Opinion
20210634
12-27-2024
State of Utah, Respondent, v. William Torres-Orellana, Petitioner.
Sean D. Reyes, Att'y Gen., David A. Simpson, Asst. Solic. Gen., Salt Lake City, for respondent Emily Adams, Freyja Johnson, Cherise Bacalski, Bountiful, for petitioner
Heard February 8, 2023
On Certiorari to the Utah Court of Appeals Second District Court, Ogden The Honorable Camille L. Neider No. 171900272
Sean D. Reyes, Att'y Gen., David A. Simpson, Asst. Solic. Gen., Salt Lake City, for respondent
Emily Adams, Freyja Johnson, Cherise Bacalski, Bountiful, for petitioner
Justice Petersen authored the opinion of the Court, in which Chief Justice Durrant, Associate Chief Justice Pearce, Justice Hagen, and Justice Pohlman joined.
OPINION
PETERSEN, JUSTICE
INTRODUCTION
¶1 A jury convicted William Torres of one count of rape. The district court had serious misgivings about the performance of Torres's trial counsel, so the court appointed post-trial counsel to investigate potential errors. Once represented by post-trial counsel, Torres moved for a new trial, arguing that he had received ineffective assistance of counsel under Strickland v. Washington, 466 U.S. 668 (1984). That case holds that a defendant's Sixth Amendment right to counsel is violated where (1) "counsel's representation fell below an objective standard of reasonableness" and (2) "the deficient performance prejudiced the defense." Id. at 687-88.
We refer to Torres by his first surname only, because that is how he refers to himself in his briefing.
¶2 The district court granted Torres a new trial. It concluded that trial counsel's performance was objectively unreasonable in several respects and that, cumulatively, those errors prejudiced Torres. Accordingly, the district court concluded that Torres received ineffective assistance of counsel and ordered a new trial under rule 24(a) of the Utah Rules of Criminal Procedure.
¶3 The State appealed, and the court of appeals reversed. See State v. Torres-Orellana, 2021 UT App 74, 493 P.3d 711. The court of appeals reviewed the district court's determination that Torres received ineffective assistance of counsel for correctness, according to existing precedent. See id. ¶ 26 & n.10 (citing Menzies v. Galetka, 2006 UT 81, ¶ 58, 150 P.3d 480). And the court of appeals concluded that Torres had failed to show that trial counsel's performance was both deficient and prejudicial under Strickland. See id. ¶ 44.
¶4 On certiorari, Torres argues that the court of appeals erred in reversing the district court's decision that his trial counsel was ineffective.
¶5 First, Torres argues that the court of appeals applied the wrong standard of review. He argues that when a defendant raises an ineffective assistance claim in a new trial motion in the district court, an appellate court should review the district court's ruling for abuse of discretion-the standard of review for a new trial order-rather than for correctness-the standard of review for an ineffective assistance claim.
¶6 We reject the premise of Torres's argument, which is that new trial orders adjudicating ineffective assistance claims are reviewed under a different standard than other new trial orders. A district court's decision to grant or deny a new trial is reviewed for an abuse of discretion. See State v. Billingsley, 2013 UT 17, ¶ 9, 311 P.3d 995. However, any legal decisions embedded in the district court's new trial ruling are reviewed for correctness. See State v. Boyden, 2019 UT 11, ¶ 21, 441 P.3d 737. Thus, when a district court rules on a criminal defendant's claim that he was deprived of his Sixth Amendment right to counsel due to ineffective assistance, the district court's determination of whether the defendant received constitutionally ineffective assistance is reviewed for correctness. See Menzies, 2006 UT 81, ¶ 58. But that is because the court's ineffective assistance determination is a legal decision underlying its new trial ruling, not because new trial orders involving ineffective assistance claims are reviewed under a different standard. And as with all discretionary decisions, if a district court's new trial ruling contains a legal error, that constitutes an abuse of discretion.
¶7 Second, Torres argues that we should overrule our holding in Menzies v. Galetka, 2006 UT 81, that ineffective assistance is a law-like mixed question that is reviewed for correctness when it arises in the circumstance presented here. He contends that when an ineffective assistance claim is raised by a defendant in the district court and ruled on by the court that presided over the trial, appellate courts should review at least the prejudice prong of the district court's Strickland analysis for abuse of discretion. We decline Torres's invitation to overrule our precedent.
¶8 And finally, we conclude that Torres has failed to show that the court of appeals erred in concluding he was not prejudiced by the performance of his counsel. We agree with the court of appeals that Torres has not established a reasonable probability of a different result at trial absent counsel's alleged errors. Accordingly, we affirm.
Torres was convicted of raping his ex-girlfriend, Tiffany (a pseudonym). We "recite the facts from the record in the light most favorable to the jury's verdict." Gregg v. State, 2012 UT 32, ¶ 2, 279 P.3d 396 (cleaned up).
¶9 Tiffany and Torres began dating in January 2017. At the time, Tiffany was seventeen years old and was recuperating from having her first child. She told Torres that she "wasn't ready" to have sex and that, "if sex is what [he] want[ed]," he "should probably be talking to someone else." Torres said he would wait until she was ready.
¶10 The two went for a drive one evening, with Tiffany expected home by 9:00 p.m. They parked the car at Sand Ridge Park where, after talking and kissing for about an hour, Torres asked Tiffany to get on top of him while he sat in the driver's seat. Tiffany refused, and Torres unsuccessfully tried to pull her over from the passenger seat. When that did not work, Torres asked her to perform oral sex. Tiffany complied but stopped shortly after, stating that she did not want to continue.
¶11 Torres again tried to pull Tiffany on top of him. This time, Tiffany was unable to fend him off. And once Tiffany was on top of him, Torres pulled her pants down and inserted his penis into her vagina. The intercourse lasted about five minutes, and all the while Tiffany "was trying to fight it."
¶12 Tiffany eventually freed herself from Torres's grip, returned to the passenger seat, and asked him to take her home. Torres asked Tiffany to "keep going" so he could finish. But when Tiffany started texting her mother-telling her that Torres would not take her home-Torres turned on the car and drove her back. On arrival, Tiffany's mother asked her if Torres had forced her to have sex. Tiffany said yes, and her mother called the police.
¶13 A sexual assault nurse examiner (SANE nurse) saw Tiffany later that night. The SANE nurse noted several bruises on Tiffany's neck, chest, and breast, which were consistent with "sucking" and "grabbing of the breasts." And there were abrasions on Tiffany's legs consistent with her allegations that Torres had pulled her from the passenger seat and that she had fought to get off of him. The SANE nurse also noted several injuries to Tiffany's genital area.
¶14 Tiffany met with police, and Torres was arrested a little over a week after the incident. The State charged him with one count of rape, a first-degree felony.
Trial
¶15 Tiffany, her mother, and the SANE nurse each testified at Torres's trial. Torres's arguments on certiorari go mainly to the State's examination of Tiffany.
¶16 The State asked Tiffany about numerous texts that she and Torres had exchanged in the days following the rape. A number of Torres's texts during this time appeared inculpatory. For instance, the next morning, Tiffany asserted in a text to Torres that he "knew [she] didn't want to have sex" and she had "told [him] multiple times and [he] still did it." Torres responded: "What do you mean? N Babe I did but I saw that you wanted too but I didn't know, I didn't wanted to do it either, n I know you told me multiple times, n I'm so sorry I should of just stoped. Forgive me, I won't do it again." The State introduced this and several other texts Torres sent Tiffany in the days after the incident with similar sentiments.
¶17 The State also asked Tiffany about certain texts she had sent Torres during this time. Some of Tiffany's texts appeared to show that she still had feelings for Torres. At one point, she texted, "I just want you but idk if my mom would let anything with us happen again." She also said, "I don't want to lose you. You're all I want. I don't want anyone else." And "I love you and I want you a lot but I can't disrespect my mom."
¶18 When asked about what she had been thinking when she sent Torres these texts, Tiffany testified, "I was super hurt, but I was also really dumb and thought that I could give him another chance. . . . [M]y hormones were all over the place from having a baby and I just wanted to feel loved and I liked him a lot before it happened." Tiffany also testified, "I didn't want to lose him but it's like [I] couldn't be with someone like that. . . . That did that to me."
¶19 The SANE nurse testified about Tiffany's injuries following the incident, explaining that they were inconsistent with consensual sex. She testified that Tiffany's "several sites of injury . . . indicate[d] that there[] [had] not [been] a correction to prevent injury"-in other words, a readjustment or change of position- which the SANE nurse said would be expected if the sex had been consensual. And when asked "how many other[] [cases the SANE nurse had] seen like this that ha[d] the severity of [Tiffany's] injuries," the SANE nurse responded that, of the hundreds of exams she had performed over the course of twelve years, "probably three."
¶20 The jury convicted Torres as charged.
Post-trial Proceedings
¶21 During sentencing, the district court identified a potential legal issue with one of the jury instructions used at trial. On its own motion, the court appointed post-trial counsel to address the issue. And during a later status conference, the court expressed concerns regarding prior counsel's performance at trial.
¶22 Through post-trial counsel, Torres moved for a new trial, arguing that he had been deprived of his Sixth Amendment right to the effective assistance of counsel. Ineffective assistance claims are governed by the test the U.S. Supreme Court set in Strickland v. Washington, which requires the defendant to "demonstrate that (1) his counsel's performance was deficient in that it 'fell below an objective standard of reasonableness' and (2) 'the deficient performance prejudiced the defense.'" State v. Ray, 2020 UT 12, ¶ 24, 469 P.3d 871 (quoting 466 U.S. 668, 687-88 (1984)).
¶23 In his motion, Torres pointed to additional texts he had exchanged with Tiffany in the days after the incident. He argued that it was objectively unreasonable for trial counsel not to have introduced them at trial. And because these additional texts would have "challenge[d] [Tiffany's] story that a rape actually occurred," Torres argued that there was a reasonable probability of a different result at trial had the jury seen them. He therefore moved for a new trial under rule 24(a) of the Utah Rules of Criminal Procedure.
¶24 The district court granted Torres's motion. In its ruling, the district court noted that it had reviewed "the entirety of the text conversations" Torres had attached in support of his new trial motion-"1,300 text messages" sent between the date of the incident and when Torres was arrested. As the court found, these texts included conversations Tiffany initiated to make Valentine's Day plans; "her thoughts on . . . moving out together, getting married, being together as a couple, Torres being a father figure to [her] new baby"; "plans to sneak out and see each other"; her "frustrations with school, the baby and her mother"; and "feelings of love and affection for Torres." Tiffany had also sent "romantic pictures" of the two of them together, as well as pictures of Torres with her baby. And there were texts that, in the district court's view, suggested that Tiffany had been in a "dark place" not because of "what happened at the park," but because of "restrictions imposed by her mom, the difficulties of being a new mom, [and] hormonal issues from giving birth and being back on birth control." In sum, these texts were of an "affectionate nature" and presented a "contrast to [Tiffany's] allegations and testimony that a violent, forceful rape . . . had just occurred."
¶25 In light of these and other findings, the district court concluded that Torres's trial counsel had provided ineffective assistance. Having presided over the trial, the court reasoned that "there [was] no reasonable trial strategy that account[ed] for [counsel's] failure to admit" the additional texts, or for not "cross-examining [Tiffany] . . . both for what she said and how she subsequently acted." Based on this error and several others identified by the court, it concluded that "the cumulative effect of trial counsel's deficient performance undermine[d] the verdict and denied Torres a fair trial."
¶26 And based upon trial counsel's ineffective assistance, the district court granted Torres's motion for a new trial.
Appeal
¶27 The State appealed. It argued that the district court "erred in granting Torres a new trial on the ground of ineffective assistance of counsel." State v. Torres-Orellana, 2021 UT App 74, ¶ 26, 493 P.3d 711.
¶28 In discussing the applicable standard of review, the court of appeals observed, "Ordinarily, we review a court's ruling on a motion for a new trial for an abuse of discretion. But when a defendant moves for a new trial on ineffective assistance of counsel grounds, [precedent requires that] we apply the standard of review set forth in Strickland v. Washington." Id. (citation omitted) (citing cases). And the court noted that appellate courts review a district court's application of the Strickland test "for correctness," even when the Strickland analysis is conducted by the district court that presided over the trial. Id. The court of appeals found this to be somewhat "perplexing" at least as to the prejudice determination, because it was a "departure from the general recognition that trial courts are in an advantaged position to that of appellate courts to determine the impact of events occurring in the courtroom on the total proceedings, [normally] resulting in a more deferential review." Id. ¶ 26 & n.10 (cleaned up).
¶29 Judge Harris explored this issue in a concurring opinion. He opined that a district court is better situated to assess whether a defendant has satisfied Strickland's prejudice element, “at least where the same judge who presided over the trial made the prejudice determination.” Id. ¶ 51 (Harris, J., concurring). And he “wonder[ed] whether our law ought to develop a standard of review in ineffective assistance of counsel cases that requires some level of deference to . . . a trial court's determination that there is (or is not) a reasonable probability that the outcome of the trial would have been different” absent counsel's error. Id.
¶30 That said, the court of appeals followed existing precedent. It reviewed the district court's ineffective assistance determination for correctness, including its assessment of prejudice. See id. ¶ 26 (majority opinion).
¶31 Regarding the additional texts, the court of appeals did not address whether trial counsel's failure to introduce them was deficient performance because it concluded that even assuming it was, the absence of the texts did not prejudice Torres. See id. ¶ 30. "[A]lthough the jury might not have known the full extent to which Tiffany expressed affection for Torres," the court of appeals noted that the State had admitted some texts indicating that Tiffany "continued to communicate with [him] and wished, at least at first, to continue her relationship with him following" the rape. Id. ¶ 33; see id. ¶ 32. The court of appeals also explained that "[a]mple additional evidence supported [the] conviction," including several "texted admissions [Torres] made in the week following the rape and [the SANE] [n]urse's testimony regarding her physical examination of Tiffany." Id. ¶ 34. In light of this evidence, the court of appeals concluded that trial counsel's failure to introduce the additional texts did not prejudice the defense, and the district court erred in concluding that Torres was entitled to a new trial based on ineffective assistance. See id. ¶¶ 36, 44.
¶32 As to the other errors identified by the district court, the court of appeals similarly held that Torres had not received ineffective assistance, either because trial counsel did not perform deficiently or because the error was not prejudicial. See id. ¶¶ 37-43. It therefore reversed the district court's decision, set aside the new trial grant, and remanded for sentencing. See id. ¶ 44.
¶33 Torres filed a petition for certiorari. We have jurisdiction under Utah Code subsection 78A-3-102(3)(a).
STANDARD OF REVIEW
¶34 "On certiorari, this court reviews the decision of the court of appeals for correctness, giving no deference to its conclusions of law." Donovan v. Sutton, 2021 UT 58, ¶ 15, 498 P.3d 382 (cleaned up). "The correctness of the court of appeals' decision turns on whether that court correctly reviewed the trial court's decision under the appropriate standard of review." State v. Dean, 2004 UT 63, ¶ 7, 95 P.3d 276.
ANALYSIS
¶35 The State appealed from the district court's order granting Torres a new trial under rule 24(a) of the Utah Rules of Criminal Procedure. Rule 24(a) allows a district court to "grant a new trial in the interest of justice if there is any error or impropriety which had a substantial adverse effect upon the rights of a party." Utah R. Crim. P. 24(a). Torres requested a new trial on the ground that he received ineffective assistance of counsel. While the district court held that Torres was entitled to a new trial based on ineffective assistance, the court of appeals disagreed and reversed.
¶36 We granted certiorari on two issues: first, whether we should "reconsider the standard of review and the level of deference afforded to a district court when considering appeals of rulings on motions for a new trial under claims of ineffective assistance of counsel"; and second, whether the court of appeals "erred in reversing the district court's grant of a new trial" in this case.
¶37 We address these issues in turn. Ultimately, we decline to revise the applicable standard of review, and we affirm the court of appeals' determination that Torres did not receive ineffective assistance of counsel.
I. We Decline to Alter the Standard of Review for New Trial Orders that Adjudicate Ineffective Assistance Claims
¶38 Torres asks us to revisit the standard of review for new trial rulings that resolve claims of ineffective assistance of counsel. He makes two different arguments that arrive at the same destination-namely, applying an abuse of discretion standard to such rulings by a district court, "especially when the judge deciding the motion for a new trial is the same judge who presided over the trial."
¶39 Torres's first argument involves the standard of review for new trial motions generally. He argues that because we review district court rulings on new trial motions for an abuse of discretion, "[t]he same deferential abuse-of-discretion review should apply to a district court's grant of a new trial on grounds of ineffective assistance of counsel as any other ground[]" available under rule 24(a). In Torres's view, "[t]he key question on appeal should not be whether the appellate court would have found ineffective assistance and granted a new trial in its independent assessment of the case, but whether the district court acted within its discretion in doing so."
Torres confines his argument to a "grant" of a new trial. But his reasoning would apply equally to a court's denial of a new trial motion.
¶40 Torres's second argument involves the standard of review for a district court's ruling on ineffective assistance of counsel specifically. He develops the point raised by Judge Harris, arguing that we should afford deference to a district court's conclusion as to the second Strickland prong-whether counsel's performance prejudiced the defense-insofar as the district court presided over the trial. Because that contravenes this court's decision in Menzies v. Galetka, 2006 UT 81, 150 P.3d 480, Torres asks us to overrule that case.
¶41 We first address Torres's argument about the standard of review for new trial rulings. We then analyze his argument that the standard of review for the prejudice prong of an ineffective assistance ruling should differ when that ruling was made by the district court that presided over the trial.
A. The Standard of Review for New Trial Rulings Under Rule 24 of the Utah Rules of Criminal Procedure
¶42 Our cases make clear that "[w]e review a [district] court's ruling on a motion for a new trial under an abuse of discretion standard." State v. Billingsley, 2013 UT 17, ¶ 9, 311 P.3d 995; see also State v. Pinder, 2005 UT 15, ¶ 20, 114 P.3d 551. But we have also stated that when the basis of a new trial order is ineffective assistance, the standard of review for ineffective assistance claims applies. See State v. Templin, 805 P.2d 182, 185-86 (Utah 1990); State v. J.A.L., 2011 UT 27, ¶ 20, 262 P.3d 1 ("Typically, we review a denial of a motion for a new trial under an abuse of discretion standard. However, when a defendant asserts a constitutional claim of ineffective assistance of trial counsel, a different standard of review applies. An ineffective assistance of counsel claim is a mixed question of law and fact.").
¶43 While we have described this as a departure from the usual standard of review for new trial rulings, a closer look reveals that it is not. This is because, while our abuse of discretion standard asks generally whether the court has "exceeded the range of discretion allowed for [a] particular act under review," Rivera ex rel. Rivera v. State Farm Mut. Auto. Ins. Co., 2000 UT 36, ¶ 7 n.2, 1 P.3d 539, it is well settled that we review any legal conclusions underlying the discretionary decision for correctness, see State v. Boyden, 2019 UT 11, ¶¶ 19-21, 441 P.3d 737; State v. De La Rosa, 2019 UT App 110, ¶ 4, 445 P.3d 955 ("[T]he abuse-of-discretion standard of review will at times necessarily include review to ensure that no mistakes of law affected a lower court's use of its discretion." (cleaned up)). In other words, a reviewing court must still ensure that the ruling is "based . . . on the law." Archuleta v. Galetka, 2011 UT 73, ¶ 152, 267 P.3d 232 (cleaned up). Even a district court's "discretionary determinations must rest upon sound legal principles." Boyden, 2019 UT 11, ¶ 21. So when an appellant challenges "a legal conclusion . . . embedded in a district court's discretionary determination, we peel back the abuse of discretion standard and look to make sure that the court applied the correct law." Id. And if a discretionary ruling is based upon a legal error, that "constitutes an abuse of discretion." Id. ¶ 19.
¶44 We have consistently reviewed new trial rulings this way. In State v. Allen, for instance, the defendant argued that the district court applied the wrong legal standard in denying his new trial motion based on juror misconduct. See 2005 UT 11, ¶ 49, 108 P.3d 730 ("[The defendant] contends that, instead of examining whether he was prejudiced by the juror misconduct, the district court should have applied [a] rebuttable presumption of prejudice standard . . . ."). We explained that we would review the new trial ruling for abuse of discretion, while "[a]t the same time" reviewing the underlying question of the applicable legal standard "for correctness." Id. ¶ 50 (cleaned up). And in State v. Centeno, although we said that we would "generally review [the] district court's denial of [the defendant's] motion for a new trial for abuse of discretion," we reviewed the legal conclusion supporting the court's decision-whether the defendant was deprived of his Sixth Amendment right to confrontation-for correctness. See 2023 UT 22, ¶¶ 44, 82-85, 537 P.3d 232.
¶45 The court of appeals illustrated this well in State v. De La Rosa, 2019 UT App 110. It noted that where a district court's ruling on a new trial motion was based on, for example, jury misconduct, the district court's determination as to whether the misconduct "merited a new trial would have been almost wholly dependent on the trial court's evaluation of the impact such misconduct had on the fairness of the proceedings." Id. ¶ 5. And a reviewing court would analyze the decision under an abuse of discretion standard and the dictates of rule 24(a)-i.e., whether there was "any error or impropriety which had a substantial adverse effect upon the rights of a party." Utah R. Crim. P. 24(a). But when the "error or impropriety" identified by the district court involves a legal conclusion-for example, because the district court determined that a jury instruction was incorrect-the court of appeals explained that a reviewing court analyzes the embedded legal conclusion for correctness. De La Rosa, 2019 UT App 110, ¶¶ 5-7 (citing Billingsley, 2013 UT 17, ¶¶ 7, 10, 17, 23). And if the ruling is legally incorrect, then that would amount to an abuse of discretion. Id. ¶ 6 (citing Billingsley, 2013 UT 17, ¶¶ 7, 10, 17, 23).
¶46 For this reason, when a district court's decision to grant or deny a new trial involves a conclusion as to whether the defendant received ineffective assistance, that embedded conclusion is reviewed for correctness. In Menzies, we explained that an ineffective assistance of counsel claim "presents a mixed question of law and fact." 2006 UT 81, ¶ 56 (citing Strickland v. Washington, 466 U.S. 668, 698 (1984)). We observed that "[t]raditionally, this court . . . has reviewed the application of the ineffective assistance standard to the facts for correctness-even when the claim is initially heard by the district court." Id. However, we noted that we had not evaluated that practice under the test we had developed in State v. Levin, 2006 UT 50, ¶ 25, 144 P.3d 1096, so we took the "opportunity to reevaluate whether correctness [was] the proper standard of review," Menzies, 2006 UT 81, ¶ 57. We determined that it was, concluding that we would continue to "review for correctness the trial court's application of the law to the facts." Id. ¶ 58.
¶47 Accordingly, just like any other legal conclusion embedded in a district court's discretionary decision, we review a court's conclusion that a defendant received constitutionally ineffective assistance for correctness, even when the issue arises in a motion for a new trial in the district court. See Centeno, 2023 UT 22, ¶ 44. We review the court's decision this way not because ineffective assistance is treated differently than any other "error or impropriety" contemplated by rule 24(a), but because that is how our abuse of discretion standard works: we "peel back the abuse of discretion standard" when reviewing a district court's underlying legal conclusions. Boyden, 2019 UT 11, ¶ 21. And under Menzies, a district court's Strickland determination resolves a law-like mixed question, and therefore we review it for correctness.
Torres makes the somewhat different argument that, because the district court can grant a new trial based on any "error or impropriety which had a substantial adverse effect upon the rights of a party," rule 24(a) "does not require a Strickland showing." In his view, rule 24(a) provides a stand-alone basis for granting a new trial, and the court of appeals therefore "misse[d] the point" by reviewing his Strickland showing for correctness, like Menzies instructs. Torres's problem, however, is that he argued to the district court that he was entitled to a new trial because his Sixth Amendment right to effective counsel had been violated. We take Torres's point that rule 24(a)'s plain language does not appear to require a party to demonstrate a constitutional violation to obtain relief. Indeed, the rule provides a remedy for "any error or impropriety which had a substantial adverse effect" on the defendant's rights. Utah R. Crim. P. 24(a) (emphasis added). We leave open the question of whether someone in Torres's shoes might fall short of his burden to establish a violation under Strickland but nevertheless demonstrate that he is entitled to a new trial based on errors or improprieties attributable to defense counsel. But because Torres asked the district court to grant a new trial based on Strickland, we leave that question for another day.
¶48 Accordingly, we do not disturb our existing practice of reviewing a legal conclusion "embedded in a district court's discretionary determination" for correctness, including when the legal conclusion is a determination of ineffective assistance. Boyden, 2019 UT 11, ¶ 21. And to the extent we have described this as applying "a different standard of review" to new trial rulings where ineffective assistance is concerned, e.g., J.A.L., 2011 UT 27, ¶ 20, we correct that mischaracterization.
B. The Standard of Review for Ineffective Assistance Determinations Under Strickland v. Washington
¶49 Torres makes a second argument, which focuses on the Strickland standard of review rather than the standard of review for new trial orders. Torres argues that when the district court that presided over a trial, rather than an appellate court, rules on a defendant's ineffective assistance claim, a reviewing appellate court should review the district court's prejudice determination with deference rather than for correctness. But that would require us to overrule what we said in Menzies about the standard of review for ineffective assistance determinations. Foreseeing this, Torres invites us to overrule our relevant holding in that case.
¶50 Ultimately, however, Torres does not persuade us to overturn the holding in Menzies. Consequently, we decline to modify the standard of review for prejudice determinations in the context of ineffective assistance claims, including in the less frequent instances when the issue arises for the first time in the district court.
¶51 As discussed above, in Menzies, we concluded that ineffective assistance is a mixed question that is reviewed for correctness. See supra ¶¶ 46-47. The U.S. Supreme Court stated in Strickland that whether counsel has rendered ineffective assistance "is a mixed question of law and fact." 466 U.S. at 698. "Mixed questions arise when a district court must apply a particular rule of law to a particular set of facts." Randolph v. State, 2022 UT 34, ¶ 20, 515 P.3d 444. They "fall somewhere in the twilight between . . . findings of fact," reviewed deferentially, and "conclusions of law," reviewed for correctness. Id. (cleaned up).
¶52 We apply a "binary method for determining the appropriate standard of review for mixed questions," which asks "whether the question is better classified as a law-like mixed question or a fact-like mixed question." Id. ¶ 22 (cleaned up). And depending on the side of the line on which it falls, we review the mixed question for correctness or with deference. See id. ¶¶ 23-24. It is important to remember that this analysis relates only to a district court's application of the law to the facts before it, not to its factual findings. See Murray v. Utah Lab. Comm'n, 2013 UT 38, ¶ 33, 308 P.3d 461 ("[A] mixed question arises when a[] . . . lower court must apply a legal standard to a set of facts unique to a particular case." (cleaned up)). Where a district court makes factual findings that are relevant to an ineffective assistance claim arising in a new trial motion, appellate courts will defer to those factual findings and "overturn [them] only if they are clearly erroneous." Menzies, 2006 UT 81, ¶ 58.
¶53 Menzies was our first opportunity to apply the Levin factors to ineffective assistance claims, although we did not use the terms "law-like" and "fact-like" at that time. See 2006 UT 81, ¶¶ 56-57. To guide the analysis of whether a mixed question should be reviewed for correctness or with deference, Levin directs courts to consider:
The State cites a case that predates Menzies, State v. Templin, where we held that "reviewing courts are free to make an independent determination of a trial court's conclusion[]" that a defendant received ineffective assistance. 805 P.2d 182, 186 (Utah 1990). The State therefore faults Torres for inviting us to overrule Menzies only and not also Templin. But we made clear in Menzies that we had "not yet evaluated the appropriate standard of review for ineffective assistance of counsel claims under" Levin, and we took the "opportunity to reevaluate whether correctness [was] the proper standard." Menzies v. Galetka, 2006 UT 81, ¶ 57, 567 P.3d 480. Consequently, we think it fair for Torres to have asked us to overrule Menzies only.
(1) the degree of variety and complexity in the facts to which the legal rule is to be applied; (2) the degree to which a trial court's application of the legal rule relies on facts observed by the trial judge, such as a witness's appearance and demeanor, relevant to the application of the law that cannot be adequately reflected in the record available to appellate courts; and (3) other policy reasons that weigh for or against granting discretion to trial courts.Id. ¶ 58 (quoting Levin, 2006 UT 50, ¶ 25); see also In re Adoption of Baby B., 2012 UT 35, ¶ 42, 308 P.3d 382 (explaining that the Levin factors are a cost-benefit analysis for determining whether a mixed question is "fact-like" or "law-like").
¶54 Applying those factors in Menzies, we concluded that a determination of ineffective assistance should be reviewed for correctness. See Menzies, 2006 UT 81, ¶ 58.
¶55 Torres asks us to overturn this holding from Menzies. Any party "asking us to overturn prior precedent ha[s] the substantial burden of persuading us" that our Eldridge factors weigh in favor of doing so. Randolph, 2022 UT 34, ¶ 66 (cleaned up) (citing Eldridge v. Johndrow, 2015 UT 21, ¶ 23, 345 P.3d 553). "Those factors include[] (1) the persuasiveness of the authority and reasoning on which the precedent was originally based, and (2) how firmly the precedent has become established in the law since it was handed down." Id. (cleaned up). Torres fails to show that either factor weighs in favor of overruling Menzies.
1. Persuasiveness
¶56 Torres contends that Menzies is unpersuasive for a few reasons. He first asserts that we "did not engage in a meaningful analysis [in Menzies] as to why motions based on claims of ineffective assistance of counsel should be treated differently" than new trial motions based on other "errors or improprieties." But the premise of Torres's argument is incorrect. As we explained above, we have not actually applied a different standard of review to new trial motions based on ineffective assistance of counsel, as opposed to such motions based on other errors or improprieties. See supra Section I.A. Rather, while a court's ultimate decision on a new trial motion is reviewed for an abuse of discretion, any underlying legal determinations are reviewed for correctness. See supra ¶ 43. And a determination of ineffective assistance of counsel is a legal conclusion. See supra at ¶¶ 46-47. Accordingly, we move to Torres's next argument.
¶57 Torres also argues that our Levin analysis in Menzies was limited and failed to distinguish between Strickland's two elements. As a reminder, Strickland requires a defendant to "demonstrate that (1) his counsel's performance was deficient in that it 'fell below an objective standard of reasonableness' and (2) 'the deficient performance prejudiced the defense.'" State v. Ray, 2020 UT 12, ¶ 24, 469 P.3d 871 (quoting Strickland, 466 U.S. at 687-88). Torres points out that Menzies "did not parse" these two elements when deciding the degree of deference owed to a district court's application of the Strickland standard to the facts before it. And he asserts that, had we done so, we would have concluded that a district court's prejudice determination is more fact-like than lawlike. We disagree.
¶58 Torres does not challenge Menzies' reasoning as far as it goes, nor does he explain why that reasoning would not also have led to the same result had we analyzed the two Strickland elements separately. Our conclusion in Menzies that ineffective assistance was a law-like mixed question hinged on the second Levin factor: "the degree to which a trial court's application of the legal rule relies on facts observed by the trial judge, such as a witness's appearance and demeanor, relevant to the application of the law that cannot be adequately reflected in the record available to appellate courts." Menzies, 2006 UT 81, ¶ 58 (cleaned up) (emphasis added). We observed that appellate courts "regularly decide[]" ineffective assistance claims on the record, because such claims most frequently arise for the first time on direct appeal. See id. On the flip side, we observed that because ineffective assistance claims do not usually arise in the district court after trial, "direct observations" of things outside the appellate record, such as a district court's firsthand observations of "a witness's appearance and demeanor," are not usually available in an ineffective assistance analysis. Id. (cleaned up). Thus, in determining that the application of the Strickland standard to particular facts is more law-like than factlike, this court's Levin analysis focused on the typical ineffective assistance posture, rather than the less frequent occurrence of such claims arising in the district court. See id. ("[I]t is unnecessary to grant deference to the district court in the minority of cases where an ineffective assistance of counsel claim is first raised before that court."). And that reasoning is equally true with respect to both Strickland elements: appellate courts frequently decide both issues in the first instance.
¶59 Taking a fresh look for ourselves, we find persuasive Menzies' holding that an ineffective assistance determination is a law-like mixed question-although the weight we place on the various Levin factors differs somewhat from the Menzies court.
¶60 We have discussed the significance of the three factors in this way:
The first and second Levin factors assess whether a particular mixed question is best resolved by either a fact-finding tribunal or an appellate court based on the relative competencies of these two types of courts. District courts and fact-finding administrative bodies are in a superior position to weigh facts that depend upon credibility determinations, the direct observation of witness testimony, and other evidence not fully captured in a written appellate record. The degree to which a mixed question is based upon facts observed by a lower tribunal determines whether the second factor weighs for or against a deferential standard of review. Appellate courts, on the other hand, have the capacity to create broad rules that can create a greater degree of consistency and predictability to future cases involving a particular mixed question. The degree to which a mixed question is based upon a complex variety of facts determines whether an appellate court can create useful precedent and, thus, whether the first factor weighs for or against de novo review.Sawyer v. Dep't of Workforce Servs., 2015 UT 33, ¶ 13, 345 P.3d 1253.
¶61 As discussed, in Menzies, we concluded that the first Levin factor-"the degree of variety and complexity in the facts to which the legal rule is to be applied," Levin, 2006 UT 50, ¶ 25-weighed in favor of viewing the application of the ineffective assistance standard as a fact-like question, see Menzies, 2006 UT 81, ¶ 58. But we concluded that the second and third Levin factors weighed in favor of a law-like classification. See Menzies, 2006 UT 81, ¶ 58.
¶62 Menzies' classification of ineffective assistance as a legal determination relied primarily on the second Levin factor-"the degree to which a trial court's application of the legal rule relies on facts observed by the trial judge, such as a witness's appearance and demeanor, . . . that cannot be adequately reflected in the record available to appellate courts." Id. ¶ 58 (cleaned up). We observed that "[i]neffective assistance of counsel claims are a unique species of claim that are frequently raised for the first time on appeal," so they are almost always decided based on the record. Id. But while that is empirically true, on reflection, we find more significant that "Strickland's requirement of a 'reasonable probability' of a different outcome" is a "high hurdle to overcome." State v. Garcia, 2017 UT 53, ¶ 44, 424 P.3d 171 (quoting 466 U.S. at 694). So we expect that an error sufficiently detrimental to the defense to undermine our confidence in the proceeding will usually be perceptible from the record.
¶63 The third factor-"other policy reasons that weigh for or against granting discretion to trial courts," Menzies, 2006 UT 81, ¶ 58 (cleaned up)-"is a catchall category under which an appellate court may weigh other considerations," Sawyer, 2015 UT 33, ¶ 14. Torres argues that this factor weighs in favor of a fact-like classification, because "the evidentiary picture presented in each case is individualized and highly variable." Torres therefore argues that "prejudice determinations do not need a defined set of specific rules," and "there is not a need for a high degree of uniformity" across Utah law on this issue. But this argument treads ground already covered by the first Levin factor: "the degree of variety and complexity in the facts to which the legal rule is to be applied." Levin, 2006 UT 50, ¶ 25. And we acknowledge here, as we did in Menzies, that the first factor "would favor granting deference to the district court['s]" prejudice determination. 2006 UT 81, ¶ 58. Accordingly, Levin's "cost-benefit analysis" already accounts for Torres's argument on this point, and we must consider other policy reasons for or against deference in this situation. See Sawyer, 2015 UT 33, ¶ 12.
¶64 The policy reason that weighs most heavily in favor of a correctness standard of review for Strickland prejudice determinations is that ineffective assistance is an issue of constitutional dimension. And when it comes to application of a constitutional standard, "we have held that policy considerations dictate that the application of the legal concept should be strictly controlled by the appellate courts." Levin, 2006 UT 50, ¶ 23; see also Randolph, 2022 UT 34, ¶ 40. So "irrespective of the difficulties inherent" in reviewing a fact-intensive question for correctness, generally, "[w]e have not ceded deference in . . . cases" involving application of a constitutional standard. Sawyer, 2015 UT 33, ¶ 14 (cleaned up); see also id. (listing "mixed questions with constitutional dimensions that we have reviewed de novo for policy reasons," including whether a police interrogation was custodial, Levin, 2006 UT 50, ¶¶ 41-42, whether a confession was voluntary, State v. Thurman, 846 P.2d 1256, 1271 (Utah 1993), whether a defamatory statement was made with actual malice, Jensen v. Sawyers, 2005 UT 81, ¶¶ 91-92, 130 P.3d 325, and whether speech is obscene, City of St. George v. Turner, 860 P.2d 929, 932-33 (Utah 1993)). This is true even where the first two Levin factors suggest we should review the issue deferentially. See, e.g., Randolph, 2022 UT 34, ¶¶ 37-40, 44 (concluding that application of the "substantial evidence" standard under article I, section 8 of the Utah Constitution presents a law-like mixed question, even though the first two Levin factors suggest it is fact-like).
¶65 And although Strickland prejudice determinations involve a wide array of factual circumstances in that the alleged errors by trial counsel come in a variety of forms, the correctness standard of review has lent a "greater degree of consistency and predictability to future cases involving" ineffective assistance. See Sawyer, 2015 UT 33, ¶ 13; accord U.S. Bank Nat'l Ass'n ex rel. CWCapital Asset Mgmt. LLC v. Vill. at Lakeridge, LLC, 583 U.S. 387, 396 n.4 (2018) ("[T]he role of appellate courts in marking out the limits of a standard through the process of case-by-case adjudication favors de novo review even when answering a mixed question primarily involves plunging into a factual record." (cleaned up)). Many cases analyzing ineffective assistance claims look to other decisions with similar circumstances as helpful data points. See, e.g., Wyatt v. State, 2021 UT 32, ¶¶ 39-40, 493 P.3d 621 (comparing defendant's argument that counsel's failure to object to jail officer's testimony prejudiced him to similar argument made in State v. Maurer, 770 P.2d 981 (Utah 1989)); State v. Newton, 2020 UT 24, ¶¶ 32-34, 466 P.3d 135 (distinguishing State v. Barela, 2015 UT 22, 349 P.3d 676, where defendant similarly argued that counsel's failure to object to erroneous jury instruction prejudiced the defense); State v. Griffin, 2016 UT 33, ¶ 59, 384 P.3d 186 (rejecting defendant's argument that counsel's failure to investigate an alleged chain of custody issue was prejudicial, based on a "similar situation" in State v. Bradshaw, 680 P.2d 1036 (Utah 1984)). If similar attorney actions could be prejudicial in one case, but not in another, based solely on the varying standard of review, this consistency would be undermined.
¶66 For these reasons, we reaffirm Menzies' holding that a court's application of the Strickland standard, including its prejudice prong, is reviewed for correctness. Although we conclude that the third Levin factor provides an even more persuasive rationale for this holding than the second-a variance from our analysis in Menzies-we continue to find Menzies' ultimate holding persuasive.
2. Establishment in the Law
¶67 Torres asserts that our second Eldridge factor also weighs in favor of abandoning Menzies. We are unpersuaded, however, by Torres's arguments on this point.
¶68 Our second Eldridge factor, "how firmly [the] precedent has become established in the law," takes into account "a variety of factors[,] including the age of the precedent, the public reliance on the precedent, the workability of the precedent, and the consistency of the precedent with other principles of law." Rutherford v. Talisker Canyons Fin., Co., 2019 UT 27, ¶ 28, 445 P.3d 474. These considerations can be critical because, even if the precedent is unpersuasive, we must assess whether "more good than harm will come by departing from" it. State v. Robertson, 2017 UT 27, ¶ 38, 438 P.3d 491 (cleaned up).
¶69 On this factor, Torres argues that Menzies "is inconsistent with caselaw that applies a deferential standard of review" to new trial rulings. He cites Judge Harris's concurrence in the court of appeals, which observed that other types of district court decisions that have a prejudice component-e.g., whether a mistrial is in order, whether the defendant's Sixth Amendment right to a fair trial was violated, and whether particular evidence is more prejudicial than probative-are reviewed for abuse of discretion, not correctness. See State v. Torres-Orellana, 2021 UT App 74, ¶ 52, 493 P.3d 711 (Harris, J., concurring) (citing State v. Maestas, 2012 UT 46, ¶ 325, 299 P.3d 892; State v. Daniels, 2002 UT 2, ¶ 19, 40 P.3d 611; Arnold v. Grigsby, 2018 UT 14, ¶ 25 n.5, 417 P.3d 606). But as we have explained, see supra Section I.A, our abuse of discretion standard requires an appellate court to review "a legal conclusion . . . embedded in a district court's discretionary determination" for correctness, Boyden, 2019 UT 11, ¶ 21. And under Menzies, we treat a Strickland determination as a legal conclusion. See 2006 UT 81, ¶ 58. So when a party appeals a district court's new trial ruling based on ineffective assistance, we review the court's Strickland determination for correctness-consistent with our abuse of discretion standard. As for the other types of discretionary decisions Judge Harris identified, see Torres-Orellana, 2021 UT App 74, ¶ 52 (Harris, J., concurring), the abuse of discretion standard similarly requires that we reverse if the decision was "premised on flawed legal conclusions," Archuleta, 2011 UT 73, ¶ 152 (cleaned up). We are therefore unpersuaded that Menzies is inconsistent with how we review new trial rulings generally or other, similar, decisions by a district court.
¶70 To be clear, however, our rejection of this argument should not be taken to minimize the importance of a district court's factual findings with respect to an ineffective assistance claim. The correctness standard applies to the district court's application of the Strickland standard to the facts. But any factual findings made by the district court are given deference. See Menzies, 2006 UT 81, ¶ 58. So to the extent that a district court sets out its firsthand observations in its findings of fact, appellate courts will give them weight and review them under the most deferential "clearly erroneous" standard. Id.
¶71 Torres also points out that Menzies is less than two decades old, and he asserts it has been relied on infrequently. But while this is often a helpful metric in determining whether a precedent is firmly established, it does not tell us much here. The relevant holding from Menzies is that the standard of review for an ineffective assistance claim is correctness. 2006 UT 81, ¶ 58. So, in every appellate decision since Menzies where a Strickland determination has been reviewed, the court should have applied a correctness standard of review. And this would be in accord with Menzies, whether or not the court cited Menzies for this proposition. Further, as we explained above, although we applied the Levin factors to ineffective assistance claims for the first time in Menzies, the result of our Levin analysis was not a departure from prior practice. See supra ¶ 46. Even before Menzies, we reviewed ineffective assistance claims for correctness. See, e.g., Templin, 805 P.2d at 186; State v. Hay, 859 P.2d 1, 4-5 (Utah 1993). So, while we re-confirmed in Menzies that correctness was the appropriate standard of review for ineffective assistance claims under our Levin factors, correctness has been the governing standard of review for such claims for decades. Regardless of the number of times that Menzies has been specifically cited for this proposition, what matters is that correctness is firmly established in Utah as the standard of review for ineffective assistance claims.
¶72 In sum, Torres fails to show that either Eldridge factor weighs in favor of abandoning Menzies. And we decline to overrule it.
II. Torres Was Not Prejudiced by the Error He Identifies
¶73 Although the district court identified several ways in which it deemed counsel's performance to be deficient, Torres's argument on certiorari focuses on his counsel's failure to introduce numerous text messages that he and Tiffany had exchanged following the rape. The court of appeals disagreed, concluding that even if counsel had performed deficiently for this reason, Torres was not prejudiced by the error. State v. Torres-Orellana, 2021 UT App 74, ¶ 33, 493 P.3d 711. It reasoned that, based on the texts the State had admitted at trial, the jury was already "aware that Tiffany had continued to communicate with Torres and wished, at least at first, to continue [the] relationship." Id. And in light of "[a]mple additional evidence support[ing] Torres's conviction," the court of appeals concluded that there was not "[a] reasonable probability" of a different result had counsel introduced the additional text messages. Id. ¶¶ 29, 34 (quoting Strickland v. Washington, 466 U.S. 668, 694 (1984)); see id. ¶¶ 34-36.
While Torres mentions several other errors identified by the district court, he does not rely upon them to challenge the court of appeals' prejudice analysis. Torres notes, for instance, that the district court concluded that counsel should have "challenge[d] the plausibility of Tiffany's allegations." He does not argue, however, that the court of appeals erred in concluding that counsel's failure to do so was not prejudicial. See State v. Torres-Orellana, 2021 UT App 74, ¶¶ 37-39, 493 P.3d 711. He also mentions the district court's observations that "counsel asked few questions" overall and "did not subject the State's case to meaningful adversarial testing." But the district court made these statements in the context of its conclusion that Torres received "a constructive denial of counsel"-a type of ineffective assistance where "counsel completely fails to subject the opposition's case to meaningful adversarial testing" and "prejudice is legally presumed." Menzies v. Galetka, 2006 UT 81, ¶ 98, 150 P.3d 480. The court of appeals did not address this specific conclusion because "[b]oth sides agree[d] that [the district court's] determination was in error." Torres-Orellana, 2021 UT App 74, ¶ 24 n.8. And Torres does not argue on certiorari that the court of appeals erred in this respect. Finally, Torres points out that the district court faulted counsel for "not properly question[ing] the SANE nurse about whether [Tiffany's] injuries [could have] come from consensual . . . sex." While that is true, this purported error was not among those the district court identified as a reason why Torres had received ineffective assistance. And Torres does not argue that, had counsel engaged in the questioning identified by the district court, the result of the proceeding would have been different. Accordingly, we focus on Torres's argument that the court of appeals erred in concluding that he was not prejudiced by his counsel's failure to introduce additional text messages.
¶74 Torres contends that the court of appeals erred in concluding that counsel's error did not prejudice him. He argues that even though "the jury knew of some post-incident contact between" him and Tiffany, "[t]he content of the full text messages [would have] affected the jury's determination of Tiffany's credibility." Torres also argues that the "additional evidence" relied on by the court of appeals is less supportive of his conviction than the court of appeals described.
¶75 We agree with the court of appeals. Even assuming that counsel's failure to introduce more of Tiffany and Torres's text messages was objectively unreasonable, Torres has failed to show that this perceived error "undermine[s] confidence in the outcome" of his trial. Strickland, 466 U.S. at 694.
¶76 First, while the additional texts may have been relevant to the jury's assessment of Tiffany's credibility, Torres oversells the effect the texts may have had on that assessment. After all, when asked about some of the texts the State had introduced at trial, Tiffany explained why she had continued to text Torres after he raped her and why her messages appeared affectionate in nature. The State asked Tiffany what had been going through her mind when, days after the rape, she texted Torres that she "just want[ed]" him, that she did not "want to lose [him]," and that she didn't "want anyone else." Tiffany explained, "I was super hurt, but I was also really dumb and thought that I could give him another chance." Tiffany further testified that her "hormones were all over the place from having a baby," that she "just wanted to feel loved," and that she had "liked [Torres] a lot before" he raped her. And she stood by her sentiment at the time, that she "didn't want to lose him." But she decided that she ultimately "couldn't be with someone like that"-someone who "did that to [her]."
¶77 Tiffany's explanation for the texts that were admitted at trial also accounts for why, as the district court found, she and Torres had exchanged numerous other texts of an "affectionate nature" following the incident. So if counsel had introduced these other texts in an attempt to discredit Tiffany's allegations of rape, the jury would already have been primed to see these texts for what Tiffany said they meant-that she had lingering feelings for someone who she came to realize she simply could not be with. And we have no reason to doubt that, if asked about why she sent these other texts, Tiffany would have provided the same explanation. Counsel's failure to introduce these other texts therefore does not undermine our confidence in the result at trial.
¶78 Torres argues that this conclusion goes against what we held in Gregg v. State, 2012 UT 32, 279 P.3d 396. In that case, the defendant was convicted of rape and claimed that he had received ineffective assistance because his attorney "fail[ed] to investigate facts that would have undermined the credibility of the victim." Id. ¶¶ 2, 23. Specifically, "only two days after the alleged rape," the victim had "sent four cheerful, light-hearted e-mails" to other men using the same online dating site where she met the defendant. Id. ¶¶ 4, 23. We noted that the victim "was the only witness that provided direct evidence" that the sex was not consensual, and that she testified that "she only logged on to [the dating site] after the alleged rape to aid the police investigation." Id. ¶ 23. Accordingly, the four emails would not only have contradicted her testimony, but they also would have "undermined her credibility regarding the severe anxiety and panic attacks she claimed to suffer as a result of the incident." Id. We held that counsel's failure to investigate these facts was not only objectively unreasonable, but also prejudicial to the defendant's case. See id. ¶¶ 27-30.
¶79 While there are some similarities between Gregg and this case, the differences carry the day. Unlike in Gregg, Tiffany's testimony is not the only direct evidence of rape; and as discussed below, the other evidence is weighty. See infra ¶¶ 81-82. Also unlike in Gregg, the content of Tiffany's texts to Torres following the incident does not directly contradict Tiffany's trial testimony. And although we reasoned in Gregg that the tone of the victim's emails would have undermined her claims of anxiety and panic attacks, Tiffany's texts to Torres are reconcilable with her trial testimony. Again, Tiffany testified that her feelings toward Torres were genuine even after the incident, but that, in time, she came to the conclusion that she could no longer be with him. Gregg is therefore distinguishable, and it does not control our decision here.
¶80 Second, like the court of appeals, we conclude that whatever effect the additional texts would have had on the jury, there was "[a]mple additional evidence support[ing] Torres's conviction." Torres-Orellana, 2021 UT App 74, ¶ 34. The State introduced several of Torres's texts to Tiffany following the incident, and they show that he acknowledged that Tiffany had not consented to sex. As the court of appeals noted, "when . . . Tiffany unambiguously confronted Torres the next day, 'How am I supposed to let someone who forced me to have sex fix it. How are you supposed to fix that,' Torres responded, 'I'm so sorry for that I didn't mean to, n I don't know, I just wish there was I way.'" Id. (emphasis omitted).
¶81 Torres argues that, "[i]n the context of dry text messages, all that can be known from [his] response is that [he] was sorry when he learned that Tiffany had been hurt after the fact" and that "it does not shed any light on . . . [his] mindset at the time of the sexual encounter." (Emphasis omitted.) But in the same exchange, Tiffany explicitly confronted Torres about how he "knew [she] didn't want to have sex" and how she had "told [him] multiple times and [he] still did it." He replied, "What do you mean? N Babe I did but I saw that you wanted too but I didn't know, I didn't wanted to do it either, n i know you told me multiple times, n I'm so sorry I should of just stoped. Forgive me, I won't do it again." And when Tiffany asked, "[W]hy did you do it even though I told you no and I didn't want to," Torres said, "I don't know babe,n I know I'm sorry, I don't know how to tell you I'm sorry, I don't want to be fighting with you, I want to be a happy couple, forgive me I won't do it again." Considering these texts in context, we disagree with Torres that the texts showed only that he was sorry for having hurt Tiffany and that they shed no light on whether he knew she had not consented to sex.
¶82 Further, the SANE nurse testified that, in her twelve years of experience, she had rarely seen injuries as severe as Tiffany's. Torres sweeps this evidence to the side on the ground that counsel should have more fully cross-examined the SANE nurse about whether Tiffany's injuries were consistent with consensual sex. But as the court of appeals pointed out, the SANE nurse in fact testified about whether the sex could have been consensual. She explained that the "several sites of injury . . . indicate[d] that there[] [was] not a correction to prevent injury"-a "correction" one would expect during consensual sex. Torres also does not separately argue that counsel provided ineffective assistance for failing to cross-examine the nurse on this point. So we take the SANE nurse's testimony as it is. And based on her testimony regarding Tiffany's injuries the night of the incident, we agree with the court of appeals that it was "particularly damaging to Torres's defense" and "provided strong evidence in support of Tiffany's rape allegation." Id. ¶ 35.
¶83 Given the weight of the other evidence that corroborated Tiffany's testimony of nonconsensual sex, we agree with the court of appeals that there is not a reasonable probability of a different result had counsel introduced additional texts between Tiffany and Torres. Accordingly, Torres fails to show that the court of appeals erred in concluding that counsel's failure to do so did not prejudice him.
CONCLUSION
¶84 While a district court's new trial ruling is reviewed for an abuse of discretion, any underlying legal conclusions-including a determination of ineffective assistance of counsel-are reviewed for correctness. The court of appeals correctly applied that standard of review, and we agree with the court that Torres was not prejudiced by counsel's failure to introduce additional texts at trial. Therefore, he fails to establish ineffective assistance under Strickland. And because ineffective assistance was the basis of the new trial ruling, Torres also fails to show that the court of appeals erred in reversing the district court's grant of a new trial. We affirm.