James Landis is widely credited with crafting the theoretical architecture supporting President Roosevelt’s radical reconstruction—and expansion—of the federal government. Landis shrewdly both established and legitimized the regulatory state, including Roosevelt’s creation of new federal administrative agencies, by offering the regulatory state as the solution to the problem of modern governance: the administrative state “is, in essence, our generation’s answer to the inadequacy of the judicial and legislative process.” The Landis premise took concrete shape through Roosevelt’s expansion of the regulatory state, and in doing so, it brought to fruition Woodrow Wilson’s progressive intellectual project: rule by experts, insulated from the popular will
Landis believed the “the administrative process” for which he advocated would “spring from the inadequacy of a simply tripartite form of government to deal with modern problems” because modern problems were simply too large and complex to be entrusted to the system based on the separation of powers instituted by our nation’s founders. Landis framed this innovation as consistent with separation of powers principles because he believed the separation of powers called both for separation but also coordination among the branches, and he saw the administrative state as essential to creating that coordination:
If the doctrine of separation of power implies division, it also implies balance, and balance calls for equality. The creation of administrative power may be the means for the preservation of that balance, so that paradoxically enough, though it may seem in theoretic violation of the doctrine of the separation of powers, it may in matter of fact be the means for the preservation of the content of that doctrine.
What the tripartite branches could not coordinate among themselves directly, Landis believed administrative agencies could coordinate as a substitute. Landis then aimed to create administrative agencies that themselves combined the three aspects of government. Years later, the Administrative Procedure Act codified this three-branches-in-one-agency approach to administrative power, defining not only rulemaking authority for federal agencies (a quasi-legislative power), but also adjudicative authority (a quasi-judicial power).
In reality, Landis’s three-branches-in-one-agency theory never comported with the separation of powers principles that the founders embedded in our Constitution. But even if it could have been reconciled with those principles as a theoretical matter, the past 100 years have demonstrated that the administrative state is the single biggest threat that faces the Constitution and the republic it establishes. What began as a type of separation-of-powers “innovation” beyond the Constitution has persisted as nothing less than tyranny. The vast majority of our governance today is created, maintained, and enforced by unelected bureaucrats who are almost entirely insulated from accountability to any branch of government, let alone the people.
This reality was never on fuller display than during the Trump Administration, as I witnessed firsthand. From President Trump’s inauguration forward, the recalcitrant federal bureaucracy slow- walked his policies, including policy promises that were central to his victorious 2016 campaign (and that therefore commanded significant support from the American people). The Army Corps of Engineers dragged its feet in finalizing plans for the construction of a border wall. The Department of Education refused to withdraw Obama-era memoranda on Title IV and disparate impact. Bureaucrats at the Department of State ultimately blocked efforts to require “extreme vetting” for foreign nationals entering the United States. The idea that the federal bureaucracy is accountable to the president is a mirage.
And yet, for decades now, conservatives have failed to mount any fundamental challenge to the central Landis claim undergirding the administrative state: the inadequacy of the self-governing tripartite branches. There lies the problem for conservative reforms of the administrative state as they have been proposed for the last 40 years. Landis believed the complexity of modern problems demanded the administrative state as a solution, and by and large, even conservatives have agreed.
In fact, when conservatives have dared to oppose the administrative state, they have framed their opposition through an economic lens: the administrative state is a vehicle for regulation and government control of the market. As such, conservatives’ tools for combatting it have focused almost exclusively on curtailing the authority of the administrative state to promulgate new regulations and affixing costs to its enactments. In this view, the administrative state as seen through green eyeshades is a problem only because it is profligate and burdens the marketplace, not because “coordination” may now work in conflict with the policy preferences and reanimated desires for political control of a free people. The tyranny of the administrative state is not merely an economic tyranny: it is a tyranny over all purposes of government, a capturing of the people’s power over all political questions, not merely pocketbook questions.
Perhaps it has been easy for conservatives to adopt the Landis premise because before FDR’s remaking of the federal government, conservatives were already committed to the idea that some modern problems were so complex they could not be resolved through the basic instruments of self-governance and instead required the intervention of experts.
When Landis was previewing his ideas publicly prior to working in the executive branch, President Herbert Hoover signed the Reconstruction Finance Corporation Act, creating a new, government-sponsored financial institution that would fit right in with the “independent agencies” of today. The Reconstruction Finance Corporation was a quasi-public corporation that borrowed its funds over its lifetime almost entirely from the federal government for the purpose of lending directly to banks and other financial institutions. The RFC was composed of professionals hired outside the civil service system, and the federal government appointed its executive officers and board of directors.
Even the leading conservative of the time, Senator Robert Taft of Ohio, favored the RFC and would later back New Deal agency programs, including subsidized loans for farmers and homeowners and accelerated public works spending. In retrospect, the RFC was a template for the New Deal federal agencies FDR later created, including the Tennessee Valley Authority, a quasi-governmental corporation, the Works Progress Administration, the Federal Communications Commission, the Federal Housing Administration, and the Securities and Exchange Commission. More importantly, it was a harbinger of decades of conservative capitulation: In creating the RFC, conservatives like Taft had essentially adopted the Progressive view that modern problems required credentialed experts and technocratic governance. As Taft would posit, laissez-faire individualism was a political-philosophical perspective that required mediation from governmental authorities.
The solution to the administrative state, however, depends on resisting the Landis premise and accepting instead that even modern problems can be solved without administrative agencies, or that the price of solving those problems is too high if administrative agencies are the only means of doing so.
Four Past Attempts to Restrain the Administrative State
The Administrative Procedure Act of 1946
The Administrative Procedure Act might be considered the first attempt at restraining the administrative state. Passed in 1946, the APA followed FDR’s Second New Deal by about a decade and came at a time of concern in the United States for the rapid rise of the administrative state. Conservatives publicly worried that its growth impaired individual liberties (by allowing federal agencies to impose regulations that burdened individuals’ freedom to work and contract, even without explicit authorization from Congress) and the free market (by allowing federal agencies to establish burdensome regulations or effectively pick “winners” and “losers” and interfere with otherwise-free markets). Liberals advocated for the administrative state based on the Landis premise—namely, that unelected experts were needed to create policies and regulations capable of meeting the demands of “modern society.”
The APA attempted to assuage concerns about the administrative state’s power by grafting onto the administrative state the same types of due process protections that applied to other branches of government. It created formal and informal rulemaking processes to regularize the administrative state’s quasi-legislative activities, and it created formal and informal adjudicative processes to regularize the administrative state’s quasi-judicial activities. It also specified conditions for review of agency action by the judicial branch.
But although the APA was seen at the time as a bipartisan compromise, it was in retrospect a compromise that leaned heavily leftward because it endorsed—and even advanced—the Landis premise. The essential compromise of the APA was biased in favor of a large administrative state: the administrative state was a necessary governmental innovation demanded by the complexity of modern society, and the only restraints Congress could place on its activities were marginal procedural protections intended to mimic the due process protections that applied to the constitutional branches of government. These protections increased public participation in rulemaking by requiring pre-rulemaking notice to and comment from the public, and they increased regularity in agency decision-making by standardizing agency processes. But they did little, if anything, to curtail the reach of federal agency power or to protect the primacy of the constitutional branches of government as set against the unelected and essentially insulated activities of the administrative state.
Many prominent conservative jurists, including Justice Antonin Scalia and D.C. Circuit Judge Kenneth Starr, spent a generation advocating for Chevron deference, which was intended to prevent liberal courts from imposing their policy preferences on the executive branch by preserving a deferentially drawn sphere of decision-making in which executive agencies were free to operate. But in protecting this deferential sphere of decision-making power, Chevron deference has ultimately proved to be incapable of checking the administrative state’s power and growth.
Chevron deference originated with the 1984 decision Chevron U.S.A. v. Natural Resources Defense Council, which created a two-part test for judicial review of the agency’s construction of a statute passed by Congress. First, a court must determine “whether Congress has directly spoken to the precise question at issue”; and if it has, and “the intent of Congress is clear, that is the end of the matter,” for both the court and the agency “must give effect to the unambiguously expressed intent of Congress.” Second, “if the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency’s answer is based on a permissible construction of the statute”; if it is, it is entitled to the court’s deference.
Chevron itself embraced the Landis premise that difficult policy questions required experts to resolve. It posited that where a statute is ambiguous, Congress might have “consciously desired . . . that those with great expertise and charged with responsibility for administering the provision would be in a better position to do so” than Congress. But even if Congress had not so determined, the opinion advocated deference to experts: “Judges are not experts in the field, and are not part of either political branch of the Government,” so it should not be for judges to resolve complex policy issues. The Chevron Court assured itself that the deference it instituted presented no separation-of-powers problem because “while agencies are not directly accountable to the people, the Chief Executive is, and it is entirely appropriate for this political branch of the Government to make such policy choice.” Today, such an argument is untenable, in light of the entrenched nature of the administrative state and the little (or, more often, utter lack) of executive control over its machinations.
Chevron deference is a legal doctrine incompatible with substantial self-governance because it translates statutory ambiguity into complete deference to the least accountable arm of modern government—the administrative state.
While conservative jurists today are more skeptical of the doctrine (and, indeed, may even be willing to replace it), the conservative jurists of yesterday embraced it. None other than Justice Scalia himself argued for a relatively expansive definition of Chevron deference. In discussing Chevron’s “step one,” Justice Scalia explained that “congressional intent must be regarded as ‘ambiguous’ not just when no interpretation is even marginally better than any other, but rather when two or more reasonable, though not necessarily equally valid, interpretations exist.” In other words, Chevron requires courts to defer to federal agencies even when those agencies adopt clearly inferior interpretations of the statutory text passed and signed by the politically accountable branches. It is no wonder, then, that the doctrine of Chevron deference has done little to check the power and proliferation of the administrative state.
More recently, conservative legislators in Congress have introduced and advocated for the REINS Act (Regulations from the Executive in Need of Scrutiny Act). Senator Rand Paul (R-Ky.) first introduced the REINS Act in 2013. The act creates categories of “major” and “nonmajor” rules and requires congressional approval by both houses of Congress before “major” rules can take effect.
The REINS Act, however, begins from the Landis premise as well—namely, that the authority to craft policy properly belongs to experts in the federal agencies. Rather than remove that power from agencies or shift lawmaking authority back to Congress in the first instance, the REINS Act leaves regulatory power with federal agencies in the very same size and scope in which it exists today and merely imposes a requirement of congressional approval on some regulatory actions. But even the definition of which regulatory actions require such approval is both ambiguous and inadequate. The REINS Act defines a “major rule” to be a rule with “an annual effect on the economy of $100 million or more,” or one that causes “major increase in costs or prices for consumers, individual industries, Federal, State, or local government agencies, or geographic regions,” or one that has “significant adverse effects on competition, employment, investment, productivity, innovation, or the ability of United States-based enterprises to compete with foreign-based enterprises in domestic and export markets.” These definitions are unsatisfactory as a drafting exercise, since they are open to interpretation and admit of ambiguities. Who will determine which rules satisfy these definitions? Who knows?
Worse, these definitions are completely inadequate because they emphasize economic impact alone, as if the administrative state poses only pocketbook harms. Edicts from the Department of Education about the treatment of trans students in the classroom; Department of Commerce regulations about the classification (and therefore, legal availability) of certain firearms and accessories; Department of Defense allowances for same-sex spouse benefits or sex-change surgeries—all of these are culturally transformative regulations that fall short of the economic impacts that trigger greater congressional oversight in the REINS Act.
The REINS Act clearly demonstrates the view of its conservative sponsors and supporters that federal agencies have too much authority to take actions with too great significance; yet rather than remove such authority from those agencies and require Congress to exercise it, these legislators are content merely to give themselves an up-or-down vote after the fact—and even then, only for regulations with considerable economic impact, not those that answer transformative cultural questions about which ordinary people and their legislators expect to express views and direct policy. Thus, even in the REINS Act, the premise that expertise, after all, lies with the agencies still reigns.
The REINS Act is notable—and rightfully criticized—for another reason, too. It provides that all other rules outside the definitions stated above are “nonmajor” rules, which Congress may disapprove under the REINS Act. But surely this is a fact that need not be stated. Of course Congress can negate an action of a regulatory agency if it chooses. The fact that legislators see the REINS Act as a vehicle to state that power is alarming, but it is also illustrative of Congress’ impotence in the face of the size and scope of the modern administrative state.
Regulatory Oversight and Deregulation
Republicans have long pursued a deregulation strategy as another antidote to the proliferation of the administrative state, although with no more success than any other strategy discussed here. Deregulation and regulatory oversight strategies are executive efforts to exert more control over agency rulemaking, but these strategies fail because the executive lacks fundamental control over the administrative state.
The Reagan Administration’s regulatory oversight required agencies to prepare cost-benefit analyses for major rules and required that agencies only issue regulations that maximize net benefits (defined as social benefits minus social costs). Similar to the REINS Act, this approach focuses not on the substance of federal regulations but only on their potential costs (and estimating costs depends on accurate forecasting—a dubious proposition). The error of this approach is on display in immigration policy. Federal regulations that grant visas to hundreds of thousands of immigrants might be economically “scored” as beneficial to the country’s gross domestic product, but that cost analysis, even if accurate, speaks to only one aspect of immigration policy and neglects the transformational effect of immigration on culture, the allocation of labor, the displacement of American workers, and domestic wages. The Reagan Administration’s regulatory policy focused myopically on the economic impact of regulation, as if regulations could only pose harm by undertaking economic decisions without the people’s participation through their elected representatives, not social, cultural, or political decisions, despite their obviously transformative nature.
Besides, the Reagan Administration’s regulatory oversight program can be judged by its fruits. By the final two years of that administration, the pace of new regulations had increased, and that increase continued into the Bush Administration. The power of the administrative state to dictate the lives of Americans, divorced from political oversight, did not shrink; it grew.
For its part, the Trump Administration attempted a new regulatory strategy targeted more precisely at deregulation. The Trump Administration pledged to remove two regulations for every one enacted, and even made the promise official by promulgating it in an executive order. The policy sounded good but faced legal and procedural hurdles. For one, deregulation requires federal agencies to go through the same notice-and-comment process that applies when affirmatively regulating, so the policy could, at most, require agencies to initiate the withdrawal of two regulations for every one proposed. From that point forward, the deregulatory and regulatory efforts had to follow different trajectories, leaving no guarantee that two regulations would actually be withdrawn for every one imposed. Nor was there any guarantee that the regulations targeted for withdrawal would be equal in significance to any new regulation being proposed.
Ultimately, the Trump Administration’s deregulatory initiatives resulted in the enactment of fewer new regulations compared to its predecessor administrations, and the Trump Administration did try to remove many regulations as well, but many of these efforts foundered on legal grounds.
Most of the Trump Administration’s important deregulatory actions, like barring asylum eligibility for certain individuals entering the United States at the southern border or rolling back the Obama Administration’s Clean Power Plan, were litigated immediately and enjoined. Overall, the Trump Administration’s track record in litigation was dismal. By one assessment carried out by the Institute for Policy Integrity, the Trump Administration succeeded in defending its regulatory actions in court 58 times but was unsuccessful 200 times. That means a mere 22 percent of the Trump Administration’s regulatory actions survived judicial review.
The Trump Administration’s deregulatory efforts come the closest of any conservative strategy to resisting the Landis premise itself: at least under President Trump, the executive branch attempted not merely to layer procedural requirements onto the regulatory process or create greater oversight for economically significant laws, but to actually reduce regulation directly. But the Landis premise is so deeply embedded in the modern regulatory state that executive action alone cannot unseat it.
Deregulation requires the same procedures as regulation, and it is subject to judicial review, which places it ultimately beyond the executive’s sole control. The administrative state results in tyranny because it operates without political oversight. Presidential oversight is an illusion. The president sits atop the bureaucracy but can have precious little effect on its conduct. The president cannot order agencies to act without following the burdensome and time-consuming notice-and-comment procedures; nor can the president rescind past agency action without undertaking the same burdens—to say nothing of the general unresponsiveness of the bureaucracy to pursuing any policy with haste or diligence.
A Proper Diagnosis
Conservatives have failed to restrain the administrative state because they have accepted the Landis premise—that the administrative state is a necessary governmental innovation required by the complexity of modern society. This intellectual capitulation is what ensures that the balance of power in this country will remain not only in Washington, D.C., but specifically with the largely unaccountable administrative state. The federal bureaucracy is the home of the most prestigious jobs in public service, the best salaries and benefits, the greatest esteem, and the most power. Educated and well-qualified individuals who aspire to power and influence want to join the administrative apparatus. These are the experts, after all, and we have entrusted to them the power to rule us.
Never before has the fallacy of expert governance been so exposed as it is today, following the emergence of COVID-19 in the United States. The problem of COVID-19 placed federal public health officials on the national stage, demanding that their expertise direct and save the nation. And they failed. They opposed masking before demanding it universally; they advocated destructive lockdowns that uncannily reflected liberal biases (like shuttering churches on account of public singing while permitting in-person alcohol sales); they ignored the science of child infection in favor of virtual schooling that has disadvantaged (or worse) a generation of children; and they opposed a vaccine as “rushed” when it was President Trump’s accomplishment, only to mandate the same vaccine at the expense of one’s livelihood once President Trump was no longer in office. These are the experts. Their training prepared them for this moment, and when the nation needed them, they proved themselves to be credentialed political hacks.
That is why any conservative response to the administrative state must begin with the counter-Landis premise: that rule by experts and technocrats is not the self-evident and necessary solution to the problem of modernity, and that in fact, rule by experts and technocrats is just as likely to harm the nation, by impeding individual freedom and restraining economic prosperity. The so-called “expertise” of the administrative state is not expertise at all but simply politics unbridled: it is liberal hegemony divorced from democratic accountability.
The only prescription for the administrative state is deconstruction. Dismantling. Eliminating at least some of the nearly 2 million civilian federal employees (let alone the legions of federal contractors) who comprise the unaccountable and uncontrolled administrative state.
A future Republican president cannot deconstruct even a portion of the federal bureaucracy without significant preplanning that begins well before assuming office. Any Republican presidential candidate must catalog a list of obsolete federal agencies and programs and articulate to the American people the waste and excess required to maintain these frivolous bureaucratic outlets.
At the same time, a future Republican president must be willing to articulate a broader vision for deconstructing significant portions of all federal agencies, including recruiting cabinet officials who are committed to downsizing their agencies.
Realistically, as the experience of the Trump Administration shows, a project to deconstruct the administrative state will depend on the participation of Congress in order to be successful. Taking down even a single regulation requires considerable effort and carries little guarantee of success, as shown by the Trump Administration’s track record in legal challenges to deregulatory efforts. Taking down entire swaths of the federal bureaucracy will face even greater obstacles, including in the form of legal challenges from career federal employees, many of whom are unionized and enjoy special employment protections. Significant policy reforms can proceed only from possession of significant political power. The greatest inroads will be made against the administrative state when the coordinated power of two branches can be brought to bear against it.
A tangible deconstruction along these lines will only be possible if conservatives begin by deconstructing the mindset of the administrative state. Rule by experts is foreign to our constitutional separation of powers; it is incompatible with democratic accountability and legitimacy; and it has proved itself a failure in our own lifetimes. The political branches and the states must be returned to their lawmaking power, and conservatives must relearn to express confidence in that power.
Conservatives must accept that some things simply will not be done by a smaller administrative state, and that is the point. Policies that can be achieved only through tyranny are too costly. To the extent that they deserve to be pursued, they must be housed in branches or levels of government sufficiently responsive to the people and their elected representatives so that tyranny is averted.
How does this translate into actionable policies for a new Republican administration? With difficulty, of course, but some measures come to mind, particularly where a Republican-led executive branch can work cooperatively with a Republican-led Congress.
First, draft and pass legislation to require a universal sunset for all agency regulations. As it stands, agencies enact regulations frequently but rarely take any down (and, as the experience of the Trump Administration shows, taking down regulations is fraught with legal challenges and is not guaranteed to succeed). Yet many good reasons exist for revisiting regulations at some point after their enactment. When regulations are enacted, predictions about their costs, benefits, and effectiveness are speculative at best. Fifteen years on, more can be said about whether a particular regulation has been justified. Mandatory sunsets also require Congress to act if a regulation is to be retained, which restores at least some measure of democratic accountability to a bureaucracy that has been allowed to otherwise run amok.
Second, repeal and reverse large portions of the Pendleton Civil Service Reform Act of 1883 and the Civil Service Reform Act of 1978, with the imposition of term limits for bureaucrats. These acts standardized federal government hiring and required that bureaucrats be primarily hired as nonpolitical positions of expertise. This has had the effect of stultifying the bureaucracy, turning hiring into a quota system and exacerbating the problem of unaccountable bureaucrats remaining in their posts for a lifetime. These reforms could have the advantage of surprise, an advantage already squandered for the Schedule F reforms, which the Trump Administration pursued by executive order and the Biden Administration immediately rescinded. Much attention has been paid to Schedule F reforms, allowing the Left to mount a public relations counterattack. But finding new ways to control the bureaucracy could allow for the element of surprise once again.
Third, Republicans should ban or restrict public-private partnerships in governance. The idea is a radical one because, at present, both the Left and the Right support these kinds of arrangements. Because government is perpetually behind the private sector in terms of technology, sophistication, innovation, and general capabilities—so the thinking goes—partnering with the private sector to provide government services allows the government to compensate for its inadequacies. But this compensation means that government remains able to grow its mandate despite its ineptitude, fanning into an ever-more-expansive oversight of Americans’ lives, and it does so at the cost of sharing data with private sector businesses that desperately seek to own and profit from it.
Consider the Obamacare exchanges, for example, which are run by private entities and host the personal health, financial, employment, and other data of millions of Americans—data that private entities are happy to contract with the federal government to control. These kinds of partnerships present increasing threats to the American people (including the threat of a growing and unaccountable federal bureaucracy) even as they decrease in visibility (think “government” websites owned and operated by private entities, with consumers none the wiser). Congress can and should exercise oversight over whether and how the federal government outsources its work to the private sector because private sector innovation and nimbleness allow the administrative state to do things that are beyond its capabilities. Obviously, some nuance is required, because the Department of Defense cannot help but contract with private entities to build military aircraft, and no one would suggest otherwise. Yet the proliferation of public-private partnerships for the purpose of growing government and ceding Americans’ data to the private sector is a real problem and one that deserves the attention of any future Republican administration.
These reforms require Congressional cooperation and significant preparation in advance of a Republican presidential administration. But if accomplished, they promise durable change to the administrative state. To be clear: their success depends on the wholesale rejection of the Landis premise and a complete commitment to the urgent necessity of dismantling the administrative state. Upending the belief that only rule by experts can accomplish the aims of modern governance must be the goal of any future Republican administration.