APA Reform, Jurisdictional Constitutional Facts, Court Agency Deference

---

Summary

• Prior to the APA, courts deferred to agencies on “ordinary” facts and independently reviewed both jurisdictional facts and constitutional facts.
• Historical evidence indicates that the APA was meant to incorporate the differentiation between ordinary, jurisdictional, and constitutional facts.
• Jurisdictional and constitutional facts are explored through a historical lens with consideration of the APA.

iStock.com/ampols

Jump to:

---

---

According to the standard historical account, courts have deferred to agencies on questions of fact since the early twentieth century. But that account is incomplete. In the world before the APA, courts generally deferred to agencies only with respect to “ordinary” facts. With respect to two particularly important types of facts—jurisdictional facts (facts on which an agency’s jurisdiction over a dispute depended) and constitutional facts (facts on which an individual litigant’s constitutional rights depended)—courts applied independent judgment in a similar manner as they did to questions of law. There is strong evidence to suggest that the APA was intended to incorporate this distinction between ordinary, jurisdictional, and constitutional facts. This Article recovers the origins of the distinction and its status at the dawn of the APA; considers its merits; and traces its evolution and diminishment in the eight decades since then.

---

---

---

---

Introduction

A century ago, the terms “jurisdictional facts” and “constitutional facts,” and the doctrines accompanying each term, were among the most important in administrative law. The two interconnected doctrines made up a large portion of the law governing judicial review of agency action. And because judicial review of agency action lay at the center of a national debate over the role of the burgeoning administrative state, the jurisdictional and constitutional fact doctrines were the issues over which many jurists and politicians of the day chose to express their positions in that debate.

Today, the terms jurisdictional facts and constitutional facts have all but disappeared from court opinions, casebooks, and the academic literature of administrative law. Their shrinking role in each edition of Gellhorn and Byse’s Administrative Law: Cases and Comments, one of the field’s most established casebooks, ** provides good evidence of their shrinking role in the law itself. In that casebook’s first edition, the discussion of jurisdictional and constitutional facts took up fifty-one pages—more than half the chapter on judicial review of agency action. By the casebook’s eleventh edition, they were not mentioned at all.

---

| Edition | Year | Pages |
| --- | --- | --- |
| First | 1940 | 51 |
| Second | 1947 | 56 |
| Third | 1954 | 40 |
| Fourth | 1960 | 20 |
| Fifth | 1970 | 18 |
| Sixth | 1974 | 20 |
| Seventh | 1979 | 19 |
| Eighth | 1987 | 8 |
| Ninth | 1995 | 6 |
| Tenth | 2003 | 6 |
| Eleventh | 2011 | 0 |
| Twelfth | 2017 | 0 |
| Thirteenth | 2023 | 0 |

Number of Pages in Gellhorn & Byse’s Administrative Law: Cases and Comments Devoted to Jurisdictional and Constitutional Facts

What were the jurisdictional and constitutional fact doctrines? Why do they appear to have disappeared from administrative law? And what can they teach us about the field today? This Article answers those questions. Although it draws on excellent discussions of the two doctrines in a variety of sources, this Article is, to my knowledge, the first piece of scholarship in fifty years to focus squarely on them.

Part I recovers the doctrines as they existed in 1946 at the dawn of the Administrative Procedure Act (APA). By that time, judicial review of agency adjudicative orders had coalesced into a framework relatively similar to today’s. Thomas Merrill describes it as the “appellate review model.” Like appeals courts reviewing trial court decisions, courts reviewing agency orders exercised de novo review of the agency’s determinations of law but deferred to the agency’s determinations of fact using the “substantial evidence” test. The APA was largely intended to codify that existing model rather than establish a new one.

But reviewing courts did not defer to all agency determinations of fact. Instead, they exercised independent judgment or de novo review when reviewing agency determinations of two particularly important types of facts: (a) jurisdictional facts, which were facts upon which the agency’s jurisdiction over the dispute depended, and (b) constitutional facts, which were facts upon which a litigant’s constitutional rights depended. The jurisdictional fact doctrine and constitutional fact doctrine were the names given to the bodies of law that instructed courts to provide these types of facts different treatment.

The jurisdictional fact doctrine was inherited from the English common law at the Founding. The constitutional fact doctrine was developed by the Supreme Court around the turn of the twentieth century. As the Court looked for ways to restrain burgeoning administrative agencies in the 1920s and 30s, it breathed new life into both doctrines. They achieved their most dramatic expression in Crowell v. Benson, the 1932 case that is often credited as the “fountainhead” of the administrative state. In Crowell, the Court held that the appellate review model was constitutionally permissible because it confined agencies to the role of “adjuncts” to the courts—so long as courts exercised independent judgment or conducted new evidentiary hearings when reviewing agency determinations of jurisdictional and constitutional facts.

Part II pauses the chronology to consider the merits of the doctrines. Although the Court never quite articulated a satisfying legal basis for them, they can be understood as expressions of either due process or Article III’s reservation of “the judicial power” to federal courts. Normatively, they assured that particularly significant questions of fact would receive a “hard look” by an independent adjudicator. That assurance protected the rights and interests of regulated parties, for whom, in many circumstances, “who finds the facts is far more important than who applies the law.” But it was never entirely clear how to draw a principled distinction between jurisdictional and nonjurisdictional facts (or, in some contexts, constitutional and nonconstitutional facts). As a result, the doctrines had the potential to make almost any issue of fact re-litigable in court, rendering agency tribunals impotent. The doctrines also raised challenging questions of comparative institutional competence: Are agencies more competent than courts to make factual determinations in the area of their expertise? If so, does that competence extend to factual determinations related to the scope of their own authority? In the 1930s and 40s, the doctrines were widely debated on these grounds.

Part III picks back up in 1946, at the dawn of the APA, and traces what has happened to the two doctrines since then. There is strong evidence that the APA itself, which was intended to “restate[] the present law as to the scope of judicial review” at the time of its enactment, incorporated the jurisdictional and constitutional fact doctrines. But they were already being cited less often at the time of the APA’s enactment than a decade earlier, and, as the analysis of Gellhorn and Byse’s evolving casebook reveals, that trend continued apace. Substantial evidence review overwhelmed them.

Nevertheless, remnants of the jurisdictional and constitutional fact doctrines continue to exist throughout administrative law. Congress has incorporated a version of the constitutional fact doctrine into several statutes by providing for de novo review of agency determinations of fact related to invidious discrimination. Federal courts often exercise independent judgment on the question of a plaintiff’s citizenship status—both a jurisdictional and a constitutional fact—in cases challenging deportation orders and military detention orders. And when a litigant brings a constitutional challenge to an agency adjudication or an adjudicative order, their claim is often scrutinized more closely than if they had brought a different type of challenge.

The constitutional fact doctrine has also made its mark outside administrative law. It has migrated to the standards governing appellate review of trial court records, where appellate courts regularly apply independent judgment to factual issues involving the First Amendment and constitutional criminal procedural rights. The Supreme Court has also drawn on the constitutional fact doctrine in asserting the importance of applying independent judgment to congressional factfinding. Although the doctrine raises different normative issues in these contexts than it does in the administrative law context in which it was born, its “wandering” across the law suggests the seriousness of the principles it embodies.

Part IV considers what to make of this history. The Article’s central historical claim is that the jurisdictional and constitutional fact doctrines were one of the earliest manifestations of concern over the accountability of executive officials. Tracing that concern across the centuries lends legitimacy to those who express it today, and it also reveals that the scope of judicial review of agency action was one of the first battlefields in the war between proponents and critics of the administrative state. But in the pre-APA era, concern over agency accountability was most often expressed by debating the scope of judicial review of factual determinations, not legal ones. Understanding the terms of that debate reveals how, and why, we ceded robust judicial review of agency factual findings in exchange for robust agency factfinding procedures. It also provides fertile ground from which to identify present and future possibilities in administrative law. Ultimately, this Article is a profile—but not an obituary—of a forgotten part of our law.

I. Recovering the Pre-APA Doctrine

Recovering the doctrine at the dawn of the APA requires recognizing that there were, in fact, two distinct but overlapping doctrines. The jurisdictional fact doctrine dictated that agency findings of fact on which an agency’s jurisdiction depended should be the subject of especially rigorous review by a court. This doctrine was inherited from the English common law, and the jurisdictional limits with which it was originally concerned were statutory limits or limits derived from the common law. The constitutional fact doctrine, by contrast, dictated that agency findings of fact on which an individual litigant’s constitutional rights depended should be the subject of especially rigorous review by a court. This doctrine was developed by the Supreme Court in the first decade of the twentieth century and grounded in due process. Crowell, decided in 1932, affirmed both doctrines and applied them to a category of facts that implicated both jurisdictional and constitutional issues: ones on which an agency’s constitutional jurisdiction depended. These facts were so important, the Court held, that reviewing courts were required to hold new evidentiary hearings before making determinations about them rather than deferring to the agency or even applying independent judgment to the agency record. Crowell was grounded in both due process and Article III.

Crowell ignited a firestorm of commentary, much of it critical. In the decade and a half following the decision, the Court affirmed the jurisdictional and constitutional fact doctrines multiple times, but it declined to apply them in new contexts as it became more solicitous of the administrative state. The underlying constitutional law also changed in ways that sharply reduced the number of circumstances in which the doctrines functioned as relevant constraints on agency action. By the time of the APA’s enactment in 1946, the jurisdictional and constitutional fact doctrines were declining in relevance. But every administrative lawyer would have understood that they were good law—so much so that they formed an essential part of the backdrop against which the statute was enacted.

A. Jurisdictional Facts

Scholars generally trace the development of the jurisdictional fact doctrine to the mid-seventeenth century. In England at that time, justice was often dispensed in summary proceedings by justices of the peace and other local, quasi-judicial officers. As the King’s Bench evolved, it began to provide an opportunity for review on the limited question of whether the officers had acted within their jurisdiction.

The 1643 case Commins v. Massam ** provides a good articulation of the rule around the time of its origin. In that case, a sewer commissioner assessed a fee on the tenant of a rental property. On a writ of certiorari to the commissioner, the tenant told the King’s Bench that the landlord should pay instead. Justice Heath responded that the commissioners were “enabled by the statute to proceed according to their discretions, & therefore if they proceed secundum aequum & bonum, we cannot correct them.” “[B]ut,” he explained, “if they proceed where they have no jurisdiction . . . then they are to be corrected here . . ..” Because the commissioner had jurisdiction to assess the fee, Justice Heath declined to correct him.

Often, but not always, application of the rule hinged on a question of fact. For example, in the 1668 case Terry v. Huntington, a statute permitted an excise commissioner to levy taxes on “strong wines perfectly made.” A winemaker sought recovery in the King’s Bench of the value of a tax assessed on his wines, which he asserted were in fact “low wines.” Chief Justice Hale provided him the opportunity to prove the fact to a jury, holding that the excise commissioner had acted beyond his “stinted, limited jurisdiction” if in fact the winemaker was correct that the taxed wines were “low” rather than “strong.” As the Massam and Huntington cases reveal, the doctrine in its earliest formulation did not provide, per se, that jurisdictional facts should be the subject of stricter review than ordinary facts. Rather, it provided that jurisdictional facts were the only kind of facts reviewable at all.

The United States imported this kind of limited jurisdictional review of governmental action from the English common law. As a result, the nineteenth century was, in Jerry Mashaw’s memorable framing, an era of “bi-polar” oversight of American administrative action. It was bipolar in the sense that “[r]eview was either de novo or nonexistent.” If a litigant wanted to challenge the substance of an officer’s decision, there was usually no forum in which to bring their claim. But if they wanted to challenge the officer’s jurisdiction to issue the decision, they could secure a full hearing in court, including the opportunity to submit new evidence and have their evidence considered de novo by the judge. Although there was no formal jurisprudence of administrative law in the nineteenth century, this “jurisdictional or ultra vires model” of judicial review was a part of the general law.

While the model was indeed limited to consideration of jurisdiction, that description is slightly misleading because jurisdiction was a more malleable concept in the nineteenth century than it is today. It included an officer’s legal authority to act, as the Huntington case reveals and as we might think of it today. But it could also include an officer’s compliance with procedural requirements or the presence of important facts related to liability. In the habeas context, a sentencing court might also be held to lack jurisdiction “if the law under which the defendant was sentenced was unconstitutional” or “if the sentence violated the rule against double jeopardy.” Because governmental action was only reviewable on the basis of jurisdictional defect, courts might characterize as jurisdictional any defect that seemed to merit court intervention. In practice, then, the “jurisdictional” model of judicial review offered courts an aggressive tool to superintend governmental action if they chose to use it.

For our purposes, what is important is that, in the nineteenth century, there was a widespread practice in American courts of exercising de novo review when litigants made fact-based assertions that agency orders had been issued without jurisdiction. For example, the federal Land Department operated a court-like tribunal in which individuals could acquire a patent over land owned by the United States. If the United States did not in fact own the land at the time it issued a patent, then the patent was issued in excess of the Land Department’s jurisdiction. A claimant making this type of claim could receive “plenary review of law and fact” in court.

Miller v. Horton is another illustrative example. In that case, municipal health commissioners killed Miller’s horse under the authority of a statute that permitted them to kill animals “infected” with the diseases “farcy or glanders.” Miller argued in court that the killing was in excess of the commissioners’ jurisdiction because his horse was not, in fact, infected with one of the requisite diseases. Oliver Wendell Holmes, writing as an Associate Justice of the Massachusetts Supreme Court, endorsed Miller’s argument and held that “it does not follow that [the commissioners] can throw the loss on the owner without a hearing” in court.

Although this widespread practice was not universal, it had cohered into something sufficiently regular by the early twentieth century that scholars began to call it a doctrine. As the next Section will discuss, it then gave birth to a parallel doctrine providing for independent judicial review of constitutional, rather than jurisdictional, facts.

B. Constitutional Facts

The early twentieth century saw the birth of modern administrative agencies and the appellate model of judicial review. Under the new model, courts made independent determinations of law but generally deferred to agency determinations of fact as long as those determinations were “supported by substantial evidence.” Two categories of facts, however, received independent judgment rather than substantial evidence review: jurisdictional facts, discussed in the previous Section, and constitutional facts, a new category composed of facts upon which an individual litigant’s constitutional rights depended.

The rise of a new model of judicial review was related to the rise of new types of administrative agencies. The first major federal regulatory agency was the Interstate Commerce Commission (ICC), established in 1887. The Food and Drug Administration (FDA) followed in 1906 and the Federal Trade Commission (FTC) in 1914. These new agencies reflected both the Progressive Era’s faith in scientific expertise and the needs of a growing industrial nation. If they were to be successful, proponents argued, their determinations of fact needed to be accorded finality rather than being subject to searching review by judges with less subject-matter expertise. Congress expressed this sentiment in the 1906 Hepburn Act.

As Thomas Merrill has documented, the Supreme Court responded to the Hepburn Act and the spirit of the Progressive Era by articulating a new standard of review for ICC orders. Here is the standard as articulated in the 1910 case ICC v. Illinois Central Railroad:

Beyond controversy, in determining whether an order of the Commission shall be suspended or set aside, we must consider (a) all relevant questions of constitutional power or right; (b) all pertinent questions as to whether the administrative order is within the scope of the delegated authority under which it purports to have been made, and (c) . . . whether, even although the order be in form within the delegated power, nevertheless it must be treated as not embraced therein, because the exertion of authority which is questioned has been manifested in . . . an unreasonable manner . . ..
Within two years, the Court would come to express category (c) in terms of the now-familiar substantial evidence standard applied to ordinary determinations of fact. Category (b), of course, is an expression of the jurisdictional fact doctrine, which was well established by 1910. But where did category (a) come from?

Although scholars have not addressed the question squarely, I believe the Court’s decision to treat “questions of constitutional power or right” separately from ordinary facts is the product of three factors. First, as the previous Section discussed, constitutionality was sometimes thought to be an issue of jurisdiction in the nineteenth century. A constitutional fact exception might, then, have been considered a subset or extension of the exception for jurisdictional facts. Second, it was the height of the Lochner era, and the Court had recently declared that railroads had a substantive due process right to rates set high enough that they could generate a fair return on their investment. Placing constitutional facts in their own category provided a way for the Court to emphasize the importance of that right. Third, the new appellate review model involved significantly more deference than courts had historically extended to agency decisions. Articulating an explicit constitutional exception allowed the Court to soften the impact of that change by preserving for the courts a larger role in fact review than they would have from a jurisdictional exception alone.

The third explanation makes particular sense when considered against the backdrop of the fierce debates that accompanied the establishment and growth of administrative agencies. By the Progressive Era, the debate was already expressed in terms familiar to us today. Proponents of the new agencies believed that they should be “qualified by experience, fortified by technical assistance, [and] free from the pulls and pressures of politics.” Critics were concerned that the agencies would “degenerate into petty tyranny.” As then-Justice Hughes declared:

The ideal which has been presented in justification of these new agencies, and that which alone holds promise of benefit rather than of hurt to the community, is the ideal of special knowledge, flexibility, disinterestedness, and sound judgment in applying broad legislative principles . . . to the intricate situations created by expanding enterprise. But mere bureaucracy—narrow, partisan, or inexpert—is grossly injurious.
The Court’s approach to judicial review was an “accommodation” of these two competing philosophies about the propriety of the new agencies. By extending deference to ordinary agency determinations of fact, the Court would permit agencies to apply their “special knowledge” and “sound judgment.” But by preserving robust judicial review of jurisdictional and constitutional questions, the Court would not permit agencies to descend into “mere bureaucracy.” Once again in Justice Hughes’s words:

It is not the proper function of the courts to fix rates or to make orders as to the facilities which should be supplied, or the safety appliances which should be used. This is the function of the Legislature or of the administrative board which it may create to aid in securing the performance of the duties it has imposed . . . . [But] where the Legislature goes beyond its constitutional powers, or where the administrative board exceeds its authority or passes its constitutional limits, the matter falls within the jurisdiction of the courts, who will declare such action null and void and prevent any attempt to enforce the provisions of the obnoxious statute or order.
By World War I, the basic pre-APA framework was in place. Courts were to conduct judicial review of an agency order on the basis of the agency record. On questions of law, they were to make an independent determination. On questions of fact, they were to broadly defer to the agency using the substantial evidence rule—except for facts upon which the agency’s jurisdiction or the litigant’s constitutional rights were contingent, which were to receive independent judgment.

The constitutional rights exception to the substantial evidence rule coalesced into a doctrine of its own ** in two decisions in the 1920s: Ohio Valley Water Co. v. Ben Avon Borough and Ng Fung Ho v. White. Ben Avon was a constitutional challenge to a run-of-the-mill ratemaking. Recall that, under the constitutional law of 1920, railroads had a right to rates set high enough that they could generate a fair return on their investment. Whether a given rate provided a fair return was a question of fact. In Ben Avon, the Court held that railroads must have “a fair opportunity for submitting” that question of fact “to a judicial tribunal for determination upon its own independent judgment . . . otherwise, the order is void because in conflict with the due process clause.”

The Ben Avon decision ** excited a barrage of commentary. Although a version of the constitutional fact doctrine had been articulated a decade earlier, Ben Avon gave it life and applied it to ratemaking, which was one of the era’s most contested exercises of government power. The decision was therefore “a nightmare” for advocates of the administrate state. “Why have an administrative state at all if courts were going to do all the work over again, but badly? Or was that the point?” they asked.

Despite that commentary, the Court expanded the constitutional fact doctrine two years later in Ng Fung Ho. Ng Fung Ho was a constitutional challenge to an administrative deportation order. Although the petitioners argued they were U.S. citizens, the Commissioner of Immigration for the Port of San Francisco determined otherwise and ordered their deportation. On judicial review, the Court held that “[j]urisdiction in the executive to order deportation exists only if the person arrested is an alien. The claim of citizenship is thus the denial of an essential jurisdictional fact.” As a result, the Fifth Amendment’s Due Process Clause entitled petitioners “to a judicial determination of their claims that they are citizens.”

Ng Fung Ho was noteworthy because it involved individual liberties dissimilar to the property rights at issue in Ben Avon. And although the decision used the language of jurisdictional facts, its logic rested in large measure on the seriousness of the affected liberty interest. Justice Brandeis, who dissented in Ben Avon but wrote the Court’s opinion in Ng Fung Ho, would come, on these grounds, to be both a critic of the jurisdictional fact doctrine and a proponent of the constitutional fact doctrine. By 1930, the Court was referring to a freestanding constitutional fact doctrine unrelated to the issue of an agency’s jurisdiction.

C. Crowell v. Benson

The jurisdictional and constitutional fact doctrines achieved their most dramatic expression in the 1932 case Crowell v. Benson. As the previous two Sections demonstrated, both doctrines predated the case. But they were articulated so explicitly in Crowell, and prompted so much discussion in the context of that case, that some commentators have described Crowell as their origin. Although Crowell was not their origin, Crowell is indeed the most important case for understanding the two doctrines and the larger “accommodation” they represented between proponents and critics of the growing administrative state.

Crowell was a worker’s compensation case that arose in a shipyard near Mobile, Alabama. Joe Knudsen, a shipyard worker, was injured while splicing cable on a river barge owned by Charles Benson. The precise relationship between Knudsen and Benson is unclear. Given the “fluid” nature of labor arrangements in the shipyard and the complex facts of the case, Knudsen can plausibly be described as either Benson’s employee, an independent contractor, or a volunteer. Nevertheless, after his injury, he filed a claim under the Longshoremen’s and Harbor Workers’ Compensation Act, which provided worker’s compensation to maritime employees.

Knudsen’s claim was adjudicated by the federal Employees’ Compensation Commission. Letus Crowell, the agency’s Deputy Commissioner, conducted an “extremely informal” adjudication in which the primary question was whether Knudsen was in fact Benson’s employee, which was a precondition to a compensation award. Crowell determined that he was. As a result, Benson was ordered to pay.

Seeking to enjoin enforcement of the compensation award, Benson challenged it in federal court. He did so using a provision of the Act that authorized courts to “suspend[] or set aside, in whole or in part” orders of the Commission that were “not in accordance with law.” The order was not in accordance with law, Benson argued, because Knudsen was not in fact his employee. After a hearing that was de novo as to both facts and law, the district court came to the opposite conclusion as the Commission, and the court of appeals affirmed.

When the case reached the Supreme Court, it posed this question: “[u]pon what record shall the district court’s review of the order of the deputy commissioner be based?” It was settled that determinations of law should be made independently by the reviewing court. The question in Crowell, then, was about determinations of fact. Knudsen’s attorneys hoped the Court would use the case to announce that courts should hold de novo hearings for most or all fact issues, and that it would invalidate the Longshoremen’s Act because it did not provide for such hearings. The agency’s attorneys hoped the Court would announce a much more deferential rule. The case thus encapsulated the central role of fact review in the era’s debates over the administrative state.

The Court, in an opinion by Chief Justice Hughes, answered the question with a grand compromise. In cases “involving public rights,” it held, the standard of judicial review “is completely within congressional control.” But in “cases of private right,” the Constitution required something like the appellate review model that had developed over the previous three decades. This model, the Court said, involved separate standards of review for ordinary, constitutional, and jurisdictional facts.

When it came to ordinary facts, “there is no requirement that, in order to maintain the essential attributes of the judicial power, all determinations of fact in constitutional courts shall be made by judges.” For this proposition, the Court cited the role of juries as fact finders and the history of “masters and commissioners” providing judges with factual advisory reports. Agencies, like juries and special masters, could make initial determinations about ordinary facts as adjuncts to a court. Those determinations could be upheld by the reviewing court as long as they were supported by substantial evidence. In Crowell, the ordinary facts were facts about “the circumstances, nature, extent, and consequences of the [employee’s] injuries.”

When it came to constitutional facts, “the judicial power of the United States necessarily extends to the independent determination of all questions.” This holding reaffirmed the Ben Avon and Ng Fung Ho rule by requiring independent judgment as to constitutional facts but permitting that judgment to be exercised on the basis of the agency record.

When it came to jurisdictional facts, the Court took a new approach. “[W]here the determinations of fact are fundamental or ‘jurisdictional’ in the sense that their existence is a condition precedent to the operation of the statutory scheme,” the Court held, a court should review such a determination “upon its own record and the facts elicited before it.” In other words, courts should hold their own evidentiary hearings rather than merely applying independent judgment to the agency record. This holding went a step further than the cases of the 1910s and 20s, and there is evidence that it made the lineup of Justices five-to-three rather than seven-to-one. But, Chief Justice Hughes explained, nothing less than a full evidentiary hearing was compelled by due process and Article III.

In worker’s compensation cases under the Longshoremen’s Act, however, the Court held that the only jurisdictional facts were the ones relating to the agency’s constitutional, as opposed to statutory, jurisdiction over the dispute. That meant (a) the question of whether Knudsen was in fact Benson’s employee, and (b) the question of whether Knudsen’s injury occurred on navigable waters. Thus, although it used the language of jurisdictional fact, the decision suggested that a new evidentiary hearing would be required only for this kind of constitutional -jurisdictional fact. In doing so, it seemed either to narrow the category of jurisdictional facts to exclude statutory-jurisdictional facts or to create a new, third category that drew on both the jurisdictional and constitutional fact doctrines. Justice Butler, reflecting this move in his docket book from the case conference, wrote the word “jurisdictional” with “fundamental” and “essential” next to it in brackets.

After all of this—affirming, in private rights cases, the substantial evidence rule for ordinary facts, independent judgment applied to the agency record for constitutional facts, and a new evidentiary hearing for constitutional-jurisdictional facts—the Court leaned on the constitutional avoidance rule to uphold the Longshoremen’s Act. By empowering courts to invalidate agency orders “not in accordance with law,” the Court held, the Act incorporated precisely the model of judicial review announced in Crowell. By setting aside Knudsen’s compensation award based on its own factual determinations after a new evidentiary hearing, the district court had complied with both the statute and the Constitution.

Justice Brandeis dissented. “Nothing in the Constitution, or in any prior decision of this Court to which attention has been called, lends support to” the requirement of a new evidentiary hearing, he wrote. Although Justice Brandeis had authored the Court’s opinion in Ng Fung Ho, he drew a distinction between that case, which involved “fundamental rights,” and this one, which did not. In the dissent’s view, the constitutional fact doctrine could be justified as an application of due process, but the jurisdictional fact doctrine could not—at least when it required a new evidentiary hearing.

D. Crowell’s Aftermath

Immediately after Crowell, the case was both celebrated and criticized as a major change in administrative law. But over the decade that followed, the Court quietly retreated from the Crowell framework, endorsing deferential substantial evidence review in a wide range of cases. By the dawn of the APA in 1946, Crowell remained good law, but the Court rarely applied it.

Crowell excited a lot of commentary—more “than any case since Ben Avon.” Some comments on the case are noteworthy for their content, others for their author. Felix Frankfurter, then a law professor at Harvard Law School, wrote to Justice Stone:

I am again in mourning. . . . Crowell v. Benson make[s] me wonder whether law is really my beat. . . . The result seems to me the evil offspring of . . . the Ben Avon case[]. It is the result of a very jejune, unreal conception of administrative law. To make the issue of employment more jurisdictional than any other fact upon which liability depends is to turn these matters into a game much more sterile than the speculations of the Schoolmen.
Judge Learned Hand wrote to Frankfurter that the fact distinctions in Crowell “seem[] . . . to me one of the most unnecessary and wanton distinctions that they have got off of late.” Frankfurter agreed, responding sarcastically:

[A]nd so it came to pass that Alexander Hamilton and James Madison and the other Fathers, by conferring the “judicial power” upon the courts, wrote into the Constitution the requirement that whether a longshoreman suffered an injury in connection with admiralty matters or was the employee of the boss or sub-boss, must forever, world without end, be tried de novo in federal court and cannot be determined upon the record of a hearing before some other functionary.
Walter Gellhorn, then a legal secretary to Justice Stone, offered perhaps the most colorful comment in a letter to his classmate Herbert Wechsler:

I think you’re dead right on Crowell v. Benson. It’s a terrible business. Hughes simply created the law of “jurisdictional” facts—they sprang full formed from his penis. It was all, primarily, to avoid the constitutional question. I believe that the Court will be swamped with cases in which counsel will seek to prove that a fact is jurisdictional and should have been tried de novo by a court after the administrative body got through with it. Of course, there will have to be limiting.
There was never an explicit “limiting.” In the 1936 case St. Joseph Stock Yards Co. v. United States, the Court did not address Crowell ’s constitutional-jurisdictional facts but reaffirmed the requirement that courts apply independent judgment to agency findings of traditional constitutional facts. To hold otherwise, the Court explained, would be “to place those rights at the mercy of administrative officials and seriously to impair the security inherent in our judicial safeguards.” That same year, in Baltimore & Ohio Railroad v. United States, the Court reiterated a muscular version of the constitutional fact doctrine in dicta. In the 1940 case United States v. Appalachian Electric Power Co., the Court upheld an agency factual determination after a de novo review of the record.

But at the same time as the Court reaffirmed Crowell, it called it into question. In several cases in the 1930s, litigants challenging agency orders argued—as Walter Gellhorn predicted they would—that new evidentiary hearings should be required for jurisdictional facts beyond the two specific facts at issue in Crowell (injury location and employment relationship). The Court simply ignored their arguments. Then, in the Rowan & Nichols cases in 1940 and 1941, lower courts exercised independent judgment on factual issues underlying an oil company’s due process challenge to its daily production limits—and the Court reversed them. Justice Frankfurter, now on the Court, took the opportunity to cast doubt on Crowell and St. Joseph Stock Yards without actually mentioning either case. “[I]t would be presumptuous for courts, on the basis of conflicting expert testimony, to deem the view of the administrative tribunal, acting under legislative authority, offensive” to due process, he wrote for the Court in Rowan & Nichols I. “Nothing in the Constitution warrants a rejection of these expert conclusions. Nor, on the basis of intrinsic skills and equipment, are the federal courts qualified to set their independent judgment on such matters against that of the” agency, he added for the Court in Rowan & Nichols II.

Why was Crowell announced with a bang but followed with a whimper? One explanation is the Court’s shifting composition in the 1930s and 40s. President Roosevelt appointed Justice Black in 1937, Justice Reed in 1938, Justices Douglas and Frankfurter in 1939, Justice Murphy in 1940, Chief Justice Stone and Justices Byrnes and Jackson in 1941, and Justice Rutledge in 1943. Many of the Roosevelt appointees were more solicitous of governmental power than their predecessors. Justice Frankfurter, who had extolled the virtues of the administrative state and passionately criticized the jurisdictional and constitutional fact doctrines as a law professor, cabined Crowell as soon as he had the opportunity.

Another explanation is the changes in constitutional law that occurred shortly after Crowell. In the late 1930s and early 1940s, the Court significantly expanded Congress’s authority to act under its Article I, Section 8 powers, especially its Commerce Clause powers; permitted Congress to broadly delegate that authority to administrative agencies; and eroded the individual economic liberties that had been grounded in substantive due process. By the end of the New Deal era, there were far fewer constitutional limitations on agency action than there were at the beginning. The Court did not need to overrule Crowell because the type of constitutional-jurisdictional facts at issue in that case were no longer constitutional at all.

In this sense, there was “a ‘doctrinal synergy’ between the decline of substantive due process and the rise of judicial deference to administrative action.” The grand compromise struck by Chief Justice Hughes in Crowell, which was intended to settle the debates about the administrative state that had raged for three decades, did not last because administrative law was overtaken by constitutional law.

By the birth of the APA in 1946, the status of the jurisdictional and constitutional fact doctrines was in transition. None of the cases establishing the doctrines had been overruled, and they still took up large sections of administrative law casebooks. But as one academic wrote, the “tendency” in the judiciary was “strongly toward” substantial evidence review. In another’s words, “constitutional facts are possibly destined to be denied special treatment in the future.” Justice Frankfurter, concurring in a case that year, referenced “the attritions of [Crowell ] through later decisions” and argued that “the doctrine had earned a deserved repose” because of its “casuistic difficulties.” Of course, Justice Frankfurter might have been indulging in as much advocacy as he was description. It would take time before anyone knew if the doctrines had in fact earned repose.

---

II. Making Sense of the Doctrine

This Part pauses the chronology in 1946 to consider the merits. As the Court developed the jurisdictional and constitutional fact doctrines over the first few decades of the twentieth century, it never quite offered a satisfying justification for them. Why should jurisdictional and constitutional facts receive different treatment on judicial review than ordinary facts? In Section II.A, this Article proposes several answers to that question. In Section II.B, it then considers counterarguments and criticisms, drawing especially on the criticisms expressed in the academic literature of the 1930s and 40s. What emerges is a set of doctrines that furthers serious constitutional and normative ends but presents similarly serious practical difficulties.

A. Motivations

First, the jurisdictional and constitutional fact doctrines might be compelled by Article III. On this account, “the judicial [p]ower” vested exclusively in the federal courts contains within it the application of independent judgment to jurisdictional and constitutional fact issues. Formally, that might be so because the jurisdictional fact doctrine was part of the common law at the Founding. Functionally, it might be so because one of the Framers’ central concerns in designing Article III was ensuring that the political branches would not be “themselves the constitutional judges of their own powers.” Granting an agency finality, or something close to finality, on its determinations of jurisdictional and constitutional facts would thus “be to sap the judicial power as it exists under the Federal Constitution, and to establish a government of a bureaucratic character alien to our system.”

This argument fits uncomfortably into our modern understanding of Article III. The entire appellate model of judicial review depends on the proposition that the judicial power extends more powerfully to findings of law than it does to findings of fact. “It is emphatically the province and duty of the judicial department to say what the law is” —not to “say what the facts are.” But the appellate model at one time treated jurisdictional and constitutional facts differently from ordinary facts. Given that rich history, our modern understanding of Article III may be missing something.

Second, the jurisdictional and constitutional fact doctrines might be compelled by the Fifth Amendment’s Due Process Clause. On this account, “a fair trial in a fair tribunal is a basic requirement of due process,” and an agency cannot, as a matter of law, be fair when deciding questions related to the scope of its own authority. Permitting an agency to do so “leaves the fox in charge of the henhouse.” Chief Justice Hughes framed the issue in these terms in a 1931 address to the American Bar Association:

The power of administrative bodies to make findings of fact which may be treated as conclusive, if there is evidence both ways, is a power of enormous consequence. An unscrupulous administrator might be tempted to say, “Let me find the facts for the people of my country, and I care little who lays down the general principles.”
Alternatively, the jurisdictional and constitutional fact doctrines might be compelled by due process simply because jurisdictional and constitutional facts are important enough to individual litigants that they should not be finally determined in a tribunal that lacks the procedures of a court. That was the logic on which Justice Brandeis seemed to base the Court’s decision in Ng Fung Ho.

It is also the logic underlying historical de novo review of habeas applications.

Third, the jurisdictional and constitutional fact doctrines might be an effective method of enforcing limits on Congress’s power. Those limits come primarily from Article I, which makes Congress’s powers “few and defined,” and from the Bill of Rights, which imposes constraints on Congress’s exercise of even its enumerated powers. Although administrative agencies live in the Executive Branch, they execute power conferred on them by Congress. Therefore, applying independent judgment to traditional jurisdictional facts and the type of constitutional-jurisdictional facts at issue in Crowell helps enforce the Article I limits, while applying independent judgment to traditional constitutional facts helps enforce the Bill of Rights limits.

Fourth, the jurisdictional and constitutional fact doctrines might be an effective method of constraining the administrative state. On this account, a sprawling administrative state is constitutionally problematic and most any doctrine that limits its power is constitutionally virtuous, as long as that doctrine is itself constitutionally permissible. Although this justification is less precise than the first three, it was the one most frequently advanced by commenters justifying the doctrines in the aftermath of Crowell. “[T]he primary concern,” wrote one, “is to save us from bureaucracy.”

Fifth, the jurisdictional and constitutional fact doctrines might be an effective method of dividing authority on the basis of institutional competence. This argument echoes the due process argument but is concerned with accuracy rather than fairness. On this account, courts are better positioned than agencies to determine jurisdictional and constitutional facts—either because agencies are less likely than courts to accurately answer questions related to the scope of their own authority or because agencies are less likely than courts to accurately answer any very important question. The appellate review model is often explained in terms of comparative institutional competence, with agencies being presumed to have “superior competence to resolve questions of fact” and courts superior competence to resolve questions of law. The jurisdictional and constitutional fact argument goes one level deeper by recognizing that agencies’ superior competence may extend to most, but not all, questions of fact.

Sixth, irrespective of comparative institutional competence, the jurisdictional and constitutional fact doctrines might simply be valuable because they provide multiple layers of review for determinations that are particularly important. On this account, two decisionmakers is better than one.

Finally, the jurisdictional fact doctrine (although not the constitutional fact doctrine) might be justified because of the unique nature of jurisdictional issues. On this account, jurisdiction is—and should be—treated differently from other legal issues because it is simply more fundamental. This type of “jurisdictional exceptionalism” informs several other doctrines. For example, courts of limited jurisdiction are presumed not to have it. They are generally supposed to address jurisdictional issues before they address merits issues. And they have an “independent obligation” to verify their jurisdiction even when the parties have not called it into question and even when verification requires expanding the record on appeal. In each of these ways, “jurisdictional questions are exceptional and escape the application of many ordinary principles that would be applicable to the resolution of nonjurisdictional questions.” It is, therefore, consistent with the rest of our law to prescribe that jurisdictional facts receive more searching treatment upon judicial review than nonjurisdictional ones.

B. Criticisms

First, the jurisdictional fact doctrine relies upon a circular conception of agency authority. If a statute provides an agency with the authority to act in particular circumstances, the statute must, by extension, provide the agency with the authority to determine when those circumstances are present. Congress, when vesting the agency with such authority, “always contemplates [liability] flowing from a finding of the facts by [the agency] tribunal, not from the existence of the facts in the absolute.” By requiring the existence of the facts in the absolute, the doctrine effectively makes the agency’s jurisdiction turn on the correctness of its findings. This criticism was levied against both the older version of the jurisdictional fact doctrine and the bolder, constitutionalized version articulated in Crowell.

Second, even if it is coherent to conceptualize an agency’s authority to act as depending upon the existence of facts in the absolute, it is difficult in practice to determine which facts should count. As Section I.A discussed, jurisdiction is a malleable concept. On a broad account of the concept, it is possible to argue that almost any fact related to liability is a jurisdictional fact. City of Arlington v. FCC made much of this challenge, holding—in the context of law, not fact—that the distinction between jurisdictional and nonjurisdictional determinations was “a mirage.” Justice Thomas also made much of this challenge in one of the jurisdictional fact doctrine’s rare mentions at the modern Court, writing in 2015 that the doctrine faded over time because courts “struggled to determine the boundary between jurisdictional and nonjurisdictional facts, and thus to determine the appropriate standard of review for administrative decisions.”

A similar criticism was levied against the constitutional fact doctrine in the 1930s, although it relied for its force on the malleable nature of the era’s constitutional law. “[U]nder the broad interpretation now placed on the Fifth and Fourteenth Amendments,” wrote John Dickinson in 1932, “there is practically no issue going to the substantial merits of a controversy which if ‘unreasonably’ decided by an administrative tribunal cannot be made the basis of a claim of constitutional right.” Today, with substantive due process accorded a much narrower “interpretation,” it is simpler—although not straightforward in every case—to distinguish constitutional from nonconstitutional facts. Of course, the challenge of identifying jurisdictional and constitutional facts is precisely what makes the doctrines such potent tools. They provide a judge reviewing an agency order the freedom, should the judge choose to use it, to exercise independent judgment on a wide range of facts.

Third, as a matter of comparative institutional competence, the argument in favor of the jurisdictional and constitutional fact doctrines might have things upside down. The raison d’etre of agency adjudication is to permit agency adjudicators to bring their expertise to bear in classes of disputes with which they have significant familiarity. It is not obvious why that expertise would not apply with the same force to a dispute’s jurisdictional and constitutional facts as it would to the dispute’s ordinary facts, especially in complex technical cases. And although agency adjudicators are incentivized to find facts in a manner favorable to the agency, the ability to act on that incentive is cabined by the highly proceduralized nature of formal agency adjudication. If that is true, then the doctrines are “indefensible from the standpoint of practical judgments as to the appropriate area of administrative activity.”

Finally, the jurisdictional and constitutional fact doctrines make our system less efficient by permitting relitigation of factual issues already decided. Perhaps the second raison d’etre of agency adjudication is to handle disputes quickly. When a larger portion of those disputes may be relitigated in court, cases take longer to resolve and court dockets swell.

The essence of this criticism, and most of the previous ones, is that the jurisdictional and constitutional fact doctrines sap agency tribunals of their utility. Because courts already exercise independent judgment when reviewing an agency’s legal conclusions, and these doctrines permit courts to exercise independent judgment on many of an agency’s factual findings as well, they erode the capacity of agency adjudication to further the ends of either efficiency or expertise. If you are persuaded that independent review of factual determinations serves positive constitutional and normative ends, then this loss in utility may well be justifiable. If, however, you are dubious of the affirmative arguments in support of the doctrines, then the loss in utility may not be.

III. Tracing the Post-APA Doctrine

Today, it is uncontroversial to state that the jurisdictional and constitutional fact doctrines were “never heard from again” after the 1940s; that they “receded into the constitutional shadows”; or that they are “as strong an example as can be found of being overruled by being ignored.” That is, of course, true if one counts pages taken up by the doctrines in administrative law casebooks or citations to Crowell by the Supreme Court. And it is certainly true that Crowell ’s precise holding, which concerned review of jurisdictional and constitutional facts in private rights cases adjudicated by agency tribunals, has more limited scope in an era in which fewer private rights cases are adjudicated by agencies and there are fewer constitutional limitations on agency jurisdiction.

But the death of the jurisdictional and constitutional fact doctrines has been overstated. The APA’s judicial review provisions may well have been intended to incorporate them. Inside administrative law, they continue to shape judicial review of agency orders in antidiscrimination cases, deportation cases, and military detention cases. And outside administrative law, the doctrines have shaped the standards governing appellate review of trial court decisions on constitutional criminal procedural issues and First Amendment issues.

The overstatement of the doctrines’ death is perhaps related to the overidentification of the doctrines with Crowell. As Part I revealed, both the doctrines predated Crowell —the jurisdictional fact doctrine by almost three hundred years and the constitutional fact doctrine by at least two decades. Crowell was simply their apex expression. It is, therefore, a mistake to conclude that the loss in force of Crowell ’s precise holding necessarily means a loss in force of the doctrines themselves. Although their fortunes are intertwined, the jurisdictional and constitutional fact doctrines continue to exert force throughout the law even though Crowell ’s holding does not. The real story is the doctrines’ mutation—not their death.

A. The APA

The adoption of the APA in 1946 serves as this Article’s fulcrum—the moment before which Part I recovers the doctrine, and after which Part III traces it—because the APA is the foundation of modern administrative law. It is, in Justice Scalia’s words, “a sort of superstatute, or subconstitution.” Although much of contemporary administrative law is administrative common law, even that common law is rooted in, or at least constrained by, the statute’s terms. This Section accordingly asks what those terms say about judicial review of jurisdictional and constitutional facts.

I believe the fairest conclusion from the APA’s text and legislative history is that the statute does not contain a unique standard of review for jurisdictional and constitutional facts. But the text contains several phrases that might plausibly be construed to suggest that it does, and the legislative history contains strong evidence that would support those constructions—so it is a close call. That makes it all the more remarkable that jurisdictional and constitutional facts are missing from the modern administrative lawyer’s toolkit.

The relevant text can be found in APA § 706, which governs the scope of review of agency action. That provision provides that a federal court shall

hold unlawful and set aside agency action, findings, and conclusions found to be—

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

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

(C) in excess of statutory jurisdiction, authority, or limitations, or short of statutory right;

(D) without observance of procedure required by law;

(E) unsupported by substantial evidence in a case subject to sections 556 and 557 of this title 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.
It also provides that “the court shall review the whole record or those parts of it cited by a party” in considering whether the agency action meets one of the standards above.

At least four phrases in § 706 evoke jurisdictional and constitutional facts. First, “not in accordance with law.” That is the precise verbal formulation used in the Longshoremen’s and Harbor Workers’ Compensation Act. In that statute, the Court had, in Crowell, interpreted the phrase to encompass a unique standard of review for jurisdictional and constitutional facts. Second and third, “contrary to constitutional right” and “in excess of statutory jurisdiction.” Those phrases, on their face, evoke the constitutional and jurisdictional fact doctrines as they were understood in the first half of the twentieth century. Fourth, “unwarranted by the facts to the extent that the facts are subject to trial de novo by the reviewing court.” That phrase suggests that the APA’s drafters intended that at least some facts be subject to de novo review, although it does not clarify which ones. All four phrases provide standards besides the substantial evidence standard according to which a reviewing court might plausibly “hold unlawful and set aside . . . findings” of fact.

The legislative history contains little evidence that clarifies the meaning of “not in accordance with law,” “contrary to constitutional right,” or “in excess of statutory jurisdiction” beyond the general statement that § 706 was intended to “declare[] the existing law concerning the scope of judicial review” at the time of the statute’s enactment. The APA’s drafters relied heavily for their understanding of “existing law” on a 1941 report prepared by the Attorney General’s Committee on Administrative Procedure. That report included an extended discussion of the jurisdictional and constitutional fact doctrines and framed them as part of “existing law,” but it also noted that “a fundamental change is taking place” and concluded that “in the future, fact issues involving due process, equal protection, and doubtless also other constitutional guarantees will in all probability no longer be subject to court review as a matter of constitutional right.” Whether the APA’s drafters intended the statute to incorporate the jurisdictional and constitutional fact doctrines thus depends on the extent to which one considers a prediction about law in “the future” to be part of “existing law.”

The legislative history contains more extended discussion about the meaning of “unwarranted by the facts to the extent that the facts are subject to trial de novo by the reviewing court.” A Senate committee print from 1945 explains that this provision did “not attempt to state” the circumstances in which facts should be subject to trial de novo “since the . . . subject is one which the courts themselves have not fully settled.” But the Attorney General’s Manual, published in 1947, observes that the provision “prescribes a judicial trial de novo only in situations where other statutes or the courts have prescribed such review.” Since the courts had prescribed trials de novo for constitutional-jurisdictional facts in Crowell, the Attorney General’s Manual suggests that § 706 incorporated Crowell.

This reasoning involves more conjecture than we might like when answering a question of such magnitude. But, as Aditya Bamzai has observed, the “fact that the APA was enacted against a shifting jurisprudential backdrop in 1946 makes understanding its terms a challenge.” One can certainly marshal the text and the legislative history into a strong argument that the APA codifies the jurisdictional and constitutional fact doctrines. Or, if those doctrines are truly compelled by the Constitution—as the Court held they were in the 1920s and 30s—then it does not much matter what the APA codified, because they supersede the superstatute.

Nevertheless, over time, the substantial evidence standard in § 706(2)(E) came to dominate almost all judicial review of agency factual findings under the APA. Today’s model of judicial review, in which facts are reviewed deferentially and law is not, is missing the pre-APA model’s two categories of nondeferential fact review. In exchange, today’s model has something the pre-APA model did not: relatively strict procedures governing the way facts are found in formal agency adjudications. Whether or not it was intentional, and whether or not the APA sanctioned the exchange, robust judicial review was traded for robust agency procedures as a method of constraining agency factfinding.

B. Inside Administrative Law

Although the jurisdictional and constitutional fact doctrines are no longer regularly cited and applied, remnants of the two doctrines continue to shape administrative law. In at least three ways, the doctrines remain “an active constraint on modern administrative adjudication.”

First, Congress has incorporated a version of the constitutional fact doctrine into several statutes that address invidious discrimination. The relevant statutes include at least Title VII of the Civil Rights Act of 1964, the Age Discrimination in Employment Act, the Fair Housing Act, the Americans with Disabilities Act, and the Rehabilitation Act of 1973. Under each of these statutes, litigants may either bring their claims directly in a court for a hearing that is de novo in the first instance or may, after submitting to an adjudication in an agency tribunal, secure de novo judicial review of the agency order. To be clear, Congress did not provide different standards of review in these statutes for the constitutional facts and the ordinary facts within each litigant’s claim. But by providing a path to de novo review for these litigants’ claims, without doing so for litigants bringing claims under many other statutes, Congress effectively created a heightened standard of review for antidiscrimination claims. In doing so, it codified the “common perception that the courts should control both fact and law in cases where the main issues are constitutional.” To be more precise, it codified the version of the constitutional fact doctrine advanced by Justice Brandeis, according to which the scope of judicial fact review depends upon the significance of the liberty interest at stake.

Second, federal courts generally conduct de novo review on the question of a litigant’s citizenship status—both a jurisdictional and a constitutional fact—in cases challenging deportation orders and military detention orders. De novo review in deportation cases is required by Agosto v. INS. Justice Marshall, writing for the Court in Agosto, traced the requirement directly back to Ng Fung Ho. And the consensus around de novo review in military detention cases is visible in Boumediene v. Bush, in which both the majority and the dissenters suggested that de novo review should be available to U.S. citizens via habeas. Commentators have generally framed these cases as applications of the constitutional, rather than jurisdictional, fact doctrine. But as Section I.B explained, Ng Fung Ho was written in “the unmitigated language of jurisdictional fact.” Because the Constitution constrains the ability of federal authorities to detain and deport U.S. citizens, citizenship status is precisely the type of hybrid constitutional-jurisdictional fact that was at issue in Crowell. The litigant’s individual constitutional rights, and the agency’s constitutional jurisdiction over the dispute, both depend on it.

Third, when a litigant brings a constitutional challenge to an agency adjudication or an adjudicative order, their claim is often scrutinized more closely than if they had brought a different type of challenge. When the litigant’s claim is a structural constitutional challenge to the adjudication itself—such as a challenge to the agency head’s removal protections— Thunder Basin Coal Co. v. Reich ** and Axon v. FTC ** provide that the litigant may bypass the agency tribunal and bring the claim directly in a federal court for de novo consideration, including on the facts. And when the litigant’s claim is a constitutional challenge to the content of a final agency order—such as a challenge to the agency’s determination of a constitutional fact on which the order depends—some lower federal courts permit the litigant to introduce new evidence not contained in the agency record. This exception to the APA’s “record rule,” which otherwise requires judicial review to be conducted on the basis of the agency record, is justified using a variety of rationales. According to one explanation, it is appropriate because “[a] direct constitutional challenge is reviewed independent of the APA.” According to another, it is appropriate because of the “sheer possibility that [an agency] has acted unlawfully.” According to yet another, it is required by Webster v. Doe, which authorized discovery on “colorable constitutional claims” in a context in which other claims were precluded. Of course, it might also be justified by the constitutional fact doctrine, although it rarely is.

Courts are not uniform in permitting a constitutional exception to the record rule. But the exception’s existence in a variety of jurisdictions and legal contexts, and its invocation on the basis of a variety of justifications, suggest that the constitutional fact doctrine is far more alive than it is usually considered to be, albeit under other names. As Martin Redish and William Gohl observe, “it is very unlikely that the Court today would refuse to recognize an obligation of de novo constitutional fact review if presented with an administrative body’s finding of constitutional facts underlying a fundamental right.”

C. Outside Administrative Law

The constitutional fact doctrine, although not the jurisdictional fact doctrine, is similarly alive and well outside administrative law. It has migrated to the standards governing appellate review of trial court records, where appellate courts regularly apply independent judgment to factual issues involving the First Amendment and constitutional criminal procedural rights. The Supreme Court has also drawn on the constitutional fact doctrine in asserting the importance of applying independent judgment to congressional factfinding.

The first area outside administrative law to which the doctrine “wandered” was constitutional criminal procedure. It did so as early as 1935, three years after Crowell, when the Court applied independent judgment in Norris v. Alabama ** to state court findings of fact underlying an equal protection claim about a racially segregated jury. Although the Court did not cite Ben Avon, Ng Fung Ho, or Crowell, those cases constitute Norris ’s “inarticulated major premise.” In the decades that followed, the Court expanded the principle to require independent judgment by appellate courts on factual issues related to forced confessions and double jeopardy determinations.

After constitutional criminal procedure, the doctrine migrated to First Amendment cases. The Court required independent judgment by appellate courts on the factual questions of whether a statement presents a clear and present danger of a substantial evil, whether it is expressive, whether it is obscene, and whether it was made with actual malice. The Court did so in cases in which the fact finder was a jury and in which it was a judge. And it did so in the context of both state court decisions and federal trial court decisions. Today, federal appeals courts generally “require plenary review in all cases involving First Amendment issues.” They do so despite Federal Rule of Civil Procedure 52(a), which requires that “[f]indings of fact” “not be set aside unless clearly erroneous.” Rule 52(a)’s default standard does not apply in the First Amendment context because “the stakes” are simply “too great to entrust them finally to the judgment of the trier of fact.”

The constitutional fact doctrine as applied to constitutional criminal procedural issues and First Amendment issues is plainly not a doctrine of administrative law. As a result, it presents a different set of legal and normative issues. For example, it is challenging to articulate an Article III justification for the doctrine in the context of factual findings by federal district courts. In the context of factual findings made by juries, moreover, it is difficult to square with the Seventh Amendment, which provides that “no fact tried by a jury, shall be otherwise re-examined in any Court of the United States, than according to the rules of the common law.” In addition to “wandering” beyond the bounds of administrative law, the doctrine also shifted its focus from economic liberties to noneconomic ones.

But it is motivated in these contexts by a similar set of concerns to the ones that prompted the development of both the jurisdictional and constitutional fact doctrines in administrative law. Here, as there, it reflects “the importance of providing substantive protection for certain constitutional rights” and “the need for independent review on appeal where there is reason to believe that the original factfinder is biased or incompetent.” And tracing the case law makes clear that it sprang directly out of the administrative law version.

IV. Making Sense of the History

History is always relevant to law. But history is particularly relevant to administrative law, both because the debates that rage over the administrative state are often cast in historical terms and because the field is dynamic enough that historical models of administration provide fertile ground from which to identify present and future possibilities. As Jerry Mashaw writes, although “there is now too much water over the dam to return to the practices of judicial review” exercised by courts a century or more ago, there is still “much to be learned” from analyzing their approach. This Part offers brief reflections on what it is that can be learned from the jurisdictional and constitutional fact doctrines.

The doctrines were one of the earliest manifestations of concern over the accountability of executive officials. In England in the seventeenth century, that concern was directed towards justices of the peace and excise commissioners, who had the opportunity and sometimes the motive to exceed the bounds of their statutory authority when dispensing justice. In the United States in the early twentieth century, that concern was directed towards the ICC and the other early federal agencies, who had a similar opportunity and motive to exceed the constitutional limits placed on them when determining rates and issuing adjudicative orders.

Tracing that concern across the centuries lends legitimacy to those who express it today. When jurists of the twenty-first century express skepticism about the capacity of agencies to self-police, they are not simply (or not always) waging an attack on government regulation. They are advancing a critique that is deeply embedded in our legal tradition. At times, the critique has been advanced by conservative proponents of economic liberties, as it was by Justice McReynolds in Ben Avon and Chief Justice Hughes in Crowell with respect to the rights of regulated corporations. At other times, the critique has been advanced by liberal proponents of noneconomic liberties, as it was by Justice Brandeis in Ng Fung Ho and the second Justice Marshall in Agosto with respect to the rights of immigrants. As is always true in administrative law, agency power may be deployed for both liberal and conservative ends. It is, therefore, most useful to debate the appropriate scope of that power on the basis of neutral principles, not politics.

Tracing the concern also reveals that the scope of judicial review of agency action was one of the first battlefields in the war between proponents and critics of the administrative state. Today, debates over the administrative state are expressed across a range of legal issues: appointment and removal, the nondelegation and major questions doctrines, the assignment of adjudicative authority to agencies, agency procedures and the separation of functions, regulatory planning and review, and so on. But the scope of judicial review was a battlefield before some or most of those issues were battlefields because it is more elemental. While it is unlikely that the English excise commissioner in Terry v. Huntington was subject to complex procedural requirements or a regulatory planning process, the legal architecture of the seventeenth century provided at least for judicial review of his orders. The scope of judicial review may also have been one of the first battlefields because our first and strongest impulse in the Anglo-American legal tradition is to look to courts to secure our liberties—whether or not they can or should provide relief in particular instances.

But in the pre-APA era, concern over agency accountability was most often expressed by debating the scope of judicial review of factual determinations, not legal ones. Our debates over judicial review of legal determinations, recently encapsulated by Chevron and Loper Bright, are a modern phenomenon. Their modernity suggests nothing about their importance. In many contexts, a court’s deference to an agency’s legal determinations is likely to be more consequential than a court’s deference to its factual determinations. But facts were the focus of our debates from the seventeenth century until the dawn of the APA because, across that long span of time, agency power was expressed most powerfully through case-by-case action. Today, deference to legal determinations captures our attention because it is rulemaking, not adjudication, that most powerfully shapes the interaction between the public and the administrative state.

It remains to be seen whether the scope of judicial review of factual determinations will capture our attention once again in the years to come. In the wake of Loper Bright, judicial review of agency action now looks something like it did prior to the enactment of the APA: deferential review of the facts and nondeferential review of the law. But it is missing a critical element of the pre-APA model: nondeferential review of jurisdictional and constitutional facts. Attorneys seeking new ways to constrain agency power, or scholars seeking to return the law to its historic foundations, could advocate for the restoration of this element of the pre-APA model. In doing so, they would have a robust history and compelling legal and normative arguments on which to draw. They might demonstrate that there is not, in fact, “too much water over the dam” to breathe new life into old administrative law practices.

Breathing new life into the jurisdictional and constitutional fact doctrines would, however, be challenging. One of the functions of revisiting history is assessing what worked well in the past and what did not. These doctrines did not work particularly well, at least as applied to the early modern administrative state in the 1930s and 40s. It was in part their “casuistic difficulties” that caused their decline and the ensuing replacement of robust judicial review of agency factual findings (an ex post constraint) with robust procedures governing agency factfinding (an ex ante one). Any effort to reinvigorate the jurisdictional and constitutional fact doctrines would need to account for lessons learned. But regardless of whether they offer a partial blueprint for the future, at least they help explain the past and present.

Conclusion

This Article began with a simple observation: a set of doctrines that was once central to the way scholars discussed administrative law no longer is. In explaining the reasons for that change, it sought to offer a useful taxonomy of the doctrines as they stood at the dawn of the APA in 1946 and as they have evolved and diminished since then. And in offering a taxonomy, it sought to reveal that the concerns fueling modern debates over the administrative state predate the administrative state itself, although they were once expressed in different forms. In a moment in which much of our law is open for reexamination, it is worth remembering jurisdictional and constitutional facts.

---

Endnotes

---

---

Committees

This content was produced by:

• Comparative Administrative Law Committee