Guaranty Trust and Outcome Determination

Guaranty Trust and Outcome Determination

Guaranty Trust Co. v. York, 326 U.S. 99 (1945), asked whether a federal court that hears a state-created claim solely because of diversity must apply the state’s statute of limitations when the claim is equitable.¹ York sued as the representative of a class of noteholders who had not accepted an exchange offer. She alleged that Guaranty, the trustee, breached its trust by failing to protect the noteholders’ interests in assenting to the offer and by failing to disclose its own self-interest when it sponsored the offer.² Diversity of citizenship was the only basis for federal jurisdiction.³

The old division between law and equity explains why a federal court might have used its own timeliness rule. In the English practice that American courts inherited, law courts awarded damages, while equity courts offered a more flexible set of remedies, such as an accounting of funds, specific performance of a contract, a constructive trust, or an injunction. Article III extends the judicial power to cases in law and equity, and federal courts have exercised both kinds of jurisdiction since their founding.⁴ York’s class action sought a remedy available on the equity side of the federal court.⁵ Equity’s own device for policing delay is laches, which resembles a limitations period but is more flexible: the court asks whether the plaintiff has “slept on her rights,” whether delay has prejudiced the defendant, and similar questions.⁶ That gave York a plausible argument, because laches might let her suit go forward where New York’s statute of limitations would end it. The district court granted Guaranty summary judgment, but the Second Circuit held that a federal court sitting in equity need not apply the state limitations statute that would govern a like suit in a state court, even when diversity is the only basis of federal jurisdiction.⁷ The Supreme Court took the case because the question mattered to the disposition of litigation in the federal courts.⁸

Erie had already settled that a federal court must respect the rights a state creates, and York kept its problem within that setting. The Court put to one side the considerations that govern a federal court adjudicating a claim based on federal law, and it confined itself to transactions whose rights and obligations a state creates and for which Congress has made a federal court another available forum.⁹ It also narrowed its task. It declined to consider whether the New York statute would actually bar the suit in a state court,¹⁰ and it described its only concern as the holding that federal courts in a suit like this are not bound by local law.¹¹

The question was therefore which of two timeliness rules governs a state-created claim heard in equity in a federal court. The labels from the last section do not answer it, because a rule limiting the time for suit can be called procedural or substantive with equal plausibility. The Court answered with a test that looks to effect. The first subsection below states that test; the second shows why, read literally, it could not stand.

The York Test

York answered the choice between federal laches and the state limitations statute with a question about effect: does it “significantly affect the result of a litigation for a federal court to disregard a law of a State that would be controlling in an action upon the same claim by the same parties in a State court”?¹² This is the outcome-determinative test, and it works by comparison. The federal court takes the same claim between the same parties, imagines it in a state court, identifies the rule that would control there, and asks whether setting that rule aside would change how the case comes out. The goal is that, so far as legal rules determine the outcome, a diversity case in federal court should come out substantially as it would in a state court.¹³

Measuring effect replaces the labels “substantive” and “procedural.” It does not matter whether state opinions call limitations statutes substantive or procedural in uses unrelated to the Erie problem.¹⁴ The question is whether a statute concerns merely the manner and means by which a right the state recognizes is enforced, or whether it is a matter of substance in the one aspect that counts here: whether disregarding it would significantly affect the result.¹⁵

The Court did not treat limitations as a special case. Setting abstractions about substance and procedure aside, it noted that federal diversity courts already follow state law on the burden of proof in Cities Service Co. v. Dunlap, on conflict of laws in Klaxon Co. v. Stentor Co., and on contributory negligence in Palmer v. Hoffman.¹⁶ Each of those rules can decide who wins, because the allocation of the burden, the choice of governing law, and a plaintiff’s own fault all bear on whether the plaintiff recovers. A federal court that ignored them would give the litigant a result different from the one a state court would have produced, and that difference is the harm the test is built to detect.

The Court cared about result rather than label because the right being enforced belongs to the state. A diversity plaintiff reaches federal court because of the parties’ citizenship, but the claim itself, including the time within which it must be brought, comes from state law. Suppose the federal court could disregard a limitations bar. A nonresident whose claim is time-barred in state court could sue in federal court and proceed, while a resident with an identical claim would stay barred, and citizenship would decide who gets a hearing. The Court said that an exception to Erie on the equity side of a federal court would “reject the considerations of policy which, after long travail, led to that decision.”¹⁷ It adopted Judge Augustus N. Hand’s objection that disregarding state limitations statutes whenever federal courts think the result inequitable “would promote the choice of United States rather than of state courts in order to gain the advantage of different laws.”¹⁸ That is the forum-shopping pattern of the Swift regime returning in a new form, now operating on timeliness rather than on common-law duties.

York compressed this reasoning into a description of the federal court’s role. When state law supplies the right and diversity alone brings the case to federal court, the court is “for that purpose, in effect, only another court of the State.”¹⁹ The two judicial systems are not identical, so the forms and modes of enforcing a right may vary between them, but the federal court “cannot afford recovery if the right to recover is made unavailable by the State nor can it substantially affect the enforcement of the right as given by the State.”²⁰ A state that defines a right also decides when the right expires, and a federal court that borrowed the right without its expiration date would enforce something the state never created.

Applied to the noteholder’s suit, the test gave a clear answer. York’s class action sat on the equity side of the federal court, where laches would ask whether delay had prejudiced the defendant and might have let the suit go forward, while the New York statute, if it applied, would end the suit by the calendar alone. If a plea of the statute of limitations would bar recovery in a state court, a federal court ought not to afford recovery.²¹ The Court reversed and remanded.²² Statutes of limitation therefore count as substantive for Erie purposes, whatever other courts call them, because disregarding one significantly affects the result.²³

The details of a state’s scheme travel with the rule. That a period may be lengthened or shortened, that a security may be foreclosed though the debt is barred, and that a barred debt may serve as a set-off are all matters of local law that a federal court sitting in the state respects when they come into play.²⁴

York rests on policy and does not claim a constitutional command. One commentator notes that Justice Frankfurter conspicuously avoided saying that the Constitution compelled the result, and that Congress surely has the constitutional power to enact a comprehensive limitations statute for federal cases, diversity cases included.²⁵ The Court’s starting point was the policy of federal jurisdiction that Erie embodies.²⁶ That policy proved broader than Erie itself: it requires federal courts to follow state practices even of “form and mode” if ignoring them would substantially affect the outcome.²⁷ A student who writes that York constitutionally required the state limitations period therefore credits the case with a ground it never claimed. And because policy is the source of the rule, Erie applies only where no controlling federal statute exists, and a Federal Rule that applies takes precedence over state policy.²⁸

The test does not require a federal court to copy every feature of state-court relief, and York said so. A federal court of equity may afford an equitable remedy for a state-recognized right even though a state court cannot give it, but a state cannot dictate the remedies a federal court must give merely because the federal court is an alternative tribunal to its own courts.²⁹ Federal equity remains bound by four restrictions the Court listed, each a rule of federal law that a state’s more permissive practice cannot lift.³⁰

• The suit must fall within the traditional scope of equity as it developed in the English Court of Chancery.
• A plain, adequate, and complete remedy at law must be lacking.
• Explicit congressional curtailment of equity powers must be respected.
• The constitutional right to trial by jury cannot be evaded.

The carve-out is narrow. It concerns the remedy a federal equity court may give for a right the state recognizes, and it leaves untouched the state’s rules on whether the right exists and whether suit came in time. Federal courts may also use procedural devices that statutes or the Federal Rules make available, but beyond the situations the Court listed, a diversity court has little room to build an independent law of remedies in equity.³¹

Stated as York stated it, the test is broad. It asks whether any state rule, once ignored, could change a result, and nothing in the question confines it to limitations periods, burdens of proof, or other rules that touch the right itself. It is hard to conceive of a rule of procedure that could not significantly affect the outcome of some case.³² This subsection has taught the test in its widest form; the next shows what happened when courts applied it literally, and why the Court had to cut it back.

Why the Test Overreached

The York test grew out of a promise of sameness. Whenever a federal court sits solely because the parties are citizens of different states, York said, “the outcome of the litigation in the federal court should be substantially the same, so far as legal rules determine the outcome of a litigation, as it would be if tried in a State court.”³³ The Court focused on the purpose of Erie, which was to avoid the potential of state and federal courts reaching different outcomes,³⁴ and it began from the policy of federal jurisdiction that Erie embodies.³⁵ In York itself the promise gave a sensible answer: the state limitations bar would end the noteholder’s claim outright, while federal laches might have let it continue. The difficulty is how much the promise covers. It is hard to conceive of any rule of procedure that cannot significantly affect the outcome of a case, and to a principle that sweeps so widely there is no apparent stopping place.³⁶

The phrase “so far as legal rules determine the outcome” reaches every rule capable of deciding a case.³⁷ Hindsight widens the reach, because the party who loses can always point to some rule that, had it been different, might have changed the result. Suppose a state rule gives fifteen days to answer and the federal rule gives twenty: a party who answers on the seventeenth day meets one deadline and misses the other. Yet no one argues that a diversity court must apply state law to every issue that arises, such as the form of pleadings or the time for filing a notice of appeal. Some questions seem naturally governed by rules devised for the tribunal that is hearing the case.³⁸

The breadth of the test mattered most because of when the Federal Rules of Civil Procedure arrived. The Court adopted the Rules on December 20, 1937, and they took effect on September 1, 1938, while Erie was decided on April 25, 1938, so the two developments landed within months of each other.³⁹ Before 1938 federal courts had conformed to state procedure under the Conformity Act while applying their own general law under Swift, and the new year reversed that pattern.⁴⁰ The Rules ushered in federal procedural uniformity after a century and a half of mixed state and federal practice,⁴¹ and their drafters expected them to apply to all disputes in the federal system.⁴² If state rules had to prevail over Federal Rules whenever the difference could affect the result, federal courts would run two procedural systems: the Federal Rules in federal question and admiralty cases, and the forum state’s procedure in diversity cases.⁴³ York’s reach did not depend on state law being constitutionally required. Under York, federal courts follow outcome-determinative state law even on procedural issues where federal authority exists to make a rule of their own, so the policy alone had to supply any limit.⁴⁴

The danger was visible when York was decided. Justice Rutledge, dissenting, pointed to the borderland where remedy and substance are so integrated that a clean break cannot be made.⁴⁵ If Erie were expanded judicially to the utmost extent in that zone, he warned, the result would be either nullifying Congress’s power to control what the federal courts may do by way of remedies or usurping it.⁴⁶ A test that asks only whether the result could change has no way to notice that such a zone exists. That gap explains why later decisions added federal interests to the analysis.

Three decisions handed down on a single day in 1949 turned the worry into fact, showing a deference to state law in matters that, for other purposes, are clearly procedural.⁴⁷

• Ragan v. Merchants Transfer & Warehouse Co. The plaintiff filed in Kansas within the two-year limitations period, and Federal Rule 3 treats an action as commenced on filing. Kansas law, however, counted an action as commenced only when the defendant was served, and service came after the two years had run.⁴⁸ The Court reasoned that local law creates the cause of action and therefore measures it, so the federal court had to follow the state service requirement, and it dismissed the case despite Rule 3.⁴⁹
• Cohen v. Beneficial Industrial Loan Corp. A federal court hearing a shareholder’s derivative suit had to apply a New Jersey statute that made an unsuccessful plaintiff liable for the defendants’ expenses, including attorney’s fees, and required security for them as a condition of suing.⁵⁰
• Woods v. Interstate Realty Co. A Mississippi rule barred a foreign corporation that had not appointed an agent for service of process from suing in the state’s courts, and the same bar applied to a diversity action in federal court.⁵¹

The reaction from those who drafted the Rules was sharp. Judge Charles E. Clark, Reporter to the committee that wrote them, observed that after these decisions “hardly a one of the heralded Federal Rules can be considered safe from attack by shrewd lawyers and obedient lower tribunals.”⁵² The alarm had a basis. Applied to the Federal Rules, the outcome-determinative test let a state rule displace a federal one,⁵³ and none of the opinions in the three cases, dissents included, so much as mentioned the Rules Enabling Act.⁵⁴ That turned supremacy on its head, because a validly promulgated Federal Rule ordinarily controls over a conflicting state rule, as Hanna later confirmed in applying Rule 4(d).⁵⁵ Many observers concluded that little room remained for independent federal regulation of procedure.⁵⁶

Setting Ragan beside the notice-of-appeal deadline shows what the literal test cannot do. Looking back from the end of each case, both rules could decide the result: the Ragan plaintiff served after the two years had run, and a party who files notice of appeal after a state deadline but within the federal one would lose the appeal if state law governed. Yet the rules play different roles. The Kansas service requirement was an integral part of the state’s statute of limitations, so it fixed when the state-created claim expired,⁵⁷ while a notice-of-appeal deadline sets the pace of a case already under way. An inquiry that asks only whether the result could change treats the two as equally decisive and cannot register the difference. Byrd later supplied a tool for drawing it, by asking whether a state rule is “bound up with the definition of the rights and obligations of the parties,” a question the next section develops.⁵⁸

Later accounts took much of the sting out of the 1949 trilogy. On one reading, the decisions did not invalidate Rule 3 or the derivative-suit rule; they required litigants to satisfy the state requirements in addition to the Rule.⁵⁹ Hanna v. Plumer offered a related explanation: each case turned on the scope of the Federal Rule, which was narrower than the losing party urged, so no Rule covered the point and Erie directed the court to state law.⁶⁰ Walker v. Armco Steel Corp. later addressed the service-and-limitations problem by asking whether the federal rule actually purports to apply in the case, the direct-collision inquiry taken up in the full analysis below.⁶¹ Those reinterpretations came after the fact. When the trilogy was decided, the opinions gave litigants no reason to expect that any Federal Rule would survive a state rule that could change the result.

The limits arrived in stages. York’s own wording supplied the first, because the test asks whether disregarding state law would “significantly affect the result of a litigation.”⁶² That wording requires a real difference in outcome and not a theoretical one.⁶³ The Court grounded the test in Erie’s purpose of avoiding different outcomes in state and federal courts,⁶⁴ and Byrd v. Blue Ridge Rural Electric Cooperative supplied the second limit. Byrd demoted outcome determination from the governing test to one of three factors, alongside the strength of the state’s interest in having its rule apply and the strength of the federal interest in having federal law control.⁶⁵ Hanna supplied the third by tying the question to Erie’s twin aims, discouraging forum shopping and avoiding inequitable administration of the laws, a step the Hanna section develops.⁶⁶ Hanna also confirmed that a valid Federal Rule on point applies even against a conflicting state rule,⁶⁷ and it left York’s own result undisturbed, since the Court continues to require resort to state law when a claim is barred under state law.⁶⁸