Opinion
No. 92 Civ. 8663 (KMW).
January 16, 1996.
Edward Scott Zas, Paul, Weiss, Rifkind, Wharton Garrison, New York City, for John F. Lyons.
Nancy D. Killian, Bronx District Attorney, Bronx, NY, for Sally B. Johnson.
OPINION AND ORDER
John Lyons brings this petition for a writ of habeas corpus pursuant to 28 U.S.C. § 2254, challenging his conviction for attempted murder in the second degree, N.Y. Penal Law §§ 110.00, 125.25, following a jury trial in New York State Supreme Court, Bronx County. Based on this conviction, Lyons is currently serving an indeterminate sentence, imposed on April 16, 1990, of from eight and one third to twenty-five years of imprisonment.
Lyons was also convicted of criminal possession of a weapon, in the third degree, N.Y.Penal Law § 265.02(4), but he does not dispute this conviction, for he concedes that he possessed a loaded weapon on the day of the crime.
Lyons was also sentenced to a concurrent, indeterminate term of from two and one-third to seven years on the weapons count, which he does not challenge.
On appeal, New York's Appellate Division, First Department, affirmed Lyons's conviction. People v. Lyons, 180 A.D.2d 440, 579 N.Y.S.2d 664 (1992). The New York Court of Appeals denied Lyons leave to pursue a further appeal. Lyons now petitions this court for a writ of habeas corpus, arguing that his Sixth Amendment right to a fair trial was violated because the trial court erroneously denied him the right to introduce relevant, exculpatory evidence.
I initially referred this case, for a Report and Recommendation, to Magistrate Judge James C. Francis IV, who recommended that I find that the trial court erred, but deem that error harmless. I agree with the Magistrate Judge's finding of trial error. However, I do not find the error to have been harmless, when viewed in the context of the record as a whole. As a consequence, I grant the writ.
I. Background
The undisputed facts of the case are these: On January 2, 1989, in the Bronx, a dispute arose in the midst of a crowd of people, which included both the petitioner, John Lyons, and another man, Kevin Moore. Lyons and Moore are both young, African-American men who were wearing black leather jackets that day. A third man, Jose Quiles, approached the crowd. Moore spoke to Lyons. Lyons entered Moore's family's residence and, a few minutes later, emerged carrying a gun. Members of the crowd began to run. Three shots were fired; two of them struck Quiles.
Lyons's defense at trial, when he took the stand and testified, was that Moore, not he, had shot Quiles. A defense witness, Pantojas, corroborated Lyons's testimony by testifying that (1) Moore was the shooter; (2) Moore wore gold "fronts" (temporary, decorative caps) on his teeth; and (3) no one else on the scene wore gold fronts. However, three prosecution witnesses — Ferrera, Poole, and Quiles — testified, to the contrary, that (1) Lyons was the shooter; (2) Lyons wore four gold fronts on his teeth; and (3) no one else on the scene wore gold fronts.
Unlike all of the other witnesses in the case, Lyons himself testified that both he and Moore wore gold fronts on the day of the crime. Moreover, in a voir dire conducted out of the presence of the jury, Moore, who was wearing gold fronts at the time of trial, corroborated Lyons's testimony by stating that the gold fronts he, Moore, was wearing that day were the same ones he, Moore, had worn on the day of the crime. However, Moore stated that he would assert his Fifth Amendment privilege against self-incrimination rather than repeat this testimony before a jury.
The voir dire examination proceeded, in relevant part, as follows:
Defense Counsel: Mr. Moore, could you please show the judge what you have on your teeth?
Prosecution: Objection.
Court: Sustained. I recognize he has teeth on February 28th, 1990 at noon.
Defense Counsel: On January 2d, 1989 [the day of the crime], did you have gold fronts on your teeth?
Moore: Yes.
Defense Counsel: And do you have those same gold fronts on your teeth today?
Moore: Yes.
Defense Counsel: Can you take those gold fronts out?
Moore: Yes.
Defense Counsel: Can you take them out for us now please?
[Moore complies]
Defense Counsel: For the record, how many teeth do those fronts cover?
Moore: Four.
Defense Counsel: Mr. Moore, if I called you to testify at this proceeding about the questions that I just asked you, if I called you to testify, would you take the Fifth Amendment about what I just asked you?
Moore: Yes.
Defense Counsel: Now if I called you to just exhibit those gold fronts in your mouth to the jury today, would you refuse to do that?
[Objection to the question sustained].
Defense Counsel: If I called you to take the witness stand and show the jury what you had in your mouth today, would you do that?
[Objection to the question sustained].
Defense Counsel: If I asked you whether or not on January 2d, 1989 [the day of the crime] you were in the vicinity of Field Place [the crime scene] with those gold fronts, would you take the Fifth Amendment or not?
Moore: Yes.
Defense Counsel: Out of the presence of the jury would you allow your gold fronts to remain here during this trial?
Moore: Yes.
Defense Counsel: If I asked you to give those gold fronts which you have now to John Lyons, would you do it?
Moore: Yes.
Defense Counsel: And would you allow John Lyons to retain those gold fronts during the pendency of this proceeding?
Moore: Yes.
Defense Counsel: And the gold fronts that you would give to John Lyons today in court, are they the same gold fronts that you had on your teeth on January 2, 1989 [the day of the crime]?
Moore: Yes.
. . . .
Prosecutor: [I]f I asked you some questions with regard to January 2d, 1989 and I'm being specific, about the defendant John Lyons and yourself, would you answer those questions for me?
Moore: No. I will take a plea.
Court: You mean you would invoke your privilege against self-incrimination?
Moore: Yes.
Prosecutor: If I asked you some questions about January 2d, 1989, in relation to whether or not the defendant had gold fronts, would you answer the question for me?
Moore: No.
Prosecutor: So what you're telling this court [and the prosecution and the defense] is that if we asked you questions concerning an incident that took place on January 2, 1989 [the day of the crime] you will choose to take your Fifth Amendment rights?
Moore: Yes.
Tr. 424-32.
The trial court indicated that it would uphold Moore's invocation of his Fifth Amendment privilege at trial. Defense counsel then requested that Moore be compelled to testify at trial, that Moore be directed to take the stand to assert his Fifth Amendment privilege, or that Moore be marked as an exhibit, entered into evidence, and displayed, with his gold fronts, to the jury. The court refused all of these requests.
The court offered, however, to show the jury a photograph of Moore without his gold fronts to allow the jury to consider whether Lyons and Moore looked alike. Defense counsel declined the court's offer, pointing out that the gold fronts were "the whole issue" in the case. The court also, at one point, asked the prosecutor whether he was willing to stipulate that Moore wore gold fronts on the day of the crime, but the prosecutor refused. As a consequence, the jury never heard Moore's testimony or saw him or any image of him, with or without his gold fronts.
The government claims that the record indicates that the judge offered to take a photograph of Moore showing Moore wearing his gold fronts. Having reviewed the full record, I reach the opposite conclusion. One colloquy between the judge and the defense lawyer was, in relevant part, as follows:
Judge: You can introduce a photograph of [Moore] here. . . . I have ruled he may physically not be introduced as a piece of evidence, as an exhibit. . . . If you want to take the photograph of the witness, take it now. . . . If you lay the proper foundation I will allow it to go into evidence as a defense exhibit.
Judge: [Moore is] here now. He's available to be photographed.
Defense Counsel: I have to lay the chain of custody as to his gold teeth. I want him to hand over those gold teeth to Mr. Lyons in court, here, right now. . . .
Judge: There is no way you can get the evidence in, so I'm not going to order the person to turn it over.
Defense Counsel: The witness has testified to the exact nature of those gold teeth. He said they're the same gold teeth.
Judge: [Moore is] going to take the Fifth Amendment. . . . I think you're waiving your right to have the photograph taken.
Defense Counsel: I'm not waiving my right. . . . I choose not to take a photograph of Mr. Kevin Moore because it's not the best evidence.
Tr. 437-39. A later colloquy clarifies the nature of the judge's offer to photograph Moore, and establishes, as well, that the defense lawyer sought a photograph which did include the fronts:
Judge: I'm ordering the Court Officers who brought Mr. Moore here to take a photograph of Kevin Moore. . . .
. . . . .
Defense Counsel: With his teeth, without his teeth, his teeth in his hands, his teeth in his mouth?
Prosecutor: Judge, what's going to happen — I want to make my objection noted for the record. When we had Mr. Moore on the stand [for a voir dire outside the jury's presence] he testified that those teeth are removable, he can take them in and out. Now if the Court plans on letting [the defense] introduce the exhibit without laying a foundation, I'm going to object to that.
Judge: I'm taking the photograph just the way he looks, I want a full length photograph. . . . I want a full length photograph of him and a head shot.
Defense Counsel: I want the record to reflect that she's not ordering a photograph of the defendant's mouth with the gold caps in his mouth: the whole issue.
Tr. 445-47. The judge's failure to correct or comment on the defense lawyer's statement, for the record, that the photograph would not include the gold fronts — combined with her statement in the earlier colloquy that there was no foundation for the gold fronts to be admitted as evidence — indicates to this court that the photograph, if taken, would not have included the gold fronts.
II. Analysis
The trial court both should have admitted Moore's voir dire testimony at trial and should have permitted defense counsel to exhibit Moore, wearing the gold fronts, to the jury, as I will explain below.
Lyons also argues that the sum of the court's rulings resulted in unequal treatment of the defense as compared to the prosecution. I hold that, as Magistrate Judge Francis concluded, this claim was not properly raised in the state system. See 28 U.S.C. § 2254(b), (c); see also Picard v. Connor, 404 U.S. 270, 275, 92 S.Ct. 509, 512, 30 L.Ed.2d 438 (1971) (exhaustion of state remedies prior to habeas review required).
Lyons argues, as well, that the trial judge erred in that it did not compel Moore to take the stand and assert his Fifth Amendment privilege. Like the Magistrate Judge, I conclude that it is generally improper to call a witness to the stand solely to cause him to assert his claim of privilege. Bowles v. United States, 439 F.2d 536 (D.C. Cir. 1970) (en banc), cert. denied, 401 U.S. 995, 91 S.Ct. 1240, 28 L.Ed.2d 533 (1971). Moreover, under New York law, the trial judge's decision whether to put a witness on the stand for this reason is highly discretionary. See, e.g., People v. Thomas, 51 N.Y.2d 466, 434 N.Y.S.2d 941, 944, 415 N.E.2d 931, 933 (1980). The trial judge's decision not to do so here was not an abuse of discretion.
A. The Court's Decision Not to Compel Moore to Testify, Not to Compel Moore to Take the Stand, And Not to Allow Moore's Voir Dire Testimony to Be Admitted At Trial
Lyons's habeas counsel contends that, by answering some questions during the voir dire, Moore waived his privilege against answering the same questions at trial. However, during the voir dire, Moore stated several times that he would claim his Fifth Amendment privilege if asked to testify at trial. Moreover, Moore's willingness to answer questions on voir dire which he adamantly refused to answer at trial suggests that he did not understand that he might waive his Fifth Amendment privilege as to his trial testimony by his answers on voir dire. Thus, like Magistrate Judge Francis, I conclude that Moore did not, during the voir dire, knowingly waive his right to assert his Fifth Amendment privilege at trial.
Moore was therefore unavailable to testify at trial, due to a claim of privilege. As a consequence, his voir dire testimony could have been admitted at trial, under two clearly established New York hearsay exceptions, which apply where the declarant is unavailable, for (1) third-party declarations against penal interest, and (2) testimony given in a prior proceeding.
1. The Hearsay Exception for Third-Party Declarations Against Penal Interest
New York's Smith case contains one of the most recent pronouncements by New York courts on the hearsay exception for third-party declarations against penal interest. People v. Smith, 195 A.D.2d 112, 606 N.Y.S.2d 656 (1st Dep't 1994). In Smith, the defendant's brother had confessed to two people that he had committed the crime charged, but he asserted his Fifth Amendment privilege against testifying about these confessions at trial. Id. 606 N.Y.S.2d at 659-60. The New York appellate court held that the trial court should have admitted the brother's confession under the hearsay exception for third party declarations against penal interest. Id. 606 N.Y.S.2d at 662.
The Smith court listed the factors that, under New York law, qualify a third-party declaration against penal interest for admission under a hearsay exception:
[F]irst, the declarant must be unavailable as a witness at trial; second, when the statement was made the declarant must be aware that it was adverse to his penal interest; third, the declarant must have competent knowledge of the facts underlying the statement; and, fourth, and most important, supporting circumstances independent of the statement itself must be present to attest to its trustworthiness and reliability.Id. 606 N.Y.S.2d at 662 (citation omitted); cf. Fed.R.Evid. 804(a).
All these factors are satisfied here. First, Moore was unavailable at trial due to his proper assertion of his Fifth Amendment privilege. Second, Moore was aware that his voir dire statements concerning the gold fronts were adverse to his penal interest, for he refused to repeat them at trial, on the ground that they might incriminate him. (Indeed, the state asserts — in arguing that Moore's voir dire statements did not waive his Fifth Amendment privilege — that Moore's voir dire statements provided a necessary basis for the assertion of his Fifth Amendment right at trial). Third, Moore obviously had a basis for knowing whether he had worn the gold fronts on the day of the crime. Fourth, several circumstances attested to Moore's statements' reliability: (1) Moore wore the gold fronts during the voir dire, which established, at least, that Moore possessed a pair of custom-made gold fronts; (2) Lyons and Pantojas testified that Moore wore the gold fronts on the day of the crime; and (3) Moore's statements were made in court, before a judge, immediately after consultation with a lawyer, and were subjected to cross-examination by the same prosecutor who was prosecuting Lyons.
This last circumstance, that the statements were made in court, is especially significant. This special guarantee of reliability was not present in the Smith case. There, in contrast, the hearsay statements at issue were made prior to trial and out of court, to private parties. Id. 606 N.Y.S.2d at 661. Yet the Smith court still held the hearsay statements at issue there to be admissible.
2. The Hearsay Exception for Statements Made at a Prior Judicial Proceeding
Because Moore's voir dire statements were made in court, another, independent New York hearsay exception also applies here: the exception for testimony given by an unavailable witness at a prior proceeding, where that testimony is accompanied by indicia of reliability such as prior cross-examination. See, e.g., People v. Phan, 150 Misc.2d 435, 568 N.Y.S.2d 498 (Kings Cty. 1990) (grand jury testimony of unavailable witness admissible, under hearsay exception, to exculpate defendant at trial); People v. Muccia, 139 A.D.2d 838, 527 N.Y.S.2d 620 (3d Dep't 1988) (testimony of co-defendant at an earlier trial admissible, under hearsay exception, to inculpate defendant at trial); People v. Okafor, 130 Misc.2d 536, 495 N.Y.S.2d 895 (Bronx Co. 1985) (prior family court testimony of murder victim admissible, under hearsay exception, to inculpate murder defendant at trial).
In determining whether this hearsay exception applies here, I must determine whether a voir dire examination counts as a "prior proceeding" for the purposes of the hearsay rule. I hold that it does. Cf. United States v. MacCloskey, 682 F.2d 468, 477-79 (4th Cir. 1982) (citing Fed.R.Evid. 804(b)(1)) (in federal system, voir dire in the same trial counts as "prior proceeding" for purposes of the hearsay exception for prior testimony, so that exculpatory voir dire testimony of witness who invokes the Fifth Amendment is admissible under the exception).
Thus, I hold that the trial court should have accepted Moore's voir dire testimony as admissible at trial under the hearsay exceptions both for third-party statements against penal interest, and for testimony at a prior proceeding. In so holding, I note that counsel did not specifically request that the court accept the voir dire testimony as trial evidence. He requested, instead, that Moore either (1) be compelled to testify, (2) be compelled to take the stand and claim his Fifth Amendment privilege, or (3) be marked as an exhibit and displayed to the jury. Given that Moore's testimony was admissible under two independent, well-established New York hearsay exceptions, these multiple requests to the court should have been sufficient to apprise the court of counsel's desire to submit Moore's voir dire testimony to the jury at trial.
Moreover, had the court properly allowed Moore to be exhibited at trial wearing his gold fronts, then the question would naturally have arisen what jury instruction concerning the gold fronts should accompany the display of Moore. The prosecution likely would have sought a cautionary instruction, to the effect that Moore's wearing the fronts at trial did not mean that he wore them the day of the crime; the defense, in response, would likely have sought to introduce Moore's testimony on voir dire that the fronts were his own, and that they were the fronts he had worn on the day of the crime. The issue of what instruction should accompany the display would, then, have naturally and directly raised the issue, before the trial court, of the admissibility of Moore's voir dire testimony at trial.
Thus, although counsel omitted to request specifically that Moore's voir dire testimony be admitted at trial, I hold that the court's failure to admit Moore's voir dire testimony at trial was in error.
B. The Court's Ruling Excluding a Display of Moore as an Exhibit
At the time of trial, a New York court had explicitly approved the practice of submitting persons into evidence in cases of mistaken identity, such as this one. People v. Diaz, 111 Misc.2d 1083, 445 N.Y.S.2d 888 (Sup.Ct.N.Y.Co. 1981). During trial, defense counsel brought the Diaz decision, complete with a citation, to the judge's attention, but the judge refused to allow Moore's person to be entered into evidence as an exhibit. Magistrate Judge Francis concluded, and I agree, that this ruling was clearly erroneous.
Lyons's and Pantojas's testimony that Moore, not Lyons, was the shooter provided a foundation for Moore to be introduced as an exhibit. This testimony helped to establish Lyons's misidentification defense, and implied that the prosecution witnesses either were lying or had misidentified Lyons as the shooter. The exhibit of Moore would have been highly relevant, because it could have substantiated Lyons's misidentification defense. See Laureano v. Harris, 500 F. Supp. 668, 672 (S.D.N Y 1980) (where defense is based on mistaken identity, evidence tending to prove that contention is highly relevant).
If Moore in fact bore a facial and/or bodily resemblance to Lyons, the exhibition of Moore's person could have supported Lyons's misidentification defense, by demonstrating that the prosecution's identification witnesses might have mistaken Lyons for Moore. Display of Moore to the jury would also have served to demonstrate directly Moore's actual height and weight, which were disputed issues relevant to Lyons's misidentification defense. One prosecution witness maintained that the men were of significantly different weights and heights, whereas the defense witness, Pantojas, testified that there was no appreciable difference between the two men's statures.
Moreover, display of Moore wearing the gold fronts would have had further relevance, especially in combination with the admission of Moore's voir dire testimony. Together, the display of Moore wearing the gold fronts and the voir dire testimony would have served to undermine the testimony of the three prosecution witnesses who said that Moore had not been wearing gold fronts on the day of the crime. At the same time, this evidence would have bolstered Lyons's contrary testimony, that Moore had been wearing gold fronts on the day of the crime. Finally, this evidence would have increased the plausibility of Lyons's misidentification defense by suggesting that the prosecution's witnesses might have confused the two men, based on their both having worn gold fronts on the day of the shooting.
In summary, then, the court's decision to forbid the exhibition of Moore was a clearly erroneous decision to exclude highly relevant evidence.
C. Harmless Error
The state urges this court to find that the trial court's decision to exclude all evidence of Moore's gold fronts — either by display of Moore or inclusion of Moore's voir dire testimony — was, if error, harmless error falling short of the standard necessary for the "great writ" of habeas corpus to issue.
Harmlessness review is applicable here, because the error that occurred below was trial error that
`occur[red] during the presentation of the case to the jury,' and is amenable to harmless-error analysis because it `may . . . be quantitatively assessed in the context of other evidence presented in order to determine [the effect it had on the trial],'Brecht v. Abrahamson, 507 U.S. 619, 629, 113 S.Ct. 1710, 1717, 123 L.Ed.2d 353 (1993) (citing Arizona v. Fulminante, 499 U.S. 279, 280-81, 111 S.Ct. 1246, 1249, 113 L.Ed.2d 302 (1991)).
1. Whether The Applicable Standard Depends on Prior State Court Review
Before considering whether the error was harmless, I must first determine the proper "harmless error" standard to be applied on federal habeas review of a state conviction where, as here, the state appellate court failed to find error, and thus never applied the Chapman harmless error standard. See Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 828, 17 L.Ed.2d 705 (1967) (unless prosecution shows that error "was harmless beyond a reasonable doubt," habeas will issue).
The Second Circuit Court of Appeals has not directly considered this issue. The other federal circuit courts that have considered the question are divided with respect to the applicable standard, and with respect to the related question of the proper interpretation of the Supreme Court's plurality opinion in Brecht v. Abrahamson, 507 U.S. 619, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993).
I note, as an initial matter, that the history of the habeas petition at issue in Brecht was very different from the history of the petition at issue here. In Brecht, as Chief Justice Rehnquist's plurality opinion twice emphasized, by the time the habeas petition reached the Supreme Court, two state appellate courts had applied harmless error review to petitioner's conviction, and two federal courts had done so as well. 507 U.S. at 628-29, 636-37, 113 S.Ct. at 1716, 1721. Brecht held that, in these circumstances, the Supreme Court was not compelled to apply Chapman's demanding standard for the fourth time, but rather could apply the more deferential Kotteakos standard in reviewing petitioner's conviction on habeas. Brecht, 507 U.S. at 636-39, 113 S.Ct. at 1721-22 ( quoting Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 1253, 90 L.Ed. 1557 (1946)). This case, where no court has ever applied Chapman review, is leagues away from Brecht itself. The Brecht petitioner had three bites at the Chapman apple; this petitioner has not yet had one.
The Court of Appeals of Wisconsin found "constitutional error" and reversed the conviction, applying a Chapman-like standard, pursuant to which an error is harmless only if "there is no reasonable possibility that the error contributed to the conviction." Wisconsin v. Brecht, 138 Wis.2d 158, 405 N.W.2d 718, 722 (1987). At the next stage of review, the Wisconsin Supreme Court cited Chapman to find that the trial errors were "harmless beyond a reasonable doubt," and reinstated the conviction. Wisconsin v. Brecht, 143 Wis.2d 297, 421 N.W.2d 96 (1988) (citing Chapman, 386 U.S. at 24, 87 S.Ct. at 828). Next, the federal district court, applying Chapman, issued habeas. Brecht v. Abrahamson, 759 F. Supp. 500, 507-08 (W.D.Wis. 1991) (citing Chapman, 386 U.S. at 24, 87 S.Ct. at 828). On appeal, the Seventh Circuit applied the Kotteakos standard to overturn the grant of habeas, after considering whether Chapman applied.
In summary, then, before it reached the Supreme Court, the petition in Brecht was considered by four courts, three of which applied Chapman or a close variant thereon, and one of which applied Kotteakos.
In reviewing this case, the Appellate Division, First Department did not properly apply the Chapman standard, as explained infra at pp. 690-691. The New York Court of Appeals, as noted earlier, declined to hear petitioner's case.
The Eighth Circuit has held that, in a situation like this one, the federal court should do what the state court failed to do, and apply the demanding standard set out in Chapman. Orndorff v. Lockhart, 998 F.2d 1426, 1429-30 (8th Cir. 1993), cert. denied, ___ U.S. ___, 114 S.Ct. 1631, 128 L.Ed.2d 354 (1994). In contrast, the Fourth, Seventh and Eleventh Circuits have held that Brecht requires a federal court in this position to apply the more deferential Kotteakos standard, pursuant to which the court determines whether the error "`had substantial and injurious effect or influence in determining the jury's verdict.'" See Smith v. Dixon, 14 F.3d 956, 979-80 (4th Cir. 1994) (citations omitted), cert. denied, ___ U.S. ___, 115 S.Ct. 129, 130 L.Ed.2d 72 (1994); Tyson v. Trigg, 50 F.3d 436, 446-47 (7th Cir. 1995); Horsley v. Alabama, 45 F.3d 1486, 1492 n. 11 (11th Cir. 1995).
The Fourth Circuit did not comment specifically on its choice to apply the Kotteakos standard despite the absence of state court Chapman review.
Writing for the Seventh Circuit, Judge Posner argued that the Brecht Court must have considered, and meant to reach, cases where, as here, there was no state court Chapman review, for, he says, such cases are the norm:
[O]rdinarily . . . the state court will not have found any error and therefore will have had no occasion to apply any standard of harmless error. Brecht itself was a fluke in this regard; the state court had considered the issue of harmless error and applied the Chapman standard.Tyson, 50 F.3d at 446. Even if this observation about the "ordinary" case is empirically true, which I think it may well be, I do not believe the Brecht plurality wrote its decision on the assumption that the average habeas case is one in which the state court ignores or misses federal constitutional error that a federal court will later recognize — as happened here, and as did not happen in Brecht itself. The plurality opinion in Brecht suggests, on the contrary, that the Brecht plurality strongly believed in the state courts' ability to render conscientious, accurate federal constitutional interpretations. See Brecht, 507 U.S. at 636, 113 S.Ct. at 1721 ("State courts are fully qualified to identify constitutional error.").
Judge Posner, for the Seventh Circuit, also defended the application of Kotteakos on the ground that "[t]he reasons the Court gave in Brecht for adopting a less stringent rule [for habeas than that of Chapman] are independent of the rule applied in the state appellate process." Tyson, 50 F.3d at 446. The Eleventh Circuit reasoned, similarly, that Brecht's "rationale (advancing comity, federalism, finality, and the importance of the trial) . . . reaches to almost all federal habeas cases." Horsley, 45 F.3d 1486, 1492 n. 11. The Eighth Circuit, on the other hand, read Brecht's holding as more limited, describing it as "based largely on the notion that because the state courts can properly apply the Chapman harmless error standard on direct review, the federal habeas courts need only review those questions under the Kotteakos harmless error standard." Orndorff, 998 F.2d at 1430 (emphasis added). As a consequence, in a situation like the one here, where there was no state court Chapman review, the Eighth Circuit applied Chapman, not Kotteakos.
This court finds the Eighth Circuit's position — that Brecht does not extend to cases like this one — more persuasive. Not only did Brecht twice note that four courts had previously applied Chapman to the petition at issue there, but it also emphasized that
[s]tate courts are fully qualified to identify constitutional error and evaluate its prejudicial effect on the trial process under Chapman, and state courts often occupy a superior vantage point from which to evaluate the effect of trial error. . . . For these reasons, it scarcely seems logical to require federal habeas courts to engage in the identical approach to harmless-error review that Chapman requires state courts to engage in on direct review."507 U.S. at 636, 113 S.Ct. at 1721 (citations omitted) (emphasis added). That is, rather than simply invoking comity and federalism alone, Brecht explicitly noted state courts' ability separately (1) to identify constitutional error, and (2) to evaluate the effect of that error under Chapman. Id. Brecht's result, then, was premised largely on respect for the actual, conscientious review of harmlessness performed by the four prior courts — both federal and state — that applied Chapman.
Brecht's rule cannot be that a federal court must blindly apply Kotteakos, rather than Chapman, as soon as it receives a § 2254 petition, utterly without regard to the conviction's history of appellate consideration in the state court system. If that were Brecht's rule, a federal court would have to apply Kotteakos simply because it was technically conducting collateral review of a criminal conviction, even if a state were to eliminate all direct appeals of criminal convictions — as it is likely that it constitutionally could do, see Pennsylvania v. Finley, 481 U.S. 551, 555-56, 107 S.Ct. 1990, 1993-94, 95 L.Ed.2d 539 (1987), quoting Ross v. Moffitt, 417 U.S. 600, 610-11, 94 S.Ct. 2437, 2443-44, 41 L.Ed.2d 341 (1974); Jones v. Barnes, 463 U.S. 745, 751, 103 S.Ct. 3308, 3312-13, 77 L.Ed.2d 987 (1983); McKane v. Durston, 153 U.S. 684, 14 S.Ct. 913, 38 L.Ed. 867 (1894); Daniel J. Meltzer, Harmless Error And Constitutional Remedies, 61 U.Chi.L.Rev. 1, 2-4 nn. 9 10 (1994). In practice, a federal court reviewing a state decision in a state without appellate courts would likely find that its "collateral" review was, in effect, direct review, and apply Chapman. But this extreme hypothetical shows that a federal court must require, as a precondition to applying Kotteakos, that at least some meaningful direct review in the state system has occurred.
Furthermore, the comity concerns that the Brecht plurality opinion raises, 507 U.S. at ___, 113 S.Ct. at 1721, are far more strongly implicated if federal courts choose to repeat Chapman analyses even when state courts have already conducted them, than if federal courts choose to perform Chapman analyses that state courts have not performed. The former choice, not the latter, can be construed to convey federal courts' lack of respect for state courts' ability as federal constitutional interpreters, and hence to violate obligations of comity and principles of federalism. Put another way, federal courts should, under Brecht, respectfully defer to state courts' analyses of the harmlessness of constitutional error — but where no prior state court Chapman holding exists, there is nothing to which to defer.
2. Prior State Court Review of the Conviction at Issue Here
The history of the petition at issue here serves as an object lesson as to why it is important for federal courts to ensure, as the Brecht Court did, that a habeas petition has in fact undergone conscientious state court Chapman review, instead of simply applying Kotteakos routinely to every state habeas petition.
With regard to the Fifth Amendment issue, the New York appellate court simply commented, without further explanation, that
we do not find that the prospective witness's assertion out of the presence of the jury that he was wearing the same gold caps on his teeth that he was wearing on the date of the incident constituted a waiver of the [Fifth Amendment] privilege.Lyons, 579 N.Y.S.2d at 666. While this conclusion is correct, the court should have gone a step further in its analysis, to consider whether, even if Moore properly asserted his Fifth Amendment privilege, Moore's voir dire testimony should nevertheless have been admissible.
With regard to the issue of whether the trial court should have exhibited Moore to the jury, the appellate court stated that
there was no reversible error in the court's refusal to permit [Moore] to be questioned about gold coverings to his teeth before the jury or to permit the introduction of the gold coverings worn by Moore into evidence. The jury was informed in both the testimony of the defendant and that of Moore's former girlfriend [Ms. Pantojas] that [Moore] wore gold coverings on the date in question.Lyons, 579 N.Y.S.2d at 666. Strikingly, then, the appellate court did not directly consider the trial court's failure to accept Moore's person into evidence as an exhibit — even though, like the trial court, it had the benefit of defense counsel's citation of directly relevant precedent, People v. Diaz, supra.
Instead, the appellate court simply implied that any "gold fronts" evidence — whether Moore's testimony, or accepting the "gold fronts" as an exhibit, or accepting Moore's person as an exhibit — would have been cumulative with Lyons's and Pantojas's testimony that Moore had been wearing gold fronts on the day of the crime.
That conclusion was in error. First, physical evidence tending to show that Moore possessed custom-made gold fronts, especially in combination with Moore's voir dire testimony that he wore the fronts on the date in question, might well have meant far more to the jury than self-serving or possibly biased testimony to this effect.
Both witnesses were open to, and suffered, credibility attacks on the ground of their interest or bias. Lyons, of course, had an interest in exculpating himself. Ms. Pantojas, as the appellate court noted, was Moore's ex-girlfriend, but was currently Lyons' friend.
Second, the display of Moore also would have been admissible to show that Moore, wearing gold fronts, resembled Lyons, wearing gold fronts. The display, insofar as it served this purpose, would not have been cumulative of the "gold fronts" testimony. Pantojas did testify at trial that Lyons and Moore were similar in height and weight. Yet neither Lyons nor Pantojas testified directly as to whether there was a facial or bodily resemblance between Lyons and Moore.
Like its review of the trial court's evidentiary rulings, the New York appellate court's application of the "harmless error" standard was also erroneous. To begin, the court should have separately inquired (1) whether the trial court's evidentiary rulings were in error; and (2) whether, if so, that error was harmless, or warranted reversal and the grant of a new trial. Instead, the court conflated the two questions, asking only whether there was "reversible error." Lyons, 579 N.Y.S.2d at 666.
This conflation may have disadvantaged petitioner. In considering in one step whether a conjunction of factors — error plus harm — is present, the deciding court need not focus clearly on either factor. Thus, the court is never forced to decide clearly, as a separate matter, whether error occurred.
In addition, although it was evaluating alleged federal constitutional error, the New York court neither cited, nor implicitly followed, either the harmless error standard of Chapman or that of Brecht. Both Chapman and Brecht require attention to whether the jury's verdict would have been swayed by introduction of the evidence erroneously excluded. See Brecht, 507 U.S. at 641-42, 113 S.Ct. at 1724 (Stevens, J., concurring), quoting Kotteakos v. United States, 328 U.S. 750, 764, 66 S.Ct. 1239, 1247-48, 90 L.Ed. 1557 (1946) (crucial focus in applying Kotteakos or other "harmless error" standard is on "`the impact of the thing done wrong on the minds of [the jurors], not on one's own.'"). Yet, instead, in concluding that there was no reversible error, the appellate court offered its own opinion that Lyons's guilt had been proven beyond a reasonable doubt by the testimony of the three prosecution witnesses, rather than focusing properly on what the jury could have found.
After the New York appellate court had rendered its decision, the New York Court of Appeals declined jurisdiction. It did so despite a clear conflict between the trial court's holding and a New York precedent, People v. Diaz, supra, and despite the First Department's clearly erroneous application of the federal constitutional law of "harmless error."
3. Application of the Chapman Standard
Because petitioner did not receive a thorough and correct review, applying Chapman, of his conviction from any New York appellate court, I now apply Chapman to the facts of this case. Chapman requires that "the beneficiary of a constitutional error" — here, the prosecution — "prove beyond a reasonable doubt that the error complained of did not contribute to the verdict obtained." Chapman, 386 U.S. at 24, 87 S.Ct. at 828. The state contends, for several reasons, that it has carried this burden.
a. The Argument that Moore Would Have Asserted a Fifth Amendment Privilege Against Being Displayed to the Jury Wearing His Gold Fronts
The state contends that the court's failure to display Moore to the jury as an exhibit was harmless because, had the trial court directed Moore to appear before the jury wearing his gold fronts, Moore would have refused, citing his Fifth Amendment privilege and contending that his wearing the fronts would constitute an admission that the fronts were his. The state, which bears the burden of proof under Chapman, offers no precedent to support this conclusion, and I reject it.
Indeed, the record indicates that Moore probably would not have refused, on Fifth Amendment grounds, to wear his gold fronts before the jury. Moore wore the gold fronts to the courthouse, while he was in the custody of federal agents and during his voir dire before the trial judge. Had Moore been subpoenaed to appear in court wearing his gold fronts, he might have been able to assert a Fifth Amendment privilege against the production of incriminating evidence. See, e.g., Curcio v. United States, 354 U.S. 118, 77 S.Ct. 1145, 1 L.Ed.2d 1225 (1957) (subpoena demanding that accused produce his own records is equivalent, for Fifth Amendment purposes, to requirement that he take the stand and admit records' genuineness). Notwithstanding this, Moore wore the gold fronts to court voluntarily. Moreover, on voir dire, although Moore claimed his Fifth Amendment privilege against testifying at trial to anything that happened on the day of the crime, he also stated that he was willing to turn over his fronts to Lyons for use during the trial.
For these reasons, I hold that the state has not conclusively established, for the purposes of "harmless error" analysis, that Moore would have asserted his Fifth Amendment privilege against being displayed to the jury wearing his gold fronts.
b. The Court's Offer to Photograph Moore
The state argues, in the alternative, that the court's erroneous refusal to exhibit Moore's person, with the gold fronts, to the jury was harmless because the court's offer to submit a head shot and a full body photograph of Moore, without the gold fronts, into evidence was an adequate substitute. In my view, a photograph of a person, even a full body photograph, is not generally an adequate substitute for the display of that person as an exhibit — where that display is clearly legally permissible, and where the person is available and, indeed, in the courtroom on the day of trial. Moreover, a photograph excluding the fronts obviously would not have adequately substituted for a display including the fronts.
c. The Weight of the Evidence
The state contends that the court's error was harmless due to the sheer weight of the evidence against Lyons. It emphasizes, in support of this view, that three prosecution witnesses — Ferrera, Poole and Quiles — confidently and unequivocally identified Lyons as the shooter. The state adds that Ferrera's and Quiles's identifications are especially reliable because both witnesses knew both Lyons and Moore prior to the shooting.
The state is correct that the testimony of the three prosecution witnesses, who corroborated each other's stories and uniformly inculpated Lyons, must have carried considerable weight with the jury. But in my view, this testimony did not carry so much weight, or bear so many of the hallmarks of credibility, as to establish, beyond a reasonable doubt, that the trial court's erroneous rulings did not contribute to the verdict obtained.
At trial, defense counsel raised colorable issues as to the accuracy and/or bias of the testimony of each of these three prosecution witnesses:
(1) Quiles, the victim, testified that he saw Lyons pull out the gun before he, Quiles, began running. However, defense counsel's cross-examination revealed that Quiles, who was fleeing, had his back turned when he was actually shot, and that Quiles told the police, in two initial interviews, that he did not know who had shot him.
(2) Defense counsel's cross-examination of Ferrera revealed that she had observed the crowded scene of the shooting from the window of an apartment across the street. It also revealed that Ferrera knew Moore and was his neighbor. "My family knows his family," she testified. "I've known him since he was a little kid." Tr. 130. Ferrera knew Lyons, as well, but a jury could reasonably have thought from her testimony that she had a much stronger allegiance to Moore. Ferrera was also effectively impeached by defense counsel as to her recollection of details of the crime — such as whether, as several other witnesses testified, there was a woman present in the disputatious crowd and what were the names of the persons present.
(3) Defense counsel's cross-examination demonstrated that Poole did not come forward until over a year after the shooting and a few days before the trial.
(4) Defense counsel's direct and cross-examinations of the trial witnesses established that the shooting took place in the midst of a frightened and fleeing crowd. Poole and Quiles were in the midst of that crowd; Ferrera was looking into the crowd from an apartment across the street. To make an identification, each of the three prosecution witnesses had to ascertain which of two young, African-American men standing close together, wearing black leather jackets, was the shooter. Moreover, the testimony of Pantojas, for the defense, suggested that the men were similar in height and weight, although this was disputed by prosecution witnesses.
Defense counsel could have attacked each of the three witnesses' testimony even more effectively, had he been able to display Moore wearing his gold fronts and to read to the jury Moore's voir dire testimony that Moore wore the fronts on the day of the crime.
Ferrera and Quiles testified confidently and unequivocally — yet inaccurately, if we believe Moore himself — that Moore did not wear gold fronts the day of the crime and, indeed, had never worn them. The introduction of Moore's person, with the gold fronts, and of Moore's voir dire testimony, would have allowed defense counsel to impeach their testimony further, on the basis of either inability to observe the scene, or lack of veracity.
First, defense counsel could have suggested — and a reasonable jury could have believed — that one or both of the two witnesses had confused the two men. Of course, these two witnesses, who knew both Lyons and Moore prior to the shooting, would not, on a day to day basis, have distinguished Moore from Lyons solely by the gold fronts. Yet in the context of the chaotic crime scene, and in memory, they might have relied more heavily on the gold fronts. If these witnesses could make a mistake in observing or recalling whether Moore wore gold fronts on the day of the crime, a jury might reason, they could also make a mistake about which of the men shot Quiles.
Second, defense counsel could have suggested — and a reasonable jury could have believed — that one or both of the witnesses had lied when they said that Moore was not wearing gold fronts, in order to implicate Lyons more clearly and to decisively defeat Lyons's misidentification defense. A jury might conclude, further, that this lie, or these lies, cast doubt on the veracity of the rest of the testimony.
Turning to the third witness, Poole, it is important to note that, although Poole did not come forward to identify Lyons until almost a year after the crime, the jury might well have placed special reliance on Poole's testimony. Unlike Quiles and Ferrera, Poole had no personal ties to Lyons, Moore, or any witness. Unlike Quiles, Poole was able to observe the very moment of the shooting. Unlike Ferrera, Poole observed the shooting from close by. For these reasons, the jury could have thought that Poole would be especially objective, truthful, and accurate in his observations.
Yet, as noted, Poole did not identify Lyons to the police as the shooter until almost a year after the crime and, as a stranger to both Moore and Lyons, he relied heavily on his memory of the gold fronts in making his identification. Like the other witnesses, Poole testified that only one man at the crime scene, the shooter, was wearing gold fronts. Poole also testified that the gold fronts were "the most descriptive thing" he remembered about the shooter and "one of the main things that [Poole] picked out about him." Tr. 211-12, 216. Poole's important identification testimony, then, would have been seriously impeached by Moore's voir dire testimony, and also by the introduction of Moore's person if Moore, wearing gold fronts, had resembled Lyons, wearing gold fronts.
For all these reasons, a jury afforded the opportunity to hear Moore's voir dire testimony, to see Moore with his gold fronts on, and to compare Moore's appearance to Lyons's, might have concluded that the three prosecution witnesses were mistaken or biased in their identifications of which man was the shooter.
Moreover, in emphasizing the weight of the three prosecution witnesses' inculpatory testimony, the state ignores the fact that evidence of Moore's gold fronts also bolstered the exculpatory testimony of Lyons and Pantojas. Lyons testified that he and Moore — two men in a small crowd — each sported four gold fronts. Even a juror who found Lyons credible, on the basis of his demeanor, might have stopped believing Lyons as soon as he said this. Absent evidence of Moore's gold fronts, a reasonable jury would likely have concluded that Lyons's testimony that he and Moore both wore gold fronts was merely wishful thinking; that Lyons's misidentification defense was a fabrication; and that, given this clear fabrication, the rest of Lyons's self-exculpatory testimony was likely to be untrue.
In addition, Pantojas, a defense witness for Lyons, testified that only Moore wore gold fronts. Evidence that Moore wore gold fronts would have at least improved the jury's assessment of the credibility of Pantojas's testimony. However, the jury would still have had to keep in mind that Pantojas testified, contrary to the testimony of all the prosecution witnesses and of Lyons himself, that Lyons was not wearing gold fronts.
If the jury reached the conclusion, as I think they must have, that Lyons's "gold fronts" story was a mere fabrication, that conclusion would have been heavily reinforced by the events at trial. The jurors saw Lyons pointed out repeatedly by prosecution witnesses, with the attendant drama of in-court identification. They saw Lyons's gold fronts entered into evidence, handed to witnesses, and identified by witnesses as his. They heard the prosecution refer several times to the "gold fronts" issue in summation. Yet they never saw Moore at all, or heard any of his voir dire testimony.
Conversely, a reasonable jury that saw Moore wearing his gold fronts, and heard Moore's voir dire testimony, would have accorded far greater credibility to all of Lyons's self-exculpatory testimony. In particular, the jury would have taken Lyons's misidentification defense far more seriously than it had done when that defense was based solely on Lyons's self-serving testimony and the testimony of Pantojas, who, as I noted earlier, had been attacked on cross-examination for her bias, because she was Moore's ex-girlfriend and Lyons's friend.
All of the trial witnesses agreed that Moore made a remark to Lyons, that Lyons then went into Moore's family's apartment to retrieve a gun, and that Lyons subsequently returned to the crime scene. A jury might have reasonably concluded, from this testimony, that Moore requested that Lyons bring him the gun, and that Lyons did exactly that.
Furthermore, a jury with some good reason to believe Lyons's testimony — the good reason that Moore's voir dire testimony and a display of Moore could have given — might also have concluded that Ferrera had confused the two men in the chaos that day, or lied to protect her friend Moore; that Quiles had seen Lyons with a gun in his hand, and failed to see Moore actually shoot him; and that Poole, thinking back a year after the shooting, had identified, not Lyons, but "the man with the gold fronts."
These conclusions, taken together, would have provided a defense theory justifying a confident "not guilty" verdict. Yet the defense need not provide a complete theory of the case; it need only raise a reasonable doubt in jurors' minds. Even a jury that did not believe this complete defense theory might still have found the prosecution and defense testimony equally credible, yet in irreconcilable conflict, and thus have been unable to find Lyons guilty beyond a reasonable doubt.
Justice Stevens, in his Brecht concurrence, admonished courts engaging in "harmless error" review to remember that "[t]he crucial thing is the impact of the thing done wrong on the minds of other men [the jury members], not on one's own, in the total setting." 507 U.S. at 643, 113 S.Ct. at 1724. With this comment in mind, I note that the jury deliberated overnight before reaching its verdict. During their deliberations, the jury members asked to see the testimony of all three prosecution identification witnesses: Poole, Ferrera and Quiles. The jury also requested a readback of the testimony concerning on which floor Ferrera's apartment was located. These requests suggest that the jury was concerned with the issue of the prosecution witnesses' capacity to observe accurately the crime scene and to identify the shooter.
Because it is, of course, impossible to read jurors' minds, I do not place undue weight on the jury's requests to have testimony read back. Indeed, even absent these requests, I would reach the same result under Chapman. Yet, in my view, the jury's requests that testimony be read back afford at least some insight into the jurors' thought processes and concerns.
For these reasons, in the context of the transcript viewed as a whole, I hold that the prosecution has not proven beyond a reasonable doubt, as Chapman requires, that the judge's evidentiary error did not contribute to the verdict the jury reached. Cf. Smith, 606 N.Y.S.2d at 662 (applying Chapman to reverse conviction on ground that decisions to (1) exclude exculpatory hearsay testimony and (2) deprive defense of ability to exhibit man whom witness identified as perpetrator together denied defendant fair trial); MacCloskey, 682 F.2d at 479 (reversing conviction where court excluded testimony vital to defense and fitting within hearsay exception). As a consequence, I hereby grant Lyons's petition and issue the writ.
I note that my analysis under Chapman would have been considerably more difficult if I had held that the court should have displayed Moore to the jury, but that, due to defense waiver, the trial court need not have admitted into evidence Moore's voir dire testimony that he had been wearing gold fronts the day of the crime, or given an instruction to this effect.
Even absent the voir dire testimony, the display of Moore still would have been relevant. See People v. Scarola, 71 N Y2d 769, 530 N.Y.S.2d 83, 86, 525 N.E.2d 728, 731, (1988) (New York's test for relevance is whether evidence has any tendency in reason to prove existence of material fact). To lay a foundation for the display, Lyons and Pantojas could have been asked whether Moore looked at trial as he had looked on the day of the crime, just as can be done with a photograph of a crime scene. See, e.g., People v. Wrigglesworth, 204 A.D.2d 758, 611 N.Y.S.2d 678, 680 (1994) (proper foundation for photos of crime scene is testimony showing they were a fair and accurate representation of scene on day of crime).
Moreover, the court could have instructed the jury — based on its taking judicial notice of Moore's having worn the fronts that day — that the fronts belonged to Moore. From Lyons's and Pantojas's testimony, and from Moore's wearing his own gold fronts on the day of trial, the jury could have permissibly inferred that Moore might have been wearing gold fronts on the day of the crime.
Moving to the Chapman analysis, even the conclusion that Moore could and might have been wearing fronts on the day of the crime would have significantly bolstered Lyons's otherwise highly implausible testimony that two men in a small crowd each wore exactly four gold fronts. It would also have bolstered the exculpatory testimony of Pantojas, who testified, not just that Moore had fronts, but also that "Every time I saw [Moore] he had them in." Tr. 409.
Evidence of Moore's mere possession of a custom-made set of fronts would also have served to undermine the testimony of Quiles and of Ferrera, Moore's neighbor, who both stated that, to their knowledge, Moore had never worn gold fronts. Thus, even had the trial court's error simply been in preventing the display of Moore, I would still apply Chapman to grant the writ, though the question would be significantly closer.
4. Application of the Kotteakos Standard
The Kotteakos standard, as I have noted, asks whether the trial court's error had a "substantial and injurious effect or influence in determining the jury's verdict." 328 U.S. at 776, 66 S.Ct. at 1253. If I were to apply that standard to these facts, I would hold, as I have held in applying Chapman, that the trial court's error here, in failing to mark Moore as an exhibit and display him to the jury, was not harmless, and I would therefore issue the writ.
As Justice Stevens emphasized in his Brecht concurrence, the Kotteakos standard — while somewhat less demanding than the Chapman standard — nevertheless requires careful judgment, and de novo review of the trial record, by the reviewing court. See 507 U.S. at 642-43, 113 S.Ct. at 1724-25 (Stevens, J., concurring). Indeed, Justice Stevens indicated that the very stringency of Kotteakos was a reason that federal courts need not always apply Chapman on collateral review:
Given the critical importance of the faculty of judgment in administering either standard . . . [the difference between them] is less significant than it might seem. . . . In the end, the way we phrase the governing standard is far less important than the quality of the judgment with which it is applied.507 U.S. at 643, 113 S.Ct. at 1725 (Stevens, J., concurring).
Because Justice Stevens provided the critical fifth vote in favor of the result in Brecht, and because, unlike the plurality opinion, his concurrence explicitly emphasizes that Brecht's substitution of Kotteakos for Chapman is a relatively small departure from previous law, his narrower reading of the departure Brecht effects should govern interpretation of Kotteakos as applied in state habeas cases. The very stringency of Kotteakos, as interpreted by Justice Stevens, may have been the reason Justice Stevens concurred in Brecht's holding in the first place. Justice Stevens also joined the O'Neal opinion, discussed below, which emphasizes that Kotteakos is demanding. O'Neal v. McAninch, ___ U.S. ___, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995).
The recent decision of the Supreme Court in O'Neal also indicates that the Kotteakos standard is demanding. O'Neal v. McAninch, ___ U.S. ___, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995). O'Neal holds that, under Kotteakos and Brecht, a federal habeas judge who reviews a state conviction and who, based on the trial record, retains "grave doubt as to the harmlessness of an error," should decide to issue the writ.
I have indicated, in the course of my Chapman analysis above, why I believe the jury members in this case might well have been influenced in their decisionmaking by the judge's erroneous decision not to exhibit Moore to them, or allow them to hear Moore's voir dire testimony. Thus, my de novo review of the record leaves me, at the very least, in grave doubt as to whether, under Kotteakos, the trial court's error had a "substantial and injurious effect or influence in determining the jury's verdict" in this case. Kotteakos, 328 U.S. at 776, 66 S.Ct. at 1253.
Conclusion
Petitioner's application for a writ of habeas corpus is granted with respect to his conviction for attempted murder, and the case is remanded for a new trial on the attempted murder charge.
SO ORDERED.