LAMBRIX v. SINGLETARY

In the sentencing phase of the trial at which petitioner Lambrix was convicted on two counts of first-degree murder, the Florida state-court jury rendered an advisory verdict recommending death sentences on both counts. Finding numerous aggravating circumstances in connection with both murders, and no mitigating circumstances as to either, the trial court sentenced Lambrix to death on both counts. After his conviction and sentence were upheld on direct and collateral review by the Florida courts, he filed a habeas petition in the Federal District Court, which rejected all of his claims. While his appeal was pending before the Eleventh Circuit, this Court decided in Espinosa v. Florida, 505 U.S. 1079, that if the sentencing judge in a "weighing" State ( i.e., a State such as Florida that requires specified aggravating circumstances to be weighed against any mitigating circumstances at a capital trial's sentencing phase) is required to give deference to a jury's advisory sentencing recommendation, then neither the jury nor the judge is constitutionally permitted to weigh invalid aggravating circumstances. Since one of Lambrix's claims was that his sentencing jury was improperly instructed on the "especially heinous, atrocious, or cruel" aggravator, Espinosa had obvious relevance to his habeas petition. The Eleventh Circuit held its proceedings in abeyance to permit Lambrix to present his Espinosa claim to the Florida Supreme Court, which rejected the claim without considering its merits on the ground that the claim was procedurally barred. Without even acknowledging the procedural bar, the Eleventh Circuit denied relief, ruling that Espinosa announced a "new rule" which cannot be applied retroactively on federal habeas under Teague v. Lane, 489 U.S. 288.

Held:

1. Although the question whether a federal court should resolve a claim of procedural bar before considering a claim of Teague bar has not previously been presented, the Court's opinions — most particularly, Coleman v. Thompson, 501 U.S. 722 — suggest that the procedural bar issue should ordinarily be considered first. The Court nonetheless chooses not to resolve this case on the procedural bar ground. Lambrix asserts several reasons why procedural bar does not apply, the validity of which is more appropriately determined by the lower federal courts, which are more familiar with the procedural practices of the States in which they sit. Rather than prolong this litigation by a remand, the Court proceeds to decide the question presented. Pp. 522-525.

2. A prisoner whose conviction became final before Espinosa is foreclosed from relying on that decision in a federal habeas proceeding. Pp. 525-540.

(a) To apply Teague, a federal habeas court must: (1) determine the date on which the defendant's conviction became final; (2) survey the legal landscape as it existed on that date to determine whether a state court then considering the defendant's claim would have felt compelled by existing precedent to conclude that the rule the defendant seeks was constitutionally required; and (3) if not, consider whether the relief sought falls within one of two narrow exceptions to nonretroactivity. Pp. 525-527.

(b) A survey of the legal landscape as of the date that Lambrix's conviction became final shows that Espinosa was not dictated by then-existing precedent, but announced a "new rule" as defined in Teague. It is significant that Espinosa, supra, at 1082, cited only a single case in support of its central conclusion, Baldwin v. Alabama, 472 U.S. 372, 382, and introduced that lone citation with a "cf." — an introductory signal indicating authority that supports the point in dictum or by analogy. Baldwin states, on the page that Espinosa cites, 472 U.S., at 382, that the defendant's Espinosa-like argument "conceivably might have merit" in circumstances not present in that case. The decisions relied on most heavily by Lambrix — Godfrey v. Georgia, 446 U.S. 420; Maynard v. Cartwright, 486 U.S. 356; and Clemons v. Mississippi, 494 U.S. 738 — do not dictate the result ultimately reached in Espinosa. Rather, a close examination of the Florida death-penalty scheme, in light of cases such as Proffitt v. Florida, 428 U.S. 242, 253 (joint opinion); id., at 260-261 (White, J., concurring in judgment); and Spaziano v. Florida, 468 U.S. 447, 451, 466, indicates that a reasonable jurist considering the matter at the time Lambrix's sentence became final could have reached a result different from Espinosa. That conclusion is confirmed by Walton v. Arizona, 497 U.S. 639, 653-654. The fact that Espinosa was handed down as a per curiam without oral argument is insignificant, since the decision followed by just three weeks Sochor v. Florida, 504 U.S. 527, in which the identical issue was fully briefed and argued, but could not be decided for jurisdictional reasons. Pp. 527-539.

(c) Espinosa's new rule does not fall within either of the exceptions to this Court's nonretroactivity doctrine. The first exception plainly has no application, since Espinosa neither decriminalizes a class of conduct nor prohibits the imposition of capital punishment on a particular class of persons. E.g., Saffle v. Parks, 494 U.S. 484, 494-495. Lambrix does not contend that the second exception — for watershed rules of criminal procedure implicating the criminal proceeding's fundamental fairness and accuracy — applies to Espinosa errors, and Sawyer v. Smith, 497 U.S. 227, 241-244, makes clear that it does not. Pp. 539-540.

72 F.3d 1500, affirmed.

SCALIA, J., delivered the opinion of the Court, in which REHNQUIST, C.J., and KENNEDY, SOUTER, and THOMAS, JJ., joined. STEVENS, J., filed a dissenting opinion, in which GINSBURG and BREYER, JJ., joined. O'CONNOR, J., filed a dissenting opinion, post, p. 546.

JUSTICE SCALIA delivered the opinion of the Court.

We granted certiorari in this case to consider whether a prisoner whose conviction became final before our decision in Espinosa v. Florida, 505 U.S. 1079 (1992), is foreclosed from relying on that decision in a federal habeas corpus proceeding because it announced a "new rule" as defined in Teague v. Lane, 489 U.S. 288 (1989).

I

On February 5, 1983, Cary Michael Lambrix and his girlfriend, Francis Smith, met Clarence Moore and Aleisha Bryant at a local tavern. The two couples returned to Lambrix's trailer for dinner, where Lambrix killed Moore and Bryant in brutal fashion. Lambrix was convicted on two counts of first-degree murder. In the sentencing phase of trial, the jury rendered an advisory verdict recommending that the trial court sentence Lambrix to death on both counts. The trial court, after finding five aggravating circumstances in connection with the murder of Moore, four aggravating circumstances in connection with the murder of Bryant, and no mitigating circumstances as to either murder, sentenced Lambrix to death on both counts. Lambrix's conviction and sentence were upheld on direct appeal by the Florida Supreme Court. Lambrix v. State, 494 So.2d 1143 (1986).

After the Florida courts denied his repeated efforts to obtain collateral relief, Lambrix v. Dugger, 529 So.2d 1110 (Fla. 1988); Lambrix v. State, 534 So.2d 1151 (Fla. 1988); Lambrix v. State, 559 So.2d 1137 (Fla. 1990), Lambrix filed a petition for a writ of habeas corpus pursuant to 28 U.S.C. § 2254 in the United States District Court for the Southern District of Florida; that court rejected all of his claims. While Lambrix's appeal was pending before the Court of Appeals for the Eleventh Circuit, this Court decided Espinosa v. Florida, supra, which held that if the sentencing judge in a "weighing" State ( i.e., a State that requires specified aggravating circumstances to be weighed against any mitigating circumstances at the sentencing phase of a capital trial) is required to give deference to a jury's advisory sentencing recommendation, then neither the jury nor the judge is constitutionally permitted to weigh invalid aggravating circumstances. Since Florida is such a State, and since one of Lambrix's claims was that his sentencing jury was improperly instructed on the "especially heinous, atrocious, or cruel" (HAC) aggravator, Espinosa had obvious relevance to his habeas petition. Rather than address this issue in the first instance, however, the Eleventh Circuit held its proceedings in abeyance to permit Lambrix to present his Espinosa claim to the Florida state courts.

The Florida Supreme Court rejected Lambrix's Espinosa claim without considering its merits on the ground that the claim was procedurally barred. Lambrix v. Singletary, 641 So.2d 847 (1994). That court explained that although Lambrix had properly preserved his Espinosa objection at trial by requesting a limiting instruction on the HAC aggravator, he had failed to raise the issue on direct appeal. 641 So.2d, at 848. The Florida Supreme Court also rejected Lambrix's claim that the procedural bar should be excused because his appellate counsel was ineffective in failing to raise the forfeited issue, explaining that this claim was itself procedurally barred and was, in any event, meritless. Id., at 848-849.

After the Florida Supreme Court entered judgment against Lambrix, the Eleventh Circuit adjudicated his habeas petition. Without even acknowledging the procedural bar — which was expressly raised and argued by the State — the Court of Appeals proceeded to address the Espinosa claim, and determined that Espinosa announced a new rule which cannot be applied retroactively on federal habeas under Teague v. Lane, 489 U.S. 288 (1989). 72 F.3d 1500, 1503 (1996). We granted certiorari. 519 U.S. 958 (1996).

II

Before turning to the question presented in this case, we pause to consider the State's contention that Lambrix's Espinosa claim is procedurally barred because he failed to contend that the jury was instructed with a vague HAC aggravator on his direct appeal to the Florida Supreme Court. According to the State, the Florida Supreme Court "has consistently required that an Espinosa issue must have been objected to at trial and pursued on direct appeal in order to be reviewed in postconviction proceedings." Brief for Respondent 30, citing Chandler v. Dugger, 634 So.2d 1066, 1069 (Fla. 1994), Jackson v. Dugger, 633 So.2d 1051, 1055 (Fla. 1993), and Henderson v. Singletary, 617 So.2d 313 (Fla.), cert. denied, 507 U.S. 1047 (1993).

In Coleman v. Thompson, 501 U.S. 722, 729 (1991), we reaffirmed that this Court "will not review a question of federal law decided by a state court if the decision of that court rests on a state law ground that is independent of the federal question and adequate to support the judgment." See also Harris v. Reed, 489 U.S. 255, 262 (1989). We in fact lack jurisdiction to review such independently supported judgments on direct appeal: since the state-law determination is sufficient to sustain the decree, any opinion of this Court on the federal question would be purely advisory. Herb v. Pitcairn, 324 U.S. 117, 125-126 (1945); see also Sochor v. Florida, 504 U.S. 527, 533-534, and n. * (1992). The "independent and adequate state ground" doctrine is not technically jurisdictional when a federal court considers a state prisoner's petition for habeas corpus pursuant to 28 U.S.C. § 2254, since the federal court is not formally reviewing a judgment, but is determining whether the prisoner is "in custody in violation of the Constitution or laws or treaties of the United States." We have nonetheless held that the doctrine applies to bar consideration on federal habeas of federal claims that have been defaulted under state law. Coleman, supra, at 729-730, 750; see also Wainwright v. Sykes , 433 U.S. 72, 81, 82 (1977), discussing Brown v. Allen, 344 U.S. 443, 486-487 (1953), and Ex parte Spencer , 228 U.S. 652 (1913); Harris, supra, at 262.

Application of the "independent and adequate state ground" doctrine to federal habeas review is based upon equitable considerations of federalism and comity. It "ensures that the States' interest in correcting their own mistakes is respected in all federal habeas cases." Coleman, 501 U.S., at 732. "[A] habeas petitioner who has failed to meet the State's procedural requirements for presenting his federal claims has deprived the state courts of an opportunity to address those claims in the first instance." Ibid. If the "independent and adequate state ground" doctrine were not applied, a federal district court or court of appeals would be able to review claims that this Court would have been unable to consider on direct review. See id., at 730-731.

We have never had occasion to consider whether a federal court should resolve a State's contention that a petitioner's claim is procedurally barred before considering whether his claim is Teague barred. Our opinions, however — most particularly, Coleman — certainly suggest that the procedural-bar issue should ordinarily be considered first. It was speculated at oral argument that the Court of Appeals may have resolved the Teague issue without first considering procedural bar because our opinions have stated that the Teague retroactivity decision is to be made as a "threshold matter." E.g., Penry v. Lynaugh, 492 U.S. 302, 329 (1989); Caspari v. Bohlen, 510 U.S. 383, 389 (1994). That simply means, however, that the Teague issue should be addressed "before considering the merits of [a] claim." 510 U.S., at 389. It does not mean that the Teague inquiry is antecedent to consideration of the general prerequisites for federal habeas corpus which are unrelated to the merits of the particular claim — such as the requirement that the petitioner be "in custody," see 28 U.S.C. § 2254(a), or that the state-court judgment not be based on an independent and adequate state ground. Constitutional issues are generally to be avoided, and as even a cursory review of this Court's new-rule cases reveals (including our discussion in Part IV, infra) the Teague inquiry requires a detailed analysis of federal constitutional law. See, e.g., Sawyer v. Smith, 497 U.S. 227, 233-241 (1990); Penry, supra, at 316-319; Gilmore v. Taylor, 508 U.S. 333, 339-344 (1993); Saffle v. Parks, 494 U.S. 484, 488-494 (1990).

We are somewhat puzzled that the Eleventh Circuit, after having held proceedings in abeyance while petitioner brought his claim in state court, did not so much as mention the Florida Supreme Court's determination that Lambrix's Espinosa claim was procedurally barred. The State of Florida raised that point before both the District Court and the Court of Appeals, going so far as to reiterate it in a postjudgment Motion for Clarification and/or Modification of Opinion before the Court of Appeals, reprinted at App. 176. A State's procedural rules are of vital importance to the orderly administration of its criminal courts; when a federal court permits them to be readily evaded, it undermines the criminal justice system. We do not mean to suggest that the procedural-bar issue must invariably be resolved first; only that it ordinarily should be. Judicial economy might counsel giving the Teague question priority, for example, if it were easily resolvable against the habeas petitioner, whereas the procedural-bar issue involved complicated issues of state law. Cf. 28 U.S.C. § 2254(b)(2) (permitting a federal court to deny a habeas petition on the merits notwithstanding the applicant's failure to exhaust state remedies).

Despite our puzzlement at the Court of Appeals' failure to resolve this case on the basis of procedural bar, we hesitate to resolve it on that basis ourselves. Lambrix asserts several reasons why his claim is not procedurally barred, which seem to us insubstantial but may not be so; as we have repeatedly recognized, the courts of appeals and district courts are more familiar than we with the procedural practices of the States in which they regularly sit, see, e.g., Rummel v. Estelle , 445 U.S. 263, 267, n. 7 (1980); County Court of Ulster Cty. v. Allen , 442 U.S. 140, 153-154 (1979). Rather than prolong this litigation by a remand, we proceed to decide the case on the Teague grounds that the Court of Appeals used.

III

Florida employs a three-stage sentencing procedure. First, the jury weighs statutorily specified aggravating circumstances against any mitigating circumstances, and renders an "advisory sentence" of either life imprisonment or death. Fla. Stat. § 921.141(2). Second, the trial court weighs the aggravating and mitigating circumstances, and enters a sentence of life imprisonment or death; if the latter, its findings must be set forth in writing. § 921.141(3). The jury's advisory sentence is entitled to "great weight" in the trial court's determination, Tedder v. State, 322 So.2d 908, 910 (Fla. 1975), but the court has an independent obligation to determine the appropriate punishment, Ross v. State, 386 So.2d 1191, 1197 (Fla. 1980). Third, the Florida Supreme Court automatically reviews all cases in which the defendant is sentenced to death. § 921.141(4).

Lambrix's jury, which was instructed on five aggravating circumstances, recommended that he be sentenced to death for each murder. The trial court found five aggravating circumstances as to Moore's murder and four as to Bryant's, including that each murder was "especially heinous and atrocious"; it found no mitigating circumstances as to either murder; it concluded that the aggravating circumstances outweighed the mitigating, and sentenced Lambrix to death on each count. App. 20-21. Although Lambrix failed to raise any claims concerning the sentencing procedure on direct appeal, the Florida Supreme Court agreed with the trial court's findings as to the aggravating circumstances. Lambrix v. State, 494 So.2d, at 1148.

Lambrix contends that the jury's consideration of the HAC aggravator violated the Eighth Amendment because the jury instructions concerning this circumstance failed to provide sufficient guidance to limit the jury's discretion. Like the Eleventh Circuit, see 72 F.3d, at 1503, we assume, arguendo, that this was so. Lambrix further contends (and this is at the heart of the present case) that the trial court's independent weighing did not cure this error. Prior to our opinion in Espinosa v. Florida, 505 U.S. 1079 (1992), the State had contended that Lambrix was not entitled to relief because the sentencing judge properly found and weighed a narrowed HAC aggravator. In Espinosa, however, we established the principle that if a "weighing" State requires the sentencing trial judge to give deference to a jury's advisory recommendation, neither the judge nor the jury is constitutionally permitted to weigh invalid aggravating circumstances. Lambrix seeks the benefit of that principle; the State contends that it constitutes a new rule under Teague and thus cannot be relied on in a federal habeas corpus proceeding.

Lambrix also contends that the trial court itself failed to apply a properly narrowed HAC aggravator. We decline to consider this contention because it is not fairly within the question presented, which asked only whether Teague v. Lane, 489 U.S. 288 (1989), bars relief based upon Espinosa v. Florida, 505 U.S. 1079 (1992), Pet. for Cert. i. See this Court's Rule 14.1(a).

In Teague we held that, in general, "new constitutional rules of criminal procedure will not be applicable to those cases which have become final before the new rules are announced." 489 U.S., at 310-311. To apply Teague, a federal court engages in a three-step process. First, it determines the date upon which the defendant's conviction became final. See Caspari v. Bohlen, 510 U.S. 383, 390 (1994). Second, it must "`[s]urve[y] the legal landscape as it then existed,' Graham v. Collins, [ 506 U.S. 461, 468 (1993)], and `determine whether a state court considering [the defendant's] claim at the time his conviction became final would have felt compelled by existing precedent to conclude that the rule [he] seeks was required by the Constitution,' Saffle v. Parks, 494 U.S. 484, 488 (1990)." Ibid. Finally, if the court determines that the habeas petitioner seeks the benefit of a new rule, the court must consider whether the relief sought falls within one of the two narrow exceptions to nonretroactivity. See Gilmore v. Taylor, 508 U.S. 333, 345 (1993).

IV

Lambrix's conviction became final on November 24, 1986, when his time for filing a petition for certiorari expired. Thus, our first and principal task is to survey the legal landscape as of that date, to determine whether the rule later announced in Espinosa was dictated by then-existing precedent — whether, that is, the unlawfulness of Lambrix's conviction was apparent to all reasonable jurists. See, e.g., Graham, 506 U.S., at 477; Butler v. McKellar, 494 U.S. 407, 415 (1990); id., at 417-418 (Brennan, J., dissenting).

In Espinosa, we determined that the Florida capital jury is, in an important respect, a co-sentencer with the judge. As we explained: "Florida has essentially split the weighing process in two. Initially, the jury weighs aggravating and mitigating circumstances, and the result of that weighing process is then in turn weighed within the trial court's process of weighing aggravating and mitigating circumstances." 505 U.S., at 1082. We then concluded that the jury's consideration of a vague aggravator tainted the trial court's sentence because the trial court gave deference to the jury verdict (and thus indirectly weighed the vague aggravator) in the course of weighing the aggravating and mitigating circumstances. Ibid. We reasoned that this indirect weighing created the same risk of arbitrariness as the direct weighing of an invalid aggravating factor. Ibid.

Our description of the holding of Espinosa in the preceding paragraph of text is so clear that we are at a loss to explain the dissent's impression that we accord Espinosa the "novel interpretation" that "the constitutional error in the jury instruction will `automatically render a defendant's sentence unconstitutional.'" Post, at 541 (dissenting opinion) (quoting infra, at 530). The sentence from which the phrase quoted by the dissent is wrenched (so violently that the word "not" which precedes it is omitted) is not discussing the holding of Espinosa; indeed, it does not even mention Espinosa; nor does the entire paragraph or the previous or subsequent paragraphs.

In our view, Espinosa was not dictated by precedent, but announced a new rule which cannot be used as the basis for federal habeas corpus relief. It is significant that Espinosa itself did not purport to rely upon any controlling precedent. The opinion cited only a single case, Baldwin v. Alabama, 472 U.S. 372, 382 (1985), in support of its central conclusion that indirect weighing of an invalid aggravator "creates the same potential for arbitrariness" as direct weighing of an invalid aggravator. Espinosa, 505 U.S., at 1082. And it introduced that lone citation with a "cf." — an introductory signal which shows authority that supports the point in dictum or by analogy, not one that "controls" or "dictates" the result.

The dissent maintains that this statement is proven wrong by Espinosa's citation of Godfrey v. Georgia , 446 U.S. 420 (1980), and Tedder v. State, 322 So.2d 908 (Fla. 1975). Post, at 541, n. 2. This is wordplay. While those two cases can be called "controlling authority" in the sense that the two propositions they established (that an instruction to the sentencing jury which fails to define the HAC aggravator violates the Eighth Amendment, and that the Florida sentencing judge must give great weight to the jury's recommendation) were among the "givens" from which any decision in Espinosa had to be derived, they assuredly were not "controlling authority" in the sense we obviously intend: that they compel the outcome in Espinosa. They do not answer the definitive question: whether the jury's advisory verdict taints the trial court's sentence, that is, whether indirect weighing of an invalid factor creates the same potential for arbitrariness as direct weighing.

Baldwin itself contains further evidence that Espinosa set forth a new rule. Baldwin considered the constitutionality of Alabama's death-sentencing scheme, in which the jury was required to "fix the punishment at death" if it found the defendant guilty of an aggravated offense, whereupon the trial court would conduct a sentencing hearing at which it would determine a sentence of death or of life imprisonment. 472 U.S., at 376. The defendant contended that because the jury's mandatory sentence would have been unconstitutional standing alone, see Woodson v. North Carolina, 428 U.S. 280, 288-305 (1976) (plurality opinion), it was impermissible for the trial court to consider that verdict in determining its own sentence. We did not reach that contention because we concluded that under Alabama law the jury's verdict formed no part of the trial judge's sentencing calculus. Id., at 382. We noted, however, on the page of the opinion that Espinosa cited, that the defendant's "argument conceivably might have merit if the judge actually were required to consider the jury's `sentence' as a recommendation as to the sentence the jury believed would be appropriate, cf. Proffitt v. Florida, 428 U.S. 242 (1976), and if the judge were obligated to accord some deference to it." Baldwin, 472 U.S., at 382 (emphasis added); see also id., at 386, n. 8 ("express[ing] no view" on the same point). This highly tentative expression, far from showing that Baldwin "dictate[s]" the result in Espinosa, see Sawyer v. Smith, 497 U.S. 227, 235 (1990), suggests just the opposite. Indeed, in Baldwin the Chief Justice, who believed that Alabama's scheme did contemplate that the trial judge would consider the jury's "sentence," nonetheless held the scheme constitutional. 472 U.S., at 392 (opinion concurring in judgment).

The Supreme Court decisions relied upon most heavily by petitioner are Godfrey v. Georgia , 446 U.S. 420 (1980); Maynard v. Cartwright, 486 U.S. 356 (1988); and Clemons v. Mississippi, 494 U.S. 738 (1990). In Godfrey, we held that Georgia's "outrageously or wantonly vile, horrible and inhuman" aggravator was impermissibly vague, reasoning that there was nothing in the words "outrageously or wantonly vile, horrible and inhuman," "that implies any inherent restraint on the arbitrary and capricious infliction of the death sentence," and concluded that these terms alone "gave the jury no guidance." 446 U.S., at 428-429 (plurality opinion). Similarly, in Maynard v. Cartwright, applied retroactively to February 1985 in Stringer v. Black, 503 U.S. 222 (1992), we held that Oklahoma's HAC aggravator, which is identically worded to Florida's HAC aggravator, was impermissibly vague because the statute gave no more guidance than the vague aggravator at issue in Godfrey and the sentencing jury was not given a limiting instruction. 486 U.S., at 363-364.

Although Godfrey and Maynard support the proposition that vague aggravators must be sufficiently narrowed to avoid arbitrary imposition of the death penalty, these cases, and others, demonstrate that the failure to instruct the sentencing jury properly with respect to the aggravator does not automatically render a defendant's sentence unconstitutional. We have repeatedly indicated that a sentencing jury's consideration of a vague aggravator can be cured by appellate review. Thus, in Godfrey itself, we were less concerned about the failure to instruct the jury properly than we were about the Georgia Supreme Court's failure to narrow the facially vague aggravator on appeal. Had the Georgia Supreme Court applied a narrowing construction of the aggravator, we would have rejected the Eighth Amendment challenge to Godfrey's death sentence, notwithstanding the failure to instruct the jury on that narrowing construction. Godfrey, supra, at 431-432. Likewise in Maynard, we stressed that the vague HAC aggravator had not been sufficiently limited on appeal by the Oklahoma Court of Criminal Appeals "to cure the unfettered discretion of the jury." 486 U.S., at 364.

We reached a similar conclusion in Clemons v. Mississippi, applied retroactively to February 1985 in Stringer. Clemons considered the question whether the sentencer's weighing of a vague HAC aggravator rendered that sentence unconstitutional in a "weighing" State. The sentencing jury in Clemons, as in Maynard, was given a HAC instruction that was unconstitutionally vague. We held that "the Federal Constitution does not prevent a state appellate court from upholding a death sentence that is based in part on an invalid or improperly defined aggravating circumstance either by reweighing of the aggravating and mitigating evidence or by harmless-error review." Clemons, 494 U.S., at 741, 745; see also Stringer, 503 U.S., at 230.

The principles of the above-described cases do not dictate the result we ultimately reached in Espinosa. Florida, unlike Oklahoma, see Maynard, supra, at 360, had given its facially vague HAC aggravator a limiting construction sufficient to satisfy the Constitution. See Proffitt v. Florida , 428 U.S. 242, 255-256 (1976) (joint opinion of Stewart, Powell, and STEVENS, JJ.); id., at 260 (White, J., concurring in judgment). Thus, unlike the sentencing juries in Clemons, Maynard, and Godfrey, who were not instructed with a properly limited aggravator, the sentencing trial judge in Espinosa did find the HAC aggravator under a properly limited construction. See Espinosa, 505 U.S., at 1082, citing Walton v. Arizona, 497 U.S. 639, 653 (1990). A close examination of the Florida death-penalty scheme persuades us that a reasonable jurist considering Lambrix's sentence in 1986 could have reached a conclusion different from the one Espinosa announced in 1992. There were at least three different, but somewhat related, approaches that would have suggested a different outcome:

The dissent says that "[g]iven that the judge's instruction to the jury failed to narrow the HAC aggravator, there is no reason to believe that [the trial judge] appropriately narrowed the [HAC] factor in his . . . deliberations." Post, at 545. Our cases establish that there is always a "reason to believe" that, which we consider fully adequate: "Trial judges are presumed to know the law and to apply it in making their decisions. If the [state] Supreme Court has narrowed the definition of the [HAC] aggravating circumstance, we presume that [state] trial judges are applying the narrower definition." Walton v. Arizona, 497 U.S. 639, 653 (1990). Without abandoning our precedent, the most the dissent can argue is that the ordinary presumption is overcome by failure to instruct. The factual support for such an argument is questionable: judges fail to instruct juries about rules of law they are aware of all the time. Moreover, if the argument were correct, the holding in Espinosa itself would have been unnecessary: we could have simply said there (as the dissent would have us say here) that the failure to instruct on the narrowing construction displayed the judge's ignorance of the narrowing construction. Instead, of course, Espinosa cited the passage from Walton quoted above. Espinosa, 505 U.S., at 1082.

(1) The mere cabining of the trial court's discretion would avoid arbitrary imposition of the death penalty, and thus avoid unconstitutionality. In Proffitt v. Florida , supra, we upheld the Florida death-penalty scheme against the contention that it resulted in arbitrary imposition of the death penalty, see Gregg v. Georgia , 428 U.S. 153, 188 (1976), because "trial judges are given specific and detailed guidance to assist them in deciding whether to impose a death penalty or imprisonment for life" and because the Florida Supreme Court reviewed sentences for consistency. Proffitt, 428 U.S., at 253 (joint opinion of Stewart, Powell, and STEVENS, JJ.); id., at 260-261 (opinion of White, J., joined by The Chief Justice and REHNQUIST, J.). (In Proffitt itself, incidentally, the jury had not been instructed on an appropriately narrowed HAC aggravator, see Proffitt v. Wainwright, 685 F.2d 1227, 1264, n. 57 (CA11 1982), cert. denied, 464 U.S. 1002 (1983).) From what was said in Proffitt it would, as the en banc Eleventh Circuit noted, "sensibly follow that the judge's proper review of the sentence cures any risk of arbitrariness occasioned by the jury's consideration of an unconstitutionally vague aggravating circumstance." Glock v. Singletary, 65 F.3d 878, 886 (1995), cert. denied, 519 U.S. 888 (1996). It could have been argued, of course, as the dissent contends, see post, at 543 (dissenting opinion), that prior constitutional error by a sentencing-determining jury would make a difference, but both the conclusion and the premise of that argument were debatable: not only whether it would make a difference, but even (as the succeeding point demonstrates) whether there was any constitutional error by a sentencing-determining jury.

(2) There was no error for the trial judge to cure, since under Florida law the trial court, not the jury, was the sentencer. In Espinosa we concluded, in effect, that the jury was at least in part a co-sentencer along with the trial court. That determination can fairly be traced to our opinion in Sochor v. Florida, 504 U.S. 527 (1992), decided just three weeks earlier, where we explained that under Florida law the trial court "is at least a constituent part of `the sentencer,'" implying that the jury was that as well. Id., at 535-536. That characterization is in considerable tension with our pre-1986 view. In Proffitt, for example, after considering Tedder v. State, 322 So.2d 908 (Fla. 1975), on which Espinosa primarily relied, the Court determined that the trial court was the sentencer. E.g., 428 U.S., at 249 (joint opinion of Stewart, Powell, and STEVENS, JJ.) ("the actual sentence is determined by the trial judge" (emphasis added)); id., at 251 (the trial court is "[t]he sentencing authority in Florida"); id., at 252 ("the sentence is determined by the judge rather than by the jury"); id., at 260 (White, J., concurring in judgment). We even distinguished the Florida scheme from the Georgia scheme on the grounds that "in Florida the sentence is determined by the trial judge rather than by the jury." Id., at 252 (joint opinion) (emphasis added). Some eight years later, just two years before petitioner's conviction became final, we continued to describe the judge as the sentencer. See Spaziano v. Florida, 468 U.S. 447 (1984); see also Barclay v. Florida, 463 U.S. 939, 952-954 (plurality opinion); id., at 962 (STEVENS, J., concurring in judgment). (Although he now believes the jury is a co-sentencer, at the time Lambrix's conviction became final JUSTICE STEVENS had explained that "the sentencing authority [is] the jury in Georgia, the judge in Florida." Ibid.) It would not have been unreasonable to rely on what we had said in Proffitt, Spaziano, and Barclay — that the trial court was the sentencer — and to conclude that where the sentencer considered properly narrowed aggravators there was simply no error under Godfrey or Maynard. The Florida Supreme Court and the Eleventh Circuit held precisely that in 1989, see Smalley v. State, 546 So.2d 720, 722; Bertolotti v. Dugger, 883 F.2d 1503, 1526-1527, cert. denied, 497 U.S. 1032 (1990); and in 1985 the Eleventh Circuit foresaw the possibility of such a holding: "[ Spaziano's] reasoning calls into question whether any given error in such a merely `advisory' proceeding should be considered to be of constitutional magnitude." Proffitt v. Wainwright, 756 F.2d 1500, 1502.

(3) The trial court's weighing of properly narrowed aggravators and mitigators was sufficiently independent of the jury to cure any error in the jury's consideration of a vague aggravator. Although the Florida Supreme Court had interpreted its statute — which provided that the judge was the sentencer, Fla. Stat. § 921.141(3) (Supp. 1992), and that the jury rendered merely an "advisory sentence," § 921.141(2) — as requiring the trial judge to give "great weight" to a jury's advisory recommendation, Tedder v. State, supra, that court nonetheless emphasized that the trial court must "independently weigh the evidence in aggravation and mitigation," and that "[u]nder no combination of circumstances can th[e] [jury's] recommendation usurp the judge's role by limiting his discretion." Eutzy v. State, 458 So.2d 755, 759 (Fla. 1984), cert. denied, 471 U.S. 1045 (1985). In one case, the Florida Supreme Court vacated a sentence because the trial court had given "undue weight to the jury's recommendation of death and did not make an independent judgment of whether or not the death penalty should be imposed." Ross v. State, 386 So.2d 1191, 1197 (1980) (emphasis added). In Spaziano v. Florida, supra, we acknowledged that the Florida trial court conducts "its own weighing of the aggravating and mitigating circumstances," id., at 451, and that "[r]egardless of the jury's recommendation, the trial judge is required to conduct an independent review of the evidence and to make his own findings regarding aggravating and mitigating circumstances," id., at 466 (emphasis added); see also Proffitt, 428 U.S., at 251. Given these precedents, it was reasonable to think that the trial court's review would at least constitute the sort of "reweighing" that would satisfy Clemons v. Mississippi, 494 U.S. 738 (1990), see also Stringer, 503 U.S., at 237. In fact, given the view of some Members of this Court that appellate reweighing was inconsistent with the Eighth Amendment, see, e.g., Cabana v. Bullock, 474 U.S. 376, 400-401, 404 (1986) (Blackmun, J., dissenting, joined by Brennan and Marshall, JJ.); Clemons, supra, at 769-772 (Blackmun, J., joined by Brennan, Marshall, and STEVENS, JJ., concurring in part and dissenting in part), it would have been reasonable to think that trial-court reweighing was preferable. As one Court of Appeals was prompted to note, " Clemons's holding, which arguably points in the opposite direction from Espinosa, indicates that even in 1990 Espinosa's result would not have been dictated by precedent." Glock v. Singletary, 65 F.3d, at 887 (en banc).

The dissent accuses us of "simply ignoring the reasoning in Tedder." Post, at 543 (dissenting opinion). We have of course not done so. See supra, at 526, 533-534 and this page. The dissent, however, fails to discuss, or indeed even mention, the cases interpreting Tedder that contradict the dissent's view — cases in both this Court and the Florida Supreme Court repeatedly emphasizing the trial judge's obligation to make an independent assessment and weighing of the aggravating and mitigating circumstances. The dissent relies, for example, upon the Florida Supreme Court's decision in Riley v. Wainwright, 517 So.2d 656 (1987), see post, at 541, n. 3 (a decision rendered after Lambrix's conviction became final and hence not technically relevant). But subsequent to that case the Florida Supreme Court summarized its jurisprudence as follows: "Our case law contains many instances where a trial judge's override of a jury recommendation of life has been upheld. Notwithstanding the jury recommendation, whether it be for life imprisonment or death, the judge is required to make an independent determination, based on the aggravating and mitigating factors. Moreover, this procedure has been previously upheld against constitutional challenge." Grossman v. State, 525 So.2d 833, 840 (Fla. 1988) (emphasis added; citations omitted). "It is clear . . . that the prosecutor correctly stated the law in Florida: the judge is the sentencing authority and the jury's role is merely advisory." Id., at 839. It is not our burden, of course, to establish that these statements in Grossman, or in the other cases we rely upon, were accurate; as we later determined, they were wrong and the dissent's (current) reading of Tedder is correct. But the question before us is whether a reasonable jurist could have disagreed with the dissent's interpretation of Tedder at the time of Lambrix's conviction. In treating as relevant to that question only that portion of precedent vindicated by later decisions, the dissent "endues the jurist with prescience, not reasonableness." Stringer v. Black, 503 U.S. 222, 244 (1992) (SOUTER, J., dissenting).

That Espinosa announced a new rule is strongly confirmed by our decision in Walton v. Arizona, 497 U.S. 639 (1990). Although decided after petitioner's conviction became final, Walton is a particularly good proxy for what a reasonable jurist would have thought in 1986, given that the only relevant cases decided by this Court in the interim were Maynard and Clemons, the holdings of both of which, we later held, were compelled by the law in 1985, see Stringer, supra. In Walton, we rejected a claim that Arizona's HAC aggravator failed sufficiently to channel the sentencer's discretion. Summarizing Godfrey and Maynard, we explained that "in neither case did the state appellate court, in reviewing the propriety of the death sentence, purport to affirm the death sentence by applying a limiting definition," and this, we said, "w[as] crucial to the conclusion we reached in Maynard." Walton, supra, at 653. This reasoning suggests that even following Maynard, a weighing-state death sentence would satisfy the Eighth Amendment so long as the vague aggravator was narrowed at some point in the process. Additionally, in the course of our opinion, we characterized Clemons as follows:

"[E]ven if a trial judge fails to apply the narrowing construction or implies an improper construction, the Constitution does not necessarily require that a state appellate court vacate a death sentence based on that factor. Rather, as we held in Clemons v. Mississippi, 494 U.S. 738 (1990), a state appellate court may itself determine whether the evidence supports the existence of the aggravating circumstance as properly defined or the court may eliminate consideration of the factor altogether and determine whether any remaining aggravating circumstances are sufficient to warrant the death penalty." Walton, 497 U.S., at 653-654 (emphasis added).

Our use of the disjunctive suggests that as late as 1990, if a Florida trial court determined that the defendant's conduct fell within the narrowed HAC aggravator, the sentence would satisfy the Eighth Amendment irrespective of whether the trial court reweighed the aggravating and mitigating factors. The holdings in Stringer, Maynard, Clemons, and Godfrey cannot be thought to suggest otherwise, because there was no indication in those cases that the state courts had found the facts of the crimes to fall within appropriately narrowed definitions of the aggravators. Before Espinosa, we had never invalidated a death sentence where a court found the challenged aggravator to be within the appellate court's narrowed definition of a facially vague aggravator.

The dissent is thus simply wrong in stating that we have confused appellate application of a limiting construction with a trial court's deference to a tainted jury recommendation, see post, at 545 (dissenting opinion). Walton indicated that our precedents provided two distinct and permissible routes to satisfy the Eighth Amendment where the sentencer considered a vague aggravator: a court's finding of the aggravator under a proper limiting construction, or independent reweighing of the circumstances.

Most of the dissent is devoted to making a forceful case that Espinosa was a reasonable interpretation of prior law — perhaps even the most reasonable one. But the Teague inquiry — which is applied to Supreme Court decisions that are, one must hope, usually the most reasonable interpretation of prior law — requires more than that. It asks whether Espinosa was dictated by precedent — i.e., whether no other interpretation was reasonable. We think it plain from the above that a jurist considering all the relevant material (and not, like the dissent, considering only the material that favors the Espinosa result) could reasonably have reached a conclusion contrary to our holding in that case. Indeed, both before and after Lambrix's conviction became final, every court decision we are aware of did so. See, e.g., Smalley v. State, 546 So.2d, at 722; Proffitt v. Wainwright, 756 F.2d, at 1502; Bertolotti v. Dugger, 883 F.2d, at 1527; Sanchez-Velasco v. State, 570 So.2d 908, 916 (Fla. 1990), cert. denied, 500 U.S. 929 (1991).

It has been suggested that Espinosa was not a new rule because our decision was handed down as a per curiam without oral argument. See, e.g., Glock v. Singletary, 65 F.3d, at 896, n. 11 (en banc) (Tjoflat, C.J., dissenting). Whatever inference of established law a summary, per curiam disposition might normally carry is precluded by the peculiar circumstances surrounding the summary per curiam in Espinosa. Just three weeks prior to our issuance of Espinosa, we had decided a case that raised the identical issue, and in which that issue had been fully briefed and argued; we found ourselves without jurisdiction to decide the point, however, because the defendant had failed to preserve his objection in the state courts. See Sochor v. Florida, 504 U.S., at 533-534. It is obvious on the face of the matter that Espinosa was only in the most technical sense an "unargued" case: we used that case, which was pending on petition for certiorari when Sochor was decided, as the vehicle for resolving a fully argued point without consuming additional resources.

V

Since we have determined that Espinosa announced a new rule under Teague, there remains only the task of determining whether that new rule nonetheless falls within one of the two exceptions to our nonretroactivity doctrine. "The first exception permits the retroactive application of a new rule if the rule places a class of private conduct beyond the power of the State to proscribe, see Teague, 489 U.S., at 311, or addresses a `substantive categorical guarante[e] accorded by the Constitution,' such as a rule `prohibiting a certain category of punishment for a class of defendants because of their status or offense.'" Saffle v. Parks, 494 U.S., at 494 (quoting Penry v. Lynaugh, 492 U.S., at 329, 330). Plainly, this exception has no application to this case. Espinosa "neither decriminalize[s] a class of conduct nor prohibit[s] the imposition of capital punishment on a particular class of persons." 494 U.S., at 495.

The second exception is for "`watershed rules of criminal procedure' implicating the fundamental fairness and accuracy of the criminal proceeding." Ibid. (quoting Teague, supra, at 311). Lambrix does not contend that this exception applies to Espinosa errors, and our opinion in Sawyer v. Smith, 497 U.S. 227, 241-244 (1990), makes it quite clear that that is so.

* * *

For the reasons stated, the judgment of the Court of Appeals for the Eleventh Circuit is

Affirmed.

JUSTICE STEVENS, with whom JUSTICE GINSBURG and JUSTICE BREYER join, dissenting.

Two propositions of law supported our holding in Espinosa v. Florida, 505 U.S. 1079 (1992): First, in a capital sentencing proceeding in a State where the sentencer weighs aggravating and mitigating circumstances, the Eighth Amendment is violated by a jury instruction that fails to define the "especially heinous, atrocious, or cruel" (HAC) aggravating circumstance. Second, in a Florida sentencing proceeding the trial court must give "`great weight'" to the jury's recommendation, whether it be for life or death. Id., at 1082. For these reasons, we concluded in Espinosa that constitutional error that taints the jury's recommendation presumptively taints the judge's sentence as well. Ibid. The two propositions supporting the Espinosa holding were well established when that case was decided. The first proposition dates back to 1980 when we decided Godfrey v. Georgia, 446 U.S. 420, 428-429, and the second was announced by the Florida Supreme Court in 1975 in Tedder v. State, 322 So.2d 908, 910. Thus I agree with Chief Judge Tjoflat that our per curiam opinion in Espinosa amounted to "nothing more than an application of well-settled principles. . . . In declaring the Florida HAC instruction unconstitutional, the Court simply applied the law as announced initially in Godfrey and later reaffirmed in Maynard [v. Cartwright, 486 U.S. 356 (1988]). The Court's conclusion — that the invalid instruction may have tainted the jury's death penalty recommendation and the trial judge's sentence — merely acknowledged what the Supreme Court of Florida has been holding for years." Glock v. Singletary, 65 F.3d 878, 896 (CA11 1995) (Tjoflat, C.J., dissenting).

Godfrey, of course, held that Georgia's "outrageously or wantonly vile, horrible and inhuman" aggravating factor failed to adequately channel the jury's discretion. See Godfrey, 446 U.S., at 428-429. We found the "heinous, atrocious or cruel" aggravator unconstitutional in Maynard v. Cartwright, 486 U.S. 356, 359 (1988), and subsequently noted that application of Godfrey to the HAC instruction did not create a new rule. See Stringer v. Black, 503 U.S. 222, 228-229 (1992).

These two "controlling precedents," both of which were cited in the Espinosa opinion, provided sufficient support for its holding. Thus the Court is simply mistaken when it asserts that " Espinosa itself did not purport to rely upon any controlling precedent." Ante, at 528.

Tedder, of course, was not an isolated decision. In Riley v. Wainwright, 517 So.2d 656 (Fla. 1987), the State Supreme Court put the point succinctly: "If the jury's recommendation, upon which the judge must rely, results from an unconstitutional procedure, then the entire sentencing process necessarily is tainted by that procedure." Id., at 659. The Riley Court relied on a pre- Tedder decision stating that the advisory opinion of the jury "is an integral part of the death sentencing process." 517 So.2d, at 657 (citing Lamadline v. State, 303 So.2d 17, 20 (Fla. 1974)).

Today the Court reaches the conclusion that Espinosa announced a new rule by placing a novel interpretation on its holding. The majority apparently construes Espinosa as holding that the constitutional error in the jury instruction will "automatically render a defendant's sentence unconstitutional." Ante, at 530. The Court suggests that our holdings in Godfrey, Maynard, and Clemons v. Mississippi, 494 U.S. 738, 745 (1990) — that an appellate court could cure a sentencing jury's weighing of an invalid aggravator — might have led a reasonable jurist down a road different from the one the Court followed in Espinosa. But in holding that a trial judge's sentence may be infected by the jury's consideration of an invalid aggravating factor, Espinosa did not address the entirely separate question of whether the jury's error could be cured or considered harmless either at the trial or the appellate level. Indeed, in subsequent proceedings the Supreme Court of Florida did conclude that the error in Espinosa's case was harmless and upheld his sentence of death. See Espinosa v. State, 626 So.2d 165, 167 (1993) (ruling that Espinosa's HAC instruction claim was procedurally barred because he had challenged the HAC factor rather than the instruction itself and, alternatively, that any error in the instruction was harmless beyond a reasonable doubt), cert. denied, 511 U.S. 1152 (1994), and affirmed Espinosa's sentence. Our decision in Espinosa did not create a new rule forbidding trial courts from curing a jury's error, rather it held that "if a weighing State decides to place capital sentencing authority in two actors rather than one, neither actor must be permitted to weigh invalid aggravating circumstances." 505 U.S., at 1082. This holding is a logical consequence of applying Godfrey to Florida's sentencing scheme.

Responding to this dissent in n. 2, ante, at 528, the Court states that the clause I have quoted was not intended to describe the Court's understanding of the holding in Espinosa. If that be so, the relevance of this portion of the Court's opinion, including its reliance on Godfrey and Maynard, is opaque, at best.

The Court also relies heavily on a passage in our opinion in Walton v. Arizona, 497 U.S. 639 (1990), noting that a trial judge's failure to apply a narrowing construction to an invalid aggravator "`does not necessarily require that a state appellate court vacate a death sentence based on that factor.'" Ante, at 537.

In a sinuous, difficult to follow argument, the Court suggests that three hypothetical propositions of law somehow demonstrate that the narrow holding in Espinosa was not dictated by Godfrey and Tedder. First, the Court posits that a reasonable jurist might have believed that "[t]he mere cabining of the trial court's discretion" was alone enough to avoid constitutional error. Ante, at 532 (emphasis deleted). A critical part of that "cabining," however, is Florida's requirement that a properly instructed jury must have an opportunity to recommend a life sentence, and that the judge must give great weight to that recommendation. The role of the jury is to provide one of the cabin's four walls. The fact that three walls remain standing hardly excuses an error that removed the wall represented by the jury's recommendation. At the time of petitioner's sentencing, the Florida Supreme Court recognized the jury's critical role, and, when error occurred before the jury, did not hesitate to remand for resentencing, even when the trial judge claimed to be unaffected by the error. For example in Messer v. State, 330 So.2d 137, 142 (1976), the State Supreme Court remanded for resentencing when the trial court failed to allow the jury to consider certain mitigating evidence. The court rejected the argument that the trial court's subsequent weighing of the mitigating evidence cured the error: the Florida scheme, the court concluded, was one of "checks and balances in which the input of the jury serves as an integral part." Ibid. Our holding in Proffitt v. Florida, 428 U.S. 242, 255 (1976) (joint opinion), that Florida's sentencing scheme is not facially unconstitutional does not suggest otherwise. There, we determined that the State's sentencing procedure provided adequate safeguards against arbitrary imposition of the death sentence in part because of the procedures followed by the trial judge in fixing the sentence. Our focus was on the adequacy of the guidance provided by the sentencing scheme; accordingly, we had no need to extensively examine or discuss the judge's relationship to the jury or Florida Supreme Court decisions like Tedder.

Second, simply ignoring the reasoning in Tedder, the Court suggests that there was "no error for the trial judge to cure, since under Florida law the trial court, not the jury, was the sentencer." Ante, at 533. It is, of course, true that the judge imposes the sentence after receiving the jury's recommendation. But this has never meant that constitutional error in the proceedings before the jury is simply irrelevant. Cf. Messer v. State, supra. As then-JUSTICE REHNQUIST noted in 1983, it is well-settled Florida law that if the jury makes a recommendation of life imprisonment, "the trial judge may not impose a death sentence unless `the facts suggesting a sentence of death [are] so clear and convincing that virtually no reasonable person could differ.' Tedder v. State, 322 So.2d 908, 910 (1975)." Barclay v. Florida, 463 U.S. 939, 955-956 (1983) (plurality opinion). Similarly, a trial judge should not disturb a jury recommendation of death "unless there appear strong reasons to believe that reasonable persons could not agree with the recommendation." See LeDuc v. State, 365 So.2d 149, 151 (Fla. 1978), cert. denied, 444 U.S. 885 (1979). Given this, it is vacuous to argue that our prior references to the judge as the sentencer somehow imply that an error before the jury would not affect the ultimate sentence. It is equally vacuous to suggest that our conclusion in Espinosa "that the jury was at least in part a co-sentencer" had its source in a case decided "just three weeks earlier," ante, at 533 (citing Sochor v. Florida, 504 U.S. 527 (1992)). In that earlier case, we cited Tedder after explaining that the jury was a constituent element of the sentencer "because the trial judge does not render wholly independent judgment, but must accord deference to the jury's recommendation." Sochor, 504 U.S., at 533.

The Florida Supreme Court has applied Tedder in numerous cases to reverse a trial judge's override of a jury's life sentence. See, e.g., Wasko v. State, 505 So.2d 1314, 1318 (1987); Goodwin v. State, 405 So.2d 170, 172 (1981); Odom v. State, 403 So.2d 936, 942-943 (1981), cert. denied, 456 U.S. 925 (1982); Neary v. State, 384 So.2d 881, 885-886 (1980); Malloy v. State, 382 So.2d 1190, 1193 (1979).

Third, the Court suggests that the trial court's "weighing of properly narrowed aggravators and migrators was sufficiently independent of the jury to cure any error in the jury's consideration of a vague aggravator." Ante, at 534 (emphasis deleted). This suggestion is doubly flawed. Given that the judge's instruction to the jury failed to narrow the HAC aggravator, there is no reason to believe that he appropriately narrowed the factor in his own deliberations. More importantly, even if he did apply a limiting definition, his sentencing decision was made without the benefit of an untainted recommendation from the jury, and, under Florida law, he could not have simply resentenced the petitioner without regard to the jury's tainted recommendation. Nor can one simply conclude that this error made no practical difference in petitioner's sentence. There is nothing in the record to suggest that had the jury recommended a life sentence, the judge would have found that "the facts suggesting a sentence of death were so clear and convincing that virtually no reasonable person could differ," as Tedder requires.

Nothing in the record indicates that the judge recognized that the jury instruction was erroneous, or that he sought to cure that error in his own weighing process. In finding that the HAC aggravator was present, the judge merely stated "[t]he facts speak for themselves." App. 20.

Here, again, the Court finds that our statements in cases like Walton v. Arizona, 497 U.S. 639 (1990), that a state appellate court may affirm a death sentence resulting from a unconstitutionally broad aggravator by applying a limiting definition, suggest that Espinosa is a new rule. The majority's analysis confuses an appellate court's application of a limiting definition on appellate review with a trial judge's deference to a tainted jury recommendation. The judge in this case did not indicate that he was applying a limiting definition of the HAC factor, or that he was in some other way curing or discounting the error in the jury instruction. At the time of petitioner's sentencing, given Godfrey and Tedder, this rendered petitioner's death sentence constitutionally defective.

As a matter of logic and law there was nothing new about Espinosa's holding that the jury plays a central role in Florida's capital sentencing scheme. Moreover, as statistics that I have previously summarized demonstrate, it was equally clear as a matter of fact that "erroneous instructions to the jury at the sentencing phase of the trial may make the difference between life or death." Sochor v. Florida,, 504 U.S., at 551.

"As a matter of fact, the jury sentence is the sentence that is usually imposed by the Florida Supreme Court. The State has attached an appendix to its brief, see App. to Brief for Respondent A1-A70, setting forth data concerning 469 capital cases that were reviewed by the Florida Supreme Court between 1980 and 1991. In 341 of those cases (73%), the jury recommended the death penalty; in none of those cases did the trial judge impose a lesser sentence. In 91 cases (19%), the jury recommended a life sentence; in all but one of those cases, the trial judge overrode the jury's recommended life sentence and imposed a death sentence. In 69 of those overrides (77%), however, the Florida Supreme Court vacated the trial judge's sentence and either imposed a life sentence itself or remanded for a new sentencing hearing.
"Two conclusions are evident. First, when the jury recommends a death sentence, the trial judge will almost certainly impose that sentence. Second, when the jury recommends a life sentence, although overrides have been sustained occasionally, the Florida Supreme Court will normally uphold the jury rather than the judge. It is therefore clear that in practice, erroneous instructions to the jury at the sentencing phase of the trial may make the difference between life or death." Sochor v. Florida, 504 U.S., at 551-552 (footnotes omitted).

I respectfully dissent.

JUSTICE O'CONNOR, dissenting.

Although I agree with much of the reasoning set forth in Part II of the Court's opinion, I disagree with its disposition of the case. I would instead vacate the judgment of the Court of Appeals and remand the case so that the Court of Appeals might consider the procedural bar issue in the first instance.

The Court holds that, as a general practice, a federal habeas court should consider whether the relief a habeas petitioner requests is a "new rule" under Teague v. Lane, 489 U.S. 288 (1989), only after resolving the State's argument that his claim is procedurally barred. Ante, at 525. Usually, then, when a federal habeas court has before it contentions that a petitioner's claim is barred both on state procedural grounds and because the petitioner seeks to rely on a "new rule" under Teague, the court should consider the Teague question only after the procedural bar issue has been resolved in the petitioner's favor. As the Court recognizes, addressing the procedural bar issue first avoids unnecessary consideration of constitutional questions and accords fitting respect to the State's procedural rules, which are indispensable to the administration of its criminal justice system. Ante, at 524-525.

With this much of the Court's opinion I agree. Of course, there may be exceptions to the rule that the procedural bar issue should be resolved first. One case might be where the procedural bar question is excessively complicated, but the Teague issue can be easily resolved. The Court of Appeals here gave no reason for its failure to consider the Florida Supreme Court's determination that petitioner's claim based on Espinosa v. Florida, 505 U.S. 1079 (1992), was procedurally barred. Indeed, the Court of Appeals did not even discuss the state court's holding, let alone decide that resolution of the procedural bar issue would be inappropriate in this case. I see no reason to think resolution of the procedural bar question would be especially troublesome, nor do I see any other reason for the Court of Appeals' failure to give priority to the State's argument that an independent and adequate state ground barred petitioner's Espinosa claim.

Accordingly, I would remand the case to the Court of Appeals for it to resolve the procedural bar issue. As the Court points out, the Court of Appeals is better suited to evaluating matters of state procedure than are we. Ante, at 525. In my view, then, it is premature to address the State's contention that petitioner's Espinosa claim is barred on Teague grounds. Nevertheless, since the Court reaches the question, I wish to express my agreement with Justice Stevens' resolution of the Teague issue.