At Jensen’s trial for the murder of his wife Julie the State introduced Julie’s handwritten letter to the police, written two weeks before her death, in which she wrote she would never take her life and that her husband should be the suspect if anything should happen to her. The Wisconsin Court of Appeals assumed the admission of the letter violated Jensen’s right to confrontation but found the error harmless. The Seventh Circuit holds that the court of appeals’ decision was an unreasonable application of the Chapman v. California, 386 U.S. 18 (1967), harmless error standard, and that the erroneous admission of the hearsay evidence had a substantial and injurious influence or effect in determining the jury’s verdict, thus satisfying the actual prejudice standard under Brecht v. Abrahamson, 507 U.S. 619 (1993).
The facts and history of this case is laid out in detail in our post on the district court’s grant of habeas relief. Here we highlight the Seventh Circuit’s conclusions about (and criticisms of) the Wisconsin Court of Appeals decision in State v. Jensen, 2011 WI App 3, 331 Wis. 2d 440, 794 N.W.2d 482 (Jensen II). As we said of the district court’s decision, this is a must-read for every litigator frustrated by the slapdash application of harmless error analysis by Wisconsin’s appellate courts, and will be worth citing as authority for the proper way to analyze whether an error is harmless.
Before turning to the decision, a note about the standard of review. A few months ago Davis v. Ayala, 134 S. Ct. 2187 (2015), clarified the relationship between Brecht‘s pre-AEDPA “actual prejudice” standard for reviewing state court determinations of harmless error and AEDPA’s general deferential standard of review, which requires the petitioner to show the state court’s decision was an unreasonable application of law. Ayala says the two standards govern together, so that a petitioner must meet the Brecht standard and must show that the state court unreasonably applied Chapman. (Slip op. at 15-16).
The central conclusion the Seventh Circuit reaches is that the court of appeals unreasonably applied Chapman by doing what amounted to a sufficiency of the evidence analysis instead of a harmless error analysis:
…. Time and again, the Supreme Court has emphasized that a harmless-error inquiry is not the same as a review for whether there was sufficient evidence at trial to support a verdict. Nearly seventy years ago, in Kotteakos v. United States, 328 U.S. 750 (1946), the Supreme Court explained as it conducted harmless-error review of jury’s decision:
And the question is, not were they right in their judgment, regardless of the error or its effect upon the verdict. It is rather what effect the error had or reasonably may be taken to have had on the jury’s decision.…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.
Id. at 764-65. The Supreme Court has reinforced this principle over and over. ….
Despite this long line of cases establishing the test for harmless error, the Wisconsin appellate court’s reasoning reads as though it is conducting an evaluation of whether there was sufficient evidence to support the verdict, not whether the error in admitting Julie’s letter and statements to police affected the jury’s verdict. Cf. Kotteakos, 328 U.S. at 764-65. Near the beginning of its analysis, the state appellate court stated, “Here, we will not attempt to catalog all the untainted evidence the State presented; however, we will summarize some of the compelling pieces in order to illustrate that the record is replete with reason to uphold the jury’s verdict, even if the assumedly tainted evidence is disregarded.” Jensen II, 794 N.W.2d at 493. The court then went through five categories of evidence presented by the State—computer evidence, motive evidence, Jensen’s incriminating statements, medical evidence, and miscellaneous evidence. Id. at 493-94. The court said this was evidence from which “a rational jury could alone conclude beyond a reasonable doubt” that Jensen murdered his wife. Id. at 494. But a statement of what a “rational jury could conclude” is not a statement of a harmless-error inquiry; it is instead the question presented when a direct appeal asks whether there is sufficient evidence to support a verdict. See State v. Kimbrough, [2001 WI App 138, 246 Wis. 2d 648,] 630 N.W.2d 752, 756 …. That is not the question here. (Slip op. at 17-19).
In addition, the court of appeals’ discussion of the trial evidence shows it unreasonably applied Chapman because it focused on the prosecution’s evidence over and above the letter but paid scant attention to the defense evidence—a mistake that also shows why its conclusion about harmless error was wrong:
The state appellate court decision contains a very detailed discussion of the State’s evidence. But its discussion does not engage with the defense evidence that goes against the evidence discussed by the court. The Supreme Court has said, however, that when a court “evaluat[es] the strength of only one party’s evidence, no logical conclusion can be reached regarding the strength of contrary evidence offered by the other side to rebut or cast doubt.” Holmes v. South Carolina, 547 U.S. 319, 331 (2006).
To be clear, if the question was whether there was sufficient evidence to convict Jensen, the answer would be “yes.” But the harmless error test does not focus just on the sufficiency of other evidence. The question as we conduct the Brecht analysis is whether we are in “‘grave doubt about whether a trial error of federal law had “substantial and injurious effect or influence in determining the jury’s verdict.”’” Ayala, 135 S. Ct. at 2198 (quoting O’Neal [v. McAninch], 513 U.S. [432,] 436 [(1995)] (emphasis added)). So we must look at the influence the improperly admitted handwritten letter and accusatory statements to the police had on the verdict. In this analysis “we look to ‘a host of factors,’ such as ‘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.’” Jones [v. Basinger], 635 F.3d [1030,] 1052 [(7th Cir. 2011)] (quoting Delaware v. Van Arsdall, 475 U.S. 673, 684 (1986)). (Slip op. at 20-21).
The opinion reviews the evidence in detail, noting that Julie’s letter was a unique piece of evidence that played a key role in the trial (it was read early in the state’s opening and at the end of the State’s rebuttal closing, and in between 12 witnesses testified about it), and that the State recognized its importance because it repeatedly fought to get the letter admitted, at one point calling it a “make or break issue.” (Slip op. at 21-24). But apart from the letter, the State’s case was circumstantial, and was “no slam dunk” (slip op. at 26), and was disputed in key points by Jensen’s defense—evidence the court of appeals didn’t pay much attention to. “As the district court observed, ‘A reader of the court of appeals’ opinion would conclude that Jensen called no witnesses, introduced no evidence, never questioned the credibility of any witness, and never even elicited helpful testimony from a prosecution witness.’ But that is far from what actually happened during the six-week trial.” (Slip op. at 25).
There is a dissent. Judge Tinder disagrees that the court of appeals’ harmless-error determination was itself unreasonable and disputes the conclusion that the court of appeals did a sufficiency rather than a harmlessness analysis:
Based on the duplicative nature of Julie’s testimonial statements and the overall strength of the prosecution’s case (even considering the defense evidence discussed by the majority), I am not convinced that the state court’s decision “‘was so lacking in justification that there was an error well understood and comprehended in existing law beyond any possibility for fairminded disagreement.’” Ayala, 135 S.Ct. at 2199 …. (Slip op. at 39).
A final note: Ayala resolves another dispute between the parties here—namely, whether Jensen II is the state court’s “last adjudication on the merits” for purposes of habeas review, or whether, as the State argues, the Wisconsin Supreme Court’s earlier decision, State v. Jensen, 2007 WI 26, 299 Wis. 2d 267, 727 N.W.2d 518 (Jensen I), is the focus of review. This matters because Jensen I came before Giles v. California, 554 U.S. 353 (2008), which made it clear that Jensen’s confrontation rights were violated, and therefore Jensen I couldn’t have been an unreasonable application of the not-yet-issued decision in Giles. The Seventh Circuit rejects the State’s attempt to get around Giles, holding that “[u]nder Ayala … it is clear that the Wisconsin appellate court decision is the last ‘adjudication on the merits’ for AEDPA purposes in Jensen’s case.” (Slip op. at 14).