 |
|||||
|
|
|
|
|
||
|
|
|||||
|
|
|
|
|
||
The leading clause of 28 U. S. C. §2254(d) establishes a general rule of claim preclusion. Unless one of the exceptions is met, a claim adjudicated on the merits in state court is not a ground for federal habeas relief. See Scheidegger, Habeas Corpus, Relitigation, and the Legislative Power, 98 Colum. L. Rev. 888, 946 (1998); 141 Cong. Rec. 15,058, col. 1-2 (1995) (statement of Sen. Biden). The exception in paragraph (2) of this subsection, relating to factual findings, is not presented in this case. The exception in paragraph (1) has two clauses, which have independent meaning. See Williams v. Taylor, 529 U. S. 362, 404-405 (2000). We address the first clause in this part and the second clause in part III, infra.
Under the statute, relief may be granted despite the prior adjudication if the state court decision "was contrary to . . . clearly established federal law, as determined by the Supreme Court of the United States . . . ." Williams endorsed the Fourth Circuit's interpretation of this clause in Green v. French, 143 F. 3d 865 (1998).
"First, a state-court decision is contrary to this Court's precedent if the state court arrives at a conclusion opposite to that reached by this Court on a question of law. Second, a state-court decision is also contrary to this Court's precedent if the state court confronts facts that are materially indistinguishable from a relevant Supreme Court precedent and arrives at a result opposite to ours." Williams, supra, 529 U. S., at 405.
The first part of this test refers to the state court's selection of the governing rule, as opposed to its application of the rule to the facts of the case. For example, a decision that correctly identified Strickland v. Washington, 466 U. S. 668 (1984) as the rule for ineffective assistance claims is not "contrary to" controlling law, regardless of how much a federal court might disagree with the state court's application of the rule. The latter question falls under the second clause. See Williams, supra, 529 U. S., at 406.
Petitioner apparently does not claim that the state court erred in rule selection. In his only mention of the governing statute, he invokes the "materially indistinguishable" and "unreasonable application" tests, but not the "contradicts the governing law" test. See Brief for Petitioner 12.
This is a wise concession, because the state court clearly did recognize and apply the correct controlling precedents. The defendant contended on appeal that the jury instructions on retrial failed to implement the requirement of Penry v. Lynaugh, 492 U. S. 302 (1989) (Penry I). Since the additional instructions were clearly intended to do just that, see Brief for Petitioner 9, the question is whether they were susceptible to misinterpretation by the jury to a degree sufficient to warrant reversal. The controlling case for that question is Boyde v. California, 494 U. S. 370 (1990). See Brief for Petitioner 12-13.
The state court recognized and applied the correct rules. It relied on its prior opinion in Coble v. State, 871 S. W. 2d 192, 207 (Tex. Crim. App. 1993). See J. A. 782. Coble relied on Fuller v. State, 829 S. W. 2d 191, 209 (Tex. Crim. App. 1992). Fuller, in turn, correctly recognized Penry as the governing rule for the substantive question and Boyde as the governing rule for the claimed ambiguity. See ibid.
Defendant contends that the present case is "materially indistinguishable" from Penry I, Brief for Petitioner 12, yet the argument in support of this contention is just a straight "merits" argument of the kind that would have been appropriate under the now-repealed regime of de novo review. See id., at 12-24. Accepting this argument would strip the term "materially indistinguishable" of any meaning. Far worse, it would strip the statute of any meaning and return to the status quo ante. It would be precisely the step the Court rejected in Williams. See 529 U. S., at 404.
The facts of the present case are more than distinguishable from Penry I, and the differences are more than material. The jury in Penry I did not receive "any jury instruction on mitigating evidence." 492 U. S., at 322 (emphasis added). The jury on retrial did. J. A. 675. One can argue until the cows come home whether the instruction was good enough, but to claim that a case of an arguably ambiguous instruction is "materially indistinguishable" from a case with no instruction at all borders on frivolous.
The two cases are also distinguishable in the arguments of counsel. These arguments are part of the context of the proceedings, and Boyde tells us that context is material. See 494 U. S., at 383. In the first trial, the defense asked the jury to answer "no" to a special issue, even if the literal answer was "yes," in response to mitigating evidence. Penry I, supra, 492 U. S., at 325. The prosecutor argued in rebuttal that they could not, and the Penry I Court thought that was significant. See id., at 325-326. In the present case, the prosecutor made the kind of argument upheld as proper in Boyde, i.e., that the proffered factors were either not mitigating or entitled to little weight. Compare Boyde, 494 U. S., at 385-386, with J. A. 665-666. To say that these are material distinctions would be an understatement. "Material differences exist between this case and [Penry I], and the [Texas Court of Criminal Appeals'] decision is not contrary to the rule [Penry I] announced." Ramdass v. Angelone, 530 U. S. 156, 147 L. Ed. 2d 125, 137, 120 S. Ct. 2113, 2120 (2000) (plurality opinion).
Under 28 U. S. C. §2254(d)(1), a federal habeas court may grant relief despite the prior adjudication of the claim "if the state court identifies the correct governing legal rule from this Court's cases but unreasonably applies it to the facts of the particular state prisoner's case." Williams v. Taylor, 529 U. S. 362, 407 (2000). Penry seeks to invoke this exception to the general rule of claim preclusion. See Brief for Petitioner 12.
As noted, supra, at 10, the state court recognized Penry v. Lynaugh, 492 U. S. 302 (1989) (Penry I) and Boyde v. California, 494 U. S. 370 (1990) as the controlling precedents. These cases are further clarified in Johnson v. Texas, 509 U. S. 350 (1993). The crux of the present case is the state court's application of the Boyde standard to the instructions given in the second trial. The state court concluded that there was no reasonable probability that the jury understood the instructions to preclude it from giving effect to Penry's mitigating evidence. See supra, at 10. Was that conclusion reasonable? If so, that is the end of the case.
In Williams, the Court distinguished "an unreasonable application of federal law" from "an incorrect application of federal law." 529 U. S., at 410 (emphasis in original). The term "incorrect" may be appropriate for the tiny handful of cases which reach this Court for decision on the merits. However, amicus suggests that a change in terminology is in order, given that the vast majority of cases do not. Questions where the state and lower federal courts disagree are likely to be close questions, "and close questions, by definition, never have clearly correct answers." Christianson v. Colt Industries Operating Corp., 486 U. S. 800, 819 (1988). When this Court answers a question, that answer is deemed correct by virtue of the Court's position as the court of last resort. See Brown v. Allen, 344 U. S. 443, 540 (1953) (Jackson, J., concurring in the result) ("we are infallible only because we are final"). However, when the state court and the lower federal court disagree and this Court denies certiorari, we do not know which is correct and may not find out until years later, if ever. Compare Dunn v. Simmons, 877 F. 2d 1275, 1279 (CA6 1989), cert. denied, 494 U. S. 1061 (1990) (Kentucky procedure violates "federal standards"), with Parke v. Raley, 506 U. S. 20, 28 (1992) (same procedure "easily passes constitutional muster"); see also Scheidegger, Habeas Corpus, Relitigation, and the Legislative Power, 98 Colum. L. Rev. 888, 942 (1998). Rather than saying that the habeas court finds that the state court "applied clearly established federal law erroneously or incorrectly," Williams, 529 U. S., at 411, it would be better to say, where otherwise appropriate, that the habeas court disagrees with the state court's application.
Disagreement is not enough. Habeas reform was expressly premised on a rejection of the thesis that the federal court resolution of the question is necessarily better. See 141 Cong. Rec. 15,062, col. 2 (1995) (statement of Sen. Hatch); see also S. O'Connor, Local Control of Crime, Address to the Attorney General's Crime Summit 5 (Mar. 4. 1991), reprinted in Habeas Corpus Issues: Hearings before the House Subcommittee on Civil and Constitutional Rights, Serial No. 39, 102d Cong., 1st Sess., 198 (1991). Congress has commanded that the writ not issue for cases within the gray zone of reasonable disagreement. See Williams, supra, 529 U. S., at 411. The inquiry is objective rather than subjective, id., at 409-411, and hence Williams rejected the "all reasonable jurists" standard. Even so, the state court decision would have to be far wide of the mark to be "unreasonable." In Williams, for example, there was "an obvious failure to consider the totality of the omitted mitigation evidence." 529 U. S., at 416 (O'Connor, J., concurring) (emphasis added). No such failure appears here.
There were several ways to fix the problem in Penry I. Another special issue could be added, as the Texas Legislature did later. See Tex. Code Crim. Proc. Art. 37.071 §2(e)(1). The term "deliberately" in the first special issue could have been defined broadly enough to encompass Penry's mitigation evidence, as Penry I indicated would be sufficient. See 492 U. S., at 323. The state is not constitutionally required to provide any particular vehicle, however, so long as it provides some method for consideration of the defendant's mitigating evidence. See Johnson, supra, 509 U. S., at 370. One such method is to supplement the special issue with an instruction telling the jurors to answer "no" to one issue if they conclude on the basis of the mitigating evidence that Penry did not deserve to be sentenced to death, even if the literal answer to the unadorned question would otherwise be yes. Penry I held that an argument of counsel precisely to that effect was insufficient "[i]n light of the prosecutor's argument [that they could not do so], and in the absence of appropriate jury instructions . . . ." 492 U. S., at 326. By implication, an appropriate instruction would have been sufficient, and it would have been more than sufficient absent a contrary prosecutor's argument.
In the present case, the jury was told the consequences of its answers to the special issues and then told "your answers . . . should be reflective of your finding as to the personal culpability of the defendant, JOHNNY PAUL PENRY, in this case." J. A. 675. The jury was then instructed to consider all the mitigating evidence, defined in accordance with Lockett v. Ohio, 438 U. S. 586, 604 (1978) (plurality opinion). See J. A. 675. Then the court instructed, "If you determine, when giving effect to the mitigating evidence, if any, that a life sentence, as reflected by a negative finding to the issue under consideration, rather than a death sentence, is an appropriate response to the personal culpability of the defendant, a negative finding should be given to one of the special issues." Ibid. (emphasis added). The plain and obvious meaning of this instruction is to endorse the argument made by counsel at the first trial. It is the "appropriate instruction" that Penry I found lacking.
To be sure, supplementing a question with an instruction to answer in a way that does not conform to the bare question is not the optimum way to write instructions. The legitimate question before this Court, however, is not whether the instruction is exemplary, but only whether it meets constitutional minimums. See Victor v. Nebraska, 511 U. S. 1, 27-28 (1994) (Ginsburg, J., concurring in part and concurring in the judgment). Instructions of this type are not unusual, and nearly all of us have encountered them at some time. Job applicants are commonly asked if they have ever been convicted of a criminal offense, yes or no, followed by instructions directing a negative answer if the offense falls below some threshold of seriousness or age. Even though the answer as directed is different from the answer to the bare question, it seems to present little difficulty.
Indeed, even the essential "reasonable doubt" instruction, given in every criminal jury trial, requires a literally false answer in a sense. If the jury believes the defendant is much more likely guilty than not, yet reasonable doubt exists, they are instructed to pronounce him "not guilty," contrary to what they believe is the truth. This instruction is not only constitutional, it is constitutionally required. See Sullivan v. Louisiana, 508 U. S. 275, 278 (1993).
A clever enough lawyer can, of course, always find some claimed ambiguity in even the clearest language. Amicus National Association of Criminal Defense Lawyers engages in a bizarre parsing of the instruction in a quest for unclarity. NACDL claims that the phrase "at the time" somehow restricts the manner in which the jury can use the mitigating evidence, rather than just designating the time for doing so. Brief for National Association of Criminal Defense Lawyers as Amicus Curiae 6 (NACDL Brief). Yet that very sentence unambiguously tells the jury to use the evidence in assessing personal culpability, exactly what Penry I requires, see 492 U. S., at 327-328, and the following sentence expressly tells them to do so by means of a negative finding if they find a life sentence is the appropriate response.
Amicus NACDL goes on to claim that the latter sentence permits the jury to give effect to mitigating evidence "only when mitigation was appropriate 'as reflected by a negative finding to the issue under consideration.' " NACDL Brief 6. That is quite simply not what the instruction says. The quoted phrase follows and unambiguously refers to "a life sentence," not to the mitigating evidence, and they had been told only three sentences earlier that the penalty "should be reflective of your finding as to the personal culpability . . . ." J. A. 675.
"In examining the charge for the purpose of ascertaining its correctness in point of law, the whole scope and bearing of it must be taken together. It is wholly inadmissible to take up single and detached passages, and to decide upon them without attending to the context, or without incorporating such qualifications and explanations as naturally flow from the language of other parts of the charge." Magniac v. Thompson, 7 Pet. (32 U. S.) 348, 390 (1833) (Story, J.); see Boyde, supra, 494 U. S., at 378.
Only by the kind of parsing that Boyde disapproved, see 494 U. S., at 381, are petitioner and amicus NACDL even able to argue that the instructions failed to comply with Penry I.
The Texas Court of Criminal Appeals referred to the instruction in question as a "nullification instruction." J. A. 782. This is a poor choice of words, but the court's choice of words does not open an otherwise final judgment to attack. Cf. Coleman v. Thompson, 501 U. S. 722, 739 (1991).
"Nullification is, by definition, a violation of a juror's oath to apply the law as instructed by the court . . . ." United States v. Thomas, 116 F. 3d 606, 614 (CA2 1997). That means the whole law, as conveyed by the whole set of instructions. An instruction to answer a question in a particular way is not a "nullification" instruction, even if the directed answer is different from the answer to the bare question. Jurors who answer according to the full instructions are obeying, not violating, their oaths.
Roberts v. Louisiana, 428 U. S. 325 (1976), relied on by petitioner, is inapposite. In that case, the Court rejected an argument that an impermissibly mandatory capital sentencing statute was actually discretionary, because the jury could return a verdict of a lesser degree of murder. See id., at 334-335. In sharp contrast to the present case, the jury was told not to do that, but only to consider the lesser offense if the evidence did not justify conviction of the greater offense. Id., at 334, n. 10. Although the Louisiana jury had discretion in practice, it was a standardless discretion and depended on their willingness "to disregard the trial judge's instructions." Id., at 335.
In the present case, the jurors were given the Lockett definition of mitigation and expressly directed to give a negative finding to one of the special issues if they found that, given the mitigation, a life sentence was appropriate. J. A. 675. No disobedience was required to render a life sentence. The jurors chose the death penalty because they believed it was the appropriate sentence, after considering all the evidence.
Amicus NACDL cites a number of minimally relevant studies that supposedly show juror confusion in capital cases. Such studies should be taken with the greatest skepticism, as they are often driven by a predetermined agenda and not subject to adversarial testing. "Mere citation of a law review to a court does not suffice to introduce into evidence the truth of the hearsay or the so-called scientific conclusions contained within it." Ramdass v. Angelone, 530 U. S. 156, 147 L. Ed. 2d 125, 139, 120 S. Ct. 2113, 2123 (2000) (plurality opinion); see also id., 147 L. Ed. 2d, at 140, 120 S. Ct., at 2123 (noting interest of sponsor and dubious methodology). It is very easy to manipulate methods or analysis to come up with the desired conclusion. For example, the notorious Baldus study of capital sentencing claimed to show a race-of-the-victim bias, but when the case was actually tried the data were found to show, if anything, just the opposite. McCleskey v. Zant, 580 F. Supp. 338, 367-368 (ND Ga. 1984), aff'd in part, rev'd in part on other grounds, 753 F. 2d 877 (CA11 1985), aff'd, 481 U. S. 279 (1987).
To take just one example from the studies cited by NACDL, we look at the methodology in Haney and Lynch, Comprehending Life or Death Matters, 18 Law and Human Behavior 411 (1994). College students were read the standard California penalty instruction and then asked in the abstract to formulate their own definitions of the terms "aggravation" and "mitigation." Id., at 419. The students' imperfect product in this sterile, artificial context, see id., at 421, should surprise no one. The students did not receive the evidence and argument or participate in the deliberation that Boyde found important. See 494 U. S., at 380-381, 383-386. Further, the task they were asked to perform, formulating a definition, is quite different from the task juries perform and not, as the authors claim, simpler. Cf. Haney & Lynch, at 419, n. 6. Defining legal terms is a task our finest legal minds find challenging. The methodology of this study appears calculated to produce a negative result, and that result therefore has little or no meaning.
Applying the Boyde rule to the facts of this case requires considering the instruction in the context of the whole trial. See Boyde, supra, 494 U. S., at 383. The present case is remarkably similar to Boyde. The instruction is not erroneous, but it is claimed to be "ambiguous and therefore subject to an erroneous interpretation." Id., at 380. The jury was presented with copious mitigating evidence and told to consider all of it. Id., at 383. Indeed, in the present case, the jury received more complete instruction on mitigation than in Boyde, using the Lockett language. Compare J. A. 675, with Boyde, 494 U. S., at 374, n. 1 (factor (k)); and id., at 374, n. 2 (subsequently amended factor (k)). In both cases, defense counsel argued to the jury the interpretation of the instructions that permitted giving effect to all the mitigating evidence. Compare J. A. 640, with 494 U. S., at 386. In both this case and Boyde, and in sharp contrast to Penry I, the prosecutor did not deny that interpretation but instead argued, entirely properly, that the circumstances were factually unfounded, not mitigating, or entitled to little weight. Compare J. A. 662-664 (abuse unsubstantiated); J. A. 665-666 (retardation mild, not causally related to criminality, not mitigating, brain damage claim unfounded); and Boyde, 494 U. S., at 385 (argument that mitigation was minimal, not that it was not a proper factor), with Penry I, supra, 492 U. S., at 325-326 (urging jurors to stick to the literal questions). In general voir dire, the prosecutor expressly endorsed the concept that the jury could answer a question "no" if the mitigating evidence warranted a sentence less than death and stated flatly that determining what was mitigating was up to them, not dictated by the instructions. J. A. 9-10. He then drove the point home during individual voir dire, as described in detail in the state's brief. The state court's application of Boyde to the facts of this case is eminently reasonable. It is the opposite conclusion, argued by petitioner, that would be unreasonable.
This case is exactly the kind of case Congress had in mind when it acted to limit the second-guessing of state decisions on federal habeas corpus. See 141 Cong. Rec. 15,064, col. 2 (1995) (statement of Sen. Hatch) (purpose of habeas to correct fundamental defects; reasonable application means no fundamental defect). Nothing horrific happened at Penry's trial. There are no confessions beaten out of prisoners, no lynch mob around the courthouse, and no denial of counsel. Cf. Rose v. Lundy, 455 U. S. 509, 544 (1982) (Stevens, J., dissenting) (proposing limiting collateral review to such fundamental violations). We have a defendant ably represented by counsel who fully presents his mitigating evidence. The prosecution rebuts firmly but fairly. The trial judge instructs the jury in a careful, conscientious effort to comply both with this Court's complex and constantly changing jurisprudence of the Eighth Amendment and with state law to the extent it is consistent with that jurisprudence. The issue in dispute here raises no question of actual innocence. The argument involves "parsing instructions for subtle shades of meaning." Cf. Boyde, supra, 494 U. S., at 381. Direct review is enough for such disputes. They do not rise to the level that justifies collateral attack on a final judgment. The state court correctly recognized and reasonably applied this Court's precedents, and that, Congress has decided, ends the case.
|
|
||||
|
|
|
|
||