The Yale Law Journal

October 2022

The Administrative Agon: A Democratic Theory for a Conflictual Regulatory State

Administrative Law

abstract. A perennial challenge for the administrative state is to answer the “democracy question”: how can the bureaucracy be squared with the idea of self-government of, by, and for a sovereign people with few direct means of holding agencies accountable? Scholars have long argued that this challenge can be met by bringing sophisticated thinking about democracy to bear on the operation of the administrative state. These scholars have invoked various theories of democracy—in particular, pluralist, civic republican, deliberative, and minimalist theories—to explain how allowing agencies to make policy decisions is consistent with core ideas about what democracy is.

There is a weakness to these theories—a weakness exposed by the deep political polarization surrounding American administrative law and the institutional fragmentation that characterizes much of the administrative state. Each of the conventional democratic theories in one way or another assumes that the goal of democracy is to reduce or settle political conflict, and that it is coherent to speak of accountability to a single mass of people we call the dêmos only after conflict has been settled. Relying on this shaky and unrealistic assumption to build an account of the administrative state’s democratic legitimacy has always been problematic, but the weakness of this standard approach is particularly glaring in the light of our polarized, conflictual politics, which makes it difficult to imagine that the assumption would be realized in administrative practice.

This Article charts a different way of looking at the democracy problem and provides a roadmap for reinforcing and building legitimacy in administrative processes. It draws on democratic agonism, an overlooked theory of democracy that assumes that political conflict is ineliminable and recognizes that every decision made in a democracy must by its very nature exclude some people or perspectives from full inclusion in the governing dêmos. With this recognition, agonism turns the conventional approach on its head. Instead of prescribing democratic processes to reduce conflict and undergird a settlement that maps onto the preferences of the people, agonism seeks to build processes that unsettle decisions and promote friendly contestation over government policies, drawing the excluded back into a conflictual process of defining the dêmos anew. Agonism’s emphasis on conflict maintenance better fosters democratic legitimacy in a deeply divided, pluralistic society like ours, where it is impossible to please every constituency with government decisions. Moreover, its resistance to settlement provides built-in safeguards against growing authoritarianism and plebiscitary presidentialism that are falsely held out as possibilities for finally settling political conflict.

Turning from theory to practice, I argue that we can imagine an “administrative agon” that incorporates agonistic elements into the institutions and practice of administrative law and public administration. The theory has prescriptions for a range of issues—from public participation to judicial review of agency action to the design and independence of agencies. In some of these areas, the agonistic democratic lens reveals ways that the administrative state might be working better than we think, at least according to agonistic metrics. In other areas, it highlights deficiencies. By bringing agonistic democratic theory into conversation with the administrative state, I aim to challenge the growing malaise about how the administrative state can fit into our conflictual politics and to point the way to reforms that could make the administrative state more genuinely democratic in practice.

author. Associate Professor of Law, Texas A&M University School of Law. I am grateful for opportunities to workshop this Article at the University of Florida Levin College of Law, the Texas A&M University School of Law, the George Washington University Law School, and the Association of American Law Schools’ New Voices in Administrative Law workshop. I especially thank Cary Coglianese, Blake Emerson, Andrew Hammond, J. Benton Heath, Luke Herrine, Michael E. Herz, Ben Johnson, Neysun A. Mahboubi, Jane J. Mansbridge, Jud Mathews, Jon D. Michaels, David L. Noll, Todd Phillips, Shalini Bhargava Ray, Mark Seidenfeld, Jocelyn Simonson, Glen Staszewski, Karen M. Tani, and Wendy E. Wagner for their thoughts and suggestions, all of which greatly benefited the paper. I also thank Daniel A. Mejia-Cruz and the rest of the editorial staff at the Yale Law Journal for their extremely helpful suggestions and diligent work in preparing this Article for publication.


James Oliver Freedman wrote almost a half century ago that the administrative state has faced a recurring sense of democratic crisis over its lifetime.1 That sentiment rings truer than ever today, as the administrative state finds itself under immense, perhaps even existential, political stress.2 This condition is evident in several parallel developments, any one of which would have been the nation’s leading political drama in a prior era.

There is, to start, the unprecedented attack on the administrative state from within the executive branch during the Trump Administration. Whether we call it “administrative sabotage,”3 “maladministration,”4 or “structural deregulation,”5 the bottom line is that President Trump opposed the administrative state, evidenced most clearly in his occasional allusions to the idea of a “deep state” out to thwart the will of the people.6 Although the Trump Administration failed to undermine the administrative state fundamentally,7 the wider populist antipathy toward government institutions that motivated the administration’s actions is alive and well, both domestically and internationally.8

A related but distinct line of attack persists in the federal courts. There, some judges have impugned the administrative state as antithetical to our constitutional democracy, the rule of law, and the liberties of individuals and businesses alike.9 A few have even discussed tearing out congressional delegation, the foundation for the administrative state, root and branch.10 At the same time, progressive activists critique the administrative state for its contributions to structural inequalities and its failures to use its authority to root out injustices.11 Progressive populists worry that “captured” agencies12 with cultures and personnel at odds with the current administration, like Immigration and Customs Enforcement, will resist the initiatives of the Biden Administration or the Congressional Progressive Caucus.13

These growing anxieties about the administrative state span the political spectrum and are in some sense epiphenomenal of deeper societal fracturing. The United States is fundamentally divided on key questions of national political valence,14 and various constituencies grow increasingly frustrated over the imperviousness of established institutions to fundamental change.15 These political tensions have thus drawn attention to problems concerning the democratic legitimacy of the administrative state—long-existing problems that our current conditions spotlight.

“Democracy,” from the Greek demokratía, concerns the legitimation of government by lodging control of the power (krátos) of the government with the people (dêmos).16 As the most practically important institution for making and implementing government policy,17 the administrative state is where we must look to know whether the dêmos truly controls government decision making. On its face, the administrative state seems to present democratic difficulties: it lacks any direct link to electoral inputs,18 and it possesses a stability and autonomy designed to make it resistant to democratic control.19 More fundamentally, it is not clear how, in a pluralistic, deeply fractured society like ours, the decisions that administrative agencies make could represent all or even most of “the people” most of the time. The perennial tension between administration and democracy—what I call the “democracy question”—increasingly feels unresolved and, perhaps, unresolvable. As a result, there is a real danger that frustration with the administrative state from all corners will only continue to grow until it experiences significant democratic delegitimation and institutional atrophy.20

Scholars have attempted to deal with the democracy question by drawing on a canon of traditional democratic theories.21 Their accounts have a common core: the idea that certain features of the administrative process help resolve or settle political conflicts that would otherwise make it difficult to say that administrative decisions represent the whole of “the people.” They argue, in other words, that the administrative state can serve as a site for “‘political will-formation’ in the public sphere,”22 although they differ on precisely how that process does and should take place. For instance, civic republicans and deliberative democratic theorists assert that certain features of the administrative process—including notice-and-comment rulemaking and agencies’ duty to provide reasons for their decisions—foster deliberation that, ideally, results in agreement (or, at least, in less disagreement).23 Other theories posit that administrative processes collectively function as a marketplace for influence by allowing all interested parties to participate.24 On this account, agencies set policy that is at least somewhat responsive to the expressed preferences of interested parties. Still other theories seek to ground administrative legitimacy in a more direct connection to electoral inputs on the theory that elections are a kind of democratic settlement (for a time, at least) and that bureaucrats are ultimately subject to presidential control.25

This Article argues that these standard approaches cannot democratically legitimize the administrative state. It is unrealistic to assume that disagreement over policy could be substantially ameliorated through administrative processes or through accountability to elected officials, such that it would be coherent to speak of a “general will” embodied in administrative action. Consider, for example, the experience of the Biden Administration’s vaccine-or-test mandates—a recent example of an attempted administrative settlement that proved to be anything but. Invoking long-standing statutory authority to issue emergency temporary standards, the Department of Labor (DOL), via the Occupational Safety and Health Administration (OSHA), issued a regulation affecting any employer with over one hundred employees.26 To decrease the risk that workers would expose each other to COVID-19, affected employers had to require their employees either to be vaccinated against the virus or to undergo weekly testing.27 There was immediate opposition to the rule.28 Just days after the policy was finalized, challengers in the U.S. Court of Appeals for the Fifth Circuit obtained a nationwide stay of enforcement of the policy.29 Although the stay was vacated by the Sixth Circuit,30 the Supreme Court intervened on short notice in a landmark shadow-docket opinion to hold that OSHA and DOL had likely exceeded their statutory authority.31 In the meantime, much of the population has stubbornly resisted the efforts of public-health experts to encourage vaccination,32 drawing the ire of “vaccinated America.”33 None of the traditional democratic justifications of administrative action could alter this deep well of resistance. President Biden’s election, to the extent that it was even accepted as legitimate,34 was not enough. Neither was the societal deliberation on vaccines or the voluminous record and thoroughly reasoned final rule issued by OSHA.35 Nor, apparently, was it persuasive to antivaxxers (or the Supreme Court) that a clear majority of people supported the rule.36 The idea that notice-and-comment processes could have led OSHA to a version of the policy that would have increased public acceptance of a mandate is similarly pollyannish. If democratic legitimacy is supposed to lead to substantial acceptance of government policy by an identifiable and singular “public,” then the administrative state clearly lacks democratic legitimacy.

These kinds of practical experiences of failure to ameliorate deep social conflict are underscored by the theoretical concern that the basic assumption of traditional democratic theories is a pipe dream in the context of administrative law. Social choice theory has long shown that under exceedingly minimal assumptions, a rational and stable form of preference aggregation is impossible, even in elections.37 Even when agencies try sincerely to aggregate the preferences of citizens, their efforts are guaranteed to devolve into arbitrariness. Deliberative theory seeks to avoid this problem by changing preferences rather than simply aggregating them.38 Yet, this project has its own problems born of ineradicable pluralism. As many commentators note, even though deliberative theories strive to be truly inclusive of all perspectives,39 in practice, the administrative state routinely makes decisions that are flatly rejected as illegitimate by one mainstream political or religious camp or another.40

Moreover, the standard answers to the democracy question—for example, rendering agencies subject to strict, hierarchical, principal-agent control by democratically accountable actors—are at odds with leading descriptive and empirical work about administrative institutions and processes.41 The picture that emerges from this work is of a bureaucracy engineered for conflict: it is often internally fragmented, interminably complex, and irreducibly diverse.42 Scholars generally praise these features of the bureaucracy, highlighting how they protect liberty or reinforce core constitutional values that underlie the separation of powers or lead to effective governance.43 Whatever the merits of these accounts, it is unclear how they are democratic defenses of the administrative state, and they are often orthogonal to the consensus-oriented project of democratic legitimation envisioned by conventional democratic theories.44 They paint a descriptive picture of an internally contest-prone administrative state that mires policy initiatives in layers of institutional combat—a picture that sits uncomfortably with the emphasis that stock democratic theories place on social consensus and public will.

These weaknesses of the traditional approaches to the democracy question might not doom the administrative state, but they do demand that we rethink democratic legitimacy and administration. We should no longer insist that some feature of the extant administrative state renders the decisions it makes congruent with the preferences or values of an identifiable dêmos that will accept and support those decisions. Rather, a convincing democratic theory must be consistent with the deep and enduring pluralism that marks American politics.

In this Article, I argue that agonistic democratic theory—agonism, for short—provides better democratic grounding for the administrative state than the conventional theories do.45 Agonism does not seek to elide political conflict through achievement or declaration of a consensus. Instead, it emphasizes the inevitability of conflict and builds democratic legitimacy around it.46 In essence, agonism turns the traditional democratic theories on their head: rather than asking what we should or must do to generate law that reflects widespread societal acceptance and then engineering administrative institutions to facilitate that buy-in, agonism finds legitimacy in “unsettlement” of the law.47 In the struggle against any temporary settlement on a particular law or policy, agonists find a different kind of democratic legitimacy—namely, the opportunity for winners and losers alike to practice democracy by defending and critiquing the status quo.48 In the commitment to this contestation, an actual dêmos is forged around a commitment to live together despite (or even because of) our irreconcilable conflicts. In short, agonists celebrate political conflict and seek to foster and sustain it, even when it does not emerge naturally.

Envisioning the administrative agon—that is, the administrative process as a meeting of administration and agonistic democracy—challenges us to rethink basic design features of the administrative state.49 The administrative agon would focus much more than existing administrative processes on forcing agencies to continually justify “settled” decisions and promote robust adversarial contestation.50 In exchange for this “unsettling” of decisions, which make it possible for multiple constituencies to “prevail,” agencies would be permitted to make decisions with far less procedural constraint than they currently can, which would in turn allow agencies to quickly pivot in response to the ebb and flow of political contest. To ensure that this contestation truly represents the diverse views of the dêmos, the administrative agon would also take far more seriously the perspectives of marginalized groups and individuals, giving them a voice in the decision making process even when they lack practical access to the levers of power.51 And, in contrast with theories that seek to clarify lines of authority tying administrative action to oversight by electorally accountable officials, the administrative agon would structure democratic legitimacy around a flattened bureaucratic hierarchy that fosters intra- and interagency competition.52 These design features would depart from conventional intuitions in many ways, but they also highlight the fact that certain existing administrative practices and institutions that have troubled theorists in the past are actually consonant with established democratic theory.

My account of the administrative agon proceeds as follows. Part I unpacks the democratic theories that have conventionally been invoked to justify the administrative state. In Part II, I highlight the many ways in which these conventional theories fail to provide a satisfying answer to the democracy question. Part III presents agonistic democratic theory as an alternative to conventional democratic theories. Then, Part IV develops a more practical account of the administrative agon, emphasizing features of the contemporary administrative process that are already agonistic while pointing the way to more thoroughly agonistic processes that have not yet been adopted or considered. Part V concludes by discussing payoffs for thinking about administration agonistically, as well as some of the drawbacks and limitations of such a conceptual shift.