Taylor v. Sturgell, Acting Administrator, Fed. Aviation Administration, 553 U.S. 880 (2008)

Justice Ginsburg

delivered the opinion of the Court.

“It is a principle of general application in Anglo-American jurisprudence that one is not bound by a judgment in personam in a litigation in which he is not designated as a party or to which he has not been made a party by service of process.” Hansberry v. Lee, 311 U. S. 32,40 (1940). Several exceptions, recognized in this Court’s decisions, temper this basic rule. In a class action, for example, a person not named as a party may be bound by a judgment on the merits of the action, if she was adequately represented by a party who actively participated in the litigation. See id., at 41. In this case, we consider for the first time whether there is a “virtual representation” exception to the general rule against precluding nonparties. Adopted by a number of courts, including the courts below in the case now before us, the exception so styled is broader than any we have so far approved.

The virtual representation question we examine in this opinion arises in the following context. Petitioner Brent Taylor filed a lawsuit under the Freedom of Information Act seeking certain documents from the Federal Aviation Administration. Greg Herrick, Taylor’s friend, had previously brought an unsuccessful suit seeking the same records. The two men have no legal relationship, and there is no evidence that Taylor controlled, financed, participated in, or even had notice of Herrick’s earlier suit. Nevertheless, the D. C. Circuit held Taylor’s suit precluded by the judgment against Herrick because, in that court’s assessment, Herrick qualified as Taylor’s “virtual representative.”

We disapprove the doctrine of preclusion by “virtual representation,” and hold, based on the record as it now stands, that the judgment against Herrick does not bar Taylor from maintaining this suit.

The Freedom of Information Act (FOIA or Act) accords “any person” a right to request any records held by a federal agency. 5 U. S. C. § 552(a)(3)(A) (2006 ed.). No reason need be given for a FOIA request, and unless the requested materials fall within one of the Act’s enumerated exemptions, see § 552(a)(3)(E), (b), the agency must “make the records promptly available” to the requester, § 552(a)(3)(A). If an agency refuses to furnish the requested records, the requester may file suit in federal court and obtain an injunction “order[ing] the production of any agency records improperly withheld.” § 552(a)(4)(B).

I

The courts below held the instant FOIA suit barred by the judgment in earlier litigation seeking the same records. Because the lower courts’ decisions turned on the connection between the two lawsuits, we begin with a full account of each action.

A

The first suit was filed by Greg Herrick, an antique aircraft enthusiast and the owner of an F-45 airplane, a vintage

model manufactured by the Fairchild Engine and Airplane Corporation (FEAC) in the 1930’s. In 1997, seeking information that would help him restore his plane to its original condition, Herrick filed a FOIA request asking the Federal Aviation Administration (FAA) for copies of any technical documents about the F-45 contained in the agency’s records.

To gain a certificate authorizing the manufacture and sale of the F-45, FEAC had submitted to the FAA’s predecessor, the Civil Aeronautics Authority, detailed specifications and other technical data about the plane. Hundreds of pages of documents produced by FEAC in the certification process remain in the FAA’s records. The FAA denied Herrick’s request, however, upon finding that the documents he sought are subject to FOIA’s exemption for “trade secrets and commercial or financial information obtained from a person and privileged or confidential,” § 552(b)(4). In an administrative appeal, Herrick urged that FEAC and its successors had waived any trade-secret protection. The FAA thereupon contacted FEAC’s corporate successor, respondent Fairchild Corporation (Fairchild). Because Fairchild objected to release of the documents, the agency adhered to its original decision.

Herrick then filed suit in the U. S. District Court for the District of Wyoming. Challenging the FAA’s invocation of the trade-secret exemption, Herrick placed heavy weight on a 1955 letter from FEAC to the Civil Aeronautics Authority. The letter authorized the agency to lend any documents in its files to the public “for use in making repairs or replacement parts for aircraft produced by Fairchild.” Herrick v. Garvey, 298 F. 3d 1184,1193 (CA10 2002) (internal quotation marks omitted). This broad authorization, Herrick maintained, showed that the F-45 certification records held by the FAA could not be regarded as “secre[t]” or “confidential” within the meaning of § 552(b)(4).

Rejecting Herrick’s argument, the District Court granted summary judgment to the FAA. Herrick v. Garvey, 200

F. Supp. 2d 1321, 1328-1329 (Wyo. 2000). The 1955 letter, the court reasoned, did not deprive the F-45 certification documents of trade-secret status, for those documents were never in fact released pursuant to the letter’s blanket authorization. See id., at 1329. The court also stated that even if the 1955 letter had waived trade-secret protection, Fairchild had successfully “reversed” the waiver by objecting to the FAA’s release of the records to Herrick. Ibid.

On appeal, the Tenth Circuit agreed with Herrick that the 1955 letter had stripped the requested documents of trade-secret protection. See Herrick, 298 F. 3d, at 1194. But the Court of Appeals upheld the District Court’s alternative determination — i. e., that Fairchild had restored trade-secret status by objecting to Herrick’s FOIA request. Id., at 1195. On that ground, the appeals court affirmed the entry of summary judgment for the FAA.

In so ruling, the Tenth Circuit noted that Herrick had failed to challenge two suppositions underlying the District Court’s decision. First, the District Court assumed trade-secret status could be “restored” to documents that had lost protection. Id., at 1194, n. 10. Second, the District Court also assumed that Fairchild had regained trade-secret status for the documents even though the company claimed that status only “after Herrick had initiated his request” for the F-45 records. Ibid. The Court of Appeals expressed no opinion on the validity of these suppositions. See id., at 1194-1195, n. 10.

B

The Tenth Circuit’s decision issued on July 24,2002. Less than a month later, on August 22, petitioner Brent Taylor— a friend of Herrick’s and an antique aircraft enthusiast in his own right — submitted a FOIA request seeking the same documents Herrick had unsuccessfully sued to obtain. When the FAA failed to respond, Taylor filed a complaint in the U. S. District Court for the District of Columbia. Like Herrick, Taylor argued that FEAC’s 1955 letter had stripped

the records of their trade-secret status. But Taylor also sought to litigate the two issues concerning recapture of protected status that Herrick had failed to raise in his appeal to the Tenth Circuit.

After Fairchild intervened as a defendant,1 the District Court in D. C. concluded that Taylor’s suit was barred by claim preclusion; accordingly, it granted summary judgment to Fairchild and the FAA. The court acknowledged that Taylor was not a party to Herrick’s suit. Relying on the Eighth Circuit’s decision in Tyus v. Schoemehl, 93 F. 3d 449 (1996), however, it held that a nonparty may be bound by a judgment if she was “virtually represented” by a party. App. to Pet. for Cert. 30a-31a.

The Eighth Circuit’s seven-factor test for virtual representation, adopted by the District Court in Taylor’s case, requires an “identity of interests” between the person to be bound and a party to the judgment. See id., at 31a. See also Tyus, 93 F. 3d, at 455. Six additional factors counsel in favor of virtual representation under the Eighth Circuit’s test, but are not prerequisites: (1) a “close relationship” between the present party and a party to the judgment alleged to be preclusive; (2) “participation in the prior litigation” by the present party; (3) the present party’s “apparent acquiescence” to the preclusive effect of the judgment; (4) “deliberate] maneuvering]” to avoid the effect of the judgment; (5) adequate representation of the present party by a party to the prior adjudication; and (6) a suit raising a “public law” rather than a “private law” issue. App. to Pet. for Cert. 31a (citing Tyus, 93 F. 3d, at 454-456). These factors, the D. C. District Court observed, “constitute a fluid test with imprecise boundaries” and call for “a broad, case-by-case inquiry.” App. to Pet. for Cert. 32a.

The record before the District Court in Taylor’s suit revealed the following facts about the relationship between Taylor and Herrick: Taylor is the president of the Antique Aircraft Association, an organization to which Herrick belongs; the two men are “close associate^],” App. 54; Herrick asked Taylor to help restore Herrick’s F-45, though they had no contract or agreement for Taylor’s participation in the restoration; Taylor was represented by the lawyer who represented Herrick in the earlier litigation; and Herrick apparently gave Taylor documents that Herrick had obtained from the FAA during discovery in his suit.

Fairchild and the FAA conceded that Taylor had not participated in Herrick’s suit. App. to Pet. for Cert. 32a. The D. C. District Court determined, however, that Herrick ranked as Taylor’s virtual representative because the facts fit each of the other six indicators on the Eighth Circuit’s list. See id., at 32a-35a. Accordingly, the District Court held Taylor’s suit, seeking the same documents Herrick had requested, barred by the judgment against Herrick. See id., at 35a.

The D. C. Circuit affirmed. It observed, first, that other Circuits “vary widely” in their approaches to virtual representation. Taylor v. Blakey, 490 F. 3d 965, 971 (2007). In this regard, the D. C. Circuit contrasted the multifactor balancing test applied by the Eighth Circuit and the D. C. District Court with the Fourth Circuit’s narrower approach, which “treats a party as a virtual representative only if the party is ‘accountable to the nonparties who file a subsequent suit’ and has ‘the tacit approval of the court’ to act on the nonpart[ies’] behalf.” Ibid, (quoting Klugh v. United States, 818 F. 2d 294, 300 (CA4 1987)).

Rejecting both of these approaches, the D. C. Circuit announced its own five-factor test. The first two factors— “identity of interests” and “adequate representation” — are necessary but not sufficient for virtual representation. 490 F. 3d, at 971-972. In addition, at least one of three other

factors must be established: “a close relationship between the present party and his putative representative,” “substantial participation by the present party in the first case,” or “tactical maneuvering on the part of the present party to avoid preclusion by the prior judgment.” Id., at 972.

Applying this test to the record in Taylor’s case, the D. C. Circuit found both of the necessary conditions for virtual representation well met. As to identity of interests, the court emphasized that Taylor and Herrick sought the same result — release of the F-45 documents. Moreover, the D. C. Circuit observed, Herrick owned an F-45 airplane, and therefore had, “if anything, a stronger incentive to litigate” than Taylor, who had only a “general interest in public disclosure and the preservation of antique aircraft heritage.” Id., at 973 (internal quotation marks omitted).

Turning to adequacy of representation, the D. C. Circuit acknowledged that some other Circuits regard notice of a prior suit as essential to a determination that a nonparty was adequately represented in that suit. See id., at 973-974 (citing Perez v. Volvo Car Corp., 247 F. 3d 303, 312 (CA1 2001), and Tice v. American Airlines, Inc., 162 F. 3d 966, 973 (CA7 1998)). Disagreeing with these courts, the D. C. Circuit deemed notice an “important” but not an indispensable element in the adequacy inquiry. The court then concluded that Herrick had adequately represented Taylor even though Taylor had received no notice of Herrick’s suit. For this conclusion, the appeals court relied on Herrick’s “strong incentive to litigate” and Taylor’s later engagement of the same attorney, which indicated to the court Taylor’s satisfaction with that attorney’s performance in Herrick’s case. See 490 F. 3d, at 974-975.

The D. C. Circuit also found its “close relationship” criterion met, for Herrick had “asked Taylor to assist him in restoring his F-45” and “provided information to Taylor that Herrick had obtained through discovery”; furthermore, Taylor “did not oppose Fairchild’s characterization of Herrick

as his ‘close associate.’” Id., at 975. Because the three above-described factors sufficed-to establish virtual representation under the D. C. Circuit’s five-factor test, the appeals court left open the question whether Taylor had engaged in “tactical maneuvering.” See id., at 976 (calling the facts bearing on tactical maneuvering “ambigu[ous]”).2

We granted certiorari, 552 U. S. 1136 (2008), to resolve the disagreement among the Circuits over the permissibility and scope of preclusion based on “virtual representation.”3

II

The preclusive effect of a federal-court judgment is determined by federal common law. See Semtek Int’l Inc. v. Lockheed Martin Corp., 531 U. S. 497, 507-508 (2001). For judgments in federal-question cases — for example, Herrick’s FOIA suit — federal courts participate in developing “uniform federal rule[s]” of res judicata, which this Court has ultimate authority to determine and declare. Id., at 508.4 The federal common law of preclusion is, of course, subject to due process limitations. See Richards v. Jefferson County, 517 U. S. 793, 797 (1996).

Taylor’s case presents an issue of first impression in this sense: Until now, we have never addressed the doctrine of “virtual representation” adopted (in varying forms) by several Circuits and relied upon by the courts below. Our inquiry, however, is guided by well-established precedent regarding the propriety of nonparty preclusion. We review that precedent before taking up directly the issue of virtual representation.

A

The preclusive effect of a judgment is defined by claim preclusion and issue preclusion, which are collectively referred to as “res judicata.”5 Under the doctrine of claim preclusion, a final judgment forecloses “successive litigation of the very same claim, whether or not relitigation of the claim raises the same issues as the earlier suit.” New Hampshire v. Maine, 532 U. S. 742, 748 (2001). Issue preclusion, in contrast, bars “successive litigation of an issue of fact or law actually litigated and resolved in a valid court determination essential to the prior judgment,” even if the issue recurs in the context of a different claim. Id., at 748-749. By “precluding] parties from contesting matters that they have had a full and fair opportunity to litigate,” these two doctrines protect against “the expense and vexation attending multiple lawsuits, conserv[e] judicial resources, and foste[r] reliance on judicial action by minimizing the possibility of inconsistent decisions.” Montana v. United States, 440 U. S. 147, 153-154 (1979).

A person who was not a party to a suit generally has not had a “full and fair opportunity to litigate” the claims and issues settled in that suit. The application of claim and issue preclusion to nonparties thus runs up against the “deep-

rooted historic tradition that everyone should have his own day in court.” Rickards, 517 U. S., at 798 (internal quotation marks omitted). Indicating the strength of that tradition, we have often repeated the general rule that “one is not bound by a judgment in personam in a litigation in which he is not designated as a party or to which he has not been made a party by service of process.” Hansberry, 311 U. S., at 40. See also, e. g., Richards, 517 U. S., at 798; Martin v. Wilks, 490 U. S. 755, 761 (1989); Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U. S. 100, 110 (1969).

B

Though hardly in doubt, the rule against nonparty preclusion is subject to exceptions. For present purposes, the recognized exceptions can be grouped into six categories.6

First, “[a] person who agrees to be bound by the determination of issues in an action between others is bound in accordance with the terms of his agreement.” 1 Restatement (Second) of Judgments §40, p. 390 (1980) (hereinafter Restatement). For example, “if separate actions involving the same transaction are brought by different plaintiffs against the same defendant, all the parties to all the actions may agree that the question of the defendant’s liability will be definitely determined, one way or the other, in a ‘test case.’” D. Shapiro, Civil Procedure: Preclusion in Civil Actions 77-78 (2001) (hereinafter Shapiro). See also California v. Texas, 459 U. S. 1096, 1097 (1983) (dismissing certain defendants from a suit based on a stipulation “that each of

said defendants ... will be bound by a final judgment of this Court” on a specified issue).7

Second, nonparty preclusion may be justified based on a variety of pre-existing “substantive legal relationship[s]” between the person to be bound and a party to the judgment. Shapiro 78. See also Richards, 517 U. S., at 798. Qualifying relationships include, but are not limited to, preceding and succeeding owners of property, bailee and bailor, and assignee and assignor. See 2 Restatement §§43-44, 52, 55. These exceptions originated “as much from the needs of property law as from the values of preclusion by judgment.” 18A C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure §4448, p. 329 (2d ed. 2002) (hereinafter Wright & Miller).8

Third, we have confirmed that, “in certain limited circumstances,” a nonparty may be bound by a judgment because she was “adequately represented by someone with the same interests who [wa]s a party” to the suit. Richards, 517 U. S., at 798 (internal quotation marks omitted). Representative suits with preclusive effect on nonparties include properly conducted class actions, see Martin, 490 U. S., at 762, n. 2 (citing Fed. Rule Civ. Proc. 23), and suits brought by trustees, guardians, and other fiduciaries, see Sea-Land Services,

Inc. v. Gaudet, 414 U. S. 573,593 (1974). See also 1 Restatement §41.

Fourth, a nonparty is bound by a judgment if she “assumed] control” over the litigation in which that judgment was rendered. Montana, 440 U. S., at 154. See also Schnell v. Peter Eckrich & Sons, Inc., 365 U. S. 260,262, n. 4 (1961); 1 Restatement § 39. Because such a person has had “the opportunity to present proofs and argument,” he has already “had his day in court” even though he was not a formal party to the litigation. Id., Comment a, at 382.

Fifth, a party bound by a judgment may not avoid its preclusive force by relitigating through a proxy. Preclusion is thus in order when a person who did not participate in a litigation later brings suit as the designated representative of a person who was a party to the prior adjudication. See Chicago, R. I. & P. R. Co. v. Schendel, 270 U. S. 611, 620, 623 (1926); 18A Wright & Miller §4454, at 433-434. And although our decisions have not addressed the issue directly, it also seems clear that preclusion is appropriate when a non-party later brings suit as an agent for a party who is bound by a judgment. See id., §4449, at 335.

Sixth, in certain circumstances a special statutory scheme may “expressly foreclos[e] successive litigation by non-litigants ... if the scheme is otherwise consistent with due process.” Martin, 490 U. S., at 762, n. 2. Examples of such schemes include bankruptcy and probate proceedings, see ibid., and quo warranto actions or other suits that, “under [the governing] law, [may] be brought only on behalf of the public at large,” Richards, 517 U. S., at 804.

Ill

Reaching beyond these six established categories, some lower courts have recognized a “virtual representation” exception to the rule against nonparty preclusion. Decisions of these courts, however, have been far from consistent.

See 18A Wright & Miller § 4457, at 513 (virtual representation lacks a “clear or coherent theory”; decisions applying it have “an episodic quality”). Some Circuits use the label, but define “virtual representation” so that it is no broader than the recognized exception for adequate representation. See, e. g., Becherer v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 193 F. 3d 415, 423,427 (CA6 1999) (en banc). But other courts, including the Eighth, Ninth, and D. C. Circuits, apply multifactor tests for virtual representation that permit non-party preclusion in cases that do not fit within any of the established exceptions. See supra, at 888-891, and n. 3.

The D. C. Circuit, the FAA, and Fairchild have presented three arguments in support of an expansive doctrine of virtual representation. We find none of them persuasive.

A

The D. C. Circuit purported to ground its virtual representation doctrine in this Court’s decisions stating that, in some circumstances, a person may be bound by a judgment if she was adequately represented by a party to the proceeding yielding that judgment. See 490 F. 3d, at 970-971. But the D. C. Circuit’s definition of “adequate representation” strayed from the meaning our decisions have attributed to that term.

In Richards, we reviewed a decision by the Alabama Supreme Court holding that a challenge to a tax was barred by a judgment upholding the same tax in a suit filed by different taxpayers. 517 U. S., at 795-797. The plaintiffs in the first suit “did not sue on behalf of a class,” their complaint “did not purport to assert any claim against or on behalf of any nonparties,” and the judgment “did not purport to bind” non-parties. Id., at 801. There was no indication, we emphasized, that the court in the first suit “took care to protect the interests” of absent parties, or that the parties to that litigation “understood their suit to be on behalf of absent [parties].” Id., at 802. In these circumstances, we held, the ap

plication of claim preclusion was inconsistent with “the due process of law guaranteed by the Fourteenth Amendment.” Id., at 797.

The D. C. Circuit stated, without elaboration, that it did not “read Richards to hold a nonparty ... adequately represented only if special procedures were followed [to protect the nonparty] or the party to the prior suit understood it was representing the nonparty.” 490 F. 3d, at 971. As the D. C. Circuit saw this case, Herrick adequately represented Taylor for two principal reasons: Herrick had a strong incentive to litigate; and Taylor later hired Herrick’s lawyer, suggesting Taylor’s “satisfaction with the attorney’s performance in the prior case.” Id., at 975.

The D. C. Circuit misapprehended Richards. As just recounted, our holding that the Alabama Supreme Court’s application of res judicata to nonparties violated due process turned on the lack of either special procedures to protect the nonparties’ interests or an understanding by the concerned parties that the first suit was brought in a representative capacity. See Richards, 517 U. S., at 801-802. Richards thus established that representation is “adequate” for purposes of nonparty preclusion only if (at a minimum) one of these two circumstances is present.

We restated Richards’ core holding in South Central Bell Telephone Co. v. Alabama, 526 U. S. 160 (1999). In that case, as in Richards, the Alabama courts had held that a judgment rejecting a challenge to a tax by one group of taxpayers barred a subsequent suit by a different taxpayer. See 526 U. S., at 164-165. In South Central Bell, however, the nonparty had notice of the original suit and engaged one of the lawyers earlier employed by the original plaintiffs. See id., at 167-168. Under the D. C. Circuit’s decision in Taylor’s case, these factors apparently would have sufficed to establish adequate representation. See 490 F. 3d, at 973-975. Yet South Central Bell held that the application of res judicata in that case violated due process. Our inquiry came

to an end when we determined that the original plaintiffs had not understood themselves to be acting in a representative capacity and that there had been no special procedures to safeguard the interests of absentees. See 526 U. S., at 168.

Our decisions recognizing that a nonparty may be bound by a judgment if she was adequately represented by a party to the earlier suit thus provide no support for the D. C. Circuit’s broad theory of virtual representation.

B

Fairchild and the FAA do not argue that the D. C. Circuit’s virtual representation doctrine fits within any of the recognized grounds for nonparty preclusion. Rather, they ask us to abandon the attempt to delineate discrete grounds and clear rules altogether. Preclusion is in order, they contend, whenever “the relationship between a party and a non-party is ‘close enough’ to bring the second litigant within the judgment.” Brief for Respondent Fairchild 20. See also Brief for Respondent FAA 22-24. Courts should make the “close enough” determination, they urge, through a “heavily fact-driven” and “equitable” inquiry. Brief for Respondent Fair-child 20. See also Brief for Respondent FAA 22 (“there is no clear test” for nonparty preclusion; rather, an “equitable and fact-intensive” inquiry is demanded (internal quotation marks omitted)). Only this sort of diffuse balancing, Fair-child and the FAA argue, can account for all of the situations in which nonparty preclusion is appropriate.

We reject this argument for three reasons. First, our decisions emphasize the fundamental nature of the general rule that a litigant is not bound by a judgment to which she was not a party. See, e. g., Richards, 517 U. S., at 798-799; Martin, 490 U. S., at 761-762. Accordingly, we have endeavored to delineate discrete exceptions that apply in “limited circumstances.” Id., at 762, n. 2. Respondents’ amorphous balancing test is at odds with the constrained approach to nonparty preclusion our decisions advance.

Resisting this reading of our precedents, respondents call up three decisions they view as supportive of the approach they espouse. Fairchild quotes our statement in Coryell v. Phipps, 317 U. S. 406, 411 (1943), that privity “turns on the facts of particular cases.” See Brief for Respondent Fair-child 20. That observation, however, scarcely implies that privity is governed by a diffuse balancing test.9 Fairchild also cites Blonder-Tongue Laboratories, Inc. v. University of III. Foundation, 402 U. S. 313, 334 (1971), which stated that estoppel questions turn on “the trial courts’ sense of justice and equity.” See Brief for Respondent Fairchild 20. This passing statement, however, was not made with non-party preclusion in mind; it appeared in a discussion recognizing district courts’ discretion to limit the use of issue preclusion against persons who were parties to a judgment. See Blonder-Tongue, 402 U. S., at 334.

The FAA relies on United States v. Des Moines Valley R. Co., 84 F. 40 (CA8 1897), an opinion we quoted with approval in Schendel, 270 U. S., at 619-620. Des Moines Valley was a quiet title action in which the named plaintiff was the United States. The Government, however, had “no interest in the land” and had “simply permitted [the landowner] to use its name as the nominal plaintiff.” 84 F., at 42. The suit was therefore barred, the appeals court held, by an earlier judgment against the landowner. As the court explained: “[W]here the government lends its name as a plaintiff... to enable one private person to maintain a suit against another,” the government is “subject to the same defenses which exist . . . against the real party in interest.” Id., at 43. Des Moines Valley, the FAA contended at oral argument, demonstrates that it is sometimes appropriate to bind

a nonparty in circumstances that do not fit within any of the established grounds for nonparty preclusion. See Tr. of Oral Arg. 31-33. Properly understood, however, Des Moines Valley is simply an application of the fifth basis for nonparty preclusion described above: A party may not use a representative or agent to relitigate an adverse judgment. See swpra, at 895-896.10 We thus find no support in our precedents for the lax approach to nonparty preclusion advocated by respondents.

Our second reason for rejecting a broad doctrine of virtual representation rests on the limitations attending nonparty preclusion based on adequate representation. A party’s representation of a nonparty is “adequate” for preclusion purposes only if, at a minimum: (1) The interests of the non-party and her representative are aligned, see Hansberry, 311 U. S., at 43; and (2) either the party understood herself to be acting in a representative capacity or the original court took care to protect the interests of the nonparty, see Richards, 517 U. S., at 801-802; supra, at 897-898. In addition, adequate representation sometimes requires (3) notice of the original suit to the persons alleged to have been represented, see Richards, 517 U. S., at 801.11 In the class-action context,

these limitations are implemented by the procedural safeguards contained in Federal Rule of Civil Procedure 23.

An expansive doctrine of virtual representation, however, would “recogniz[e], in effect, a common-law kind of class action.” Tice, 162 F. 3d, at 972 (internal quotation marks omitted). That is, virtual representation would authorize preclusion based on identity of interests and some kind of relationship between parties and nonparties, shorn of the procedural protections prescribed in Hansberry, Rickards, and Rule 23. These protections, grounded in due process, could be circumvented were we to approve a virtual representation doctrine that allowed courts to “create de facto class actions at will.” Tice, 162 F. 3d, at 973.

Third, a diffuse balancing approach to nonparty preclusion would likely create more headaches than it relieves. Most obviously, it could significantly complicate the task of district courts faced in the first instance with preclusion questions. An all-things-considered balancing approach might spark wide-ranging, time-consuming, and expensive discovery tracking factors potentially relevant under seven- or five-prong tests. And after the relevant facts are established, district judges would be called upon to evaluate them under a standard that provides no firm guidance. See Tyus, 93 F. 3d, at 455 (conceding that “there is no clear test for determining the applicability of” the virtual representation doctrine announced in that case). Preclusion doctrine, it should be recalled, is intended to reduce the burden of litigation on courts and parties. Cf. Montana, 440 U. S., at 153-154. “In this area of the law,” we agree, “ ‘crisp rules with sharp corners’ are preferable to a round-about doctrine of opaque standards.” Bittinger v. Tecumseh Products Co., 123 F. 3d 877, 881 (CA6 1997).

c

Finally, relying on the Eighth Circuit’s decision in Tyus, 93 F. 3d, at 456, the FAA maintains that nonparty preclusion should apply more broadly in “public law” litigation than in “private law” controversies. To support this position, the FAA offers two arguments. First, the FAA urges, our decision in Richards acknowledges that, in certain cases, the plaintiff has a reduced interest in controlling the litigation “because of the public nature of the right at issue.” Brief for Respondent FAA 28. When a taxpayer challenges “an alleged misuse of public funds” or “other public action,” we observed in Richards, the suit “has only an indirect impact on [the plaintiff’s] interests.” 517 U. S., at 803. In actions of this character, the Court said, “we may assume that the States have wide latitude to establish procedures ... to limit the number of judicial proceedings that may be entertained.” Ibid.

Taylor’s FOIA action falls within the category described in Richards, the FAA contends, because “the duty to disclose under FOIA is owed to the public generally.” Brief for Respondent FAA 34. The opening sentence of FOIA, it is true, states that agencies “shall make [information] available to the public.” 5 U. S. C. § 552(a) (2006 ed.). Equally true, we have several times said that FOIA vindicates a “public” interest. E. g., National Archives and Records Admin, v. Favish, 541 U. S. 157, 172 (2004). The Act, however, instructs agencies receiving FOIA requests to make the information available not to the public at large, but rather to the “person” making the request. § 552(a)(3)(A). See also § 552(a)(3)(B) (“In making any record available to a person under this paragraph, an agency shall provide the record in any [readily reproducible] form or format requested by the person. ...” (emphasis added)); Brief for National Security Archive et al. as Amici Curiae 10 (“Government agencies do not systematically make released records available to the general public.”). Thus, in contrast to the public-law litiga

tion contemplated in Richards, a successful FOIA action results in a grant of relief to the individual plaintiff, not a decree benefiting the public at large.

Furthermore, we said in Richards only that, for the type of public-law claims there envisioned, States are free to adopt procedures limiting repetitive litigation. See 517 U. S., at 803. In this regard, we referred to instances in which the first judgment foreclosed successive litigation by other plaintiffs because, “under state law, [the suit] could be brought only on behalf of the public at large.” Id., at 804.12 Richards spoke of state legislation, but it appears equally evident that Congress, in providing for actions vindicating a public interest, may “limit the number of judicial proceedings that may be entertained.” Id., at 803. It hardly follows, however, that this Court should proscribe or confine successive FOIA suits by different requesters. Indeed, Congress’ provision for FOIA suits with no statutory constraint on successive actions counsels against judicial imposition of constraints through extraordinary application of the common law of preclusion.

The FAA next argues that “the threat of vexatious litigation is heightened” in public-law cases because “the number of plaintiffs with standing is potentially limitless.” Brief for Respondent FAA 28 (internal quotation marks omitted). FOIA does allow “any person” whose request is denied to resort to federal court for review of the agency’s determination. 5 U. S. C. § 552(a)(3)(A), (4)(B) (2006 ed.). Thus it is theoretically possible that several persons could coordinate to mount a series of repetitive lawsuits.

But we are not convinced that this risk justifies departure from the usual rules governing nonparty preclusion. First, stare decisis will allow courts swiftly to dispose of repetitive suits brought in the same circuit. Second, even when stare

decisis is not dispositive, “the human tendency not to waste money will deter the bringing of suits based on claims or issues that have already been adversely determined against others.” Shapiro 97. This intuition seems to be borne out by experience: The FAA has not called our attention to any instances of abusive FOIA suits in the Circuits that reject the virtual representation theory respondents advocate here.

IV

For the foregoing reasons, we disapprove the theory of virtual representation on which the decision below rested. The preclusive effects of a judgment in a federal-question case decided by a federal court should instead be determined according to the established grounds for nonparty preclusion described in this opinion. See Part II-B, supra.

Although references to “virtual representation” have proliferated in the lower courts, our decision is unlikely to occasion any great shift in actual practice. Many opinions use the term “virtual representation” in reaching results at least arguably defensible on established grounds. See 18A Wright & Miller §4457, at 535-539, and n. 38 (collecting cases). In these cases, dropping the “virtual representation” label would lead to .clearer analysis with little, if any, change in outcomes. See Tice, 162 F. 3d, at 971 (“[T]he term ‘virtual representation’ has cast more shadows than light on the problem [of nonparty preclusion].”).

In some cases, however, lower courts have relied on virtual representation to extend nonparty preclusion beyond the latter doctrine’s proper bounds. We now turn back to Taylor’s action to determine whether his suit is such a case, or whether the result reached by the courts below can be justified on one of the recognized grounds for nonparty preclusion.

A

It is uncontested that four of the six grounds for nonparty preclusion have no application here: There is no indication

that Taylor agreed to be bound by Herrick’s litigation, that Taylor and Herrick have any legal relationship, that Taylor exercised any control over Herrick’s suit, or that this suit implicates any special statutory scheme limiting relitigation. Neither the FA A nor Fairchild contends otherwise.

It is equally clear that preclusion cannot be justified on the theory that Taylor was adequately represented in Herrick’s suit. Nothing in the record indicates that Herrick understood himself to be suing on Taylor’s behalf, that Taylor even knew of Herrick’s suit, or that the Wyoming District Court took special care to protect Taylor’s interests. Under our pathmarking precedent, therefore, Herrick’s representation was not “adequate.” See Richards, 517 U. S., at 801-802.

That leaves only the fifth category: preclusion because a nonparty to an earlier litigation has brought suit as a representative or agent of a party who is bound by the prior adjudication. Taylor is not Herrick’s legal representative and he has not purported to sue in a representative capacity. He concedes, however, that preclusion would be appropriate if respondents could demonstrate that he is acting as Herrick’s “undisclosed agen[t].” Brief for Petitioner 23, n. 4. See also id., at 24, n. 5.

Respondents argue here, as they did below, that Taylor’s suit is a collusive attempt to relitigate Herrick’s action. See Brief for Respondent Fairchild 32, and n. 18; Brief for Respondent FAA 18-19, 33, 39. The D. C. Circuit considered a similar question in addressing the “tactical maneuvering” prong of its virtual representation test. See 490 F. 3d, at 976. The Court of Appeals did not, however, treat the issue as one of agency, and it expressly declined to reach any definitive conclusions due to “the ambiguity of the facts.” Ibid. We therefore remand to give the courts below an opportunity to determine whether Taylor, in pursuing the instant FOIA suit, is acting as Herrick’s agent. Taylor concedes that such a remand is appropriate. See Tr. of Oral Arg. 56-57.

We have never defined the showing required to establish that a nonparty to a prior adjudication has become a litigating agent for a party to the earlier case. Because the issue has not been briefed in any detail, we do not discuss the matter elaboratively here. We note, however, that courts should be cautious about finding preclusion on this basis. A mere whiff of “tactical maneuvering” will not suffice; instead, principles of agency law are suggestive. They indicate that preclusion is appropriate only if the putative agent’s conduct of the suit is subject to the control of the party who is bound by the prior adjudication. See 1 Restatement (Second) of Agency § 14, p. 60 (1957) (“A principal has the right to control the conduct of the agent with respect to matters entrusted to him.”).13

B

On remand, Fairchild suggests, Taylor should bear the burden of proving he is not acting as Herrick’s agent. When a defendant points to evidence establishing a close relationship between successive litigants, Fairchild maintains, “the burden [should] shif[t] to the second litigant to submit evidence refuting the charge” of agency. Brief for Respondent Fairchild 27-28. Fairchild justifies this proposed burden-shift on the ground that “it is unlikely an opposing party will have access to direct evidence of collusion.” Id., at 28, n. 14.

We reject Fairchild’s suggestion. Claim preclusion, like issue preclusion, is an affirmative defense. See Fed. Rule Civ. Proc. 8(c); Blonder-Tongue, 402 U. S., at 350. Ordinarily, it is incumbent on the defendant to plead and prove such a defense, see Jones v. Bock, 549 U. S. 199, 204 (2007), and we have never recognized claim preclusion as an exception to that general rule, see 18 Wright & Miller § 4405, at 83 (“[A] party asserting preclusion must carry the burden of establishing all necessary elements.”). We acknowledge that direct evidence justifying nonparty preclusion is often in the hands of plaintiffs rather than defendants. See, e. g., Montana, 440 U. S., at 155 (listing evidence of control over a prior suit). But “[v]ery often one must plead and prove matters as to which his adversary has superior access to the proof.” 2 K. Broun, McCormick on Evidence §337, p. 475 (6th ed. 2006). In these situations, targeted interrogatories or deposition questions can reduce the information disparity. We see no greater cause here than in other matters of affirmative defense to disturb the traditional allocation of the proof burden.

For the reasons stated, the judgment of the United States Court of Appeals for the District of Columbia Circuit is vacated, and the case is remanded for further proceedings consistent with this opinion.