Johnson v. Williams, 568 U.S. ___ (2013)

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.

SUPREME COURT OF THE UNITED STATES

_________________

No. 11–465

_________________

DEBORAH K. JOHNSON, ACTING WARDEN, PE- TITIONER v. TARA SHENEVA WILLIAMS

on writ of certiorari to the united states court of appeals for the ninth circuit

[February 20, 2013]

Justice Alito delivered the opinion of the Court.

The Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA) restricts the circumstances under which a federal habeas court may grant relief to a state prisoner whose claim has already been “adjudicated on the merits in State court.” 28 U. S. C. §2254(d). Specifically, if a claim has been “adjudicated on the merits in State court,” a federal habeas court may not grant relief unless “the adjudication of the claim—

“(1) resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States; or

“(2) resulted in a decision that was based on an unreasonable determination of the facts in light of the evidence presented in the State court proceeding.” Ibid.

Because the requirements of §2254(d) are difficult to meet, it is important whether a federal claim was “adjudicated on the merits in State court,” and this case requires us to ascertain the meaning of the adjudication-on-the merits requirement. This issue arises when a defendant convicted in state court attempts to raise a federal claim, either on direct appeal or in a collateral state proceeding, and a state court rules against the defendant and issues an opinion that addresses some issues but does not expressly address the federal claim in question. If this defendant then raises the same claim in a federal habeas proceeding, should the federal court regard the claim as having been adjudicated on the merits by the state court and apply deference under §2254(d)? Or may the federal court assume that the state court simply overlooked the federal claim and proceed to adjudicate the claim de novo, the course taken by the Court of Appeals in the case at hand?

We believe that the answer to this question follows logically from our decision in Harrington v. Richter, 562 U. S. ___ (2011). In that case, we held that, when a state court issues an order that summarily rejects without discussion all the claims raised by a defendant, including a federal claim that the defendant subsequently presses in a federal habeas proceeding, the federal habeas court must presume (subject to rebuttal) that the federal claim was adjudicated on the merits. We see no reason why this same rule should not apply when the state court addresses some of the claims raised by a defendant but not a claim that is later raised in a federal habeas proceeding.

Applying this rule in the present case, we hold that the federal claim at issue here (a Sixth Amendment jury trial claim) must be presumed to have been adjudicated on the merits by the California courts, that this presumption was not adequately rebutted, that the restrictive standard of review set out in §2254(d)(2) consequently applies, and that under that standard respondent is not entitled to habeas relief. We therefore reverse the judgment of the Court of Appeals.

I

A

In October 1993, respondent Tara Williams took two of her friends for a drive in southern California with the objective of committing a robbery. They stopped at a liquor store in Long Beach, and while Williams waited in the getaway car, her friends stole money from the cash register and fatally shot the store’s owner. Williams then drove one of her friends away, and the other fled on foot. Williams avoided capture for five years but was ultimately apprehended and charged with first-degree murder.

At trial, Williams admitted that she had served as the getaway driver but claimed that she did not know that her friends were going to rob the liquor store at the particular time in question. Instead, she contended that the three friends had agreed only that they would “case” the store and would possibly return later that evening to rob it. The State countered that, regardless of whether Williams knew precisely when and where the robbery was to take place, she had agreed to help commit a robbery and that this was sufficient to provide the predicate for felony murder under California law.

After deliberating for about three hours, the jury foreman sent the judge two notes. The first note asked the following question:

“ ‘Is it legally permissible for a juror to interpret . . . the jury instructions to mean that the conspiracy should involve a plan to commit a specific robbery rather than a general plan to commit robberies in the future?’ ” Tr. 1247.

The second note stated:

“I wish to inform you that we have one juror who . . . has expressed an intention to disregard the law . . . and . . . has expressed concern relative to the severity of the charge (first degree murder).” Id., at 1246.

The judge told the jury that the answer to the question in the first note was “no.” Id., at 1249. Then, over Williams’ objection, the judge briefly questioned the foreman outside the presence of the rest of the jury about the second note. The foreman said that he thought the judge’s answer to the first note might resolve the problem, and the judge instructed the jury to resume its deliberations.

The next morning, once again over Williams’ objection, the judge decided to inquire further about the foreman’s second note. On questioning by the judge and lawyers for both parties, the foreman testified that Juror 6 had brought up past instances of jury nullification. The foreman also expressed doubt about whether Juror 6 was willing to apply the felony-murder rule. The trial judge then ordered questioning of Juror 6, who first denied and then admitted bringing up instances of nullification. Juror 6 also testified that this was a serious case and that he would vote to convict only if he was “very convinced . . . beyond a reasonable doubt.” Id., at 1280. He later clarified that in his view “convinced beyond a reasonable doubt” and “very convinced beyond a reasonable doubt” meant the same thing. Id., at 1281. After taking testi- mony from the remaining jurors, who corroborated the foreman’s testimony to varying degrees, the trial judge dis- missed Juror 6 for bias. With an alternate juror in place, the jury convicted Williams of first-degree murder.

B

On appeal to the California Court of Appeal, Williams argued, among other things, that the discharge of Juror 6 violated both the Sixth Amendment and the California Penal Code, which allows a California trial judge to dismiss a juror who “upon . . . good cause shown to the court is found to be unable to perform his or her duty.” Cal. Penal Code Ann. §1089 (West 2004). Although Williams’ brief challenged the questioning and dismissal of Juror 6 on both state and federal grounds, it did not clearly distinguish between these two lines of authority.

In a written opinion affirming Williams’ conviction, the California Court of Appeal devoted several pages to discussing the propriety of the trial judge’s decision to dis- miss the juror. People v. Taylor, No. B137365 (Mar. 27, 2001). The court held that Juror 6 had been properly dismissed for bias and quoted this Court’s definition of “impartiality” in United States v. Wood, 299 U. S. 123, 145–146 (1936). But despite its extended discussion of Juror 6’s dismissal and the questioning that preceded it, the California Court of Appeal never expressly acknowledged that it was deciding a Sixth Amendment issue.

Williams petitioned the California Supreme Court for review, and while her petition was pending, that court decided People v. Cleveland, 25 Cal. 4th 466, 21 P. 3d 1225 (2001), which held that a trial court had abused its discretion by dismissing for failure to deliberate a juror who appeared to disagree with the rest of the jury about the evidence. The California Supreme Court granted Williams’ petition for review and remanded her case for further consideration in light of this intervening authority. People v. Taylor, No. S097387 (July 11, 2001).

On remand, the California Court of Appeal issued a revised opinion holding that the trial court had not abused its discretion by questioning the jury and dismissing Juror 6. Williams argued that Juror 6—like the holdout juror in Cleveland—was dismissed because he was uncooperative with other jurors who did not share his view of the evidence. But the California Court of Appeal disagreed, explaining that Williams’ argument “not only misstate[d] the evidence,” but also “ignore[d] the trial court’s explanation that it was discharging Juror No. 6 because he had shown himself to be biased, not because he was failing to deliberate or engaging in juror nullification.” People v. Taylor, No. B137365 (Jan. 18, 2002), App. to Pet. for Cert. 105a. As in its earlier opinion, the California Court of Appeal quoted our definition of juror bias in Wood, but the court did not expressly acknowledge that Williams had invoked a federal basis for her argument. Despite that omission, however, Williams did not seek rehearing or other- wise suggest that the court had overlooked her federal claim. Instead, she filed another petition for review in the California Supreme Court, but this time that court denied relief in a one-sentence order. People v. Taylor, No. S104661 (Apr. 10, 2002), App. to Pet. for Cert. 85a.

Williams sought but failed to obtain relief through state habeas proceedings, and she then filed a federal habeas petition under 28 U. S. C. §2254. The District Court applied AEDPA’s deferential standard of review for claims previously adjudicated on the merits and denied relief. Williams v. Mitchell, No. 03–2691 (CD Cal., May 30, 2007), App. to Pet. for Cert. 57a. In so holding, the District Court adopted a Magistrate Judge’s finding that the evidence “amply support[ed] the trial judge’s determination that good cause existed for the discharge of Juror 6.” Williams v. Mitchell, No. 03–2691 (CD Cal., Mar. 19, 2007), id., at 70a.

The Ninth Circuit reversed. Unlike the District Court, the Ninth Circuit declined to apply the deferential standard of review contained in §2254(d). The Ninth Circuit took this approach because it thought it “obvious” that the State Court of Appeal had “overlooked or disregarded” Williams’ Sixth Amendment claim.[1] Williams v. Cavazos, 646 F. 3d 626, 639 (2011). The Ninth Circuit reasoned that Cleveland, the State Supreme Court decision on which the State Court of Appeal had relied, “was not a constitutional decision,” 646 F. 3d, at 640, and the Ninth Circuit attributed no significance to the state court’s citation of our decision in Wood. Reviewing Williams’ Sixth Amendment claim de novo, the Ninth Circuit applied its own precedent and held that the questioning and dismissal of Juror 6 violated the Sixth Amendment. 646 F. 3d, at 646–647. We granted the warden’s petition for a writ of certiorari, 565 U. S. ___ (2012), in order to decide whether the Ninth Circuit erred by refusing to afford AEDPA deference to the California Court of Appeal’s decision.

II

A

As noted above, AEDPA sharply limits the circum- stances in which a federal court may issue a writ of habeas corpus to a state prisoner whose claim was “adjudicated on the merits in State court proceedings.” 28 U. S. C. §2254(d). In Richter, 562 U. S., at ___ (slip op., at 10), we held that §2254(d) “does not require a state court to give reasons before its decision can be deemed to have been ‘adjudicated on the merits.’ ” Rather, we explained, “[w]hen a federal claim has been presented to a state court and the state court has denied relief, it may be presumed that the state court adjudicated the claim on the merits in the absence of any indication or state-law procedural principles to the contrary.” Id., at ___ (slip op., at 9).

Our reasoning in Richter points clearly to the answer to the question presented in the case at hand. Although Richter itself concerned a state-court order that did not address any of the defendant’s claims, we see no reason why the Richter presumption should not also apply when a state-court opinion addresses some but not all of a defendant’s claims. There would be a reason for drawing a distinction between these two situations if opinions issued by state appellate courts always separately addressed every single claim that is mentioned in a defendant’s papers. If there were such a uniform practice, then federal habeas courts could assume that any unaddressed federal claim was simply overlooked.

No such assumption is warranted, however, because it is not the uniform practice of busy state courts to discuss separately every single claim to which a defendant makes even a passing reference. On the contrary, there are several situations in which state courts frequently take a different course.

First, there are circumstances in which a line of state precedent is viewed as fully incorporating a related federal constitutional right. In California, for example, the state constitutional right to be present at trial “ ‘is generally coextensive with’ ” the protections of the Federal Constitution. People v. Butler, 46 Cal. 4th 847, 861, 209 P. 3d 596, 606 (2009); see also, e.g., Commonwealth v. Prunty, 462 Mass. 295, 305, n. 14, 968 N. E. 2d 361, 371, n. 14 (2012) (standard for racial discrimination in juror selection “ ‘is the same under the Federal Constitution and the [Massachusetts] Declaration of Rights’ ”); State v. Krause, 817 N. W. 2d 136, 144 (Minn. 2012) (“ ‘The due process protection provided under the Minnesota Constitution is identical to the due proces[s] guaranteed under the Constitution of the United States’ ”); State v. Engelhardt, 280 Kan. 113, 122, 119 P. 3d 1148, 1158 (2005) (observing that a Kansas statute is “analytically and functionally identical to the requirements under the Confrontation Clause and the Due Process Clause of the federal Constitution”). In this situation, a state appellate court may regard its discussion of the state precedent as sufficient to cover a claim based on the related federal right.

Second, a state court may not regard a fleeting reference to a provision of the Federal Constitution or federal precedent as sufficient to raise a separate federal claim. Federal courts of appeals refuse to take cognizance of arguments that are made in passing without proper development. See, e.g., United States v. Cloud, 680 F. 3d 396, 409, n. 7 (CA4 2012); United States v. Mitchell, 502 F. 3d 931, 953, n. 2 (CA9 2007); United States v. Charles, 469 F. 3d 402, 408 (CA5 2006); Reynolds v. Wagner, 128 F. 3d 166, 178 (CA3 1997); Carducci v. Regan, 714 F. 2d 171, 177 (CADC 1983). State appellate courts are entitled to follow the same practice.

Third, there are instances in which a state court may simply regard a claim as too insubstantial to merit discussion. Indeed, the California Court of Appeal has expressly stated that it has no obligation to address claims that lack arguable merit. See People v. Rojas, 118 Cal. App. 3d 278, 290, 173 Cal. Rptr. 91, 93 (1981). That court has explained: “In an era in which there is concern that the quality of justice is being diminished by appellate backlog with its attendant delay, which in turn contributes to a lack of finality of judgment, it behooves us as an appellate court to ‘get to the heart’ of cases presented and dispose of them expeditiously.” Ibid. See also People v. Burke, 18 Cal. App. 72, 79, 122 P. 435, 439 (1912) (“The author of an opinion . . . must follow his own judgment as to the degree of elaboration to be accorded to the treatment of any proposition and as to the questions which are worthy of notice at all” (emphasis added)). While it is preferable for an appellate court in a criminal case to list all of the arguments that the court recognizes as having been properly presented, see R. Aldisert, Opinion Writing 95–96 (3d ed. 2012), federal courts have no authority to impose mandatory opinion-writing standards on state courts, see Coleman v. Thompson, 501 U. S. 722, 739 (1991) (“[W]e have no power to tell state courts how they must write their opinions”). The caseloads shouldered by many state appellate courts are very heavy,[2] and the opinions issued by these courts must be read with that factor in mind.

In sum, because it is by no means uncommon for a state court to fail to address separately a federal claim that the court has not simply overlooked, we see no sound reason for failing to apply the Richter presumption in cases like the one now before us. When a state court rejects a federal claim without expressly addressing that claim, a federal habeas court must presume that the federal claim was adjudicated on the merits—but that presumption can in some limited circumstances be rebutted.

B

Not satisfied with a strong but rebuttable presumption, petitioner urges us to make the presumption irrebuttable. Specifically, petitioner contends that a state court must be regarded as having adjudicated a federal claim on the merits if the state court addressed “the substance of [an] asserted trial error.” Brief for Petitioner 27. Suppose, for example, that a defendant claimed in state court that something that occurred at trial violated both a provision of the Federal Constitution and a related provision of state law, and suppose further that the state court, in denying relief, made no reference to federal law. According to petitioner’s argument, a federal habeas court would be required to proceed on the assumption that the federal claim was adjudicated on the merits.

This argument goes too far. To be sure, if the state-law rule subsumes the federal standard—that is, if it is at least as protective as the federal standard—then the federal claim may be regarded as having been adjudicated on the merits. See Early v. Packer, 537 U. S. 3, 8 (2002) (per curiam). But what if, for example, in at least some circumstances the state standard is less protective? Or what if the state standard is quite different from the federal standard, and the defendant’s papers made no effort to develop the basis for the federal claim? What if a provision of the Federal Constitution or a federal precedent was simply mentioned in passing in a footnote or was buried in a string cite? In such circumstances, the presumption that the federal claim was adjudicated on the merits may be rebutted—either by the habeas petitioner (for the purpose of showing that the claim should be considered by the federal court de novo) or by the State (for the purpose of showing that the federal claim should be regarded as procedurally defaulted). See Coleman, supra, at 739 (rebuttable presumption of no independent and adequate state ground applies so long as “it fairly appears that a state court judgment rested primarily on federal law or was interwoven with federal law”). Thus, while the Richter presumption is a strong one that may be rebutted only in unusual circumstances, it is not irrebuttable.[3] “Per se rules should not be applied . . . in situations where the generalization is incorrect as an empirical matter,” Coleman, 501 U. S., at 737, and an irrebuttable presumption that state courts never overlook federal claims would occasionally miss the mark.

The language of 28 U. S. C. §2254(d) makes it clear that this provision applies only when a federal claim was “adjudicated on the merits in State court.” A judgment is normally said to have been rendered “on the merits” only if it was “delivered after the court . . . heard and evaluated the evidence and the parties’ substantive arguments.” Black’s Law Dictionary 1199 (9th ed. 2009) (emphasis added). And as used in this context, the word “merits” is defined as “[t]he intrinsic rights and wrongs of a case as determined by matters of substance, in distinction from matters of form.” Webster’s New International Dictionary 1540 (2d ed. 1954) (emphasis added); see also, e.g., 9 Oxford English Dictionary 634 (2d ed. 1989) (“the intrinsic ‘rights and wrongs’ of the matter, in contradistinction to extraneous points such as the competence of the tribunal or the like” (emphasis added)); Random House Dictionary of the English Language 897 (1967) (“the intrinsic right and wrong of a matter, as a law case, unobscured by procedural details, technicalities, personal feelings, etc.” (emphasis added)). If a federal claim is rejected as a result of sheer inadvertence, it has not been evaluated based on the intrinsic right and wrong of the matter. Justice Scalia is surely correct that such claims have been adjudicated and present federal questions we may review, post, at 3–4, but it does not follow that they have been adjudicated “on the merits.” By having us nevertheless apply AEDPA’s deferential standard of review in such cases, petitioner’s argument would improperly excise §2254(d)’s on-the-merits requirement.

Nor does petitioner’s preferred approach follow inexorably from AEDPA’s deferential architecture. Even while leaving “primary responsibility” for adjudicating federal claims to the States, Woodford v. Visciotti, 537 U. S. 19, 27 (2002) (per curiam), AEDPA permits de novo review in those rare cases when a state court decides a federal claim in a way that is “contrary to” clearly established Supreme Court precedent, see Panetti v. Quarterman, 551 U. S. 930, 953 (2007). When the evidence leads very clearly to the conclusion that a federal claim was inadvertently overlooked in state court, §2254(d) entitles the prisoner to an unencumbered opportunity to make his case before a fed- eral judge.

We are not persuaded that applying a rebuttable presumption in this context will be unduly burdensome for federal courts. Before Richter, every Court of Appeals to consider the issue allowed a prisoner to argue that a state court had overlooked his federal claim.[4] That approach did not prompt an unmanageable flood of litigation, and we see no reason to fear that it will do so now.

III

Applying the presumption of merits adjudication to the facts of this case, we hold that the Ninth Circuit erred by finding that the California Court of Appeal overlooked Williams’ Sixth Amendment claim. Several facts make this conclusion inescapable.

Most important is the state court’s discussion of Cleveland, 25 Cal. 4th 466, 21 P. 3d 1225, a California Supreme Court decision on which the Court of Appeal solicited briefing. Cleveland held that a California trial court, “if put on notice that a juror is not participating in delib- erations,” may “conduct ‘whatever inquiry is reasonably necessary to determine’ whether such grounds exist and . . . discharge the juror if it appears as a ‘demonstrable reality’ that the juror is unable or unwilling to deliberate.” Id., at 484, 21 P. 3d, at 1237 (citations omitted). The Cleveland court acknowledged “[t]he need to protect the sanctity of jury deliberations,” id., at 476, 21 P. 3d, at 1231, and included a lengthy discussion of three Federal Court of Appeals cases that it said had “considered these issues in depth,” id., at 480–484, 21 P. 3d, at 1234–1237. Those three cases—United States v. Symington, 195 F. 3d 1080 (CA9 1999), United States v. Thomas, 116 F. 3d 606 (CA2 1997), and United States v. Brown, 823 F. 2d 591 (CADC 1987)—concern the discharge of holdout jurors in federal court. Each case discusses the Sixth Amendment right to a jury trial and concludes that a trial court should not inquire further if it appears that there is “ ‘any reasonable possibility that the impetus for a juror’s dismissal stems from the juror’s views on the merits of the case.’ ” Cleveland, supra, at 484, 21 P. 3d, at 1237 (quoting Symington, supra, at 1087); see also Thomas, supra, at 621–622; Brown, supra, at 596. Though the Cleveland court found much to praise in these decisions, it expressly declined to follow them on this point. 25 Cal. 4th, at 483–484, 21 P. 3d, at 1236–1237.

Cleveland did not expressly purport to decide a federal constitutional question, but its discussion of Symington, Thomas, and Brown shows that the California Supreme Court understood itself to be deciding a question with federal constitutional dimensions. See 25 Cal. 4th, at 487, 21 P. 3d, at 1239 (Werdegar, J., concurring) (emphasizing importance of careful appellate review in juror discharge cases in light of the “constitutional dimension to the problem”). Indeed, it is difficult to imagine the California Supreme Court announcing an interpretation of Cal. Penal Code Ann. §1089 that it believed to be less protective than the Sixth Amendment, as any such interpretation would provide no guidance to state trial judges bound to follow both state and federal law.

The Ninth Circuit’s conclusion to the contrary rested on the fact that Cleveland refused to follow Symington, Brown, and Thomas. 646 F. 3d, at 640. But the views of the federal courts of appeals do not bind the California Supreme Court when it decides a federal constitutional question, and disagreeing with the lower federal courts is not the same as ignoring federal law. The Ninth Circuit’s apparent assumption that the California Supreme Court could not refuse to follow federal court of appeals precedent without disregarding the Federal Constitution would undo §2254(d)’s “contrary to” provision, which requires deference unless a state court fails to follow Supreme Court precedent. 28 U. S. C. §2254(d)(1).

Regardless of whether a California court would consider Williams’ §1089 and Sixth Amendment claims to be perfectly coextensive, the fact that these claims are so similar makes it unlikely that the California Court of Appeal decided one while overlooking the other. Indeed, it is dif- ficult to imagine any panel of appellate judges reading Cleveland and passing on the propriety of dismissing a holdout juror under §1089 without realizing that such situations also bear on the federal constitutional right to a fair trial. The California Court of Appeal’s quotation of our definition of “impartiality” from Wood, 299 U. S., at 145–146, points to the same conclusion, confirming that the state court was well aware that the questioning and dismissal of Juror 6 implicated both state and federal law.

Williams’ litigation strategy supports the same result. Throughout her state proceedings, Williams treated her state and federal claims as interchangeable, and it is hardly surprising that the state courts did so as well. See Brief for Appellant in No. B137365 (Cal. App.), App. 29 (citing §1089 precedent and concluding that Williams “was accordingly denied her Sixth Amendment right to a unanimous jury”). After the California Court of Appeal rendered its decision, Williams neither petitioned that court for rehearing nor argued in the subsequent state and federal proceedings that the state court had failed to adjudicate her Sixth Amendment claim on the merits. The possibility that the California Court of Appeal had simply overlooked Williams’ Sixth Amendment claim apparently did not occur to anyone until that issue was raised by two judges during the oral argument in the Ninth Circuit. See 646 F. 3d, at 638, n. 7. Williams presumably knows her case better than anyone else, and the fact that she does not appear to have thought that there was an oversight makes such a mistake most improbable.

We think it exceedingly unlikely that the California Court of Appeal overlooked Williams’ federal claim, and the Ninth Circuit’s judgment to the contrary is reversed. The case is remanded for further proceedings consistent with this opinion.

It is so ordered.

SUPREME COURT OF THE UNITED STATES

_________________

No. 11–465

_________________

DEBORAH K. JOHNSON, ACTING WARDEN, PE- TITIONER v. TARA SHENEVA WILLIAMS

on writ of certiorari to the united states court of appeals for the ninth circuit

[February 20, 2013]

Justice Scalia, concurring in the judgment.

I agree with the Court’s rejection of the proposition that a judgment denying a federal claim is irrebuttably presumed to have been “adjudicated on the merits” within the meaning of 28 U. S. C. §2254(d). I disagree, however, that one of the grounds on which the rebuttal may rely is that the federal claim was “inadvertently overlooked.” Ante, at 13. In my view the rebuttal must consist of a show- ing, based on the explicit text of the court’s order, or upon standard practice and understanding in the jurisdiction with regard to the meaning of an ambiguous text, that the judgment did not purport to decide the federal question. “Decided after due consideration” is not, and has never been, the meaning of the legal term of art “decided on the merits,” and giving it that meaning burdens our lower courts with an unusual subjective inquiry that demeans state courts and will be a fertile source of litigation and delay.

In the Court’s view, a habeas petitioner receives de novo review if he can prove that the state court, although addressing his state claim, overlooked his federal claim. A nonexhaustive list of factors, we are told, may bear on the analysis: state-court opinion-writing practices, ante, at 8, 9; state-law precedents and whether and how they incorporate federal law, ante, at 8; substantiality of the federal claim, ante, at 9; citations to federal cases in state-court opinions (or citations to state cases that contain citations to federal cases), ante, at 13–14; the degree of similarity between the federal and state claim, ante, at 15; a petitioner’s “litigation strategy,” ante, at 15–16; and other clues that may possibly illuminate the inner thought processes of a state-court judge. Only after conducting its own detective work does the Court conclude that the federal claim was not overlooked in this case.

This complex exercise is unnecessary. A judgment that denies relief necessarily denies—and thus adjudicates—all the claims a petitioner has raised. See 1 H. Black, Law of Judgments §1, p. 2 (2d ed. 1902) (“[T]he judgment necessarily affirms, or else denies, that [an alleged] duty or . . . liability rests upon the person against whom the aid of the law is invoked”); id., §24, at 37. The judgment itself gives conclusive expression that the claims have been con- sidered and rejected—whatever the individual judge might have been pondering (or not pondering). At common law the formal language traditionally preceding the announcement of a court’s judgment was “consideratum est per curiam” (“It is considered by the court”). See Black’s Law Dictionary 349–350 (9th ed. 2009); 1 Bouvier’s Law Dictionary 619 (8th ed. 1914).

The Court maintains that “[i]f a federal claim is rejected as a result of sheer inadvertence, it has not been evalu- ated based on the intrinsic right and wrong of the matter,” ante, at 12. Perhaps not, but it nonetheless may have been rejected “on the merits.” That phrase does not suggest a line between a considered rejection of a claim and an unconsidered, inadequately considered, or inadvertent rejection. Rather, it refers to a “determination that there exist or do not exist grounds entitling a petitioner” to relief under his claim, as contrasted with a “denial for such rea- sons as failure to exhaust, procedural default, or statute-of-limitations bar.” Gonzalez v. Crosby, 545 U. S. 524, 532, n. 4 (2005). An “adjudication on the merits” is “best understood by stating what it is not: it is not a resolution of a claim on procedural grounds.” Muth v. Frank, 412 F. 3d 808, 815 (CA7 2005). And, as we have affirmed and reaffirmed recently, where a claim has been denied, but it is unclear from the record whether the denial was on the merits or on another basis, we presume the former. Harrington v. Richter, 562 U. S. ___, ___ (2011) (slip op., at 9–10) (citing Harris v. Reed, 489 U. S. 255, 265 (1989)); see also Coleman v. Thompson, 501 U. S. 722, 732–733 (1991).

We apply a presumption of merits determination in that sense not just with respect to §2254(d) but for other purposes as well. We have long applied it, for example, in determining whether a claim is barred by res judicata:

“Ordinarily, such a question is answered by a mere inspection of the decree—the presumption being that a dismissal in equity, without qualifying words, is a final decision on the merits. That presumption of finality . . . disappears whenever the record shows that the court did not pass upon the merits but dismissed the bill because of a want of jurisdiction, for want of parties, because the suit was brought prematurely, because the plaintiff had a right to file a subsequent bill on the same subject-matter, or on any other ground not going to the merits.” Swift v. McPherson, 232 U. S. 51, 55–56 (1914) (emphasis added); see also Hubbell v. United States, 171 U. S. 203, 207 (1898); Durant v. Essex Co., 7 Wall. 107, 109 (1868).

We also apply a presumption of merits determination in the sense I have described for purposes of 28 U. S. C. §1257, which imposes a federal-question requirement as a condition of this Court’s appellate jurisdiction. Michigan v. Long, 463 U. S. 1032, 1040–1041 (1983). Indeed, the application of the presumption in direct-review cases was the genesis of the presumption in federal habeas cases. The condition for federal habeas—that the federal question must have been addressed on the merits by the state courts—did not originate with the enactment of the Antiterrorism and Effective Death Penalty Act in 1996, but was established as early as 1977 in Wainwright v. Sykes, 433 U. S. 72, 81, 86–87. We described the assessment of whether that requirement was met as presenting “the same problem of ambiguity that this Court resolved in Michigan v. Long.” Harris, 489 U. S., at 262. And indeed, we described the habeas requirement as an application of the “adequate and independent state ground doctrine,” which inquires whether a “finding of procedural default will bar federal habeas review.” Ibid. It is of course unthinkable that a state-court resolution of a federal question will escape our review under §1257 if it is inadvertent rather than intentional.

Given this background, there is no reason to believe that AEDPA established a new and peculiar regime in which the federal habeas court must make one assessment of whether the federal question has been decided “on the merits” for purposes of determining its authority to review the question (a Long assessment which counts, as §1257 cases count, inadvertent resolution of a federal question); and then must proceed to a different assessment of “on the merits” (one that does not count inadvertent resolution) for purposes of determining whether deference to the state-court judgment is required.

But, it will be argued, how can a court “defer” to a state-court determination that was in fact never made? Must not one first be sure it exists before one can accord it respect? The answer is no; according respect only to determinations that have for-sure been made is demonstrably not the scheme that AEDPA envisions. Federal ha- beas courts defer to state determinations that may in fact never have been made whenever they find a summary, unexplained rejection of a federal claim to be sustainable (e.g., not contrary to clearly established federal law as determined by this Court). The validating basis that the federal habeas court posits need not have been the one that the state court actually relied upon; the state court may well have applied a theory that was flat-out wrong, and may not have made the subsidiary determinations (including factual assessments) necessary to support the correct theory. That does not matter. For what is accorded deference is not the state court’s reasoning but the state court’s judgment, which is presumed to be supported by whatever valid support was available. See Harrington, supra, at ___ (slip op., at 12) (“Under §2254(d) a habeas court must determine what arguments or theories supported or, as here, could have supported, the state court’s decision”). Indeed, the deference with regard to the basis of decision is much more “blind” than the deference I assert is necessary in the present case. I demand a state-court statement (contained in the unqualified terms of its judgment) that it has rejected the federal claim; I sim- ply refuse to question the veracity of that statement. By contrast, no statement is ever even required that the state court relied upon the theory of federal law that the habeas court finds validating.

I doubt that the Court is prepared to abide by its novel interpretation of “on the merits” for purposes of §2254(d). Imagine that the state court formulated its judgment as follows: “All claims raised by the defendant have been considered and denied.” I cannot believe that the Court would require federal courts to test the veracity of that statement. Yet, as we have described, that is precisely what an (unadorned) judgment denying relief already conveys. Although the Court acknowledges that “ ‘[w]e have no power to tell state courts how they must write their opinions,’ ” ante, at 9, its analysis would turn solely on how the order of judgment is styled.

Resolution of this case is direct: Respondent’s claim was “adjudicated on the merits,” because the state court rendered a judgment rejecting all her claims, and the judgment gave no indication (such as a statement that it was “without prejudice”) that it was based on a procedural or other nonmerits ground.

The Court’s novel resolution of the “on the merits” question produces a clear enough answer in this case. The weight of the evidence demonstrated that it was “exceedingly unlikely” that the state court overlooked the federal claim. Ante, at 16. But such ready resolution will not be commonplace. Consider another case, where the federal and state claims are not related, where there is no relevant state precedent referring to federal law, where state law might be interpreted as less defendant-friendly than the federal standard, or where a confluence of such factors exists. The answer to whether the federal claim has been “evaluated based on the intrinsic right and wrong of the matter” is anybody’s guess. One thing, however, is certain: The Court’s case-by-case approach will guarantee protracted litigation over whether a state-court judge was aware of a claim on the day he rejected it.

The Court tells us not to worry about a flood of liti- gation, because the Courts of Appeals have previously al- lowed arguments from petitioners that the state courts overlooked their federal claims. Ante, at 13, and n. 4 (citing cases). But many of those cases applied a much simpler (and even less justifiable) test than the one adopted today: if the federal claim was not addressed in the opinion, then it was not adjudicated on the merits. See, e.g., Lyell v. Renico, 470 F. 3d 1177, 1181–1182 (CA6 2006); Fortini v. Murphy, 257 F. 3d 39, 47 (CA1 2001). And even those courts that attempted to “divin[e] the thought processes of” the judge limited their inquiry to “what a state court has said.” Brown v. Luebbers, 371 F. 3d 458, 461 (CA8 2004) (emphasis added); see also, e.g., Chadwick v. Janecka, 312 F. 3d 597, 606 (CA3 2002). By contrast, the Court today asks whether a judge thought about the merits of an unaddressed claim, and leaves on the table any evidence relevant to that inquiry.

This newly-sponsored enterprise of probing the judicial mind is inappropriately intrusive upon state-court processes. Are federal habeas courts now to consider evidence relevant to the internal deliberations of the state judiciary? Can a petitioner introduce testimony showing that state-court judges—because of time constraints, heavy caseloads, or other reasons—fail to read the briefs but leave that to their assistants, whose recommendations they rarely reject? Or testimony showing that, typically, only one judge on the state-court appellate panel reads the briefs and considers all the claims, and the others simply join the drafted order? Has there been an “adjudication on the merits” then? Future litigation will supply the answers.

For these reasons, I do not join the opinion of the Court

and concur only in the judgment.