Vol. 99, No. 3

To Defer or Not to Defer: Squaring the Administrative Procedure Act’s “Substantial Evidence” Standard of Review with the Demise of Agency Deference

Lindsey Kate Miller*
Agency Deference, Constitutional Law, Standard of Review
Notes

While the 2024 Supreme Court decision overruling the Chevron doctrine, Loper Bright Enterprises v. Raimondo, held that Article III courts cannot defer to agencies’ findings of law, it was silent on whether judicial deference is owed to agencies’ findings of fact. In reaching this decision, the Court interpreted language from the Administrative Procedure Act (“APA”), a comprehensive legislative scheme that governs administrative agencies’ rulemaking and adjudicative processes and prescribes the applicable standards of review for such actions, to mean that courts must review agencies’ findings of law de novo. Given this holding, one could infer that the APA, which says that courts should apply the deferential “substantial evidence” standard of review to formal agency adjudications, should be interpreted to mean that Article III courts ought to defer to agencies’ findings of fact. However, scholars have, on occasion, argued that this standard of review is unconstitutional. For these scholars, it violates the Constitution’s separation of powers principle, especially in cases involving private rights, in which Article III courts must adjudicate matters from start to end. However, Loper Bright and another 2024 Court case SEC v. Jarkesy both implicitly acknowledged and responded to some of these scholars’ concerns.

This Note builds on these developments. Specifically, it identifies one implication of the Loper Bright holding: Congress has discretion to set standards of review for agency decisions, meaning the substantial evidence standard is lawful in some cases. Additionally, it explores Jarkesy’s revival of the public/private rights distinction, in which public rights cases may be freely adjudicated by agency tribunals. This Note concludes that this distinction is crucial. Under Jarkesy, the APA’s substantial evidence standard attaches to judicial review in public rights cases and implicitly excludes cases that implicate the Seventh Amendment and private rights, since agency tribunals cannot decide Seventh Amendment, private rights cases in the first place. By clarifying the constitutional limits of the APA’s substantial evidence standard, this Note provides courts and litigators with guidance on navigating administrative law challenges in a post–Loper Bright landscape.

INTRODUCTION

It would take roughly 6,500 Supreme Courts to handle the caseload of one of the busiest federal administrative agencies¹ alone.² For many legal commentators, agencies’ significant role in defining and protecting individuals’ rights by adjudicating hundreds of thousands of cases annually is reason to grant extensive powers to these agencies.³ For others, agencies are constitutionally defective, and dangerously so, given how much influence agencies have over our daily lives.⁴ The Supreme Court has recently weighed in on this debate, giving indications that it may fall into the latter camp.⁵ Among its recent criticisms of administrative law, the Supreme Court has derided judicial deference to agencies’ interpretations of law.⁶ What the Supreme Court has yet to conclusively decide, among other issues, is whether this concept of judicial independence extends to agencies’ findings of fact.⁷

The Administrative Procedure Act (“APA”) provides a comprehensive scheme for conducting and reviewing adjudications by federal agencies, including setting the standards of judicial review for different types of findings made by agencies.⁸ When Article III courts⁹ review agencies’ decisions made pursuant to “formal” hearings prescribed by the APA, they generally examine agencies’ findings of fact under a deferential “substantial evidence” standard.¹⁰ One who reads the 2024 Supreme Court case overturning the Chevron deference doctrine, Loper Bright Enterprises v. Raimondo, may conclude that the APA’s substantial evidence standard of review passes constitutional muster: there, Chief Justice Roberts noted that section 706(2)(E) of the APA (which codified the substantial evidence standard of review) did mandate deferential judicial review of agencies’ findings of fact.¹¹ Yet, this decision did not explicitly answer whether the APA’s substantial evidence standard itself is constitutional. Rather, the Court merely asserted that Congress intended Article III courts, in applying section 706(2)(E) of the APA to reviews of formal agency adjudications, to defer to agencies’ findings of fact.¹²

In contrast, in interpreting another clause from section 706 of the APA, which says that a “reviewing court shall decide all relevant questions of law” and “interpret constitutional and statutory provisions,”¹³ the Loper Bright Court read in a “de novo” standard of review.¹⁴ This holding stands in stark contrast to decades of judicial deference to agencies’ interpretations of law.¹⁵ Furthermore, this holding brings into relief the fact that the Supreme Court can, and will, interpret the APA to meet constitutional demands. Finally, this holding begs the question: Is the substantial evidence standard of review as applied to formal agency adjudications constitutional?

Some scholars believe it is not. According to these critics, judicial deference to agencies’ findings of fact offends the Constitution’s separation of powers principle, including the judicial branch’s historical primacy, if not exclusivity, over resolving private disputes.¹⁶ However, much of this criticism came before Loper Bright and another Supreme Court case Jarkesy v. SEC¹⁷ were decided, which, taken together, shed light on these issues and contextualize these critics’ arguments.

In Jarkesy, the Supreme Court recognized that the Seventh Amendment right to a jury trial attaches when the cause of action litigated before an administrative tribunal sounds in “common law,” or is analogous to a common law claim, and the case does not involve resolution of a “public right[].”¹⁸ By implication, this allows the executive and legislative branches to freely adjudicate public rights. Consequently, in having the power to regulate public rights, Congress may also limit its power. This is supported by the Court’s finding in Loper Bright that when Congress exercises its discretionary powers, its findings in court are conclusive.¹⁹ One way Congress may limit its power is through setting the applicable standard of review for its findings of fact. Here, Congress creates agencies, meaning it can delegate to agency tribunals the power to make findings of fact and tell courts to scrutinize these findings according to the applicable standard of review.

This Note’s conclusion draws support from recent Supreme Court decisions that have, broadly speaking, cut the power of administrative agencies to make legal determinations while also yielding to agency tribunals’ findings of fact. Part I first explores background constitutional principles, including the separation of powers doctrine and scholarly criticisms of administrative law. In reviewing this material, this Part explains one of the rationales behind the APA—to bring administrative law in conformity with these constitutional principles. It also explores whether the APA, according to some scholars, fails to fully vindicate these constitutional concerns because it codifies the appellate review model,²⁰ or the idea that agency tribunals engage in initial fact-finding and law-finding, the former of which is treated deferentially by reviewing courts.²¹ After defining the substantial evidence standard of review in Part I, this Note then proceeds, in Part II, to compare Supreme Court jurisprudence on agencies’ quasi-legislative²² and quasi-judicial power,²³ concluding that, while the Roberts Court has curtailed agency power in many respects, it has spoken little on the issue of whether agency tribunals can make findings of fact, which Article III courts then deferentially review. Then, Part III synthesizes the Court’s holdings in Loper Bright and Jarkesy and concludes that, given the Court’s current formulation of permissible agency adjudication, the APA’s substantial evidence standard of review is constitutional in public rights cases.

I. BACKGROUND

A. Separation of Powers, Checks and Balances, and the Administrative State

In recent discussions of administrative powers, the Supreme Court has often invoked the Framers’ belief in a government that is divided into three branches, each possessing purely legislative, executive, or judicial powers with the ability to “check and balance” the other branches’ excesses.²⁴ According to the Court, administrative agencies, which fall under the purview of the chief executive, challenge this framework, especially when agencies exercise “quasi-legislative” and “quasi-judicial power.”²⁵

Under the separation of powers theory, the division of power between three branches of government serves two purposes: (1) it promotes each branch’s integrity through denying each branch the ability to wield the other branches’ powers, and (2) it uplifts individual liberty through curtailing government overreach.²⁶ As to the first point, the separation of powers principle “preserve[s] the integrity of” the judicial branch, for example, because it affords “judicial Power” to this branch and denies such power to the legislative and executive branches, which gives the courts and their acts credibility.²⁷ In other words, the judiciary as an institution would lose all meaning if Congress or the president could simply do what a court does. Hayburn’s Case makes this clear.²⁸ In that case, the Supreme Court took up the issue of whether Article III courts could, pursuant to a congressional act, hold hearings to determine injured veterans’ eligibility for certain government benefits.²⁹ It answered in the negative.³⁰ There, the Court noted that “the business directed by” Congress was “not of a judicial nature,” meaning that courts had no “constitutional authority” to conduct such hearings.³¹ As Mark Chenoweth noted, the holding of Hayburn’s Case “applied equally to all of the branches”: “No branch could exercise a type of power other than that vested in it by the Constitution.”³²

Additionally, the Founders recognized that when “the power of judging [was] not separated from the legislative and executive powers,” there would be “no liberty.”³³ If Congress or the president could offer their input on “Cases” and “Controversies,”³⁴ the risk of arbitrary decision-making grew. Congress and the president were political actors who were susceptible to temporary “passions” that did not necessarily promote individual rights;³⁵ meanwhile, judges had an office of “judgment,” meaning they were supposed to be independent arbiters in search of the law.³⁶ According to the Founders, separation of powers insulated courts from the political branches and their susceptibility to public pressure, such that judges could exercise their independent judgment and promote individual liberty.³⁷

Separation of powers works in tandem with the checks and balances principle, which further enhances individual liberty. In arriving at this conclusion, the Founders looked to the English court system in which judges had no tenure or salary protections, which created arbitrary decision-making marked by executive pressure on judges.³⁸ Thus, the Founders surmised that these constitutional guarantees, in addition to other structural checks and balances such as executive and legislative involvement in the selection of Article III judges, would reduce the risk of each branch wielding power without limit.³⁹ Meanwhile, Philip Hamburger, an administrative law skeptic, argues that the separation of powers and checks and balances system promotes individual liberty because liberty can be “constrain[ed]” only in certain situations: here, an individual’s liberty is constrained inasmuch as Congress can pass legislation and judges can issue judgments determining what they can and cannot lawfully do.⁴⁰ Furthermore, the executive branch binds individuals only in the sense that it implements congressional acts, such as disbursing government benefits, or seeks enforcement of the laws in the courts.⁴¹

The Three Branches of Government

The Constitution incorporates the separation of powers and checks and balances principles by vesting “legislative Powers” in Congress, “executive Power . . . in a President,” and “judicial Power” in “one supreme Court[] and . . . inferior Courts as . . . [prescribed by] Congress,”⁴² and enumerates some of these “Powers.”⁴³ Yet, the Constitution is relatively short and somewhat vague regarding the three branches’ functions, which has resulted in gap-filling by the courts to define the precise contours of legislative, executive, and judicial powers.⁴⁴

For the legislative branch, the “intelligible principle” doctrine has informed much of what Congress can do. Here, Congress’s ability to implement laws depends on telling the president what to execute.⁴⁵ However, given that laws require the vote of a majority of members of Congress to enact, sometimes have technical aspects that Congress is ill-equipped to handle, and often entail unforeseen consequences, Congress may not necessarily write laws with clarity and may leave the details to the executive to hash out.⁴⁶ So, when the executive fills in statutory gaps, they may be engaging in lawmaking power in contravention of the separation of powers principle.⁴⁷ In response to this concern, the Supreme Court in J.W. Hampton, Jr., & Co. v. United States held that not all delegations were unconstitutional; only when Congress failed to provide sufficient instructions to the executive branch, or what the Court called an “intelligible principle,” did the president impermissibly engage in lawmaking.⁴⁸ This was because an intelligible principle authorized the executive branch to do no more than act as a “mere agent of the law-making department to ascertain and declare the event upon which [Congress’s] expressed will” or policy “was to take effect.”⁴⁹ Notwithstanding recent attacks on the intelligible principle doctrine,⁵⁰ this constraint on congressional power has helped determine what legislative power is and the executive branch’s role in it.

As J.W. Hampton suggested, executive power often depends on whether Congress has given the executive branch any laws to execute. In Youngstown Sheet & Tube Co. v. Sawyer, the Supreme Court held that President Truman’s executive order, which directed the secretary of commerce to seize steel mills in order to stave off a potential strike, was unlawful because no congressional act permitted him to do so.⁵¹ Yet, presidents are not completely reliant on Congress in order to exercise executive power. As Justice Jackson noted in his influential concurrence, in some situations, the president may be able to act without congressional authorization, such as in the foreign affairs realm, because the Constitution granted the president sole discretion over this realm.⁵² In the case of a steel mill seizure, however, the Constitution did not give the president sole authority to act.⁵³

Finally, the 1803 Supreme Court case Marbury v. Madison established the core of judicial power, in which Chief Justice Marshall declared that “[i]t is emphatically the province and duty of the judicial department to say what the law is.”⁵⁴ For the Marbury Court, this “duty” meant courts had to resolve the laws in favor of the Constitution.⁵⁵ So, while Congress prevents unlawful exercises of legislative power by the executive branch by supplying it intelligible principles, courts also constrain executive action by serving as the final arbiter of the laws’ meaning.⁵⁶ This means the executive branch is necessarily constrained by courts’ definition of (1) the laws the executive branch is charged with carrying out and (2) its own executive power.

Since Marbury, the Supreme Court has continued to refine its definitions of legislative, executive, and judicial power. Yet, perhaps the most enduring theme in these cases is Marbury’s framing of judicial power as a responsibility—one that may not be abdicated⁵⁷—that courts independently resolve “[C]ases” and “[C]ontroversies.”⁵⁸ While the intelligible principle doctrine may authorize some quasi-legislative actions by the executive branch, courts’ continued insistence on this judicial duty to decide Cases and Controversies, coupled with the Supreme Court’s attacks on agencies’ rulemaking powers in Loper Bright,⁵⁹ sits uncomfortably with the idea of administrative agencies that are empowered to adjudicate individual rights.

Administrative Agencies: A Fourth Branch of Government?

As James Madison warned in 1787, “where the whole power of one department is exercised by the same hands which possess the whole power of another . . . , the fundamental principles of a free constitution[] are subverted.”⁶⁰ Many critics of agency power argue that the rapid creation and expansion of agencies, or what they call the administrative state, during the New Deal in the twentieth century embodies this risk because many modern-day agencies (arguably) perform legislative, executive, and judicial powers in contravention of the Constitution.⁶¹

However, in the same passage, Madison observed that there was “not a single instance” in state constitutions in which the branches were “absolutely separate and distinct”; in fact, he encouraged some overlapping functions between the three branches of government.⁶² Indeed, agencies have been around since our nation’s founding, which may be indicative of the Founders’ understanding of constitutionally permissible distribution of powers. For example, the First Congress authorized executive officers to decide matters such as those relating to taxation and foreign commerce.⁶³ While many administrative state opponents concede this point, they note that such categories fall within the ambit of executive discretion and were never contemplated by the Founders as unconstitutionally encroaching on the legislative or judicial branches.⁶⁴

On the other hand, proponents of agency power maintain that agencies are perfectly consistent with the Constitution because agencies actually promote the two concerns addressed by the separation of powers and checks and balances principles. To this end, such proponents argue that agencies help carry out Congress’s will (given that Congress creates agencies in the first place) while also promoting the president’s interests (through presidential appointment of agency heads).⁶⁵ Furthermore, agencies may enhance individual liberty through making “reasonable . . . judgments about the common good and the general welfare.”⁶⁶ For example, agencies oversee programs intended to prevent or curtail “subordinat[ion] [caused] by market exploitation” and harm caused by “ill health, poverty, pollution, and old age.”⁶⁷

Hamburger rejects the agency power proponents’ arguments and posits that agencies unconstitutionally “exercise . . . binding legislative and judicial powers,” given that the Constitution implicitly rejected government structures from medieval England and eighteenth-century America, which were marked by decision-making bodies that wielded legislative, executive, and judicial powers.⁶⁸ For Hamburger, agencies become unconstitutional when they have “absolute” power, or power that is (1) “extralegal,” meaning it “runs outside the law”; (2) “supralegal,” meaning that agencies rely on “judicial deference” for legitimacy; and (3) “consolidat[ed],” meaning powers that are supposed to be split between three branches are housed in one.⁶⁹ As Hamburger contends, the Founders’ beliefs were informed by English pushback against the Crown’s “prerogative” legislation, called proclamations, and “prerogative courts,” such as the Star Chamber and High Commission.⁷⁰ The Star Chamber, for example, could bind parties and prosecute crimes arising under Crown-issued proclamations.⁷¹ The Star Chamber was authorized to do so because Parliament recognized the Star Chamber’s many advantages, including its ability to dispose of cases swiftly without interference from juries that could be bribed and prosecutors who could be scared off from prosecuting.⁷² Against this backdrop, English courts, unlike the Star Chamber, refused to limit their review of prerogative actions on account of their “office of judgment, in which” judges recognized and adhered to their “duty to follow the law of the land.”⁷³ Eventually, Parliament abolished these prerogative courts in response to public criticism over such courts’ unchecked, extralegal nature and in recognition of the fact that such courts contravened the Magna Carta’s decree that “no one was to be deprived of his liberty or property, other than by the judgment of his peers or the law of the land.”⁷⁴

The Constitution corrected these unchecked dangers, Hamburger argues, through housing judicial power in Article III courts, guaranteeing juries, and providing tenure and salary to judges, among other structural protections.⁷⁵ However, agencies may threaten to undo this constitutional balance by binding people through executive-made legislation and adjudication.⁷⁶ Hamburger’s worry lies partly in the fact that, like the prerogative actors of medieval England, executive officers lack structural protections that promote independent judgment and often opt to work for the president because of personal precommitments favoring the government.⁷⁷

Cass R. Sunstein and Adrian Vermeule contest Hamburger’s version of events. Sunstein and Vermeule argue that administrative state critics like Hamburger focus too narrowly on the risk of executive overreach that they fail to recognize the Founders had other goals in mind when creating a national government.⁷⁸ For Sunstein and Vermeule, the Constitution struck a balance between checking “official abuse” and promoting “public participation and accountability, . . . efficiency in government, . . . rational and coordinated policymaking[,] and . . . the common good and overall welfare.”⁷⁹ The Founders recognized that the executive was well suited to promote these latter goals, which is why they created a powerful executive branch.⁸⁰ While recognizing that some limits to agency power are important, Sunstein and Vermeule oppose eliminating administrative agencies wholesale on constitutional grounds.⁸¹ They argue that too much judicial interference in the realm of administrative law harms individual liberty because it risks the possibility that Article III judges will weigh in on matters

that they are ill-suited to resolve, such as those involving “technocratic expertise and political accountability.”⁸²

Some Supreme Court Justices echo Hamburger’s concerns and question agencies’ constitutionality,⁸³ and the Court as a whole has a “fraught” relationship regarding its view of executive power.⁸⁴ However, as it currently stands, agencies may permissibly wield some power.⁸⁵ To what extent, including whether agencies may conclusively determine facts in some adjudications, remains to be seen.

B. The Administrative Procedure Act and the Standards of Judicial Review

The APA embodies the “appellate review model” in which agencies, much like trial-level courts, hold initial adjudicatory proceedings that courts review using different levels of deference depending on whether agencies’ findings of fact or law are at issue: in the former case, courts grant moderate deference to agencies’ findings, whereas in the latter case, no deference is given.⁸⁶ The appellate review model predates the APA and arose to prominence in the early twentieth century;⁸⁷ it also shaped “widespread demand for legislation to settle and regulate . . . administrative law and procedure” in the 1930s and 1940s.⁸⁸

1. The Origins of the APA

As Evan D. Bernick, a critic of judicial deference to agency fact-finding, notes, from the nation’s founding until the early twentieth century, “judicial review of governmental action was both narrow and deep”: “It was narrow in that complaints about governmental activity often did not entitle individuals to judicial review”; and “[i]t was deep in that, if individuals did have a complaint that triggered judicial review, that complaint was adjudicated in an especially rigorous way.”⁸⁹ Individuals were precluded from judicial interference unless they could trace government action “upon core private rights.”⁹⁰ This was because the Magna Carta, which in turn inspired the Constitution, held that if the government sought to deprive an individual of certain rights—that is, the private rights of “life, liberty, or property”—it could only do so in accordance with due process considerations and “the law[s] of the land.”⁹¹ In other words, potential deprivation of one’s private rights triggered court intervention that followed certain procedural requirements,⁹² including an obligation to “interpret[] . . . the relevant law,” which was performed by judges, and to “determin[e] . . . the relevant facts,” which was a jury task.⁹³ Congress could not run around these requirements by creating “entities called ‘courts’ ” that could not otherwise exercise judicial power or did not follow these procedural requirements.⁹⁴

Thus, when the executive branch sought to affect an individual’s private rights, it had to do so through an Article III court; it could not bind the individual pursuant to agency proceedings.⁹⁵ Once in an Article III court, any findings of fact made by the executive beforehand were treated as “merely the records of one of the parties in the case” rather than “records at law” with “presumptive verity or validity.”⁹⁶ An 1843 Supreme Court case United States v. Irving⁹⁷ is one such time in which the government had to litigate individuals’ private rights through an Article III court. In that case, the government sought to recover a federal customs collector’s “debt[s] from” the collector’s “sureties” after he had fled the country.⁹⁸ Here, the Court held that the United States Treasury’s various factual findings, including a transcript it created describing the customs collector’s account, were to be treated as “prima facie evidence,” that is, such determinations could be rebutted by other evidence and weighed by juries.⁹⁹ In other words, because the Treasury’s transcript had no “presumptive effect [like] a court record”—notwithstanding it being “an executive record . . . kept by officers who,” like the court, “were expected to exercise judgment rather than will”—the jury was free to determine the transcript’s validity and persuasiveness.¹⁰⁰

Meanwhile, the executive branch could freely adjudicate matters concerning “public rights and quasi-private privileges.”¹⁰¹ This was because “disposition of those legal interests did not require the use of ‘judicial’ power.”¹⁰² The Supreme Court “has not ‘definitively explained’ the distinction between public and private rights,”¹⁰³ but precedent has generally included among the category of public rights “claims that were understood to be held by the people as a whole,” such as

(1) proprietary rights held by government on behalf of the people, such as the title to public lands or the ownership of funds in the public treasury; (2) servitudes that every member of the body politic could use but that the law treated as being collectively held, such as rights to sail on public waters or to use public roads; and (3) less tangible rights to compliance with the laws established by public authority “for the government and tranquility of the whole.”¹⁰⁴

Privileges were created by Congress to “carry out public ends[,] . . . ‘originat[ing] with the state rather than the individual.’ ”¹⁰⁵

In 2018, the Supreme Court in Oil States Energy Services, LLC v. Greene’s Energy Group, LLC affirmed the significance of the public/private rights distinction but essentially collapsed rights held by the people as a whole and quasi-private privileges into the umbrella of public rights.¹⁰⁶ In that case, the Court considered a challenge to the Leahy-Smith America Invents Act, which created an “inter partes” review process.¹⁰⁷ Through the inter partes review process, an individual could petition the United States Patent and Trademark Office to cancel a patent’s claims without needing to resort to a jury trial.¹⁰⁸ “[P]ublic rights,” the Court confirmed, “cover[ed] matters ‘which ar[ose] between the Government and persons subject to its authority in connection with the performance of the constitutional functions of the executive or legislative departments.’ ”¹⁰⁹ There, the Court held that the granting of—and therefore revocation of—a patent constituted a public right because patents (1) were “ ‘public franchises’ that the Government,” and therefore the public, “grant[ed] ‘to the inventors of new and useful improvements’ ”;¹¹⁰ and (2) could be granted by the executive or legislative branches “without ‘judicial determination.’ ”¹¹¹ Therefore, like all other public rights, the granting and revocation of a patent through the inter partes review process could be adjudicated by any branch of government.¹¹²

Historically, when individuals wanted to challenge agencies’ decisions in public rights cases, they had to bring separate lawsuits—rather than appeal the decisions within the agencies—in Article III courts where they would be entitled to “damages or other remedies against wayward executive officers.”¹¹³ However, as Hamburger points out, litigants had to show injury in order to recover.¹¹⁴ Once they demonstrated harm “such as distraint or payment under protest,” then courts would, like in private rights cases, treat the agencies’ findings of fact as prima facie evidence.¹¹⁵ This was because the “judicial proceedings were not extensions of the executive proceedings.”¹¹⁶ Meanwhile, judicial deference, if any were given, applied to the agencies’ findings of law. But Hamburger argues deference is too strong a term since what the courts were doing was “defer[ring] to the law,” which gave the executive discretion to determine public rights, not the agencies themselves.¹¹⁷ The 1840 decision Decatur v. Paulding illustrates this point.¹¹⁸ There, the secretary of the navy, interpreting two statutes, refused to grant a widow two pensions even though the statutes apparently required payment of both.¹¹⁹ As the pensions fell under the public rights doctrine, the executive branch had authority to regulate their disbursement.¹²⁰ Although the Supreme Court recognized it had authority to invalidate an agency’s construction of a statute, the relevant statutes permitted the secretary to interpret them.¹²¹ Thus, the Court merely deferred to the law when it upheld the secretary’s interpretation because the law said the secretary could exercise discretion in disbursing pensions.¹²²

Around the turn of the twentieth century, the appellate review model began to take shape, ushering in an era of judicial deference to agencies’ findings of fact.¹²³ In 1906, Congress passed the Hepburn Act, which modified the process by which the Interstate Commerce Commission (“ICC”) prescribed rate orders.¹²⁴ Before the Hepburn Act, the ICC had to file bills of equity in appellate-level courts for its rate orders to have binding force; afterwards, its rate orders were self-executing thirty days after their issuance.¹²⁵ Another aspect of ICC rate orders that the Hepburn Act changed was how courts reviewed challenges to such orders. Although appellate courts had previously reviewed ICC orders under a de novo standard of review, the Hepburn Act failed to prescribe the applicable standard of review for self-executing ICC rate orders.¹²⁶ In a series of Supreme Court cases following passage of the Hepburn Act, the Court, borrowing from the standard of review that courts applied to jury findings of facts, effectively read in a substantial evidence standard for reviewing self-executing ICC rate orders.¹²⁷ Under this standard, reviewing courts independently analyzed the ICC’s legal findings but deferred to its findings of fact.¹²⁸ This contrasted with decades of precedent in which courts had treated the ICC’s findings of fact as merely prima facie evidence.¹²⁹ For the Court, the substantial evidence standard made sense because the Hepburn Act, in making ICC orders self-executing, “endowed the [ICC] ‘with large administrative functions,’ ”¹³⁰ that is, the ICC had the “institutional competence to determine ‘whether [its] administrative power ha[d] been wisely exercised.’ ”¹³¹

In 1932, the Court assessed the constitutionality of a statute that required reviewing courts to defer to an agency’s findings of fact involving two private parties.¹³² In Crowell v. Benson, an employer challenged the Longshore and Harbor Workers’ Compensation Act (“LHWCA”), which authorized the United States Employees’ Compensation Commission to issue workers’ compensation awards against employers for employee injuries occurring on navigable waters.¹³³ According to the employer, the LHWCA violated the Fifth and Seventh Amendments’ due process and jury trial rights.¹³⁴ The Crowell Court found that the LHWCA’s requirement that reviewing courts defer to an agency’s findings of fact in workers’ compensation matters was “constitutionally unproblematic” because “[t]o hold otherwise would be to defeat the obvious purpose of the legislation to furnish a prompt, continuous, expert and inexpensive method for dealing with a class of questions of fact which are peculiarly suited to examination and determination by an administrative agency.”¹³⁵ Thus, Crowell expanded an agency’s power by permitting it to adjudicate matters involving private rights with at least some binding effect (inasmuch as an agency’s findings of fact were supported by substantial evidence).¹³⁶ Yet, the Court noted that it still had an obligation to independently review (1) an agency’s findings of law as well as to (2) determine “jurisdictional” facts, or facts that had to exist for an agency to not exceed its statutory authority, and “constitutional” facts, or facts that had to exist if a challenged action was to be upheld as constitutional.¹³⁷ This, according to Chief Justice Hughes, was because in private rights cases, “the ‘essential attributes’ of the judicial decision [had to] remain in an [A]rticle III enforcement court,” which courts ensured through reviewing issues of law as well as jurisdictional and constitutional facts de novo.¹³⁸ Yet, the Court’s distinct treatment of jurisdictional and constitutional facts would not last; today, most courts do not review jurisdictional and constitutional facts de novo.¹³⁹

Around the time that the Court decided Crowell, scholars, lawmakers, and interested parties began calling for administrative reform.¹⁴⁰ Although contemporary scholars debate the driving force behind this movement,¹⁴¹ they agree that this movement coincided with a rapid expansion of administrative agencies and an agency-friendly Supreme Court as exemplified by the Crowell decision.¹⁴² For many reformers, Congress could ensure agencies acted lawfully by creating a comprehensive body of law to guide their actions while also correcting potential constitutional violations that they perceived existed in agencies.¹⁴³ One potential violation of prime importance to these reformers was some agencies’ tendency to act as “prosecutor, judge, jury, and executioner” in making the rules and enforcing them.¹⁴⁴ Thus, what would become the APA emerged as a solution to the issue of unchecked administrative power.¹⁴⁵

One thing the APA did to ensure that agencies did not exceed their constitutional limits was to distinguish formal from informal adjudication.¹⁴⁶ It subjected proceedings of the former type to stringent requirements, reflecting the idea that such cases typically needed more than objective tests, measurements, or standards to resolve, and implicated agencies’ executive and quasi-judicial powers.¹⁴⁷

Sections 554, 556, and 557 of the APA governed formal adjudications.¹⁴⁸ Under these sections, agencies conducted “trial-type, adversarial hearing[s]” that were overseen by impartial decisionmakers, typically administrative law judges (“ALJs”).¹⁴⁹ Parties had certain procedural rights in formal adjudications, including the opportunity to appear with counsel, present evidence, and cross-examine adverse parties and witnesses.¹⁵⁰ In enforcing ALJs’ independence, the APA prohibited prosecuting officers from supervising ALJs when they were involved in the same matters, and required that ALJs only communicate with interested parties “openly and upon notice” or on “[e]x parte matters authorized by law.”¹⁵¹ ALJs had other duties under the APA, such as administering oaths, issuing subpoenas, receiving and weighing evidence, and making initial or recommended decisions.¹⁵² An agency was required to conduct formal adjudications when a statute either “explicitly state[d]” that its hearings would be held “on the record” or that the agency would follow the procedures outlined by sections 554, 556, and 557 of the APA.¹⁵³

The APA also provided for judicial review of agencies’ actions.¹⁵⁴ Under the APA, an individual who suffered a legal wrong that was caused by an agency action or as authorized by statute had a right of appeal.¹⁵⁵ Often, these individuals were required to exhaust their administrative remedies before they could obtain judicial review, such as appealing an ALJ’s decision within the agency.¹⁵⁶ Once a case came before an Article III court for review, the court was required to examine the administrative record according to certain standards of review.¹⁵⁷ Per section 706 of the APA, “the reviewing court” was required to “decide all relevant questions of law, interpret constitutional and statutory provisions, . . . determine the meaning or applicability of the terms of an agency action,” and “hold unlawful and set aside agency action, findings, and conclusions” that were

(A) arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law;

(B) contrary to constitutional right, power, privilege, or immunity;

(D) without observance of procedure required by law;

(E) unsupported by substantial evidence in a case subject to sections 556 and 557 of th[e] [APA] or otherwise reviewed on the record of an agency hearing provided by statute; or

(F) unwarranted by the facts to the extent that the facts are subject to trial de novo by the reviewing court.¹⁵⁸

Under the APA, a reviewing court could apply more than one standard of review depending on the issues that came before it.¹⁵⁹ For example, while the “arbitrary and capricious” standard applied to appeals of all agency actions, an Article III court only applied the substantial evidence standard when it reviewed agency decisions made pursuant to formal agency adjudications or otherwise required by statute.¹⁶⁰ This meant the level of judicial deference afforded to agency actions turned on the issues before Article III courts, in which a court applying de novo review engaged in a more searching examination of the record than a court applying the arbitrary and capricious standard of review.¹⁶¹

The Substantial Evidence Standard of Review

The APA did not actually define the substantial evidence standard of review beyond instructing courts to apply it when reviewing formal adjudications.¹⁶² Despite the APA’s minimal guidance to reviewing courts, legislative history offers some, albeit conflicting, clues on the meaning of this standard of review. As the Senate Judiciary Committee exalted in reporting the bill to the full Senate for consideration, “[t]he ‘substantial evidence’ rule . . . [was] exceedingly important” in the context of judicial review of agency decisions and administrative law generally because this standard embodied many of the concerns underlying the APA, such as ensuring a separation of an agency’s investigatory division from its adjudicatory division, streamlining agency procedures, and promoting fairness to all parties.¹⁶³ Per this view, the substantial evidence standard of review had some constitutional significance (especially as it related to separation of powers concerns) and was meant to make clear to courts that it was “the[ir] duty . . . to determine in the final analysis and in the exercise of their independent judgment[] whether[,] on the whole record[,] the evidence in a given instance [was] sufficiently substantial to support a[n agency] finding, conclusion, or other . . . action as a matter of law.”¹⁶⁴

Yet, the committee’s emphasis on independent judgment seemed to contradict courts,¹⁶⁵ scholars, and even some parts of the Senate Report’s view of the substantial evidence standard, who argued it (1) required a considerable degree of judicial deference to agencies and (2) applied in limited circumstances, namely, to judicial review of agencies’ findings of facts.¹⁶⁶ A prominent proponent of this view was then-Attorney General Tom C. Clark, who wrote a letter in 1945 interpreting some of the APA’s provisions. Tellingly, his letter was appended to the official Senate Report, which complicates the idea that the substantial evidence standard accomplished the “important” goals articulated by the APA’s drafters, such as ensuring separation of powers through judges exercising independent judgment.¹⁶⁷

In his letter, the attorney general argued the substantial evidence standard was “intended to embody” common law, defining it as “more than a mere scintilla[,] [that is,] . . . such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.”¹⁶⁸ In other words, the attorney general treated the substantial evidence standard of review for agency decisions as a corollary to the substantial evidence standard of review that appellate courts applied in evaluating jury determinations of fact.¹⁶⁹

In the jury context, the substantial evidence standard was a “highly deferential” one in which “[t]he verdict [would] stand unless [an] appellant c[ould] show that there [was] no substantial evidence to support it, considering the evidence in the light most favorable to the verdict and with all reasonable inferences deducible from the evidence drawn in support of the verdict.”¹⁷⁰ Judicial review of jury findings of fact was deferential given the Constitution expressly guaranteed resolution of civil and criminal disputes by juries.¹⁷¹ For the Founders, juries “protecte[d] American liberty” because they represented common interests, “guarding against arbitrary rule” marked by judges who could be prone to “develop[ing]” biases “in favor of government.”¹⁷² Thus, the substantial evidence standard, in the jury context, preserved the Constitution’s respect for the jury’s unique, unbiased position. Moreover, judicial deference to juries’ findings of fact was warranted given that juries were closer to the evidence.¹⁷³

Yet, if the APA codified the latter view, then the APA requires courts to apply a standard of review that might not necessarily reap the benefits or address the constitutional concerns the substantial evidence standard was designed to create or resolve, respectively. Perhaps a different standard would be warranted to capitalize on agencies’ unique, often technical, expertise or address the separation of powers concerns that agency adjudication may raise. For administrative law critics like Hamburger, the fact that “an administrative agency is not a jury” is another reason why agencies may be constitutionally defective.¹⁷⁴ Whereas juries do not “have . . . precommitments” to the government, ALJs often do.¹⁷⁵ Therefore, as agency skeptics’ theory goes, ALJs can actually threaten, rather than protect or promote, individual liberty. Other scholars counter Hamburger’s fears because, like juries, ALJs are close to the evidence that parties present.¹⁷⁶ Also, ALJs often have specialized knowledge and expertise that make them particularly suited to resolve the matters that come before them.¹⁷⁷ Meanwhile, a jury, which is comprised of laypersons, is less likely to understand the facts that come before agency tribunals.¹⁷⁸ Thus, a jury may misinterpret technical facts, passing misguided (that is, biased) judgment.

No matter the concerns the latter view may raise, the attorney general’s characterization of the substantial evidence standard has won the day.¹⁷⁹ Indeed, for several decades after the APA’s passage, “agencies proceeded as if” the law had not “change[d] much of anything.”¹⁸⁰ Around the 1960s, as agencies began to rely more heavily on an informal rulemaking procedure created by the APA called “notice-and-comment” rulemaking, courts began to develop common law doctrines to constrain this power.¹⁸¹ On the other hand, courts seldom addressed agencies’ adjudicatory powers, including clarifying what the substantial evidence standard required, given the trend toward notice-and-comment rulemaking.¹⁸²

Universal Camera Corp. v. NLRB¹⁸³ and Allentown Mack Sales & Service, Inc. v. NLRB¹⁸⁴ are two such cases that interpreted the APA’s substantial evidence standard of review. In both, the Supreme Court characterized the standard as akin to the standard applicable to jury findings of facts.

In Universal Camera, the Court reviewed a National Labor Relations Board (“NLRB”) order that had found an employer’s dismissal of an employee was not retaliatory even though the dismissal followed the employee testifying against his employer in a collective bargaining matter.¹⁸⁵ The relevant statute, the Wagner Act, merely said that reviewing courts were to treat the NLRB’s findings of fact as conclusive “if supported by evidence.” Although the words substantial evidence were absent from the Wagner Act, the Court, citing precedent, nevertheless determined this language meant that the relevant standard of review for NLRB orders was the substantial evidence standard of review.¹⁸⁶ To this end, the Court characterized substantial evidence as meaning “more than a mere scintilla. It mean[t] such relevant evidence as a reasonable mind m[ay] accept as adequate to support a conclusion.”¹⁸⁷ This characterization of the substantial evidence standard of review tracked with the substantial evidence standard used in jury trials, in which courts also treated it as a “reasonableness test.”¹⁸⁸ To be sure, the Court went on to explain that substantial evidence meant, regarding judicial review of agencies’ findings of facts, as

[doing] more than creat[ing] a suspicion of the existence of the fact[s] to be established. . . . [I]t must be enough to justify, if the trial were to a jury, a refusal to direct a verdict when the conclusion sought to be drawn from it is one of fact for the jury.¹⁸⁹

Yet, because section 706 of the APA required reviewing courts to analyze the whole record, the Court determined that Congress intended to reverse common law only with respect to the fact that “substantiality [had to] be determined in . . . light of” the whole record.¹⁹⁰ This meant reviewing courts had to “take into account whatever in the record fairly detract[ed] from its weight.”¹⁹¹

This rule was probably not much different from the substantial evidence standard in the jury trial context. Under the latter standard, courts had to draw “all reasonable inferences” in favor of the jury.¹⁹² Similarly, as noted above, the Universal Camera Court used the term “fairly detracting from its weight” in defining the substantiality test.¹⁹³ This suggests courts may not adopt the agencies’ findings of fact if doing so means ignoring relevant context. After all, drawing reasonable inferences requires courts to consider the validity of facts in light of other facts.¹⁹⁴ Also, Universal Camera did not require courts to draw inferences against the jury, only that courts must take conflicting facts into account.¹⁹⁵

Allentown Mack, decided forty years after Universal Camera, similarly suggested the APA’s substantial evidence standard operated like the substantial evidence standard applicable to jury findings of fact. In that case, a union successfully argued before the NLRB that a factory’s decision to no longer recognize the union after the factory conducted a poll gauging employee support for the union constituted an “unfair labor practice.”¹⁹⁶ The Court reviewed the NLRB’s order under the substantial evidence standard of review and held that the substantial evidence standard was an “objective test.”¹⁹⁷ As an objective test, a reviewing court employing the substantial evidence standard of review had “no room” to afford “deference to an agency’s eccentric view of what a reasonable factfinder ought to demand.”¹⁹⁸ Despite this limitation, Justice Scalia noted that the substantial evidence test still required considerable deference to an agency’s findings of fact: a court’s role was merely “to ensure” that an agency “achieve[d] minimal compliance with [its] obligation” to “draw all those inferences that the evidence fairly demand[ed] . . . , which [was] the foundation of all honest and legitimate adjudication.”¹⁹⁹ Using the same “fairly” language used in the Universal Camera decision, Allentown Mack confirmed that the substantial evidence standard of review was a reasonableness test.

Yet, Allentown Mack concluded that the NLRB’s findings of fact were not supported by substantial evidence.²⁰⁰ For Justice Scalia, the NLRB improperly ignored some evidence before it, including an employee’s statement to the factory that he was not being adequately represented “for the [union dues that] he was paying.”²⁰¹ Taken together, these facts constituted “reasonable, good-faith grounds” on the factory’s part to doubt union support and therefore supported its decision to survey employee support of the union and revoke its recognition of the union.²⁰² Allentown Mack’s holding that the factory’s decision to revoke its recognition of the union did not constitute an unfair labor practice, contrary to the NLRB’s finding, fits uncomfortably with the Court’s description of the substantial evidence standard of review.²⁰³ This suggests the Court may actually apply the standard less deferentially than it claims it should. This holding may also represent a growing trend toward judicial scrutiny of agency power, or it may simply embody a typical application of the substantial evidence standard of review. Whatever proposition Allentown Mack may stand for, however, is beyond the scope of this Note; what is clear is that the Court continues to recognize a duty to give agencies’ findings of fact some “benefit of the doubt” and that the substantial evidence standard of review largely mirrors the standard for reviewing jury findings of fact.²⁰⁴ In fact, as Justice Breyer argued in a separate opinion in Allentown Mack, the majority’s holding effectively changed the standard of review to one that “look[ed] like an ordinary jury standard that . . . reflect[ed] not an agency’s specialized knowledge of the workplace, but a court’s common understanding of human psychology.”²⁰⁵

Agency skeptics have responded to Universal Camera and Allentown Mack’s endorsements of the deferential substantial evidence standard by raising several constitutional concerns. For one, agency skeptics worry that agency fact-finding may be limited since agencies have discretion over whether to permit discovery.²⁰⁶ Furthermore, even when agencies grant discovery, they argue that discovery may be nominal, inequitable, or less effective as compared with discovery in Article III adjudications: in agency proceedings, although private parties “can ask an agency to issue a subpoena,” they have no guarantee that it will “be answered until the hearing”; meanwhile, an agency “can demand testimony and documents long before any hearing—even before [it] brings charges.”²⁰⁷ This stands in contrast to Article III adjudication, in which civil litigants have “equal access to discovery,” and criminal prosecutors have a duty to share information with the defendants.²⁰⁸ Additionally, unlike federal courts, critics suggest that agencies are not required to adhere to federal rules of evidence or procedure, and can “make up certain rules as they go.”²⁰⁹ For example, the Securities and Exchange Commission (“SEC”) has a “Commission[] Rules of Practice, [in which] hearsay is generally permitted, depositions are limited, discovery is generally not permitted,” and ALJs must make initial decisions within “aggressive timelines.”²¹⁰ This is particularly alarming for these agency skeptics, since the purpose of evidence rules, for example, is to regulate the fact-finding process and improve its reliability.²¹¹ Compounding the issue is that parties might not have control over what agencies include in the factual record or be able to supplement it on appeal.²¹² This could be concerning because Article III courts’ review of agencies’ findings of facts are based on the administrative record.²¹³ Moreover, parties often have no opportunity for de novo review at any time in the appeals process, including during intermediate appeals at the agency level, since “many agencies operate under a presumption that . . . [its] internal appellate authority . . . will adopt the ALJ’s ruling.”²¹⁴ Lastly, this process may be circular because “agencies have a vested interest in protecting their power”; as agency skeptics charge, anyone who challenges agencies’ procedures may be out of luck because “no agency is going to rule that its procedures or structure are unconstitutional.”²¹⁵

Since the rise of the appellate review model, the Supreme Court largely ignored these concerns and had instead actually expanded agency power in several respects. That all changed in 2024 when the Court first overruled a landmark 1984 decision that had held that courts had to defer to agencies’ findings of law in certain circumstances,²¹⁶ and then revived the public/private rights doctrine, finding that claims involving private rights must be resolved by a jury in an Article III court.²¹⁷ The next Part explores these developments.

II. THE SUPREME COURT ON ADMINISTRATIVE DEFERENCE

The Supreme Court’s jurisprudence on administrative law tends to fall into one of two categories depending on whether an agency’s actions can be characterized as quasi-legislative or quasi-judicial in nature. This tracks the APA’s distinction of agency rulemaking from adjudication.²¹⁸ While this Note is primarily interested in quasi-judicial power, given that appeals of formal adjudications under the APA trigger the substantial evidence standard of review,²¹⁹ a line of cases dealing with judicial deference to agencies’ quasi-legislative power suggests constitutional limits to agencies’ fact-finding. Section II.A analyzes three deference doctrines that the Supreme Court has applied to judicial review of agencies’ findings of law and has subsequently refined or dismantled entirely. Section II.B then reviews recent quasi-judicial decisions that pave the way for this Note’s conclusion, that the APA’s substantial evidence standard of review is constitutional.

A. Quasi-Legislative Decisions

Skidmore v. Swift & Co., which was decided while Congress debated the merits of administrative reform like the APA, was the first of three forms of deference²²⁰ to agencies’ findings of law that the Court has recognized.²²¹ It is also the most enduring common law deference doctrine as it continues to be cited and applied by the Court today.²²² The other deference doctrines include Chevron deference, which was overruled in Loper Bright, and Auer deference, which the Court may also discard if one takes Loper Bright’s logic to its natural limit.²²³ At their most basic level, Skidmore, Chevron, and Auer deference doctrines answer the question of whether and when courts should take into account agencies’ interpretations of ambiguous statutes and how much weight such interpretations should hold.²²⁴

Skidmore Deference

Skidmore involved a challenge to the meaning of “overtime” as set out by the Fair Labor Standards Act (“FLSA”).²²⁵ Here, the plaintiffs worked at a packing plant as daytime firefighters but also agreed to work several nights to respond to fire alarms.²²⁶ Although the plaintiffs were paid each time they answered alarms, the alarms seldom went off; this meant the plaintiffs spent hours at the plant without compensation.²²⁷ The plaintiffs alleged this nighttime work qualified as “overtime,” and thus they were entitled to compensation for both waiting for and responding to fire alarms.²²⁸ Although the FLSA did not define overtime in this context, the Court looked to an interpretive bulletin promulgated by the administrator of the Department of Labor’s Wage and Hour Division for guidance.²²⁹ According to the administrator, whether waiting time counted as working time (and therefore counted toward overtime compensation) depended “upon the degree to which [an] employee [was] free to engage in personal activities during periods of idleness when he [was] subject to call and the number of consecutive hours that the employee [was] subject to call without being required to perform active work.”²³⁰ The Court held this interpretation was “entitled to respect” given the administrator’s “policies [were] made in pursuance of official duty,” one in which the administrator had much more extensive knowledge of labor issues than would “likely . . . come to a judge in a particular case.”²³¹ However, the Skidmore Court recognized that judicial deference to an agency’s “interpretations and opinions” had limits: deference could be understood as a sliding scale, depending on a number of factors including (1) the “thoroughness evident in [an agency’s] consideration”; (2) the “validity of [an agency’s] reasoning”; (3) an interpretation or opinion’s “consistency with earlier and later pronouncements”; and (4) “all those factors which give [an interpretation or opinion] power to persuade, if lacking power to control.”²³²

Chevron Deference and the Loper Bright Response

Forty years later, the Court revisited the question of judicial deference to agency interpretations of the law in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc. In what would become known as Chevron deference, the Court dramatically increased agencies’ power to expound on the meaning of ambiguous statutes by mandating that reviewing courts adopt agencies’ reasonable interpretations of such statutes when agencies defined the statutes through regulations.²³³ Whereas Skidmore deference was narrow (applying to nonbinding agency materials that interpreted ambiguous statutes) and flexible (giving courts discretion over how much weight they could afford agency interpretations), Chevron deference was wide (applying to agency regulations and orders) and inflexible (requiring courts to adopt agency interpretations of ambiguous statutes).²³⁴

Chevron dealt with a 1977 amendment to the Clean Air Act (“CAA”).²³⁵ Under the CAA, “ ‘nonattainment’ States,” or states that had not fulfilled certain air quality requirements set by a previous statute, could authorize “ ‘new or modified major stationary sources’ of air pollution” only in certain situations.²³⁶ In enforcing the CAA, the Environmental Protection Agency (“EPA”) eventually promulgated a regulation that interpreted the meaning of a “stationary source” under the statute.²³⁷ In Chevron, the Supreme Court recognized the term stationary source as used in the CAA was ambiguous and thus answered whether the EPA could validly interpret such term through its regulation that would be binding on permit applicants.²³⁸ The Court concluded in the affirmative.²³⁹ Specifically, the Court held that regulations interpreting statutes “control[ed]” if two conditions were met: (1) Congress did not unambiguously express its intent regarding how a statute was to be construed, and (2) an agency’s construction of a statutory term was reasonable.²⁴⁰ In arriving at this conclusion, the Court pointed to a history of judicial deference to agencies’ interpretations of statutes in cases involving technical knowledge and competing policy interests.²⁴¹ Like precedent, the CAA involved complex subject matter that agencies were uniquely qualified to address; the Court suggested that Congress may not have had the requisite expertise to adequately define stationary sources and wanted to delegate this task to a more knowledgeable body, the EPA.²⁴² Moreover, in enacting the CAA, Congress sought to balance economic and environmental interests.²⁴³ Congress may not have had the time, foresight, or votes to sort out minute details of contested legislation and so either intentionally or unintentionally left this gap to the EPA to fill.²⁴⁴ Given these considerations, the EPA reasonably interpreted an ambiguous term because it promulgated the regulation pursuant to extensive agency studies and was consistent with a concern for “economic growth.”²⁴⁵ Thus, the Court upheld the regulation.²⁴⁶

Shortly after Chevron, Justice Breyer and Justice Scalia further clarified the Court’s holding. According to the Justices, Chevron deference was not motivated by separation of powers concerns, or the idea that judicial deference to agencies’ interpretations of statutory ambiguities “guard[ed] against judicial displacement of political judgments.”²⁴⁷ Rather, “Chevron . . . rest[ed] on [one] simple idea: Courts defer[red] to agenc[ies’] interpretations of law when, and because, Congress has told them to do so. . . . [In other words,] the national legislature retain[ed] control of the deference question.”²⁴⁸

Around the turn of the twenty-first century, the Supreme Court revisited the Chevron doctrine and created what would be known as Chevron “[s]tep [z]ero.”²⁴⁹ Step zero of the Chevron doctrine required that, for an agency interpretation to be entitled to judicial deference, such interpretation needed to “carry the force of law.”²⁵⁰ In United States v. Mead Corp., the Court explained that, as Congress authorized agencies to carry out specific but complex procedures, such as notice-and-comment rulemaking, it followed that Congress probably did not intend that agencies could also conclusively bind individuals through less specific or simpler procedures, such as “policy statements, . . . manuals, and enforcement guidelines,” too.²⁵¹ Instead, these kinds of agency actions should be analyzed through the less deferential Skidmore lens.²⁵²

Although Mead appeared to set the stage for overturning Chevron deference, the Court’s 2005 decision in National Cable & Telecommunications Ass’n v. Brand X Internet Services actually reinforced and expanded it, marking the doctrine’s high-water mark. In Brand X, the Supreme Court held that courts must always apply the Chevron analysis to agencies’ interpretation of statutes—even if prior court holdings interpreted the same statutes—unless prior courts found the relevant statutory terms were unambiguous.²⁵³ Justice Scalia dissented, arguing Brand X’s holding was “bizarre,” if not “unconstitutional,” in light of Mead.²⁵⁴ Justice Scalia noted that if an agency, for example, interpreted an ambiguous statute in a manual, but the Supreme Court rejected this interpretation because it lacked the force of law, this holding would not have stare decisis effect.²⁵⁵ This was because the agency could simply issue a subsequent regulation or order (that is, an act that carried the force of law) interpreting the statute the same way, and a subsequent court applying the Chevron analysis would uphold it if the interpretation was reasonable, notwithstanding the fact that the Supreme Court previously held that same interpretation was unlawful.²⁵⁶

Justice Scalia’s dissent was a sign of future Chevron criticisms. Future decisions chipped away at the Chevron doctrine²⁵⁷ until it was finally overruled in Loper Bright.²⁵⁸ In Loper Bright, the Court decided whether regional fishery management councils under the National Marine Fisheries Service (“NMFS”) could require Atlantic herring fishermen to have observers aboard their vessels and to bear these costs.²⁵⁹ Although the Magnuson-Stevens Fishery Conservation and Management Act governed this situation because it enumerated certain situations in which the management counsels could impose observer-related costs on vessels, it was silent as to whether this option extended to Atlantic herring fishing operations.²⁶⁰ Rather than determine whether the NMFS reasonably interpreted an ambiguous statute, the Court threw out the Chevron analysis altogether. “Article III . . . assign[ed] to the Federal Judiciary the responsibility and power to adjudicate ‘Cases’ and ‘Controversies,’ ” Chief Justice Roberts declared.²⁶¹ This duty extended to interpreting the laws.²⁶² The APA “incorporate[d] th[is] traditional understanding of the judicial function” because it codified that courts decided “all relevant questions of law, interpret[ed] constitutional and statutory provisions, and determine[d] the meaning or applicability of the terms of . . . agency action[s].”²⁶³ That statutes were sometimes ambiguous did not mean agencies had the authority to clarify them.²⁶⁴ Indeed, the Framers recognized that, although laws were thoroughly deliberated with “the greatest technical skill,” they would still contain ambiguities.²⁶⁵ To address this, Article III of the Constitution vested courts with the authority—affirmed in Marbury and the APA—to serve as the final arbiter of the laws’ meaning.²⁶⁶ To the Court, Chevron deference subverted the Constitution and the APA’s directives by requiring courts to abdicate their responsibility to interpret the laws.²⁶⁷ In effect, the Constitution and the APA required courts to uncover a statute’s “best” meaning rather than one of many “reasonable” meanings.²⁶⁸

Loper Bright did not prohibit judicial deference to agencies’ interpretations of law wholesale. It suggested Congress could still enact statutes delegating authority to agencies to “exercise [some] discretion.”²⁶⁹ In such a case, “the best reading of a statute” was that an agency had statutory authority to determine what certain statutory terms meant.²⁷⁰ Ultimately, a court still determined what a statute meant: it just decided in that scenario that a statute delegated discretionary power to agencies to carry out the laws.²⁷¹ While reaching a different outcome, the Loper Bright Court ostensibly endorsed the justification offered in favor of Chevron deference by Justice Breyer and Justice Scalia, that is, that Congress created legislation telling courts that they were to defer to agencies’ interpretations of statutes; here, the APA told reviewing courts that they were not to defer to agencies’ findings of law. Thus, Congress had power to create statutes, including modifying the APA, so that it delegated to agencies the responsibility to define statutory terms.²⁷² Finally, the Court clarified that Skidmore deference was still good law since agencies’ interpretations and opinions “constitute[d] a body of experience and informed judgment” that could guide a court’s analysis of ambiguous statutes.²⁷³

Auer Deference

While Loper Bright clarified the limits of Chevron and Skidmore deference, it did not discuss the fate of Auer deference. Creating a third form of judicial deference to agencies’ interpretations of law, Auer v. Robbins held that an agency’s interpretation of an ambiguous regulation was “controlling unless [it was] ‘plainly erroneous or inconsistent with the regulation.’ ”²⁷⁴ Arguably, Auer deference is still good law since the Supreme Court had the opportunity to overrule Auer in the 2019 case Kisor v. Wilkie but declined to do so.²⁷⁵ Rather, Kisor supplied a limiting principle to Auer deference, noting that a court could defer to an agency’s interpretation of an ambiguous regulation only after concluding the regulation was “genuinely ambiguous.”²⁷⁶ A regulation was genuinely ambiguous if a court “exhaust[ed] all [its] ‘traditional tools’ of construction” to no avail.²⁷⁷

On the other hand, Auer deference could be on its way out if one considers Chief Justice Roberts’s reasons for overruling Chevron deference in Loper Bright. Whereas, for the Roberts Court, Chevron deference contravenes the APA’s command that courts independently determine questions of law, Auer deference also seems to contradict the APA because the APA “directs courts to ‘determine the meaning or applicability of the terms of . . . agency action[s],’ which includes regulations.”²⁷⁸ Another issue with Auer deference, like Chevron deference, involves separation of powers concerns. Justice Scalia questioned Auer deference because it “arrogat[ed]” executive power by enabling agencies to “write . . . law[s] and . . . interpret” such laws.²⁷⁹ Similarly, Chevron deference interfered with “judges[’] [ability] to exercise . . . judgment independent of influence from the political branches.”²⁸⁰ This could explain why Justice Breyer and Justice Scalia did not defend Chevron deference on separation of powers grounds.

B. Quasi-Judicial Decisions

Loper Bright represents the Roberts Court’s view that, regarding agencies’ quasi-legislative powers, agencies have little power to say what the law is. Meanwhile, as to agencies’ quasi-judicial powers, the Court tends to posit that agencies have little power to say who the law applies to: in Axon Enterprise, Inc. v. FTC, the Court held that agencies’ ability to channel certain claims through agency tribunals—in which ALJs promulgate orders subject to review by intra-agency appeals boards—before these claims are entitled to Article III review is limited in certain circumstances;²⁸¹ SEC v. Jarkesy then found that the Seventh Amendment sometimes forbade agencies from binding individuals altogether.²⁸²

Axon Enterprise and Exhaustion of Administrative Remedies

Axon Enterprise traces its holding to Thunder Basin Coal Co. v. Reich.²⁸³ In Thunder Basin, the Court considered a coal company’s challenge to a regulation under the Mine Act, which required mine operators to “post . . . information” indicating whom miners had selected to represent them during required site “safety inspections.”²⁸⁴ Initially, the coal company asked the Mine Safety and Health Administration (“MSHA”) to find that the miners representative requirement violated the National Labor Relations Act’s collective-bargaining provisions, but the MSHA declined.²⁸⁵ Before the secretary of labor sought enforcement of the regulation, the company sought injunctive relief in a district court.²⁸⁶ Additionally, the company argued that the Mine Act’s judicial review scheme—in which ALJs under the Federal Mine Safety and Health Review Commission were supposed to resolve mine operators’ complaints, subject to review by the full Commission, before employers could seek review by Article III appeals courts—violated its Fifth Amendment due process rights.²⁸⁷ Once before the Supreme Court, the Court held that Congress could channel certain claims through agency tribunals and preclude Article III jurisdiction until agencies had issued final orders.²⁸⁸ In so holding, the Court articulated three factors that should aid courts’ inquiry in determining whether Congress “inten[ded]” to preclude review: (1) whether “a finding of preclusion [would] foreclose all meaningful judicial review”; (2) whether a party’s “claims were ‘wholly collateral’ to a statute’s review provisions”; and (3) whether these “claims were ‘outside [an] agency’s expertise.’ ”²⁸⁹

Here, circumstances indicated that Congress intended the Mine Act to preclude district court jurisdiction.²⁹⁰ For one, the Mine Act’s comprehensive review scheme applied to all Mine Act violations and did not “distinguish between pre-enforcement and post-enforcement challenges.”²⁹¹ That the Mine Act provided for appellate court review suggested that the coal company, even if forced to bring its claim to the Commission first, could eventually obtain meaningful judicial review.²⁹² Although the Commission probably could not resolve the company’s constitutional claim, the Court was satisfied that the company had recourse given the Mine Act’s appellate review provision.²⁹³ After all, Congress created the Commission to resolve all claims that fell under the Mine Act and, as such, had requisite jurisdiction and experience to handle the company’s miners representative’s claim.²⁹⁴ This meant that the company’s purpose for filing its suit in the district court, to ask a court to declare that they did not need to post information regarding the miners’ representative, meant it brought a claim that fell within the Mine Act’s review provisions and implicated the Commission’s expertise.²⁹⁵

Axon Enterprise found the Thunder Basin factors counseled in favor of finding initial district court jurisdiction when individuals alleged the SEC and Federal Trade Commission (“FTC”) were structurally defective or that any intra-agency proceedings brought against them were unlawful.²⁹⁶ In this case, the Court handled claims brought by Michelle Cochran, whom the SEC charged with “fail[ure] to comply with” statutory “auditing standards,” and Axon Enterprise, which the FTC alleged violated its “ban on unfair methods of competition.”²⁹⁷ After the SEC and FTC initiated agency proceedings against Cochran and Axon Enterprise, Cochran and Axon Enterprise each brought lawsuits in federal district court alleging the agencies’ double-for-cause removal tenure protections for ALJs meant ALJs were unconstitutionally “insulate[d] . . . from presidential supervision.”²⁹⁸ Axon Enterprise also argued that the FTC’s dual “prosecutorial and adjudicative functions” violated the separation of powers principle as well.²⁹⁹

The Court explained that 28 U.S.C. § 1331, which codified federal question jurisdiction, gave federal district courts jurisdiction to hear parties’ challenges to agency actions, and Congress could take away jurisdiction in whole or in part.³⁰⁰ It thus explained that Thunder Basin was actually a two-step process in which courts first asked whether Congress created a “comprehensive review process” that deprived ordinary district courts of jurisdiction, and, if so, whether the claims brought before them were “of the type[s that] Congress intended to be reviewed within” such scheme.³⁰¹ It was during this second step that courts would apply the Thunder Basin factors.³⁰²

The Court analyzed Cochran and Axon Enterprise’s claims by comparing its findings in Thunder Basin and other cases, Elgin v. Department of the Treasury³⁰³ and Free Enterprise Fund v. Public Co. Accounting Oversight Board.³⁰⁴ Unlike Thunder Basin and Elgin, in which the parties challenged agencies’ “specific substantive decision[s],” such as fining a company or firing an employee, here, Cochran and Axon Enterprise made generalized claims about the validity of the SEC and FTC’s “structure or very existence.”³⁰⁵ Thus, the Court reasoned Cochran and Axon Enterprise’s double-for-cause removal claim (and Axon Enterprise’s prosecutor-and-adjudicator claim) were likely not those that Congress sought to filter through the agencies’ tribunal systems.³⁰⁶

To be sure, the Court applied the Thunder Basin factors and arrived at the same conclusion. Unlike Thunder Basin and Elgin, in which appeals courts could effectively reverse the harm suffered by the coal company and the discharged government employee in Elgin, Michael Elgin, by revoking the company’s fine or reinstating Elgin, here, an appeals court could not undo Cochran and Axon Enterprise’s harm of “having to appear in [potentially unlawful] proceedings.”³⁰⁷ In other words, the Court found that Cochran and Axon Enterprise’s “here-and-now injury” precluded any meaningful review of their constitutional claims.³⁰⁸ Moreover, like Free Enterprise, in which Free Enterprise challenged a tribunal’s existence, Cochran and Axon Enterprise’s constitutional claims were wholly collateral to the SEC and FTC’s proceedings because Congress established these agencies’ tribunals to merely adjudicate claims relating to auditing practices and business mergers, not constitutional issues.³⁰⁹ Finally, Cochran and Axon Enterprise’s claims “raise[d] ‘standard questions of administrative’ and constitutional law, detached from ‘considerations of agency policy,’ ” meaning such claims did not implicate the SEC or FTC’s expertise.³¹⁰ This was unlike Thunder Basin and Elgin, in which the coal company’s and employee’s claims were “intertwined with . . . matters on which the [agencies were] expert.”³¹¹

The Axon Enterprise majority opinion did not actually rule on the merits of the constitutional claims brought by Cochran or Axon Enterprise. In fact, Justice Kagan expressly avoided answering whether the SEC or FTC tribunal structures violated the Constitution’s separation of powers principles.³¹² However, Justice Thomas did address this issue in a concurring opinion.³¹³ For Justice Thomas, Congress impermissibly encroached on Article III courts’ judicial power when it channeled adjudication of certain claims through administrative tribunals with deferential judicial review on the back end.³¹⁴ Whether Congress could empower agencies to conduct hearings that would result in the “impos[ition of] orders and penalties on private parties” hinged on the types of rights it sought to regulate.³¹⁵ The types of rights Congress implicated when it empowered the SEC and FTC to adjudicate claims like those the agencies brought against Cochran and Axon Enterprise were private in nature, and therefore the only way such claims could be resolved was by complete Article III adjudication.³¹⁶ Like Hamburger and Bernick, Justice Thomas looked to history for his analysis.³¹⁷ As Justice Thomas explained, courts historically adjudicated private rights, or rights invoking the Constitution’s guarantees of “life, liberty, and property,” from start to end; meanwhile, agencies could resolve matters involving public rights, or those “belong[ing] to the people at large.”³¹⁸ “[M]onetary fines,” including those that the SEC and FTC sought to impose on Cochran and Axon Enterprise, were historically regarded by courts as “property” interests; accordingly, “[a] civil penalty was a type of remedy at common law that could only be enforced in court[].”³¹⁹ That the APA codified the appellate review model did not mean courts no longer had a duty to resolve claims involving private rights.³²⁰ Rather, the Constitution obligated Article III courts to resolve these claims; Article II tribunals had no role in such proceedings.³²¹ This meant that deferential judicial review of agencies’ findings of fact in these cases was constitutionally questionable because it asked judges to exercise independent judgment in only some respects, that is, only as to matters of law.³²² Yet, as Justice Thomas argued, the Constitution “treat[ed] factfinding [no] differently from deciding questions of law” because Article III explicitly granted appellate jurisdiction to the Supreme Court on matters of fact and law.³²³ This suggested courts, like their eighteenth- and nineteenth-century predecessors, should not “defer to the executive record[s] or the facts supposedly established by it” in cases involving private rights.³²⁴

Jarkesy and the Seventh Amendment

Jarkesy responded to some of Justice Thomas’s concerns regarding agencies’ power to adjudicate claims involving private rights.³²⁵ In Jarkesy, the Supreme Court assessed the constitutionality of the Dodd-Frank Act, which gave the SEC the choice to adjudicate an “enforcement action” within the agency or “in federal court.”³²⁶ In this case, the SEC opted to resolve securities fraud charges it brought against George Jarkesy, Jr. and Patriot28, LLC within the SEC.³²⁷ The Court concluded that the SEC could not choose where it adjudicated securities fraud claims because these claims clearly implicated the Seventh Amendment and its guarantee of “trial[s] by jury” in Article III courts.³²⁸

The Court noted that agencies could sometimes adjudicate claims in executive branch tribunals.³²⁹ Specifically, agencies could adjudicate claims it brought against third parties (and vice versa) if (1) such claims did not “implicate the Seventh Amendment” or, (2) if such claims did implicate the Seventh Amendment, such claims fell under the “ ‘public rights’ exception.”³³⁰ The Court explained the Seventh Amendment covered “[s]uits at common law,” which meant a claim implicated the Constitution’s

guarantee of a jury trial if the suit was “not of equity or admiralty jurisdiction.”³³¹

Scholars have noted that “suits at common law” refers to the types of cases that courts heard at the time of the Seventh Amendment’s enactment, with the Federal Judiciary Act of 1789 serving as a key reference point.³³² Section 9 of the Federal Judiciary Act of 1789 gave district courts jurisdiction over “all suits for penalties and forfeitures incurred,” and that “trial of issues in fact . . . in all causes except civil causes of admiralty and maritime jurisdiction[] shall be by jury.”³³³ As John Gibbons pointed out, that “the Framers considered it appropriate for cases involving ‘penalties and forfeitures incurred . . . ’ to be heard by civil juries [was] powerful evidence that the common law presumption was that such cases would be brought in common law courts, rather than equity courts.”³³⁴

In addition to common law claims, the Jarkesy Court found that “statutory claim[s]” that were “legal in nature” also implicated the Seventh Amendment.³³⁵ A statutory claim was legal in nature if the cause of action and remedy, much like common law claims, “sound[ed] in . . . law.”³³⁶ A claim’s remedy was the “ ‘more important’ consideration,” since some causes of action sounded in both law and equity.³³⁷ In Jarkesy, the statutorily created cause of action, securities fraud, resembled common law fraud since “[b]oth target[ed] . . . misrepresent[ation] or conceal[ment of] material facts” and also used similar terms, such as “fraud.”³³⁸ Moreover, the SEC sought money damages against Jarkesy and Patriot28, which was a “common law remedy” because it served to “punish culpable individuals,” rather than to restore parties to the “status quo.”³³⁹ Indeed, the relevant laws, the Securities Act, the Securities Exchange Act, and the Investment Advisers Act, “establish[ed] three ‘tiers’ of civil penalties” in securities fraud cases, in which each tier was a function of how culpable a defendant was.³⁴⁰ Thus, given that the securities fraud charges brought against Jarkesy and Patriot28 were punitive, the Court concluded that these claims implicated the Seventh Amendment.³⁴¹

Since the SEC’s claims implicated the Seventh Amendment, the Court turned to its analysis of the public rights exception.³⁴² The Court noted the public rights exception was narrow and historically applied in cases involving “the collection of revenue[,] . . . foreign commerce[,] . . . immigration[,] . . . relations with [Native American] tribes, . . . the administration of public lands, . . . and the granting of public benefits such as payment to veterans, . . . pensions, . . . and patent rights.”³⁴³ Typically, public rights traced their origins to some “government[] prerogative[]”; for example, in the context of tariffs, the political branches “traditionally held exclusive power over” whether to impose tariffs on other governments.³⁴⁴ In effect, the Court adopted Oil States’ definition of public rights as those rights that belonged to the public at large along with other privileges given to individuals for public policy reasons.³⁴⁵ Here, the SEC’s claim did not implicate this definition because it sought to “target the same basic conduct as common law fraud,” rather than regulate some matter implicating the legislature or executive’s prerogative powers.³⁴⁶ Thus, the Court held that the SEC was required to bring its claims against Jarkesy and Patriot28 in an Article III court lest it violate their Seventh Amendment right to a jury trial.³⁴⁷

Importantly, this holding brought the Supreme Court’s understanding of Article II adjudication in line with pre-twentieth-century courts’ understanding, at least when Seventh Amendment interests are involved: when a claim implicates one’s private rights, an agency cannot adjudicate it. Only an Article III court can.

III. MAPPING THE LIMITS OF THE APA’S SUBSTANTIAL EVIDENCE STANDARD OF REVIEW

If Jarkesy held that claims that invoke the Seventh Amendment and involve private rights must be adjudicated by Article III courts, does that mean agency tribunals can hear claims that involve public rights? If agency tribunals have power to adjudicate at least some claims, can Article III courts hear appeals to such tribunals’ decisions? Moreover, how much deference should these courts accord these tribunals’ findings of fact? This Note argues that Jarkesy and Loper Bright, taken together, affirmatively resolve these questions: agencies may freely adjudicate matters involving public rights and their decisions are subject to review using the APA’s substantial evidence standard of review.

A. Resolution of Public Rights Claims by Agency Tribunals

Taking Jarkesy’s holding to its natural limit, it is clear that agencies may continue to adjudicate public rights claims. As Jarkesy recognized that the public rights exception attaches when Congress or the president have prerogative to act or bind individuals, such as in the tariffs realm,³⁴⁸ it follows that such cases can be resolved by the executive or legislative branches since they have the requisite power to resolve them. For example, Oil States’s holding that a patent is a public right—and therefore, that the executive branch can freely grant or revoke it—fits nicely within this framework.³⁴⁹ As the Constitution granted Congress the power to issue patents, and Congress in turn passed legislation enabling executive officers to issue and revoke patents within certain statutorily defined limits, it follows that the inter partes review process constitutes a government prerogative.³⁵⁰ Here, an Article III court need not involve itself in the inter partes review process because the executive branch has the necessary powers to determine whether to revoke a patent.

Not only may agency tribunals adjudicate claims involving public rights without violating the Constitution, but practical considerations also support this conclusion. The Social Security Administration, for example, handles over 650,000 social security benefits claims a year, meaning that it would be challenging for federal district courts to suddenly take on this caseload.³⁵¹ Additionally, transferring this caseload to Article III courts could harm litigants, since agency adjudication tends to promote consistency among decisions and leads to quicker and more affordable resolution of claims.³⁵² Fortunately, social security benefits claims properly fall under the category of public rights since they are “public benefits.”³⁵³ Therefore, for Jarkesy to stand for the proposition that agency tribunals could not resolve these types of claims would be to create an unmanageable situation in Article III courts.

B. To Review or Not to Review: Agency Tribunals’ Findings of Fact in Public Rights Cases

Since agency tribunals can adjudicate public rights cases, the logical follow-up question is whether Article III courts can hear appeals to decisions made by agency tribunals. It is clear that these courts can. After all, the Supreme Court routinely weighs in on agencies’ decisions, including those made by agency tribunals, suggesting that it considers its jurisdiction over these matters a given.³⁵⁴ Consider the Thunder Basin line of cases: in determining whether Congress intended to deprive Article III courts of jurisdiction over certain claims, courts must consider whether preclusion would foreclose all meaningful judicial review.³⁵⁵ This makes it clear that Article III courts must have some power to hear or review cases that agencies may also have an interest in resolving, especially when important constitutional issues are involved. After all, the Axon Enterprise and Free Enterprise Courts permitted original district court jurisdiction when the parties’ claims involved interpreting constitutional and statutory text to determine the validity of agencies’ structures and existences. Likewise, Thunder Basin assumed that Article III courts would have appellate jurisdiction over the coal company’s constitutional claims.³⁵⁶

Also, it would make little sense for courts to abdicate this kind of responsibility to review agency tribunals’ decisions, especially given the Court’s recent skepticism toward agency power. A more pressing question—with broader implications—is the extent to which reviewing courts should defer to agency tribunals’ findings of fact. This Section argues the APA’s substantial evidence standard of review, in light of the Court’s recent administrative law jurisprudence, should apply to claims involving public rights; otherwise, claims should be handled by Article III courts in the first instance, rendering the APA’s substantial evidence irrelevant in such cases.

Revisiting the Appellate Review Model

As some scholars point out, before the twentieth century, courts treated agencies’ findings of fact as rebuttable, prima facie evidence, even in public rights cases.³⁵⁷ This indicates that, if history accurately describes constitutional limits, Congress cannot relax the applicable standard of review. In other words, if the agency skeptics’ argument is true, then the APA’s substantial evidence standard of review may be unlawful.

However, Loper Bright likely neutralizes this argument. In Loper Bright, to determine whether the Chevron doctrine could be reconciled with the Constitution, the Court surveyed judicial treatment of questions of fact and law during the New Deal era.³⁵⁸ This implies that, at least in the Court’s mind, the New Deal era faithfully adhered to and administered the Constitution regarding judicial review of agency decisions, and so how courts reviewed agency decisions then is how courts should review such decisions today.³⁵⁹ Specifically, the Court noted that, even amid relatively significant government expansion, New Deal–era courts reviewed questions of law de novo and questions of fact as “binding on the courts, provided that there was ‘[substantial] evidence to support the findings.’ ”³⁶⁰ Of course, this describes the appellate review model, which, as many agency skeptics posited, was not the model used by courts at the time of the Constitution’s adoption. So, perhaps the Court is incorrect. Perhaps the pre–appellate review model era, rather than the New Deal era, more faithfully adhered to the Constitution. But what matters is not history but the Court’s version of the law since it is its job to say what the law is.

More precisely, the “law,” as conceived by the Court, might be best captured by Crowell. To see how this may be true, it is important to note that Justice Roberts cited the 1936 decision St. Joseph Stock Yards Co. v. United States for his proposition that New Deal–era courts did (and should) apply the substantial evidence test to questions of fact. In St. Joseph Stock Yards, the Supreme Court contemplated the issue of whether the secretary of agriculture effected an unconstitutional deprivation of property in “prescrib[ing] maximum charges for” a stock yard’s services.³⁶¹ In answering this question in the negative,³⁶² the Court distinguished ordinary fact-finding from constitutional fact-finding: whereas courts could defer to agencies’ findings of fact made pursuant to agencies’ lawfully delegated quasi-legislative power, courts had to independently assess agencies’ findings of fact when “constitutional rights . . . [were] involved.”³⁶³ This harkens back to Crowell, in which the Supreme Court concluded, four years earlier in 1932, that courts had to independently assess what facts had to exist for an agency action to be upheld as constitutional.³⁶⁴ Loper Bright’s reference to St. Joseph Stock Yards thus suggests that the appellate review model, specifically that envisioned by Crowell and applied in St. Joseph Stock Yards, offers the appropriate analysis.³⁶⁵ Although Loper Bright did not explicitly distinguish ordinary facts from jurisdictional and constitutional facts (which may bolster scholars’ claim that de novo review of constitutional and jurisdictional facts is no more), its reference to St. Joseph Stock Yards, its suggestion that New Deal–era courts applied the appropriate standards of review to agencies’ findings of fact, and its general embrace of the substantial evidence standard challenges the view expressed by some scholars and Supreme Court Justices that agencies’ findings of fact cannot be accorded deference.

Congress’s Power to Prescribe the Applicable Standard of Review

Elsewhere in the Loper Bright decision, the Court pointed out that when Congress exercised its discretionary powers, “its determinations” were “conclusive” in court proceedings.³⁶⁶ One area in which the government has discretion to act, as Jarkesy made clear, is in public rights cases. So, when Congress creates, limits, or revokes public rights, its determinations are controlling.³⁶⁷ But while discretion implies a right to do whatever one pleases, it also implies a right to constrain one’s own actions, delegate this power to constrain to others, and determine the extent to which oneself or one’s agent will be constrained.

Loper Bright took for granted that Congress can delegate its discretionary powers to agencies.³⁶⁸ So, when Congress delegates its discretionary powers to agencies, it can also decide how much agencies will be constrained.³⁶⁹

Setting the standard of judicial review is one way Congress can constrain agencies’ actions. To understand how this works, suppose Congress wants to create a social assistance program and passes a law that establishes an agency and intra-agency tribunal to regulate these benefits. As social benefits would fall under the public rights doctrine, the political branches would have discretion to police them. Therefore, Congress could instruct courts to review agency tribunals’ findings of fact under the substantial evidence test since it has the power to constrain itself, via the agency’s actions. Here, a reviewing court merely “defers to the law,” to use Hamburger’s words (and the idea espoused by Justice Breyer and Justice Scalia in defending Chevron deference, for that matter), when it analyzes the agency’s findings of fact for substantial evidence because that is what the statute tells it to do.³⁷⁰

Of course, this is exactly what the APA does: it constrains Congress’s discretionary authority. Indeed, as Loper Bright explained, under the APA, courts decide “all relevant questions of law, interpret[ations of] constitutional and statutory provisions, and determin[ations of] the meaning or applicability of the terms of an agency action.”³⁷¹ If Congress wanted agencies to have more or less authority to declare what the facts were in public rights cases with little interference from reviewing courts, it could. Apparently, the APA’s drafters opted to give agencies moderate discretion to find facts by holding that reviewing courts were to uphold such facts if supported by substantial evidence.

That the substantial evidence standard has its roots in judicial review of jury findings of fact is of no consequence here given that Congress has wide discretion to prescribe the applicable standard of review when it has discretionary power to act, such as in public rights cases. Thus, the substantial evidence standard of review can apply in public rights cases because Congress, through agencies, may regulate public rights and tell courts that the law is to review its findings of fact made pursuant to formal agency adjudications for substantial evidence. This remains true no matter how impractical this standard of review may be in the eyes of agency critics.³⁷²

On the other hand, if a claim implicates the Seventh Amendment and involves a private right, an agency has no power to decide it. This means that the APA’s substantial evidence standard of review would not apply in Seventh Amendment, private rights cases since appellate review would necessarily follow resolution by an Article III court, the only body capable of resolving such matters. In other words, the APA’s substantial evidence standard is triggered by review of an agency’s formal adjudication, but an agency cannot even formally adjudicate cases implicating the Seventh Amendment and private rights in the first place.

As for when the Seventh Amendment is not implicated but one’s private rights are involved, the current Supreme Court does not definitively answer the constitutionality issue. On one hand, Loper Bright could be interpreted to authorize review of agencies’ findings of fact for substantial evidence because (1) Loper Bright assumed St. Joseph Stock Yards correctly states the law, (2) St. Joseph Stock Yards relied on Crowell’s distinction of ordinary from jurisdictional and constitutional fact-finding, and (3) Crowell itself did not implicate the Seventh Amendment as it was an admiralty case and permitted agency adjudication of private rights.³⁷³ Other support for this position comes from Justice Breyer’s concurrence in Oil States: in that case, Justice Breyer noted that precedent, such as Stern v. Marshall, “should not be read to say that matters involving private rights may never be adjudicated other than by Article III courts.”³⁷⁴ But, on the other hand, the majority in Oil States pointed out that, “[w]hen determining whether a proceeding involves an exercise of Article III judicial power, this Court’s precedents have distinguished between ‘public rights’ and ‘private rights,’ ” which could be read to either say the existence of private rights (1) is dispositive in determining whether agencies may adjudicate a matter or (2) merely a factor to consider.³⁷⁵ Considering that Oil States permitted agency adjudication of patent cancellations because this involved public rights, (1) seems to be the more plausible reading. While this Note does not purport to answer this question in particular, it is clear that, at least in public rights cases, the APA’s substantial evidence standard of review is constitutional.

CONCLUSION

Judicial deference to agencies’ findings of law has been on the retreat since Loper Bright sounded the death knell on Chevron deference. However, the Supreme Court has not exhibited a similar level of distrust toward agencies’ findings of fact. Indeed, since the early days of the APA, the Court has maintained that courts are to uphold agencies’ findings of fact so long “as a reasonable mind might accept [these findings] as adequate to support a conclusion.”³⁷⁶ Yet, an important constitutional issue arises when courts defer to agencies’ findings of fact when both the Seventh Amendment and private rights are implicated. In these cases, agencies have no authority to find facts because a jury must find such facts. Meanwhile, agencies may freely adjudicate public rights cases.

Therefore, when an agency adjudicates claims implicating individuals’ public rights, its findings of fact are entitled to some judicial deference because Congress has somewhat curtailed its discretionary powers through requiring reviewing courts, applying the APA, to analyze these findings according to the substantial evidence test. To the extent that Jarkesy nullified the APA’s substantial evidence standard, it did so when agencies adjudicated matters implicating the Seventh Amendment and private rights. This is because there can be no lawful agency proceeding to review. So, while the APA’s substantial evidence standard of review raises important constitutional questions, any concerns are probably limited to application of the standard in Seventh Amendment, private rights cases. The substantial evidence standard may also raise concerns in non-Seventh Amendment, private rights cases too. Otherwise, the substantial evidence standard of review is constitutional.

99 S. Cal. L. Rev. 725

Download

Lindsey Kate Miller*

* Executive Senior Editor, Southern California Law Review, Volume 99; J.D. Candidate 2026, University of Southern California Gould School of Law; B.A. Political Science 2021, University of California, Los Angeles. Many thanks to my Note Advisor, Professor Jonathan Barnett, for his thoughtful suggestions and encouragement throughout the process. I am also grateful to the members of the Southern California Law Review for their helpful comments and care in preparing this Note for publication. To my family, thank you for your love and support, and for always believing in me.

To Defer or Not to Defer: Squaring the Administrative Procedure Act’s “Substantial Evidence” Standard of Review with the Demise of Agency Deference

INTRODUCTION

I. BACKGROUND

A. Separation of Powers, Checks and Balances, and the Administrative State

The Three Branches of Government

Administrative Agencies: A Fourth Branch of Government?

B. The Administrative Procedure Act and the Standards of Judicial Review

1. The Origins of the APA

The Substantial Evidence Standard of Review

II. THE SUPREME COURT ON ADMINISTRATIVE DEFERENCE

A. Quasi-Legislative Decisions

Skidmore Deference

Chevron Deference and the Loper Bright Response

Auer Deference

B. Quasi-Judicial Decisions

Axon Enterprise and Exhaustion of Administrative Remedies

Jarkesy and the Seventh Amendment

III. MAPPING THE LIMITS OF THE APA’S SUBSTANTIAL EVIDENCE STANDARD OF REVIEW

A. Resolution of Public Rights Claims by Agency Tribunals

B. To Review or Not to Review: Agency Tribunals’ Findings of Fact in Public Rights Cases

Revisiting the Appellate Review Model

Congress’s Power to Prescribe the Applicable Standard of Review

CONCLUSION

99 S. Cal. L. Rev. 725

Like this article?