with whom Justice Ginsburg joins, concurring in part and dissenting in part.
I agree that 18 U. S. C. § 3626(e)(2) (1994 ed., Supp. IV) is unambiguous and join Parts I and II of the majority opinion. I also agree that applying the automatic stay may raise the due process issue, of whether a plaintiff has a fair chance to preserve an existing judgment that was valid when entered. Ante this page. But I believe that applying the statute may *351also raise a serious separation-of-powers issue if the time it allows turns out to be inadequate for a court to determine whether the new prerequisite to relief is satisfied in a particular case.1 I thus do not join Part III of the Court’s opinion and on remand would require proceedings consistent with this one. I respectfully dissent from the terms of the Court’s disposition.
A prospective remedial order may rest on at least three different legal premises: the underlying right meant to be secured; the rules of procedure for obtaining relief, defining requisites of pleading, notice, and so on; and, in some cases, rules lying between the other two, such as those defining a required level of certainty before some remedy may be ordered, or the permissible scope of relief. At issue here are rules of the last variety.2
Congress has the authority to change rules of this sort by imposing new conditions precedent for the continuing enforcement of existing, prospective remedial orders and requiring courts to apply the new rules to those orders. Cf. Plant v. Spendthrift Farm, Inc., 514 U. S. 211, 282 (1995). If its legislation gives courts adequate time to determine the applicability of a new rule to an old order and to take the action necessary to apply it or to vacate the order, there seems little basis for claiming that Congress has crossed *352the constitutional line to interfere with the performance of any judicial function. But if determining whether a new rule applies requires time (say, for new factfinding) and if the statute provides insufficient time for a court to make that determination before the statute invalidates an extant remedial order, the application of the statute raises a serious question whether Congress has in practical terms assumed the judicial function. In such a ease, the prospective order suddenly turns unenforceable not because a court has made a judgment to terminate it due to changed law or fact, but because no one can tell in the time allowed whether the new rule requires modification of the old order. One way to view this result is to see the Congress as mandating modification of an order that may turn out to be perfectly enforceable under the new rule, depending on judicial factfinding. If the facts are taken this way, the new statute might well be treated as usurping the judicial function of determining the applicability of a general rule in particular factual circumstances.3 Cf. United States v. Klein, 13 Wall. 128, 146 (1872).
Whether this constitutional issue arises on the facts of this action, however, is something we cannot yet tell, for the *353District Court did not address the sufficiency of the time provided by the statute to make the findings required by § 3626(b)(3) in this particular action.4 Absent that determination, I would not decide the separation-of-powers question, but simply remand for further proceedings. If the District Court determined both that it lacked adequate time to make the requisite findings in the period before the automatic stay would become effective, and that applying the stay would violate the separation of powers, the question would then be properly presented.
The Court forecloses the possibility of a separation-of-powers challenge based on insufficient time under the Prison Litigation Reform Act of 1995 (PLRA): “In this action, we have no occasion to decide whether there could be a time constraint on judicial action that was so severe that it implicated these structural separation of powers concerns. The PLRA does not deprive courts of their adjudicatory role, but merely provides a new legal standard for relief and encourages courts to apply that standard promptly.” Ante, at 350.
Other provisions of the PLRA narrow the scope of the underlying entitlements that an order can protect, but some orders may have been issued to secure constitutional rights unaffected by the PLRA. In any event, my concern here is solely with the PLRA’s changes to the requisites for relief
The constitutional question inherent in these possible circumstances does not seem to be squarely addressed by any of our cases. Congress did not engage in discretionary review of a particular judicial judgment, cf. Plant v. Spendthrift Farm, Inc., 514 U. S. 211, 218, 226 (1995) (characterizing Hayburn’s Case, 2 Dall. 409 (1792)), or try to modify a final, non-prospeetive judgment, cf. 514 U. S., at 218-219. Nor would a stay result from the judicial application of a change in the underlying law, ef. Pennsylvania v. Wheeling & Belmont Bridge Co., 18 How. 421, 431 (1856); Plant, supra, at 218 (characterizing United States v. Klein, 13 Wall. 128 (1872)). Instead, if the time is insufficient for a court to make a judicial determination about the applicability of the new rules, the stay would result from the inability of the Judicial Branch to exercise the judicial power of determining whether the new rules applied at all. Cf Marbury v. Madison, 1 Cranch 137,177 (1803) (“It is emphatically the province and duty of the judicial department to say what the law is”).
Neither did the Court of Appeals. It merely speculated that “[i]t may be . . . that in some cases the courts will not be able to carry out their adjudicative function in a responsible way within the time limits imposed by (e)(2),” French v. Duckworth, 178 F. 3d 437, 447 (CA7 1999), without deciding whether this action presented such a situation. The court then concluded that “under Klein [the Congress] cannot take away the power of the court in a particular case to preserve the status quo while it ponders these weighty questions.” Ibid.