Ylst v. Nunnemaker

*799 Justice Scalia

delivered the opinion of the Court.

In this case we decide whether the unexplained denial of a petition for habeas corpus by a state court lifts a state procedural bar imposed on direct appeal, so that a state prisoner may then have his claim heard on the merits in a federal ha-beas proceeding.

I

In 1975, respondent Nunnemaker was tried in California state court for murder. He raised a defense of diminished capacity and introduced psychiatric testimony in support. In response, the State introduced — without objection from respondent — the testimony of a psychiatrist based upon a custodial interview. The jury found respondent guilty. He appealed, claiming for the first time that the State’s psychiatric testimony was inadmissible because the interview had not been preceded by a Miranda warning, see Miranda v. Arizona, 384 U. S. 436 (1966). In addition, he alleged that his attorney’s failure to object to the psychiatric testimony amounted to ineffective assistance of counsel, and raised other claims not relevant here.

The California Court of Appeal affirmed the conviction. The sole basis for its rejection of the Miranda claim was the state procedural rule that “an objection based upon a Miranda violation cannot be raised for the first time on appeal.” App. 15. See People v. Bennett, 60 Cal. App. 3d 112, 116, 131 Cal. Rptr. 305, 306-307 (1976); In re Dennis M., 70 Cal. 2d 444, 461-462, 450 P. 2d 296, 306-307 (1969). The California Supreme Court denied discretionary review on September 27, 1978.

*800 In 1985, respondent filed a petition for collateral relief in California Superior Court. The petition was denied without opinion. Respondent then filed a similar petition for relief in the California Court of Appeal, invoking that court’s original jurisdiction. That petition was also denied without opinion. Finally, respondent filed a petition for habeas corpus in the California Supreme Court, invoking the original jurisdiction of that tribunal. That petition was denied on December 3, 1986, with citation of In re Swain, 34 Cal. 2d 300, 304, 209 P. 2d 793, 796 (1949), and In re Waltreus, 62 Cal. 2d 218, 225, 397 P. 2d 1001, 1005 (1965). App. 82. No opinion or other explanation accompanied these citations.

Respondent next filed a petition for writ of habeas corpus in the United States District Court for the Northern District of California. The court dismissed the petition without prejudice, ruling that it was not clear whether respondent had exhausted his state remedies with respect to all his claims.1 See Rose v. Lundy, 455 U. S. 509 (1982). Respondent then filed a second petition for habeas relief in the California Supreme Court, again invoking that court’s original jurisdiction. That petition was denied, without opinion or case citation, on April 7, 1988.

Respondent then filed a second petition for habeas relief in the Northern District of California, raising the Miranda claim and the ineffectiveness claim. The court rejected the ineffectiveness claim on the merits. As to the Miranda claim, the court found that respondent’s state procedural default barred federal review. Respondent appealed. The Court of Appeals for the Ninth Circuit reversed in part. The court agreed that the ineffective-assistance claim was *801 meritless. However, relying upon our intervening opinion in Harris v. Reed, 489 U. S. 255 (1989), the court held that the California Supreme Court’s “silent denial” of respondent’s second state habeas petition to that court lifted the procedural bar arising from the decision on direct review. Specifically, the Ninth Circuit held that because the California Supreme Court did not “clearly and expressly state its reliance on Nunnemaker’s procedural default,” the federal court could not say that the Supreme Court’s order “was based on a procedural default rather than on the underlying merits of Nunnemaker’s claims.” 904 F. 2d 473, 476 (1990). We granted certiorari, 498 U. S. 957 (1990).

1 — I I — I

The last state court to render a judgment on the Miranda claim as of 1978, the California Court of Appeal, expressly found a procedural default. When a state-law default prevents the state court from reaching the merits of a federal claim, that claim can ordinarily not be reviewed in federal court. Wainwright v. Sykes, 433 U. S. 72, 87-88 (1977); Murray v. Carrier, 477 U. S. 478, 485-492 (1986). Thus, had respondent proceeded to federal habeas on the basis of the Miranda claim upon completing his direct review in 1978, federal review would have been barred by the state-law procedural default.

State procedural bars are not immortal, however; they may expire because of later actions by state courts. If the last state court to be presented with a particular federal claim reaches the merits, it removes any bar to federal-court review that might otherwise have been available. See Harris, supra, at 262. We consider, therefore, whether the California Supreme Court’s unexplained order denying his second habeas petition to that court, which according to the Ninth Circuit sought relief on the basis of his Miranda claim, constituted a “decision on the merits” of that claim sufficient to lift the procedural bar imposed on direct appeal.

*802 The Ninth Circuit concluded that it did constitute a decision on the merits by applying a presumption that when a federal claim is denied without explicit reliance on state grounds, the merits of the federal claim are the basis for the judgment. Petitioner argues that that was error,2 and we agree. The Ninth Circuit thought itself to be following our decision in Harris v. Reed, supra, at 263. As we have since made clear, however, see Coleman v. Thompson, ante, p. 722, the Harris presumption is to be applied only after it has been determined that “the relevant state court decision . . . fairly appear[s] to rest primarily on federal law or [is] interwoven with [federal] law.” Ante, at 740.

The consequent question presented by the present case, therefore, is how federal courts in habeas proceedings are to determine whether an unexplained order (by which we mean an order whose text or accompanying opinion does not disclose the reason for the judgment) rests primarily on federal law. The question is not an easy one. In Coleman itself, although the order was unexplained, the nature of the disposition (“dismissed” rather than “denied”) and surrounding circumstances (in particular the fact that the State had rested its argument entirely upon a procedural bar), indicated that the basis was procedural default. But such clues *803 will not always, or even ordinarily, be available. Indeed, sometimes the members of the court issuing an unexplained order will not themselves have agreed upon its rationale, so that the basis of the decision is not merely undiscoverable but nonexistent.

The problem we face arises, of course, because many formulary orders are not meant to convey anything as to the reason for the .decision. Attributing a reason is therefore both difficult and artificial. We think that the attribution necessary for federal habeas purposes can be facilitated, and sound results more often assured, by applying the following presumption: Where there has been one reasoned state judgment rejecting a federal claim, later unexplained orders upholding that judgment or rejecting the same claim rest upon the same ground. If an earlier opinion “fairly appear[s] to rest primarily upon federal law,” Coleman, ante, at 740, we will presume that no procedural default has been invoked by a subsequent unexplained order that leaves the judgment or its consequences in place. Similarly where, as here, the last reasoned opinion on the claim explicitly imposes a procedural default, we will presume that a later decision rejecting the claim did not silently disregard that bar and consider the merits. This approach accords with the view of every Court of Appeals to consider the matter, save the court below. See Prihoda v. McCaughtry, 910 F. 2d 1379, 1383 (CA7 1990) (dicta); Harmon v. Barton, 894 F. 2d 1268, 1272 (CA11 1990); Evans v. Thompson, 881 F. 2d 117, 123, n. 2 (CA4 1989); Ellis v. Lynaugh, 873 F. 2d 830, 838 (CA5 1989).

This presumption assists, as we have said, not only administrability but accuracy as well — unlike the application of Harris to unexplained orders, which achieves the former at the expense of the latter. As applied to an unexplained order leaving in effect a decision (or, in the case of habeas, the consequences of a decision) that expressly relies upon procedural bar, the Harris presumption would interpret the order as rejecting that bar and deciding the federal question *804 on the merits. That is simply a most improbable assessment of what actually occurred. The maxim is that silence implies consent, not the opposite — and courts generally behave accordingly, affirming without further discussion when they agree, not when they disagree, with the reasons given below. The essence of unexplained orders is that they say nothing. We think that a presumption which gives them no effect— which simply “looks through” them to the last reasoned decision-most nearly reflects the role they are ordinarily intended to play.3

Respondent poses various hypotheticals in which this presumption would not produce a correct assessment of the state-court disposition. We need not consider them, because we do not suggest that the presumption is irrebuttable; strong evidence can refute it. It might be shown, for example, that even though the last reasoned state-court opinion had relied upon a procedural default, a retroactive change in law had eliminated that ground as a basis of decision, and the court which issued-the later unexplained order had directed extensive briefing limited to the merits of the federal claim. Or it might be shown that, even though the last reasoned state-court opinion had relied upon a federal ground, the later appeal to the court that issued the unexplained order was plainly out of time, and that the latter court did not ordinarily waive such a procedural default without saying so. *805 While we acknowledge that making the presumption rebuttable will make it less efficient than the categorical approach taken by the Courts of Appeals that have adopted the “look-through” methodology, see Prihoda, supra, at 1383; Harmon, supra, at 1272; Evans, supra, at 123, n. 2; Ellis, supra, at 838, we think it will still simplify the vast majority of cases. The details of state law need not be inquired into unless, if they should be as the habeas petitioner asserts, they would constitute strong evidence that the presumption, as applied, is wrong.

To decide the present case, therefore, we begin by asking which is the last explained state-court judgment on the Miranda claim. Obviously it is not the second denial of habeas by the California Supreme Court; although that was the last judgment, it said absolutely nothing about the reasons for the denial. The first denial of habeas by that court, on December 3, 1986, did cite (without any elaboration) two state cases, Swain and Waltreus. The former holds that facts relied upon in a habeas petition must be alleged with particularity, and the latter that claims presented on direct review ordinarily may not be relitigated on state habeas. Even if we knew that the court intended to apply both of these cases to the Miranda claim (as opposed to the other claims raised by the same petition), that would be irrelevant to the point before us here. Respondent had exhausted his Miranda claim by presenting it on direct appeal, and was not required to go to state habeas at all, see Castille v. Peoples, 489 U. S. 346, 349-350 (1989); state rules against that superfluous recourse have no bearing upon his ability to raise the Miranda claim in federal court. Thus, although the California Supreme Court’s denial of respondent’s first habeas petition to it was not utterly silent, neither was it informative with respect to the question before us.

The prior denials of respondent’s state habeas petitions by the two lower California courts were silent; and, as discussed above, the discretionary denial of review on direct appeal by *806 the California Supreme Court is not even a “judgment.” Thus, the last state opinion on the Miranda claim is that of the Court of Appeal on direct review, and that opinion unequivocally rested upon a state procedural default. We look through the subsequent unexplained denials to that opinion, unless respondent has carried his burden of adducing strong evidence that one of the subsequent courts reached the merits of the federal claim. He has not done so. He claims to be able to show that California habeas courts could have allowed him to relitigate his Miranda claim, in spite of the ordinary state rule barring relitigation of claims raised on direct appeal. See, e. g., Waltreus, 62 Cal. 2d, at 225, 397 P. 2d, at 1005. But even if he established that, to prove that they could do so is not to prove that they did do so — much less to prove that, having done so, they decided the relitigated point on the merits rather than on the basis of the procedural default relied upon in 1978. Respondent has adduced nothing to show that any California court actually reached the merits of his federal claim. The presumption that the California Supreme Court’s last unexplained order did not reach the merits, and that the bar of procedural default subsists, has not been overcome. Federal-court review of the claim is therefore barred unless respondent can establish “cause and prejudice” for the default, see Murray v. Carrier, 477 U. S., at 493, 495-496. The District Court specifically found no cause and prejudice, but, since the Court of Appeals had no occasion to review that holding, we remand for that purpose.

The judgment of the Court of Appeals is reversed, and the case is remanded for further proceedings consistent with this opinion.

So ordered.

In fact he had. The California Court of Appeal decision on direct review shows that all claims, including the Miranda claim and the ineffectiveness claim, were presented to, and specifically addressed by, that court. See App. 15, 17. The District Court’s mistake on this point was apparently caused by respondent’s own statement “that none of his claims were [sic] raised by way of direct appeal.” Id., at 83.

Petitioner also argues that in California original habeas corpus jurisdiction is discretionary, so that denial of a petition is not a “judgment,” and the last state-court “judgment” to which we should look is that of the Court of Appeal on direct review. Respondent concedes that a discretionary denial of review cannot lift a pre-existing procedural bar, and the federal courts are in accord. See Goodwin v. Collins, 910 F. 2d 185, 187 (CA5 1990); Prihoda v. McCaughtry, 910 F. 2d 1379, 1382-1383 (CA7 1990). Respondent denies, however, that California courts have any discretion not to entertain habeas corpus petitions. The state law on this question is not clear, and we shall assume for purposes of this case that respondent is right. We also assume, since the point has not been argued, that Miranda claims such as that raised by respondent are cognizable in federal habeas corpus. See Duckworth v. Eagan, 492 U. S. 195, 205-214 (1989) (O’Connor, J., concurring); cf. Stone v. Powell, 428 U. S. 465 (1976).

The only common circumstance in which the presumption is unrealistic is that in which the later state decision rests upon a prohibition against further state review — for example, an unexplained denial of state habeas resting in fact upon a rule (such as petitioner contends exists in California) preventing the relitigation on state habeas of claims raised on direct appeal. In that circumstance, even though the presumption does not posit the real reason for the later denial, it does produce a result (“looking through” to the last reasoned decision) that is the correct one for federal habeas courts. Since a later state decision based upon ineligibility for further state review neither rests upon procedural default nor lifts a pre-existing procedural default, its effect upon the availability of federal habeas is nil — which is precisely the effect accorded by the “look-through” presumption.