Concept
Glidden and Article III Status
Organization and relationship of state and federal courts in a federal system
Glidden and Article III Status
Glidden Co. v. Zdanok asked whether Congress can make a court an Article III court by saying so, and the Court answered by examining what Congress had actually built. The yardstick is the text of Article III, Section 1: judges of the Supreme Court and of every inferior court hold office during good behavior and receive compensation that cannot be diminished while they serve.1 The judges of the Court of Claims and the Court of Customs and Patent Appeals had statutory assurances of life tenure and undiminished pay, yet Ex parte Bakelite Corp. and Williams v. United States had treated both courts as legislative tribunals outside Article III.2 Congress later disagreed and provided as to each that “such court is hereby declared to be a court established under article III of the Constitution of the United States.”3
The petitioners asked the Court to reaffirm Bakelite and Williams and so to hold those declarations for naught, at least as applied to judges appointed before Congress enacted them.4 Their request put the judges' own protections at stake. A statutory declaration that judges will serve during good behavior and with undiminished salary binds no later Congress unless the judges received Article III protections at appointment, so each judge's tenure depended on the constitutional status of the court to which he was appointed.5 Two consolidated cases raised the issue.6 In both, the petitioners attacked the participation of judges drawn from these two courts, Judges Madden and Jackson, who sat on regular federal courts by assignment from the Chief Justice.7 Justice Harlan announced the judgment of the Court and an opinion joined by Justices Brennan and Stewart, so the reasoning that follows is that of a plurality.8
The plurality's analysis runs in a fixed order. It first decided that the petitioners could raise their objection at all. It then explained why Canter, the territorial precedent behind Bakelite and Williams, does not govern courts that decide national business, and why the reasoning of those cases failed on its own terms. Next it stated a test for Article III status, weighed Congress's declarations, and applied the test to each court. It closed with the small amount of doubtful jurisdiction each court still held. The concurring and dissenting opinions follow.
Late Challenges to Judges
The petitioners raised the Article III defect for the first time in the Supreme Court. No one had challenged the judges' authority in either case below, and the Solicitor General argued that the de facto doctrine therefore barred the petitioners from questioning the designations on appeal.9 The doctrine rests on a sound policy: it keeps litigants from abiding the outcome of a lawsuit and then overturning an adverse result on a technicality they knew about all along.10 A litigant who spots a flaw in a judge's authority and holds it back gets a free option, keeping a favorable ruling and voiding an unfavorable one. The doctrine denies private litigants that option, although a United States Attorney may be permitted to upset an order issued on defective authority on behalf of the public.11
The doctrine gives way when the defect is more than technical. Where the statute claimed to restrict a judge's authority embodies a strong policy about the proper administration of judicial business, the Court has treated the alleged defect as “jurisdictional” and considered it on direct review even though no one raised it at the first opportunity.12 It gives way more readily when the challenge rests on nonfrivolous constitutional grounds.13 The Court's own decisions on judicial assignments show both sides. In McDowell v. United States, a defect in the statutory authorization for an intracircuit assignment was immunized from examination by the de facto doctrine, yet the Court specifically passed upon and upheld Congress's constitutional authority to provide for such an assignment.14 In Lamar v. United States, the claim that an intercircuit assignment violated the Sixth Amendment's criminal venue restrictions and usurped the presidential appointing power under Article II was heard and decided on the merits, although it first appeared in a supplemental brief filed in the Supreme Court.15
The Court drew one more analogy, this time to diversity jurisdiction, where the existence of diversity of citizenship may be questioned for the first time in the Supreme Court on direct review in order to protect the area the Constitution reserves to the States.16 It concluded: “We hold that it is similarly open to these petitioners to challenge the constitutional authority of the judges below.”17
The ruling left the Court no way around the merits. Given the highly theoretical character of the problem in this setting, the Court said it would be well advised to decide on narrower grounds if any were fairly available, and it found none.18 The point to carry forward is the line between two kinds of defect. Silence below forfeits a technical statutory defect, but it does not forfeit a nonfrivolous constitutional objection to a judge's authority. A fact pattern that stresses the litigants' failure to object invites a waiver answer, and that answer is right only for the technical defect.
Why Canter Did Not Control
Bakelite and Williams leaned on the territorial courts, and the territorial courts rest on Chief Justice Marshall's decision in American Insurance Co. v. Canter. Canter held that the jurisdiction of a territorial court “is not a part of that judicial power which is defined in the 3d article of the Constitution, but is conferred by Congress, in the execution of those general powers which that body possesses over the territories.”20 Admiralty jurisdiction can be exercised in the states only by Article III courts, but “the same limitation does not extend to the territories,” because in legislating for them “Congress exercises the combined powers of the general and of a State government.”21 The question in Glidden was whether that reasoning reaches courts that serve the nation rather than a territory.22
The plurality first explained why it was free to revisit Bakelite and Williams. The Court applies stare decisis less rigidly in constitutional cases than in others, and the doctrine's practical purpose is to protect the expectations people have formed in reliance on a decision. Here the expectations ran the other way: Congress had acted on its own understanding that these were Article III courts and had provided for the assignment of judges whose decisions were now said to be open to attack. Reexamination therefore served the purpose of the doctrine, and the plurality treated the questions those cases decided as entirely open.23
The plurality traced Canter to conditions peculiar to the territories. A territory had no state government, so the national government had to establish courts with enough judges to handle the general jurisdiction that a state's courts would otherwise exercise.24 Territories soon became states, and when they did, their territorial courts lost authority over matters of state concern. That left Congress with territorial judges and no place to put them. When Florida was admitted, for example, Congress replaced three territorial courts of general jurisdiction, with five judges, by one federal district court with one judge.25 A guarantee of life tenure would have saddled the government with judges it could not use. Independence also mattered less in the territories. Poor roads and slow mails meant that Washington did not rule them directly, and Congress left municipal law largely to territorial legislatures acting under organic acts and subject to a retained veto. The self-government they exercised was nearly as broad as that of the states.26
The Court of Claims and the Court of Customs and Patent Appeals share none of those conditions. They were created to carry into effect national powers over subject matter, roughly the payment of debts and the collection of customs revenue, and not over localities.27 They serve no territory awaiting statehood, so statehood will never strip them of their business. Bakelite sketched the history of the territorial and consular courts and moved at once to the customs court without asking whether that difference mattered.28 The Court had presumed that territorial judges who received statutory assurances of life tenure and undiminished pay held them as legislative grace and not constitutional compulsion. By a parity of reasoning, the plurality reversed that presumption for courts whose continuing exercise of jurisdiction is unembarrassed by such practical difficulties.29
Bakelite and Williams also made a logical error, which the plurality called their crucial non sequitur. Each reasoned that because Congress might have assigned the jurisdiction to an administrative agency, it must be deemed to have done so, even though it assigned the jurisdiction to a tribunal having every appearance of a court and composed of judges with statutory assurances of life tenure and undiminished compensation.30 The power to choose among forums does not show which forum Congress chose, and what Congress actually built is the evidence of its choice.
Both earlier opinions had also invoked Justice Curtis's observation in Murray's Lessee v. Hoboken Land & Improvement Co. that Congress has a choice about whether to bring certain matters before the courts. The plurality read that observation as describing a choice, not a prohibition. The very statute that authorized the summary distress proceedings sustained in Murray's Lessee also let the distrainee sue in a federal district court to arrest the levy against the United States.31 Murray's Lessee therefore supports the Court of Claims's Article III status. To deny that Congress may create Article III tribunals for matters it might have reserved for legislative or executive decision would deprive it of the very choice Justice Curtis described.32
One objection remained. A court created for matters Congress might have kept from the courts altogether seems precarious, because Congress could withdraw its entire jurisdiction. The plurality answered that this threat is not different in kind or effect from the one every inferior federal court faces. The great compromise at the Convention authorized Congress to create inferior federal courts without obliging it to, and those courts passed almost a century without exercising any very significant jurisdiction.33
Williams had added a separate barrier. Every case before the two courts has the United States as a party, and Article III extends the judicial power to controversies to which the United States is a party, but Williams held that suits against the United States fall outside that grant. Williams relied in part on the omission of the word “all” before “Controversies” in that clause. The plurality refused to give that semantic circumstance controlling significance, which seemed hardly faithful to Marshall's admonition that “it is a constitution we are expounding.” It treated the real basis of Williams as the rule, well settled when the Convention met, that the sovereign power is immune from suit, and it asked whether that rule makes suits against the United States nonjusticiable in an Article III court.34
The answer was no. Sovereign immunity is one of the postulates behind the words of the constitutional provisions that limit and control, but once the sovereign consents, the postulate is satisfied and no barrier to justiciability remains.35 Consent does not turn the dispute into something foreign to courts. Suits by monstrans de droit and other forms of action designed to gain redress against unlawful action of the English Crown had developed over centuries before the Revolution, and the American states provided judicial remedies against themselves immediately after it.36 Earlier authority such as Minnesota v. Hitchcock had likewise spoken of suits against the United States as controversies within the judicial power. The consequences of the contrary view confirm the point. District courts determine the tort liability of the United States under the Federal Tort Claims Act and so exercise functions akin to those of the Court of Claims, as the statutory authorization of appeals from their judgments to that court shows. Denying that such suits fall within the judicial power would call into question a large measure of the district courts' own jurisdiction.37 The plurality concluded that the presence of the United States as a defendant does not debar either court from exercising the judicial power.38
Courts now read this part of Glidden as having overruled Williams, even though no opinion commanded a majority.39 Justiciability did not displace sovereign immunity. The federal government is immune from civil suit as a default, and Congress must waive that immunity unequivocally and may attach conditions to the waiver, including a requirement of trial before non-Article III judges.40
Status Test and Declarations
Having set aside the reasoning of Bakelite and Williams, the plurality owed an affirmative answer to the petitioners' question: what makes a court an Article III court? Whether a tribunal was created under Article III depends basically on whether its establishing legislation complies with that article's limitations. The plurality put the question this way: whether “its business is the federal business there specified and its judges and judgments are allowed the independence there expressly or impliedly made requisite.” Answering it requires examining each court's history, the development of its functions, and its present characteristics.43
The test has two prongs, and each blocks a different failure. The business prong keeps a court within the federal business Article III specifies. The independence prong looks to the judges and to what they decide: the judges need tenure and pay of the kind Section 1 guarantees, and their judgments must be ones that no officer of another branch can revise. A court meeting only the first prong would decide proper federal cases through judges who depend on the other branches. A court meeting only the second would give insulated judges work Article III does not assign to courts. Both prongs are read from what the court is actually built to do, so a label on the statute answers neither one.
Congress had already spoken to the question. It declared the Court of Claims an Article III court in 1953 and did the same for the Court of Customs and Patent Appeals in 1958.44 The plurality said it could not disregard those declarations, although Congress “may not by fiat overturn the constitutional decisions of this Court.” The legislative history showed that Congress was attempting nothing of the kind. The 1958 House Report described the purpose of the legislation as to “declare which of the powers Congress was intending to exercise when the court was created.”45 That explains why the declarations count. They speak to which power Congress used in creating each court. Later legislation declaring the intent of an earlier law is not conclusive about what the earlier Congress meant, but it is entitled to weight, especially when Congress has been stimulated by the Court's decisions to investigate the historical materials and has drawn a contrary conclusion. The plurality took the 1953 and 1958 declarations as persuasive evidence on the question before it.46
A question that quotes a statutory declaration of Article III status invites the conclusion that the label settles the court's status. The better answer examines the court's business, its judges, and its judgments, and gives the declaration the weight of persuasive evidence.
The Court of Claims meets the business prong on its docket. Its jurisdiction covers claims founded on the Constitution or any law of Congress (except pensions), on any regulation of an executive department, or on any contract with the government, along with certain damages claims not sounding in tort. The statute describes these as claims on which the claimant would be entitled to redress against the United States in a court of law, equity, or admiralty if the United States were suable.47 When the law provides a remedy enforceable in the courts by the regular course of legal procedure and the claimant pursues it, a case arises within the meaning of the Constitution, whether the subject of the suit is property or status.48 The plurality held that the government's presence as defendant does not disqualify either court. It added that all of the business before the two courts is susceptible of disposition in a judicial manner, which left only the question how far it is in fact so disposed of.49
The independence prong turns on judges and judgments. As to the judges, a majority of the Court of Claims's proponents insisted on life tenure when the court was created, as a means of assuring independence of judgment, and the creating Act accepted their proposal. The debates also contain substantial indications that Congress thought it was establishing an Article III court.50 As to the judgments, the question is whether an officer of another branch can revise them. Judicial finality forbids review by the political branches, although superior courts in the Article III hierarchy may review a court's decisions. The circuit-riding Justices of the 1790s refused to hear veterans' claims for just that reason, because the scheme subjected their decisions to review by the executive and legislative branches.51 The Court of Claims's history follows the same principle. In Gordon v. United States, the Supreme Court declined to review the court's judgments while an executive officer could revise them before payment. After Congress removed that authority, United States v. O'Grady treated an unappealed Court of Claims judgment as “just as conclusive under existing laws as the judgment of the Supreme Court, until it is set aside on a motion for a new trial.”52
Payment raised a related worry about finality. Congress had established a standing appropriation, the Judgment Fund, for judgments against the government up to $100,000, and the plurality concluded that the fund, together with Congress's good record of paying, gave a sufficiently firm prospect of payment to satisfy any requirement of finality. Today the fund has no limit and judgments require no separate appropriation bill.53 The plurality even saw the court's creation as a fulfillment of the design of Article III, since sovereign immunity had long kept the laws governing the government's rights and obligations from receiving a fully definitive exposition.54
The Court of Customs and Patent Appeals met the same two prongs by a different route. Under the Customs Administrative Act of 1890, review of decisions of the Board of General Appraisers lay in the circuit courts, which are undoubted Article III courts, and that jurisdiction is what Congress proposed to transfer to the new court. The debates asked only whether the federal system needed a specialized court for customs matters.55 Moving Article III business from one federal court to another leaves the business what it was. Congress had also treated a comparable specialized court as an Article III court. When it abolished that court in 1913, three years after creating it, Congress concluded after extensive debate that the judges then serving were protected in tenure by Article III and assigned them to sit on other constitutional courts.56 After Bakelite, Congress gave the customs judges express good-behavior tenure in the Tariff Act of 1930, and a sponsor explained that when the court was established it was believed to be a constitutional court, so that it had seemed unnecessary to fix a term.57
The petitioners attacked judges, so the plurality also asked whether the judges' protections could be assessed apart from the status of their courts. They could not. The Chief Justice's power to revoke assignments is wholly inconsistent with reading the assignment statute as a power to appoint inferior Article III judges, so a judge assigned by the Chief Justice who lacks constitutional tenure and compensation gains nothing from the designation.58 The constitutional quality of the tenure and compensation extended to Judges Madden and Jackson at confirmation therefore depended on the constitutional status of the courts to which they were primarily appointed.59 Congress had not enacted the present broad assignment statute until after it declared the two courts to be constitutional courts. A major purpose of those declarations was to eliminate uncertainty whether regular Article III judges could be assigned to help these courts reach a quorum when disability or disqualification made that difficult. Holding the judges protected while leaving the courts' own status unsettled would have expanded those doubts rather than allayed them.60
Nonjudicial Remnants and Result
Two small slices of jurisdiction seemed to defeat the business prong. The customs court had jurisdiction over tariff references, and the Court of Claims could receive references from Congress.63 The Supreme Court had questioned the judicial quality of the first kind of jurisdiction in Bakelite without resolving the point. It had treated the second as nonjudicial, so far as susceptibility to Supreme Court review is concerned, in In re Sanborn.64 The plurality declined to decide their validity “in a vacuum.” It noted the doubt attending the jurisdiction and proceeded on the assumption that an Article III court cannot entertain it.65
The assumption mattered only if it could cost the courts their status, so the plurality framed the question precisely. Article III's restriction to cases or controversies reflects more than one concern, among them the Framers' desire to safeguard the judiciary's independence from the other branches by confining it to cases of a judiciary nature. That concern remains fully applicable at least to courts invested with jurisdiction solely over matters of national import. The question was whether either court's independence had been so compromised by its investiture with these particular heads of jurisdiction as to destroy its eligibility for recognition as an Article III court.66
The answer was no. Neither court resembles the Interstate Commerce Commission or the Federal Trade Commission, tribunals a substantial and integral part of whose business is nonjudicial.67 A district court or court of appeals would keep its status if Congress tried to give it such insignificant nonjudicial business, and it would be equally perverse to make the status of these two courts turn on so minuscule a portion of their purported functions.68 Justice Clark's figures show how small the portion was. He counted one tariff reference in 27 years and six in 40.69
Proportion and character decide whether a nonjudicial remnant matters. A body whose substantial and integral business is nonjudicial is an administrative agency, and a court whose work is overwhelmingly cases and controversies remains a court despite a few doubtful matters. If a single rarely used statute could strip a court of Article III status, the court's classification, and with it the tenure and pay of its judges, would hang on a provision that almost never came into play. The plurality therefore treated the doubtful jurisdiction as severable: “if necessary, the particular offensive jurisdiction, and not the courts, would fall.” The history supported that choice. The Congress that enacted the assignment statute and its declarations had been warned that reexamining Bakelite and Williams might lead to disallowance of some of this jurisdiction, and it passed the statute anyway.70
The result followed from that classification. Since the Court of Claims and the Court of Customs and Patent Appeals are courts created under Article III, their judges, retired judges included, are and have been constitutionally protected in tenure and compensation.72 The plurality added that this conclusion is not an ex post facto resurrection of a banished independence. The judges of these courts never accepted the dependent status that Bakelite and Williams assigned them, and the factors those decisions did not consider show that the Court's conclusion confers legal recognition on an independence long exercised in fact.73 Judges Madden and Jackson therefore held Article III protections from their confirmations, and the judgments of the courts of appeals were affirmed.74
The courts Glidden classified no longer exist in their 1962 form. Congress created the Federal Circuit in 1982 as the successor to the Court of Customs and Patent Appeals and to the appellate division of the Court of Claims.75 The Federal Circuit is an Article III court and reviews decisions of the Court of Federal Claims.76 The trial forum for claims against the government moved the other way. The Court of Claims briefly held Article III status after Glidden, and Congress later re-established the tribunal under Article I as the Court of Federal Claims, with judges who lack life tenure.77 Those judges serve fifteen-year terms and may be appointed to a second term.78 Tax Court judges likewise serve fifteen-year terms, must retire at 70, and lack protection against reduction of their compensation.79 Because the plurality's test looks to a court's establishing legislation, Glidden fixed the status of the two courts as Congress had built them, and every later tribunal is classified under its own statute. Congress may place the same kind of business in an Article III court or an Article I court, and the judges' protections follow whichever it chooses.
Concurrence and Dissent
Two separate opinions measured the Court of Claims and the Court of Customs and Patent Appeals against Article III's tenure and compensation guarantees and reached different results.81 Justice Clark concurred in the result, and Justice Douglas would have reversed.82 They disagreed about the same two matters, how much weight a congressional declaration deserves and whether a court's class is fixed by its judges' tenure or by the work the court does.
Justice Clark did not treat the declarations as self-executing. The 1953 statute was, in his words, an ipse dixit that could not by itself make the Court of Claims an Article III court, because the court “must be examined in light of the congressional power exercised and the jurisdiction enjoyed, together with the characteristics of its judges.” But the Act “did definitely establish the intent of the Congress, which prior to that time was not clear in light of the Williams holding 20 years earlier that it was not an Article III court.”83
Clark's account of the courts was developmental. Because the Court of Claims's jurisdiction “was then almost entirely over Article III cases,” he read the 1953 Act as unequivocally expressing Congress's intent that the court be an Article III court, thereby irrevocably establishing life tenure and irreducible salaries for its judges. The customs court followed the same arc. Contrary to the assumption in Bakelite, its business had come to consist exclusively of Article III cases, with tariff references practically nonexistent, and in view of that evolution he believed it became an Article III court upon Congress's clear manifestation of intent in 1958.84 On this view, status moves with a court's jurisdiction and with congressional intent, and the earlier cases need not have been wrong when decided.
The remaining nonjudicial work got a practical answer. Tariff references, “numbering only 6 in 40 years,” are not an Article III court function, so the customs court should, with due deference, refuse to exercise that jurisdiction. If a congressional reference ever reached the Court of Claims, it should advise Congress, as the Supreme Court had advised the President 169 years earlier, that it cannot render advisory opinions.85
Justice Douglas rejected the premise that declarations or statutory tenure could do this work. The Court of Claims, he said, was created under the Necessary and Proper Clause “to pay the Debts,” and the Court of Customs and Patent Appeals in furtherance of the collection of duties, which makes it an administrative court performing essentially an executive task.86 The tenure Congress gave their judges was “of necessity statutory only,” because Article I imposes no requirement of tenure, and “a constitutional amendment would be necessary to supply Article I judges with the guarantees of tenure and salary that Article III gives its judges.” Looking to congressional intent, he said, evades that problem, since Congress always understood it was creating Article I courts. He added that Hamilton, in The Federalist No. 79, had stated the entire case for these guarantees.87
On the declarations, Douglas was categorical. Whether a court belongs to one class or the other does not depend on Congress's intention, he said. The true test lies in the power under which the court was created and the jurisdiction conferred.88 The Court is the expositor of the Constitution, and a congressional enactment in the Article III field deserves no greater weight than in other areas. The declarations would be determinative only if Congress could modify the concepts that radiate throughout Article III and those provisions of the Bill of Rights that specify how the judicial power is to be exercised.89
Douglas also pointed to function. Article III courts of law must sit with a jury in suits where the value in controversy exceeds $20, but the Court of Claims, an Article I court, is not so confined by the Seventh Amendment. It hears claims the government has consented to have heard, and Congress may prescribe the rules and procedures. The customs court hears appeals that “include nothing which inherently or necessarily requires judicial determination, but only matters the determination of which may be, and at times has been, committed exclusively to executive officers.”90 Appointments, he added, follow duties. The President nominates and the Senate confirms in light of the duties of the particular office, so an appointee suited to an Article I tribunal is not chosen for his qualifications to preside at jury trials or to resolve constitutional questions in Article III cases.91 Congress could give Interstate Commerce Commission members lifetime tenure, yet no one would say that a member could sit on a district court, and no difference in function separates the Commission's adjudication from that of these two tribunals.92 His conclusion was that “we subtly undermine the constitutional system when we treat federal judges as fungible.”93
The three approaches to the declarations form a spectrum. The plurality treated them as persuasive evidence of the power Congress meant to exercise, Clark as evidence of intent that the court's real business must confirm, and Douglas as powerless to change what Article III means. Douglas's insistence that a tribunal's work matters as much as its judges' statutory promises lost as to these two courts. The question he pressed, what work a tribunal may do without Article III judges, returns in Northern Pipeline, the next case in this chapter, which also produced a plurality opinion and a separate concurrence in the judgment.94