concurred with the plurality’s judgment because he disagreed with the substantive rule announced in Scheiner, but he did not agree with the plurality’s reasoning. After stating that his views on retroactivity diverged from the plurality’s “in a fundamental way,” Justice Scalia explained:
“I share [the dissent’s] perception that prospective decisionmaking is incompatible with the judicial role, which is to say what the law is, not to prescribe what [the law] shall be. The very framing of the issue that
*286we purport to decide today — whether our decision in Scheiner shall ‘apply’ retroactively — presupposes a view of our decisions as creating the law, as opposed to declaring what the law already is. Such a view is contrary to that understanding of ‘the judicial Power,’ U. S. Const., Art. Ill, § 1, which is not only the common and traditional one, but which is the only one that can justify courts in denying force and effect to the unconstitutional enactments of duly elected legislatures, see Marburg v. Madison, 1 Cranch 137 (1803) — the very exercise of judicial power asserted in Scheiner. To hold a governmental Act to be unconstitutional is not to announce that we forbid it, but that the Constitution forbids it; and when, as in this case, the constitutionality of a state statute is placed in issue, the question is not whether some decision of ours ‘applies’ in the way that a law applies; the question is whether the Constitution, as interpreted in that decision, invalidates the statute. Since the Constitution does not change from year to year; since it does not conform to our decisions, but our decisions are supposed to conform to it; the notion that our interpretation of the Constitution in a particular decision could take prospective form does not make sense. Either enforcement of the statute at issue in Scheiner (which occurred before our decision there) was unconstitutional, or it was not; if it was, then so is enforcement of all identical statutes in other States, whether occurring before or after our decision; and if it was not, then Scheiner was wrong, and the issue of whether to ‘apply’ that decision needs no further attention.” American Trucking Assns., Inc. v. Smith, 496 U. S., at 201.
Because Justice Scalia’s vote rested on his disagreement with the substantive rule announced in Scheiner — rather than with the retroactivity analysis in the dissenting opin*287ion — there were actually five votes supporting the dissent’s views on the retroactivity issue. Accordingly, it is the dissent rather than the plurality that should inform our analysis of the issue before us today.22
Moreover, several years later, a majority of this Court explicitly adopted the Smith dissent’s reasoning in Harper v. Virginia Dept. of Taxation, 509 U. S. 86 (1993). Harper, like Smith, involved a request for a refund of taxes paid before we declared a similar Michigan tax unconstitutional. We held that the Virginia tax at issue in Harper was in fact invalid — even before we declared the similar tax unconstitutional — but that this did not necessarily entitle petitioners to a full refund. We explained that the Constitution required Virginia to “'provide relief consistent with federal due process principles,’ ” 509 U. S., at 100 (quoting American Trucking Assns., Inc. v. Smith, 496 U. S., at 181 (plurality opinion)), but that “ ‘a State found to have imposed an impermissibly discriminatory tax retains flexibility in responding to this determination’ ” under the Due Process Clause, 509 U. S., at 100 (quoting McKesson Corp. v. Division of Alcoholic Beverages and Tobacco, Fla. Dept. of Business Regulation, 496 U. S. 18, 39-40 (1990)). We left to the “Virginia courts this question of state law and the performance of other tasks pertaining to the crafting of any appropriate remedy.” 509 U. S., at 102. And we specifically noted that Virginia “ ‘is free to choose which form of relief it will provide, so long as that relief satisfies the minimum federal requirements we have outlined.’” Ibid, (quoting McKesson, 496 U. S., at 51-52); see also 509 U. S., at 102 (“State law may provide relief beyond the demands of federal due process, but under no circumstances may it confine petitioners to a lesser remedy” (citation omitted)).
*288Thus, to the extent that these civil retroactivity decisions are relevant to the issue before us today,23 they support our conclusion that the remedy a state court chooses to provide its citizens for violations of the Federal Constitution is primarily a question of state law. Federal law simply “sets certain minimum requirements that States must meet but may exceed in providing appropriate relief.” American Trucking Assns., Inc. v. Smith, 496 U. S., at 178-179 (plurality opinion). They provide no support for the proposition that federal law places a limit on state authority to provide remedies for federal constitutional violations.
VI
Finally, while the State acknowledges that it may grant its citizens broader protection than the Federal Constitution requires by enacting appropriate legislation or by judicial interpretation of its own Constitution, it argues that it may not do so by judicial misconstruction of federal law. Oregon v. Hass, 420 U. S. 714 (1975) - like our early decisions in Ableman v. Booth, 21 How. 506 (1859), and Tarble’s Case, 13 Wall. 397 (1872) — provides solid support for that proposition. But *289the States that give broader retroactive effect to this Court’s new rules of criminal procedure do not do so by misconstruing the federal Teague standard. Rather, they have developed state law to govern retroactivity in state postconviction proceedings. See, e. g., State v. Whitfield, 107 S. W. 3d 253, 268 (Mo. 2003) (“[A]s a matter of state law, this Court chooses not to adopt the Teague analysis . . . ”). The issue in this case is whether there is a federal rule, either implicitly announced in Teague, or in some other source of federal law, that prohibits them from doing so.
The absence of any precedent for the claim that Teague limits state collateral review courts’ authority to provide remedies for federal constitutional violations is a sufficient reason for concluding that there is no such rule of federal law. That conclusion is confirmed by several additional considerations. First, if there is such a federal rule of law, presumably the Supremacy Clause in Article V of the Federal Constitution would require all state entities — not just state judges — to comply with it. We have held that States can waive a Teague defense, during the course of litigation, by expressly choosing not to rely on it, see Collins v. Young-blood, 497 U. S. 37, 41 (1990), or by failing to raise it in a timely manner, see Schiro v. Farley, 510 U. S. 222, 228-229 (1994). It would indeed be anomalous to hold that state legislatures and executives are not bound by Teague, but that state courts are.
Second, the State has not identified, and we cannot discern, the source of our authority to promulgate such a novel rule of federal law. While we have ample authority to control the administration of justice in the federal courts — particularly in their enforcement of federal legislation — we have no comparable supervisory authority over the work of state judges. Johnson v. Fankell, 520 U. S. 911 (1997). And while there are federal interests that occasionally justify this
*290Court’s development of common-law rules of federal law,24 our normal role is to interpret law created by others and “not to prescribe what it shall be.” American Trucking Assns., Inc. v. Smith, 496 U. S., at 201 (Scalia, J., concurring in judgment). Just as constitutional doubt may tip the scales in favor of one construction of a statute rather than another, so does uncertainty about the source of authority to impose a federal limit on the power of state judges to remedy wrongful state convictions outweigh any possible policy arguments favoring the rule that respondent espouses.
Finally, the dissent contends that the “end result [of this opinion] is startling” because “two criminal defendants, each of whom committed the same crime, at the same time, whose convictions became final on the same day, and each of whom raised an identical claim at the same time under the Federal Constitution” could obtain different results. Post, at 292. This assertion ignores the fact that the two hypothetical criminal defendants did not actually commit the “same crime.” They violated different state laws, were tried in and by different state sovereigns, and may — for many reasons — be subject to different penalties. As previously noted, such nonuniformity is a necessary consequence of a federalist system of government.
VII
It is important to keep in mind that our jurisprudence concerning the “retroactivity” of “new rules” of constitutional law is primarily concerned, not with the question whether a *291constitutional violation occurred, but with the availability or nonavailability of remedies. The former is a “pure question of federal law, our resolution of which should be applied uniformly throughout the Nation, while the latter is a mixed question of state and federal law.” American Trucking Assns., Inc. v. Smith, 496 U. S., at 205 (Stevens, J., dissenting).
A decision by this Court that a new rule does not apply retroactively under Teague does not imply that there was no right and thus no violation of that right at the time of trial— only that no remedy will be provided in federal habeas courts. It is fully consistent with a government of laws to recognize that the finality of a judgment may bar relief. It would be quite wrong to assume, however, that the question whether constitutional violations occurred in trials conducted before a certain date depends on how much time was required to complete the appellate process.
Accordingly, the judgment of the Supreme Court of Minnesota is reversed, and the case is remanded for further proceedings not inconsistent with this opinion. As was true in Michigan v. Payne, the Minnesota court is free to reinstate its judgment disposing of the petition for state postconviction relief.
It is so ordered.
While the opinions discussed at great length our earlier cases raising retroactivity issues, none of them suggested that federal law would prohibit Arkansas from refunding the taxes at issue if it wanted to do so.
The petitioners and the dissenters in American Trucking Assns., Inc. v. Smith relied heavily on separate opinions authored by Justice Harlan, and on the Court’s then-recent opinion in Griffith, 479 U. S. 314, supporting the proposition that a new constitutional holding should be applied not only in cases that had not yet been tried, but also in all cases still pending on direct review. The plurality, however, declined to follow Griffith because of its view that “there are important distinctions between the retroactive application of civil and criminal decisions that make the Griffith rationale far less compelling in the civil sphere.” 496 U. S., at 197. While Justice Harlan would probably disagree with the suggestion that the distinction between civil and criminal eases provided an acceptable basis for refusing to follow Griffith in the American Trucking Assns., Inc. v. Smith litigation, see Mackey, 401 U. S., at 683, n. 2 (Harlan, J., concurring in judgments in part and dissenting in part), if relevant, that same distinction would make it appropriate to disregard the plurality’s opinion in American Trucking Assns., Inc. v. Smith in this case.
See Boyle v. United Technologies Corp., 487 U. S. 500, 504 (1988) (“[W]e have held that a few areas, involving ‘uniquely federal interests,’ are so committed by the Constitution and laws of the United States to federal control that state law is pre-empted and replaced, where necessary, by federal law of a content prescribed ... by the courts — so-called ‘federal common law’” (citation omitted)); United States v. Kimbell Foods, Inc., 440 U. S. 715 (1979); Banco Nacional de Cuba v. Sabbatino, 376 U. S. 398 (1964).