He therefore acknowledged that this aspect of the culture would be a difficult hurdle for his vision of administration to overcome. “Our governmental organization developed,” he explained, “at a time when expert service could not be obtained, when the expert as we now understand him did not exist.”¹⁰¹

B. Discretion for administrators and the separation of politics and administration

Since administrative experts were now available, Goodnow urged that they be employed and empowered with significant discretion to manage the new tasks that Progressives had in mind for the national government. In Politics and Administration, Goodnow featured chapters on the “Function of Politics” and the “Function of Administration.” Both chapters have exactly the same aim: to show that politics ought to be kept out of administration's way. Put differently, the function of politics is to refrain from influencing administration, and the function of administration is to exercise discretion by avoiding the influence of politics. Goodnow's distinction between the roles of politics and administration was a response to what he believed had been the corrupting influence of politics on national administration. While Goodnow, for example, championed the potential of the party system (as a means of circumventing the separation of powers), he did so only under the condition that party politics be kept entirely out of administration.¹⁰²

Administrative decision making, Goodnow argued, was politically neutral and required expertise. Where administration had been “subjected too much to the control of politics in the United States,” he claimed, the effect had been the “decreasing of administrative efficiency.”¹⁰³ Thus, political institutions, Goodnow concluded, ought to concern themselves only with ensuring that administrators remained neutral, and such institutions were to be kept out of influencing the substance of administrative decisions. He explained: “All that the legislature, or any political body, can do is to see to it, through the exercise of its control, that persons discharging these administrative functions are efficient and impartial. Their general conduct, but not their concrete actions, should be subject to control.”¹⁰⁴

Goodnow was well aware that insulating administration from political control in this way ran up against the traditional, constitutional role for administration, where administrators are subservient to the chief executive. He explained that his conception of administration was novel, since it considered the sphere of administration to lie outside the sphere of constitutional law; indeed, this new conception was exactly what Wilson had given Goodnow credit for in 1894. Emphasizing the distinction between the constitutional and administrative spheres, Goodnow remarked that the student of government “is too apt to confine himself to constitutional questions, perhaps not considering at all the administrative system.”¹⁰⁵

It is for this reason of considering administration as an object of study outside of the Constitution that Goodnow's landmark book on administrative law, Comparative Administrative Law, relies almost entirely upon an account of foreign administrative systems.¹⁰⁶ He knew, as Wilson did, that such a concept was a novelty in the American political tradition. Modern administrative law, therefore, would take it for granted that the political branches of government had to cede significant discretion to administrative agencies; the new body of law would be dedicated to establishing a framework for governing the extent and organization of this discretion.¹⁰⁷

Careful readers of Goodnow's work will note that in making his case for freeing administration from political influence, he did not speak of a strict or rigid separation between politics and administration; indeed, Goodnow noted that the boundary between the two is difficult to define and that there would inevitably be overlap.¹⁰⁸

This overlap, however, seems to be in one direction only, in a manner that enlarges the orbit of administration; that is, Goodnow seemed to contemplate instances where administrative organs would exercise political functions, but he apparently did not contemplate instances of political organs engaging in administrative activity. He characterized the function of politics as “expressing” the will of the state, while the function of administration is to “execute” the will of the state. He made it clear, however, that the overlap between politics and administration would come in the form of administrative agencies taking a share in “expressing” as well as “executing” state will:

The notion that Goodnow might see administration as subordinate to politics—as confined only to executing previously expressed will¹¹⁰

—is hereby called into question. Moreover, Goodnow's statement essentially laid the foundation for the abandonment of the nondelegation doctrine. He elaborated: “As a result, either of the provisions of the constitution or of the delegation of the power by the legislature, the chief executive or subordinate executive authorities may, through the issue of ordinances, express the will of the state as to details where it is inconvenient for the legislature to act.”¹¹¹

Goodnow shed further light on the role administration might play in “expressing” the will of the state when he categorized the kinds of decisions in which administrative agencies are legitimately involved. When administrators make “executive” decisions, he explained, they are subordinate to previously expressed will. But in reasoning that would later be adopted by the Supreme Court in the Humphrey's case, Goodnow contended that administrators engage not only in “executive” decisions but also in decisions that are quasi-legislative and quasi-judicial. In doing so, administrators not only execute the will of the state, but also play a role in its expression. When administrators are merely executing state will, Goodnow explained, it is appropriate for them to be subject to the supervision of politics. When they are expressing state will in their quasi-legislative or quasi-judicial role, however, administrators require freedom from political influence. As Goodnow wrote, the “executive function must therefore of necessity be subordinated to the function of politics. No such close connection, however, exists between the function of politics and the other branches of the administration of government. No control of a political character can bring it about that administrative officers will discharge better their quasi-judicial duties, for example….”¹¹²

In other words, when administrators are merely functioning as executives by following previously expressed will, political accountability can be brought to bear; but it is at the point when they exercise even more power, acting as legislators or judges by aiding the expression of state will, that they must be the most free from political accountability.

C. Why trust administrators?

The key to trusting administrators with the kind of discretion that Goodnow envisioned was his profound faith in the expertise and objectivity of the administrative class—just as it had been for Wilson and would be for Landis. Administrators could be freed from political control because they were “neutral”: their salary and tenure would take care of any self-interested inclinations that might corrupt their decision making, liberating them to focus solely on truth and the good of the public as a whole. As Goodnow explained, “such a force should be free from the influence of politics because of the fact that their mission is the exercise of foresight and discretion, the pursuit of truth, the gathering of information, the maintenance of a strictly impartial attitude toward the individuals with whom they have dealings, and the provision of the most efficient possible administrative organization.”¹¹³

A natural objection here would be that freeing administrators from political control is a recipe for corruption: that it is precisely through the electoral connection of public officials that we “make their interest coincide with their duty,” as Hamilton puts it in The Federalist.¹¹⁴ For Goodnow, though, it was just this connection to electoral politics that would make administrators corrupt, while the absence of accountability to the electorate would somehow make them pure. Politics, Goodnow explained, is “polluted” and full of “bias,” whereas administration is all about the “truth.”¹¹⁵

Goodnow's confidence in the objectivity of administrators, like Wilson's, is a sign of his Hegelianism, and it shows that he accepted Hegel's premise that bureaucrats could be freed of their particularity and could will only the objective good of the state.¹¹⁶

In his introduction to the recent edition of Goodnow's Politics and Administration, bureaucracy scholar John A. Rohr concludes that Goodnow's “Hegelian starting point” doomed his attempt at crafting a realistic administrative reform.¹¹⁷ Goodnow himself worried, in 1900, that the emphasis in American political culture on popular sovereignty and individual rights would pose a formidable obstacle to the enactment of his vision for administration. The people, given their historical background, might resist deferring to a permanent, administrative class:

Goodnow also worried that this public hostility to “experts” would make it hard to attract talented individuals into the bureaucracy, although he subsequently consoled himself with the possibility that change was on the horizon. In 1916, Goodnow hoped he was seeing a change in attitude, where the people were becoming less attached to their sovereignty so that experts could be allowed to rule:

D. Goodnow's new categories of state power

Goodnow's vision for administrative governance is grounded in a set of categories he employs to describe the different types of state power. Contrary to the American separation-of-powers framework, which divides power into three categories, Goodnow's framework contains only two. When the state exercises power, Goodnow argued in Politics and Administration, it does it either by “expressing” will or by “executing” will that has previously been expressed.¹²⁰

This is an elaboration of an expression that Goodnow had employed in his first book, Comparative Administrative Law, where he referred to the “expression of the will of the state.”¹²¹ The distinction between “expressing” and “executing” translates directly into Goodnow's distinction between “politics” and “administration”—the very distinction at the heart of the administrative state today. As he explained, “There are, then, in all governmental systems two primary or ultimate functions of government, viz. the expression of the will of the state and the execution of that will. There are also in all states separate organs, each of which is mainly busied with the discharge of one of these functions. These functions are, respectively, Politics and Administration.”¹²² As Wilson had done in his essay “The Study of Administration,” Goodnow emphasized the separation between politics and administration in order to pave the way for importing a foreign administrative science into the American system of government. Since administration, under this scheme, is not involved in politics—since it is only involved in “executing” previously expressed will—Goodnow saw no danger in adopting a monarchic system of bureaucracy in a democratic regime. As I have noted, however, Goodnow did not always talk so strictly or rigidly about the separation between politics and administration, and he acknowledged that there are many instances where administration will have a role in “expressing” as well as “executing” state will.

Within that part of the state responsible for “executing” will, Goodnow made a further division into three parts. It is this division of the executive branch of government that is central to his argument for the independence of administration and that represents the most distinctive part of his overall argument for reform. The three elements of the executive are: (1) the judiciary; (2) the administrative agencies; and (3) what would traditionally be called the constitutional executive—the president, state governors, etc.¹²³

What most immediately stands out here, of course, is the inclusion of the judiciary within the executive. Goodnow explained that the founders had erred in following Montesquieu's portrayal of the judiciary as its own independent branch of government. Judges, Goodnow reasoned, were essentially involved in the execution of previously expressed will or law, just like the other elements of the executive.¹²⁴ In addition to the inclusion of the judiciary within the executive, it is also worth noting the distinction that Goodnow made between the president and the administrative agencies. Whereas the Constitution would have to consider agencies simply as an extension of the presidency (the president being vested with all of “the executive power”), Goodnow clearly separated agencies from the political office of president.

Goodnow's conception of the executive is also unique in that it considers both courts and agencies to be part of administration. The executive, he explained, involves both the “administration of justice” (by the courts) and the “administration of government” (by agencies). And within the “administration of government,” agencies can perform different functions; some of these are purely executive, but others take on more of a judicial or even legislative character.¹²⁵

What matters most is that neither kind of administration—neither that done by the courts nor that done by agencies—is confined to the mere execution of previously expressed will. These traditionally administrative elements commonly cross over into the realm of politics, helping to express as well as execute the will of the state. This is true of courts in particular, since they interpret both the law and the Constitution itself.¹²⁶

Goodnow emphasized the independence from political influence that American courts had traditionally enjoyed. But whereas the original idea behind this independence had been the securing of individual rights,¹²⁷

for Goodnow it served as a vehicle for granting vast discretion to administration. The key is Goodnow's placing of the courts and agencies alongside one another in the realm of administration. Since the courts traditionally enjoyed independence from politics, and since agencies are partners with the courts in the realm of administration, it follows that agencies (like courts) ought to enjoy independence from political control. Administration—whether it be “administration of justice” in the courts, or “administration of government” in the bureaucracy—is politically neutral and ought not be influenced by the vicissitudes of elections, even though, as Goodnow himself made clear, administration must necessarily involve itself in not only executing but giving expression to the will of the state.¹²⁸ As Rohr puts it, “by assimilating administrators to judges whom he includes within the concept of executive and excluding constitutional executives from administration, Goodnow neatly carves a niche for the administrator to exercise legitimately a certain degree of independence.”¹²⁹ Put another way, Goodnow tapped into a traditional feature of American constitutionalism—the relative independence of the judiciary—to argue for something quite untraditional—the independence of administration from political control.

Goodnow's employment of independent courts to secure a similar independence for administrative agencies is useful in placing in its proper perspective the “capture” phenomenon of modern administrative law. As I have noted, much of modern administrative law consists of fights between agencies and judges, who are spurred on by interest groups to intervene in agency decision making out of a belief that agencies are too lenient on regulated industries. From one point of view, such a phenomenon demonstrates that agencies no longer have the discretion that they once enjoyed. Understood from Goodnow's perspective, however, such wrangling between courts and agencies would be largely inconsequential, or merely at the level of an intrafamily squabble, since both courts and agencies were considered by Goodnow to be part of administration. What mattered for Goodnow was not whether courts or agencies had the upper hand, but rather that both of them, as part of administration, were independent of politics. This is also the important point for those of us who are interested in the fate of traditional constitutionalism in today's administrative state: it ought to matter less exactly which unelected entity is most influential in policymaking (agencies or courts), and it ought to matter more that such power has been delegated in the first place by the elected branches of government.