Concept
Legislative Courts
Organization and relationship of state and federal courts in a federal system
Legislative Courts
Article III’s text describes a closed system. Section 1 vests the judicial power in one Supreme Court and in the inferior courts Congress chooses to establish, and its next sentence gives the judges of all those courts tenure during good behavior and compensation that cannot be diminished while they remain in office.1 Nothing in that language mentions a second kind of federal court. Hamilton explained why the guarantees matter: “The complete independence of the courts of justice is peculiarly essential in a limited constitution.”2 Read literally, the Constitution reserves federal judicial power to judges who carry those protections, and practice has never followed that reading.
Congress has kept a great deal of federal adjudication outside the system. It has created tribunals beyond Article III since the early Republic, among them courts-martial, territorial courts, and administrative agencies, and their judges often lack Article III’s tenure and salary protection.3 The numbers show how far practice has traveled from the text: in 2021 Congress had authorized 860 Article III judgeships, while nearly 3000 non-Article III adjudicators were at work, about 1930 of them administrative law judges.4 The Madisonian Compromise lets Congress decline to create inferior federal courts at all, so a Congress that wants federal rights adjudicated may rely on state courts or build a non-Article III tribunal. The long history of such tribunals belies, in practice if not in theory, any claim that every court of Congress’s own creation must meet Article III’s standards.5
A legislative court, also called an Article I court, is a tribunal that Congress creates under its legislative powers rather than under Article III. Its judges lack Article III’s salary and tenure protections, and Congress may direct it to resolve disputes without regard to the limits Article III places on the judicial power.6 A court that does meet Article III’s standards is a constitutional court, or Article III court.
The names mislead. Not every such court is created under Article I, and these courts usually sit in the executive branch rather than the legislative one.8 Chief Justice Marshall called them “legislative courts” in an 1828 opinion, and they have existed for more than 200 years.9 The doctrine took shape in three stages: territorial courts in Canter, subject matter in Bakelite, and the District of Columbia and Court of Claims decisions of 1933. Each stage stretched the category, and by the end the Court’s explanations no longer fit together.
Canter and the Territories
The doctrine begins with American Insurance Co. v. Canter. Insurers to whom the owners had abandoned a cargo sued to recover it, claiming that an earlier decree was void because it had not been rendered in a court created by Congress, as Article III requires for the exercise of admiralty jurisdiction.10 Chief Justice Marshall swept the objection aside. Congress had created the Superior Courts of Florida, their judges served only four years, and Article III, he concluded, did not apply in the territories.11 Those courts, he wrote, “are not constitutional Courts, in which the judicial power conferred by the Constitution on the general government, can be deposited. They are incapable of receiving it.” They were instead “legislative Courts, created in virtue of the general right of sovereignty which exists in the government,” or under the clause that lets Congress make all needful rules and regulations for the territory.12
Glidden later read Canter narrowly. All the Chief Justice meant, and what the case has ever after been taken to establish, is that in the territories cases within Article III’s enumeration may be heard and decided in courts “constituted without regard to the limitations of that article”: courts with judges of limited tenure that entertain business beyond conventional cases and controversies.13 The rule is therefore geographic. In the states, admiralty jurisdiction can be exercised only by courts established under Article III, but that limit does not extend to the territories, where Congress exercises the combined powers of the general and of a state government.14 The jurisdiction of a territorial court is not part of the judicial power defined in Article III; Congress confers it in executing its general powers over the territories.15
Two features of territorial government explain why Article III’s guarantees yielded there. First, the absence of a federal structure in the territories produced problems the Framers had not foreseen. Second, independence from Congress and the President mattered less. The territories were not ruled directly from Washington; in a day of poor roads and slow mails, that was unthinkable. Congress left municipal law largely to the territorial legislatures, within the framework of organic acts and subject to a retained veto, and the self-government they exercised was nearly as broad as that of the states.16 The cost of legislative status falls on the judges, who lack Article III’s salary and tenure protections.17 Canter thus teaches where to look: at where the court sits and under what power Congress created it.
Bakelite and Subject Matter
Ex parte Bakelite Corp., decided in 1929, concerned the status of a court Congress had created to hear customs appeals, and it moved the inquiry from geography to function. Justice Van Devanter rejected an argument drawn from tenuous evidence of congressional understanding because it “mistakenly assumes that whether a court is of one class or the other depends on the intention of Congress, whereas the true test lies in the power under which the court was created and in the jurisdiction conferred.”19 The test no longer asks whether the court sits in a territory. It asks what power Congress used and what business the court was given. The opinion in Glidden, discussed below, softens the point about intent: Van Devanter would hardly have denied that explicit evidence of legislative intent about the controlling factors may be relevant and indeed highly persuasive.20
Bakelite did not dissolve the tenure requirement. The Court noted that non-Article III courts could have jurisdiction only over cases involving “public rights” to which the government was a party, though that party factor is no longer imperative. “Inherently judicial” controversies, whatever they might be, were reserved for Article III courts.21
The statutory label does not settle the first grant. Bakelite looks to the power under which Congress created the board and the jurisdiction it conferred, so a tribunal confined to the valuation review can be a legislative court, although explicit evidence of what Congress intended about those factors may be relevant and persuasive.22 The importers’ contract suit differs in kind. It is an inherently judicial controversy, and Bakelite reserved such controversies for Article III courts.23
Hybrid Courts and Williams
Two companion decisions of 1933 showed how unsettled the category had become. Judges of the District of Columbia courts and judges of the Court of Claims, facing a salary reduction, each claimed Article III’s protection of judicial pay, and the Court answered the two claims differently. In O’Donoghue v. United States it sustained the Article III status of the District of Columbia courts, treating them as constitutional courts to which Congress could also assign legislative business, so their judges kept the protection.25
In Williams v. United States the Court went the other way and held the Court of Claims a legislative court, so its judges’ pay could be cut.26 That holding required a textual repair. Article III extends the judicial power to “Controversies to which the United States shall be a party,” and the government is a party in every Court of Claims case. To keep the court legislative, Williams read the clause as if it said “Controversies to which the United States shall be a party plaintiff or petitioner,” which placed suits against the government outside the judicial power.27
Later readers found the pair hard to reconcile. Glidden observed that the Bakelite and Williams decisions “have long been considered of questionable soundness” and “stand uneasily next to O’Donoghue, much of whose reasoning in sustaining the Article III status of the District of Columbia superior courts seems applicable to the Court of Claims and the Court of Customs and Patent Appeals.”28 Set side by side, the four decisions answer different questions:
- Canter held that territorial courts are legislative courts, created under Congress’s authority over the territory.
- Bakelite held that a court’s class depends on the power under which it was created and the jurisdiction conferred, and it reserved inherently judicial controversies for Article III courts.
- O’Donoghue sustained the Article III status of the District of Columbia courts, on reasoning that seemed to reach other courts as well.
- Williams held the Court of Claims legislative by reading the party clause to cover only suits in which the United States is plaintiff or petitioner.
The sections ahead follow two efforts to impose order on this record. Glidden begins from that unease. Justice Brennan later undertook to rationalize the earlier cases as establishing a “constitutional command that the judicial power of the United States must be vested in Art. III courts,” subject to “three narrow situations” in which Congress’s power to create legislative courts was consistent with the separation of powers.29