Concept
Whose Preclusion Law Applies
Claim preclusion
Whose Preclusion Law Applies
A judgment pleaded in a second lawsuit raises a question that purely local litigation never presents: whose preclusion rules decide what the first judgment settled? The court that decided case one is the rendering court, and the court asked to honor its judgment is the second court. When both belong to one system, the second court applies the rules it already knows. The difficulty begins when case two is filed in a different system from case one, as when a state judgment is pleaded in federal court, a federal judgment in a state court, or one state’s judgment in the courts of another. Congress answered part of that question for state and territorial judgments in 28 U.S.C. § 1738.1 By its terms the statute names only the courts of a State, Territory, or Possession, and it says nothing about judgments of federal courts.2
The answer matters because preclusion law varies from one jurisdiction to another. Jurisdictions differ on how they define a claim, on who counts as a privy, on whether an omitted counterclaim is lost, and on which dismissals count as on the merits. Semtek shows the last difference at work. A federal court sitting in diversity had dismissed a claim as barred by the statute of limitations, and the Supreme Court held that Rule 41(b)’s statement that a dismissal operates as an adjudication on the merits does not itself decide whether the judgment is on the merits for claim-preclusion purposes, because jurisdictions may treat limitations dismissals differently.3 The same two lawsuits can therefore be barred under one body of preclusion law and open under another, and the body the second court applies decides the outcome. If every second court applied its own law, a judgment would be worth different amounts in different places, and a loser could hunt for the forum with the narrowest rule. That hunt would defeat the repose the doctrine exists to give the winner.
Three scenarios follow from the identity of the rendering court: a first judgment from a state court, from a federal court sitting in diversity, and from a federal court deciding a federal question. In each, the second court asks the same underlying question: what effect would the rendering court’s judgment have under the law that governs it? The scenarios differ only in where that law comes from. Section 1738 supplies the answer for the first. Because the statute speaks only of state and territorial courts, the two federal scenarios need another source, and Semtek frames the diversity case. Throughout, keep one question in view: what preclusion law would the rendering court apply? The state-court scenario comes first, because the statute states its rule most directly and the federal scenarios are easiest to see against it.
State Judgments and § 1738
The governing text is 28 U.S.C. § 1738, which gives a state court’s judgment “the same full faith and credit in every court within the United States” that it has “by law or usage” in the courts of the state from which it is taken.4 The statute carries out the Full Faith and Credit Clause, which leaves to Congress the power to prescribe the effect of a state’s records and judicial proceedings, and Congress chose to make a state judgment worth the same in every other court as it is worth at home.5 A state judgment therefore travels with the preclusion law of the state that rendered it. The earlier sections of this chapter explained what a judgment settles; this section identifies the system whose answer controls.
In practice the court in case two asks what effect the judgment would have in the courts of the rendering state and gives it exactly that effect. The rendering state supplies the content of the answer, whether it defines a claim by the transaction, by the same evidence, or by primary rights, but the command that this content governs elsewhere comes from federal law.6 That division of labor has a consequence for the states themselves. No state can independently decide how far its judgments reach into other states, so a state statute declaring that its judgments bar more broadly in sister-state courts than they bar at home would have no effect, because the interstate reach of a state judgment is fixed by § 1738 and only indirectly by the state’s own rules.7
The same rule governs whether case two is filed in a state court or a federal one. The Clause by its terms commands the States, not the federal government, and § 1738 closes the gap by extending the command to every court within the United States, federal courts included.8 A federal court hearing case two therefore asks what a sister-state court would ask and takes its answer from the same place, the law of the court that decided case one.
Three reasons support the rule. The first is certainty. When the first suit is filed, no one knows where a later suit might be brought, but because every second court looks back to the rendering court’s rules, the parties know, as soon as the judgment is rendered, how far it will bar further litigation. That knowledge bears on the plaintiff’s choice of forum, on the plaintiff’s willingness to sue at all, and on the defendant’s strategy inside case one.9 The second is forum shopping. A loser gains nothing by bringing the second action in a state with less restrictive res judicata rules, since all courts look back to the rules of the rendering state.10 The third explains why the rule can be so firm: errors in the first judgment can be corrected by appeal within the rendering state’s own system, so a second court has no need to second-guess the first. Certainty carries most of the weight, and the availability of appeal makes the rule tolerable, since a litigant who thinks the first court erred already had the remedy.
An exam stem that carries a case across a state line asks one question: which state’s court rendered case one?
The forum and the plaintiff’s choice of forum decided nothing in either scene; the rendering court decided everything. The instinct that loses points is to apply the forum’s own rule, which would leave the first suit open and bar the second, the reverse of the correct results. The Tellis scene is the one that feels wrong, because a West Dakota judge must permit a suit that West Dakota law would forbid had its own courts decided case one. The result follows from the design of the rule: the East Dakota judgment never carried a broader reach than East Dakota gave it.11
The command binds a federal court in case two with equal force, even when the second suit rests on federal law. In Frier, a federal court faced a § 1983 suit after Frier had lost state replevin actions over his towed cars. The court of appeals looked to Illinois law, which under § 1738 governed the preclusive effect of the replevin judgments, and concluded that Illinois would bar the suit.12 In Gargallo v. Merrill Lynch, Pierce, Fenner & Smith, Inc., an Ohio court had dismissed Gargallo’s counterclaim with prejudice for noncompliance with a discovery rule. When he later asserted federal securities claims in federal court, the court of appeals asked whether Ohio claim preclusion law would bar the claim had the suit been filed in an Ohio court.13 Both courts applied what the Frier majority called the generally accepted rule: look to the jurisdiction that rendered the first judgment.14 The rule holds even when the later claim arises under federal law, although whether an exception applies in that setting depends on congressional intent, as Marrese v. American Academy of Orthopaedic Surgeons held.15 A student who sees a federal court in case two and reaches for federal preclusion law has made the usual mistake, because federal preclusion law matters only when the judgment came from a federal court, the subject of the next two subsections.
The borrowing covers every variable the earlier sections taught, not only the definition of a claim. If the rendering state counts a default judgment, a two-dismissal rule, or an omitted compulsory counterclaim as a bar, a sister-state court must treat the judgment as on the merits and give it that effect even though the claim was never actually litigated.16 The rendering state’s rules on finality and on who counts as a privy travel the same way. A court asked to honor a state judgment may not replace those rules with its own, because the parties litigated in reliance on the rendering court’s law.
The command is demanding, and Baker v. General Motors Corp. marks its edges by separating the credit owed a sister state’s laws from the credit owed its judgments. A forum may be guided by its own public policy when it decides which state’s law governs a controversy, but as to judgments the full faith and credit obligation is “exacting”: a final judgment rendered by a court with adjudicatory authority over the subject matter and the persons bound gains nationwide force for claim and issue preclusion.17 The Court recalled that a Missouri judgment had to be honored in Mississippi even if it rested on a misapprehension of Mississippi law.18 It found no roving “public policy exception” to the credit due a judgment, so a forum’s disagreement with the first court’s result, or its contrary policy, is no ground for refusing the judgment its effect.19 The distinction makes structural sense. A forum with a contrary policy usually has some local interest that justifies applying its own law to a dispute, but it can resist a judgment only by revisiting what the rendering court decided, and the rendering state’s appeals already exist to correct mistakes of that kind.
Firm does not mean unlimited. A court may decline in some circumstances to enforce a sister-state judgment, and Baker itself marks two limits.20
The first limit concerns reach. Elwell and General Motors submitted an agreed injunction to a Michigan court, and the presiding judge signed it. That order was claim preclusive between them: Elwell’s wrongful-discharge, contract, and tort claims had “merged in the judgment,” and he could not sue again to recover more, although no issue was ever joined, litigated, or determined in Michigan.21 The judgment could not reach further. General Motors relied on the injunction to keep Elwell from testifying in a wrongful-death action brought by the Bakers, and the Court held that the Michigan judgment could not control proceedings brought by strangers to it or dictate that evidence in another jurisdiction’s courts be inadmissible. This is consistent with the rule on nonparties earlier in this chapter: a judgment carries its effect between the parties and their privies, and it cannot bind a person Michigan had no power over.
The second limit separates preclusive effect from enforcement. Full faith and credit does not require a state to adopt another state’s practices on the time, manner, and mechanisms for enforcing judgments, because enforcement measures do not travel with a sister-state judgment the way preclusive effects do and remain subject to the forum’s own law.22 A sister state’s decree about land in another state, for example, cannot transfer title there, although it may still settle the rights and obligations between the parties to that litigation.23 The two questions differ in kind. Preclusion asks what has already been decided between these parties, which only the rendering system could have answered. Enforcement asks how a particular court’s officers and processes will be used to collect or compel, and the forum controls that evenhandedly. A judgment creditor who carries a State X judgment to State Y therefore finds the claim conclusively resolved there and must still collect by State Y’s methods.
Federal Diversity Judgments
In Semtek International Inc. v. Lockheed Martin Corp. the Supreme Court held that federal common law, meaning rules the federal courts develop for themselves when no statute or constitutional text supplies one, governs the claim-preclusive effect of a dismissal by a federal court sitting in diversity, and it adopted as the content of that rule the law that state courts in the state where the federal court sits would apply.25 The holding has two parts that are easy to blur. The source of the rule is federal, because the Supreme Court has the last word on what any federal judgment is worth.26 The content of the rule is borrowed from the state, because the Court judged that reference to the rendering court’s home state is the right federal answer when the case rests on state substantive law. The preceding subsection showed that a state judgment is worth in every court what the rendering state’s law says it is worth; this subsection answers the parallel question when the first judgment came from a federal court.
The Court reached that answer only after rejecting the two sources the parties could point to. As the earlier discussion of Rule 41(b) showed, the rule’s statement that a dismissal operates as an adjudication on the merits does not itself decide the dismissal’s claim-preclusive effect in other courts, and reading it to do so would arguably violate the jurisdictional limitation of the Rules Enabling Act and in many cases offend Erie’s federalism principle.27 That left a gap, because some body of law still had to say what a California federal judgment means to a Maryland court. The obvious candidates were the Full Faith and Credit Clause and § 1738, but the Court found that neither addresses the question. By their terms they govern only the effects of state-court judgments and, in the statute’s case, judgments of courts of territories and possessions, and no other federal textual provision addresses the claim-preclusive effect of a federal diversity judgment.28 The statute’s text bears that out, because it gives full faith and credit to the records and judicial proceedings of the courts of a State, Territory, or Possession and says nothing about the courts of the United States.29
The absence of a governing text does not leave a federal judgment without a rule. No federal text addresses the effect of a federal-question judgment either, yet the Court has long held that States cannot give those judgments merely whatever effect they would give their own, and must accord them the effect the Court prescribes.30 The Court drew a conclusion for diversity judgments from that line of cases. In Dupasseur v. Rochereau a state was allowed, indeed required, to give a federal diversity judgment no more effect than it would give one of its own, but only because the Supreme Court had deemed reference to state law the appropriate federal rule.31 State law therefore never acquires independent authority over federal judgments. The question is one of federal common law, even though the answer comes from the states.
Having placed the source in federal law, the Court still had to choose the content, and it chose to borrow. The first reason comes from what a diversity case is. The claim arises under state substantive law, so no federal policy calls for a uniform national rule on how far a judgment reaches, and nationwide uniformity is better served when a dismissal carries the same claim-preclusive effect whether a state court or a federal court ordered it.32 The second reason is the forum-shopping concern behind Erie. Any other rule would encourage filing in, or removing to, federal court, because litigants would foresee different effects from the likely grounds of dismissal.33 A rule that mirrors the home state removes that incentive, since the choice between the two courthouses no longer changes what a dismissal accomplishes.
The facts of Semtek show the rule at work. A federal court in California dismissed Semtek’s diversity action on the basis of a California statute of limitations, and Semtek renewed the suit against the same defendant in a Maryland state court, whose limitations period had not yet run.34 The Maryland Court of Special Appeals held that the federal dismissal necessarily precluded the new suit. The Supreme Court reversed, because the governing federal rule incorporates California’s law of claim preclusion and the Maryland courts had to ask what California would do.35 Nothing federal stood in the way of that borrowing. The California federal court dismissed this state-law claim only because a California statute required it, and the Court saw no conceivable federal interest in giving that time bar more effect in other courts than California’s own courts would give it.36
The Court was careful about how far it went. It did not decide what California’s law of claim preclusion provides, and it sent the case back for further proceedings.37 Semtek’s position was that a California state court would not give a limitations dismissal claim-preclusive effect.38 If that is right, the Maryland suit may go forward; if California treats such a dismissal as a bar, Maryland must dismiss. Either way, Maryland’s longer limitations period decides nothing about preclusion.
An exam problem can test that point by changing the rendering state.
Miguel’s reply decides the motion. The first judgment came from a federal court sitting in North Carolina in a diversity case, so the federal rule asks what North Carolina’s courts would do with it, and they would bar the refiled claim. Virginia’s limitations period and Virginia’s view of limitations dismissals describe the court hearing case two, which rendered nothing, so the court should grant the motion. The same kind of dismissal and the same refiling in a state with a longer period can come out either way, and what changes the result is the preclusion law of the state where the rendering federal court sat.
The identity of the second court does not change the analysis. When case two is in state court, as in Semtek, the state court applies the federal common-law rule, which for a diversity judgment points to the preclusion law of the state where the first federal court sat. When case two is in federal court, the effect of the diversity judgment is again decided by the preclusion rules of the state where the rendering court sits.41 Wrong answers go astray at the source. Section 1738 gives no answer for a federal judgment, because its text reaches only the proceedings of state, territorial, and possession courts.42 Rule 41(b)’s on-the-merits default produces no nationwide bar either, since that premise is the one Semtek rejected.43 Nor does the law of the state where case two is pending supply the rule, since that state’s courts did not render the judgment.
The borrowing has a limit, which the Court stated in the same passage. The reference to state law does not apply where state law is incompatible with federal interests. The Court’s example is a dismissal for willful violation of discovery orders: if state law gave such a dismissal no claim-preclusive effect, the federal courts’ interest in the integrity of their own processes might justify a contrary federal rule.44 The reason follows from what a sanction is for. A discovery sanction works by making a litigant fear the consequence of defying the court, and a dismissal that the sanctioned plaintiff could undo by refiling in a court that gives it no preclusive weight would leave the court’s orders without force. The interest belongs to the federal court, because its own orders and procedures produced that dismissal. A limitations dismissal sits at the other end, since the time bar is a creature of California law and the federal order added nothing federal to it. The Court said only that such an interest might justify a departure, and it applied no such departure in Semtek, so an answer should present the limit as a recognized qualification without a developed test behind it.45 When case one was a federal-question judgment, no borrowing of state law occurs, and the next subsection takes up that rule.
Federal-Question Judgments
The full faith and credit statute reaches only the records and judicial proceedings of the courts of a State, Territory, or Possession, so a federal-court judgment falls outside its text.47 What a federal-question judgment is worth in a later suit must therefore come from somewhere else. Semtek observed that no federal textual provision addresses the claim-preclusive effect of a federal-court judgment in a federal-question case, and it recalled that the Court has long held that States cannot give those judgments merely whatever effect they would give their own, but must accord them the effect the Court prescribes.48 Uniform federal common law governs, and the Supreme Court has the last word on its content.
The reason for a national rule here, where the diversity case borrowed a state rule, lies in whose substantive law the first court applied. In a diversity case the claim arises under state law, and the Court reasoned that there is then no need for a uniform federal rule, so it adopted the preclusion law of the state where the federal court sits.49 A federal-question case reverses that premise, because the claim arises under federal law and the federal interest in a single answer is at its strongest. Two reasons follow. The first is uniformity. If each state’s preclusion rules fixed what a federal judgment settles, a federal court’s rejection of a § 1983 claim would end the dispute in a state that defines a claim broadly and leave it open in a state that defines it narrowly, so the same federal right would be enforced unevenly and a losing plaintiff could shop for the friendliest state. The second concerns authority. A state that wished to blunt a federal judgment could adopt a stingy preclusion rule and apply it in its own courts, which is why the Court treats the question whether a state court gave a federal judgment its due force as a federal question that it decides for itself.50 Uniformity explains why the rule is national, and authority explains who has the last word about it.
The content of the federal rule will look familiar from earlier in this chapter. The federal common law standard is virtually identical to the version of claim preclusion prevalent among the states: it asks for a final judgment on the merits in the earlier action, an identity of the cause of action in the two suits, and an identity of parties or privies.51 For causes of action arising under federal law, the federal courts define the same claim by the same-transaction test.52 What changes in this setting is the source of the rule, and with it the court that decides what the rule means. The federal common law of preclusion is also subject to due process limits.53 The nonparty material earlier in this chapter illustrates the point. In Taylor v. Sturgell the first suit was Herrick’s Freedom of Information Act action, a federal-question case, so the Court treated the effect of that judgment as governed by a uniform federal rule that it has ultimate authority to declare, and the categories of nonparty preclusion it catalogued belong to that rule.54
River Park, Inc. v. City of Highland Park shows what the rule means for a later state-court suit. The plaintiffs there lost a federal-question suit in a federal district court and then sued the city again in an Illinois court. The Illinois Supreme Court decided whether the two suits were one claim by choosing among the definitions available under Illinois law, yet under existing precedents it should have looked to federal law for that definition, because a federal district court had entered the original judgment on a federal cause of action and not under its diversity jurisdiction.55 The detour probably cost little in that case, because Illinois chose the transactional test that the federal courts already use. A state that kept the same-evidence or primary-rights definition would have to put it aside, since the federal definition travels with the federal-question judgment into the state courthouse.
The hinge in the two scenes is the basis of federal jurisdiction in case one. The second court, the second suit, and the firing are the same in both, and only a federal-question first judgment brings the federal definition of a claim along. A diversity first judgment brings the law of the state where the federal court sat, which here is State X’s own narrower rule. An exam stem that places a federal court in case one and a state court in case two therefore turns on one question: did the federal court’s authority rest on a federal claim or on diversity?
The three scenarios in this section sort by the court that rendered case one:
| Case one was decided by | Preclusion law that governs its effect |
|---|---|
| A state or territorial court | The law of the rendering state, through § 1738 |
| A federal court sitting in diversity | Federal common law, borrowing the preclusion law of the state where the rendering federal court sits |
| A federal court deciding a federal question | Uniform federal common law |
A federal judgment carries its rules into a later suit in either system. Federal preclusion principles govern, except that the effect of a federal judgment based on diversity is again decided by the preclusion rules of the state in which the rendering court sits. The question that keeps all three straight is the one this section began with: what preclusion law would the rendering court apply?56