with whom
Justice Thomas joins, dissenting.In Lawrence v. Chater, 516 U. S. 163 (1996) (per curiam), we greatly expanded our “no-fault V & R practice” (GVR) *871beyond its traditional bounds. Id., at 179 (Scalia, J., dissenting). At the time, I remarked that “[t]he power to ‘revise and correct’ for error has become a power to void for suspicion” of error, id., at 190 (quoting Marbury v. Madison, 1 Cranch 137, 175 (1803); alterations omitted). And I predicted that “ ‘GVR’d for clarification of —’ ” would “become a common form of order, drastically altering the role of this Court.” 516 U. S., at 185. Today, by vacating the judgment of a state court simply because “[i]f this Court is to reach the merits of this case, it would be better to have the benefit of the views of the full Supreme Court of Appeals of West Virginia on the Brady issue,” ante, at 870, the Court brings this prediction to fulfillment.
In Lawrence, I identified three narrow circumstances in which this Court could, consistent with the traditional understanding of our appellate jurisdiction (or at least consistent with entrenched practice), justify vacating a lower court’s judgment without first identifying error: “(1) where an intervening factor has arisen [e. g., new legislation or a recent judgment of this Court] that has a legal bearing upon the decision, (2) where, in a context not governed by Michigan v. Long, 463 U. S. 1032 (1983), clarification of the opinion below is needed to assure our jurisdiction, and (3) (in acknowledgment of established practice, though not necessarily in agreement with its validity) where the respondent or appellee confesses error in the judgment below.” 516 U. S., at 191-192 (dissenting opinion). Needless to say, today’s novel GVR order falls into none of these categories. There has been no intervening change in law that might bear upon the judgment. Our jurisdiction is not in doubt, see ante, at 870; State v. Frazier, 162 W. Va. 935, 942, n. 5, 253 S. E. 2d 534, 538, n. 5 (1979) (petitioner’s Brady claim was properly presented in his motion for a new trial). And the State has confessed no error — not even on the broadest and least supportable theory of what constitutes an error justifying vacatur. See, e. g., Alvarado v. United States, 497 U. S. 543, 545 *872(1990) (Rehnquist, C. J., dissenting) (vacating when the Solicitor General confessed error in the lower court’s “ ‘analysis,’ ” but not its judgment); Stutson v. United States, 516 U. S. 193 (1996) (per curiam) (vacating when the Solicitor General confessed error in a position taken before the Court of Appeals, on which the court might have relied; discussed in Lawrence, supra, at 184-185 (Scalia, J., dissenting)); Department of Interior v. South Dakota, 519 U. S. 919, 921 (1996) (Scalia, J., dissenting) (vacating when “the Government, having lost below, wishes to try out a new legal position”). Here, the Court vacates and remands in light of nothing.
Instead, the Court remarks tersely that it would be “better” to have “the benefit” of the West Virginia court’s views on petitioner’s Brady claim, should we eventually decide to take the case. Ante, at 870. The Court thus purports to conscript the judges of the Supreme Court of Appeals of West Virginia to write what is essentially an amicus brief on the merits of an issue they have already decided, in order to facilitate our possible review of the merits at some later time. It is not at all clear why it would be so much “better” to have the full court below address the Brady claim. True, we often prefer to review reasoned opinions that facilitate our consideration — though we may review even a summary disposition. See Lawrence, supra, at 186 (Scalia, J., dissenting). But the dissenting judges in the case below discussed petitioner’s Brady claim at some length (indeed, at greater length than appears in many of the decisions we agree to review), and argued that it was meritorious. See 217 W. Va. 535, 549-552, 618 S. E. 2d 544, 558-561 (2005) (Davis, J., joined by Starcher, J., dissenting). Since we sometimes review judgments with no opinion, and often review judgments with opinion only on one side of the issue, it is not clear why we need opinions on both sides here.
To tell the truth, there is only one obvious sense in which it might be “better” to have the West Virginia court revisit *873the Brady issue: If the majority suspects that the court below erred, there is a chance that the GVR-in-light-of-nothing will induce it to change its mind on remand, sparing us the trouble of correcting the suspected error. It is noteworthy that, to justify its GVR order, the Court does not invoke even the flabby standard adopted in Lawrence, namely, whether there is “a reasonable probability that the decision below rests upon a premise that the lower court would reject if given the opportunity for further consideration,” 516 U. S., at 167. That is because (there being no relevant intervening event to create such a probability) the only possibility that the West Virginia court will alter its considered judgment is created by this Court’s GVR order itself. A case such as this, which meets none of the usual, outcome-neutral criteria for granting certiorari set forth in this Court’s Rules 10(a)-(c), could attract our notice only if we suspected that the judgment appealed from was in error. Those whose judgments we review have sometimes viewed even our legitimate, intervening-event GVR orders as polite directives that they reverse themselves. See, e. g., Sharpe v. United States, 712 F. 2d 65, 67 (CA4 1983) (Russell, J., dissenting) (“Once again, I think the majority has mistaken gentleness in instruction for indefiniteness in command. The Supreme Court was seeking to be gentle with us but there is, I submit, no mistaking what they expected us to do”). How much more is that suspicion justified when the GVR order rests on nothing more than our statement that it would be “better” for the lower court to reconsider its decision (much as a mob enforcer might suggest that it would be “better” to make protection payments).
Even when we suspect error, we may have many reasons not to grant certiorari outright in a case such as this — an overcrowded docket, a reluctance to correct “the misapplication of a properly stated rule of law,” this Court’s Rule 10, or (in this particular case) even a neo-Victorian desire to keep the lurid phrases of the “graphically explicit note,” *874ante, at 868, out of the U. S. Reports. But none of these reasons justifies “a tutelary remand, as to a schoolboy made to do his homework again.” Lawrence, 516 U. S., at 185-186 (Scalia, J., dissenting). In “the nature of the appellate system created by the Constitution and laws of the United States,” id., at 178, state courts and lower federal courts are constitutionally distinct tribunals, independently authorized to decide issues of federal law. They are not, as we treat them today, “the creatures and agents of this body,” id., at 178-179. If we suspect that a lower court has erred and wish to correct its error, we should grant certiorari and decide the issue ourselves in accordance with the traditional exercise of our appellate jurisdiction.
It is particularly ironic that the Court inaugurates its “GVR-in-light-of-nothing” practice by vacating the judgment of a state court. Our no-fault GVR practice had its origins “in situations calling forth the special deference owed to state law and state courts in our system of federalism.” Id., at 179. We first used it to allow the state court to decide the effect of an intervening change in state law. Ibid. (citing Missouri ex rel. Wabash R. Co. v. Public Serv. Comm’n, 273 U. S. 126 (1927)). Likewise, our other legitimate category of no-fault GVR — to ensure our own jurisdiction — “originate[d] in the special needs of federalism.” Lawrence, 516 U. S., at 181. In vacating the judgment of a state court for no better reason than our own convenience, we not only fail to observe, but positively flout the “special deference owed to . . . state courts,” id., at 179. Like the Ouroboros swallowing its tail, our GVR practice has ingested its own original justification.
Chief Justice Marshall wrote in Marbury v. Madison that “[i]t is the essential criterion of appellate jurisdiction, that it revises and corrects the proceedings in a cause already instituted ....” 1 Cranch, at 175. At best, today’s unprecedented decision rests on a finding that the state court’s “opinion, though arguably correct, [is] incomplete and un*875workmanlike,” Lawrence, 516 U. S., at 189 (Scalia, J., dissenting) — which all Members of the Court in Lawrence agreed was an illegitimate basis for a GVR, see id., at 173 (per curiam). At worst, it is an implied threat to the lower court, not backed by a judgment of our own, that it had “better” reconsider its holding.
I suppose it would be available to the West Virginia Supreme Court of Appeals, on remand, simply to reaffirm its judgment without further elaboration. Or it could instead enter into a full discussion of the Brady issue, producing either a reaffirmance or a revision of its judgment. The latter course will of course encourage and stimulate our new “GVR-in-light-of-nothing” jurisprudence. Verb. sap.
For these reasons, I respectfully dissent.