State of Wisconsin, Plaintiff-Respondent, v. Daniel J. Coughlin, Defendant-Appellant. (2026)

¶1 Daniel J. Coughlin appeals an order denying his motion for postconviction relief following convictions on five counts of child sexual assault after a jury trial. In his postconviction motion, Coughlin claims that his trial counsel was constitutionally ineffective in multiple respects, requiring that the postconviction court vacate the judgment of conviction on all five counts.

¶2 In an earlier appeal addressing rulings by the postconviction court, this court affirmed a portion of the order denying the postconviction motion, but we concluded that Coughlin was entitled to a Machner hearing in which he could attempt to show that trial counsel was ineffective.¹ State v. Coughlin, No. 2021AP1416-CR, unpublished slip op. (WI App March 2, 2023) (hereafter, “2021AP1416-CR”), petition for review denied, 411 Wis. 2d at 689 (WI August 17, 2023). Following remand, the postconviction court held a Machner hearing and the court denied the postconviction motion in its entirety. Coughlin appeals.

¶3 We conclude that Coughlin has shown that the combined effects of deficient acts or inactions of trial counsel create a reasonable likelihood that, but for counsel's deficiencies, the result of the proceeding would have been different. Accordingly, we reverse that portion of the order denying Coughlin's motion for a new trial and remand with directions that the circuit court vacate the judgment of conviction.²

BACKGROUND

¶4 The State charged Coughlin with sexually assaulting two child victims during the period 1989 through 1992. 2021AP1416-CR, ¶1. The allegations were brought to the attention of authorities about 20 years later. Id., ¶7. There were three trials on the same charges, but the first two ended in mistrials, while the third in 2015 resulted in the guilty verdicts that Coughlin challenges. Id., ¶1. After the third trial, appellate counsel for Coughlin pursued a direct appeal, raising a double jeopardy argument that the case should have been dismissed with prejudice after the first mistrial. Id., ¶¶1, 20 (referring to the appeal in State v. Coughlin, No. 2017AP1394-CR, unpublished op. and order (WI App June 13, 2018)). This court rejected that argument as meritless in a summary order issued in 2018. Id. Our supreme court denied Coughlin's petition for review of that order. Coughlin, No. 2017AP1394-CR, petition for review denied, 384 Wis. 2d at 773 (WI October 9, 2018).

¶5 Represented by new counsel, Coughlin brought the postconviction motion under Wis. Stat. § 974.06 (2023-24) that is at issue in this appeal.³ He claims constitutional violations, namely, multiple instances of ineffective assistance of counsel. The postconviction court denied the postconviction motion, without conducting a Machner hearing, and Coughlin appealed. 2021AP1416-CR, ¶¶15-16. We affirmed some specific rulings of the postconviction court, but we reversed its order denying Coughlin's postconviction motion.⁴ Id., ¶¶2-6, 115, 118. We remanded for a Machner hearing, which the postconviction court held in March 2024. Id. After the Machner hearing, the postconviction court denied the postconviction motion in its entirety, and Coughlin now appeals.

¶6 We proceed under the assumption that the reader of this opinion is familiar with our lengthy prior opinion in 2021AP1416-CR, issued in March 2023. This eliminates the need for us to reproduce here significant portions of that opinion covering many details that are relevant to this appeal. Instead, we cite to that opinion only as needed to generally explain our conclusions.

¶7 In this appeal, we address only Coughlin's claims of ineffective assistance against his trial counsel, and not his additional claims of ineffective assistance by the attorney who represented him in the direct appeal. Coughlin accurately points out that, at the Machner hearing, appellate counsel testified that he had no strategic reasons or other reasonable grounds for any of his relevant actions or inactions in handling the direct appeal.⁵ Further, the State concedes that Coughlin's claim of ineffective assistance by appellate counsel “rises and falls with trial counsel's performance”—in other words, if trial counsel was constitutionally ineffective, then appellate counsel was also ineffective in the first appeal for not arguing that trial counsel was ineffective. Accordingly, we do not address appellate counsel's performance further in this opinion.

¶8 Summarizing broadly, in 2021AP1416-CR we affirmed a portion of the circuit court's order denying the postconviction motion, but we concluded that a Machner hearing was required. This was because Coughlin successfully identified the following three categories of ineffective assistance claims that either appeared to establish prejudicial error or that might establish prejudicial error depending on further relevant evidence.

● Unanimity. Jury instructions regarding Counts 1-4 (the counts involving the older of the two victims) that did not clearly require unanimous jury verdicts on issues that require unanimity. 2021AP1416-CR, ¶¶26-58.

● “Intimate parts” omission. A jury instruction regarding Count 5 (the count involving the younger victim in the calf-pen incident) that erroneously omitted reference to the intentional touching of “the complainant's or defendant's intimate parts.” Id., ¶¶11, 59-60, 73-84.

● Four additional topics. The prosecution's introduction and use of a letter containing prejudicial material; references to conduct of, and child sexual assault convictions of, Coughlin's brother, Donald Coughlin; and the failure to impeach witnesses in two areas. Id., ¶¶85-117.6

¶9 At the Machner hearing, trial counsel testified in pertinent part as follows on these issues.

Unanimity

¶10 As explained in our prior opinion, Coughlin argues that three flaws in the jury instructions deprived him of the right to a unanimous verdict for each of Counts 1-4: ambiguity about the modes of sexual assault alleged; the fact that three separate types of acts of sexual assault were alleged for each count; and the trial court's modification of Wis JI-Criminal 517. 2021AP1416-CR, ¶¶37-58.

¶11 Explaining more fully, the first two flaws are similar in nature, but they are distinct. The first involves disjunctive modes of sexual assault. This was the failure to clearly inform jurors that they needed to unanimously agree, for each of the four counts, whether Coughlin had sexual contact or instead sexual intercourse with the older victim, even though there was testimony about both forms of sexual assault during pertinent time periods. Id., ¶¶31, 38-41. The second flaw involves a failure to clearly inform jurors that they needed to unanimously agree on a specific act involving sexual contact or sexual intercourse that Coughlin committed with respect to the older victim. Id., ¶¶31, 42-46. The third flaw is the trial court's omission of a portion of Wis JI-Criminal 517, which we concluded permitted the interpretation that individual jurors had to be certain only that the State had proven that a single act occurred, even if one or more other jurors were persuaded by evidence bearing on a distinct act. See 2021AP1416-CR, ¶¶47-58.

¶12 With that background, trial counsel testified at the Machner hearing that he raised an issue with the trial court on the topic of disjunctive instructions about sexual contact and sexual intercourse, in what would have been an off-the-record discussion.⁷ Counsel further testified that the court rejected whatever specific point counsel made. Counsel testified in pertinent part,

My vague recollection of it is that [when counsel for both sides and the trial court] were in [the trial court's] chambers, we talked about this, and that for whatever reason it was dropped after that. And I believe it's because [the court] said [it] was going to include it in the disjunctive. But ․ I don't specifically recall that.

Counsel testified that he could not recall the specific nature of his objection, which “for whatever reason” counsel “dropped,” and that counsel also could not recall any rationale given by the trial court in making whatever ruling it made. When specifically asked if he challenged any instruction regarding the need for jury unanimity, counsel responded, “I remember having some discussion with” the court. Counsel testified that he was “certain that we had a discussion about it,” but that he had no memory of taking any steps to make a record of that discussion.

¶13 Trial counsel further testified that he did not recall whether he noticed the modifications to pattern jury instructions that the trial court made that contributed to ambiguity about whether juror unanimity was required. He also did not dispute that the trial court's modifications could have made it more likely that the jury would find Coughlin guilty based on non-unanimous findings about specific acts, and counsel did not assert any strategic basis for failing to address this issue.

¶14 In responding to questions on this topic, as with other topics, trial counsel emphasized that his overarching trial strategy was to try to persuade the jury that the incriminating statements made by the two victims were false in all respects. “So when you talk about ․ carving up these little pieces of the jury instructions, these really don't affect what the strategy of ours was. And that would be that this never happened and these people were lying.” Even after it was pointed out to counsel at the Machner hearing how Wis JI-Criminal 517, as modified by the trial court, is ambiguous regarding unanimity, he testified that he could see no reason to object to the instruction.

¶15 The postconviction court appeared to credit trial counsel's testimony that counsel raised some form of objection involving unanimity, and that counsel's point was rejected in some manner by the trial court. While discussing this topic, the court also observed that, in general, trial attorneys often choose to avoid the risk of irritating a trial judge as can happen when the attorney renews an objection that the judge has already denied.

¶16 The court observed further that trial counsel testified that this instructional error was not “a big factor” to counsel at the time of trial, because counsel's efforts were primarily devoted to the overarching defense theory that all of the incriminating statements by the victims were false. The court said that the instructions “may have confused one or two jurors that they had to find intercourse” in different “ways.” At the same time, the court credited counsel's testimony that counsel believed, at the time of trial, that the trial court “would never allow a non-unanimous verdict,” and that counsel personally “believed that the jury instructions showed that the jury needed to be unanimous.”

¶17 The postconviction court did not make a determination, or make any factual finding, on the prejudice prong of the test that was specific to the issue of juror unanimity. Instead, in a summary fashion the court said that Coughlin had not shown prejudice on any of his ineffective assistance claims.

“Intimate parts” omission

¶18 As stated above, the issue here involves the fact that a jury instruction addressing Count 5 erroneously omitted reference to the intentional touching of “the complainant's or defendant's intimate parts.” See 2021AP1416-CR, ¶¶59-60, 73-84.

¶19 When trial counsel was asked at the Machner hearing whether he realized at trial that this language would be, or had been, omitted from the instruction, counsel testified, “I don't know if I did or not. But again, I will tell you that that didn't really affect our defense, because our position was that [the alleged offense involving the calf or calves] never happened.” Counsel further testified that he did not make a strategic decision to advocate for, or permit without objection, the omission of that language.

¶20 The postconviction court did not make any clear finding or determination regarding the Machner testimony on this issue, aside from briefly suggesting the views that this issue involves mere “semantics,” and that any error did not undermine trial counsel's overarching defense theory.

Four additional topics

¶21 Our prior opinion included the following overview of the four additional topics:

Coughlin argues that, in addition to the topics addressed above, trial counsel was also constitutionally ineffective when he: (1) did not properly address the prosecution's introduction and use of a letter containing what Coughlin submits were provably false allegations that Coughlin molested or “probably” molested more than three other children, in addition to molesting the [alleged victims in this case]; (2) failed to object to prejudicial evidence and arguments regarding conduct of and child sexual assault convictions of Coughlin's brother, Donald Coughlin; (3) failed to impeach witnesses called by the prosecution with evidence supporting the defense theory that the mother of the [victims] prompted her sons to falsely accuse Coughlin; and (4) failed to impeach the older [victim] regarding “drastic differences in the number and location of assaults [against the older victim] alleged [by the older victim] over time.”

2021AP1416-CR, ¶85; see also id., ¶¶88-117 (identifying and then discussing each issue in more detail). We now summarize the pertinent Machner testimony on each issue in turn.

The letter

¶22 Trial counsel testified that he consulted with Coughlin about concerns that prejudice to the defense could result from allowing the jury to read the June 19, 2008 letter from the mother of the victims to Coughlin. Counsel also testified, however, that Coughlin urged counsel to allow the jury to read the letter. Counsel did not dispute that the letter alleges in direct, accusatory terms that Coughlin not only sexually assaulted the two victims in this case, but also three to five other children. Counsel also testified that he was concerned about the potential prejudicial effects that this might have on jurors. Counsel further testified that he made a strategic decision that the possibility of prejudice was outweighed by counsel's hope that the letter would be understood by jurors as a “crazy” attempt by the mother to “blackmail” Coughlin into backing down in a church-related dispute that Coughlin was having with the mother. If the jury took this view, counsel believed that this might lend credence to the defense theory that the mother, animated by her anger over her church dispute, encouraged the victims to falsely accuse Coughlin with the allegations underlying this case.

¶23 Trial counsel testified that he “may have” considered moving the trial court to order redactions of the more prejudicial aspects of the letter—while still supporting the defense theory that the mother encouraged false allegations against Coughlin—but counsel did not explain what he thought or did at the time of trial based on his possible consideration of this issue. Counsel suggested that he did not affirmatively try to disprove the allegations in the letter that Coughlin had sexually assaulted victims other than the two alleged victims in this case, in part because Coughlin did not want counsel to bring the other alleged victims into “the middle of this.”

¶24 The postconviction court concluded that the strategy testified to by trial counsel regarding the letter was a reasonable strategy.

References to Donald Coughlin

¶25 Trial counsel acknowledged that the prosecutor, without objection, informed the jury that: the victims in this case were sexually, physically, and emotionally abused by Donald Coughlin, who was Coughlin's brother and the stepfather of the victims; Donald Coughlin started sexually assaulting the older victim when he was seven years old; Donald Coughlin's assaults of the older victim occurred in a residence, a vehicle, and a fire station; and Donald Coughlin caused the younger victim to masturbate Donald Coughlin and Donald Coughlin masturbated the younger victim.⁸ In addition, counsel acknowledged that the prosecutor, again without objection, made references to the jury that could reasonably be interpreted as suggesting that Coughlin and Donald Coughlin operated as joint or conspiring abusers. Further, the jury in this case learned that Donald Coughlin had recently been convicted of child sexual assaults in the separate case, without any record being made of the substance of a defense objection.

¶26 Trial counsel testified that he recognized at the time of trial that it could increase Coughlin's chances of being found guilty if jurors believed that Donald Coughlin had in fact sexually assaulted the same victims as in this case. There was “obviously [a] risk,” counsel testified, that jurors would infer that, if these same witnesses told the truth when they said that Donald Coughlin sexually assaulted them, then the victims were also telling the truth when they said that Coughlin sexually assaulted them. Counsel further testified that he had recognized a risk that the State might persuade the jury that Donald Coughlin and Coughlin, having grown up together and having “both been abused as children,” both grew up to sexually abuse children as adults.

¶27 Trial counsel testified to the “belief”—although he did not “know” this to be a “fact”—that he made a substantive objection, during an off-the-record sidebar conference, to the trial court allowing the jury to learn that Donald Coughlin had been convicted of sexually assaulting the victims, and he also testified that the trial court overruled this objection. As discussed in more detail below, the record reflects that: counsel failed to object to the State's pretrial motion to allow the jury to learn of Donald Coughlin's convictions; despite that fact, counsel initially objected when the prosecutor raised the issue with a witness; and there was then a sidebar conference, but there is no record of any arguments or of any ruling by the trial court as a result of the sidebar.

¶28 Trial counsel testified that his approach to issues surrounding Donald Coughlin was influenced in part by counsel's belief that some Juneau County jurors would arrive at the courthouse for Coughlin's trial having, in counsel's words, “heard about” the “generally well known” allegations against Donald Coughlin. One consequence of counsel's approach, counsel testified, was that he did not file a pretrial motion in limine to preclude or limit evidence related to Donald Coughlin. Postconviction counsel asked trial counsel why trial counsel's concern about some jurors being familiar with allegations of criminal conduct by Donald Coughlin, together with the danger of prejudice to Coughlin that could arise from references to Donald Coughlin's criminal conduct, did not provide all the more reason to pursue a motion seeking to limit references to Donald Coughlin. Trial counsel responded that such a motion “wouldn't erase the fact that some of the jurors knew about Donald.”

¶29 Trial counsel also testified that he tried to take advantage of evidence that could support a theory that the victims wanted to spend time at Coughlin's farm (during periods in which, according to the State, Coughlin sexually assaulted them). This was because, according to this defense theory, Coughlin's residence represented a refuge from abuse by Donald Coughlin—that is, “Daniel's house was a safe place for” the victims to spend their summers. Consistent with this “safe place” theory, the trial court overruled an objection by counsel to questions that the prosecutor posed to the older victim about abuse by Donald Coughlin, on the ground that testimony by the older victim on this topic was relevant to the older victim's motivations to spend time at Coughlin's farm.

¶30 Trial counsel asserted in his testimony that, given the way that the evidence and arguments came in at trial, he did not believe that jurors confused Coughlin with Donald Coughlin. “I think ultimately it was pretty clearly established that Don and Daniel were different.” At the same time, counsel acknowledged that he did not consider the possibility of objecting at trial to specific details regarding the assaultive conduct by Donald Coughlin that might have risked inflaming or misleading the jury about Coughlin's alleged conduct, while allowing limited evidence consistent with the defense safe-place theory. That is, counsel did not consider asking the trial court to allow the jury to be exposed to only a more limited, less prejudicial amount of evidence of abuse by Donald Coughlin that was still sufficient to show that Donald Coughlin was feared by the victims to the point where the victims sought out Coughlin's residence as a refuge from abuse.

¶31 The postconviction court credited trial counsel's testimony to the effect that counsel saw no point in trying to limit references to Donald Coughlin because some of the jurors selected to serve in this case would likely have been aware of at least some of the allegations against Donald Coughlin. Further, the court attributed to counsel a reasonable strategy to try to portray Donald Coughlin as “a scapegoat”—that is, to pursue a strategy of encouraging the jury to think that it was only ever Donald Coughlin, and never Coughlin, who sexually assaulted any children.

Motives to Lie: Potential Impeachment and Evidence Regarding Defense Theory that the Victims’ Mother Caused the Older Victim to Falsely Accuse Coughlin

¶32 We summarized in 2021AP1416-CR Coughlin's three closely related arguments in support of his position regarding the significance of evidence related to the motives for various people to make false allegations against him. Coughlin's argument is that trial counsel was deficient in failing to attempt to impeach the victims’ mother and the older victim, and to introduce related evidence, with the goal of bolstering the defense theory that the mother encouraged the older victim to falsely accuse Coughlin. See 2021AP1416-CR, ¶¶102-111. Coughlin contends that, through this deficient performance, counsel doomed any chance for the defense to convince the jury that the mother encouraged the older victim to fabricate the sexual assault allegations against Coughlin in 2008 in connection with the church dispute.

¶33 When trial counsel was asked a series of questions on this topic, he initially testified consistently that he was “not sure” whether he had considered these issues, and that he could not recall relevant details. In later testimony, counsel shifted ground, testifying that he “[p]robably” considered impeaching the older victim about the timing of the older victim's first disclosure to the mother, which is one of the key topics here. But even then, his only testimony about a related strategy was to reemphasize his overarching theory that the sexual assaults charged in this case “never happened.”

¶34 On this topic, the postconviction court characterized trial counsel as having testified that counsel “wanted to prove that there was a conspiracy by [the victims’ mother] against” Coughlin, and that Coughlin “firmly” expressed to counsel, in then-confidential conversations, Coughlin's belief that there had been such a conspiracy against him.

Potential Impeachment Regarding Inconsistent Victim Statements

¶35 We summarized in 2021AP1416-CR Coughlin's basis for arguing that trial counsel performed deficiently by failing to impeach the older victim regarding his inconsistent statements pertaining to the charged assaults, and we characterized the inconsistencies as “significant.” See 2021AP1416-CR, ¶¶112-14.

¶36 Trial counsel testified that it was “absolutely” part of his strategy to challenge the credibility of the victims based on prior inconsistent statements. Consistent with that, he testified that he was “sure” that he considered impeaching the older victim based on inconsistencies in the accounts that the older victim gave about the alleged assaults. But, when asked why he did not impeach the older victim to a greater extent on these topics, counsel repeatedly testified that he could not recall why he did not do so. In later testimony, counsel replied “No” in response to each of the following questions: (1) “[Postconviction counsel] asked you some questions about impeaching [the older victim] on the frequency of the assaults. Would the frequency of assaults have impacted your theory of defense?”; and (2) “What about the location? Would the location of the assaults have impacted your theory of the defense?” Along similar lines, trial counsel testified that his strategy “wasn't so much” to undermine the victims’ specific testimony about “what happened and how it happened,” but instead to generally challenge the victims’ “credibility” and to raise doubts with the jury that any such things had “ever happened.”

¶37 Trial counsel also suggested that he was leery about attempting to impeach the victims extensively or aggressively, because this might turn the jury against the defense, out of sympathy for the victims. Counsel further suggested that this general concern, common in cases of this type, was amplified here because counsel interpreted the facial expressions or the body language of a subset of the jury to be “very hostile” to the defense.

¶38 The postconviction court did not make any clear findings or determinations specific to this topic.

DISCUSSION

I. Legal standards

¶39 A defendant shows ineffective assistance of counsel by establishing both that counsel's performance was deficient and that the deficient performance was prejudicial, and if the defendant fails to satisfy either prong of the test, the court need not address the other prong. State v. Pinder, 2018 WI 106, ¶26, 384 Wis. 2d 416, 919 N.W.2d 568 (citing Strickland v. Washington, 466 U.S. 668, 687 (1984)).

¶40 “ ‘To establish that counsel's performance was deficient, the defendant must show that it fell below ‘an objective standard of reasonableness.’ In general, there is a strong presumption that trial counsel's conduct falls within the wide range of reasonable professional assistance.’ ” Id. (quoted source omitted). A reviewing court can determine that defense counsel's performance was “objectively reasonable,” even if counsel offers no sound strategic reason for the challenged action or inaction. State v. Honig, 2016 WI App 10, ¶¶24, 28, 366 Wis. 2d 681, 874 N.W.2d 589.

¶41 In assessing the prejudice prong of the Strickland test, the defendant can prevail only by showing that it is “reasonably likely” that the result would have been different. Harrington v. Richter, 562 U.S. 86, 111-12 (2011). “This does not require a showing that counsel's actions ‘more likely than not altered the outcome,’ but the difference between Strickland’s prejudice standard and a more-probable-than-not standard is slight and matters ‘only in the rarest case.’ ” Id. (quoted source omitted). “The likelihood of a different result must be substantial, not just conceivable.” Id.

¶42 The determination of whether a defendant in a criminal case was denied effective assistance of counsel presents issues of fact and law. Pinder, 384 Wis. 2d 416, ¶26. When relevant facts, including those involving counsel's conduct and strategies, are found by the circuit court, then a reviewing court will rely on those findings of fact unless they are shown to be clearly erroneous. Id. But whether counsel rendered ineffective assistance presents an issue of law that we review de novo. Id. That is, putting aside findings of fact articulated by a circuit court, a reviewing court accords no deference to a circuit court's “ ‘ultimate determination of whether counsel's performance was deficient and prejudicial to the defense.’ ” State v. Thiel, 2003 WI 111, ¶¶22-24, 264 Wis. 2d 571, 665 N.W.2d 305 (declining invitation to modify the standard of review to “accord some degree of deference to a trial judge's assessment of counsel's deficient performance and the prejudicial effect of counsel's errors,” based on the “unique vantage point” that circuit courts have on relevant evidence and on the conduct and demeanor of witnesses, counsel, and members of the jury); see also id., ¶¶26-32, 46 (overturning convictions based on facts found by the circuit court and de novo determination by appellate court that trial counsel provided ineffective assistance).

II. Law of the case

¶43 The parties dispute the implications of some of our conclusions in 2021AP1416-CR about Coughlin's ability to show prejudice on various of his claims, based on the record as it existed before the Machner hearing, the allegations in Coughlin's postconviction motion, and the specific arguments advanced by the parties in the prior appeal. But we now explain why this dispute does not matter to our resolution of this appeal.

¶44 Coughlin acknowledges that this court in 2021AP1416-CR could not conclusively determine that he had shown prejudice for purposes of ultimately resolving any of his claims of ineffective assistance of trial counsel in advance of a Machner hearing, at which evidence could be adduced about the corresponding actions or inactions of counsel that allegedly constituted deficient performance. See State v. Sholar, 2018 WI 53, ¶54, 381 Wis. 2d 560, 912 N.W.2d 89 (“For purposes of determining whether counsel was ineffective, prejudice cannot exist without being attached to an error on the part of counsel.”). Bearing the concept from Sholar in mind, Coughlin contends that this court in 2021AP1416-CR essentially made permissible conditional rulings to the following effect. Given the inherently prejudicial nature of some of the claimed deficient performance here, if the Machner hearing were to reveal deficient performance, then Coughlin would show both deficient performance and prejudice on those claims. Further, Coughlin argues that this conditional ruling should be treated as law of the case, binding in this appeal. See State v. Casteel, 2001 WI App 188, ¶15, 247 Wis. 2d 451, 634 N.W.2d 338 (“A decision on a legal issue by an appellate court establishes the law of the case that must be followed in all subsequent proceedings in the case in both the circuit and appellate courts.”).

¶45 Based on the concept in Sholar that we have just referred to, the State is critical of some of our statements in 2021AP1416-CR regarding Coughlin's showings of prejudice on various of his ineffective assistance claims in advance of the Machner hearing. The result, the State argues, is that we should not apply the law of the case doctrine to any statements in our prior opinion to the effect that Coughlin definitively or automatically satisfied the prejudice prong of the test on some of his claims.

¶46 In favor of the State, we assume without deciding that, regardless how these issues were briefed by the parties in 2021AP1416-CR or the precise wording used in our prior opinion, the law of the case doctrine does not apply to any such statements in our opinion. All conclusions stated in this opinion are based on our current consideration of the record as a whole, including the Machner hearing.

III. Unanimity

Deficient performance

¶47 We apply, as we must, a highly deferential mode of evaluation on all deficiency issues, including the unanimity issues. See State v. Carter, 2010 WI 40, ¶22, 324 Wis. 2d 640, 782 N.W.2d 695 (“In evaluating the reasonableness of counsel's performance,” reviewing court “must be ‘highly deferential.’ ” (quoting Strickland, 466 U.S. at 689)). Even when we do so, however, we conclude that under the totality of the circumstances as they existed at the time of trial, including as reflected in trial counsel's testimony from the Machner hearing, counsel performed deficiently when he failed to reasonably address the multiple problems that could have contributed to the jury returning verdicts on Counts 1-4 that were not based on unanimous votes.

¶48 To begin, Coughlin had a Sixth Amendment right to unanimous verdicts, as well as a Fifth Amendment due process right to verdict specificity, and, because he was charged with multiple crimes, a constitutional right to have a unanimous jury determination of which act resulted in which count in the jury's verdict. See State v. Marcum, 166 Wis. 2d 908, 912, 919, 480 N.W.2d 545 (Ct. App. 1992).

¶49 We review jury instructions “as a whole to determine whether they fully and fairly convey the applicable rules of law to the jury.” Dakter v. Cavallino, 2015 WI 67, ¶87, 363 Wis. 2d 738, 866 N.W.2d 656. We explained in our prior opinion why the only relevant instructions, as they were given to the jury here, were defective, and nothing in the Machner hearing changes our view on that topic. See 2021AP1416-CR, ¶¶26-31, 37-58.

¶50 With those points as background, Coughlin has shown that, if trial counsel had made judgments based on the exercise of reasonable competence, counsel would have recognized the three unanimity problems and he would have taken effective steps to address them. See Carter, 324 Wis. 2d 640, ¶22 (counsel's representation is deficient only if it is shown to fall “ ‘below an objective standard of reasonableness’ considering all the circumstances” (quoting Strickland, 466 U.S. at 688)). Counsel had only to consider the pattern instructions in light of the nature of the evidence in this case, and then consider the way in which they were modified by the trial court, to see that jurors might easily misunderstand their obligation to return verdicts that were unanimous as to each mode of alleged sexual assault and each alleged act during a relevant time period, while acting as a jury as a whole, and not while acting as individual jurors. See Hinton v. Alabama, 571 U.S. 263, 274-75 (2014) (per curiam) (stating that it was deficient performance for trial counsel to commit an “inexcusable mistake of law” involving a failure to conduct the “basic research” that would have been necessary to understand that he could hire an adequate expert under state law); Cates v. United States, 882 F.3d 731, 736-37 (7th Cir. 2018) (stating that it was deficient performance for trial counsel to fail to recognize that a jury instruction on aggravated sexual assault misstated the law; “[t]here is no conceivable strategic reason for a defense lawyer to forgo a challenge to a prejudicial jury instruction”).

¶51 Our conclusion does not depend on “ ‘the distorting effects of hindsight,’ ” nor does it diverge from the particular “ ‘perspective at the time’ ” that trial counsel should have had if he acted based on an objective standard of reasonableness. Carter, 324 Wis. 2d 640, ¶22 (quoting Strickland, 466 U.S. at 689). The combined effect of these three unanimity problems should have been obvious from core trial materials that were (or should have been) squarely before counsel: the specific charges and how the prosecution planned to prove them through the statements of the victims; the pattern instructions; and the instructions as modified. Counsel did not need to conduct investigation or difficult legal research. Further, counsel needed only to take the elementary step of calling these problems to the trial court's attention. In sum, both the pertinent law and counsel's duties were clear.

¶52 This brings us to the issue of whether trial counsel was deficient in the specific actions he took or failed to take on this issue. To recap, the postconviction court credited counsel's testimony that he had some “discussion” with the trial court, apparently amounting to an objection on the topic of sexual contact as opposed to sexual intercourse, and that the trial court rejected whatever point or points counsel raised, albeit on an unknown basis. There is no record of the nature of any objection or how the trial court ruled.

¶53 This creates two possibilities, either of which establishes deficient performance by trial counsel on the unanimity issues.

¶54 One possibility is that trial counsel made a valid, sufficiently well-articulated objection out of the presence of the jury, which should have been granted by the trial court. In that case, even if the trial court denied the objection, counsel would have done all that could reasonably be expected of him, at least in terms of making the objection. But that is not the end of the analysis. Because, if that occurred, it was deficient for counsel not to make a record of that objection and the court's ruling, because counsel did not provide any reason, framed as strategic or otherwise, for failing to make such a record. And doing so would have been an obvious and simple task for counsel to undertake. If trial counsel had made such a record, it would have allowed Coughlin to successfully pursue relief on this issue in postconviction or appellate proceedings. That is, because only those objections that appear in the record can provide grounds for appellate review, counsel's failure to make a record left Coughlin without a basis to establish that the trial court erred in denying the objection, if that is what happened. See State v. Dowdy, 2012 WI 12, ¶5, 338 Wis. 2d 565, 808 N.W.2d 691 (“As a general rule, issues not raised in the circuit court will not be considered for the first time on appeal.”).

¶55 Trial counsel took the position at the Machner hearing that it was “everybody's responsibility” to create a record of counsel's objections and the resulting rulings by the trial court. Whatever responsibilities other system actors did or did not have, however, it remained one important responsibility of trial counsel. See State v. Wedgeworth, 100 Wis. 2d 514, 528, 302 N.W.2d 810 (1981) (“Counsel who rely on unrecorded sidebar conferences do so at their own peril.”). Counsel testified that he had no memory of taking any steps to make a record.⁹

¶56 The other possibility is that trial counsel failed to make a valid, sufficiently well-articulated objection. This would have been deficient performance, given the significant issue presented, and counsel's testimony that he had no reasonable strategy or other legitimate basis to justify failing to do so.

¶57 We observe that, for the following reasons, the most likely possibility appears to be that trial counsel failed to make a valid, sufficiently well-articulated objection, even if, as the postconviction court found, counsel did raise some form of objection. As noted, the postconviction court essentially found that counsel made some form of objection that was denied by the trial court, but the postconviction court did not made detailed findings revealing the nature of the objection or the ruling of the trial court. For example, counsel testified at the Machner hearing that he did not recall whether he noticed the modifications to instructions that the trial court made which contributed to ambiguity, suggesting that he paid insufficient attention to the proposed instructions. Further, counsel appeared to testify that his approach at trial was that, while he was confident that the trial court would ensure that the jury received the necessary instructions about unanimity, it did not matter to counsel precisely how the jury was instructed, because all that could matter was whether the jury accepted the overarching defense theory that all incriminating statements made by the two victims were false in all respects. He referred to the unanimity problems identified by Coughlin as being merely “little pieces of the jury instructions,” which counsel testified “really” did not “affect what the strategy of ours was,” namely, “that this never happened and these people were lying.” Indeed, even by the time of the Machner hearing, counsel expressed the view that Wis JI-Criminal 517, as modified by the trial court, is not ambiguous. The postconviction court found that counsel personally “believed that the jury instructions showed that the jury needed to be unanimous,” but we conclude that, counsel's beliefs aside, this was not the case, for the same reasons we have previously explained. 2021AP1416-CR, ¶¶47, 52-58.

¶58 It would be illogical to conclude that trial counsel was not deficient on this issue because counsel concentrated his efforts on the overarching defense theory that all of the incriminating statements by the victims were false. Counsel did not testify to, and the record does not reflect any, inconsistency between counsel presenting the overarching defense theory and his also taking adequate steps (outside the presence of the jury) to ensure that the jurors were clearly instructed that complete unanimity was required. This should have included taking the simple steps to show that a sufficient objection was made and that it was rejected by the trial court on a specific ground or grounds. Therefore, it would have been irrational for counsel to fail to address the unanimity issues on the grounds that doing so would have jeopardized the defense in some way. See State v. Breitzman, 2017 WI 100, ¶65, 378 Wis. 2d 431, 904 N.W.2d 93 (courts do not second-guess a reasonable trial strategy unless the strategy was “irrational” or “based upon caprice” instead of “judgment”); Cates, 882 F.3d at 736-37 (concluding that it was deficient performance to fail to object to a critical jury instruction contrary to law).

¶59 A contrary approach would imply a general rule that, so long as defense counsel in any case can identify a rational, overarching defense theory, then no other aspect of counsel's performance could be deficient. Courts are to consider “all the circumstances” in assessing whether deficiency has been shown, not solely whether a defense attorney has pursued a rational overall theory of the defense. See State v. Carter, 324 Wis. 2d 640, ¶22; Strickland, 466 U.S. at 688 (“In any case presenting an ineffectiveness claim, the performance inquiry must be whether counsel's assistance was reasonable considering all the circumstances.” (emphasis added)).

¶60 The postconviction court found that trial counsel believed, at the time of trial, that the trial court “would never allow a non-unanimous verdict.” Whatever the court precisely meant to convey, it is irrelevant under the objective test for deficient performance. Indeed, to the extent that counsel was not attentive to the unanimity issues arising from what we conclude were the self-evident, prejudicial ambiguities in the jury instructions identified above, based on counsel's confidence that the trial court would ensure that unanimity was adequately addressed in the instructions, this would not have been competent performance.

¶61 The State makes two categories of arguments on this issue, which we address in turn.

¶62 The State renews its arguments from the prior appeal to the effect that there were no unanimity problems with the jury instructions as modified by the trial court, and therefore trial counsel could not have been deficient in failing to adequately address the issue. We explained in detail in 2021AP1416-CR why we reached the opposite conclusion, and the State fails to persuade us with the benefit of the Machner hearing that this conclusion was wrong. For example, the State notes that we concluded that “one feasible interpretation” of the modified Wis JI—Criminal 517 instruction would be “that the jurors had to agree on a particular mode and act of sexual assault for each count,” and emphasizes in its arguments the reasons that jurors might have gone with that interpretation. See 2021AP1416-CR, ¶52. But we are not persuaded by the State's limited points attempting to counter our further point that “there is at least one other feasible interpretation, one that did not clarify that unanimity was required and in fact could have misled the jury to think that unanimity was not required, particularly in light of” two problems discussed elsewhere in that opinion and noted above. See id., ¶53. A competent attorney would have identified and adequately addressed the risk that jurors would rely on the second interpretation.

¶63 The State emphasizes that deficient performance is not necessarily shown simply because trial counsel fails to identify a strategy to support counsel's challenged action or inaction. See Roe v. Flores-Ortega, 528 U.S. 470, 481 (2000) (“The relevant question is not whether counsel's choices were strategic, but whether they were reasonable.”). With this emphasis, the State argues that counsel's testimony that he held the “belief that the jury was adequately instructed on unanimity” was sufficient to support the presumption that his performance was reasonable. We are persuaded by the reasoning of the federal circuit court of appeals that, as a general matter, a defense counsel's failure to object to a clearly prejudicial jury instruction is not reasonable performance, see Cates, 882 F.3d at 736-37, and we conclude that this general rule applies here.

¶64 As its second category of arguments, the State argues in the alternative that, even if there were unanimity problems, the law was unsettled and therefore trial counsel was not deficient in failing to identify a problem.¹⁰ See Breitzman, 378 Wis. 2d 431, ¶49 (the failure to raise arguments “ ‘that require the resolution of unsettled legal questions generally does not render a lawyer's services ‘outside the wide range of professionally competent assistance’ sufficient to satisfy the Sixth Amendment.’ ” (quoted sources omitted)). This is based on three opinions of our supreme court and one opinion of this court addressing unanimity issues, which the State asserts created especially difficult law for counsel to understand and apply. We now explain why the State does not persuade us that the relevant law was unsettled for purposes of determining whether counsel was deficient.

¶65 The earliest opinion among those cited by the State is State v. Lomagro, 113 Wis. 2d 582, 335 N.W.2d 583 (1983). In Lomagro, numerous acts of sexual violence were alleged to have occurred over a two-hour period and were presented to the jury as one count of sexual assault, which was deemed by our supreme court to have been a single encounter involving “one continuing criminal episode and properly chargeable as one offense.” Id. at 583-86, 598. As we explained in 2021AP1416-CR, based on these facts, our supreme court in Lomagro concluded that the circuit court did not err in failing to instruct the jury that unanimity was required regarding the specific act constituting charged sexual assault based on evidence of multiple acts that could constitute the offense, because “[i]f there is only one crime, jury unanimity on the particular alternative means of committing the crime is required only if the acts are conceptually distinct.”¹¹ Lomagro, 113 Wis. 2d at 592-93; see also 2021AP1416-CR, ¶40. This discussion in Lomagro could not have created unsettled law for current purposes, because the alleged acts in this case were conceptually distinct. The alleged acts here were separated by time and included allegations of various types of sexual intercourse and sexual contact.

¶66 We also referred in 2021AP1416-CR to the second opinion by date cited by the State, State v. McMahon, 186 Wis. 2d 68, 519 N.W.2d 621 (Ct. App. 1994). See 2021AP1416-CR, ¶40. We cited McMahon for the proposition that there was, at the time of McMahon, no Wisconsin case law concluding that the allegation of multiple sexual acts occurring over the course of more than one month could amount to a single, continuous transaction for unanimity purposes. McMahon, 186 Wis. 2d at 82; 2021AP1416-CR, ¶40. Again here, the State fails to develop a supported argument that McMahon created unsettled law in the instant case. This court explained in McMahon that, when the events described by a victim “occurred over a month and one-half,” this not could be considered “one continuous story with various chapters,” which is the scenario that would allow a jury to find guilt on a charged count of sexual intercourse even if all members have not agreed on the specific acts that satisfy the elements of that charge. McMahon, 186 Wis. 2d at 84. The applicable charging periods here were separated by considerably more time.

¶67 The third opinion by date cited by Coughlin is State v. Derango, 2000 WI 89, 236 Wis. 2d 721, 613 N.W.2d 833, which we also relied on in our prior opinion.¹² See 2021AP1416-CR, ¶¶34, 45. We relied on Derango for its summary of the requirement of juror unanimity, and also for its discussion of cases in which multiple modes of committing an offense are charged in each count, as opposed to being charged as multiple offenses. See 2021AP1416-CR, ¶¶34, 45. Derango also does not create unsettled law for current purposes. This is because the court in Derango determined that child enticement is one crime for unanimity purposes, with multiple modes of commission based on varying mental states. See Derango, 236 Wis. 2d 721, ¶¶17-25 (observing that Wis. Stat. § 948.07 (1999-2000), criminalized causing or attempting to cause a child to go into various types of locations with any of six prohibited mental states). That bears no relevant resemblance to elements of the charge here, child sexual assault in violation of Wis. Stat. § 948.02(2), because that crime requires proof of a single mental purpose for the act of sexual contact and no mental purpose for the act of sexual intercourse. See Wis. Stat. § 939.22(34) and (36).

¶68 Finally in this list, the State cites State v. Johnson, 2001 WI 52, 243 Wis. 2d 365, 627 N.W.2d 455. In this opinion, our supreme court discussed the degree to which Wis. Stat. § 948.025—which requires that a defendant committed three or more acts of sexual assault against the same child within a charging period—constitutes a “continuing course of conduct” crime, such that the requirement of jury unanimity is satisfied even when jurors need not all agree about which underlying act or acts constitute the crime. Johnson, 243 Wis. 2d 365, ¶17. The State fails to develop a supported argument that Johnson’s discussion of § 948.025 could have created unsettled law in this case, which involves charges of specific acts of child sexual assault, not one or more continuing courses of conduct.

Prejudice

¶69 In 2021AP1416-CR we explained, at length and in detail, why we conditionally concluded that, under the standards articulated in Strickland and Marcum, if trial counsel was deficient in addressing the unanimity issues, this caused our confidence in the outcome to be placed in doubt. See 2021AP1416-CR, ¶¶32-58. We explained why we concluded that the cumulative effects of the instructional errors could have deprived Coughlin of his right to unanimous verdicts with respect to Counts 1-4.

¶70 We now conclude that Coughlin has shown that trial counsel's deficiencies on the unanimity issues substantially contribute to a reasonable likelihood that the result on Counts 1-4 would have been different, based on the entire record, taking into account the testimony at the Machner hearing and the arguments that the State now advances. We do not conclude that the unanimity problems, standing alone, necessarily create that reasonable likelihood. But a reasonable likelihood arises from the combination with the other deficiencies by counsel, as discussed below, involving references made at trial to Donald Coughlin and significant inconsistencies in victim testimony. It is not merely a conceivable likelihood but a substantial one, given the nature of the errors.

¶71 The State does not attempt to suggest that trial counsel could not easily have requested more specific instructions that would have resolved these issues for Counts 1-4, fixing the modifications that Coughlin has identified as depriving him of his constitutional rights to verdict specificity. That distinguishes the situation here from the one in which a defendant claims prejudice based on deficient performance of defense counsel, but without providing an adequate picture of what counsel should or should not have done that could have changed the outcome of the trial.

¶72 As we explained in our prior opinion, this court has summarized the general standards regarding ineffective assistance in the context of a jury instruction error. See 2021AP1416-CR, ¶32 (quoting State v. Ziebart, 2003 WI App 258, ¶¶14-16, 268 Wis. 2d 468, 673 N.W.2d 369). Consistent with those standards, the analysis in persuasive authority that is also rooted in the Strickland analysis is helpful here. Specifically, we observe that the prejudice prong in this context has been helpfully refined by the Utah courts. When there is an instructional error resulting in a claim of ineffective assistance, those courts address the prejudice prong by asking the following: did the error create a possibility that the jury convicted the defendant based on factual findings that would not have led to conviction if a correct instruction had been given and, if so, is there a reasonable probability that at least one juror voted guilty based on those factual findings? State v. Grunwald, 2020 UT 40, ¶22, 478 P.3d 1. Here, there is a reasonable probability that jurors who contemplated different acts by Coughlin returned guilty verdicts based on those acts even though the jurors did not all agree that the State proved the same act, occurring at or around the same time, with proof beyond a reasonable doubt.

¶73 When we put to the side the State's arguments regarding prejudice that challenge application of the law of the case doctrine—to repeat, we assume that issue in favor of the State—what remains on this issue are the following points by the State.

¶74 The State places weight on State v. Becker, 2009 WI App 59, 318 Wis. 2d 97, 767 N.W.2d 585, but we now explain why that opinion does not support the State's argument.

¶75 In Becker, the defendant was charged with two counts of first-degree sexual assault of the same child, and he was found guilty on both counts. Id., ¶¶1-2. The criminal complaint alleged two distinct acts: touching the victim's vagina, and causing the victim to touch the defendant's penis. Id., ¶2. The criminal information also charged two counts, but it used identical language to describe the charges—that is, with language that did not distinguish between the two distinct acts—and the court repeated this language in instructing the jury. Id., ¶3. The verdict forms also failed to tie a particular act of sexual contact to each count. Id., ¶5. But, significantly, the jury was explicitly instructed, “Each Count charges a separate crime and you must consider each one separately.” Id., ¶24. In a note sent out during deliberations, the jury asked, “Does count one and count two correspond to the specific events? i.e., is one the vaginal contact and two the penis contact?” Id., ¶6. The circuit court answered, “No.” Id., ¶7.

¶76 The defendant in Becker argued that trial counsel was constitutionally ineffective in failing to recognize that, because the criminal information, the instructions, and the verdicts all failed to tie a particular act of sexual contact to a particular count, this created the possibility that the jury's verdicts were not unanimous. Id., ¶14. This court determined that the defendant could not have been prejudiced because the jury, in unanimously agreeing that the defendant was guilty on both counts, had to have unanimously agreed that he committed each of the two acts at issue. Id., ¶¶23-24. It did not matter how individual jurors voted with respect to each count, because they were instructed that each count charged a separate crime and therefore all jurors must have agreed that he committed both alleged acts. Id. Further, this court specifically rejected the defendant's argument that the victim had described three, not two, acts of sexual contact. Id., ¶¶25-26.

¶77 The State fails to identify any point made in Becker that could assist the State here on the prejudice issue related to unanimity. Indeed, the sharp contrast between the facts in Becker and the facts in this case serves to reinforce Coughlin's argument on prejudice. In Becker, there were just the two specific alleged criminal acts (touching the victim's vagina and causing the victim to touch his penis), but here, the victim testified to three separate types of assaultive conduct for each count of Counts 1-4, each of which allegedly occurred multiple times within each charging period, some of which were sexual contact and some of which were sexual intercourse. See 2021AP1416-CR, ¶¶8, 42.

¶78 The State also relies on this court's decision in State v. Gustafson, 112 Wis. 2d 369, 332 N.W.2d 848 (Ct. App. 1983) (Gustafson I), for its broad statements, made in the course of addressing the issue of harmlessness, based on the particular facts of that case and which do not include analogs to all of the issues we address in this opinion. See Gustafson I, 112 Wis. 2d at 380-81 (“There is no way the jury could have found from the evidence that the defendant committed the first set of assaults but not the second or vice versa.”). Gustafson I was modified on review by our supreme court in State v. Gustafson, 119 Wis. 2d 676, 350 N.W.2d 653 (1984) (Gustafson II). And, as our supreme court clarified in Gustafson II, that case, unlike this one, involved the allegation of one alleged offense committed in a compressed time period. Gustafson II, 119 Wis. 2d at 696 (“[A]s in Lomagro, the sexual assault” in Gustafson “may be viewed as a single, continuous criminal transaction that occurred within a short time period. Under these circumstances, the general unanimity instruction that the trial court gave in this case was sufficient to assure a unanimous verdict.”).¹³ The State does not persuade us that any aspect of Gustafson I controls the prejudice analysis here.

IV. “Intimate parts” omission

Deficient performance

¶79 Again applying the highly deferential mode of evaluation, we conclude that, under the totality of the circumstances as they existed at the time of trial, including as reflected in trial counsel's testimony from the Machner hearing, counsel performed deficiently when he failed to address the fact that the jury instruction regarding Count 5 erroneously omitted reference to the intentional touching of “the complainant's or defendant's intimate parts.” The postconviction court did not make any material findings of fact on this issue.

¶80 Here again, Coughlin has shown that, if trial counsel had made reasonable judgments and exercised reasonable competence, he would have recognized that omitting the key requirement of “sexual contact” as an element carried significant risks for the defense, and then he would have taken elementary steps to address that omission. As with the unanimity issues, the law and counsel's duty were both clear.

¶81 There can be no reasonable debate about whether trial counsel was deficient in the specific actions he took or failed to take on this issue, because he testified that he had no memory of even being aware of the omission and that he made no strategic decision on this issue. As with other issues, his position was that it did not matter whether this instruction could mislead or confuse the jury, “because our position was that [the alleged offense] never happened.” But here again, counsel failed to explain how avoiding the significant problem with this instruction, through discussion with the trial court out of the presence of the jury, would have been even remotely in conflict or tension with fully pursuing his overarching defense theory. To the contrary, allowing the omission of the intimate parts requirement from the sexual contact element only served to directly undermine the defense theory that there was no touching of the victim's intimate parts. It was not objectively reasonable for counsel to fail to address this issue.

¶82 In arguing against deficiency on this issue, the State emphasizes statements made by the prosecutor and trial counsel during closing arguments referring to the younger victim's penis in discussing the Count 5 allegations, which the State suggests eliminated the possibility that the jury could have failed to understand that the “sexual contact” in Count 5 depended on a calf or calves touching the younger victim's penis.

¶83 The State's point is reasonable as far as it goes. These arguments reduced to some degree the chances of confusion or error by the jury in applying the correct legal standard. But this would not have eliminated the danger that the instructional error contributed to potential confusion surrounding the “calf-pen incident,” and it remains that trial counsel should have taken note that the jury instruction, as given, did not require proof that the touching was of an intimate part. See Geise v. American Transmission Co., 2014 WI App 72, ¶26, 355 Wis. 2d 454, 853 N.W.2d 564 (courts presume that jurors follow the jury instructions as given by the circuit court). For example, the definition of “sexual contact” omitting that requirement, as provided by the trial court, required proof of intentional touching by the defendant, either directly or through clothing, by the use of any object. A juror faithfully applying that instruction to the allegations could have thought that all that was required was proof that, through intentional action by Coughlin, a calf or calves came close to or risked touching the victim's penis, including through his clothing, and that this is why counsel during argument referred to the younger victim's penis.

Prejudice

¶84 In 2021AP1416-CR, ¶¶73-84, we explained, again in detail, why we rejected the State's argument that the omission to the Count 5 instruction was not harmless error. The State had argued harmless error to show why appellate counsel was not ineffective in failing to raise this issue in the direct appeal. Applying the same reasoning now as we did then, based on the entire record and considering the arguments that the parties now advance, we conclude that Coughlin has shown that trial counsel's deficiencies on this issue created a reasonable likelihood that the result on Count 5 would have been different.

¶85 Applying the standards in Ziebart, and the helpful Utah framework cited above from Grunwald, for reasons we have already explained, any of the jurors presented with the evidence about the calf-pen incident could have voted guilty based on evidence that a calf or calves came close to or risked touching the victim's penis, including through his clothing. The State does not even attempt to dispute Coughlin's well-supported argument that the State's evidence on Count 5 was not strong and could have generated confusion for jurors.

¶86 The State criticizes the way we framed this issue in 2021AP1416-CR. But, when that framing is put to the side, the State fails to develop any argument that we did not address and reject in substance in our prior opinion based on reasoning to which we continue to subscribe. Coughlin has shown that it is reasonably likely, to a substantial degree, that the result would have been different if the jury had been correctly instructed. For example, the State does not attempt to break down the four separate reasons we provided that support a determination of prejudice, albeit as framed using harmless error terminology. See 2021AP1416-CR, ¶83.

¶87 Consistent with its argument on deficiency addressed above, the State contends that Coughlin cannot show prejudice on this issue because the parties disputed at trial whether Coughlin intentionally caused sexual contact with the younger victim's penis. But again, the instructional error was sufficiently stark to support a determination of prejudice given the nature of the evidence and arguments made by the parties at trial regarding the allegations in Count 5.

V. Four additional topics

A. Deficiency

¶88 We conclude that Coughlin has failed to show deficient performance on the following topics: the letter; Donald Coughlin's so-called “other acts” that did not involve references to joint abusive conduct of Donald Coughlin and Coughlin or to the fact that Donald Coughlin had been convicted of sexually assaulting the same victims; and the evidence supporting a motive for the victims to lie. Therefore, these issues could not have directly contributed to any determination of prejudice. See Thiel, 264 Wis. 2d 571, ¶61.

¶89 Separately, we conclude that Coughlin has shown deficient performance in the following ways: failing to prevent or object to statements by the prosecutor to the jury that could reasonably be interpreted as describing joint, abusive conduct by Donald Coughlin and Coughlin; failing to make, or at a minimum failing to create a sufficient record of, an objection to the jury being informed that Donald Coughlin had been convicted of sexually assaulting the same victims; and failing to impeach the older victim regarding significant inconsistencies in his account. Therefore, these issues contribute to the prejudice determination.

¶90 Because they have overlapping elements, we begin with the two sets of Donald Coughlin-related issues. Then we address the remaining issues on which we conclude that Coughlin has not shown deficiency, followed by those on which we reach the opposite conclusion.

¶91 The following two sets of topics related to Donald Coughlin have at times been intertwined when referred to by the parties on appeal and at the Machner hearing, but we conclude that the record calls for a line to be drawn between them.

¶92 One set of topics related to Donald Coughlin, on which we conclude Coughlin fails to show deficient performance, involves references to which the jury was exposed regarding the specifics of criminal conduct engaged in by Donald Coughlin without any involvement by Coughlin, excluding reference to Donald Coughlin's convictions for sexually assaulting the same victims. Coughlin now categorizes this set of topics as “other acts” evidence. See supra note 8.

¶93 The other set of topics related to Donald Coughlin, which we conclude were the subjects of deficient performance, as explained below, involve (1) statements by the prosecutor that could reasonably be interpreted as explicitly describing the brothers as criminal co-actors who victimized the same children in the same ways; and (2) allowing the jury to be informed that Donald Coughlin had recently been convicted in a court of law of sexually assaulting the same two victims as in this case.

Donald Coughlin's “other acts” not directly involving Coughlin or the convictions

¶94 Regarding the first set of topics, we conclude that the record does not reveal that trial counsel took an approach that no competent lawyer would have chosen. See Dunn v. Reeves, 594 U.S. 731, 739 (2021) (“[E]ven if there is reason to think that counsel's conduct ‘was far from exemplary,’ a court still may not grant relief if ‘[t]he record does not reveal’ that counsel took an approach that no competent lawyer would have chosen.” (quoted sources omitted)). At the Machner hearing, counsel articulated a pertinent strategy, which was credited by the postconviction court, and which we cannot say was irrational or based on mere caprice.

¶95 This strategy involved two premises and a goal, all of which we conclude fell within the broad zone of rational understandings and choices. One premise was that at least some jurors would have a degree of familiarity with allegations about Donald Coughlin's criminal conduct, and that no amount of care taken during jury selection or through motions in limine could prevent that. Another premise was that the jury could keep track of references to Donald Coughlin as distinct from references to Coughlin. The goal was to turn this situation into an advantage for Coughlin, by allowing some evidence showing that Donald Coughlin abused the victims and trying to convince the jury that the victims were abused only by Donald Coughlin, and that they sought refuge with Coughlin to escape Donald Coughlin. As noted above, trial counsel referred to this goal by talking about a “safe place,” and the postconviction court referred to this as a “scapegoat” theory. It is not our role to decide if this theory appears from the record to have been strong or weak, only whether it was rational.

¶96 Further, as the State points out, trial counsel was focused on these issues at the time of trial, telling the trial court that counsel would “object[ ] at reasonable times” “if we get too far into Donny,” and counsel did object to some of this evidence as “other acts” evidence. Objections by counsel prompted the trial court at one point, outside the earshot of the jury, to caution the prosecutor against pursuing “any more gruesome detail about what's happened, what the sexual assaults were” that Donald Coughlin committed.

¶97 In addition, as we discussed in our prior opinion, the trial court gave the jury a cautionary instruction explaining that testimony by the victims about Donald Coughlin's physical and sexual assaults of them were to be considered solely as an explanation of why the victims worked on Coughlin's farm during the summers. 2021AP1416-CR, ¶98. This instruction weighs against a determination that trial counsel performed deficiently with regard to the jury simply hearing allegations of criminal conduct by Donald Coughlin alone.

¶98 Given trial counsel's pertinent strategies and the steps taken by the trial court to try to prevent or limit unfair prejudice based on allegations against Donald Coughlin, Coughlin does not persuade us that it was deficient performance for counsel to fail to attempt to prevent the jury from hearing allegations of criminal conduct engaged in by Donald Coughlin that did not directly involve Coughlin or include reference to Donald Coughlin's convictions for sexually assaulting the same victims.¹⁴

¶99 It is true that trial counsel acknowledged at the Machner hearing that he did not consider the possibility of objecting at trial to specific allegations of assaultive conduct by Donald Coughlin alone that might have risked inflaming or misleading the jury about Coughlin's alleged conduct. But Coughlin does not now show how it was deficient performance for counsel not to have moved the trial court to bar the admission of only limited evidence of abuse by Donald Coughlin that would still have been sufficient to show that Donald Coughlin was feared by the victims and therefore they sought out Coughlin's residence as a refuge from abuse. If there was a reasonable way that counsel could have threaded that needle, Coughlin does not now explain what it would have been, much less how it was deficient performance not to do so.

Joint criminal conduct; Donald Coughlin's convictions

¶100 We have just explained why the strategy just referred to is sufficient to justify failures to object at trial to specific allegations of criminal conduct by Donald Coughlin. In contrast, however, we conclude that it fell below an objective standard of reasonableness for trial counsel to allow the jury to hear without objection (1) explicit references by the prosecutor to what, as we explained in our prior opinion, “could reasonably be interpreted as describing the brothers as criminal co-actors who victimized the same children in the same ways,” 2021AP1416-CR, ¶97; and (2) that Donald Coughlin had recently been convicted in a court of law of sexually assaulting the same two victims. Counsel should have recognized at the time of trial that this was a potentially devastating combination for the defense.

¶101 Regarding the explicit references by the prosecutor, as discussed in our prior opinion, it is self-evident that these details created unfair prejudice. This included the prosecutor telling the jury in opening statement, without objection by the defense, that Donald Coughlin subjected the victims to hard blows to their genitals, which the victims would return in kind, “because that's how [the victims] were taught, that's how they were raised by Donald Coughlin and by the defendant.” See id., ¶97 (emphasis omitted). The prosecutor also told the jury that the victims would testify at trial that the victims “had been groomed to be the perfect sexual assault victims by Donald and the defendant.” See id. (emphasis omitted). This further included the prosecutor telling the jury that “something that was done regularly by” both Coughlin and Donald Coughlin was to grab the genitals of the victims, “squeezing hard,” “from the time they were young boys,” and the prosecutor also said that Coughlin and Donald Coughlin both commented on the size of the older victim's penis. Id.

¶102 Trial counsel testified that he did not view the prosecutor's statements as portraying Donald Coughlin and Coughlin as criminal co-actors. But the prosecutor's words were reasonably open to the interpretation that the two brothers were criminal co-actors who jointly victimized the two children. Interpreted this way, these statements undermined the overarching theory of the defense that counsel repeatedly relied on at the Machner hearing to explain his actions and inactions at trial. This is because, if Coughlin joined Donald Coughlin in abusing the victims in various ways, then it would be easier for the State to prove the allegations of the specific sexual assaults charged against Coughlin in this case.

¶103 The prosecutor's statements also undermined the specific defense strategy, which we have just described above, that is sufficient to justify allowing allegations of criminal conduct by Donald Coughlin when he acted alone. It became much harder, if not impossible, to portray Coughlin as providing a “safe space” from Donald Coughlin if the jury accepted the prosecution's suggestion that the two men acting together victimized the two children.

¶104 For these reasons, it was deficient performance for trial counsel not to have taken the simple step of making a motion, which likely would have been granted by the trial court if made, to bar the prosecution from making statements that would be open to the reasonable interpretation that the Coughlin brothers joined together in abusing these victims. Such references could only be prejudicial and would have carried no discernable, fair probative content. The State does not suggest any fact or legal principle that would have prevented the trial court from granting such a motion if counsel had made it.

¶105 Trial counsel suggested in his Machner hearing testimony that he did not consider any statements made by the prosecutor in the opening statement to go “beyond the [pale] and contrary to our kind of overview of the case,” such that they required an objection. But this misses the point that counsel should have, but did not, take the practical step of trying to prevent potentially explosive references to Donald Coughlin from reaching the jury. A simple pretrial motion in limine would have given the trial court an opportunity to foreclose the kind of joint-criminality references that the prosecutor made. See Thiel, 264 Wis. 2d 571, ¶51 (concluding that the failure to file a clearly meritorious motion was deficient performance). It was easily foreseeable that the State might, either intentionally or inadvertently, make improper references to Donald Coughlin or elicit improper references by witnesses. It supports this conclusion that counsel was able to make some objections that resulted in the court admonishing the prosecution from improperly delving too far into topics related to Donald Coughlin.

¶106 Further, the negative effect of these references was amplified by the fact that the jury learned that Donald Coughlin had recently been convicted of sexually assaulting these same victims. By being informed of the conviction, the jury would have understood that Donald Coughlin had either recently admitted in court to sexually assaulting these victims, or else he had been found guilty of doing so beyond a reasonable doubt and that the victims’ allegations had been credited in a court. It is true that, as we explain above, we cannot say that trial counsel made an irrational decision to allow into evidence, without objection, references to some of the allegations related to Donald Coughlin's conviction. But mere allegations are not the same as convictions.¹⁵ When considered together with the co-abuser statements by the prosecutor, counsel should have recognized the risk that the jury might have considered itself invited to think that convictions would be as appropriate for Coughlin as for his purported co-actor Donald Coughlin.

¶107 Our conclusion on this point requires additional background. Trial counsel testified that he had no memory of what he thought at trial about the prosecution's proposal that the jury learn of Donald Coughlin's convictions. But counsel suggested that he would have seen this as fitting into his general safe-place strategy. The following are pertinent events from the trial.

¶108 The State filed a pretrial motion for an order permitting the prosecution to attempt to impeach a witness, Coughlin's sister, if the sister were called by the defense at trial, by using the fact that Donald Coughlin had been convicted of sexually assaulting the two victims in this case. The sister was prepared to testify (and then did testify at trial) to the following exculpatory information: that she had overheard the younger victim apologize to Coughlin and tell Coughlin that the victims’ mother “is on a mission to destroy” Donald Coughlin and Coughlin, even though “nothing ever happened.” The State's pretrial motion asked for permission to attempt to impeach the sister by asking her about her “failure to disclose [her] alleged conversation with [the younger victim] until after Donald Coughlin had been convicted of sexually assaulting” the two victims, in order to challenge the sister's credibility, because “this case has been pending for over four years, the witness is the sister of [Coughlin], and thus [she is] likely to have known of the existence of this case for a substantial period of time and yet only disclosed this information after Donald Coughlin had been convicted.”

¶109 At the hearing that the trial court held on this motion, trial counsel did not object to the State's motion, which the trial court granted. Instead, counsel suggested that he would not object to the prosecution's proposed impeachment of the sister with the fact of Donald Coughlin's convictions, so long as counsel was permitted to ask the sister on redirect examination why she delayed in making the disclosure.

¶110 At trial, the defense called the sister, who as predicted testified on direct examination to the victim's alleged exculpatory statements. As contemplated in the State's pretrial motion granted by the trial court, in cross examining the sister, the prosecutor began to ask her about the criminal charges against Donald Coughlin. Trial counsel objected. The prosecutor reminded the trial court that the defense had not objected to the State's pretrial motion to allow the jury to learn of Donald Coughlin's convictions in this precise context. The trial court called for a sidebar conference. But there was never any record made of what was said by anyone during this sidebar.

¶111 After the sidebar, the prosecution was allowed to inform the jury, through questioning of the sister, that Donald Coughlin “was convicted of assaulting” the same two victims as in the instant case against Coughlin, in what the prosecution presented as an attempt to impeach the sister's testimony. The State took the position, as it had in advance of trial, that there was impeachment value for the prosecution in highlighting that Coughlin's sister waited until after Donald Coughlin was convicted to report to those involved in the investigations of Donald Coughlin and Coughlin that the younger victim had made these exculpatory statements.

¶112 On redirect examination, trial counsel elicited testimony from the sister that she did not reveal the exculpatory conversation that she allegedly overheard before Donald Coughlin's conviction because she was concerned about possible retribution by the victims or some of their family members, implying that she feared for her physical safety.

¶113 During closing and rebuttal argument, the prosecutor referred to the allegations against Donald Coughlin as if they were established facts, telling the jury for example, “Well, we know that Donald Coughlin started” sexually assaulting the two victims when they were each only seven years old. The danger of guilt-by-association was increased by the fact that jurors were informed that Donald Coughlin had recently been convicted.

¶114 With that as additional background, when the State teed up the Donald Coughlin conviction issue before trial, trial counsel should have recognized how deeply prejudicial it would be for the jury to learn of these convictions. This was particularly so given the possibility that the State or a witness might say or imply that Donald Coughlin and Coughlin jointly abused the alleged victims. Then, when the issue arose at trial, again we have a failure of counsel to take the elementary step of creating a record of his objection to critical information being presented to the jury and of how the trial court ruled. As with his testimony regarding the unanimity issues, counsel testified at the Machner hearing that he believed that he made a substantive objection during the sidebar conference to the trial court allowing the jury to learn of the convictions, and that the trial court overruled this objection, but there is no record of any arguments or of any ruling by the trial court as a result of the sidebar. Here again, if counsel did raise the issue in an adequate way, then he was deficient in failing to make a record of it. Or, if he did not raise the issue in an adequate way, then that was deficient performance.

¶115 The postconviction court appeared to determine that trial counsel applied a rational strategy on the issue of Donald Coughlin's convictions. The strategy, as described by the court, was that counsel decided that, because at least some jurors would be familiar with the allegations against Donald Coughlin, it was “better to just deal with [the Donald Coughlin allegations] early on.” But even putting to the side the critical difference between mere allegations and the fact of criminal convictions on this highly sensitive topic, counsel did not testify that counsel applied such a strategy to this issue. Instead, counsel raised an objection, although without making a record of it.

¶116 If trial counsel had properly pursued the objection, this could have been a reasonable defense strategy in response to the prosecution seeking to impeach the sister in this especially prejudicial way. Counsel could have pointed out to the trial court that the State had ways to highlight the delay in the sister's reporting of the alleged statements by the younger victim other than by informing the jury of Donald Coughlin's convictions. In addition, as Coughlin points out, the logic of this theory of impeachment by the prosecution was not strong. That is, there is only a vague potential link between the fact of Donald Coughlin's convictions and the timing of the sister coming forward with this information shortly before Coughlin's third trial.

¶117 The State now argues that it was objectively reasonable for trial counsel not to object to the jury learning of Donald Coughlin's convictions because of the way this issue played out during the defense redirect examination of the sister. Counsel was permitted to elicit what the State now refers to as “otherwise inadmissible character evidence regarding the victims and their family members.” We are not persuaded. It is unduly speculative to suggest that it assisted counsel in persuading the jury to entertain reasonable doubt about the charges here that the jury heard the sister testify that she was fearful of the victims and some of their relatives.

The letter

¶118 We conclude that the record does not reveal that trial counsel took an approach that no competent lawyer would have chosen on the topic of the letter. After consulting with Coughlin on the issue, the two agreed on a strategy, which counsel attempted to pursue, of risking prejudicial fallout from introducing the letter based on a meaningful goal. To recap, the goal was to convince the jury that the letter showed, in the words of counsel, that the victims’ mother took a “crazy blackmail approach,” and that she was motivated to encourage the alleged victims to falsely accuse Coughlin. Coughlin argues that this rationale defies common sense. We disagree. It was not irrational to adopt a strategy that the mother's decision to, as the State now frames the issue, “bury numerous allegations of child sexual assault” in a letter about a church dispute, when presented in its entirety, could reasonably raise questions about the mother's motives and reliability from the defense point of view.

¶119 Coughlin urges us to conclude that it fell below an objective standard of reasonableness for trial counsel not to request redactions to the letter. But the record does not reflect any specific redactions that clearly would have helped the defense and that also would likely have been approved by the trial court. It does not matter if we could imagine various proposed redactions that perhaps might have been positive moves for the defense. Again, we do not second-guess a reasonable trial strategy unless the strategy was “irrational” or “based upon caprice” instead of “judgment.” Breitzman, 378 Wis. 2d 431, ¶65. As the State points out, counsel suggested in his testimony that he thought at the time of trial that the “bizarreness of the letter”—its value in feeding into the defense theory that the mother encouraged the victims to falsely accuse Coughlin—would be best appreciated by the jury if jurors read the entire letter, unredacted. Coughlin does not provide us with a clear basis to essentially second-guess counsel's assessment about potential redaction.

¶120 Separately, Coughlin argues that it was deficient performance for trial counsel not to affirmatively contest the additional assault allegations made in the letter. Coughlin makes reasonable points on this topic. But we conclude that it fell within the wide range of reasonable professional assistance for counsel to determine that such an effort might complicate the trial in a way that could hurt the defense. Counsel testified that trying to disprove the allegations in the letter “would have highlighted” the allegations, and “I frankly didn't want to bring any other evidence up.” This could be interpreted as reflecting a strategy to avoid further calling the jury's attention to the potentially highly prejudicial allegations that Coughlin sexually assaulted children in addition to the two victims in this case.

Evidence supporting a motive for the victims to lie

¶121 This brings us to what appears to present a close call. We conclude that, under the highly deferential review that we must apply, the impeachment and motive-based failures alleged by Coughlin involved sufficiently delicate judgments for trial counsel that Coughlin has not overcome the strong presumption that counsel's conduct fell within the wide range of reasonable professional assistance. This involves Coughlin's argument that counsel was ineffective by failing to make use of impeachment material that would have helped the jury to understand when and why the victims, primarily the older one, reported that Coughlin, as opposed to Donald Coughlin, had sexually assaulted them and regarding motives of various witnesses, in particular the mother of the victims, to lie or encourage others to lie about Coughlin. See 2021AP1416-CR, ¶¶102-111.

¶122 We explained in our prior opinion why we concluded that Coughlin has raised significant issues on these topics. See id. Further, trial counsel failed to explain at the Machner hearing why it would have been in any way inconsistent with his overarching defense—that the victims were making the allegations out of whole cloth—to pursue the lines of impeachment now argued by Coughlin, and counsel had nothing to say about his reasoning on the specific points raised by Coughlin. The postconviction court found only that it was sufficient that counsel pursued a theory that there was a “conspiracy” to falsely accuse him.

¶123 At the same time, however, when we consider each of the alleged failures by trial counsel to attempt to impeach the victims’ mother and the older victim, and to introduce related evidence to bolster the defense theory that the mother caused the older victim to falsely accuse Coughlin, we conclude that these carried potential risks of prejudicing the defense to at least some degree. This includes drawing attention to unfavorable facts for Coughlin or to criminal conduct of Donald Coughlin and appearing to treat the victims unfairly or too harshly. Put differently, Coughlin fails to provide us with a clear roadmap of what particular actions counsel was obligated to take in these areas that includes addressing the potential downsides of taking those particular actions.

¶124 For example, in order to accomplish a goal that Coughlin suggests was essential for the defense—namely, to show that the victims had only ever made incriminating statements about Donald Coughlin, and never about Coughlin, before the mother got into the church dispute with Coughlin—trial counsel would necessarily have had to make references to, and to cause witnesses to testify about, Donald Coughlin. But every additional reference to Donald Coughlin needed to be carefully considered, for reasons we made obvious in 2021AP1416-CR and in discussion above.

Failure to impeach victims regarding significant inconsistencies

¶125 We conclude that trial counsel's approach to the significant discrepancies in statements made by the older victim—which counsel had in hand at the time of trial and did not need to discover through additional investigation—about details involving the alleged assaults, as explained in our prior opinion, fell below an objective standard of reasonableness. See 2021AP1416-CR, ¶¶112-115. Notably, the differences between what the older victim initially told police in a recorded interview and the older victim's testimony at trial was quite different in its details. See id., ¶113. And, as summarized above, by counsel's own account, the trial largely came down to how the jury assessed the credibility of the victims, and counsel failed to identify a sufficient reason not to raise doubts about essential facts of the alleged sexual assaults in the older victim's incriminating accounts.

¶126 At the Machner hearing, trial counsel had little to say on this topic. To some extent, in responding to inquiries on this topic, counsel again fell back on his overarching defense theory, asserting that it did not matter what the victims specifically testified to about “what happened and how it happened,” because the defense theory was that nothing incriminating “ever happened.” But one obvious way to show that nothing “ever happened” would have been to point out that what the victims testified to was significantly inconsistent. See Thiel, 264 Wis. 2d 571, ¶46 (when credibility of alleged victim “was paramount to” the case and trial counsel was “aware of the need to locate any evidence or information to impeach” the victim, counsel's “failure to delve further into the circumstances of the charges and the background of [the defendant's] accuser is objectively unreasonable.”); see also State v. Jeannie M.P., 2005 WI App 183, ¶25, 286 Wis. 2d 721, 703 N.W.2d 694 (determining that defendant showed that trial counsel was aware of evidence that could be used to impeach the credibility of the State's two key witnesses and failed to investigate readily available facts, and concluding that it was deficient performance to fail to use those facts to undermine the credibility of the two key witnesses by showing their motives to fabricate an assault allegation). Further, in this case, unlike in Thiel and Jeannie M.P., none of the impeachment material at issue needed to be obtained through investigative or research efforts by trial counsel.

¶127 It is true that trial counsel asserted briefly in his testimony that he was leery about turning the jury against the defense, out of sympathy for the victims. But he did not explain how raising clear inconsistencies on critical aspects of their accounts would have carried a meaningful risk of alienating the jury, given the defense, which was to argue that the victims were making up all aspects of their incriminating allegations.

¶128 Trial counsel at the Machner hearing agreed with the prosecutor that it would be a “bad look” for counsel in a sexual assault case to impeach a victim “on every single impeachable statement.” This general point is fair. In general, it will not likely constitute deficient performance for a defense attorney to overlook or decline to pursue relatively minor details of potential impeachment material, even if the details could be used to directly challenge relevant testimony of an alleged victim of child sexual assault. See State v. Coleman, 2015 WI App 38, ¶39, 362 Wis. 2d 447, 865 N.W.2d 190 (stating that one “misstep” by trial counsel in failing to impeach the victim with a statement of a witness directly contradicting the victim's account might be considered to involve one “minor detail” that may not “rise to deficient performance”). But this court has determined that it contributed to a determination of deficient performance when defense counsel failed to impeach an alleged child sexual assault victim's credibility in two specific ways that were significant under the facts of the case, and it was not enough for the defense attorney to testify that he did not want to cross examine a child about distasteful subjects or to raise “every little inconsistency.” Id., ¶¶33-39. And here, as we have noted, the nature of the impeachment material left untouched by counsel is striking. See 2021AP1416-CR, ¶¶113-114. The State's attempt to imply that counsel decided not to impeach on merely trivial or cumulative points is not borne out by the record.

¶129 The State acknowledges that trial counsel did not take a full opportunity to challenge the older victim's credibility, but it contends that the impeachment of the older victim by counsel was adequate. The State highlights the fact that counsel impeached the older victim with his statements about when alleged assaults occurred relative to when Coughlin's barn blew down in 1990, which allowed counsel to argue that this raised doubts that the sexual assaults occurred as charged in 1989 and 1990. The State also calls our attention to cross examination by counsel that caused the older victim to acknowledge that some details of his interactions with Coughlin were “sketchy” and “cloudy,” given the passage of time.

¶130 But if the fact that trial counsel conducted some impeachment automatically settles the issue, then the right to effective assistance of counsel would be a hollow promise. It would always be enough for counsel to merely touch on an issue to qualify as constitutionally competent counsel on that issue. Coughlin has identified significant omissions that could not reasonably be justified by concern about creating a “bad look” with the jury when compared with the impeachment that counsel did conduct. See 2021AP1416-CR, ¶¶112-115. This included the most basic of facts regarding the allegation that Coughlin performed oral sex on the older victim. As Coughlin now points out, the older victim initially told police that oral sex happened once or twice, and only in the shower. Then, at the first trial, the victim testified that it occurred once or twice a month, but “could have been less.” At the second trial, the victim testified that this occurred “often” throughout each of the summers at issue, usually in the bedroom or living room of Coughlin's residence. At the third trial, the victim testified on direct examination that it occurred every other week or weekly, every summer except the first one, and in at least four different rooms, and that “it happened so often it's hard to say how often exactly.” But on cross examination, counsel made no attempt to address these significant differences.

¶131 To repeat, trial counsel testified time and again that the overarching defense theory was that every bit of the incriminating testimony by the victims was false, and that distinguishes this sexual assault case from those in which the overarching defense is that an alleged victim was confused, intoxicated, mistaken, or misinterpreted events. Indeed, one aspect of the defense here was that the mother encouraged the victims to lie about the alleged assaults. Significant inconsistencies in critical details of their accounts could have strongly supported that theory. See Jeannie M.P., 286 Wis. 2d 721, ¶22 (trial counsel's choice of strategy “in no way explain[ed]” decision not to impeach key witness for the prosecution, which was deficient performance); Honig, 366 Wis. 2d 681, ¶¶29-30 (trial counsel performed deficiently by failing to impeach or introduce testimony because it “was a failure to present available evidence that went to the core of the defense theory”); cf. State v. Teynor, 141 Wis. 2d 187, 211-12, 414 N.W.2d 76 (Ct. App. 1987) (failure to impeach witness was reasonable because impeachment at issue was inconsistent with defense theory).

¶132 This leaves trial counsel's testimony that he was concerned about what he interpreted to be signs of hostility in the facial expressions or body language of a subset of the jury. To clarify, counsel did not tie this observation to anything specifically having to do with his cross examination of the victims. Rather, he testified that he sensed that some jurors were hostile to the defense in general. Assuming without deciding that the postconviction court implicitly credited counsel's testimony on this point, it is not clear how it could support the State's position on this issue. We do not discern what it could have added to counsel's reasoning regarding this potential impeachment to suspect that some jurors were predisposed to favor the prosecution. As far as counsel explained his thinking, this would seem to have been all the more reason to give seemingly defense-hostile jurors concrete reasons to doubt the victims’ accounts that the defense was claiming were made up from whole cloth, at the urging of the mother. See Honig, 366 Wis. 2d 681, ¶30 (“reviewing court does not ratify a lawyer's decision simply because it is labeled ‘trial strategy’ by postconviction court”).

B. Prejudice

¶133 Having concluded that Coughlin has shown deficient performance by trial counsel on the Donald Coughlin-related issues that we identify, and on the significant inconsistencies in the older victim's accounts, we conclude that Coughlin has shown that these deficiencies substantially contribute to a reasonable likelihood that the result on Counts 1-4 would have been different, based on the entire record, taking into account the testimony at the Machner hearing and the arguments that the State now advances. As with the unanimity issue, we do not conclude that these deficiencies, standing alone, necessarily create that reasonable likelihood. Instead, the strikingly improper implications of the references to joint conduct of Donald Coughlin and Coughlin, and to Donald Coughlin's convictions, together with the failure to raise significant doubts about details of the older victim's accounts, add meaningfully to the likelihood that jurors did not reach completely unanimous verdicts on Counts 1-4. For example, impeachment directed toward the timing, frequency, or location of particular alleged acts could have convinced some jurors that the State could not prove beyond a reasonable doubt that a particular act occurred as alleged, but the jury might still have returned a guilty verdict because other jurors were thinking of a different alleged act. It only adds to the problem that jurors were invited to see Donald Coughlin and Coughlin as co-actors and were told that Donald Coughlin had recently been convicted for sexually assaulting the same victims.

¶134 Coughlin characterizes the weight of the evidence against him as “not remotely overwhelming,” because it depended almost entirely on the specific recollections of the victims as they testified at trial. The State responds in part by commenting that it is “unsurprising that there weren't [other] witnesses” to the alleged sexual assaults, given the secretive nature of child sexual assaults. The State also observes that “it isn't surprising that the victims would have trouble remembering the exact details of the assaults when they reported the abuse some 20 years later.” But the issue is not whether the relative weaknesses of the case is surprising given its nature. It is also not whether those weaknesses are in any sense attributable to the alleged victims. The issue is whether Coughlin has shown a reasonable likelihood that the result would have been different. Citing reasons for the weaknesses does not advance the State's position on this issue. The State also notes that some of the victim testimony at trial was unambiguous and consistent with prior statements they had made. These are reasons that the result of a new trial could be the same. But there is also a reasonable likelihood of the opposite result on each of the charged counts. See State v. Dillard, 2014 WI 123, ¶103, 358 Wis. 2d 543, 859 N.W.2d 44 (stating that a determination of prejudice under Strickland “ ‘does not require certainty or even a preponderance of the evidence that the outcome would have been different with effective assistance of counsel,’ ” but instead “requires only ‘reasonable probability’ ” (quoted source omitted)).

CONCLUSION

¶135 For all of these reasons, we conclude that Coughlin has shown that the performance of trial counsel on multiple issues was deficient, and that this resulted in prejudice on each charged count. Accordingly, we reverse the contrary portions of the order denying the postconviction motion and remand for the circuit court to vacate each count in the judgment of conviction.

By the Court.—Order reversed and cause remanded with directions.

This opinion will not be published. See Wis. Stat. Rule 809.23(1)(b)5.

FOOTNOTES

1.   See State v. Machner, 92 Wis. 2d 797, 804, 285 N.W.2d 905 (Ct. App. 1979) (describing the need for an evidentiary hearing to give the circuit court a chance to hear testimony from counsel explaining the reasons for counsel's relevant actions or failures to act, and for the court to rule on the ineffective assistance claim).The Hon. John Pier Roemer (“the trial court”) presided at trial, and the Hon. Stacy A. Smith (“the postconviction court”) handled all of the postconviction proceedings at issue in this appeal.

2.   Coughlin argues exclusively that there should be a new trial on all counts. The State argues exclusively that there should not be a new trial on any counts. Neither side makes an alternative argument for vacating only a subset of the five counts of conviction.

3.   All references to the Wisconsin Statutes are to the 2023-24 version unless otherwise noted.

4.   In 2021AP1416-CR, we rejected the State's argument that the claims in Coughlin's postconviction motion are not “clearly stronger” than the claims that appellate counsel raised earlier in the direct appeal. See 2021AP1416-CR, ¶¶17-25; see also State v. Romero-Georgana, 2014 WI 83, ¶¶36, 45-46, 360 Wis. 2d 522, 849 N.W.2d 668 (ineffective assistance of counsel may be a sufficient reason for a defendant pursuing a Wis. Stat. § 974.06 motion to have failed to raise an available claim in an earlier motion or on direct appeal, but the defendant must demonstrate that the new claims are “clearly stronger” than the claims previously presented). The State does not now renew its argument on this point, nor does the State argue that the law of the case doctrine does not apply to this issue. Accordingly, we do not revisit the “clearly stronger” issue.

5.   Accordingly, the transcript of the Machner hearing refutes a finding made by the postconviction court to the effect that appellate counsel testified that he made a conscious choice not to address the juror unanimity issue now raised by Coughlin.

6.   In this opinion, Daniel Coughlin, the defendant in this case, is referred to exclusively by his last name and his brother, Donald Coughlin, is referred to by his full name.

7.   Trial counsel, who represented Coughlin at all three trials in this case, testified that he was not sure in which of the three trials he raised the issue. But Coughlin points out in this appeal that, if counsel did raise the issue, then this must have been at the third trial, because that was the only trial at which these jury instructions were discussed by the parties and the trial court.

8.   The separate prosecution of Donald Coughlin reached our supreme court. State v. Donald Coughlin, 2022 WI 43, ¶¶1, 5-6, 54, 402 Wis. 2d 107, 975 N.W.2d 179 (affirming Donald Coughlin's convictions on 14 counts of first-degree and second-degree sexual assault of a child and one count of repeated sexual assault of a child, involving the same two victims as in the instant case, plus a third victim).Separately, Coughlin argues in part that references to criminal conduct of Donald Coughlin, or implied joint, bad conduct of Donald Coughlin and Coughlin, amounted to inadmissible “other acts” evidence at the trial in the instant case, or at least that it was inadmissible for the same reasons that “other acts” evidence is generally barred. See Wis. Stat. § 904.04.

9.   An attorney's failure to make an adequate record should not be confused with an attorney's failure to renew an objection that has already been rejected by the circuit court in an earlier ruling. Here, as referenced above, the postconviction court pointed out that it can be a common, reasonable strategy for an attorney engaged in litigation to forego renewing an objection that the court has previously overruled. The postconviction court's observation was apparently related to trial counsel's testimony that he believed that he “dropped” one unanimity issue after unsuccessfully raising it with the trial court in some manner. The postconviction court's observation is appropriate, as far as it goes. See State v. Venema, 2002 WI App 202, ¶25 n.6, 257 Wis. 2d 491, 650 N.W.2d 898 (a definitive pretrial ruling preserves an objection to the admissibility of evidence, without the need for a renewed objection at trial, if the facts and law presented to the court in the pretrial motion are the same as those that arise at trial). But, in general, it is not a reasonable strategy for an attorney litigating a case to decide not to make an adequate record of a previously made, off-the-record objection (along with the basis for the court's denial of the objection) out of fear that doing so might irritate the circuit court. And here, the record provides no reason to think that counsel had any reasonable basis for not making an adequate record.

10.   Coughlin argues that the State has repeatedly forfeited an argument based on unsettled law, including by implicitly conceding in 2021AP1416-CR that it had no such arguments to make. But, in the interest of judicial efficiency, we exercise our discretion to consider the State's new alternative argument and we reject it on the merits.

11.   We observe that, while our supreme court applied this “conceptually distinct” test to determine whether unanimity is required in State v. Lomagro, 113 Wis. 2d 582, 335 N.W.2d 583 (1983), and in other opinions, our supreme court later explained that the test was modified by the U.S. Supreme Court “in favor of a constitutional test for unanimity challenges that focuses on an evaluation of the fundamental fairness and rationality of the legislature's choice to provide for a single offense with alternate modes of commission.” See State v. Derango, 2000 WI 89, ¶¶22-25, 236 Wis. 2d 721, 613 N.W.2d 833 (discussing Schad v. Arizona, 501 U.S. 624, 635 (1991) (plurality opinion), abrogated on other grounds in Ramos v. Louisiana, 590 U.S. 83 (2020)). But the degree of conceptual distinctness or equivalency remains part of the test, see Derango, 236 Wis. 2d 721, ¶24, and in any case the State does not develop an argument based on this modification of the test from the time of Lomagro.

12.   Derango, 236 Wis. 2d 721, was modified on other grounds in State v. Davison, 2003 WI 89, ¶¶34, 43-44, 263 Wis. 2d 145, 666 N.W.2d 1, namely, regarding the proper analysis of multiplicity claims.

13.   To further clarify, as discussed in State v. Gustafson, 121 Wis. 2d 459, 359 N.W.2d 920 (1985) (Gustafson III) (per curiam, granting motion for reconsideration), the earlier State v. Gustafson, 119 Wis. 2d 676, 350 N.W.2d 653 (1984) (Gustafson II) was a fractured opinion and the mandate changed between Gustafson II and Gustafson III. But four justices in Gustafson II (Heffernan, Bablitch, Abrahamson, and Ceci), appeared to join in the characterization that the offense at issue involved “a single, continuous criminal transaction,” and this characterization was not altered by the court in Gustafson III.

14.   We reach this decision even though, as we recognized in our prior opinion, the prosecution exceeded the scope of the evidence that the trial court had permitted. See 2021AP1416-CR, ¶96.

15.   Our supreme court has developed case law to address concerns that jurors can be unfairly swayed by their exposure to information about prior convictions. See State v. Warbelton, 2009 WI 6, ¶¶44-45, 48, 315 Wis. 2d 253, 759 N.W.2d 557 (discussing doctrines addressing the dangers of unfair prejudice arising from jurors learning of a defendant's prior convictions, suspensions, or revocations). Of course, the convictions here were Donald Coughlin's, not Coughlin's. But, given the way in which the prosecution discussed the two men, the information about Donald Coughlin's convictions was highly prejudicial.

PER CURIAM.