Whatever their merits or flaws, the protections of Article III make it difficult for Congress to create a flexible, ad hoc judiciary — life tenure means that if you create a temporary court to fill a temporary need, when that court is no longer needed, you’re still stuck with these federal judges on the payroll.
This was a serious issue early in our history, when the federal judiciary as a whole was much smaller, and when the U.S. was constantly acquiring new territories that needed judges. The territories, which lacked the governmental infrastructure of states, were managed directly by Congress — and courts were an important tool in that management. Territorial courts often filled many administrative as well as purely judicial functions, and of course all law in the territories was federal law. So a territory needed many more federal judges than did the states they eventually became. To have hired a bunch of Article III judges at the outset would have left an awful judge surplus upon the transition to statehood. To remedy this problem, Congress relied on its plenary powers over the territories (under Article IV) and the Article I power to “constitute tribunals” to create “territorial courts” with non-life-tenured judges. The same type of court was also appointed Washington, D.C. And this is the first category of non-Article III court. It is completely uncontroversial and has been since the Supreme Court first passed on it in 1828.
The second category of non-Article III court is military courts. These are also relatively uncontroversial. Congress has the power, under Article I, to “make rules for the government and regulation of the land and naval forces,” and this is understood to include providing for military discipline. Of course, one can still ask, why don’t Article III protections apply to the judges in courts martial? Several reasons suggest themselves. First, the Army needs the flexibility to move its JAG personnel around to different roles as military necessity dictates. Some days you’re a judge; some days you’re a battalion commander. A too-rigid scheme of tenured judgeships could interfere with military operations. Second, the ranks of the military swell and ebb depending on the needs of the day; thus, the number of judges will similarly swell and ebb. Third, it may be argued that too much independence on the part of military judges would be counter to the principles of both civilian command of the military and chain-of-command hierarchy within the military. For all these reasons, the military has its own courts.
(Military courts, it should be said, are empowered only to try military members — not, for example, their civilian spouses. There is, however, some question regarding the trial of non-citizens in “military commissions” — though after Hamdan v. Rumsfeld it seems likely that the commissions must at least be authorized by Congress; it is not enough for the President to create them unilaterally.)
The third “category” of legislative courts, such as it is, is both much more controversial than the other two and much less well-defined. It’s generally said to be the category of courts that adjudicate “public rights,” as opposed to “private rights.” At first, the distinction seems sensible — perhaps the Article III courts are only needed to adjudicate traditional disputes between private parties, and Congress may create separate courts to settle disputes between private citizens and the government. After all, the government need not deign to be sued at all (thanks to sovereign immunity), so perhaps if it wishes it may condition its consent to be sued on the case being heard in a somewhat less independent tribunal.
But a moment’s reflection suggests the obvious answer to that proposal — if there’s anywhere we most need a judiciary that’s independent of the other two branches, it is in adjudicating disputes between the people and the government. And even if we accept the “public rights” scheme on the level of political philosophy, the implementation undermines the rationale on multiple levels. First, although it seems clear that we want criminal matters adjudicated by independent courts, criminal prosecutions are probably not really a matter of private law or private rights — suggesting that the public/private distinction is not the only axis along which we are measuring the need for Article III courts. Second, in practice legislative courts are frequently used to decide disputes between private parties. The bankruptcy process, for example, mediates between creditors and debtors — yet bankruptcy courts are legislative, not Article III, courts. Third, in addition to legislative “courts,” this category is often understood as encompassing adjuncts to Article III courts, such as magistrate judges, who often fulfill important district court functions like fact-finding even in criminal and “private rights” cases. Thus, the “public rights” label, as a full justification of all uses of non-Article III judges outside the territorial and military contexts, seems less than helpful.
In any event, let us say that Congress periodically establishes courts in the course of conducting business under its Article I, Section 8 powers, and that because those courts are doing work that seems more like the administration of government business than common-law dispute resolution or criminal fact-finding, they need not meet the Article III requirements. These can include immigration courts, tax courts, “commissions” to decide matters in areas like worker’s compensation, and, as mentioned before, bankruptcy courts.
How narrowly those courts must hew to their particular mission, and how much oversight they must have from the Article III courts, is a confusing question. In Crowell v. Benson (1932), the Supreme Court found that Congress could use commissioners to make decisions in workers’ compensation claims. Noting that the Article III courts frequently used jurors, special masters, and assessors to make determinations of fact, the Court saw no reason why Congress could not appoint commissioners to make “what is virtually an assessment of damages.” However, the Court held that de novo Article III court review of facts necessary to the commission’s jurisdiction was appropriate:
A different question is presented where 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. These fundamental requirements are that the injury occur upon the navigable waters of the United States and that the relation of master and servant exist. These conditions are indispensable to the application of the statute, not only because the Congress has so provided explicitly (§ 3), but also because the power of the Congress to enact the legislation turns upon the existence of these conditions.
In amending and revising the maritime law, the Congress cannot reach beyond the constitutional limits which are inherent in the admiralty and maritime jurisdiction.
In other words, under the Constitution as it was understood at the time, Congress could only regulate workers’ compensation in maritime employment. (It’s helpful to understand that this case was decided during the so-called “Lochner era,” when the Court took a very narrow view of the scope of the Commerce Clause. Today, the Commerce Clause is understood much more expansively, and forms the constitutional basis for a great deal of federal regulation.) Thus, the commission’s findings could not be part of Congress’s lawful regulation of maritime activity unless the incident actually took place in the context of maritime employment — and so a finding a maritime employment was a “fundamental” or “constitutional” or “jurisdictional” fact, not a regular fact. And the big “constitutional” facts, if they were in dispute, had to be determined by Article III courts, because all constitutional issues are decided through use of the judicial power, NOT Congress’s power to regulate or the executive’s power to carry out that regulation. That is perhaps the most definitive thing we know about Article III versus non-Article III courts, so latch onto it.
Later, in Northern Pipeline Construction Company v. Marathon Pipe Line Company (1982), the Court declared unconstitutional a statute creating a new system of bankruptcy courts. Congress vested these courts with the authority to decide, not just bankruptcy issues themselves (which, like maritime issues, plausibly fall under Congress’s Article I, Sec. 8 powers), but “civil proceedings . . . arising in or related to” bankruptcy cases. (In the actual case, the creditor sued a third party for breach of contract under state law in the bankruptcy court, in order to recover money to pay its debts.) Congress explicitly bestowed on these courts the “powers of a court of equity, law, and admiralty.” Yet the bankruptcy judges were not Article III judges: they were appointed to 14-year terms, and their salaries were adjustable by statute.
The Court held that Congress lacked the constitutional authority to vest such broad powers in a non-Article III federal court. Beyond that, however, there was a split of opinion. Justice Brennan, writing for a plurality of four, stated that it was not enough that there be some appellate review available in the Art. III courts: “[T]he constitutional requirements for the exercise of the judicial power must be met at all stages of adjudication.” This suggests that although the boundaries of the judicial power may be hard to determine, once you’ve encroached on that territory, the constitutional requirements are absolute.
He then dispatched the idea that bankruptcy courts fit into one of the three recognized categories in which Congress could create “legislative courts.” Bankruptcy courts were clearly not territorial or military courts, and because they resolved private disputes between individuals, they were also not “public rights” courts.
Brennan also rejected the contention that, under Crowell, bankruptcy courts might best be thought of as adjuncts to the Article III courts — no different from magistrates or the commissioners whose operation Crowell judged constitutional. He distinguished Crowell in four ways. First, the bankruptcy courts were not limited in their scope to narrow factual findings, but made judgments of law as well. Second, the powers given these courts were broad enough to encompass all the powers of a federal district court. Third, the bankruptcy courts’ judgments were self-executing, rather than requiring a district court order to give them effect. And finally, the standard of deference the statute accorded the bankruptcy courts’ decisions — district courts were to overturn them only when they were “clearly erroneous” — was clearly much greater than that in Crowell, where the Article III courts could review at least jurisdictional and constitutional facts de novo. (The canonical example of a court adjunct, the magistrate judge, gets no deference at all — although in practice district judges often adopt the recommendations of the magistrate, they may conduct de novo review at any time.) For all these reasons, Brennan wrote, the bankruptcy courts could not plausibly be seen as mere “adjuncts” to the Article III courts.
Finally, Brennan’s opinion articulated two principles that “aid us in determining the extent to which Congress may constitutionally vest traditionally judicial functions in non-Art. III officers.” First, Congress has more power to vest those functions in non-Article III judges when Congress created the substantive rights at issue as part of its Article I power. The idea, loosely, is that because Congress might not have created the right at all, it has the right to tack on some conditions to the exercise of that right. Because bankruptcy disposes of state-created property rights, Congress has less authority to make conditions for the exercise of the right, including assigning decision-making to non-Article III officers. And second, Congress may not assign essential judicial functions to such officers. Of course, these statements of principle, coming in the text of a mere plurality opinion, are non-binding on future Courts.
Justices Rehnquist and O’Connor concurred with the plurality’s ultimate conclusion — that the statutory scheme was unconstitutional — but rested their votes on narrower and different grounds: namely, that Congress may not assign such unfettered authority over state law claims to a non-Article III court.
Justice White, on the other hand, dissented, arguing that the Court should not create per se rules about non-Article III courts, but should instead weigh the benefits of the use of such a court against the threat it posed to “Article III values” and the separation of powers. Where the subject of litigation was one that was unlikely to involve a lot of encroachment or interference by the political branches, White thought, Congress had greater latitude to employ non-Article III officers and courts.
After the decision in Northern Pipeline, Congress reconstituted the bankruptcy courts so that the bankruptcy judges were now officers of the district courts, like magistrates. They would be appointed by the Courts of Appeals. And although they retained the power to “hear and decide” issues regarding the disposition of property in bankruptcy, in all other ancillary matters they would make recommendations to the district court, which would have the final say. Finally, bankruptcy judges may conduct jury trials (where the Seventh Amendment requires it) in the course of deciding matters that are at the “core” of bankruptcy proceedings, as long as they obtain the consent of all parties.
All this is quite similar to the rules governing the work of a magistrate judge — making it fairly clear that bankruptcy judges are “adjuncts” to the Article III courts and not “legislative courts” like the territorial and military courts, or even the “public rights” courts. Thus Justice Brennan’s “two principles” in Northern Pipeline are, perhaps, simply describing two different phenomena. On the one hand, where Congress creates a substantive right against the government, it may have some leeway to create moderately independent “legislative” courts to administer the right. On the other hand, Congress has general and broad powers to create “adjuncts” to the Article III courts to achieve its legitimate Article I ends — and as long as those adjuncts are sufficiently subordinate to Article III courts, they may be assigned to do fact-finding and carry out some judicial functions regardless of whether the rights involved are “public” or “private.” Thus it may be helpful to think of adjuncts as a separate category, not really connected to “public rights” courts at all.
Perhaps we have, then two broad categories — relatively independent “legislative courts” whose validity rests on certain special administrative powers of Congress, and the “adjuncts,” who are closely monitored by, and often work directly under, the Article III courts. Subsequent decisions have tended to uphold the authority of non-Article III judges who can fit the latter description.
In Commodity Futures Trading Commission v. Schor (1986) the Court again tried to sort out the limits of non-Article III adjudication. The CFTC was empowered to provide reparations to investors swindled by fraudulent brokers. In Schor, the supposedly injured investor lodged a reparations claim against the broker, while at the same time the broker lodged a claim for fees owed in federal district court. The investor submitted a motion to the district court arguing that the court should let the CFTC proceed, since the reparations claim would resolve all the same issues. The broker subsequently agreed to drop its claim in the district court and have it heard as a counterclaim in the CFTC. The CFTC found for the broker and against the investor. After all this, the investor then claimed that the CFTC could not hear the counterclaim, because it was not an Article III court.
The Supreme Court ruled against the investor and found that the CFTC could rule on the counterclaim. Justice O’Connor, writing for a majority, declined to employ Brennan’s two principles or to define a bright-line test for when non-Article III judges or officers may hear common law claims. Rather, she suggested that the inquiry was ad hoc and multi-factor in nature:
Among the factors upon which we have focused are the extent to which the “essential attributes of judicial power” are reserved to Article III courts, and, conversely, the extent to which the non-Article III forum exercises the range of jurisdiction and powers normally vested only in Article III courts, the origins and importance of the right to be adjudicated, and the concerns that drove Congress to depart from the requirements of Article III.
Here, O’Connor wrote, the intrusion on the “essential attributes” of judicial power was relatively minimal.
The CFTC’s adjudicatory powers depart from the traditional agency model in just one respect: the CFTC’s jurisdiction over common law counterclaims . . . .
[W]e are . . . persuaded that there is little practical reason to find that this single deviation from the agency model is fatal to the congressional scheme . . . .
The CFTC, like the agency in Crowell, deals only with a “particularized area of law . . . .” CFTC orders, like those of the agency in Crowell, but unlike those of the bankruptcy courts [in Northern Pipeline], are enforceable only by order of the district court. CFTC orders are also reviewed under the same “weight of the evidence” standard sustained in Crowell, rather than the more deferential standard found lacking in Northern Pipeline. The legal rulings of the CFTC, like the legal determinations of the agency in Crowell, are subject to de novo review. Finally, the CFTC, unlike the bankruptcy courts under the 1978 Act, does not exercise “all ordinary powers of district courts . . . .”
O’Connor declined to distinguish between “public” and “private” rights in this case — further suggesting that this opinion rested primarily on Crowell-type “adjunct” reasoning. It may also have rested, to some degree on consent. It was, initially, investor Schor who sought to have the matter resolved in the CFTC rather than a federal district court. Indeed, he explicitly made the argument that the CFTC’s determination would resolve the broker’s claim as well. Thus, even if he might have been able to object at the beginning to CFTC jurisdiction, by consenting to — in fact, embracing — the Commission’s jurisdiction he essentially waived that objection. Of course, waiver seems quite relevant to the “adjunct” theory, since adjuncts may take on some judicial functions — such as administration of jury trials — with the consent of the parties. Thus, to my eyes, Schor looks very much like an “adjuncts” case, rather than a “legislative courts” case.
Finally, in the recent case of Stern v. Marshall (2011) — one small part of a Dickensian probate dispute between Anna Nicole Smith and her husband’s family — Chief Justice Roberts had to decide whether to consider the judgment of a bankruptcy court a “final judgment” for the purposes of having preclusive effect on other claims. Roberts once again grappled with the weird “public rights” idea before concluding, sensibly, that a dispute between two private parties grounded in state law probably did not meet the definition. (He noted, in particular, that it did not stem from a federal regulatory scheme, and it did not depend on adjudication of a purely federal statutory claim.) Whatever else the “public rights” courts are, they don’t resolve this sort of dispute.
Roberts then found that, once again, Congress had failed to constitute the bankruptcy courts so as to make them properly “adjuncts” to the Article III courts. Although in theory the bankruptcy courts were supposed to limited their authoritative decision-making to “core” bankruptcy issues, the case before the case demonstrated that in fact they were exercising broad jurisdiction over a variety of claims. However, in closing he suggested that bankruptcy courts could still “propos[e] findings of fact and conclusions of law” and present them to the district court judge for approval. It was simply the bankruptcy court’s ability to render final judgments that was constitutionally problematic. Thus, again, the Court seemed quite comfortable with non-Article III judges functioning as adjuncts to the Article III courts, but much less comfortable with the idea of them functioning independently.
I’d also like quote, in passing, Justice Scalia’s concurrence, in which he suggests that the whole “public rights” doctrine has become unwieldy and should be re-thought:
I adhere to my view . . . that — our contrary precedents notwithstanding — “a matter of public rights . . . must at a minimum arise between the government and others.”
The sheer surfeit of factors that the Court was required to consider in this case should arouse the suspicion that something is seriously amiss with our jurisprudence in this area. I count at least seven different reasons given in the Court’s opinion for concluding that an Article III judge was required to adjudicate this lawsuit . . . .
Apart from their sheer numerosity, the more fundamental flaw in the many tests suggested by our jurisprudence is that they have nothing to do with the text or tradition of Article III . . . . The multifactors relied upon today seem to have entered our jurisprudence almost randomly.
Leaving aside certain adjudications by federal administrative agencies, which are governed (for better or worse) by our landmark decision in Crowell v. Benson, 285 U.S. 22, 52 S.Ct. 285, 76 L.Ed. 598 (1932), in my view an Article III judge is required in all federal adjudications, unless there is a firmly established historical practice to the contrary. For that reason — and not because of some intuitive balancing of benefits and harms — I agree that Article III judges are not required in the context of territorial courts, courts-martial, or true “public rights” cases.
I don’t know exactly what Justice Scalia means by “an Article III judge is required in all federal adjudications.” I highly doubt that he is calling for the Court to do away with magistrate judges. So it seems likely, in the end, that he is simply saying that Article III judges should oversee and render final judgment in cases where much of the work may actually be done by adjuncts — whether magistrates, bankruptcy judges, or something else. Scalia, however, holds out the possibility that bankruptcy courts might actually exercise independent jurisdiction as “legislative courts” rather than as adjuncts — but only if “there is a firmly established historical practice” justifying it.