Hackett v. Price

BECKER, Circuit Judge,

dissenting.

The majority rejects Richard Hackett’s habeas petition by concluding that the state court did not reach a result that was *304 contrary to, or an unreasonable application of, the Eighth Amendment rule expressed in Mills v. Maryland, 486 U.S. 367 , 108 S.Ct. 1860 , 100 L.Ed.2d 384 (1988). The majority makes no showing, however, that the instructions and verdict slip in this case were at all different from, or any less confusing than, the instructions and verdict slips which led this Court to grant relief under Mills in Banks v. Horn, 271 F.3d 527 (3d Cir.2001), and Frey v. Fulcomer, 132 F.3d 916 (3d Cir.1997). Instead, the majority justifies its result on the sole ground that Hackett’s jury unanimously found “no mitigating circumstance.” The majority infers from this finding that Hackett’s jury must not have been confused by the instructions and verdict slip with regard to its role in finding mitigating circumstances. Like the able District Judge who granted Hackett’s petition, I find this analysis unconvincing.

In Banks , this Court granted relief under AEDPA upon concluding that essentially the same instructions and verdict slip made it likely a jury would believe that it could not consider mitigating evidence unless the whole jury was unanimous as to the existence of a particular mitigating factor. 16 The majority tries to evade Banks by noting that Hackett’s jury found “no mitigating circumstance,” whereas in Banks (and in Frey) the jury found at least one mitigating circumstance and reached the weighing stage. In my view, this distinction between the cases is immaterial. The trial court never explained what it meant to “unanimously” find “no mitigating circumstance,” and its instructions, as we have held in Banks and Frey , were reasonably likely to lead a jury to believe that unanimity was required before it could even consider a mitigating factor. Thus, contrary to the majority’s suggestion that each juror in Hackett’s case must have considered and rejected all mitigating evidence, the record more accurately reflects that this jury likely was unanimous only in finding that there was “no mitigating circumstance” upon which all jurors could agree.

Such an understanding of its instructions and verdict slip violated Hackett’s rights under the Eighth Amendment, as individual jurors were prevented from considering a mitigating factor, and possibly voting for a lesser sentence, absent the agreement of all jurors as to the presence of that factor. Put another way, a single juror, insisting Hackett failed to prove any mitigating circumstance by a preponderance of the evidence, could compel the death sentence despite other jurors who felt mitigating evidence was proven and sufficient to warrant a life sentence. This is precisely the situation that the Mills rule seeks to prevent. Indeed, the majority observes only in passing that Mills itself was a case in which the jury found no *305 mitigating circumstance and the Supreme Court vacated the sentence due to the likelihood that the jury thought it was precluded from considering mitigating evidence absent unanimity. See Mills, 486 U.S. at 370, 384 , 108 S.Ct. 1860 .

In short, there is a reasonable likelihood, under the standard of Boyde v. California, 494 U.S. 370 , 110 S.Ct. 1190 , 108 L.Ed.2d 316 (1990), that Hackett’s jury understood its instructions in a manner that violated Mills . Because the Pennsylvania Supreme Court’s rejection of this claim resulted in an unreasonable application of federal law, I respectfully dissent.

I.

As a threshold matter, I agree with the majority that the “unreasonable application of’ prong of 28 U.S.C. § 2254 (d)(1) appears to provide the best fit for analyzing the Pennsylvania Supreme Court’s decision to reject Hackett’s Mills claim. 17

*306 An AEDPA merits analysis of the Mills claim must begin, of course, with a review of the Supreme Court’s clearly established holdings in this area. Fortunately, our task in identifying the controlling law is made easier by the fact that our Court has fully explored the Mills question in the past. For example, in Frey , we summarized the relevant law as follows:

Under the Supreme Court’s current construction of the Eighth Amendment, the sentencer in a death penalty case must be permitted to consider all relevant mitigating evidence that the defendant proffers as counseling less than a sentence of death. Eddings v. Oklahoma, 455 U.S. 104 , 102 S.Ct. 869 , 71 L.Ed.2d 1 (1982); Lockett v. Ohio, 438 U.S. 586 , 98 S.Ct. 2954 , 57 L.Ed.2d 973 (1978). Accordingly, it is well established that the sentencer cannot be precluded from considering any such evidence. Skipper v. South Carolina, 476 U.S. 1 , 106 S.Ct. 1669 , 90 L.Ed.2d 1 (1986); Eddings, 455 U.S. at 114 , 102 S.Ct. 869 . The source of this preclusion is irrelevant; whether its source is statutory (Lock-ett ), the sentencing court (Eddings), or an evidentiary ruling (Skipper), the result is the same.

In Mills , the Supreme Court relied on these precedents to conclude that a death sentence should be vacated if there is a substantial probability that reasonable jurors, upon receiving the judge’s instructions and attempting to complete the verdict form based on those instructions, may have thought that they could only consider those mitigating factors which they unanimously found to exist. Put differently, if the jurors were led to believe that they could not each individually consider certain mitigating circumstances because there was not unanimous agreement as to the existence of those circumstances, then “some jurors were prevented from considering factors which may call for a less severe penalty, and petitioner’s sentence cannot stand.” Id. at 376 , 108 S.Ct. 1860 (internal citations omitted). See also Zettlemoyer v. Fulcomer, 923 *307 F.2d 284, 306-07 (1991) (discussing Mills).

Frey, 132 F.3d at 920-21 .

In Boyde v. California, 494 U.S. 370 , 110 S.Ct. 1190 , 108 L.Ed.2d 316 (1990), as the majority explains, the Supreme Court clarified that the proper standard for assessing a Mills challenge is “whether there is a reasonable likelihood that the jury has applied the challenged instruction in a way that prevents the consideration of constitutionally relevant evidence.” Id. at 380 , 108 S.Ct. 1860 . Thus, as we made clear in Frey , “while our inquiry is directed toward whether the [ ] instruction suffers from the same type of defect discussed in Mills (i.e. that the instruction could be read to require a unanimous finding of mitigating circumstances), our standard is that of Boyde , not Mills. ” Frey, 132 F.3d at 921 .

Frey was a pre-AEDPA case in which we found a Mills violation upon review of Pennsylvania jury instructions that mirrored, in all pertinent respects, the instructions at issue here. Cf. Frey, 132 F.3d at 922 (“The Crimes Code provides that the verdict must be a sentence of death if the jury unanimously finds at least one aggravating circumstance and no mitigating circumstances, or if the jury unanimously finds one or more aggravating circumstances which outweigh any mitigating circumstances.”) with Hackett, 212 F.Supp.2d. at 400-01 (“The Crimes Code provides that a verdict must be the sentence of death if the jury unanimously finds at least one aggravating circumstance and no mitigating circumstances, or if the jury unanimously finds one or more aggravating circumstance [sic] which outweigh any mitigating circumstances that you may determined.”). The majority understandably makes no attempt to distinguish the language of the Frey instructions from the instructions to Hackett’s jury.

We observed in Frey that this instruction “emphasize[d] the importance of a unanimous finding, using the phrase frequently and in close proximity to within seven words of the mitigating circumstances clause.” 132 F.3d at 923 (quoting the pertinent language as “if the jury unanimously finds at least one aggravating circumstance and no mitigating circumstance”). We noted that this “clause is, to the ear and to the mind, one sound bite,” id., and given the use of “unanimously” in close proximity to “mitigating circumstances,” we held that, when viewed in the context of the entire charge, “it is quite possible that a juror would, regardless of other qualifying language, believe that mitigating circumstances had to be found unanimously.” Id.

In Banks , a post-AEDPA case, we confronted a “very similar charge” to that issued in Frey and held that it, too, ran afoul of Mills. Banks, 271 F.3d at 547-49 . The instructions to the jury in Banks provided in relevant part as follows: “The Crime [sic] Code in this Commonwealth provides that the verdict must be a sentence of death if the jury unanimously finds at least one aggravating circumstance and no mitigating circumstances, or if the jury unanimously finds one or more aggravating circumstances which outweigh any mitigating circumstance or circumstances.” Id. at 546-47 . The majority makes no attempt at a distinction between the Banks and Hackett instructions, and I see no basis for one: in both cases, the instructions, viewed in their entirety, were likely understood to require unanimity to consider a mitigating factor. 18

*308 We concluded in Banks that the state court’s rejection of the petitioner’s claim resulted in an unreasonable application of federal law. Banks, 271 F.3d at 545 . We observed that the state court failed (as it did here) to analyze how the jury might have interpreted its instructions and verdict slip, merely holding that there was no error because the instructions tracked the language of 42 Pa. Cons.Stat. § 9711(c)(l)(iv). We found this determination insufficient because Mills , in conjunction with Boyde , requires a reviewing court to examine the entire charge and ask whether a reasonable jury could have concluded that unanimity was required to find a mitigating circumstance. Id. at 546. In fact, we quoted in Banks the District Court’s reasoning in Hackett’s case to explain that the Pennsylvania Supreme Court in both cases

“misconstrue[d] the court’s task in examining for Mills error by focusing on the meaning of the statute rather than on the issue of jury confusion. As Mills instructs, it is the danger of jury misinterpretation of the statutory scheme, rather than the existence of a constitutional interpretation of the statute by the courts, that creates the Mills problem.”

Id. at 544-45 (quoting Hackett v. Price, 212 F.Supp.2d 382, 404 (E.D.Pa.2001)).

Moreover, when discussing in Banks the reasons why the state court’s decision to reject the Mills claim amounted to an unreasonable application of federal law, we observed that the Pennsylvania Supreme Court’s analysis there contained the same flaw as its analysis in Hackett’s case. We explained as follows:

In Hackett , the Pennsylvania Supreme Court had attempted to distinguish Mills based on the Maryland statute, reasoning that since the Maryland statute barred consideration of mitigating evidence unless there was unanimous agreement and the Pennsylvania statute required unanimity as to the absence of a mitigating circumstance, the Pennsylvania statute allowed individual jurors to prevent death sentences but not to compel them. See Hackett, 212 F.Supp.2d 382 , 2001 WL 884721 at *19 . The differences in the statutes were not enough to render the Pennsylvania statute constitutional, since the danger of jury misinterpretation was present in both statutes, and the Pennsylvania Supreme Court had examined only the statute, not the possibility that the jury had been confused by the instructions given. Boyde v. California, 494 U.S. 370 , 110 S.Ct. 1190 , 108 L.Ed.2d 316 (1990), reh’g denied 495 U.S. 924 , 110 S.Ct. 1961 , 109 L.Ed.2d 322 , established that the jury instructions must be considered in the context of the entire proceeding, and the Pennsylvania Supreme Court’s failure to do so was contrary to clearly established federal law. Id. *309 at 21 . Here, the Pennsylvania Supreme Court has essentially ignored the teachings of Boyde and engaged the same reasoning regarding the constitutionality of the instructions as in Hackett - itself constitutionally defective.

Banks, 271 F.3d at 545-46 .

II.

It is against this background that we reach our consideration of the instructions and verdict slip employed in Hackett’s case. As noted, the majority does not, and cannot, distinguish Frey and Banks on whether the jury likely understood the instructions here to require unanimity to consider mitigating evidence. Like those earlier cases, the Hackett court used the word “unanimously” in close proximity to “mitigating circumstances,” instructing the jury that it must return a death sentence “if you unanimously find at least one aggravating circumstance and no mitigating circumstance.” As we explained in Frey , it is likely “a juror would, regardless of other qualifying language, believe that mitigating circumstances had to be found unanimously.” Frey, 132 F.3d at 923 . The Hackett trial court also repeated the instruction near the end of its charge, thus twice reinforcing the notion that the jury needed unanimity. See Maj. Op. at 297. Moreover, as in Banks and Frey , nowhere did the court explain or even hint that unanimity was unnecessary to find a mitigating factor. I note that this flaw in the jury instructions is seemingly remedied by Pennsylvania’s current model instructions, quoted in the margin. 19

The language of the verdict slip is also ambiguous, at best, as to the need for unanimity. The preface to the jury’s finding provides, “We, the jury, have found unanimously ...,” following which the jury placed a check to reflect “at least one aggravating circumstance and no mitigating circumstance.” In Banks , the corresponding language was identical, see Banks, 271 F.3d at 549 , and we held this language “at least confusing, and more likely suggestive, regarding the need for unanimity as to mitigating circumstances.” Id. at 550 . Moreover, the verdict slip here, as in Banks , had “no additional language that would imply that there is a different standard for aggravating circumstances than there is for mitigating circumstances.” Id. “There is also no language anywhere on the form from which the jury could infer that a mitigating circumstance might be marked if only one juror had found that circumstance to exist.” Id.

Under Boyde , the issue before us is whether it is reasonably likely Hackett’s jury interpreted its instructions and verdict slip to mean that it could find a mitigating circumstance only if all twelve jurors agreed upon its existence. This Court has twice resolved that issue in favor of the habeas petitioner with our holdings in Banks and Frey , and there is no basis for a different outcome here. Moreover, our analysis in Banks calls for relief under AEDPA, as we fully explained why the state court’s flawed application of Mills is unreasonable under clearly established federal law. See 271 F.3d at 543-551 .

*310 The majority, nevertheless, holds that there can be no Mills error because Hack-ett’s jury “unanimously” found “no mitigating circumstance,” a finding that the majority believes can mean one, and only one, thing- that each juror considered the mitigating evidence and agreed with all other jurors that no mitigating circumstance was proven, resulting in no Mills error. The record reflects, however, the trial court’s lone pertinent instruction on the finding of mitigating evidence was its directive to return a death sentence “if you unanimously find at least one aggravating circumstance and no mitigating circumstance.” What did it mean to Hackett’s jury to “unanimously” find “no mitigating circumstance”? The majority merely offers its view as to what the instruction “plainly” must have meant, but its analysis fails to comply with the Supreme Court’s directive in Boyde that “a single instruction to a jury may not be judged in isolation, but must be viewed in the context of the overall charge.” Boyde, 494 U.S. at 378 , 110 S.Ct. 1190 ; see also Banks, 271 F.3d at 546 (explaining that “the Boyde standard requires that the court view the instruction in its totality, not examine in isolation a few sentences that reference the Crimes Code.”). We must look, therefore, to the full context here rather than offer a facile interpretation of a single instruction viewed in isolation.

As noted, the court’s charge on mitigating evidence was at best ambiguous. Indeed, we expressly found in Banks that, based on these instructions, “[tjhere is no way that a juror would understand that a mitigating circumstance could be considered by less than all jurors.” Banks, 271 F.3d at 548 (emphasis added). The Hack-ett trial court also did not explain that a juror was free to find a mitigating circumstance despite the views of other jurors, and it issued no instruction suggesting that a finding of “no mitigating circumstance” should reflect that each juror had considered and rejected all mitigating evidence individually rather than reflecting that there was no mitigating circumstance of common agreement. The language of the verdict slip was likewise ambiguous, if not suggestive, as to a need for unanimity on mitigating circumstances. The verdict slip also contained no language that would imply that mitigating circumstances were subject to a different standard or that mitigation might be found by only one juror.

Hackett’s jury, then, was left to decide whether to check the box reflecting a finding of “at least one aggravating circumstance and no mitigating circumstance” based solely on its understanding that it could find a mitigating circumstance only with the unanimous agreement of all jurors. In this light, it seems far more likely, if not outright certain, that the jury believed itself required to return a death verdict if it unanimously found an aggravating circumstance and unanimously found “no mitigating circumstance” upon which all jurors could agree. As Judge Padova observed:

With respect to the specific question of whether the clause requires the unanimous rejection of all mitigating factors, it is simply not clear, in the context in which it was given, whether the instruction means that a death sentence is warranted if the jury unanimously finds at least one aggravating circumstance and unanimously finds that no mitigating circumstances exist (as the [majority] contends), or if the phrase means a death sentence is warranted if the jury unanimously finds at least one aggravating circumstance and finds no unanimously agreed upon mitigating circumstances. The jury’s interpretation of the unanimity requirement as to the rejection *311 of mitigating circumstances would depend on how the jury understood and employed the concept of unanimity. If the jury believed that all jurors would have to agree as to the existence of the same particular mitigating circumstance, then the jury could reach a death verdict even if individual jurors believed mitigating factors existed, so long as all the jurors did not agree as to the existence of a particular mitigating circumstance. Such a jury would meet the requirement of the “no mitigating circumstance” prong on the basis that all the jurors agreed, unanimously, that there was no mitigating circumstance that was commonly agreed upon by all twelve jurors.

The ambiguity of the language of the sentencing statute, the additional jury instructions, and the verdict form all contribute to the conclusion that it was reasonably likely the jury believed it must find mitigating circumstances unanimously before considering them, and that it could reach a verdict of death under the “no mitigating circumstance” prong even if one or more jurors believed that a mitigating circumstance existed without performing the weighing inquiry.

Hackett, 212 F.Supp.2d at 411-412 .

The District Court was correct that we must infer from the record that this jury acted in accordance with a belief that unanimity was required to find a mitigating factor, and that its finding of “no mitigating circumstance” reflects that erroneous belief. The majority’s alternative view is untenable, as it merely assumes, with no record support, that the jury applied its instructions so as not to violate Mills . I find this assumption a troubling basis upon which to affirm a sentence of death.

In addition to ignoring the context of this case, the majority, as noted above, observes only in passing, that the jury in Mills likewise found no mitigating circumstance to exist and imposed a death sentence without reaching the weighing stage. Mills, 486 U.S. at 370 , 108 S.Ct. 1860 . The petitioner in Mills argued that the court’s instructions and verdict slip served “to require the imposition of a death sentence if the jury unanimously found an aggravating circumstance, but could not agree unanimously as to the existence of any particular mitigating circumstance.” Id. at 371 , 108 S.Ct. 1860 . The petitioner thus claimed that “even if some or all of the jurors were to believe that some mitigating circumstance or circumstances were present, unless they could unanimously agree on the existence of the same mitigating factor, the sentence necessarily would be death.” Id. The Maryland Court of Appeals rejected this claim, concluding, inter alia, that because the instructions and verdict slip expressly required unanimity in the acceptance or rejection of a mitigating circumstance, the jury’s finding likely reflected (as the majority concludes here) a unanimous determination that a mitigating circumstance did not exist, not a failure to find unanimously the existence of a mitigating circumstance. Id. at 372 , 108 S.Ct. 1860 .

The Supreme Court rejected the Maryland court’s conclusion as to what the jury likely understood its task to be. The Court distilled the question before it as follows:

If the jury understood the verdict form as the Court of Appeals asserted it should have, then every time it marked “no” beside a mitigating circumstance it indicated its unanimous conclusion that petitioner had not proved the relevant facts by a preponderance of the evidence, and thus the *312 court properly upheld the judgment. On the other hand, if the jury understood that it should mark “no” when it failed to agree unanimously that a mitigating circumstance existed, then some jurors were prevented from considering “factors which may call for a less severe penalty,” Lockett v. Ohio, 438 U.S. at 605 , 98 S.Ct. 2954 , and petitioner’s sentence cannot stand.

Mills, 486 U.S. at 376 , 108 S.Ct. 1860 (footnote omitted). The Court conceded that the Maryland Court of Appeals’ interpretation of the record was “plausible,” but it found, after reviewing the instructions as a whole and the language of the verdict slip, that it could not reject the petitioner’s interpretation “with any degree of certainty.” Id. at 377 , 108 S.Ct. 1860 . The Court thus vacated the sentence because the jurors “well may have thought they were precluded from considering mitigating evidence unless all 12 jurors agreed in the existence of a particular such circumstance.” Id. at 384 , 108 S.Ct. 1860 .

While the language of the instructions in Mills took a different form than those at issue here, see 486 U.S. at 378-80 , 108 S.Ct. 1860 , the practical effect of the jury’s finding of no mitigation in each case is the same. As in Mills , the instructions here likely prevented the jury from considering mitigating evidence absent unanimity, and in both cases the juries found no mitigating factor to exist. Unlike my colleagues, however, the Supreme Court refused to infer from the finding of no mitigation that the jury must have reached a unanimous conclusion that the petitioner failed to prove the relevant facts by a preponderance of the evidence. Such an inference could not be drawn in Mills after a consideration of the context of the jury instructions and verdict slip, and the same is true here. Of course, we cannot verify whether Hackett’s jury in fact believed itself precluded from considering mitigating evidence absent unanimity, but we can safely conclude that there is a reasonable likelihood the jury interpreted its task in that manner, and that this death sentence, therefore, cannot stand under Mills .

The majority’s attempt to find support for its conclusion in the different sentencing verdicts reached for Hackett’s co-defendants rings hollow. The jury acted upon the same instructions in its consideration of the sentencing evidence as to all four defendants. The jury sentenced one defendant (Spence) to death after finding that aggravation outweighed mitigation, but given its instructions and verdict slip, the jury likely reached a unanimous finding as to a mitigating circumstance (in violation of Mills) and thereby reached the weighing stage. This does nothing to diminish the showing of a Mills error as to Hackett. The same can be said for the two co-defendants (Gray and Barrett) who received life sentences given its instructions, the jury likely was either unanimous in its agreement that the Commonwealth failed to prove an aggravating circumstance, or unanimous as to particular mitigating circumstances and found that those mitigating circumstances outweighed the aggravating circumstances, thereby resulting in life sentences. The different sentences thus do nothing to bolster the majority’s speculation that there was no Mills error.

It should be added that Hackett’s jury was not without substantial mitigating evidence to consider. The Pennsylvania death penalty statute specifies eight categories of mitigating evidence, and Hackett argued mitigation under five of those categories: “[t]he defendant has no significant history of prior criminal convictions”; “[t]he defendant was under the influence of extreme mental or emotional disturbance”; “[t]he capacity of the defendant to *313 appreciate the criminality of his conduct or to conform his conduct to the requirements of law was substantially impaired”; “[t]he age of the defendant at the time of the crime”; and “[a]ny other evidence of mitigation concerning the character and record of the defendant and the circumstances of his offense.” 42 Pa. Cons.Stat. § 9711(e)(l)-(4) and (8).

Hackett was twenty years old at the time of the offense. He had one prior conviction in Montgomery County at age nineteen for conspiracy and four counts of burglary, for which he was sentenced to concurrent terms of ten and one-half to twenty-three months imprisonment. His mother, Bonnie Hackett, testified that her son twice failed the first grade and suffered from severe learning problems that resulted in his attending a special school throughout childhood for the learning and emotionally disabled. She further testified, and documentary evidence reflected, that Hackett was diagnosed with perception problems and as functioning at a “low average” or “dull normal” range of intelligence based on separate full-scale IQ scores of 80 and 85. She testified that her son began to abuse alcohol around age eighteen. Ann Marie Clay, a long-time neighbor and family friend, also testified to Hackett’s childhood difficulties.

Doctor Albert Levitt, a psychologist, testified as to the results of various tests he performed on Hackett before trial. He explained the results as showing “very regressed, immature perception of reality situations, which is a sign of immaturity and possibly illness.” App. at 48. Hack-ett also displayed a low-level of maturity, signs of depression, suicidal problems, an immature view of social relationships, anxiety, and “psychosexual difficulties.” Id. at 51-52. Doctor Levitt concluded that Hackett “was limited in terms of coping mechanisms, limited in terms of the way he functions, limited in understanding social relationships. There was a significant degree of psychopathology when I saw him, with a lot of morbid and negative thought processes.” Id. at 54.

Doctor Levitt also described the results of the standardized Minnesota Multiphas-ing Inventory, a computer-graded true/ false test. He explained the results as indicating, among other things: “[limited interpersonal resources”; “[h]igh clinical depression, which means that his depression has reached pathological limits”; “[pjrojective of blame and hostility”; “[p]aranoid features”; “[l]ook[s] for delusions of persecution and maltreatment”; “[m]ay have a thought disorder[, which] would be schizophrenia”; and “[hyperactive and agitated.” Id. at 55-56. Doctor Levitt found these results consistent with his own finding that “the clinical impression was schizoaffective disorder and major depression.” Id. at 56. He opined that Hackett “is an emotionally disturbed individual, especially when he has to interact in social situations,” and that if “he has to interact socially, he tends to get anxious, confused and develops all kinds of distorted ideations.” Id. at 57. Doctor Levitt added that “[ajlcohol would have a significant effect” on Hackett because “[h]is perceptions of reality are not good in the first place, and alcohol would cause all these things to become more distorted.” Id. at 59.

The likelihood that Hackett’s jury misunderstood its task, resulting as it did in the inability of individual jurors to consider the evidence in mitigation and possibly impose a lesser sentence, poses precisely the type of Eighth Amendment problem that the Mills rule seeks to eliminate. I again refer to Judge Padova’s thoughtful analysis:

Consider, for example, the following scenario. All twelve jurors agree that *314 a particular aggravating factor exists. Two of the jurors conclude that age is a mitigating factor in the case; however, the other ten jurors do not agree. None of the jurors believe that any other mitigating factors exist, and so the jury is not unanimous as to the existence of the individual mitigating factor. Believing that unanimity is required in order to find that a mitigating factor exists, the jurors “find” one aggravating factor, and fail to find any commonly agreed upon mitigating factors. Rather than moving on to the weighing prong and determining if the aggravating circumstance outweighs the mitigating factor, the jury uses its findings and renders a verdict of death on the basis that the jury unanimously agreed to the existence of “at least one aggravating circumstance” and that the jury unanimously agreed that there was “no mitigating circumstance” that was commonly agreed upon by all twelve jurors. Notwithstanding the fact that two of the jurors believed that a mitigating factor existed, because of the jury’s erroneous interpretation of the instructions with respect to the requirement of unanimity, the result would be a sentence of death. This scenario even creates the possibility of a death verdict under the “no mitigating circumstance” prong where all twelve jurors believe a mitigating factor to exist, but there is no agreement as to the particular factor. For example, two jurors could believe age is a mitigating factor and the other ten jurors could believe the defendant’s lack of significant criminal history is a mitigating factor. All twelve jurors thus believe that a mitigating factor exists, but because they do not all agree as to the same particular mitigating factor, they conclude that there are no commonly agreed upon mitigating factors. Based on their findings, the jurors thus conclude, unanimously, that there are no mitigating factors that were unanimously agreed upon by all the jurors. The jury therefore delivers a verdict of death without ever weighing the aggravating circumstance against the mitigating factors.

Hackett, 212 F.Supp.2d at 411-12 .

Finally, the majority correctly notes that Pennsylvania adopted changes to its verdict slip subsequent to Hackett’s trial, see Maj. Op. at 301 n. 14, but the majority seems to believe that these changes pertain to “weighing cases” only and that “no similar change was made to strengthen the directive that a jury’s finding of no mitigating circumstance must be unanimous.” Id. The changes to Pennsylvania’s model jury instructions suggest otherwise.

The present model instructions, adopted after Hackett’s trial, provide in pertinent part as follows with regard to how a Pennsylvania jury should reach its capital sentencing findings and record them on the verdict slip:

You have been given a sentencing verdict slip on which to record your verdict and findings. I shall now give you further directions on how to go about reaching a verdict, making findings, and using the verdict slip.

As I told you earlier, and as the verdict slip indicates, you must agree unanimously on one of two general findings before you can sentence the defendant to death. They are a finding that there is at least one aggravating circumstance and no mitigating circumstances or a finding that there are one or more aggravating circumstances which outweigh any mitigating *315 circumstances. (In deciding whether aggravating circumstances outweigh mitigating circumstances, do hot simply count their number. Compare the seriousness and importance of the aggravating with the mitigating circumstances.) If you all agree on either one of the two general findings then you can and must sentence the defendant to death.

When voting on the general findings, you are to regard a particular aggravating circumstance as present only if you all agree that it is present. On the other hand, each of you is free to regard a particular mitigating circumstance as present despite what other jurors may believe. (This different treatment of aggravating and mitigating circumstances is one of the law’s safeguards against unjust death sentences. It gives a defendant the full benefit of any mitigating circumstances ....)

Pa. Suggested STANDARD Criminal Jury Instructions Nos. 15.2502F(7) & 15.2502H(l)-(2) (2000). The Committee Note to instruction 15.2502H explains that the final portion of the above-quoted language is intended to “tell the jurors that they are to make collective decisions about the aggravating circumstances but that they ultimately are to make individual decisions about mitigating circumstances.”

This addition to the model instructions coupled with the decision no longer to use the word “unanimously” in close proximity to “mitigating circumstances” plainly seeks to eliminate the previous ambiguity as to whether all jurors had to agree on a mitigating factor. The new model instructions state that “each of you is free to regard a particular mitigating circumstance as present despite what other jurors may believe” language that should reduce the risk, present even in a “non-weighing case” like Hackett’s, that the jury would interpret its mandate as requiring a death sentence only if it unanimously agreed that there was no mitigating circumstance of common agreement. The new instructions would leave a unanimous finding of “no mitigating circumstance” more likely to reflect that each juror actually considered and rejected all mitigating evidence. Notably, in Commonwealth v. Cross, 555 Pa. 603, 726 A.2d 333 (Pa.1999), two concurring Justices of the Pennsylvania Supreme Court opined that the jury instruction in Frey , which mirrors the Banks and Hack-ett instruction, “contains a degree of imprecision that can be avoided merely by employing Pennsylvania Suggested Standard Criminal Jury Instruction 15.2502H(3) [currently 15.2502H(2) ], which removes any doubt concerning a juror’s role in assessing mitigating circumstances.” Id. at 338-39. These Justices noted that they “would affirmatively require Pennsylvania trial courts to instruct capital sentencing juries in accordance with the standard instruction.”

Pennsylvania’s well-intended effort to clarify its sentencing procedure does not in itself indicate that its former instructions and verdict slip were constitutionally infirm. Nevertheless, the above-noted revisions, as well as the views expressed by the concurring Pennsylvania Justices in Cross, lend support to the conclusion that an instruction on unanimity that merely tracks the language of 42 Pa. Cons.Stat. § 9711(c)(l)(iv) is imprecise and subject to an impermissible risk of jury misinterpretation on whether unanimity is required to consider mitigating evidence. Moreover, these revisions further support the view that there was insufficient clarity in the instructions and verdict slip at Hackett’s trial to permit this Court to uphold his sentence of death on the sole ground that the jury found “no mitigating circumstance.” At the very least, my colleagues *316 in the majority should find themselves in agreement with the following observation from Mills :

The decision to exercise the power of the State to execute a defendant is unlike any other decision citizens and public officials are called upon to make. Evolving standards of societal decency have imposed a correspondingly high requirement of reliability on the determination that death is the appropriate penalty in a particular case. The possibility that [Haekettj’s jury conducted its task improperly certainly is great enough to require resentencing.

Mills, 486 U.S. at 383-84 , 108 S.Ct. 1860 .

III.

For these reasons, I conclude that Hack-ett has established a Mills error, and that the state court’s rejection of this claim resulted in an unreasonable application of clearly established federal law. I therefore would affirm the District Court’s order vacating Hackett’s sentence and allow Pennsylvania to resentence him to life imprisonment or conduct a new capital sentencing proceeding.

. As the majority notes, our decision in Banlcs to grant relief under Mills is no longer precedential given the Supreme Court’s holding in Beard v. Banks,-U.S. --, 124 S.Ct. 2504 , 159 L.Ed.2d 494 (2004), that the Mills rule is barred under Teague v. Lane, 489 U.S. 288 , 109 S.Ct. 1060 , 103 L.Ed.2d 334 (1989), from retroactive application. See Maj. Op. at 295 n. 9. Hackett’s conviction became final after Mills was decided, and thus there is no Teague problem here. The majority correctly observes that our discussion in Banlcs regarding the Mills issue, although not precedential, is nevertheless "instructive and relevant to our current inquiry,” Maj. Op. at 295 n. 9, particularly given that the analysis in Banlcs reflects the unanimous judgment of three members of this Court in granting relief on a Mills issue that is, I submit, indistinguishable from the one presented here. It also bears mention that Justice Stevens, joined by Justices Souter, Ginsburg, and Breyer, filed a dissent in Beard v. Banks in which he noted that the four dissenters would have affirmed the grant of habeas relief on the Mills question in Banks ”[f]or the reasons identified by the Third Circuit[.]” - U.S. at -, 124 S.Ct. at 2517 (Stevens, J., dissenting).

. I do not concede, however, that the Pennsylvania Supreme Court's decision is necessarily immune from analysis under the "contrary to” test, or that its decision would survive such scrutiny. As Justice O'Connor explained for the majority in Williams v. Taylor, 529 U.S. 362 , 120 S.Ct. 1495 , 146 L.Ed.2d 389 (2000), "[a] state-court decision will certainly be contrary to our clearly established precedent if the state court applies a rule that contradicts the governing law set forth in our cases.” 529 U.S. at 405 , 120 S.Ct. 1495 . Justice O'Connor offered the following example of a "contrary to” application of federal law:

Take, for example, our decision in Strickland v. Washington, 466 U.S. 668 , 104 S.Ct. 2052 , 80 L.Ed.2d 674 (1984). If a state court were to reject a prisoner's claim of ineffective assistance of counsel on the grounds that the prisoner had not established by a preponderance of the evidence that the result of his criminal proceeding would have been different, that decision would be "diametrically different,” "opposite in character or nature,” and "mutually opposed” to our clearly established precedent because we held in Strickland that the prisoner need only demonstrate a "reasonable probability that ... the result of the proceeding would have been different.” Id., at 694 , 104 S.Ct. 2052 .

Williams, 529 U.S. at 405-06 , 120 S.Ct. 1495 .

A similar flaw can be found in the state court analysis here. The Pennsylvania Supreme Court cited Mills and purported to analyze Hackett’s claim under the Mills rule, but the court failed to consider, as is required by Mills and Boyde , whether the jury likely interpreted its instructions and verdict slip in a manner that prevented its consideration of relevant mitigating evidence. The state court instead reasoned that, because the instructions mirrored the language of 42 Pa. Cons. Stat. § 9711 (c)(1)(iv), and because it had previously held that statute constitutional by finding that it does not expressly permit a single juror who refuses to agree that mitigation exists to force a death verdict on the other jurors, there can be no Mills violation. See Commonwealth v. Hackett, 534 Pa. 210 , 627 A.2d 719, 725 (1993). This analysis employed a rule {i.e., no Mills violation if the challenged instructions can be read in a constitutional manner) that did not comport with the governing rule set forth in Mills and Boyde {i.e., whether there is a reasonable likelihood that the jury applied the challenged instructions in a way that prevented consideration of mitigating evidence). In Mills itself, the state court interpreted its statute as constitutional, but the Supreme Court vacated the state court decision because there was a substantial risk that the jury did not understand that it could consider mitigating evidence absent unanimous agreement. Mills, 486 U.S. at 371, 384 , 108 S.Ct. 1860 .

The majority finds that the state court "applied Mills , but did not apply Boyde," and therefore the state court’s analysis was merely "incomplete” but not necessarily contrary to Supreme Court precedent. See Maj. Op. at 300. The majority fails to acknowledge, however, that the Pennsylvania Supreme Court never applied Mills in the sense that it failed to identify the correct rule from Mills , not to mention its failure to apply the Boyde standard. Thus, as in Justice O'Connor’s example involving a misapprehension of the Strickland rule, the state court's decision arguably fits the mold of "contrary to” federal law. Cf. Bell v. Cone, 535 U.S. 685, 698 , 122 S.Ct. 1843 , 152 L.Ed.2d 914 (2002) (“We hold, *306 therefore, that the state court correctly identified the principles announced in Strickland as those governing the analysis of respondent's claim. Consequently, we find no merit in respondent’s contention that the state court's adjudication was contrary to our clearly established law.”). Moreover, Hackett can show not only that the state court failed to apply the correct Mills rule, but that he is entitled to relief under a correct application of Mills and Boyde . Thus, although I would agree with the majority that ”[i]f the Pennsylvania Supreme Court denied Hackett relief under Mills , it logically would have done the same if applying Boyde,” Maj. Op. at 300, the Pennsylvania Supreme Court’s failure to apply Mills correctly precludes an inference that it also would have denied relief under the Boyde standard.

In the final analysis, however, this case may fall in the overlap between the “contrary to” and "unreasonable application of” standards, given that the state court assessed and sought to distinguish Hackett's claim by analogy to Mills , despite its failure to conduct a proper Mills inquiry. See Moore v. Morton, 255 F.3d 95, 104 (3d Cir.2001) (noting that "[i]n Williams v. Taylor, 529 U.S. 362, 404-06 , 120 S.Ct. 1495 , 146 L.Ed.2d 389 (2000), the Supreme Court held that 'contrary to’ and ‘unreasonable application of' have independent, if overlapping meanings.”). I see no need to explore this "overlap” here. In Banlcs, we held that the Pennsylvania Supreme Court rendered an unreasonable application of federal law on indistinguishable instructions, and we opted not to analyze the claim under the "contrary to” test. Banks, 271 F.3d at 545 n. 21. Because I believe the same approach can be taken here, there is little need to engage in a separate contrary to analysis for Hackett's claim. Moreover, the majority denies relief to Hackett primarily by finding that the state court's decision survives the unreasonable application test. Thus, I too focus upon the unreasonableness standard.

. The Commonwealth’s only argument in this regard is that Banks was wrongly decided on the Mills question, as it contends that Banks “did not apply the correct evaluative inquiry under Boyde , but rather, consistently and improperly employed possibility-based, *308 and single-juror based, standards of evaluation that Boyde explicitly rejected.” Appellant’s Br. at 51. The Commonwealth levels a similar charge against Frey . Appellant's Br. at 56-58, 61 -62. However, a close review of Banks and Frey shows that this Court did not misapply the Boyde standard. The Court was careful in Frey to note the difference between Mills and Boyde , and we held, in accordance with Boyde , that "it was reasonably likely that the jury could have believed that it was required to find the existence of mitigating circumstances unanimously....” Frey, 132 F.3d at 924 . Likewise, we held in Banks that, "[c]onsidered as a whole, the jury instructions leave no doubt ‘there is a reasonable likelihood that the jury has applied the challenged instruction in a way that prevents consideration of constitutionally relevant evidence.' " Banks, 271 F.3d at 549 (quoting Boyde, 494 U.S. at 380 , 110 S.Ct. 1190 ).

.

When voting on the general findings, you are to regard a particular aggravating circumstance as present only if you all agree that it is present. On the other hand, each of you is free to regard a particular mitigating circumstance as present despite what other jurors may believe.

Pa. Suggested Standard Criminal Jury Instructions No. 15.2502H(2) (2000).