Full opinion text
ORDER GRANTING PETITIONER’S MOTION FOR SUMMARY JUDGMENT AND PETITION FOR WRIT OF HABEAS CORPUS WHALEY, District Judge. The issue before the Court is whether Petitioner’s murder conviction was the result of a fair trial. After a full consideration of the parties’ extensive written and oral submissions on the questions addressed in this order, the Court concludes constitutional errors at Petitioner’s trial undermined the legitimacy of his conviction. Thus, Petitioner’s request for a writ of habeas corpus shall be granted, and the State of Washington must either grant Petitioner a new trial or forfeit the right to hold him in custody. The constitutional errors that infected Petitioner’s trial involved the federal constitution’s Confrontation Clause, which -requires that “[i]n all criminal prosecutions, the accused shall enjoy the right ... to be confronted with the witnesses against him.” U.S. Const, amend. VI. This right of confrontation is a fundamental precept of our legal tradition, with roots dating back to ancient Rome. As a protection against the once common practice of allowing individuals to be convicted on the basis of anonymous or ex parte affidavits, the ability to face one’s accuser in open court has long been recognized as an important safeguard against the mistaken conviction of innocent persons. Ensuring that persons accused of crimes are afforded this right serves the truth-finding function of the jury trial process in a number of ways, including: 1) ensuring that testimony is given under an oath that impresses on witnesses the importance of their testimony and that they are subject to penalties for perjury should they give false testimony; 2) forcing witnesses to face the accused in person when they testify, making it more difficult to lie “to their face” and reducing the risk that the wrong person is convicted; 3) subjecting testimony to the truth-exposing rigors of cross-examination; and 4) allowing the jury to personally observe the witness’s demeanor and other aspects of the testimony that may bear on its credibility. All of these benefits are lost, and the risk of erroneous conviction is increased, when witnesses are not required to testify at trial. It is true, of course, that a defendant is entitled only to a fair trial, not one that is perfect. In this case, however, Petitioner was unable to confront not one, but three separate individuals who provided damaging testimony against him. Two of these witnesses had been charged with the very same crime as Petitioner and, thus, had a strong incentive to shift the blame for that murder to Petitioner. The third individual, the father of one of Petitioner’s codefendants, had a similar incentive and provided testimony that helped to undermine Petitioner’s claim that he was not present when the murder occurred. Finally, this Court is not the first to conclude that Petitioner’s right to confront his accusers was violated. On Petitioner’s initial appeal, the Supreme Court of Washington determined that the admission of the testimony of Petitioner’s codefendants violated the Confrontation Clause. Washington’s supreme court also let stand a similar court of appeals ruling regarding the testimony of the father of Petitioner’s codefendant. While this decision agrees that Petitioner’s right of confrontation was violated, it comes to a different conclusion than that reached by the Washington courts on the question of whether these constitutional violations affected the jury’s verdict. Because the Court concludes this testimony had a substantial and injurious effect on the jury’s verdict, it also concludes that Petitioner’s conviction was the product of an unfair trial. Thus, his petition must be granted and he must be awarded a writ of habeas corpus. I. Background The background of this case is described in detail in the Court’s first summary judgment order. See Ct. Rec. 91 (entered November 26, 1996). Rather than repeat that discussion, the factual and procedural section (Section I) of that order is incorporated herein, as modified by the discussion below. The Court’s initial summary judgment order granted judgment to Respondent on 14 of the 18 grounds for relief presented by Petitioner. This order addresses the four remaining grounds for relief, which Petitioner and Respondent both have requested be decided as a matter of law based on the record before the Court. Also addressed in this order are Petitioner’s contention that the Court’s prior summary judgment order erroneously dismissed his other 14 grounds for relief and his request to amend his petition to add an additional claim. Because the Court’s decision regarding Petitioner’s remaining four grounds for relief is dispositive of this case, those claims are addressed first. II. Summary Judgment Four of Petitioner’s grounds for relief remain pending at this stage of this ease. As alleged by Petitioner, those four grounds are: 1. The admission of codefendants’ taped interrogation statements after codefendants (McKee; Flota) invoked their Fifth Amendment rights. 2. The admission of [the] videotaped deposition of George Flota; 5. Trial court erred in not admitting relevant evidence [of] the crime, a blue blanket sequestered from the crime scene ____ 18. Accumulation of the errors — violation of due process under state and federal constitutions deprived petitioner of a fair trial. Petition Under 28 U.S.C. § 2254 for Writ of Habeas Corpus by a Person in State Custody (Ct.Ree.2). For the reasons described more fully below, the Court concludes that Petitioner’s first and eighteenth grounds for relief are meritorious. Of the other two grounds for relief, the failure to admit the blue blanket offered by Petitioner at trial did not involve constitutional error, while the admission of George Flota’s videotaped testimony involved constitutional error that does not, on its own, warrant granting Petitioner a writ of habeas corpus. A. Codefendants Mckee’s and Flota’s Taped Statements Petitioner contends that the admission at his trial of three taped statements given to law enforcement authorities by Jerry McKee and Jeffrey Flota violated the Confrontation Clause. Violation of the right of confrontation is considered to be “trial error” since it “occurred during the presentation of the case to the jury and may [] therefore be quantitatively assessed in the context of other evidence presented to determine whether its admission was harmless.” Arizona v. Fulminante, 499 U.S. 279, 307-08, 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991). Reversal for trial error is not automatic. Instead, if a court finds that such a violation has occurred, the court must also determine whether the error was harmless. Id., 499 U.S. at 308; see also Lee v. Illinois, 476 U.S. 530, 547, 106 S.Ct. 2056, 90 L.Ed.2d 514 (1986) (remanding case for determination of whether Confrontation Clause error was harmless). To prevail, therefore, Petitioner must show that admission of these taped statements violated the Confrontation Clause, and their admission cannot have been harmless when considered in the context of his entire trial. These conditions have been met in this case. In reaching this conclusion, the Court joins the Supreme Court of Washington, which determined on direct review that McKee’s and Flota’s statements were not sufficiently reliable to render their admission at trial permissible despite the denial of Petitioner’s right of confrontation. See State v. Whelchel, 115 Wash.2d 708, 801 P.2d 948 (1990). The Court parts company with that court on the question of whether these violations were harmless and concludes that these statements had a substantial and injurious effect on the jury’s verdict. The discussion below details the analysis underlying these conclusions, following a brief summary of the facts relevant to this ground for relief. 1. Factual and procedural background During the early morning hours of September 27, 1986, Emargo McKee (Ms. McKee) was murdered near the shore of Moses Lake in Grant County, Washington. Her body was hidden in nearby brush and was not discovered until 18 days later, on October 15,1986. The next day, October 16, 1986, the victim’s husband, Jeny McKee (“McKee”), came to the Moses Lake Police Department to file a missing persons report on his wife. McKee was interviewed in the investigator’s office of the Moses Lake Police Department by a Moses Lake police officer and a Grant County deputy sheriff. During that interview, McKee admitted that his wife was murdered and implicated the five people who were eventually tried for crimes relating to the murder: himself, Petitioner, Jeffrey Flo-ta (“Flota”), Beth Massey (“Massey”), and Nancy Hughes (“Hughes”). Later that evening, Flota was arrested and made a post-arrest statement at the Moses Lake Police Department to the same Grant County deputy sheriff and an additional investigator from that office. On the next day, October 17, 1986, McKee made an additional post-arrest statement in Grant County jail to the Grant County deputy sheriff and another investigator from that office. Extensive portions of the McKee and Flota statements were audiotaped. See Respondent’s Exhibits 26-28 (transcribed statements of McKee and Flota). Both McKee and Flota admitted in their statements that they were present at Ms. McKee’s murder and that they knew beforehand that she was going to be killed. Each also contended that Petitioner was the moving force behind the murder and that he inflicted the injuries that killed Ms. McKee. McKee and Flota both refused to testify at Petitioner’s trial, exercising their Fifth Amendment right not to incriminate themselves. A preliminary hearing was held to determine whether admission at trial of McKee’s and Flota’s taped statements would violate the Confrontation Clause. After hearing argument on this issue, the trial court ruled that the statements were sufficiently reliable to fall within the Confrontation Clause exception for trustworthy statements made by individuals who are legally unavailable to appear at trial. See Respondent’s Exhibit Twenty-Four (Verbatim Report of the Proceedings; hereafter “Trial Record”) at 205-221, 244-253. Pursuant to this ruling, the audiotaped statements were played for the jury at trial. Id. at 1161-62, 1167-68. Prior to the commencement of the jury’s deliberations, the trial court instructed the jury to subject the testimony of alleged accomplices to “careful examination.” Respondent’s Exhibit 29 (Instruction No. 6). The court also told the jurors that they should act on such testimony with “great caution” and should not find Petitioner guilty based on such testimony alone unless they were “satisfied beyond a reasonable doubt of its truth.” Id. 2. Confrontation Clause Violation a. Legal Standard There is ho dispute that the Confrontation Clause’s requirement that “[i]n all criminal prosecutions, the accused shall enjoy the right ... to be confronted with the witnesses against him” was not expressly complied with in this case. McKee’s and Flota’s taped statements were played for the jury at Petitioner’s trial, and Petitioner had no opportunity to confront them at any point during his trial or its preliminary proceedings because they exercised their Fifth Amendment right not to be compelled to incriminate themselves. Strict compliance with the confrontation requirement is not always required, however. Because the underlying reason for the right of confrontation is to help ensure the accuracy of incriminating trial testimony, statements made by individuals who are legally unavailable to testify at trial may be admitted where those statements are sufficiently reliable that confrontation at trial would be unlikely to enhance the testimony’s accuracy. Ohio v. Roberts,. 448 U.S. 56, 65-66, 100 S.Ct. 2531, 65 L.Ed.2d 597 (1980). In this ease, McKee and Flota both were legally unavailable because they exercised their Fifth Amendment right not to be compelled to incriminate themselves. See California v. Green, 399 U.S. 149, 168 n. 17, 90 S.Ct. 1930, 26 L.Ed.2d 489 (1970) (defendant who invokes the Fifth Amendment right against compelled self-incrimination is unavailable for purposes of the Confrontation Clause). Thus, the question of whether McKee’s and Flota’s statements were properly admitted hinges on their reliability. There are two ways to meet the reliability requirement of the Confrontation Clause. The first involves statements that correspond to “firmly rooted hearsay exceptions.” Idaho v. Wright, 497 U.S. 805, 815, 110 S.Ct. 3139, 111 L.Ed.2d 638 (1990). This approach has not been advanced by Respondent, and it remains an open question in the Ninth Circuit whether the hearsay exception under which these statements were admitted — declarations against penal interest — constitutes a firmly rooted exception. United States v. Nazemian, 948 F.2d 522, 530 (9th Cir.1991); see also Williamson v. United States, 512 U.S. 594, 605, 114 S.Ct. 2431, 129 L.Ed.2d 476 (1994) (declining to reach issue). Instead, Respondent contends McKee’s and Flota’s statements satisfy the second means of meeting the reliability requirement because they are possessed of “particularized guarantees of trustworthiness.” Wright, 497 U.S. at 815. The “particularized guarantees of trustworthiness” inquiry is a totality of the circumstances test that balances factors relevant to the statements’ trustworthiness. Id., 497 U.S. at 819. This inquiry does not begin with a clean slate when the statements at issue are self-inculpatory statements by co-defendants that also inculpate the defendant at issue. Instead, such statements are “presumptively unreliable.” Lee v. Illinois, 476 U.S. 530, 539, 106 S.Ct. 2056, 90 L.Ed.2d 514 (1986); see also Bruton v. United States, 391 U.S. 123, 135-36, 88 S.Ct. 1620, 20 L.Ed.2d 476 (1968). Thus, at Petitioner’s trial, the State was required to demonstrate that McKee’s and Flota’s statements regarding Petitioner’s role in the murder of Ms. McKee were possessed of particularized guarantees of trustworthiness that were strong enough to overcome this presumption of unreliability. Lee, 476 U.S. at 539. In considering this question, this Court acts de novo, giving no deference to the legal conclusions of the state courts that have preceded it. Webb v. Lewis, 44 F.3d 1387, 1392 (9th Cir.1994). Although the trustworthiness inquiry is a totality of the circumstances test, there are limits on the circumstances that a court may consider. As the Supreme Court recently explained: We agree that “particularized guarantees of trustworthiness” must be shown from the totality of the circumstances, but we think the relevant circumstances include only those that surround the making of the statement and that render the declarant particularly worthy of belief. Wright, 497 U.S. at 819. Thus, a court cannot look to factors unrelated to the nature of the declarant and to the circumstances in which the statement was made, factors such as whether extrinsic evidence relating to the crime corroborates the accuracy of the challenged statement. Id. at 823 (“we think the presence of corroborating evidence more appropriately indicates that any error in admitting the statement might be harmless, rather than that any basis exists for presuming the declarant to be trustworthy”); Webb, 44 F.3d at 1392 (“Wright forbids recourse to other evidence to corroborate the trustworthiness of the hearsay offered”). Within this framework, the factors relevant to a statement’s trustworthiness are context dependent; there is no definitive, exhaustive list of factors that applies across the wide variety of contexts to which this test applies. In the context of declarations against penal interest, however, the United States Court of Appeals for the Ninth Circuit has described a non-exclusive list of questions that generally are relevant to assessing the trustworthiness of such statements. These include: 1. Was the statement made voluntarily? 2. Was the statement made contemporaneously with the occurrence of the events it references? 3. Did the declarant admit that he committed acts contrary to his penal interest or likely to bring him into disrepute? 5. Did the declarant have personal knowledge of the matters addressed in the statement? 6. Was the statement uttered spontaneously? 7. Was the person to whom the statement was made someone to whom the declarant would likely speak truthfully? United States v. Layton, 855 F.2d 1388, 1405 (9th Cir.1988) (citations and reference to extrinsic corroborating evidence omitted), overruled on different grounds, People of the Territory of Guam v. Ignacio, 10 F.3d 608 612 (9th Cir.1993). Other Supreme Court and Ninth Circuit cases indicate that relevant factors may also include: 1) whether the individual had a motive to lie or shift blame; Wright, 497 U.S. at 821-22; 2) whether the statement was genuinely self-inculpatory, Williamson, 512 U.S. at 605; 3) the extent to which the statement was subjected to contemporaneous cross-examination or its equivalent; Lee, 476 U.S. at 544; and 4) the declarant’s mental state at the time the statement was made, Wright, 497 U.S. at 821. Because neither the parties nor the Court have identified additional factors relevant to this particular case, this general list of factors has been adopted as a framework for the trustworthiness inquiry in this case. b. Trustworthiness of McKee’s and Flota’s Statements Viewed through the prism of these factors, there is little about McKee’s and Flota’s statements that indicates they were trustworthy enough to rebut their presumed unreliability. At most, only two factors support a conclusion that these statements were trustworthy. First, McKee and Flota each admitted that they committed acts that were contrary to their penal interest. Each stated he had prior knowledge that Ms. McKee was going to be murdered before the group went down to the lake. Each also stated that they were either physically or audibly present when the injuries that killed Ms. McKee were inflicted, and that they either inflicted some of these injuries or did nothing to stop them. Each also admitted to helping to eliminate evidence of the crime. Second, it is undisputed that by virtue of their admitted presence while the murder occurred, McKee and Flota had personal knowledge of the key events they described in their statements. The relatively limited period of time that had passed since the crime makes it unlikely that they were unable to remember the type of details contained in their statements. In contrast, none of the remaining factors supports a finding that these statements were trustworthy. First, it is questionable whether these statements can be viewed as ones that were made “voluntarily.” While the statements do not appear to have been the product of the type of undue coercion or duress that would render a confession “involuntary” under the Fifth Amendment, all three statements were made while McKee and Flota were at the Moses Lake Police Department, and the latter two statements were made after McKee and Flota had been arrested and were being investigated for the murder of Ms. McKee.- The psychological duress that is inherent in such circumstances counsels caution in accepting such statements as having been made “voluntarily.” See, e.g., Nazemian, 948 F.2d at 580 (recognizing fact that statement is made while in custody is a circumstance that casts doubts on its reliability). This level of duress was enhanced by the fact that both McKee and Flota were told during their interrogations that they had been implicated in the crime by others. See, Lee, 476 U.S. at 544 (fact that statement was made after declarant was told that he had been implicated in crime is circumstance that undermines its reliability). These statements, which were made more than two weeks after Ms. McKee’s murder, also are not the type of “spontaneous” or “contemporaneous” statements that the law deems reliable. Additionally, the statements were not made to trusted confidants or other persons to whom McKee and Flota were unlikely to lie. Nor can the manner in which McKee and Flota were questioned by the investigating officers, who had only recently learned of Ms. McKee’s death and of McKee’s and Flota’s potential involvement in her death, be fairly construed as equivalent to contemporaneous cross-examination. Moreover, there was nothing about McKee and Flota or their mental state that suggests that these statements were especially likely to be trustworthy. Indeed, the traumatic nature of the events in question and the fact the statements were made in the course of custodial interrogation suggests the opposite. Most importantly, McKee and Flota had a strong incentive to misrepresent their role in the murder and shift the blame to Petitioner. Indeed, this is the very basis of the presumption of unreliability that accompanies those portions of confessions that incriminate individuals other than the confessor. In McKee’s case, there was a powerful motive to shift blame to others because the murder victim was his wife and he had repeatedly indicated that he wished to either kill her or have her killed. See, e.g., Trial Record at 1493 (McKee offered friend $4,000 to kill Ms. McKee shortly after their marriage). The motive to lie and shift blame is equally pronounced with respect to Flota, who admitted that he personally participated in the assault on Ms. McKee, but claimed that his conduct caused her no harm. McKee’s and Flota’s motives were even more prominent given that there was evidence that tensions had arisen between them and Petitioner and that Petitioner had threatened to turn in his four codefendants for murdering Ms. McKee. See Trial Record at 584 (testimony of Hughes). Moreover, even the most cursory review of McKee’s and Flota’s statements reveals that they minimized their own role in the murder and placed the blame squarely on Petitioner’s shoulders. In his initial statement, for example, McKee stated: He [Whelchel] told me if I tried to do anything, he’d he’d do the same thing to me. He said he would have killed me right there on the spot too____I wanted to stop him, I wanted to and that but he wouldn’t, he wouldn’t stop you know, so I took off ____I kept blackin’ out and I was shaking and everything and then about I don’t know, an hour to a half an hour later, they, he um Beth came up and that and said, “Don’t worry, every thing’s okay. Steve won’t do anything to you, you know. Nothing will happen to us.” I, I just, I couldn’t handle it. Respondent’s Exhibit 26 at A10. The lack of trustworthiness of McKee’s statements is also demonstrated by the fact that he initially came to the police station not to make a clean breast of his role in his wife’s murder, but to file a missing person’s report. Additionally, a comparison of McKee’s two statements reveals substantial inconsistencies that call into question the veracity of his version of the murder. Compare, e.g., Respondent’s Exhibit 26 at A26 (contending that McKee did not believe his wife would be assaulted) with Respondent’s Exhibit 27 at A47 (acknowledging that McKee participated in game to choose who would murder Ms. McKee). Flota’s single statement also contains similar passages that downplay his role and portray Whelchel as the moving force behind the assault on Ms. McKee. See, e.g., Respondent’s Exhibit 28 at A70: Flota: ... it was the three of us [McKee, Flota and Hughes] sitting on the hill and we weren’t sure what was going on and at that point, I didn’t really care, okay, I was almost to the point where I was going to puke, just the first impact, I got this queasy feeling in my stomach and just, you know, oooo, turned and walked out____And after that point, Steve sort of started getting meaner, his voice started getting meaner and he just changed quickly, started to wig out on us. And after that point we knew that if we didn’t do what Steve was going to say, we were going to get the shit beat out of us, not killed or any- ■ thing, just get beat up for a while____ All three of these statements are replete with similar examples of McKee and Flota painting a picture of themselves as the unwilling pawns of the Petitioner in the assault on Ms. McKee. Balancing these - factors, there is little about these statements — statements in which two admitted participants in Ms. McKee’s murder place the blame for that murder on Petitioner — that suggests they were made in circumstances endowing them with particularized guarantees of trustworthiness. Thus, the admission of these statements violated Petitioner’s right to confrontation under the Sixth Amendment. 3. Prejudice The heart of this case is the question of .whether this error was sufficiently prejudicial to Petitioner to warrant granting his petition. This question was also reached by the Supreme Court of Washington in its consideration of Petitioner’s initial appeal, and Washington’s justices were divided on the question of whether this error was harmless. For reasons similar to those that were elaborated persuasively in the supreme court’s dissenting opinion, this Court concludes that this error was not harmless. a. Legal Standard The question of prejudice posed in this ease differs somewhat from that decided by Washington’s courts, which were required to determine whether the admission of these statements was “harmless beyond a reasonable doubt.” See Chapman v. California, 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 (1967). In contrast, because Petitioner now is attacking his conviction collaterally based on trial error, the question is whether that error had a “substantial and injurious effect or influence in determining the jury’s verdict.” Brecht v. Abrahamson, 507 U.S. 619, 622, 640-41, 113 S.Ct. 1710, 123 L.Ed.2d 353 (1993) (plurality opinion and Stevens, J., concurring); see also O’Neal v. McAninch, 513 U.S. 432, 115 S.Ct. 992, 130 L.Ed.2d 947 (1995) (following Brecht); Kotteakos v. United States, 328 U.S. 750, 776, 66 S.Ct. 1239, 90 L.Ed. 1557 (1946) (origin of Brecht standard). Neither party bears the burden of proof on this question. Instead, a court must conduct a de novo review of the entire record and determine this question “without the benefit of such aids as presumptions or allocated burdens of proof.” O’Neal, 513 U.S. at 437 (internal quotation omitted). If a court is convinced that error was not harmless, or if the court is in grave doubt because the evidence relating to harmlessness is in virtual equipoise, a petitioner’s conviction must be overturned. Id. at 432; Thompson v. Borg, 74 F.3d 1571, 1574-75 (9th Cir.1996). While it is undisputed that the Brecht-Kotteakos harmless error standard governs this case, the Court has found it difficult to come to a precise understanding of what it means to say that an error had a “substantial and injurious effect or influence in determining the jury’s verdict.” From this standard’s plain language, it is apparent that it must have had some effect or influence on the jury’s “verdict;” i.e., whether the jury ultimately would have convicted or acquitted if the error had not occurred. It is also well established that this inquiry focuses on the jury that decided the petitioner’s fate rather than what a court thinks a hypothetical “rational” jury would have done had the error not occurred. See, e.g., Kotteakos, 328 U.S. at 765 (“The inquiry can not be merely whether there was enough to support the result ____ It is rather, ... whether the error itself had substantial influence.”); see also California v. Roy, 519 U.S. 2, -, 117 S.Ct. 337, 339, 136 L.Ed.2d 266 (1996) (Scalia, J., concurring) (trial error “can not be rendered harmless by the fact that, given the evidence no reasonable jury would have found otherwise. To allow error to be cured in this fashion would be to dispense with trial by jury.”). Additionally, it is a matter of common sense that the “injurious” component of this standard means the error must have made it more likely that the jury would convict the petitioner; errors that are helpful or noninjurious are, by definition, harmless. Less clear is what it means to say an injurious effect or influence on the jury’s verdict was “substantial.” Virtually all of the published judicial decisions identifying or applying this standard simply recite the standard and then conclude that the facts of the case either meet or fail to meet it. In the Supreme Court decision that originally enunciated this standard, for example, the Court described this standard in the following circular fashion: If, when all is said and done, the conviction is sure that the error did not influence the jury, or had but very slight effect, the verdict and the judgment should stand ____But if one cannot say, with fair assurance, after pondering all that happened without stripping the erroneous action from the whole, that the judgment was not substantially swayed by the error, it is impossible to conclude that substantial rights were not affected. The inquiry cannot be merely whether there was enough to support the result, apart from the phase affected by the error. It is rather, even so, whether the error itself had substantial influence. If so, or if one is left in grave doubt, the conviction cannot stand. Kotteakos, 328 U.S. at 764-65. One possibility is that this embodies the type of “but for” requirement that is commonly found in causation inquiries; i.e., that but for the error, the jury would not have convicted the petitioner. Because the standard focuses on whether the injurious effect or influence was “substantial,” however, such an approach would be inconsistent with the general recognition that the “substantial factor” approach to causation involves a different inquiry than “but for” causation. See, e.g., Price Waterhouse v. Hopkins, 490 U.S. 228, 293 n. 4, 109 S.Ct. 1775, 104 L.Ed.2d 268 (1989) (Kennedy, J. dissenting); SEC v. Murphy, 626 F.2d 633, 650 (9th Cir.1980). Moreover, the circuits that have addressed this question have concluded that the BrechtKotteakos standard does not require “but for” causation. See; e.g., Robinson v. Arvonio, 27 F.3d 877, 885 (3d Cir.1994). Instead, this test appears to require only that the error made it substantially more likely that the jury would convict rather than acquit. This approach is consistent with the common understanding of the term “substantial” as describing a point between “de minimis” and “necessary” that tends toward the more demanding end of that scale. See, e.g., Webster’s New Third International Dictionary at 2280 (unabridged 1967) (defining substantial as “something of moment: an important or material matter.”); see also Black’s Law Dictionary 1428 (6th ed.1990) (defining substantial as “[o]f real worth or importance; of considerable value ... actually existing; real; not seeming or imaginary, not illusive”). It is also consistent with the infrequent discussions of the term “substantial” in the Supreme Court’s and Ninth Circuit’s case law, which have tended to characterize an effect as substantial when its impact is important rather than de minimis. See, e.g., United States v. Lopez, 514 U.S. 549, 558, 115 S.Ct. 1624, 131 L.Ed.2d 626 (1995); United States v. Juvenile Male, 118 F.3d 1344, 1348 (9th Cir.1997). Most importantly, however, it also is consistent with the Supreme Court’s characterization of this standard in Brecht and Kotteakos as being less stringent than Chapman’s “harmless beyond a reasonable doubt” inquiry, but more demanding than a simple inquiry as to whether enough evidence was presented to support the conviction. Moving to the more specific context of this case, the Supreme Court has elaborated an illustrative set of factors that should be weighed when assessing the harmlessness of Confrontation Clause violations. These factors include the importance of the witness’ testimony in the prosecution’s case, whether the testimony was cumulative, the presence or absence of evidence corroborating or contradicting the testimony of the witness on material points, the extent of cross-examination otherwise permitted, and, of course, the overall strength of the prosecution’s case. Delaware v. Van Arsdall, 475 U.S. 673, 684, 106 S.Ct. 1431, 89 L.Ed.2d 674 (1986); see also Brecht, 507 U.S. at 639 (looking at frequency of error, relevance of error to state’s case, whether error was cumulative of properly admitted evidence, and strength of state’s ease). The consideration of these and other relevant factors necessarily involves some level of guesswork because a court cannot simply ask jury members why they reached a particular verdict. A court’s inquiry must be based on evidence, however, not simply speculation as to what went through the jury’s mind. For example, in a Confrontation Clause case, it is not permissible to base an assessment as to harmlessness on speculation as to whether the witness’s testimony would have been the same, or the jury’s verdict unaltered, if the defendant had been able to confront the witness at trial. Coy v. Iowa, 487 U.S. 1012, 1021-22, 108 S.Ct. 2798, 101 L.Ed.2d 857 (1988); see also Kotteakos, 328 U.S. at 750 (“it is not the [] court’s function to determine guilt or innocence. Nor is it to speculate upon probable reeonvietion and decide according to how the speculation comes out.”). b. The Effect of McKee’s and Flota’s Statements on the Verdict Having applied these factors and considered the relevant evidence, the Court concludes that the admission of McKee’s and Flota’s statements had a substantial and injurious effect or influence on the jury’s verdict. This is true regardless of whether this test entails a “but for” inquiry or a lesser showing of effect on the jury’s verdict. Moreover, while this is a close ease, it is not one in which the Court is in grave doubt or in which the evidence is in virtual equipoise. The record indicates that the first factor listed above — the importance of these statements to the prosecutor’s case — weighs in favor of a finding that the jury’s verdict was substantially and injuriously affected by these statements. At trial, the prosecutor presented three basic types of evidence implicating Petitioner in the murder of Ms. McKee: 1) the testimony of each of Petitioner’s four codefendants, all of whom implicated him as the person most responsible for Ms. McKee’s death; 2) testimony of third parties, who either placed Petitioner near the scene of the murder or heard him make incriminating statements regarding the murder; and 3) physical evidence, such as clothes and a knife that may have been one of the murder weapons that were found in Petitioner’s ear when he was arrested. The most important leg of this triad was the testimony and statements of the codefendants, each of whom was present at the murder and participated in the events directly preceding the murder. Such testimony is almost always powerfully incriminating, so much so that the Supreme Court has concluded that it is beyond the capability of most jurors to ignore such testimony when it is improperly admitted. See Bruton, 391 U.S. at 135-36. Here, in contrast to the situation that the Supreme Court found objectionable in Bruton, the jurors were expressly permitted to rely on McKee’s and Flota’s testimony in determining whether Petitioner was guilty or innocent. Moreover, without these statements, the prosecution’s eyewitness testimony would have been limited to the testimony of Hughes and Massey. As discussed in the Court’s first summary judgment order, Hughes’ role in the actual murder was very limited as she quickly fled the scene after the assault began and never actually saw Petitioner strike Ms. McKee. Massey, by her own testimony, participated directly in the assault on Ms. McKee, and, thus, as defense counsel pointed out to the jury, she had a powerful motive to shift the principal blame for murder to Petitioner. By introducing the statements of McKee and Flota, the State not only presented two additional eyewitnesses to the assault, but also was able to highlight the interlocking nature of their versions of the murder, all of which asserted that Petitioner was present and directly participated in the assault on Ms. McKee. Indeed, the prosecutor placed such emphasis on the importance of McKee’s and Flota’s statements that he referred to them extensively in his closing argument, including direct quotations from the statements and assertions that the four stories “dovetailed” on their material points. See, e.g., Trial Record at 2067, 2071, 2075-76, 2160-61. The second Van Arsdall factor — whether the impermissible testimony was cumulative of other evidence — favors a finding of harmlessness, but not in a dispositive fashion. In many regards, the McKee and Flota statements are cumulative of the live testimony of Massey and Hughes. Both Massey and Hughes, for example, testified that Petitioner indicated well in advance of the murder that he intended to kill Ms. McKee and that he was both present at the time of the murder and inflicted the first blow to the victim. Massey also testified that Petitioner repeatedly beat and eventually stabbed the victim, something that McKee and Flota also stated in their taped statements. On a broader plane, the McKee and Flota statements were also cumulative of the testimony of third parties, described more fully below, who either placed Petitioner near the scene of the murder or heard him make incriminating statements regarding the murder. There are, however, a number of significant elements that are contained in McKee’s and Flota’s statements but not in Massey’s or Hughes’ testimony. For example, both McKee and Flota state that shortly after arriving at the murder scene, they and Petitioner participated in a macabre game of “odd man out” to determine who would murder Ms. McKee, and that Petitioner decided to begin the assault even though he was not selected. Additionally, while both Flota and Massey stated that Flota placed the victim in a chokehold, only Flota contended that Ms. McKee was still alive when he finished, thereby implying that the injuries that killed her were inflicted by Petitioner. Finally, it also is significant that only Flota and Massey testified they were present during the assault after the initial blows were inflicted. Without Flota’s statement regarding what occurred during this period, the only direct evidence detailing Petitioner’s alleged assault on Ms. McKee was the testimony of Massey, who, as an admitted assailant, had an obvious motive to shift blame. In such a context, McKee’s and Flota’s statements cannot fairly be construed as being simply “cumulative.” The third factor — the presence or absence of evidence corroborating or contradicting the testimony of the witness on material points — is more difficult to assess. Generally, courts have viewed the existence of corroborating evidence as a factor that decreases the likelihood that improperly admitted testimony affected or influenced the jury’s verdict. See, e.g., United States v. Sasson, 62 F.3d 874, 885 (7th Cir.1995), cert. denied, 516 U.S. 1131, 116 S.Ct. 953, 133 L.Ed.2d 876 (1996). The underlying rationale of this view appears to be that the existence of corroborative evidence makes it more likely that the jury based its verdict on the properly admitted evidence rather than the tainted testimony. In this sense, however, this factor adds little to the harmlessness inquiry that is not already accomplished by considering whether the improperly admitted testimony was cumulative and whether the prosecution’s case was otherwise strong. In other words, for corroborative evidence to reduce the likelihood that improperly admitted testimony affected or influenced the jury’s verdict, the corroborative evidence must possess some independent prejudicial value apart from its relationship to the tainted evidence; i.e., it must be cumulative to some extent. If not, its primary value to the prosecution is to enhance the credibility of the tainted testimony in the jurors’ eyes, thereby increasing the likelihood that this testimony affected or influenced the verdict. By the same token, the existence of evidence contradicting the tainted testimony decreases the likelihood that the testimony affected or influenced the verdiet because contradictory evidence undermines the probative value of the erroneously admitted testimony. In this case, the statements of McKee and Flota were corroborated by other evidence on most of their key points. The testimony of Massey and Hughes corroborated those statements on the crucial questions of whether Petitioner premeditated Ms. McKee’s murder and whether he inflicted the injuries that resulted in her death. To a lesser extent, their testimony was also corroborated by the testimony of third parties, described more fully below, who either placed Petitioner near the scene of the murder or heard him make incriminating statements regarding the murder. The principal impact of this evidence, in the Court’s view, was to increase the likelihood that Petitioner was convicted on the basis of the improperly admitted McKee and Flota statements. As already discussed, the existence of this evidence did, to some extent, increase the probability that the jury based its decision on properly admitted evidence. This consideration, however, is already taken into’ account by other Van Arsdall factors. As yet unaccounted for is the harmful effect that the existence of evidence corroborating McKee’s and Flota’s statement had on the jury. There is little doubt the corroborating evidence had such an effect in this case. This was not a case where the prosecution presented a powerfully incriminating pattern of physical and testimonial evidence. Rather, this case was built almost entirely on the testimony of the four codefendants, each of whom had a strong motive to shift the blame to someone else. Thus, that testimony was naturally suspect because of this motive, a point that Petitioner’s trial counsel repeatedly brought home to the jury. The fact that all four told essentially the same story greatly enhanced the weight of this evidence. Thus, the corroborative effect of Hughes’ and Massey’s testimony enhanced the credibility of McKee’s and Flota’s statements, just as those statements enhanced the prejudicial value of Hughes’ and Massey’s testimony. The net effect was that the prosecutor was able to present the interlocking testimony of four codefendants, rather than two, a factor that almost certainly was harmful to Petitioner’s chances of gaining an acquittal. On the other hand, Petitioner also presented evidence that contradicted McKee’s and Flota’s statements. Specifically, Petitioner testified that he played no part in the murder, and his father, mother, and sister all testified that he was at home when the murder occurred. Because this testimony gave the jury a basis for questioning the credibility of McKee’s and Flota’s statements, it increases the likelihood that the admission of those statements was harmless error. Ultimately, however, the effect of this testimony on the jury’s consideration of McKee’s and Flota’s statements was probably minimal because, as discussed below, Petitioner’s alibi was effectively impeached by other evidence. In sum, this third factor was one that increased the likelihood that McKee’s and Flota’s statements had a substantial and injurious effect or influence on the jury’s verdict. The existence of evidence corroborating McKee’s and Flota’s statements served principally to enhance the harmfulness of those statements, and Petitioner’s contradictory evidence had minimal effect on the harmfulness of these statements. The next Van Arsdall consideration — the extent to which some cross-examination was permitted — works unambiguously against a finding of harmlessness. McKee and Flota never testified in any pretrial proceeding or at trial, other than to indicate that they would invoke their Fifth Amendment right not to incriminate themselves. Thus, the testimony that the jurors viewed was solely their potentially self-serving version of the events, unchecked by the truth-enhancing protections inherent in the right of confrontation: testimony under oath, in front of a jury and the accused, subject to cross-examination. The final Van Arsdall factor — the overall strength of the prosecution’s case — favors a finding that Petitioner was unfairly prejudiced by the admission of the McKee and Flota statements. This was not a ease where the untainted evidence was so strong that the erroneous admission of McKee’s and Flota’s statements likely had little or no effect on the jury. Close scrutiny of the prosecution’s case reveals that the evidence against Petitioner was far from overwhelming. The strongest pillar supporting the prosecution’s case was the direct testimony of Hughes and Massey, which clearly implicated Petitioner in the murder of Ms. McKee. Such testimony is inherently suspect, a consideration that the trial court properly instructed the jury they should take into account when considering how much weight it should be given in their deliberations. Moreover, defense counsel described to the jury numerous reasons why they should give this testimony little weight. First, as participants in the murder conspiracy, both Hughes and Massey had a substantial motive to downplay their role and shift the blame to the other participants, such as Petitioner. This motive was most pronounced for Massey, who provided the most incriminating testimony because she was present throughout the entire assault on Ms. McKee. Hughes had less of a motive to shift blame from herself because of her more limited role in the murder, but the limited nature of her participation also undercut the strength of her testimony because, unlike Massey, Hughes never actually saw Petitioner strike or stab Ms. McKee. Additionally, Massey and Hughes both had additional motives to shift the blame for the murder from themselves and McKee and Flota. At the time of the murder and the arrests, for example, Hughes was Flota’s girlfriend. Petitioner and Massey also had been involved at the time of the murder, but had become estranged in the weeks thereafter because Petitioner believed Massey was dating one of his friends. Trial Record at 584, 613, 636, 699. More significantly, there was evidence that tensions had developed between Petitioner and all of his eodefendants by the time of their arrests because Petitioner had threatened to turn them in for their role in the murder. Finally, both Hughes and Massey testified pursuant to agreements with the State under which they received lenient treatment in return for their promise to testify against McKee, Flota, and Petitioner. In sum, as the defense and trial court pointed out, there was ample reason for the jury to treat the testimony of Hughes and Massey with caution. There was also testimony from a number of third parties who either saw Petitioner near the scene of the murder or heard him make incriminating statements regarding the murder. As with Hughes’ and Massey’s testimony, however, the strength of this evidence was substantially undermined at trial. For example, McKee’s apartment manager testified that he saw Petitioner with the group that went down to Moses Lake shortly before Ms. McKee’s murder, but he also testified that: 1) he was sleepy at the time; 2) the group consisted of five persons, not the six that all four of Petitioner’s eodefendants claimed were present; 3) he only saw these people from behind and identified Petitioner based on his build and hair; 4) he had only seen Petitioner and Flota in passing on a small number of occasions prior to the murder; and 5) he did not identify Petitioner until several months after the murder. Trial Record at 732, 737-38, 740-41. Additionally, although witnesses testified that Petitioner made statements before and after the murder indicating he had helped kill, or would like to kill, Ms. McKee, there are strong reasons to question the accuracy or probative value of these statements. Many of these individuals were persons who had reasons to carry a grudge against Petitioner. See, e.g., id. at 850-51 (admitting Petitioner had previously “put me in the hospital” in a fight). In addition, most of the statements were generally ambiguous and potentially consistent with his admitted role in the murder conspiracy after the fact. See, e.g., id. at 753, 765 (Flota and Petitioner casually told two friends that “we killed” and “they killed” Ms. McKee.); id. at 896-901 (witness admitted he could not recall Petitioner’s exact words and that witness “crashed” due to heavy drinking after talking with Petitioner). In fact, the only witness who testified that Petitioner unequivocally admitted he had personally participated in Ms. McKee’s murder was a former cell mate of McKee and Petitioner who testified at the request of McKee. Trial Record at 917-26. Moreover, there was evidence in the record that indicated Petitioner had a penchant for claiming involvement in murders that he did not commit, evidence from which the jury could reasonably have inferred that Petitioner’s statements were merely a perverted form of puffery that were not based in fact. See Respondent’s Exhibit 26 at A31 (statement by McKee that Petitioner bragged about being involved with organized crime and of killing 64 people). This also was not a case where overwhelming physical evidence linked Petitioner to the crime. According to the prosecution, the clubs that were used to assault Ms. McKee were never recovered. No blood was found on Petitioner’s clothes, and the prosecution was unable to establish conclusively that Petitioner’s knife was the weapon that was used to stab Ms. McKee. While evidence relating to the crime, including Ms. McKee’s purse, was found in the car in which Petitioner was stopped prior to his arrest, this is entirely consistent with his version of his role in the murder as being one of an accomplice after the fact rather than a direct participant. Additionally, the strength of the prosecution’s ease was undercut, to some extent, by the fact that Petitioner mounted a vigorous, affirmative defense of his innocence. Petitioner himself took the stand and described for the jury a markedly different version of events than was described by his codefendants. According to Petitioner, he returned home from the McKee’s apartment several hours before the murder and did not learn of the murder until the next day. In support of this story, three members of Petitioner’s family testified that they saw him at home on the morning of the murder, a time when, according to his codefendants, Petitioner was either killing Ms. McKee or sleeping after-wards at the McKees’ home. Additionally, a former cell mate of McKee’s testified that McKee had bragged and joked with him about his involvement in Ms. McKee’s murder without ever mentioning involvement by Petitioner. Trial Record at 1793-96. While this evidence is riddled with many of the inconsistencies and flaws that existed with regard to the prosecution’s case-in-chief, this evidence gave the jury an additional reason to question whether Petitioner had actually participated in the murder of Ms. McKee. Finally, there is an additional indicator that the prosecution’s ease was not strong: the extent to which the jury struggled to reach its verdict. In most cases, the hallowed principle of jury secrecy combined with the many potential factors that might underlie a jury’s verdict make it difficult or impossible to divine how a jury viewed the evidence in a case. In this ease, however, the jury broke from its deliberations at one point to tell the trial court that it was unable to reach a unanimous verdict after taking three separate votes and weighing all of the evidence. While not a controlling factor, the jurors’ difficulty in reaching a verdict even when they had before them the improperly admitted statements suggests that the prosecution’s untainted evidence was not so strong that the improper admission of McKee’s and Flota’s statements was harmless. In sum, having considered the factors detañed above, the Court concludes that the admission of McKee’s and Flota’s statements had a substantial and injurious effect on the jury’s verdict. B. George Flota’s Testimony Petitioner also contends that the trial court erroneously añowed a videotape of Flota’s father — George Flota — to be played for the jury. As with Petitioner’s claims regarding the admission of Flota’s and McKee’s audio-taped statements, this ground for relief arises out of the Confrontation Clause. Thus, this claim is subject to the same two-part analysis as to whether the alleged error violated the Confrontation Clause, and whether any such error was sufficiently prejudicial to Petitioner to warrant relief. This issue was raised in Petitioner’s direct appeal and was considered by the Court of Appeals of Washington. That court, the only state court to reach this issue, held that the admission of George Flota’s videotaped testimony violated the Confrontation Clause, but that this violation was harmless. Washington’s supreme court declined to review this issue. Though this analysis, described below, differs somewhat from that of the court of appeals, its conclusion is the same. 1. Factual and Procedural Background Petitioner testified at this trial that he was not present when Ms. McKee was murdered. According to Petitioner, he had been at the McKees’ apartment on the night of September 26,1986, but he left several hours before the murder occurred, returning home by around 1:30 a.m. He remained awake for several hours, watching television, then feñ asleep and slept until 10:30 or 11:00 a.m. on September 27, 1986. After rising, he ate breakfast, then remained at home until sometime that afternoon, when he went to the McKee’s apartment. Later that evening, McKee and Flota told him they, Massey, and Hughes had kified Ms. McKee. Three members of Petitioner’s famfly testified that Petitioner was at home on the morning of the murder. Petitioner’s mother testified that she awoke around 3:30 a.m. and found Petitioner watching television in her living room. Petitioner’s sister testified that she was awakened at the same time by a conversation that Petitioner had with his mother. Petitioner’s mother and father also testified that when they arose for the day at 6:00 a.m., Petitioner was asleep in their living room. They also testified that he rose shortly before noon and remained at home until later that afternoon. Petitioner’s sister also testified that Petitioner was asleep at home when she awoke that morning. In addition to the statements and testimony of Petitioner’s four eodefendants, four witnesses directly contradicted his story that he was at the Whelchel home on the morning of September 27, 1986. The first witness, the apartment manager at the McKees’ apartment complex, testified that he saw Petitioner leave the McKees’ apartment at 3:30 a.m. on September 27,1986 in the company of Ms. McKee, her husband, Massey and Hughes. Two additional witnesses — Jeanetta Massey (Massey’s mother), and Deputy Sheriff Detrolio (“Detrolio”) of the Grant County Sheriffs Office — testified that they went to Petitioner’s family home between noon and 1:00 p.m. that day in connection with a runaway complaint that Jeanetta Massey had filed regarding her daughter Beth. Detrolio testified that he saw no sign of Petitioner at the home and that he was told Petitioner was not living there. Detrolio also testified that he stopped McKee and Petitioner in McKee’s ear in a separate part of town at 2:10 p.m. and that their car was full of laundry. Jeanetta Massey testified that she went to Petitioner’s home after Detrolio, where she saw George Flota and talked to the Whelchels in their kitchen. She also saw no sign of Petitioner and was told by the Whelchels that he was not living there at that time. The fourth witness was George Flota, Jeffrey Flota’s father. Flota testified that he went to the Whelchel home during the late morning of September 27, 1986, and that he was there in the Whelehel’s kitchen when Detrolio and Massey arrived. Like Detrolio and Massey, he testified that he saw no sign of Petitioner at Whelchels’ residence, and that Petitioner’s family gave him no indication of Petitioner’s whereabouts. Flota also testified that he arrived earlier than Detrolio and Massey, but could not place the time exactly. Flota also provided additional testimony regarding the relationship between the Flotas and Petitioner, including a description of a fight in which the three of them used clubs made from table legs to defend themselves from others. In their trial testimony, Petitioner’s father and mother acknowledged that Flota, Detrolio, and Jeanetta Massey visited their residence and were told that Petitioner was not at home. They contended, however, that these visits occurred during the late morning or early afternoon of September 24-25,1986, several days before the murder occurred. Contrary to this testimony, however, records from Grant County’s 911 Dispatch center and the testimony of Grant County 911 Dispatch Supervisor Rowland corroborated that Detrolio was dispatched to the Whelchel home at 12:15 p.m. on September 27,1986. George Flota’s testimony was presented by means of a videotape of his deposition. Portions of this videotaped testimony were played for the jury, once in the prosecution’s case-in-chief, and once for the purpose of impeaching Petitioner’s alibi. Trial Record at 928, 2054; see also Respondent’s Exhibit 25 (transcript of videotape). Ultimately, the entire videotape was played for the jury. Petitioner objected to the admission of this testimony on the ground that the prosecution had not shown that George Flota would be unavailable as a witness. Id. at 227, 927. The trial court, in a preliminary proceeding, ruled that because George Flota would be moving to Arkansas for a new job prior to the trial, he would be unavailable to testify at trial. Id. at 229. Petitioner challenged this ruling on appeal. The Court of Appeals of Washington concluded that the admission of this testimony violated the Confrontation Clause. Respondent’s Exhibit 2 (unpublished opinion) at A7-A8. That court predicated its ruling on its finding that the State had made no effort to secure George Flota’s live testimony at trial and, therefore, had not demonstrated that he was legally unavailable to testify at trial. The court of appeals declined to reverse Petitioner’s conviction on this ground because it viewed the error as harmless. Id. at A9. Although Petitioner appealed this ruling to the Supreme Court of Washington, that court exercised its discretion not to review this aspect of the court of appeals’ decision. 2. Confrontation Clause Violation The admission of George Flota’s videotaped deposition testimony presents the reverse question from that posed by the admissions of McKee’s and Flota’s statements: whether testimony that fits a firmly rooted exception violates the Confrontation Clause if the declarant was not unavailable to testify at trial. In this case, the answer to this question is yes. George Flota’s testimony was taken at a prior judicial proceeding presided over by the trial judge, at which Petitioner was present and was represented by counsel who cross-examined the witness on the relevant points. As such, it falls within the “firmly rooted exception” for former judicial testimony, and there is no need to do an independent inquiry into its reliability. Ohio v. Roberts, 448 U.S. 56, 66, 72, 100 S.Ct. 2531, 65 L.Ed.2d 597 (1980); United States v. Koon, 34 F.3d 1416, 1426 (9th Cir.1994), affirmed in relevant part, 518 U.S. 81, 116 S.Ct. 2035, 135 L.Ed.2d 392 (1996); Fed.R.Evid. 804(b)(1). The issue here is the proscription on admission of prior judicial testimony if the declarant is available to testify at tr