In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 22-2833 CHRISTOPHER ROALSON, Petitioner-Appellant, v.

JON NOBLE, Warden, Respondent-Appellee. ____________________

Appeal from the United States District Court for the Eastern District of Wisconsin. No. 2:18-cv-01831-PP — Pamela Pepper, Chief Judge. ____________________

ARGUED MAY 15, 2024 — DECIDED AUGUST 28, 2024 ____________________

Before BRENNAN, KIRSCH, and LEE, Circuit Judges. BRENNAN, Circuit Judge. Christopher Roalson is serving a life sentence for stabbing and bludgeoning a 93-year-old woman to death. At his trial, a DNA analyst testified about evidence left behind on the two knives and barstool used to commit the murder. The testifying analyst did not swab the items and develop the sample left on the weapons—another analyst performed that work but was unavailable at trial. In- stead, the analyst who took the stand testified that Roalson 2 No. 22-2833

was a possible contributor after comparing the sample from the weapons to a sample of his DNA. Roalson was convicted and now collaterally challenges that conviction. He argues that the trial court denied his right to confront a witness by allowing the substitute analyst to testify. The district court de- nied his habeas petition, and he asks us to reverse that deci- sion. I. In 2009, Christopher Roalson and Austin Davis broke into the Radisson, Wisconsin home of a 93-year-old woman. Davis was in the kitchen looking for items to steal when Roalson emerged from the woman’s bedroom holding a bloody knife. Roalson and Davis had each taken a knife from Davis’s cousin’s house, their last stop before the break-in. Wordlessly, Roalson grabbed Davis’s knife from his hand, picked up a wooden kitchen stool, and went back into the bedroom. Davis heard the woman screaming from the kitchen. Then, he heard something break and saw Roalson running out of the bed- room. Roalson kicked down a screen door and ran from the house, and Davis followed. As the two fled, Roalson broke the silence. He said that he stabbed the woman “a bunch of times” and “broke the chair over her,” that “he was Satan’s son,” and that the woman “would [have] been saved if God was here.” A few days later, Roalson told his friend Jacqueline Walczak that he stabbed a woman. According to Walczak, Roalson said he and Davis set out “to rob a lady,” they broke into the house, the “lady … caught them,” “he took a chair and he hit her and he hit her and he hit her,” “he stabbed her and he stabbed her and he stabbed her and he stabbed her,” “he said hail, Satan” while stabbing her, and “if he got away with it, he’d do it again.” When Walczak heard that a woman No. 22-2833 3

had been found murdered and the cause of death was not be- ing disclosed, she contacted police and told them what Roalson had told her. The state charged Roalson with burglary and first-degree intentional homicide. Davis pleaded guilty to second-degree intentional homicide in exchange for cooperating with law enforcement and testifying at Roalson’s trial. Several witnesses testified at Roalson’s trial in Sawyer County, Wisconsin Circuit Court. Davis shared what he saw and what Roalson told him after the murder. Walczak de- scribed Roalson’s confession to her. And Carly Leider, a DNA analyst at the Wisconsin State Crime Laboratory in Madison, also testified. Leider’s testimony is at issue here. Another analyst, Ryan Gajewski, had swabbed evidence collected from the scene, tested the DNA samples he recovered, and concluded that Roalson’s DNA was a possible contributor to some of the sam- ples, including the knives. But Gajewski was unavailable to testify at trial because he was employed elsewhere and was in Afghanistan. So Leider appeared at trial. Leider testified that she looked at Gajewski’s notes and was able “to reach [her] own conclusions based on developed profiles[,]” which she compared to “standards” (that is, a per- son’s DNA sample) to identify potential contributors. She ex- plained her analysis was just like that of a peer reviewer, who examines the principal analyst’s work. But in the peer review process, the reviewer does not retest the sample because the initial swab generates “the best collection of that DNA.” During Leider’s testimony, counsel for the state presented her with several pieces of evidence—in particular, the two 4 No. 22-2833

knives and the barstool—and asked if she was able “to reach an opinion regarding the profile that was developed … versus the standards that were … developed.” Each time, she testi- fied as to her own conclusions. For example, Leider said she reached a conclusion about the DNA collected from swabbing the handle and the blade of one of the knives. The DNA de- tected from the handle included four “possible contributors to this DNA mixture profile,” including the victim, Roalson, and Davis. And the DNA detected from the blade included “a female DNA profile,” of which the victim “was the source.” The jury found Roalson guilty, and he was sentenced to life in prison. He appealed his state conviction, arguing the trial court violated his right to confront Gajewski by allowing Leider to testify instead. The Wisconsin Court of Appeals affirmed, ap- plying a rule from the Wisconsin Supreme Court’s decision in State v. Luther Williams, 644 N.W.2d 919 (Wis. 2002). Citing that case, the court explained the Confrontation Clause is not always violated when one analyst testifies to his own conclu- sions about samples tested by another analyst. When the tes- tifying analyst can provide an independent evaluation of the initial report, the original analyst need not be called. After the Wisconsin and United States Supreme Courts de- nied certiorari, Roalson began his collateral attacks on his con- viction. The state trial court denied his challenge without a hearing, and the Wisconsin Court of Appeals affirmed with- out addressing his Confrontation Clause claim. Roalson then petitioned for habeas corpus in the Eastern District of Wisconsin. The court dismissed the petition. In just a few sentences, the court explained there is no federal law No. 22-2833 5

“clearly holding” that the Confrontation Clause bars a testify- ing analyst from testifying to “her own independent opinions and conclusions regarding the DNA collected … .” Roalson was free to confront Leider “about those opinions and conclu- sions,” so the decision of the Wisconsin Court of Appeals was not unreasonable. Roalson appeals. II. To grant a writ of habeas corpus, the adjudication of the prisoner’s claim must have resulted in a decision that was ei- ther: (1) “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States,” or (2) “based on an un- reasonable determination of the facts in light of the evidence presented in the State court proceeding.” 28 U.S.C. § 2254(d)(1) & (2). We defer to the state-court decision if it is reasonable and review the district court’s decision de novo. Gonzales v. Eplett, 77 F.4th 585, 591 (7th Cir. 2023). Under § 2254(d)(1), an application of federal law is unrea- sonable if it is “so erroneous that ‘there is no possibility fair- minded jurists could disagree that the state court’s decision conflicts with [the Supreme] Court’s precedents.’” Nevada v. Jackson, 569 U.S. 505, 508–09 (2013) (quoting Harrington v. Richter, 562 U.S. 86, 102 (2011)). The federal law itself must be “clearly established,” 28 U.S.C. § 2254(d)(1), which “refers to the holdings, as opposed to the dicta, of [the Supreme] Court’s decisions.” Carey v. Musladin, 549 U.S. 70, 74 (2006) (quota- tions omitted); Shirley v. Tegels, 61 F.4th 542, 545 (7th Cir. 2023); see also Bland v. Hardy, 672 F.3d 445, 448 (7th Cir. 2012) (“Until the Supreme Court has made a right concrete, it has not been ‘clearly established.’”); White v. Woodall, 572 U.S. 415, 427 (2014) (“[C]ourts must reasonably apply the rules ‘squarely 6 No. 22-2833

established’ by [the Supreme] Court’s holdings to the facts of each case.” (quoting Knowles v. Mirzayance, 556 U.S. 111, 122 (2009))). Our first step is to “determin[e] the relevant clearly estab- lished law.” Yarborough v. Alvarado, 541 U.S. 652, 660 (2004). Broadly speaking, the federal law at issue is the Confrontation Clause and the Supreme Court’s interpretation of it. That Clause provides “the accused [in a criminal prosecution] shall enjoy the right … to be confronted with the witnesses against him.” U.S. CONST. amend. VI. “Testimonial statements of wit- nesses absent from trial [may be] admitted only where the de- clarant is unavailable, and only where the defendant has had a prior opportunity to cross-examine.” Crawford v. Washing- ton, 541 U.S. 36, 59 (2004). The state cannot introduce a report with testimonial conclusions into evidence without produc- ing the analyst who prepared the report. See Melendez-Diaz v. Massachusetts, 557 U.S. 305, 308–09 (2009); Bullcoming v. New Mexico, 564 U.S. 647, 657–58 (2011). But it is less clear whether a state may allow an analyst to testify to his own conclusions about data another analyst collected. In Williams v. Illinois, 567 U.S. 50 (2012), the Court’s most recent and relevant Confrontation Clause decision, 1 the Court

1 The Supreme Court recently decided another Confrontation Clause

case that touches on the question here. Smith v. Arizona, 144 S. Ct. 1785 (2024). In that case, an analyst who did not testify collected some samples, tested them, and concluded that they tested positive for certain drugs. An- other analyst reviewed this report to reach his own, independent conclu- sion about what the samples were, but also testified to the substance of the other analyst’s report. Id. at 1795. The Court held that the testifying analyst testified to the truth of the other analyst’s report and remanded for the No. 22-2833 7

held that the Confrontation Clause was not violated when an analyst testified that swabs from a rape victim matched the defendant’s DNA. Id. at 61–62, 71. Like Leider here, the ana- lyst in Williams had not collected the swab, but reviewed the work of the analyst who did collect it. Id. at 61–62. From that review, she formed her own conclusion about whether the swab and the defendant’s DNA matched. See id. A fragmented Court decided Williams. Justice Alito wrote the plurality opinion joined by three other justices. He ex- plained that the original analyst’s report “was not to be con- sidered for its truth but only for the distinctive and limited purpose of seeing whether it matched something else.” 567 U.S. at 79 (quotations omitted). Justice Thomas, writing alone, agreed that the scheme did not violate the Confrontation Clause, but only as applied to Williams’s case. The original report, he explained, “lacked the requisite formality and solemnity to be considered testimonial for purposes of the Confrontation Clause.” Id. at 104 (Thomas, J., concurring) (quotations omitted). Justice Kagan, in dissent and joined by the remaining justices, saw no difference between introducing an unavailable analyst’s report and allowing another analyst

state court to determine whether the report was testimonial. Id. at 1799– 1802. That Supreme Court case does not affect our analysis, as the Wiscon- sin Court of Appeals issued its decision in 2014. See Greene v. Fisher, 565 U.S. 34, 38 (2011) (“[Section] 2254(d)(1) requires federal courts to focu[s] on what a state court knew and did, and to measure state-court decisions against this Court's precedents as of the time the state court renders its decision.” (quotations and emphasis omitted) (second alteration in origi- nal)). 8 No. 22-2833

to read that report and testify to her own conclusions. Id. at 125 (Kagan, J., dissenting). We do not conclude that Williams clearly established a rule that helps us decide whether to grant Roalson’s petition. “When a fragmented Court decides a case and no single ra- tionale explaining the result enjoys the assent of five Justices, ‘the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds[.]’” Marks v. United States, 430 U.S. 188, 193 (1977) (quoting Gregg v. Georgia, 428 U.S. 153, 169 n.15 (1976) (opinion of Stewart, Powell, and Stevens, JJ.)). If “a concur- rence that provides the fifth vote necessary to reach a majority does not provide a ‘common denominator’ for the judgment,” the rule set out in Marks “does not help to resolve the ultimate question.” United States v. Heron, 564 F.3d 879, 884 (7th Cir. 2009); see id. (listing cases). Justice Thomas’s concurrence and the plurality opinion do not share a “common denominator.” Id. Justice Thomas fo- cused on the formality of the original report, while the plural- ity opinion addressed how the report was presented at trial. It is true that Justice Thomas “share[s] the dissent’s view of the plurality’s flawed analysis.” Williams, 567 U.S. at 104 (Thomas, J., concurring). But “under Marks, the positions of those Justices who dissented from the judgment are not counted in trying to discern a governing holding from di- vided opinions.” Gibson v. Am. Cyanamid Co., 760 F.3d 600, 620 (7th Cir. 2014). For one, Marks is expressly limited to the jus- tices who “concurred in the judgments.” 430 U.S. at 193 (quo- tations omitted); see Gibson, 760 F.3d at 620. Further, as this court explained in Gibson, “the dissenters have disagreed with the plurality and the concurrence on how the governing No. 22-2833 9

standard applies to the facts and issues at hand []even if there is agreement” on some other issue. 760 F.3d at 620. We are not the only court scratching its head at Williams. Other circuits have applied Marks to Williams and been left wanting for clarity, United States v. Duron-Caldera, 737 F.3d 988, 994 & n.4 (5th Cir. 2013); Garlick v. Lee, 1 F.4th 122, 133 (2d Cir. 2021), including in the habeas context, Washington v. Grif- fin, 876 F.3d 395, 409 (2d Cir. 2017). Marks aside, neither this court nor the Supreme Court has suggested that Williams clearly established a rule. This court has applied Williams to reach a holding once. United States v. Maxwell, 724 F.3d 724, 727–28 (7th Cir. 2013). In that case, the defendant did not object to the Confrontation Clause issue and did not dispute the chemical makeup of the material tested. Id. at 727. The other two times this court has been pre- sented with the opportunity to apply Williams, it has assumed a Confrontation Clause violation and focused instead on the harmless error question. United States v. Turner, 709 F.3d 1187, 1194 (7th Cir. 2013); United States v. Garvey, 688 F.3d 881, 885 (7th Cir. 2012). The Supreme Court has also noted the lack of clarity sur- rounding Williams. Dissenting in Williams, Justice Kagan rec- ognized the “uncertainty” of the opinions of the plurality and Justice Thomas and stated that she believes the earlier cases “continu[e] to govern, in every particular, the admission of forensic evidence.” 567 U.S. at 141 (Kagan, J., dissenting). In- deed, the Court recognized the “muddle” Williams caused in lower courts just over a month after this case was argued, in Smith, 144 S. Ct. at 1794. 10 No. 22-2833

Recognizing Williams’s cloudiness, we are left with Melendez-Diaz and Bullcoming as the clearly established law governing Roalson’s habeas appeal. These cases hold that a state cannot introduce a report with testimonial statements into evidence without producing the analyst who prepared the report. Melendez-Diaz, 557 U.S. at 308, 329; Bullcoming, 564 U.S. at 658. Thus, the next question is whether the rule that the Wis- consin Court of Appeals applied is an unreasonable applica- tion of clearly established law. That rule, from Luther Williams, provides “one expert cannot act as a mere conduit for the opinion of another” and must instead “render[] her own ex- pert opinion.” Luther Williams, 644 N.W.2d at 926; see State v. Deadwiller, 834 N.W.2d 362, 377 (Wis. 2013) (applying rule). The testifying expert cannot be just anyone. The expert must be “highly qualified[,]” “familiar with the procedures at hand[,]” and must have “supervise[d] or review[ed] the work of the testing analyst.” Luther Williams, 644 N.W.2d at 926. That rule does not contradict Melendez-Diaz or Bullcoming. 2 To the contrary, it expressly prohibits a state from introducing an underlying report through testimony and requires that an

2 The Wisconsin Supreme Court implicitly reached the same conclu-

sion in State v. Deadwiller by holding the Luther Williams rule does not contradict Williams. In Deadwiller, an analyst testified at trial to his own conclusion that the defendant’s DNA matched a profile another analyst collected from semen from sexual assault victims. 834 N.W.2d at 365. The jury found the defendant guilty. Id. at 368. The court in Deadwiller dis- cussed Williams at length and concluded that the overlap among the facts of Williams, Deadwiller, and Luther Williams meant that the Luther Williams rule does not contradict Williams. Deadwiller, 834 N.W.2d at 375–77. No. 22-2833 11

analyst form an independent opinion and testify to that inde- pendent opinion. See Deadwiller, 834 N.W.2d at 370 n.7. So, as he must, Roalson argues that the state broke that rule and did introduce Gajewski’s report through Leider’s tes- timony. But this is not so. First, Roalson says Leider testified several times that she was “able to look at the … materials that were prepared by another analyst in the lab to reach [her] own conclusions.” A peer reviewer would have to look at the notes to locate the profile and conduct her own comparison. What matters is whether the peer reviewer testified to the original reviewer’s conclusions. Leider did not. Second, Roalson argues Leider testified to Gajewski’s pro- cess when asked: “[F]rom the review of the notes, does it in- dicate how the profiles were developed, meaning was there just one or were there … multiple?” Leider answered “yes” and explained, “[t]he knife was processed by swabbing the blade of the knife and the handle of the knife separately.” This is not a case where one analyst testified that another analyst “had followed standard procedures in testing the substances and that he reached the same conclusion based on the result- ing data that [the original analyst] had … .” Turner, 709 F.3d at 1191. Further, this case comes to us on habeas review, where we defer to a state court’s reasonable application of federal law. The Wisconsin Court of Appeals reasonably con- cluded, based on Leider’s testimony, that “[t]he opinions [she] reached on the basis of the materials she reviewed were her own.” State v. Roalson, 855 N.W.2d 492 (unpublished table de- cision) (Wis. Ct. App. July 15, 2014). Melendez-Diaz and Bull- coming do not establish that this kind of testimonial corollary is problematic. 12 No. 22-2833

Even if Roalson were correct that the state court commit- ted an error under § 2254(d) because the state introduced Gajewski’s report for its truth via Leider’s testimony, ample evidence supported his conviction, so any error did not have a “substantial and injurious effect or influence” on the verdict. Brecht v. Abrahamson, 507 U.S. 619, 637 (1993) (quotations omitted); Brown v. Davenport, 596 U.S. 118, 133 (2022). On ha- beas review, a reversible error must be harmful by more than a reasonable possibility. The court must find that the defend- ant was “actually prejudiced by the error.” Davis v. Ayala, 576 U.S. 257, 268 (2015) (quotations omitted); see Rhodes v. Dittmann, 903 F.3d 646, 665–66 (7th Cir. 2018). When the error is a deprivation of an opportunity to cross-examine, this court considers several factors: [T]he importance of the witness’ testimony in the prosecution’s case, whether the testimony was cumulative, the presence or absence of evi- dence corroborating or contradicting the testi- mony of the witness on material points, the extent of cross-examination otherwise permit- ted, and, of course, the overall strength of the prosecution’s case. Jones v. Basinger, 635 F.3d 1030, 1052 (7th Cir. 2011) (quotations omitted); see Rhodes, 903 F.3d at 666. These factors favor the state. Leider’s testimony was not the most important evidence for the state. Far more impactful was Davis’s testimony—which placed Roalson at the scene and provided Roalson’s statements in the moments following the murder—and Walczak’s testimony as to what Roalson confessed to her. Fingerprint evidence corroborated Leider’s testimony in part; forensic examiners found a fingerprint No. 22-2833 13

matching Roalson’s left pinky finger on one of the knives. And Roalson does not suggest that the court restricted his opportunity to cross-examine Leider. The prosecution sup- ported its charge that Roalson committed the murder by call- ing Davis and Walczak, who had not spoken to each other and yet testified that Roalson reported the same details to each of them. Therefore, even if the Wisconsin Court of Appeals was incorrect and the Constitution prohibited Leider’s testimony, the error did not have a “substantial and injurious effect or influence” on the jury’s verdict. Brecht, 507 U.S. at 637 (quota- tions omitted). III. The Wisconsin Supreme Court’s decision in Luther Wil- liams is in line with the United States Supreme Court’s hold- ings in Melendez-Diaz and Bullcoming. So, the court’s rule is a reasonable application of clearly established law. Even if there were an error, it is not substantial enough to justify releasing Roalson. The district court’s judgment denying Roalson’s pe- tition for a writ of habeas corpus is therefore AFFIRMED.