United States Parole Commission v. Geraghty

445 U.S. 388 (1980) UNITED STATES PAROLE COMMISSION ET AL. v. GERAGHTY. No. 78-572. Supreme Court of United States. Argued October 2, 1979. Decided March 19, 1980. CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE THIRD CIRCUIT. *389 Kent L. Jones argued the cause pro hac vice for petitioners. With him on the briefs were Solicitor General McCree, Assistant Attorney General Heymann, Deputy Solicitor General Easterbrook, Jerome M. Feit, and Elliott Schulder. Kenneth N. Flaxman argued the cause for respondent. With him on the brief was Thomas R. Meites.[*] *390 MR. JUSTICE BLACKMUN delivered the opinion of the Court. This case raises the question whether a trial court's denial of a motion for certification of a class may be reviewed on appeal after the named plaintiff's personal claim has become "moot." The United States Court of Appeals for the Third Circuit held that a named plaintiff, respondent here, who brought a class action challenging the validity of the United States Parole Commission's Parole Release Guidelines, could continue his appeal of a ruling denying class certification even though he had been released from prison while the appeal was pending. We granted certiorari, 440 U. S. 945 (1979), to consider this issue of substantial significance, under Art. III of the Constitution, to class-action litigation,[1] and to resolve the conflict in approach among the Courts of Appeals.[2] *391 I In 1973, the United States Parole Board adopted explicit Parole Release Guidelines for adult prisoners.[3] These guidelines establish a "customary range" of confinement for various classes of offenders. The guidelines utilize a matrix, which combines a "parole prognosis" score (based on the prisoner's age at first conviction, employment background, and other personal factors) and an "offense severity" rating, to yield the "customary" time to be served in prison. Subsequently, in 1976, Congress enacted the Parole Commission and Reorganization Act (PCRA), Pub. L. 94-233, 90 Stat. 219, 18 U. S. C. §§ 4201-4218. This Act provided the first legislative authorization for parole release guidelines. It required the newly created Parole Commission to "promulgate rules and regulations establishing guidelines for the powe[r] . . . to grant or deny an application or recommendation to parole any eligible prisoner." § 4203. Before releasing a prisoner on parole, the Commission must find, "upon consideration of the nature and circumstances of the offense and the history and characteristics of the prisoner," that release "would not depreciate the seriousness of his offense or promote disrespect for the law" and that it "would not jeopardize the public welfare." § 4206 (a). Respondent John M. Geraghty was convicted in the United States District Court for the Northern District of Illinois of *392 conspiracy to commit extortion, in violation of 18 U. S. C. § 1951, and of making false material declarations to a grand jury, in violation of 18 U. S. C. § 1623 (1976 ed. and Supp. II).[4] On January 25, 1974, two months after initial promulgation of the release guidelines, respondent was sentenced to concurrent prison terms of four years on the conspiracy count and one year on the false declarations count. The United States Court of Appeals for the Seventh Circuit affirmed respondent's convictions. United States v. Braasch, 505 F. 2d 139 (1974), cert. denied sub nom. Geraghty v. United States, 421 U. S. 910 (1975). Geraghty later, pursuant to a motion under Federal Rule of Criminal Procedure 35, obtained from the District Court a reduction of his sentence to 30 months. The court granted the motion because, in the court's view, application of the guidelines would frustrate the sentencing judge's intent with respect to the length of time Geraghty would serve in prison. United States v. Braasch, No. 72 CR 979 (ND Ill., Oct. 9, 1975), appeal dism'd and mandamus denied, 542 F. 2d 442 (CA7 1976). Geraghty then applied for release on parole. His first application was denied in January 1976 with the following explanation: "Your offense behavior has been rated as very high severity. You have a salient factor score of 11. You have been in custody for a total of 4 months. Guidelines established by the Board for adult cases which consider the above factors indicate a range of 26-36 months to be served before release for cases with good institutional program performance and adjustment. After review of all relevant factors and information presented, it is found *393 that a decision at this consideration outside the guidelines does not appear warranted." App. 5. If the customary release date applicable to respondent under the guidelines were adhered to, he would not be paroled before serving his entire sentence minus good-time credits. Geraghty applied for parole again in June 1976; that application was denied for the same reasons. He then instituted this civil suit as a class action in the United States District Court for the District of Columbia, challenging the guidelines as inconsistent with the PCRA and the Constitution, and questioning the procedures by which the guidelines were applied to his case. Respondent sought certification of a class of "all federal prisoners who are or who will become eligible for release on parole." Id., at 17. Without ruling on Geraghty's motion, the court transferred the case to the Middle District of Pennsylvania, where respondent was incarcerated. Geraghty continued to press his motion for class certification, but the court postponed ruling on the motion until it was prepared to render a decision on cross-motions for summary judgment. The District Court subsequently denied Geraghty's request for class certification and granted summary judgment for petitioners on all the claims Geraghty asserted. 429 F. Supp. 737 (1977). The court regarded respondent's action as a petition for a writ of habeas corpus, to which Federal Rule of Civil Procedure 23 applied only by analogy. It denied class certification as "neither necessary nor appropriate." 429 F. Supp., at 740. A class action was "necessary" only to avoid mootness. The court found such a consideration not comprehended by Rule 23. It found class certification inappropriate because Geraghty raised certain individual issues and, inasmuch as some prisoners might be benefited by the guidelines, because his claims were not typical of the entire proposed class. 429 F. Supp., at 740-741. On the merits, the court ruled that the guidelines are consistent with the PCRA and *394 do not offend the Ex Post Facto Clause, U. S. Const., Art. I, § 9, cl. 3. 429 F. Supp., at 741-744. Respondent, individually "and on behalf of a class," appealed to the United States Court of Appeals for the Third Circuit. App. 29. Thereafter, another prisoner, Becher, who had been denied parole through application of the guidelines and who was represented by Geraghty's counsel, moved to intervene. Becher sought intervention to ensure that the legal issue raised by Geraghty on behalf of the class "will not escape review in the appeal in this case." Pet. to Intervene After Judgment 2. The District Court, concluding that the filing of Geraghty's notice of appeal had divested it of jurisdiction, denied the petition to intervene. Becher then filed a timely notice of appeal from the denial of intervention. The two appeals were consolidated. On June 30, 1977, before any brief had been filed in the Court of Appeals, Geraghty was mandatorily released from prison; he had served 22 months of his sentence, and had earned good-time credits for the rest. Petitioners then moved to dismiss the appeals as moot. The appellate court reserved decision of the motion to dismiss until consideration of the merits. The Court of Appeals, concluding that the litigation was not moot, reversed the judgment of the District Court and remanded the case for further proceedings. 579 F. 2d 238 (CA3 1978). If a class had been certified by the District Court, mootness of respondent Geraghty's personal claim would not have rendered the controversy moot. See, e. g., Sosna v. Iowa, 419 U. S. 393 (1975). The Court of Appeals reasoned that an erroneous denial of a class certification should not lead to the opposite result. 579 F. 2d, at 248-252. Rather, certification of a "certifiable" class, that erroneously had been denied, relates back to the original denial and thus preserves jurisdiction. Ibid. On the question whether certification erroneously had been denied, the Court of Appeals held that necessity is not a prerequisite *395 under Rule 23. 579 F. 2d, at 252. The court expressed doubts about the District Court's finding that class certification was "inappropriate." While Geraghty raised some claims not applicable to the entire class of prisoners who are or will become eligible for parole, the District Court could have "certif[ied] certain issues as subject to class adjudication, and . . . limite[d] overbroad classes by the use of sub-classes." Id., at 253. Failure "to consider these options constituted a failure properly to exercise discretion." Ibid. "Indeed, this authority may be exercised sua sponte." Ibid. The Court of Appeals also held that refusal to certify because of a potential conflict of interest between Geraghty and other members of the putative class was error. The subclass mechanism would have remedied this problem as well. Id., at 252-253. Thus, the Court of Appeals reversed the denial of class certification and remanded the case to the District Court for an initial evaluation of the proper subclasses. Id., at 254. The court also remanded the motion for intervention. Id., at 245, n. 21.[5] In order to avoid "improvidently dissipat[ing] judicial effort," id., at 254, the Court of Appeals went on to consider whether the trial court had decided the merits of respondent's case properly. The District Court's entry of summary judgment was found to be error because "if Geraghty's recapitulation of the function and genesis of the guidelines is supported by the evidence," the guidelines "may well be" unauthorized or unconstitutional. Id., at 259, 268. Thus, the dispute on the merits also was remanded for further factual development. II Article III of the Constitution limits federal "judicial Power," that is, federal-court jurisdiction, to "Cases" and "Controversies." This case-or-controversy limitation serves *396 "two complementary" purposes. Flast v. Cohen, 392 U. S. 83, 95 (1968). It limits the business of federal courts to "questions presented in an adversary context and in a form historically viewed as capable of resolution through the judicial process," and it defines the "role assigned to the judiciary in a tripartite allocation of power to assure that the federal courts will not intrude into areas committed to the other branches of government." Ibid. Likewise, mootness has two aspects: "when the issues presented are no longer `live' or the parties lack a legally cognizable interest in the outcome." Powell v. McCormack, 395 U. S. 486, 496 (1969). It is clear that the controversy over the validity of the Parole Release Guidelines is still a "live" one between petitioners and at least some members of the class respondent seeks to represent. This is demonstrated by the fact that prisoners currently affected by the guidelines have moved to be substituted, or to intervene, as "named" respondents in this Court. See n. 1, supra. We therefore are concerned here with the second aspect of mootness, that is, the parties' interest in the litigation. The Court has referred to this concept as the "personal stake" requirement. E. g., Franks v. Bowman Transportation Co., 424 U. S. 747, 755 (1976); Baker v. Carr, 369 U. S. 186, 204 (1962). The personal-stake requirement relates to the first purpose of the case-or-controversy doctrine—limiting judicial power to disputes capable of judicial resolution. The Court in Flast v. Cohen, 392 U. S., at 100-101, stated: "The question whether a particular person is a proper party to maintain the action does not, by its own force, raise separation of powers problems related to improper judicial interference in areas committed to other branches of the Federal Government. . . . Thus, in terms of Article III limitations on federal court jurisdiction, the question of standing is related only to whether the dispute sought to be adjudicated will be presented in an adversary *397 context and in a form historically viewed as capable of judicial resolution. It is for that reason that the emphasis in standing problems is on whether the party invoking federal court jurisdiction has `a personal stake in the outcome of the controversy,' Baker v. Carr, [369 U. S.], at 204, and whether the dispute touches upon `the legal relations of parties having adverse legal interests,' Aetna Life Insurance Co. v. Haworth, [300 U. S.], at 240-241." See also Schlesinger v. Reservists to Stop the War, 418 U. S. 208, 216-218 (1974). The "personal stake" aspect of mootness doctrine also serves primarily the purpose of assuring that federal courts are presented with disputes they are capable of resolving. One commentator has defined mootness as "the doctrine of standing set in a time frame: The requisite personal interest that must exist at the commencement of the litigation (standing) must continue throughout its existence (mootness)." Monaghan, Constitutional Adjudication: The Who and When, 82 Yale L. J. 1363, 1384 (1973). III On several occasions the Court has considered the application of the "personal stake" requirement in the class-action context. In Sosna v. Iowa, 419 U. S. 393 (1975), it held that mootness of the named plaintiff's individual claim after a class has been duly certified does not render the action moot. It reasoned that "even though appellees . . . might not again enforce the Iowa durational residency requirement against [the class representative], it is clear that they will enforce it against those persons in the class that appellant sought to represent and that the District Court certified." Id., at 400. The Court stated specifically that an Art. III case or controversy "may exist . . . between a named defendant and a member of the class represented by the named plaintiff, even *398 though the claim of the named plaintiff has become moot." Id., at 402.[6] Although one might argue that Sosna contains at least an implication that the critical factor for Art. III purposes is the timing of class certification, other cases, applying a "relation back" approach, clearly demonstrate that timing is not crucial. When the claim on the merits is "capable of repetition, yet evading review," the named plaintiff may litigate the class certification issue despite loss of his personal stake in the outcome of the litigation. E. g., Gerstein v. Pugh, 420 U. S. 103, 110, n. 11 (1975). The "capable of repetition, yet evading review" doctrine, to be sure, was developed outside the class-action context. See Southern Pacific Terminal Co. v. ICC, 219 U. S. 498, 514-515 (1911). But it has been applied where the named plaintiff does have a personal stake at the outset of the lawsuit, and where the claim may arise again with respect to that plaintiff; the litigation then may continue notwithstanding the named plaintiff's current lack of a personal stake. See, e. g., Weinstein v. Bradford, 423 U. S. 147, 149 (1975); Roe v. Wade, 410 U. S. 113, 123-125 (1973). Since the litigant faces some likelihood of becoming involved in the same controversy in the future, vigorous advocacy can be expected to continue. When, however, there is no chance that the named plaintiff's expired claim will reoccur, mootness still can be avoided through certification of a class prior to expiration of the named plaintiff's personal claim. E. g., Franks v. Bowman Transportation Co., 424 U. S., at 752-757. See Kremens v. Bartley, *399 431 U. S. 119, 129-130 (1977). Some claims are so inherently transitory that the trial court will not have even enough time to rule on a motion for class certification before the proposed representative's individual interest expires. The Court considered this possibility in Gerstein v. Pugh, 420 U. S., at 110, n. 11. Gerstein was an action challenging pretrial detention conditions. The Court assumed that the named plaintiffs were no longer in custody awaiting trial at the time the trial court certified a class of pretrial detainees. There was no indication that the particular named plaintiffs might again be subject to pretrial detention. Nonetheless, the case was held not to be moot because: "The length of pretrial custody cannot be ascertained at the outset, and it may be ended at any time by release on recognizance, dismissal of the charges, or a guilty plea, as well as by acquittal or conviction after trial. It is by no means certain that any given individual, named as plaintiff, would be in pretrial custody long enough for a district judge to certify the class. Moreover, in this case the constant existence of a class of persons suffering the deprivation is certain. The attorney representing the named respondents is a public defender, and we can safely assume that he has other clients with a continuing live interest in the case." Ibid. See also Sosna v. Iowa, 419 U. S., at 402, n. 11. In two different contexts the Court has stated that the proposed class representative who proceeds to a judgment on the merits may appeal denial of class certification. First, this assumption was "an important ingredient," Deposit Guaranty Nat. Bank v. Roper, ante, at 338, in the rejection of interlocutory appeals, "as of right," of class certification denials. Coopers & Lybrand v. Livesay, 437 U. S. 463, 469, 470, n. 15 (1978). The Court reasoned that denial of class status will not necessarily be the "death knell" of a small-claimant action, since there still remains "the prospect of prevailing on *400 the merits and reversing an order denying class certification." Ibid. Second, in United Airlines, Inc. v. McDonald, 432 U. S. 385, 393-395 (1977), the Court held that a putative class member may intervene, for the purpose of appealing the denial of a class certification motion, after the named plaintiffs' claims have been satisfied and judgment entered in their favor. Underlying that decision was the view that "refusal to certify was subject to appellate review after final judgment at the behest of the named plaintiffs." Id., at 393. See also Coopers & Lybrand v. Livesay, 437 U. S., at 469. And today, the Court holds that named plaintiffs whose claims are satisfied through entry of judgment over their objections may appeal the denial of a class certification ruling. Deposit Guaranty Nat. Bank v. Roper, ante, p. 326. Gerstein, McDonald, and Roper are all examples of cases found not to be moot, despite the loss of a "personal stake" in the merits of the litigation by the proposed class representative. The interest of the named plaintiffs in Gerstein was precisely the same as that of Geraghty here. Similarly, after judgment had been entered in their favor, the named plaintiffs in McDonald had no continuing narrow personal stake in the outcome of the class claims. And in Roper the Court points out that an individual controversy is rendered moot, in the strict Art. III sense, by payment and satisfaction of a final judgment. Ante, at 333. These cases demonstrate the flexible character of the Art. III mootness doctrine.[7] As has been noted in the past, *401 Art. III justiciability is "not a legal concept with a fixed content or susceptible of scientific verification." Poe v. Ullman, 367 U. S. 497, 508 (1961) (Plurality opinion). "[T]he justiciability doctrine [is] one of uncertain and shifting contours." Flast v. Cohen, 392 U. S., at 97. IV Perhaps somewhat anticipating today's decision in Roper, petitioners argue that the situation presented is entirely different when mootness of the individual claim is caused by "expiration" of the claim, rather than by a judgment on the claim. They assert that a proposed class representative who individually prevails on the merits still has a "personal stake" in the outcome of the litigation, while the named plaintiff whose claim is truly moot does not. In the latter situation, where no class has been certified, there is no party before the court with a live claim, and it follows, it is said, that we have no jurisdiction to consider whether a class should have been certified. Brief for Petitioners 37-39. We do not find this distinction persuasive. As has been noted earlier, Geraghty's "personal stake" in the outcome of the litigation is, in a practical sense, no different from that of the putative class representatives in Roper. Further, the opinion in Roper indicates that the approach to take in applying Art. III is issue by issue. "Nor does a confession of judgment *402 by defendants on less than all the issues moot an entire case; other issues in the case may be appealable. We can assume that a district court's final judgment fully satisfying named plaintiffs' private substantive claims would preclude their appeal on that aspect of the final judgment; however, it does not follow that this circumstance would terminate the named plaintiffs' right to take an appeal on the issue of class certification." Ante, at 333. See also United Airlines, Inc. v. McDonald, 432 U. S., at 392; Powell v. McCormack, 395 U. S., at 497. Similarly, the fact that a named plaintiff's substantive claims are mooted due to an occurrence other than a judgment on the merits does not mean that all the other issues in the case are mooted. A plaintiff who brings a class action presents two separate issues for judicial resolution. One is the claim on the merits; the other is the claim that he is entitled to represent a class. "The denial of class certification stands as an adjudication of one of the issues litigated," Roper, ante, at 336. We think that in determining whether the plaintiff may continue to press the class certification claim, after the claim on the merits "expires," we must look to the nature of the "personal stake" in the class certification claim. Determining Art. III's "uncertain and shifting contours," see Flast v. Cohen, 392 U. S., at 97, with respect to nontraditional forms of litigation, such as the class action, requires reference to the purposes of the case-or-controversy requirement. Application of the personal-stake requirement to a procedural claim, such as the right to represent a class, is not automatic or readily resolved. A "legally cognizable interest," as the Court described it in Powell v. McCormack, 395 U. S., at 496, in the traditional sense rarely ever exists with respect to the class certification claim.[8] The justifications that led to the development of the class action include the protection of *403 the defendant from inconsistent obligations, the protection of the interests of absentees, the provision of a convenient and economical means for disposing of similar lawsuits, and the facilitation of the spreading of litigation costs among numerous litigants with similar claims. See, e. g., Advisory Committee Notes on Fed. Rule Civ. Proc. 23, 28 U. S. C. App., pp. 427-429; Note, Developments in the Law, Class Actions, 89 Harv. L. Rev. 1318, 1321-1323, 1329-1330 (1976). Although the named representative receives certain benefits from the class nature of the action, some of which are regarded as desirable and others as less so,[9] these benefits generally are by-products of the class-action device. In order to achieve the primary benefits of class suits, the Federal Rules of Civil Procedure give the proposed class representative the right to have a class certified if the requirements of the Rules are met. This "right" is more analogous to the private attorney general concept than to the type of interest traditionally thought to satisfy the "personal stake" requirement. See Roper, ante, at 338. As noted above, the purpose of the "personal stake" requirement is to assure that the case is in a form capable of judicial resolution. The imperatives of a dispute capable of judicial resolution are sharply presented issues in a concrete factual setting and self-interested parties vigorously advocating opposing positions. Franks v. Bowman Transportation Co., 424 U. S., at 753-756; Baker v. Carr, 369 U. S., at 204; Poe v. Ullman, 367 U. S., at 503 (plurality opinion). We conclude that these elements can exist with respect to the class certification issue notwithstanding the fact that the named plaintiff's claim on the merits has expired. The question whether class certification is appropriate remains as a concrete, sharply presented *404 issue. In Sosna v. Iowa it was recognized that a named plaintiff whose claim on the merits expires after class certification may still adequately represent the class. Implicit in that decision was the determination that vigorous advocacy can be assured through means other than the traditional requirement of a "personal stake in the outcome." Respondent here continues vigorously to advocate his right to have a class certified. We therefore hold that an action brought on behalf of a class does not become moot upon expiration of the named plaintiff's substantive claim, even though class certification has been denied.[10] The proposed representative retains a "personal stake" in obtaining class certification sufficient to assure that Art. III values are not undermined. If the appeal results in reversal of the class certification denial, and a class subsequently is properly certified, the merits of the class claim then may be adjudicated pursuant to the holding in Sosna. Our holding is limited to the appeal of the denial of the class certification motion. A named plaintiff whose claim expires may not continue to press the appeal on the merits until a class has been properly certified. See Roper, ante, at 336-337. If, on appeal, it is determined that class certification properly was denied, the claim on the merits must be dismissed as moot.[11] *405 Our conclusion that the controversy here is not moot does not automatically establish that the named plaintiff is entitled to continue litigating the interests of the class. "[I]t does *406 shift the focus of examination from the elements of justiciability to the ability of the named representative to `fairly and adequately protect the interests of the class.' Rule 23 (a)." *407 Sosna v. Iowa, 419 U. S., at 403. We hold only that a case or controversy still exists. The question of who is to represent the class is a separate issue.[12] We need not decide here whether Geraghty is a proper representative for the purpose of representing the class on the merits. No class as yet has been certified. Upon remand, the District Court can determine whether Geraghty may continue to press the class claims or whether another representative would be appropriate. We decide only that Geraghty was a proper representative for the purpose of appealing the ruling denying certification of the class that he initially defined. Thus, it was not improper for the Court of Appeals to consider whether the District Court should have granted class certification. V We turn now to the question whether the Court of Appeals' decision on the District Court's class certification ruling was proper. Petitioners assert that the Court of Appeals erred in requiring the District Court to consider the possibility of certifying *408 subclasses sua sponte. Petitioners strenuously contend that placing the burden of identifying and constructing subclasses on the trial court creates unmanageable difficulties. Brief for Petitioners 43-51. We feel that the Court of Appeals' decision here does not impose undue burdens on the district courts. Respondent had no real opportunity to request certification of subclasses after the class he proposed was rejected. The District Court denied class certification at the same time it rendered its adverse decision on the merits. Requesting subclass certification at that time would have been a futile act. The District Court was not about to invest effort in deciding the subclass question after it had ruled that no relief on the merits was available. The remand merely gives respondent the opportunity to perform his function in the adversary system. On remand, however, it is not the District Court that is to bear the burden of constructing subclasses. That burden is upon the respondent and it is he who is required to submit proposals to the court. The court has no sua sponte obligation so to act. With this modification, the Court of Appeals' remand of the case for consideration of subclasses was a proper disposition. It would be inappropriate for this Court to reach the merits of this controversy in the present posture of the case. Our holding that the case is not moot extends only to the appeal of the class certification denial. If the District Court again denies class certification, and that decision is affirmed, the controversy on the merits will be moot. Furthermore, although the Court of Appeals commented upon the merits for the sole purpose of avoiding waste of judicial resources, it did not reach a final conclusion on the validity of the guidelines. Rather, it held only that summary judgment was improper and remanded for further factual development. Given the interlocutory posture of the case before us, we must defer decision on the merits of respondent's case until after it is determined affirmatively that a class properly can be certified. *409 The judgment of the Court of Appeals is vacated, and the case is remanded for further proceedings consistent with this opinion. It is so ordered. MR. JUSTICE POWELL, with whom THE CHIEF JUSTICE, MR. JUSTICE STEWART, and MR. JUSTICE REHNQUIST join, dissenting. Respondent filed this suit as a class action while he was serving time in a federal prison. He sought to represent a class composed of "all federal prisoners who are or who will become eligible for release on parole." App. 17. The District Court denied class certification and granted summary judgment for petitioners. Respondent appealed, but before briefs were filed, he was unconditionally released from prison. Petitioners then moved to dismiss the appeal as moot. The Court of Appeals denied the motion, reversed the judgment of the District Court, and remanded the case for further proceedings. Conceding that respondent's personal claim was moot, the Court of Appeals nevertheless concluded that respondent properly could appeal the denial of class certification. The Court today agrees with this conclusion. The Court's analysis proceeds in two steps. First, it says that mootness is a "flexible" doctrine which may be adapted as we see fit to "nontraditional" forms of litigation. Ante, at 400-402. Second, the Court holds that the named plaintiff has a right "analogous to the private attorney general concept" to appeal the denial of class certification even when his personal claim for relief is moot. Ante, at 402-404. Both steps are significant departures from settled law that rationally cannot be confined to the narrow issue presented in this case. Accordingly, I dissent. I As the Court observes, this case involves the "personal stake" aspect of the mootness doctrine. Ante, at 396. There *410 is undoubtedly a "live" issue which an appropriate plaintiff could present for judicial resolution. The question is whether respondent, who has no further interest in this action, nevertheless may—through counsel—continue to litigate it. Recent decisions of this Court have considered the personal stake requirement with some care. When the issue is presented at the outset of litigation as a question of standing to sue, we have held that the personal stake requirement has a double aspect. On the one hand, it derives from Art. III limitations on the power of the federal courts. On the other, it embodies additional, self-imposed restraints on the exercise of judicial power. E. g., Singleton v. Wulff, 428 U. S. 106, 112 (1976); Warth v. Seldin, 422 U. S. 490, 498 (1975). The prudential aspect of standing aptly is described as a doctrine of uncertain contours. Ante, at 402. But the constitutional minimum has been given definite content: "In order to satisfy Art. III, the plaintiff must show that he personally has suffered some actual or threatened injury as a result of the putatively illegal conduct of the defendant." Gladstone, Realtors v. Village of Bellwood, 441 U. S. 91, 99 (1979).[1] Although noneconomic injuries can confer standing, the Court has rejected all attempts to substitute abstract concern with a subject— or with the rights of third parties—for "the concrete injury required by Art. III." Simon v. Eastern Kentucky Welfare Rights Org., 426 U. S. 26, 40 (1976).[2] *411 As the Court notes today, the same threshold requirement must be satisfied throughout the action. Ante, at 397; see Sosna v. Iowa, 419 U. S. 393, 402 (1975). Prudential considerations not present at the outset may support continuation of an action in which the parties have invested substantial resources and generated a factual record.[3] But an actual case or controversy in the constitutional sense "`must be extant at all stages of review.'" Preiser v. Newkirk, 422 U. S. 395, 401 (1975), quoting Steffel v. Thompson, 415 U. S. 452, 459, n. 10 (1974). Cases that no longer "`touc[h] the legal relations of parties having adverse legal interests'" are moot because "federal courts are without power to decide questions that cannot affect the rights of litigants in the case before them." North Carolina v. Rice, 404 U. S. 244, 246 (1971) (per curiam), quoting Aetna Life Ins. Co. v. Haworth, 300 U. S. 227, 240-241 (1937). The limitation flows directly from Art. III. DeFunis v. Odegaard, 416 U. S. 312, 316 (1974) (per curiam).[4] Since the question is one of power, the practical importance of review cannot control. Sosna v. Iowa, supra, at 401, n. 9; Richardson v. Ramirez, 418 U. S. 24, 36 (1974); United States v. Alaska S. S. Co., 253 U. S. 113, 116 (1920). Nor can public interest in the resolution of an issue replace the *412 necessary individual interest in the outcome. See DeFunis v. Odegaard, supra, at 316. Collateral consequences of the original wrong may supply the individual interest in some circumstances. Sibron v. New York, 392 U. S. 40, 53-58 (1968). So, too, may the prospect of repeated future injury so inherently transitory that it is unlikely to outlast the normal course of litigation. Super Tire Engineering Co. v. McCorkle, 416 U. S. 115 (1974); Southern Pacific Terminal Co. v. ICC, 219 U. S. 498, 515 (1911). The essential and irreducible constitutional requirement is simply a nonfrivolous showing of continuing or threatened injury at the hands of the adversary. These cases demonstrate, contrary to the Court's view today, that the core requirement of a personal stake in the outcome is not "flexible." Indeed, the rule barring litigation by those who have no interest of their own at stake is applied so rigorously that it has been termed the "one major proposition" in the law of standing to which "the federal courts have consistently adhered . . . without exception." Davis, Standing: Taxpayers and Others, 35 U. Chi. L. Rev. 601, 617 (1968) (emphasis deleted).[5] We have insisted upon the personal stake requirement in mootness and standing cases because it is embedded in the case-or-controversy limitation imposed by the Constitution, "founded in concern about the proper—and properly limited—role of the courts in a democratic society." Warth v. Seldin, supra, at 498. In this *413 way we have, until today, "prevent[ed] the judicial process from becoming no more than a vehicle for the vindication of the value interests of concerned bystanders." United States v. SCRAP, 412 U. S. 669, 687 (1973); see Simon v. Eastern Kentucky Welfare Rights Org., 426 U. S., at 60 (BRENNAN, J., concurring in judgment); Sierra Club v. Morton, 405 U. S. 727, 740 (1972). II The forgoing decisions establish principles that the Court has applied consistently. These principles were developed outside the class action context. But Art. III contains no exception for class actions. Thus, we have held that a putative class representative who alleges no individual injury "may [not] seek relief on behalf of himself or any other member of the class." O'Shea v. Littleton, 414 U. S. 488, 494 (1974). Only after a class has been certified in accordance with Rule 23 can it "acquir[e] a legal status separate from the interest asserted by [the named plaintiff]." Sosna v. Iowa, supra, at 399. "Given a properly certified class," the live interests of unnamed but identifiable class members may supply the personal stake required by Art. III when the named plaintiff's individual claim becomes moot. Franks v. Bowman Transportation Co., 424 U. S. 747, 755-756 (1976); Sosna v. Iowa, supra, at 402. This case presents a fundamentally different situation. No class has been certified, and the lone plaintiff no longer has any personal stake in the litigation.[6] In the words of his own *414 lawyer, respondent "can obtain absolutely no additional personal relief" in this case. Tr. of Oral Arg. 25. Even the lawyer has evinced no interest in continuing to represent respondent as named plaintiff, as distinguished from other persons presently incarcerated. Ibid.[7] In these circumstances, Art. III and the precedents of this Court require dismissal. But the Court views the case differently, and constructs new doctrine to breathe life into a lawsuit that has no plaintiff. The Court announces today for the first time—and without attempting to reconcile the many cases to the contrary—that there are two categories of "the Art. III mootness doctrine": "flexible" and "less flexible." Ante, at 400, and n. 7. The Court then relies on cases said to demonstrate the application of "flexible" mootness to class action litigation. The cases principally relied upon are Gerstein v. Pugh, 420 U. S. 103, 110-111, n. 11 (1975), United Airlines, Inc. v. McDonald, 432 U. S. 385 (1977), and today's decision in Deposit Guaranty Nat. Bank v. Roper, ante, p. 326. Each case is said to show that a class action is not mooted by the loss of the class representative's personal stake in the outcome of the lawsuit, even though no class has been certified. Ante, at 400. Sosna itself is cited for the proposition that the requirements of Art. III may be met "through means other than the traditional requirement of a `personal stake in the outcome.'" Ante, at 404. In my view, the Court misreads these precedents. *415 A In Sosna, the Court simply acknowledged that actual class certification gives legal recognition to additional adverse parties. Cf. Aetna Life Ins. Co. v. Haworth, 300 U. S., at 240.[8] And in Gerstein, the Court applied a rule long established, outside the class action context, by cases that never have been thought to erode the requirement of a personal stake in the outcome. Gerstein held that a class action challenging the constitutionality of pretrial detention procedures could continue after the named plaintiffs' convictions had brought their detentions to an end. The Court did not suggest that a personal stake in the outcome on the merits was unnecessary. The action continued only because of the transitory nature of pretrial detention, which placed the claim within "that *416 narrow class of cases" that are "distinctly `capable of repetition, yet evading review.'" 420 U. S., at 110, n. 11.[9] McDonald and Roper sanction some appeals from the denial of class certification notwithstanding satisfaction of the class representative's claim on the merits. But neither case holds that Art. III may be satisfied in the absence of a personal stake in the outcome. In McDonald, a putative class member intervened within the statutory time limit to appeal the certification ruling. 432 U. S., at 390.[10] Because the Court found that her claim was not time-barred, the intervenor in McDonald possessed the stake necessary to pursue the action. Indeed, the Court devoted its entire opinion to showing that the intervenor's claim for relief had not expired.[11] At most, McDonald holds only that an action which is kept alive by interested parties within prescribed periods of limitations does not "die" in an Art. III sense. There is dictum in McDonald that the "refusal to certify was subject to appellate review after final judgment at the behest of the named plaintiffs. . . ." 432 U. S., at 393. That gratuitous sentence, repeated in Coopers & Lybrand v. Livesay, *417 437 U. S. 463, 469, 470, n. 15 (1978), apparently is elevated by the Court's opinion in this case to the status of new doctrine. There is serious tension between this new doctrine and the much narrower reasoning adopted today in Roper. In Roper the Court holds that the named plaintiffs, who have refused to accept proffered individual settlements, retain a personal stake in sharing anticipated litigation costs with the class. Ante, at 334, n. 6, 336. Finding that Art. III is satisfied by this alleged economic interest, Roper reasons that the rules of federal practice governing appealability permit a party to obtain review of certain procedural rulings that are collateral to a generally favorable judgment. See ante, at 333-334, 336. The Court concludes that the denial of class certification falls within this category, as long as the named plaintiffs "assert a continuing stake in the outcome of the appeal." Ante, at 336. It is far from apparent how Roper can be thought to support the decision in this case. Indeed, the opinion by THE CHIEF JUSTICE in Roper reaffirms the obligation of a federal court to dismiss an appeal when the parties no longer retain the personal stake in the outcome required by Art. III. Ibid. Here, there is not even a speculative interest in sharing costs, and respondent affirmatively denies that he retains any stake or personal interest in the outcome of his appeal. See supra, at 413-414. Thus, a fact that was critical to the analysis in Roper is absent in this case. One can disagree with that analysis yet conclude that Roper affords no support for the Court's ruling here. B The cases cited by the Court as "less flexible"—and therefore less authoritative—apply established Art. III doctrine in cases closely analogous to this one. Indianapolis School Comm'rs v. Jacobs, 420 U. S. 128 (1975) (per curiam); Weinstein v. Bradford, 423 U. S. 147 (1975) (per curiam); Pasadena City Board of Education v. Spangler, 427 U. S. 424, 430 *418 (1976). As they are about to become second-class precedents, these cases are relegated to a footnote. Ante, at 400-401, n. 7. But the cases are recent and carefully considered decisions of this Court. They applied long-settled principles of Art. III jurisprudence. And no Justice who participated in them suggested the distinction drawn today. The Court's backhanded treatment of these "less flexible" cases ignores their controlling relevance to the issue presented here. In Jacobs, six named plaintiffs brought a class action to challenge certain high school regulations. The District Court stated on the record that class treatment was appropriate and that the plaintiffs were proper representatives, but the court failed to comply with Rule 23. After this Court granted review, we were informed that the named plaintiffs had graduated. We held that the action was entirely moot because the "class action was never properly certified nor the class properly identified by the District Court." 420 U. S., at 130.[12] Since the faulty certification prevented the class from acquiring separate legal status, Art. III required a dismissal. We reached precisely the same conclusion in Spangler, an action saved from mootness only by the timely intervention of a third party. 427 U. S., at 430-431. See also Baxter v. Palmigiano, 425 U. S. 308, 310, n. 1 (1976). And in Bradford, where the District Court had denied certification outright, the Court held that the named plaintiff's release from prison required the *419 dismissal of his complaint about parole release procedures. 423 U. S., at 149. See also Memphis Light, Gas & Water Div. v. Craft, 436 U. S. 1, 8 (1978). The Court suggests that Jacobs and Spangler may be distinguished because the plaintiffs there were not appealing the denial of class certification. The Court overlooks the fact that in each case the class representatives were defending a judgment on the merits from which the defendants had appealed. The plaintiffs/respondents continued vigorously to assert the claims of the class. They did not take the procedural route of appealing a denial of certification only because the District Court had granted—albeit defectively—class status. We chose not to remand for correction of the oral certification order in Jacobs because we recognized that the putative class representative had suffered no injury that could be redressed by adequate certification. Underlying Jacobs, and Bradford as well, is the elementary principle that no one has a personal stake in obtaining relief for third parties, through the mechanism of class certification or otherwise.[13] The Court rejects that principle today. III While the Court's new concept of "flexible" mootness is unprecedented, the content given that concept is even more disturbing. The Court splits the class aspects of this action into two separate "claims": (i) that the action may be maintained by respondent on behalf of a class, and (ii) that the class is entitled to relief on the merits. Since no class has been certified, the Court concedes that the claim on the merits is moot. Ante, at 404, 408. But respondent is said to *420 have a personal stake in his "procedural claim" despite his lack of a stake in the merits. The Court makes no effort to identify any injury to respondent that may be redressed by, or any benefit to respondent that may accrue from, a favorable ruling on the certification question.[14] Instead, respondent's "personal stake" is said to derive from two factors having nothing to do with concrete injury or stake in the outcome. First, the Court finds that the Federal Rules of Civil Procedure create a "right," "analogous to the private attorney general concept," to have a class certified. Second, the Court thinks that the case retains the "imperatives of a dispute capable of judicial resolution," which are identified as (i) a sharply presented issue, (ii) a concrete factual setting, and (iii) a self-interested party actually contesting the case. Ante, at 403.[15] *421 The Court's reliance on some new "right" inherent in Rule 23 is misplaced. We have held that even Congress may not confer federal-court jurisdiction when Art. III does not. Gladstone, Realtors v. Village of Bellwood, 441 U. S., at 100; O'Shea v. Littleton, 414 U. S., at 494, and n. 2; see Marbury v. Madison, 1 Cranch 137, 175-177 (1803). Far less so may a rule of procedure which "shall not be construed to extend . . . the jurisdiction of the United States district courts." Fed. Rule Civ. Proc. 82. Moreover, the "private attorney general concept" cannot supply the personal stake necessary to satisfy Art. III. It serves only to permit litigation by a party who has a stake of his own but otherwise might be barred by prudential standing rules. See Warth v. Seldin, 422 U. S., at 501; Sierra Club v. Morton, 405 U. S., at 737-738. Since neither Rule 23 nor the private attorney general concept can fill the jurisdictional gap, the Court's new perception of Art. III requirements must rest entirely on its tripartite test of concrete adverseness. Although the components of the test are no strangers to our Art. III jurisprudence, they operate only in "`cases confessedly within [the Court's] jurisdiction.'" Franks v. Bowman Transportation Co., 424 U. S., at 755-756, and n. 8, quoting Flast v. Cohen, 392 U. S. 83, 97 (1968). The Court cites no decision that has premised jurisdiction upon the bare existence of a sharply presented issue in a concrete and vigorously argued case, and I am aware of none.[16] Indeed, each of these characteristics is *422 sure to be present in the typical "private attorney general" action brought by a public-spirited citizen.[17] Although we have refused steadfastly to countenance the "public action," the Court's redefinition of the personal stake requirement leaves no principled basis for that practice.[18] The Court reasons that its departure from precedent is compelled by the difficulty of identifying a personal stake in a "procedural claim," particularly in "nontraditional forms of litigation." Ante, at 402. But the Court has created a false dilemma. As noted in Roper, class certification issues are "ancillary to the litigation of substantive claims." Ante, *423 at 332. Any attempt to identify a personal stake in such ancillary "claims" often must end in frustration, for they are not claims in any ordinary sense of the word. A motion for class certification, like a motion to join additional parties or to try the case before a jury instead of a judge, seeks only to present a substantive claim in a particular context. Such procedural devices generally have no value apart from their capacity to facilitate a favorable resolution of the case on the merits. Accordingly, the moving party is neither expected nor required to assert an interest in them independent of his interest in the merits. Class actions may advance significantly the administration of justice in appropriate cases. Indeed, the class action is scarcely a new idea. Rule 23 codifies, and was intended to clarify, procedures for dealing with a form of action long known in equity. See 1 H. Newberg, Class Actions § 1004 (1977). That federal jurisdiction can attach to the class aspect of litigation involving individual claims has never been questioned. But even when we deal with truly new procedural devices, our freedom to "adapt" Art. III is limited to the recognition of different "`means for presenting a case or controversy otherwise cognizable by the federal courts.'" Aetna Life Ins. Co. v. Haworth, 300 U. S., at 240 (Declaratory Judgment Act), quoting Nashville, C. & St. L. R. Co. v. Wallace, 288 U. S. 249, 264 (1933) (emphasis added). The effect of mootness on the vitality of a device like the class action may be a relevant prudential consideration.[19] But it *424 cannot provide a plaintiff when none is before the Court, for we are powerless to assume jurisdiction in violation of Art. III.[20] IV In short, this is a case in which the putative class representative— respondent here—no longer has the slightest interest in the injuries alleged in his complaint. No member of the class is before the Court; indeed, none has been identified. The case therefore lacks a plaintiff with the minimal personal stake that is a constitutional prerequisite to the jurisdiction of an Art. III court. In any realistic sense, the only persons before this Court who appear to have an interest are the defendants and a lawyer who no longer has a client.[21] I would vacate the decision of the Court of Appeals and remand with instructions to dismiss the action as moot. NOTES [*] Robert J. Hobbs filed a brief for the National Client Council, Inc., et al. as amici curiae urging affirmance. [1] The grant of certiorari also included the question of the validity of the Parole Release Guidelines, an issue left open in United States v. Addonizio, 442 U. S. 178, 184 (1979). We have concluded, however, that it would be premature to reach the merits of that question at this time. See infra, at 408. While the petition for a writ of certiorari was pending, respondent Geraghty filed a motion to substitute as respondents in this Court five prisoners, then incarcerated, who also were represented by Geraghty's attorneys. In the alternative, the prisoners sought to intervene. We deferred our ruling on the motion to the hearing of the case on the merits. 440 U. S. 945 (1979). These prisoners, or most of them, now also have been released from incarceration. On September 25, 1979, a supplement to the motion to substitute or intervene was filed, proposing six new substitute respondents or intervenors; each of these is a presently incarcerated federal prisoner who, allegedly, has been adversely affected by the guidelines and who is represented by Geraghty's counsel. Since we hold that respondent may continue to litigate the class certification issue, there is no need for us to consider whether the motion should be granted in order to prevent the case from being moot. We conclude that the District Court initially should rule on the motion. [2] See, e. g., Armour v. City of Anniston, 597 F. 2d 46, 48-49 (CA5 1979); Susman v. Lincoln American Corp., 587 F. 2d 866 (CA7 1978), cert. pending, No. 78-1169; Goodman v. Schlesinger, 584 F. 2d 1325, 1332-1333 (CA4 1978); Camper v. Calumet Petrochemicals, Inc., 584 F. 2d 70 (CA5 1978); Roper v. Consurve, Inc., 578 F. 2d 1106 (CA5 1978), aff'd sub nom. Deposit Guaranty Nat. Bank v. Roper, ante, p. 326; Satterwhite v. City of Greenville, 578 F. 2d 987 (CA5 1978) (en banc), cert. pending, No. 78-1008; Vun Cannon v. Breed, 565 F. 2d 1096 (CA9 1977); Winokur v. Bell Federal Savings & Loan Assn., 560 F. 2d 271 (CA7 1977), cert. denied, 435 U. S. 932 (1978); Lasky v. Quinlan, 558 F. 2d 1133 (CA2 1977); Kuahulu v. Employers Ins. of Wausau, 557 F. 2d 1334 (CA9 1977); Boyd v. Justices of Special Term, 546 F. 2d 526 (CA2 1976); Napier v. Gertrude, 542 F. 2d 825 (CA10 1976), cert. denied, 429 U. S. 1049 (1977). [3] 38 Fed. Reg. 31942-31945 (1973). The guidelines currently in force appear at 28 CFR § 2.20 (1979). [4] The extortion count was based on respondent's use of his position as a vice squad officer of the Chicago police force to "shake down" dispensers of alcoholic beverages; the false declarations concerned his involvement in this scheme. [5] Apparently Becher, too, has now been released from prison. [6] The claim in Sosna also fit the traditional category of actions that are deemed not moot despite the litigant's loss of personal stake, that is, those "capable of repetition, yet evading review." See Southern Pacific Terminal Co. v. ICC, 219 U. S. 498, 515 (1911). In Franks v. Bowman Transportation Co., 424 U. S. 747, 753-755 (1976), however, the Court held that the class-action aspect of mootness doctrine does not depend on the class claim's being so inherently transitory that it meets the "capable of repetition, yet evading review" standard. [7] Three of the Court's cases might be described as adopting a less flexible approach. In Indianapolis School Comm'rs v. Jacobs, 420 U. S. 128 (1975), and in Weinstein v. Bradford, 423 U. S. 147 (1975), dismissal of putative class suits, as moot, was ordered after the named plaintiffs' claims became moot. And in Pasadena City Bd. of Education v. Spangler, 427 U. S. 424, 430 (1976), it was indicated that the action would have been moot, upon expiration of the named plaintiffs' claims, had not the United States intervened as a party plaintiff. Each of these, however, was a case in which there was an attempt to appeal the merits without first having obtained proper certification of a class. In each case it was the defendant who petitioned this Court for review. As is observed subsequently in the text, appeal from denial of class classification is permitted in some circumstances where appeal on the merits is not. In the situation where the proposed class representative has lost a "personal stake," the merits cannot be reached until a class properly is certified. Although the Court perhaps could have remanded Jacobs and Weinstein for reconsideration of the class certification issue, as the Court of Appeals did here, the parties in those cases did not suggest "relation back" of class certification. Thus we do not find this line of cases dispositive of the question now before us. [8] Were the class an indispensable party, the named plaintiff's interests in certification would approach a "legally cognizable interest." [9] See, e. g., Landers, Of Legalized Blackmail and Legalized Theft: Consumer Class Actions and the Substance-Procedure Dilemma, 47 S. Cal. L. Rev. 842 (1974); Simon, Class Actions—Useful Tool or Engine of Destruction, 55 F. R. D. 375 (1972). [10] We intimate no view as to whether a named plaintiff who settles the individual claim after denial of class certification may, consistent with Art. III, appeal from the adverse ruling on class certification. See United Airlines, Inc. v. McDonald, 432 U. S. 385, 393-394, and n. 14 (1977). [11] MR. JUSTICE POWELL, in his dissent, advocates a rigidly formalistic approach to Art. III, post, at 412, and suggests that our decision today is the Court's first departure from the formalistic view. Post, at 414-419. We agree that the issue at hand is one of first impression and thus, in that narrow sense, is "unprecedented," post, at 419. We do not believe, however, that the decision constitutes a redefinition of Art. III principles or a "significant departur[e]," post, at 409, from "carefully considered" precedents, post, at 418. The erosion of the strict, formalistic perception of Art. III was begun well before today's decision. For example, the protestations of the dissent are strikingly reminiscent of Mr. Justice Harlan's dissent in Flast v. Cohen, 392 U. S. 83, 116, in 1968. Mr. Justice Harlan hailed the taxpayer-standing rule pronounced in that case as a "new doctrine" resting "on premises that do not withstand analysis." Id., at 117. He felt that the problems presented by taxpayer standing "involve nothing less than the proper functioning of the federal courts, and so run to the roots of our constitutional system." Id., at 116. The taxpayers were thought to complain as "private attorneys-general," and "[t]he interests they represent, and the rights they espouse, are bereft of any personal or proprietary coloration." Id., at 119. Such taxpayer actions "are and must be . . . `public actions' brought to vindicate public rights." Id., at 120. Notwithstanding the taxpayers' lack of a formalistic "personal stake," even Mr. Justice Harlan felt that the case should be held nonjusticiable on purely prudential grounds. His interpretation of the cases led him to conclude that "it is . . . clear that [plaintiffs in a public action] as such are not constitutionally excluded from the federal courts." Ibid. (emphasis in original). Is it not somewhat ironic that MR. JUSTICE POWELL, who now seeks to explain United Airlines, Inc. v. McDonald, supra, as a straightforward application of settled doctrine, post, at 416-417, expressed in his dissent in McDonald, 432 U. S., at 396, the view that the holding rested on a fundamental misconception about the mootness of an uncertified class action after settlement of the named plaintiffs' claims? He stated: "Pervading the Court's opinion is the assumption that the class action somehow continued after the District Court denied class status. But that assumption is supported neither by the text nor by the history of Rule 23. To the contrary, . . . the denial of class status converts the litigation to an ordinary nonclass action." Id., at 399. The dissent went on to say: "[Petitioner] argues with great force that, as a result of the settlement of their individual claims, the named plaintiffs `could no longer appeal the denial of class' status that had occurred years earlier. . . . Although this question has not been decided by this Court, the answer on principle is clear. The settlement of an individual claim typically moots any issues associated with it. . . . This case is sharply distinguishable from cases such as Sosna v. Iowa . . . and Franks v. Bowman Transp. Co. . . . where we allowed named plaintiffs whose individual claims were moot to continue to represent their classes. In those cases, the District Courts previously had certified the classes, thus giving them `a legal status separate from the interest[s] asserted by [the named plaintiffs]' Sosna v. Iowa, supra, at 399. This case presents precisely the opposite situation: The prior denial of class status had extinguished any representative capacity." Id., at 400 (footnote omitted). Thus, the assumption thought to be "[p]ervading the Court's opinion" in McDonald, and so vigorously attacked by the dissent there, is now relegated to "gratuitous" "dictum," post, at 416. MR. JUSTICE POWELL, who finds the situation presented in the case at hand "fundamentally different" from that in Sosna and Franks, post, at 413, also found the facts of McDonald "sharply distinguishable" from those previous cases. 432 U. S., at 400. We do not recite these cases for the purpose of showing that our result is mandated by the precedents. We concede that the prior cases may be said to be somewhat confusing, and that some, perhaps, are irreconcilable with others. Our point is that the strict, formalistic view of Art. III jurisprudence, while perhaps the starting point of all inquiry, is riddled with exceptions. And, in creating each exception, the Court has looked to practicalities and prudential considerations. The resulting doctrine can be characterized, aptly, as "flexible"; it has been developed, not irresponsibly, but "with some care," post, at 410, including the present case. The dissent is correct that once exceptions are made to the formalistic interpretation of Art. III, principled distinctions and bright lines become more difficult to draw. We do not attempt to predict how far down the road the Court eventually will go toward premising jurisdiction "upon the bare existence of a sharply presented issue in a concrete and vigorously argued case," post, at 421. Each case must be decided on its own facts. We hasten to note, however, that this case does not even approach the extreme feared by the dissent. This respondent suffered actual, concrete injury as a result of the putatively illegal conduct, and this injury would satisfy the formalistic personal-stake requirement if damages were sought. See, e. g., Powell v. McCormack, 395 U. S., at 495-500. His injury continued up to and beyond the time the District Court denied class certification. We merely hold that when a District Court erroneously denies a procedural motion, which, if correctly decided, would have prevented the action from becoming moot, an appeal lies from the denial and the corrected ruling "relates back" to the date of the original denial. The judicial process will not become a vehicle for "concerned bystanders," post, at 413, even if one in respondent's position can conceivably be characterized as a bystander, because the issue on the merits will not be addressed until a class with an interest in the outcome has been certified. The "relation back" principle, a traditional equitable doctrine applied to class certification claims in Gerstein v. Pugh, 420 U. S. 103 (1975), serves logically to distinguish this case from the one brought a day after the prisoner is released. See post, at 420-421, n. 15. If the named plaintiff has no personal stake in the outcome at the time class certification is denied, relation back of appellate reversal of that denial still would not prevent mootness of the action. [12] See, e. g., Comment, A Search for Principles of Mootness in the Federal Courts: Part Two—Class Actions, 54 Texas L. Rev. 1289, 1331-1332 (1976); Comment, Continuation and Representation of Class Actions Following Dismissal of the Class Representative, 1974 Duke L. J. 573, 602-608. [1] See, e. g., Duke Power Co. v. Carolina Environmental Study Group, Inc., 438 U. S. 59, 72 (1978); Arlington Heights v. Metropolitan Housing Dev. Corp., 429 U. S. 252, 260-261 (1977); Warth v. Seldin, 422 U. S. 490, 499 (1975); Linda R. S. v. Richard D., 410 U. S. 614, 617 (1973). Each of these cases rejects the view, once expressed by Mr. Justice Harlan and now apparently espoused by the Court, that the personal stake requirement lacks constitutional significance. Ante, at 404-407, n. 11; Flast v. Cohen, 392 U. S. 83, 120 (1968) (Harlan, J., dissenting); see also United States v. Richardson, 418 U. S. 166, 180 (1974) (POWELL, J., concurring). Until today, however, that view never had commanded a majority. [2] See, e. g., Schlesinger v. Reservists to Stop the War, 418 U. S. 208, 227 (1974); O'Shea v. Littleton, 414 U. S. 488, 494 (1974); Moose Lodge No. 107 v. Irvis, 407 U. S. 163, 166-167 (1972); Sierra Club v. Morton, 405 U. S. 727, 736-738 (1972); Tileston v. Ullman, 318 U. S. 44, 46 (1943) (per curiam). The rule is the same when the question is mootness and a litigant can assert no more than emotional involvement in what remains of the case. Ashcroft v. Mattis, 431 U. S. 171, 172-173 (1977) (per curiam). [3] See 13 C. Wright, A. Miller, & E. Cooper, Federal Practice and Procedure § 3533, p. 265 (1975); Note, The Mootness Doctrine in the Supreme Court, 88 Harv. L. Rev. 373, 376-377 (1974). [4] See, e. g., Preiser v. Newkirk, 422 U. S. 395, 401-402 (1975); SEC v. Medical Comm. for Human Rights, 404 U. S. 403, 407 (1972); Powell v. McCormack, 395 U. S. 486, 496, n. 7 (1969); Liner v. Jafco, Inc., 375 U. S. 301, 306, n. 3 (1964). [5] The Court states that "the erosion of the strict, formalistic perception of Art. III was begun well before today's decision," and that the Art. III personal stake requirement is "riddled with exceptions." Ante, at 404-405, 406, n. 11. It fails, however, to cite a single Court opinion in support of either statement. To the extent that the decision in Flast v. Cohen, supra, supports the position ascribed to it in the dissent, 392 U. S., at 117-120, it does not survive the long line of express holdings that began with Warth v. Seldin, supra, and were reaffirmed only last Term. Gladstone, Realtors v. Village of Bellwood, 441 U. S. 91, 99 (1979). See nn. 1 and 2, supra. Even before Warth, Professor Davis observed that the personal stake requirement had no exceptions. 35 U. Chi. L. Rev., at 616, 617. [6] No one suggests that respondent could be affected personally by any ruling on the class certification question that is remanded today. In fact, the Court apparently concedes that respondent has no personal stake—"in the traditional sense"—in obtaining certification. Ante, at 402. Several prisoners now in federal custody have filed a motion to intervene as parties respondent in this Court. Although the Court does not rule on that motion, I note that the motion was received well over a year after respondent was released from prison. In the interim, respondent obtained a ruling from the Court of Appeals and filed his petition for certiorari in this Court. Such untimely intervention comes too late to save the action under United Airlines, Inc. v. McDonald, 432 U. S. 385 (1977). [7] Respondent's lawyer opened his argument by saying that "[t]he mootness question in this case is, from a practical standpoint, not very significant." If the action is dismissed as moot he plans simply to "file a new case" on behalf of prisoners serving longer terms. Tr. of Oral Arg. 25. On the basis of this representation by counsel, there is reason to believe that members of the putative class at issue ultimately will be included in a class action that will not moot out. [8] Certification is no mere formality. It represents a judicial finding that injured parties other than the named plaintiff exist. It also provides a definition by which they can be identified. Certification identifies and sharpens the interests of unnamed class members in the outcome; only thereafter will they be bound by the outcome. After certification, class members can be certain that the action will not be settled or dismissed without appropriate notice. Fed. Rule Civ. Proc. 23 (c); 3 H. Newberg, Class Actions § 5050 (1977); cf. Almond, Settling Rule 23 Class Actions at the Precertification Stage: Is Notice Required?, 56 N. C. L. Rev. 303 (1978). Vigorous advocacy is assured by the authoritative imposition on the named plaintiffs of a duty adequately to represent the entire class. If the named plaintiff's own claim becomes moot after certification, the court can re-examine his ability to represent the interests of class members. Should it be found wanting, the court may seek a substitute representative or even decertify the class. Fed. Rules Civ. Proc. 23 (c) (1), 23 (d); see 1 Newberg, supra, § 2192; Comment, Continuation and Representation of Class Actions Following Dismissal of the Class Representative, 1974 Duke L. J. 573, 589-590, 602-603. After certification, the case is no different in principle from more traditional representative actions involving, for example, a single party who cannot participate himself because of his incompetence but is permitted to litigate through an appointed fiduciary. [9] The Court's Gerstein analysis, which emphasized that "[p]retrial detention is by nature temporary" and that "[t]he individual could. . . suffer repeated deprivations" with no access to redress, falls squarely within the rule of Southern Pacific Terminal Co. v. ICC, 219 U. S. 498, 515 (1911). See Roe v. Wade, 410 U. S. 113, 125 (1973). In similar cases we have noted that the continuation of the action will depend "`especially [upon] the reality of the claim that otherwise the issue would evade review.'" Swisher v. Brady, 438 U. S. 204, 213, n. 11 (1978), quoting Sosna v. Iowa, 419 U. S. 393, 402, n. 11 (1975). These limitations are inconsistent with the concept of "flexible" mootness and the redefinition of "personal stake" adopted today. [10] The individual claims of the original named plaintiffs had been settled after judgment on the question of liability. 432 U. S., at 389, 393, n. 14. [11] This extensive inquiry would have been unnecessary if, as the Court holds today, the intervenor had a personal stake in the class certification issue itself. Since the present respondent's claim long since has "expired," he stands in the same position as a member of the putative class whose claim has "expired" by reason of the statute of limitations. [12] The vitality of the Jacobs result is underscored by the repeated dictum that a properly certified class is necessary to supply adverseness once the named plaintiff's claim becomes moot. East Texas Motor Freight v. Rodriguez, 431 U. S. 395, 406, n. 12 (1977); Franks v. Bowman Transportation Co., 424 U. S. 747, 754, n. 6, 755-756 (1976); see Kremens v. Bartley, 431 U. S. 119, 129-130 (1977); Richardson v. Ramirez, 418 U. S. 24, 39 (1974). Conversely, we have often stated that the named plaintiff's individual claim must be a live one both at the time the action is filed and at the time of certification. Kremens v. Bartley, supra, at 143, n. 6 (BRENNAN, J., dissenting); Sosna v. Iowa, supra, at 402, 403; see Bell v. Wolfish, 441 U. S. 520, 526, n. 5 (1979); Zablocki v. Redhail, 434 U. S. 374, 382, n. 9 (1978). [13] In some circumstances, litigants are permitted to argue the rights of third parties in support of their claims. E. g., Singleton v. Wulff 428 U. S. 106, 113 (1976); Barrows v. Jackson, 346 U. S. 249, 255-256 (1953). In each such case, however, the Court has identified a concrete, individual injury suffered by the litigant himself. Ibid.; see n. 2, supra, and accompanying text. [14] In a footnote, ante, at 406, n. 11, the Court states: "This respondent suffered actual, concrete injury as a result of the putatively illegal conduct, and this injury would satisfy the formalistic personal-stake requirement if damages were sought. See, e. g., Powell v. McCormack, 395 U. S., at 495-500." This appears to be a categorical claim of the actual, concrete injury our cases have required. Yet, again, the Court fails to identify the injury. The reference to damages is irrelevant here, as respondent sought no damages—only injunctive and declaratory relief. Moreover, counsel for respondent frankly conceded that his client "can obtain absolutely no additional personal relief" in this case. Tr. of Oral Arg. 25. If the Court seriously is claiming concrete injury "at all stages of review," see supra, at 411, it would be helpful for it to identify specifically this injury that was not apparent to respondent's counsel. Absent such identification, the claim of injury is indeed an empty one. [15] The Court attempts to limit the sweeping consequences that could flow from the application of these criteria, see infra, at 421-422, and n. 18, by asserting that "[e]ach case must be decided on its own facts" on the basis of "practicalities and prudential considerations." Ante, at 406, n. 11. The Court long has recognized a difference between the prudential and constitutional aspects of the standing and mootness doctrines. See supra, at 410. I am not aware that the Court, until today, ever has merged these considerations for the purpose of eliminating the Art. III requirement of a personal stake in the litigation. The Court cites no prior case for this view. Moreover, the Court expounds no limiting principle of any kind. Adverse practical consequences, even if relevant to Art. III analysis, cannot justify today's holding as none whatever would flow from a finding of mootness. See n. 18, infra. Nor does the Court's reliance upon a "`relation back' principle," ante, at 407, n. 11, further the analysis. Although this fiction may provide a shorthand label for the Court's conclusion, it is hardly a principle and certainly not a limiting one. [16] The Court often has rejected the contention that a "spirited dispute" alone is sufficient to confer jurisdiction. E. g., Richardson v. Ramirez, 418 U. S., at 35-36; Hall v. Beals, 396 U. S. 45, 48-49 (1969) (per curiam). [17] The Court's assertion to the contrary notwithstanding, there is nothing in the record to suggest that respondent has any interest whatever in his new-found "right to have a class certified." Ante, at 403. In fact, the record shows that respondent's interest in the merits was the sole motivation for his attempt to represent a class. The class claims were added to his complaint only because his lawyer feared that mootness might terminate the action. App. 17; Brief for Respondent 23, 33. The record does not reveal whether respondent—as distinguished from his lawyer—now wishes to continue with the case. If he does, it is clear that his interest has nothing to do with the procedural protections described by the Court as the "primary benefits of class suits." Ante, at 403. It is neither surprising nor improper that respondent should be concerned with parole procedures. But respondent's actual interest is indistinguishable from the generalized interest of a "private attorney general" who might bring a "public action" to improve the operation of a parole system. [18] The Court's view logically cannot be confined to moot cases. If a plaintiff who is released from prison the day after filing a class action challenging parole guidelines may seek certification of the class, why should a plaintiff who is released the day before filing the suit be barred? As an Art. III matter, there can be no difference. Even on prudential grounds, there is little difference between this action and one filed promptly after the named plaintiff's release from prison. In the present case, this Court has ruled on neither the merits nor the propriety of the class action. At the same time, it has vacated a judgment by the Court of Appeals that in turn reversed the judgment of the District Court. No determination on any issue is left standing. For every practical purpose, the action must begin anew—this time without a plaintiff. The prudential considerations in favor of a finding of mootness could scarcely be more compelling. [19] I do not imply that the result reached today is necessary in any way to the continued vitality of the class action device. On the contrary, the practical impact of mootness in this case would be slight indeed. See n. 18, supra. And this may well be typical of class actions brought under Rule 23 (b) (1) or (2) to seek injunctive or declaratory relief. Such actions are not subject to frustration through sequential settlement offers that "buy off" each intervening plaintiff. Cf. Deposit Guaranty Nat. Bank v. Roper, ante, at 339. Nor will substitute plaintiffs be deterred by the notice costs that attend certification of a class under Rule 23 (b) (3). [20] The Court's efforts to "save" this action from mootness lead it to depart strikingly from the normal role of a reviewing court. The Court fails to identify how, if at all, the District Court has erred. Nothing is said about the District Court's ruling on the merits or its refusal to certify the broad class sought by respondent. Nor does the Court adopt the Court of Appeals' conclusion that the District Court erred in failing to consider the possibility of subclasses sua sponte. Nevertheless, respondent— or his lawyer—is given the opportunity to raise the subclass question on remand. That result cannot be squared with the rule that a litigant may not raise on appeal those issues he has failed to preserve by appropriate objection in the trial court. The Court intimates that the District Court waited too long to deny the class certification motion, thus making a motion for subclasses a "futile act." Ante, at 408. But nothing in the record suggests that the District Court would not have entertained such a motion. Since respondent sought certification in the first place only to avoid mootness on appeal, the entry of an order against him on the merits provides no excuse for his subsequent failure to present a subclass proposal to the District Court. [21] I imply no criticism of counsel in this case. The Court of Appeals agreed with counsel that the certification issue was appealable, and the case was brought to this Court by the United States.