dissenting.
On December 19, 1991, Congress enacted § 27A of the Securities Exchange Act of 1934, 15 U. S. C. § 78aa-1 (1988 ed., Supp. V) (hereinafter 1991 amendment), to remedy a flaw in the limitations rule this Court announced on June 20, 1991, in Lampf, Pleva, Lipkind, Prupis & Petigrow v. Gilbertson, 501 U. S. 350 (1991). In Lampf the Court replaced the array of state statutes of limitations that had governed shareholder actions under the Securities Exchange Act of 1934, 15 U.S.C. § 78j(b), and Rule 10b-5, 17 CFR 240.10b-5 (1994) (hereinafter 10b-5 actions), with a uniform federal limitations rule. Congress found only one flaw in the Court’s new rule: its failure to exempt pending cases from its operation. Accordingly, without altering the prospective effect of the Lampf rule, the 1991 amendment remedied its flaw by providing that pre-Lampf law should determine the limitations period applicable to all cases that had *247been pending on June 20, 1991 — both those that remained pending on December 19,1991, when § 27A was enacted, and those that courts dismissed between June 20 and December 19, 1991. Today the Court holds that the 1991 amendment violates the Constitution’s separation of powers because, by encompassing the dismissed claims, it requires courts to reopen final judgments in private civil actions.
Section 27A is a statutory amendment to a rule of law announced by this Court. The fact that the new rule announced in Lampf was a product of judicial, rather than congressional, lawmaking should not affect the separation-of-powers analysis. We would have the same issue to decide had Congress enacted the Lampf rule but, as a result of inadvertence or perhaps a scrivener’s error, failed to exempt pending cases, as is customary when limitations periods are shortened.1 In my opinion, if Congress had retroactively restored rights its own legislation had inadvertently or unfairly impaired, the remedial amendment’s failure to exclude dismissed cases from the benefited class would not make it invalid. The Court today faces a materially identical situation and, in my view, reaches the wrong result.
Throughout our history, Congress has passed laws that allow courts to reopen final judgments. Such laws characteristically apply to judgments entered before as well as after their enactment. When they apply retroactively, they may raise serious due process questions,2 but the Court *248has never invalidated such a law on separation-of-powers grounds until today. Indeed, only last Term we recognized Congress’ ample power to enact a law that “in effect ‘restored’ rights that [a party] reasonably and in good faith thought he possessed before the surprising announcement” of a Supreme Court decision. Rivers v. Roadway Express, Inc., 511 U. S. 298, 310 (1994) (discussing Frisbie v. Whitney, 9 Wall. 187 (1870)). We conditioned our unambiguous restatement of the proposition that “Congress had the power to enact legislation that had the practical effect of restoring the status quo retroactively,” 511 U. S., at 310, only on Congress’ clear expression of its intent to do so.
A large class of investors reasonably and in good faith thought they possessed rights of action before the surprising announcement of the Lampf rule on June 20,1991. When it enacted the 1991 amendment, Congress clearly expressed its intent to restore the rights Lampf had denied the aggrieved class. Section 27A comported fully with Rivers and with other precedents in which we consistently have recognized Congress’ power to enact remedial statutes that set aside classes of final judgments. The only remarkable feature of *249this enactment is the fact that it remedied a defect in a new judge-made rule rather than in a statute.
The familiar history the Court invokes, involving colonial legislatures’ ad hoc decisions of individual cases, “ ‘unfettered by rules,’ ” ante, at 220 (quoting Vermont State Papers 1779-1786, p. 540 (Slade ed. 1823)), provides no support for its holding. On the contrary, history and precedent demonstrate that Congress may enact laws that establish both substantive rules and procedures for reopening final judgments. When it enacted the 1991 amendment to the Lampf rule, Congress did not encroach on the judicial power. It decided neither the merits of any 10b-5 claim nor even whether any such claim should proceed to decision on the merits. It did provide that the rule governing the timeliness of 10b-5 actions pending on June 19, 1991, should be the pre-Lampf statute of limitations, and it also established a procedure for Article III courts to apply in determining whether any dismissed case should be reinstated. Congress’ decision to extend that rule and procedure to 10b-5 actions dismissed during the brief period between this Court’s law-changing decision in Lampf and Congress’ remedial action is not a sufficient reason to hold the statute unconstitutional.
I
Respondents conducted a public offering of common stock in 1983. Petitioners, suing on behalf of themselves and other purchasers of the stock, filed a 10b-5 action in 1987 in the United States District Court for the Eastern District of Kentucky, alleging violations of substantive federal rules that had been in place since 1934. Respondents moved to dismiss the complaint as untimely because petitioners had filed it more than three years after the events in dispute. At that time, settled law in Kentucky and elsewhere in the United States directed federal courts to determine statutes of limitations applicable to 10b-5 actions by reference to *250state law.3 The relevant Kentucky statute provided a 3-year limitations period,4 which petitioners contended ran from the time the alleged fraud was or should have been discovered. A Magistrate agreed with petitioners and recommended denial of respondents’ motion to dismiss, but by 1991 the District Court had not yet ruled on that issue. The factual question whether petitioners should have discovered respondents’ alleged 10b-5 violations more than three years before they filed suit remained open for decision by an Article III judge on June 20, 1991.
On that day, this Court’s decision in Lampf changed the law. The Court concluded that every 10b-5 action is time barred unless brought within three years of the alleged violation and one year of its discovery. Moreover, it applied that novel rule to pending cases. As Justice O’Connor pointed out in her dissent, the Court held the plaintiffs’ suit “time barred under a limitations period that did not exist before,” a holding that “departed] drastically from our established practice and inflict[ed] an injustice on the [plaintiffs].” Lampf, 501 U. S., at 369.5 The inequitable consequences of Lampf reached beyond the parties to that case, *251injuring a large class of litigants that includes petitioners. Without resolving the factual issue that would have determined the timeliness of petitioners’ complaint before Lampf the District Court dismissed the instant action as untimely under the new limitations period dictated by this Court. Because Lampf had deprived them of any nonfrivolous basis for an appeal, petitioners acquiesced in the dismissal, which therefore became final on September 12,1991.
Congress responded to Lampf by passing §27A, which became effective on December 19, 1991. The statute changed the substantive limitations law, restoring the pre-Lamp/limitations rule for two categories of 10b-5 actions that had been pending on June 19, 1991. Subsection (a) of § 27A applies to cases that were still pending on December 19, 1991. The Courts of Appeals have uniformly upheld the constitutionality of that subsection,6 and its validity is not challenged in this case. Subsection (b) applies to actions, like the instant case, that (1) were dismissed after June 19, 1991, and (2) would have been timely under the pr e-Lampf regime. This subsection authorized the district courts to reinstate dismissed cases if the plaintiff so moved within 60 days after the effective date of §27A. The amendment was not self-executing: Unless the plaintiff both filed a timely motion for reinstatement and then satisfied the court that the complaint had been timely filed under applicable pr e-Lampf law, the dismissal would remain in effect.
In this case petitioners made the required showing, but the District Court refused to reinstate their case. Instead, *252it held § 27A(b) unconstitutional. 789 F. Supp. 231 (ED Ky. 1992). The Court of Appeals for the Sixth Circuit, contrary to an earlier decision of the Fifth Circuit, affirmed. 1 F. 3d 1487 (1993).
II
Aside from §27A(b), the Court claims to “know of no instance in which Congress has attempted to set aside the final judgment of an Article III court by retroactive legislation.” Ante, at 230. In fact, Congress has done so on several occasions. Section 27A(b) is part of a remedial statute. As early as 1833, we recognized that a remedial statute authorizing the reopening of a final judgment after the time for appeal has expired is “entirely unexceptionable” even though it operates retroactively. “It has been repeatedly decided in this court, that the retrospective operation of such a law forms no objection to it. Almost every law, providing a new remedy, affects and operates upon causes of action existing at the time the law is passed.” Sampeyreac v. United States, 7 Pet. 222, 239 (1833). We have upheld remedial statutes that carried no greater cause for separation-of-powers concerns than does §27A(b); others have provoked no challenges. In contrast, the colonial directives on which the majority relies were nothing like remedial statutes.
The remedial 1830 law we construed in Sampeyreac strongly resembled § 27A(b): It authorized a class of litigants to reopen claims, brought under an 1824 statute, that courts had already finally adjudicated. The 1824 statute authorized proceedings to establish title to certain lands in the State of Missouri and the territory of Arkansas. It provided for an appeal to this Court within one year after the entry of the judgment or decree, “and should no appeal be taken, the judgment or decree of the district court shall in like manner be final and conclusive.” 7 Pet., at 238. In 1827 the Arkansas Territorial Court entered a decree in favor of one Sampeyreac, over the objection of the United States that the nominal plaintiff was a fictitious person. Because no appeal *253was taken from that decree, it became final in 1828. In 1830 Congress passed a special statute authorizing the Arkansas court to reopen any decree entered under the 1824 statute if, prior to July 1, 1831, the United States filed a bill of review alleging that the decree had been based on forged evidence of title. The United States filed such a bill and obtained a reversal of the 1827 decree from the Arkansas court.
The successors in interest of the fictitious Mr. Sampeyreae argued in this Court that the Arkansas court should not have entertained the Government’s bill of review because the 1830 statute “was the exercise of a judicial power, and it is no answer to this objection, that the execution of its provisions is given to a court. The legislature of the union cannot use such a power.” Id., at 229. We categorically rejected that argument: “The law of 1830 is in no respect the exercise of judicial powers.” Id., at 239. Of course, as the majority notes, ante, at 232-233, the particular decree at stake in Sampeyreae had issued not from an Article III court but from a territorial court. However, our opinion contains no suggestion that Congress’ power to authorize the reopening of judgments entered by the Arkansas court was any broader than its power to authorize the reopening of judgments entered under the same statute by the United States District Court in Missouri. Moreover, the relevant judicial power that the 1830 statute arguably supplanted was this Court’s Article III appellate jurisdiction — which, prior to the 1830 enactment, provided the only avenue for review of the trial courts’ judgments.
Similarly, in Freeborn v. Smith, 2 Wall. 160 (1865), the Court rejected a challenge to an Act of Congress that removed an accidental impediment to the exercise of our appellate jurisdiction. When Congress admitted Nevada into the Union as a State in March 1864, ch. 36, 13 Stat. 30, it neglected to provide for the disposition of pending appeals from final judgments previously entered by the Supreme Court of *254the Nevada Territory. Accordingly, the Freeborn defendants in error moved to dismiss a writ of error to the territorial court on the ground that we had no power to decide the case. At the suggestion of plaintiffs in error, the Court deferred ruling on the motion until after February 27, 1865, when Congress passed a special statute that authorized the Court to decide this and similar cases.7 Defendants in error renewed their motion, arguing that Congress could not reopen judgments that were already final and unreviewable because Congress was not competent to exercise judicial power.
Defendants in error argued that, “[i]f it be possible for a right to attach itself to a judgment, it has done so here, and there could not be a plainer case of an attempt to destroy it by legislative action.” 2 Wall., at 165. The Court, however, noted that the omission in the 1864 statute had left the case “in a very anomalous situation,” id., at 174, and that passage of the later statute “was absolutely necessary to remove an impediment in the way of any legal proceeding in the case.” Id., at 175. It concluded that such “acts are of a remedial character, and are peculiar subjects of legislation. They are not liable to the imputation of being assumptions of judicial power.” Ibid. As in Sampeyreac, although Freeborn in*255volved the review of a judgment entered by a territorial court, the “judicial power” to which the opinion referred was this Court’s Article III appellate jurisdiction. If Congress may enact a law authorizing this Court to reopen decisions that we previously lacked power to review, Congress must have the power to let district courts reopen their own judgments.
Also apposite is United States v. Sioux Nation, 448 U. S. 371 (1980), which involved the Sioux Nation’s longstanding claim that the Government had in 1877 improperly abrogated the treaty by which the Sioux had held title to the Black Hills. The Sioux first brought their claim under a special 1920 jurisdictional statute. The Court of Claims dismissed the suit in 1942, holding that the 1920 Act did not give the court jurisdiction to consider the adequacy of the compensation the Government had paid in 1877. Congress passed a new jurisdictional statute in 1946, and in 1950 the Sioux brought a new action. In 1975 the Court of Claims, although acknowledging the merit of the Sioux’s claim, held that the res judicata effect of the 1942 dismissal barred the suit. In response, Congress passed a statute in 1978 that authorized the Court of Claims to take new evidence and instructed it to consider the Sioux’s claims on the merits, disregarding res judicata. The Sioux finally prevailed. We held that the 1978 Act did not violate the separation of powers. 448 U. S., at 407.
The Court correctly notes, see ante, at 230-231, and n. 5, that our opinion in Sioux Nation prominently discussed precedents establishing Congress’ power to waive the res judicata effect of judgments against the United States. We never suggested, however, that those precedents sufficed to overcome the separation-of-powers objections raised against the 1978 Act. Instead, we made extensive comments about the propriety of Congress’ action that were as necessary to our holding then as they are salient to the Court’s analysis today. In passing the 1978 Act, we held, Congress
*256“only was providing a forum so that a new judicial review of the Black Hills claim could take place. This review was to be based on the facts found by the Court of Claims after reviewing all the evidence, and an application of generally controlling legal principles to those facts. For these reasons, Congress was not reviewing the merits of the Court of Claims’ decisions, and did not interfere with the finality of its judgments.
“Moreover, Congress in no way attempted to prescribe the outcome of the Court of Claims’ new review of the merits.” 448 U. S., at 407.
Congress observed the same boundaries in enacting §27A(b).
Our opinions in Sampeyreac, Freeborn, and Sioux Nation correctly characterize statutes that specify new grounds for the reopening of final judgments as remedial. Moreover, these precedents correctly identify the unremarkable nature of the legislative power to enact remedial statutes. “[A]cts . . . of a remedial character . . . are the peculiar subjects of legislation. They are not liable to the imputation of being assumptions of judicial power.” Freeborn, 2 Wall., at 175. A few contemporary examples of such statutes will suffice to demonstrate that they are ordinary products of the exercise of legislative power.
The most familiar remedial measure that provides for reopening of final judgments is Rule 60(b) of the Federal Rules of Civil Procedure. That Rule both codified common-law grounds for relieving a party from a final judgment and added an encompassing reference to “any other reason justifying relief from the operation of the judgment.”8 Not a *257single word in its text suggests that it does not apply to judgments entered prior to its effective date. On the contrary, the purpose of the Rule, its plain language, and the traditional construction of remedial measures all support construing it to apply to past as well as future judgments. Indeed, because the Rule explicitly abolished the common-law writs it replaced, an unintended gap in the law would have resulted if it did not apply retroactively.9
*258Other examples of remedial statutes that resemble §27A include the Soldiers’ and Sailors’ Civil Relief Act of 1940, 50 U. S. C. App. § 520(4), which authorizes members of the Armed Forces to reopen judgments entered while they were on active duty; the Handicapped Children’s Protection Act of 1986, 20 U. S. C. § 1415(e)(4)(B) (1988 ed. and Supp. V), which provided for recovery of attorney’s fees under the Education for All Handicapped Children Act of 1975, 20 U. S. C. § 1411 et seq. (1988 ed. and Supp. V);10 and the federal habeas corpus statute, 28 U. S. C. § 2255, which authorizes federal courts to reopen judgments of conviction. The habeas statute, similarly to Rule 60(b), replaced a common-law writ, see App. to H. R. Rep. No. 308, 80th Cong., 2d Sess., A180 (1947), and thus necessarily applied retroactively.11 State statutes that authorize the reopening of various types of default judgments 12 and judgments that became final before a party re*259ceived notice of their entry,13 as well as provisions for motions to reopen based on newly discovered evidence,14 further demonstrate the widespread acceptance of remedial statutes that allow courts to set aside final judgments. As in the case of Rule 60(b), logic dictates that these statutes be construed to apply retroactively to judgments that were final at the time of their enactments. All of these remedial statutes announced generally applicable rules of law as well as establishing procedures for reopening final judgments.15
In contrast, in the examples of colonial legislatures’ review of trial courts’ judgments on which today’s holding rests, the legislatures issued directives in individual cases without purporting either to set forth or to apply any legal standard. Cf. ante, at 219-225; see, e. g., INS v. Chadha, 462 U. S. 919, 961-962 (1983) (Powell, J., concurring in judgment). The principal compendium on which the Court relies, ante, at 219, accurately describes these legislative directives:
“In these records, which are of the first quarter of the 18th century, the provincial legislature will often be found acting in a judicial capacity, sometimes trying causes in equity, sometimes granting equity powers to some court of the common law for a particular temporary purpose, and constantly granting appeals, new trials, and other relief from judgments, on equitable *260grounds.” Judicial Action by the Provincial Legislature of Massachusetts, 15 Harv. L. Rev. 208, n. 1 (1902).
The Framers’ disapproval of such a system of ad hoc legislative review of individual trial court judgments has no bearing on remedial measures such as Rule 60(b) or the 1991 amendment at issue today. The history on which the Court relies provides no support for its holding.
Ill
The lack of precedent for the Court’s holding is not, of course, a sufficient reason to reject it. Correct application of separation-of-powers principles, however, confirms that the Court has reached the wrong result. As our most recent major pronouncement on the separation of powers noted, “we have never held that the Constitution requires that the three branches of Government ‘operate with absolute independence.’” Morrison v. Olson, 487 U. S. 654, 693-694 (1988) (quoting United States v. Nixon, 418 U. S. 683, 707 (1974)). Rather, our jurisprudence reflects “Madison’s flexible approach to separation of powers.” Mistretta v. United States, 488 U. S. 361, 380 (1989). In accepting Madison’s conception rather than any “hermetic division among the Branches,” id., at 381, “we have upheld statutory provisions that to some degree commingle the functions of the Branches, but that pose no danger of either aggrandizement or encroachment,” id., at 382. Today’s holding does not comport with these ideals.
Section 27A shares several important characteristics with the remedial statutes discussed above. It does not decide the merits of any issue in any litigation but merely removes an impediment to judicial decision on the merits. The impediment it removes would have produced inequity because the statute’s beneficiaries did not cause the impediment. It requires a party invoking its benefits to file a motion within a specified time and to convince a court that the statute entitles the party to relief. Most important, §27A(b) specifies *261both a substantive rule to govern the reopening of a class of judgments — the pr e-Lampf limitations rule — and a procedure for the courts to apply in determining whether a particular motion to reopen should be granted. These characteristics are quintessentially legislative. They reflect Congress’ fealty to the separation of powers and its intention to avoid the sort of ad hoc excesses the Court rightly criticizes in colonial legislative practice. In my judgment, all of these elements distinguish §27A from “judicial” action and confirm its constitutionality. A sensible analysis would at least consider them in the balance.
Instead, the Court myopically disposes of § 27A(b) by holding that Congress has no power to “requir[e] an Article III court to set aside a final judgment.” Ante, at 240. That holding must mean one of two things. It could mean that Congress may not impose a mandatory duty on a court to set aside a judgment even if the court makes a particular finding, such as a finding of fraud or mistake, that Congress has not made. Such a rule, however, could not be correct. Although Rule 60(b), for example, merely authorizes federal courts to set aside judgments after making appropriate findings, Acts of Congress characteristically set standards that judges are obligated to enforce. Accordingly, Congress surely could add to Rule 60(b) certain instances in which courts must grant relief from final judgments if they make particular findings — for example, a finding that a member of the jury accepted a bribe from the prevailing party. The Court, therefore, must mean to hold that Congress may not unconditionally require an Article III court to set aside a final judgment. That rule is both unwise and beside the point of this case.
A simple hypothetical example will illustrate the practical failings of the Court’s new rule. Suppose Congress, instead of endorsing the new limitations rule fashioned by the Court in Lampf, had decided to return to the pre-Lampf regime (or perhaps to enact a longer uniform statute). Subsection *262(a) of § 27A would simply have provided that the law in effect prior to June 19, 1991, would govern the timeliness of all 10b-5 actions. In that event, subsection (b) would still have been necessary to remedy the injustice caused by this Court’s failure to exempt pending cases from its new rule. In my judgment, the statutory correction of the inequitable flaw in Lampf would be appropriate remedial legislation whether or not Congress had endorsed that decision’s substantive limitations rule. The Court, unfortunately, appears equally consistent: Even though the class of dismissed 10b-5 plaintiffs in my hypothetical would have been subject to the same substantive rule as all other 10b-5 plaintiffs, the Court’s reasoning would still reject subsection (b) as an impermissible exercise of “judicial” power.
The majority’s rigid holding unnecessarily hinders the Government from addressing difficult issues that inevitably arise in a complex society. This Court, for example, lacks power to enlarge the time for filing petitions for certiorari in a civil case after 90 days from the entry of final judgment, no matter how strong the equities. See 28 U. S. C. § 2101(c). If an Act of God, such as a flood or an earthquake, sufficiently disrupted communications in a particular area to preclude filing for several days, the majority’s reasoning would appear to bar Congress from addressing the resulting inequity. If Congress passed remedial legislation that retroactively granted movants from the disaster area extra time to file petitions or motions for extensions of time to file, today’s holding presumably would compel us to strike down the legislation as an attack on the finality of judgments. Such a ruling, like today’s holding, would gravely undermine federal courts’ traditional power “to set aside a judgment whose enforcement would work inequity.” Ante, at 234.16
*263Even if the rule the Court announces today were sound, it would not control the case before us. In order to obtain the benefit of §27A, petitioners had to file a timely motion and persuade the District Court they had timely filed their complaint under pr e-Lampf law. In the judgment of the District Court, petitioners satisfied those conditions. Congress reasonably could have assumed, indeed must have expected, that some movants under § 27A(b) would fail to do so. The presence of an important condition that the District Court must find a movant to have satisfied before it may reopen a judgment distinguishes §27A from the unconditional congressional directives the Court appears to forbid.
Moreover, unlike the colonial legislative commands on which the Court bases its holding, § 27A directed action not in “a civil case,” ante, at 223 (discussing Calder v. Bull, 3 Dall. 386 (1798)), but in a large category of civil cases.17 The Court declares that a legislative direction to reopen a class of 40 cases is 40 times as bad as a direction to reopen a single final judgment because “power is the object of the separation-of-powers prohibition.” See ante, at 228. This self-evident observation might be salient if § 27A(b) unconditionally commanded courts to reopen judgments even absent findings that the complaints were timely under pr e-Lampf law. But Congress did not decide — and could not know how any court would decide — the timeliness issue in any particu*264lar case in the affected category. Congress, therefore, had no way to identify which particular plaintiffs would benefit from § 27A. It merely enacted a law that applied a substantive rule to a class of litigants, specified a procedure for invoking the rule, and left particular outcomes to individualized judicial determinations — a classic exercise of legislative power.
“All we seek,” affirmed a sponsor of §27A, “is to give the victims [of securities fraud] a fair day in court.”18 A statute, such as § 27A, that removes an unanticipated and unjust impediment to adjudication of a large class of claims on their merits poses no danger of “aggrandizement or encroachment.” Mistretta, 488 U. S., at 382.19 This is particularly true for §27A in light of Congress’ historic primacy over statutes of limitations.20 The statute contains several checks against the danger of congressional overreaching. The Court in Lampf undertook a legislative function. Essentially, it supplied a statute of limitations for 10b-5 ac*265tions. The Court, however, failed to adopt the transition rules that ordinarily attend alterations shortening the time to sue. Congress, in §27A, has supplied those rules. The statute reflects the ability of two coequal branches to cooperate in providing for the impartial application of legal rules to particular disputes. The Court’s mistrust of such cooperation ill serves the separation of powers.21
IV
The Court has drawn the wrong lesson from the Framers’ disapproval of colonial legislatures’ appellate review of judicial decisions. The Framers rejected that practice, not out of a mechanistic solicitude for “final judgments,” but because they believed the impartial application of rules of law, rather *266than the will of the majority, must govern the disposition of individual cases and controversies. Any legislative interference in the adjudication of the merits of a particular case carries the risk that political power will supplant evenhanded justice, whether the interference occurs before or after the entry of final judgment. Cf. United States v. Klein, 13 Wall. 128 (1872); Hayburn’s Case, 2 Dall. 409 (1792). Section 27A(b) neither commands the reinstatement of any particular case nor directs any result on the merits. Congress recently granted a special benefit to a single litigant in a pending civil rights case, but the Court saw no need even, to grant certiorari to review that disturbing legislative favor.22 In an ironic counterpoint, the CourLtodav places a higher priority on protecting the'Kepublic from the restoration to a large class of litigants of the opportunity to have Article III courts resolve the merits of their claims.
“We must remember that the machinery of government would not work if it were not allowed a little play in its joints.” Bain Peanut Co. of Tex. v. Pinson, 282 U. S. 499, 501 (1931) (Holmes, J.). The three branches must cooperate in order to govern. We should regard favorably, rather than with suspicious hostility, legislation that enables the judiciary to overcome impediments to the performance of its mission of administering justice impartially, even when, as here, this Court has created the impediment.23 Rigid rules often make good law, but judgments in areas such as the review of potential conflicts among the three coequal branches of the *267Federal Government partake of art as well as science. That is why we have so often reiterated the insight of Justice Jackson:
“The actual art of governing under our Constitution does not and cannot conform to judicial definitions of the power of any of its branches based on isolated clauses or even single Articles torn from context. While the Constitution diffuses power the better to secure liberty, it also contemplates that practice will integrate the dispersed powers into a workable government. It enjoins upon its branches separateness but interdependence, autonomy but reciprocity.” Youngstown Sheet & Tube Co. v. Sawyer, 343 U. S. 579, 635 (1952) (concurring opinion).
We have the authority to hold that Congress has usurped a judicial prerogative, but even if this case were doubtful I would heed Justice Iredell’s admonition in Calder v. Bull, 3 Dall., at 399, that “the Court will never resort to that authority, but in a clear and urgent case.” An appropriate regard for the interdependence of Congress and the judiciary amply supports the conclusion that §27A(b) reflects constructive legislative cooperation rather than a usurpation of judicial prerogatives.
Accordingly, I respectfully dissent.
Because the Court finds a separation-of-powers violation, it does not reach respondents’ alternative theory that §27A(b) denied them due process under the Fifth Amendment, a theory the Court of Appeals did not identify as an alternative ground for its holding. In my judgment, the statute easily survives a due process challenge. Section 27A(b) is ration*248ally related to a legitimate public purpose. Cf., e. g., Concrete Pipe & Products of Cal., Inc. v. Construction Laborers Pension Trust for Southern Cal., 508 U. S. 602, 636-641 (1993). Given the existence of statutes and rules, such as Rule 60(b), that allow courts to reopen apparently “final” judgments in various circumstances, see infra, at 256-259, respondents cannot assert an inviolable “vested right” in the District Court’s postLampf dismissal of petitioners’ claims. In addition, § 27A(b) did not upset any “settled expectations" of respondents. Cf. Landgraf v. USI Film Products, 511 U. S. 244, 266 (1994). In Landgraf, we concluded that Congress did not intend § 102 of the Civil Rights Act of 1991, 42 U. S. C. § 1981a (1988 ed., Supp. V), to apply retroactively because retroactive application would have placed a new legal burden on past conduct. 511 U. S., at 280-286. Before 1991 no one could have relied either on the yet-to-be-announced rule in Lampf or on the Court’s unpredictable decision to apply that rule retroactively. All of the reliance interests that ordinarily support a presumption against retroactivity militate in favor of allowing retroactive application of § 27A.
“Federal judges have ‘borrowed’ state statutes of limitations because they were directed to do so by the Congress of the United States under the Rules of Decision Act, 28 U. S. C. § 1652.” Lampf, 501 U. S., at 367, n. 2 (Stevens, J., dissenting) (citations omitted); see, e. g., Stull v. Bayard, 561 F. 2d 429, 431-432 (CA2 1977), cert. denied, 434 U. S. 1035 (1978); Roberts v. Magnetic Metals Co., 611 F. 2d 450, 456 (CA3 1979); Robuck v. Dean Witter & Co., 649 F. 2d 641, 644 (CA9 1980) (borrowing state statutes of limitations for 10b-5 actions).
See Ky. Rev. Stat. Ann. § 292.480(3) (Michie 1988).
The Lampf opinion drew two other dissents. Justice Kennedy, joined by Justice O’Connoe, would have adopted a different substantive limitations rule. See 501 U. S., at 374. Justice Souter and I would have adhered to “four decades of . . . settled law” and maintained the existing regime until Congress enacted a new federal statute of limitations. Id., at 366-367 (Stevens, J., dissenting). No one dissented from the proposition that a uniform federal limitations period would be wise policy.
See Cooperativa de Ahorro y Credito Aguada v. Kidder, Peabody & Co., 993 F. 2d 269 (CAI), cert. pending, No. 93-564; Axel Johnson Inc. v. Arthur Andersen & Co., 6 F. 3d 78 (CA2 1993); Cooke v. Manufactured Homes, Inc., 998 F. 2d 1256 (CA4 1993); Berning v. A G. Edwards & Sons, Inc., 990 F. 2d 272 (CA7 1993); Gray v. First Winthrop Corp., 989 F. 2d 1564 (CA9 1993); Anixter v. Home-Stake Production Co., 977 F. 2d 1533 (CA10 1992), cert. denied sub nom. Dennler v. Trippet, 507 U. S. 1029 (1993); Henderson v. Scientific-Atlanta, Inc., 971 F. 2d 1567 (CA11 1992), cert. denied, 510 U. S. 828 (1993).
The Act provided, in part:
“That all cases of appeal or writ of error heretofore prosecuted and now pending in the supreme court of the United States, upon any record from the supreme court of the Territory of Nevada, may be heard and determined by the supreme court of the United States, and the mandate of execution or of further proceedings shall be directed by the supreme court of the United States to the district court of the United States for the district of Nevada, or to the supreme court of the State of Nevada, as the nature of said appeal or writ of error may require, and each of these courts shall be the successor of the supreme court of Nevada Territory as to all such cases, with full power to hear and determine the same, and to award mesne or final process thereon. . . . Provided, That said appeals shall be prosecuted and said writs of errors sued out at any time before the first day of July, eighteen hundred and sixty-six.” Ch. 64, § 8, 13 Stat. 441.
The full text of Rule 60(b) provides:
“On motion and upon such terms as are just, the court may relieve a party or a party’s legal representative from a final judgment, order, or proceeding for the following reasons: (1) mistake, inadvertence, surprise, or excusable neglect; (2) newly discovered evidence which by due diligence could not have been discovered in time to move for a new trial under Rule 59(b); (3) fraud (whether heretofore denominated intrinsic or extrinsic), *257misrepresentation, or other misconduct of an adverse party; (4) the judgment is void; (5) the judgment has been satisfied, released, or discharged, or a prior judgment upon which it is based has been reversed or otherwise vacated, or it is no longer equitable that the judgment should have prospective application; or (6) any other reason justifying relief from the operation of the judgment. The motion shall be made within a reasonable time, and for reasons (1), (2), and (3) not more than one year after the judgment, order, or proceeding was entered or taken. A motion under this subdivision (b) does not affect the finality of a judgment or suspend its operation. This rule does not limit the power of a court to entertain an independent action to relieve a party from a judgment, order, or proceeding, or to grant relief to a defendant not- actually personally notified as provided in Title 28, U. S. C. § 1655, or to set aside a judgment for fraud upon the court. Writs of coram nobis, coram vobis, audita querela, and bills of review and bills in the nature of a bill of review, are abolished, and the procedure for obtaining any relief from a judgment shall be by motion as prescribed in these rules or by an independent action.”
This Court adopted the Federal Rules of Civil Procedure and submitted them to Congress as the Rules Enabling Act required. They became effective after Congress adjourned without altering them. See generally 308 U. S. 647 (letter of transmittal to Congress, Jan. 3,1938).
In its criticism of this analysis of Rule 60(b), the majority overstates our holdings on retroactivity in Landgraf, 511 U. S., at 280, and Rivers v. Roadway Express, Inc., 511 U. S. 298 (1994). Our opinion in Landgraf nowhere says “that statutes do not apply retroactively unless Congress expressly states that they do.” Ante, at 237. To the contrary, it says that, “[w]hen ... the statute contains no such express command, the court must determine whether the new statute would have retroactive effect,” an inquiry that requires “clear congressional intent favoring such a result.” Landgraf, 511 U. S., at 280 (emphasis added); see also id., at 273-275; Rivers, 511 U. S., at 304-309. In the case of Rule 60(b), the factors I have identified, taken together, support a finding of clear congressional *258intent. Moreover, neither Landgraf nor Rivers “rejected” consideration of a statute’s remedial purpose in analyzing Congress’ intent to apply the statute retroactively. Compare ante, at 237, with Landgraf, 611 U. S., at 281-286, and n. 37, and Rivers, 511 U. S., at 304-311.
When it enacted the Handicapped Children’s Protection Act, Congress overruled our contrary decision in Smith v. Robinson, 468 U. S. 992 (1984), by applying the Act retroactively to any action either pending on or brought after July 4,1984, the day before we announced Smith. See 100 Stat. 798. Accordingly, a court has applied the Act retroactively to a case in which the parties had entered into a consent decree prior to its enactment. See Counsel v. Dow, 849 F. 2d 731, 738-739 (CA2 1988). The Court’s attempts to explain away the retroactivity provision, ante, at 235-236, simply do not comport with the plain language of the Act.
The Government also calls our attention to 28 U. S. C. § 1655, a statute that requires courts to reopen final in rem judgments upon entries of appearance by defendants who were not personally served. See Brief for United States 24-25, and n. 17. While that statute had only prospective effect, the Court offers no reason why Congress could not pass a similar statute that would apply retroactively to judgments entered under preexisting procedures.
See, e.g., Del. Code Ann., Tit. 18, § 4418 (1989); Fla. Stat. § 631.734 (1984); N. Y. Ins. Law § 7717 (McKinney Supp. 1995); 40 Pa. Cons. Stat. § 991.1716 (Supp. 1994).
For example, a Virginia statute provides that, when a pro se litigant fails to receive notice of the trial court’s entry of an order, even after the time to appeal has expired, the trial judge may within 60 days vacate the order and grant the party leave to appeal. Va. Code Ann. § 8.01-428(0 (Supp. 1994).
See Herrera v. Collins, 506 U. S. 390, 410-411, and nn. 8-11 (1993) (citing state statutes).
The Court offers no explanation of why the Constitution should be construed to interpose an absolute bar against these statutes’ retroactive application. Under the Court’s reasoning, for example, an amendment that broadened the coverage of Rule 60(b) could not apply to any inequitable judgments entered prior to the amendment. The Court’s rationale for this formalistic restriction remains elusive.
The Court also appears to bar retroactive application of changes in the criminal law. Its reasoning suggests that, for example, should Congress one day choose to abolish the federal death penalty, the new statute *263could not constitutionally save a death row inmate from execution if his conviction had become final before the statute was passed.
At the time Congress was considering the bill that became §27A, a House Subcommittee reported that Lampf had resulted in the dismissal of 15 cases, involving thousands of plaintiffs in every State (of whom over 32,000 had been identified) and claims totaling over $692.25 million. In addition, motions to dismiss based on Lampf were then pending in 17 cases involving thousands of plaintiffs in every State and claims totaling over $4,578 billion. Hearing on H. R. 3185, before the Subcommittee on Telecommunications and Finance of the House Committee on Energy and Commerce, 102d Cong., 1st Sess., 1-4 (1991).
137 Cong. Rec. S18624 (Nov. 27, 1991) (statement of Sen. Bryan).
Today’s decision creates a new irony of judicial legislation. A challenge to the constitutionality of § 27A(a) could not turn on the sanctity of final judgments. Section 27A(a) benefits litigants who had filed appeals that Lampf rendered frivolous; petitioners and other law-abiding litigants whose claims Lampf rendered untimely had acquiesced in the dismissal of their actions. By striking down §27A(b) on a ground that would leave §27A(a) intact, the Court indulges litigants who protracted proceedings but shuts the courthouse door to litigants who proceeded with diligence and respect for the Lampf judgment.
“Statutes of limitation find their justification in necessity and convenience rather than in logic. They represent expedients, rather than principles. . . . They are by definition arbitrary, and their operation does not discriminate between the just and the unjust claim, or the voidable and unavoidable delay. They have come into the law not through the judicial process but through legislation. They represent a public policy about the privilege to litigate____[T]he history of pleas of limitation shows them to be good only by legislative grace and to be subject to a relatively large degree of legislative control.” Chase Securities Corp. v. Donaldson, 325 U. S. 304, 314 (1945) (Jackson, J.) (footnote and citation omitted).
Although I agree with Justice Breyer’s general approach to the separation-of-powers issue, I believe he gives insufficient weight to two important features of § 27A. First, he fails to recognize that the statute restored a pre-existing rule of law in order to remedy the manifest injustice produced by the Court’s retroactive application of Lampf. The only ‘“substantial deprivation’” Congress imposed on defendants was that properly filed lawsuits proceed to decisions on the merits. Cf. ante, at 242 (Breyer, J., concurring in judgment) (quoting INS v. Chadha, 462 U. S. 919, 962 (1983) (Powell, J., concurring in judgment)). Second, he understates the class of defendants burdened by §27A: He finds the statute underinclusive because it provided no remedy for potential plaintiffs who may have failed to file timely actions in reliance on pr e-Lampf limitations law, but he denies the importance of §27A(a), which provided a remedy for plaintiffs who appealed dismissals after Lampf. See ante, at 243-244 (Breyer, J., concurring in judgment). The coverage of § 27A is coextensive with the retroactive application of the general rule announced in Lampf. If Congress had enacted a statute providing that the Lampf rule should apply to all cases filed after the statute’s effective date and that the pr e-Lampf rule should apply to all cases filed before that date, Justice Breyer could not reasonably condemn the statute as special legislation. The only difference between such a statute and §27A is that §27A covered all cases pending on the date of Lampf— June 20, 1991— rather than on the effective date of the statute — December 19, 1991. In my opinion, §27A has sufficient generality to avoid the characteristics of a bill of attainder.
See Atonio v. Wards Cove Packing Co., 513 U. S. 809 (1994); see also Landgraf 511 U. S., at 258 (“The parties agree that § 402(b) [of the Civil Rights Act of 1991] was intended to exempt a single disparate-impact lawsuit against the Wards Cove Packing Company”).
Of course, neither the majority nor I would alter its analysis had Congress, rather than the Court, enacted the Lamp/ rule without any exemption for pending cases, then later tried to remedy such unfairness by enacting §27A. Thus, the Court’s attribution of §27A to “the legislature’s genuine conviction (supported by all the law professors in the land) that [Lampf] was wrong,” ante, at 228, is quite beside the point.