INTRODUCTION
¶1 In 2012, a jury convicted Caroline Ashby of two counts of aggravated sexual abuse of a child for allegedly abusing her son, Kevin,
1
while they were bathing together. Ashbys conviction rested on allegations Kevin made first during a Childrens Justice Center (CJC) interview when he was eight years old, and later at trial when he was ten.
¶2 About a decade later, having had no contact with his mother during that period, Kevin recanted his statements. Kevin divulged that he had lied to the CJC interviewer and at trial, and that Ashby had never sexually abused him. Based on Kevins recantation, Ashby filed a petition for a post-conviction determination of factual innocence pursuant to Utahs Post-Conviction Remedies Act (PCRA).
¶3 After holding an evidentiary hearing, the district court denied the petition, ruling that Ashby had failed to prove her factual innocence by clear and convincing evidence. The court indicated that it would be difficult to meet the clear and convincing burden of proof with a recantation, even if that recantation was reconcilable with the undisputed surrounding facts. But the court found Kevins recantation could not be reconciled with the “undisputed facts.”
¶4 On appeal, Ashby argues that the district court erred in its application of the clear and convincing evidence standard. To the extent the court held Ashby to a higher standard because she sought to prove her factual innocence with a recantation, it incorrectly inflated the burden of proof. Where a defendant is convicted based on uncorroborated witness testimony and that witness later recants under oath, that recantation, if credible, is sufficient to prove factual innocence by clear and convincing evidence. Although the district court must carefully assess the circumstances surrounding the recantation and the witnesss credibility to determine if the recantation is believable, the district court made no express findings that Kevins testimony at the evidentiary hearing was false.
¶5 The State argues that, by finding Kevins recantation to be “irreconcilable” with the “surrounding undisputed facts,” the district court implicitly found that Kevin was not credible. But even assuming the district court envisioned that finding as an implicit credibility determination, the finding that Kevins testimony conflicted with the “undisputed facts” is not supported by the record. In any event, proof by clear and convincing evidence does not require the petitioner to eliminate or reconcile all conflicts in the evidence. While such discrepancies may bear on the credibility of the recanting witness, the existence of conflicting evidence is not determinative.
¶6 Both parties agree that Kevins recantation, if believable, is sufficient to prove Ashbys factual innocence by clear and convincing evidence. We therefore remand to the district court to determine whether the recantation is credible in accordance with the legal standards set forth in this opinion.
BACKGROUND
¶7 After a jury trial in 2012, Ashby was convicted of two counts of aggravated sexual abuse of her son, Kevin. At trial, no physical evidence of abuse or eyewitness testimony was presented. The only evidence of abuse was Kevins testimony that Ashby had taken indecent liberties with him while bathing together.
A. Events Leading up to the CJC Interview
¶8 Kevin was born to Caroline Ashby and David Ashby (Father) in 2002. Eventually the couple divorced, and Father married Stepmother. After the divorce, Ashby and Father engaged in a high-conflict relationship regarding custody and visitation, which Kevin often observed.
¶9 When Kevin was between six and eight years old, he began exhibiting behavioral issues, including throwing tantrums, being argumentative, lying, saying “outlandish” things, and acting out against his younger brother. Kevin also began acting out sexually toward other children.
¶10 As a result of his behavioral issues, Kevin began seeing a therapist. In therapy, Kevin candidly disclosed his sexual behavior with other children but never disclosed any abuse or sexual behavior with his mother. Kevin told his therapist that he bathed with Ashby and it made him uncomfortable, but he specified that Ashby wore clothing when she bathed with him. At some point, Father and Stepmother learned that Kevin was bathing with Ashby, which Father and Stepmother thought was inappropriate. Eventually, either Father and Stepmother or Kevin himself told the therapist that Ashby and Kevin bathed together naked.
¶11 Kevins ongoing sexual behaviors prompted Stepmother to call the Division of Child and Family Services, which resulted in an interview at the CJC.
B. The CJC Interview
¶12 About two years before Ashbys trial, Father and Stepmother took Kevin to the CJC to be interviewed. Kevin knew Father and Stepmother were concerned about his sexual behaviors. Stepmother explained to the CJC interviewer that she and Father thought Ashbys bathing with Kevin was inappropriate and that it was related to why he was having behavior issues. Stepmother told the interviewer that she felt Kevin, who had just turned eight at the time, was too old to see his mother naked, and that she was concerned it was “becoming a problem with him noticing different [body] parts.”
¶13 The interviewer began by asking Kevin about school, but then turned to what he knew about “privates” and how he knew about them. For the remainder of the interview, the interviewer primarily questioned Kevin about bathing with Ashby. Kevin reported that he and Ashby would take baths together and that he “used to do it naked.” When the interviewer asked Kevin to tell her more, Kevin sighed and said, “Its embarrassing to me.” Kevin sighed again and explained that it was embarrassing because “its just really bad and I didnt know that then.”
¶14 When asked what he and Ashby did in the bathtub, Kevin reported that they played with toys and washed themselves. Kevin explained that he would wash Ashby, and she would “wash everything except [his] private parts,” which he washed himself. Later, he said that Ashby did wash his “private parts” but that he “didnt wash hers.” The interviewer then showed Kevin some drawings of children with no clothes on and asked Kevin to identify by name various body parts. Kevin then reported that Ashby had him wash her “private parts,” including her “boob[s]” and “nipple[s],” “some of the butt cheek,” “inside [the] bum,” and “a little inside” her vagina. Kevin first claimed that he had washed inside her vagina with his “whole hand,” but adjusted his answer to indicate that it was just one finger.
¶15 Kevin was never asked during the interview whether he engaged in sexual behavior with other children, nor did he disclose such behavior to the interviewer.
¶16 Shortly after the CJC interview, Kevin began seeing a second therapist. With this therapist, Kevin denied that he had engaged in any sexual behavior with other children, but he claimed that his mother had sexually abused him over 100 different times.
C. Ashbys Trial
¶17 Among other witnesses, Kevin, Father, Stepmother, and Kevins therapists testified at Ashbys trial, and the jury was shown a video of Kevins CJC interview. No physical evidence of abuse was presented, and Father and Stepmother testified that, while they disapproved of Ashby bathing with Kevin, they never suspected Ashby of abuse. The only evidence of abuse was Kevins testimony.
¶18 Kevin was ten years old when he testified, and his testimony was given outside the presence of the jury and Ashby. When asked about “private parts,” Kevin testified that he had seen Ashbys private parts, but no one elses, and that Ashbys private parts had no hair. He described a vagina as “[t]wo ovals stick together” with a hole, and “boobs” as “two bouncy balls” with “[a] nipple.”
¶19 Kevin testified that he would usually see Ashbys private parts in the bathtub, but that they would also sometimes wear swimsuits. He testified that while in the bathtub, he would wash the “outside” of Ashbys breasts, buttocks, and vagina. He denied washing inside her vagina or her anus, although he admitted that he said the opposite in his CJC interview.
¶20 The jury convicted Ashby on both counts of aggravated sexual abuse of a child, and she was sentenced to prison.
D. Kevins Recantation
¶21 Sometime after Ashbys trial, both of Kevins biological parents relinquished their parental rights, and Kevin was eventually adopted by his paternal aunt and uncle. When Kevin was seventeen years old, he disclosed to his adoptive mother that he had lied about Ashbys conduct during the CJC interview and at trial. At that point, Kevin had not had any contact with Ashby since he was eight years old.
¶22 Kevins adoptive mother took him to see a psychologist, Dr. Goldsmith, to “determine whether [Kevin] is changing his story on his own free will or, whether he has been pressured to change his narrative.” Dr. Goldsmith reviewed transcripts of the jury trial and sentencing hearing and interviewed Kevin twice.
¶23 Kevin told Dr. Goldsmith that, at the time he accused Ashby, a neighbor boy had been sexually abusing Kevin since he was about six years old. Kevin explained that he began acting out with other children because of the sexual behaviors he learned from the neighbor boy. Kevin recalled going to the CJC as a result of his sexual behaviors and said he “didnt want [his] friend to get in trouble, so [he] said it was [Ashby].” Kevin told the psychologist that it “was all a lie.” He remembered bathing with Ashby but said “there was nothing sexual about it.” At the time, Kevin “wasnt scared for [his] mom. [He] was getting through the interview and told them what they wanted to know.”
¶24 Kevin told Dr. Goldsmith that the one time he had asked to do something inappropriate with Ashby, she had corrected him. Once, after watching babies nursing, Kevin asked to suck on Ashbys nipples, and she said no. He was embarrassed and thought he shouldnt have asked that.
¶25 Dr. Goldsmith prepared a report in which he opined that Kevin was not “under any external pressure to recant” and was “acting of his own free will.” He found Kevin to be “forthright throughout the interviews” and “was exposing himself to deep emotional risks by attempting to set the record straight.”
¶26 Dr. Goldsmith also noted that Kevin has ADHD. During the interview with Dr. Goldsmith, Kevin “provide[d] short answers and often struggle[d] with accessing his memories[,] which is not atypical for children with ADHD.” Dr. Goldsmith concluded that “it is very likely that as a ten-year-old child in court, [Kevin] provided quick answers in order to put an end to a stressful interview” and offered excerpts from Kevins trial testimony as examples of such an exchange.
E. Letter to the Board of Pardons
¶27 Shortly after visiting Dr. Goldsmith, Kevin wrote a letter to the Utah Board of Pardons asserting that he had lied in the CJC interview and at Ashbys trial when he accused her of sexual conduct. In his letter, Kevin explained that his sexual behavior with other children was actually the result of being abused by a neighbor boy, something that Kevin had never disclosed. The “sexual play” with the neighbor boy “taught [him] that horrible sexual things were the way kids played together.” As a child, Kevin believed adults could not “get in trouble.” So he reasoned that he could protect the neighbor boy, whose sexual conduct he had come to enjoy, by blaming his mother for his sexual reactivity with other children. Kevin told the Board of Pardons that Ashby was innocent and should not be in prison.
¶28 Kevin also recalled in his letter some specific memories with Ashby, including bathing with her. His letter did not indicate whether they were naked or wearing swimsuits during the baths, but he affirmatively remembered that there was nothing sexual about them. He recalled, “When I was questioned in court I know I lied about this and remember thinking that I should make it sound convincing.”
¶29 The Board of Pardons held a hearing in which Kevin was placed under oath and testified. Kevin read his letter and maintained that his previous allegations against his mother were false. Following the hearing, the Board of Pardons granted Ashby parole.
F. Ashbys Innocence Petition
¶30 After the Board of Pardons hearing, Ashby filed a petition for the determination of her factual innocence under Utah Code section 78B-9-402 and rule 65C of the Utah Rules of Civil Procedure, attaching Kevins letter to the Board of Pardons and a declaration from Dr. Goldsmith.
¶31 The State moved for summary judgment, asserting that recantations are viewed with suspicion and are a disfavored basis for innocence claims, and that Kevins recantation in particular did not ring true. The State argued that (1) Kevins recantation never affirmatively disavowed the specific conduct underlying Ashbys conviction and only expressed his subjective belief about not being sexually abused; and (2) it is “implausible” that an eight-year-old would be able to provide accurate descriptions of female anatomy without having engaged in the conduct of seeing and touching female genitals.
¶32 Ashby opposed the States summary judgment motion, asserting that it is inappropriate for a court to weigh credibility in summary judgment proceedings. Ashby also argued that (1) Kevins statements to Dr. Goldsmith and the Board of Pardons did disavow the conduct underlying Ashbys convictions; and (2) because more than one inference could be drawn from Kevins testimony about female anatomy, summary judgment was not appropriate. Ashby also identified a number of disputed material facts that made summary judgment inappropriate.
¶33 The district court never entered an order regarding the summary judgment motion. Instead, it lifted the no-contact order that had been in place so the parties could depose Kevin. The State subsequently withdrew its summary judgment motion.
G. Kevins Deposition
¶34 The parties deposed Kevin, then eighteen years old, almost a year after his hearing with the Board of Pardons. Again, Kevin maintained that he had lied when he accused Ashby of sexual conduct in an attempt to protect a neighbor boy who had been sexually abusing him. Kevin explained that he knew he was being taken to the CJC because of his sexual behavior toward other children and that Father and Stepmother wanted to know why he was behaving that way.
¶35 Ashbys counsel questioned Kevin regarding the specific conduct underlying Ashbys conviction, and Kevin responded that she had never asked him to touch her genitals, nor had she ever tried to sexually stimulate him. When asked about how he came to know about female anatomy, Kevin could not remember, but thought it was possible he saw female nudity in an anatomy book, artwork, or pornography.
¶36 Kevin also recalled bathing with Ashby—she would wash him, but there was nothing sexual in nature about it. He never said whether they were naked or wearing swimsuits, but described the bathing as being more like a pool day in the bathtub and he remembered playing with toys.
¶37 After Kevins deposition, the State maintained that Kevins testimony was insufficient to prove Ashbys factual innocence. In the States view, Kevins lack of memory about how he came to know about female anatomy necessarily meant he had engaged in sexual conduct with Ashby.
H. Evidentiary Hearing
¶38 The court then held an evidentiary hearing in which Kevin testified. Consistent with the statements he made to Dr. Goldsmith, to the Board of Pardons, and in his deposition, Kevin testified that he had been sexually abused by a neighbor boy and that, as a result, he began acting out sexually with other children. Kevin described the abuse from the neighbor boy as “sexual touching,” but he was not asked about any details. He explained that he blamed Ashby in order to protect the neighbor boy, not realizing that Ashby could get into any trouble. Kevin was never asked about why he thought he needed to place blame on anyone in the CJC interview. Nor was he questioned regarding what he understood as the reason why he was being taken to the CJC.
¶39 Kevin affirmatively testified that he had lied at the CJC interview and at trial when he accused Ashby of sexual conduct. Kevin testified about bathing with Ashby, but he maintained that nothing sexual in nature ever occurred. When asked whether he remembered bathing nude with his mother, Kevin said that he did not and that his only clear memory of bathing together was when they had worn swimsuits and played with toys as if they were in a swimming pool.
¶40 Kevin could not specifically remember how he came to know about female anatomy as a child, but he knew it was not because of Ashby. Kevin testified that he could have snuck onto the internet, he probably saw female babies getting diaper changes, and he could have looked at pornography. He remembered being shown drawings of female anatomy during his CJC interview. He was never specifically asked whether he could remember ever seeing Ashby naked.
¶41 Following the evidentiary hearing, the district court found that Ashby had failed to show her factual innocence by clear and convincing evidence and denied her petition. The court noted that it would be difficult for Ashby to prove her innocence by clear and convincing evidence with recantation testimony because Kevin “testified that he has been a prevaricator.” Under these circumstances, the court thought it would be difficult to meet the clear and convincing standard even if Kevins “testimony [was] reconcilable with the undisputed surrounding facts.” But the court concluded that Kevins “testimony is irreconcilable.”
¶42 Specifically, the court identified three ways in which it deemed Kevins recantation “inconsistent with undisputed facts.” First, the district court found that it was “undisputed that [Ashby] bathed naked with [Kevin] during his childhood.” As support, the court pointed to the statements Kevin made as a child and his more recent statements to Dr. Goldsmith and the Board of Pardons in which Kevin “implied that they were naked during the baths.” Specifically, Kevin told the Board of Pardons that Ashby washed him while he played with toys, told Dr. Goldsmith that they washed each others backs, and told both that he had once asked to suck on Ashbys nipples. The court contrasted these statements with Kevins testimony at the evidentiary hearing “that there were only a couple of times they ever bathed together, that they wore swimming suits on both occasions, that he only played with toys, and that neither of them washed the other.”
¶43 Second, the court concluded that it was undisputed that Kevin “had not, at least up to the time of the CJC interview, ever seen pornography or female nudity other than [Ashbys].” The court observed that, at the CJC interview, Kevin “knew that a finger could be inserted into a vagina” and, at trial, “was able to accurately describe not only female anatomy in general” but also that his mothers pubic area was hairless. The court found that, at the evidentiary hearing, Kevin was “unable to explain how he acquired this knowledge” and “did not even admit to seeing [Ashbys] nude body.”
¶44 Third, the court noted that “the primary premise of [Kevins] recantation is that he impugned his mother to protect his friend.” But, the court found, “nothing in the CJC interview applied any pressure on [Kevin] to implicate anyone of anything or divert attention from the friend.” The court found that Kevins “description of his baths with [Ashby] bears no indications that he was covering for, or protecting, a friend who was abusing him” and that it was apparent that “he did not consider the baths to be abusive at the time of the CJC interviews.” On these bases, the court concluded that Ashby had failed to carry her burden to prove her factual innocence by clear and convincing evidence.
¶45 Ashby appeals.
STANDARD OF REVIEW
¶46 The ultimate determination of factual innocence is a mixed question of fact and law because it requires a district court to apply the clear and convincing standard to the evidence presented in each case. See Randolph v. State, 2022 UT 34, ¶¶ 30, 45, 515 P.3d 444; State ex rel. E.R., 2021 UT 36, ¶ 17, 496 P.3d 58. “Because a trial court is in a better position to judge credibility and resolve evidentiary conflicts, an appellate court reviews the trial courts findings of fact for clear error.” Brown v. State, 2013 UT 42, ¶ 37, 308 P.3d 486 (cleaned up). “We will set aside a district courts factual finding as clearly erroneous only if it is against the clear weight of the evidence, or if we otherwise reach a definite and firm conviction that a mistake has been made.” Id. (cleaned up). On the other hand, “we review the post-conviction courts legal conclusions for correctness, granting no deference to the district court.” Oseguera v. State, 2014 UT 31, ¶ 9, 332 P.3d 963 (cleaned up).
ANALYSIS
¶47 We begin with an overview of the statutory scheme under which claims of factual innocence are evaluated. Under Part 4 of the PCRA, titled “Postconviction Determination of Factual Innocence,” a person convicted of a felony offense may petition the district court “for a hearing to establish that the person is factually innocent of the crime or crimes of which the person was convicted.” UTAH CODE § 78B-9-402(1). The petition must demonstrate, among other things, that “newly discovered material evidence exists that, if credible, establishes that the petitioner is factually innocent.” Id. § 78B-9-402(2)–(3). If the court determines that the statutes threshold requirements are met and that “there is a bona fide and compelling issue of factual innocence regarding the charges of which the petitioner was convicted,” the district court must hold a hearing. Id. § 78B-9-402(9)(c)(i).
¶48 Section 78B-9-404 “sets forth how the evidentiary hearing is to proceed and gives direction to courts on how to determine factual innocence.” Brown v. State, 2013 UT 42, ¶ 40, 308 P.3d 486. “The burden is upon the petitioner to establish the petitioners factual innocence by clear and convincing evidence.” UTAH CODE § 78B-9-404(1)(b). As relevant here, “factual innocence” means that the petitioner did not “engage in the conduct for which the person was convicted.” Id. § 78B-9-401.5(2)(a).
¶49 Unlike a motion for a new trial, in which the district court only assesses the “probable weight” that a jury would afford the recantation on retrial, see State v. Loose, 2000 UT 11, ¶ 18, 994 P.2d 1237, a factual innocence petition requires the district court itself to act as factfinder—to weigh the evidence, assess credibility, and determine whether the petitioner has proven factual innocence by clear and convincing evidence. In determining whether the petitioner has met that burden, “the court shall consider, in addition to the evidence presented at the hearing under this part, the record of the original criminal case and at any postconviction proceedings in the case.” UTAH CODE § 78B-9-404(3).
¶50 If, “after considering all the evidence,” the court “determines by clear and convincing evidence that the petitioner” is factually innocent, the court shall order the conviction vacated with prejudice and expunged from the petitioners record. Id. § 78B-9-404(4)(a). Under the statutory scheme, a successful petitioner is also entitled to assistance payments based on the time spent incarcerated for a crime the petitioner did not commit. Id. § 78B-9-405.
¶51 In this case, the district court found that Ashby had satisfied the threshold showing under section 78B-9-402 and was entitled to an evidentiary hearing. After hearing Kevins sworn testimony and reviewing the record, the district court concluded that Ashby had not proven her factual innocence by clear and convincing evidence.
I. WHERE A CONVICTION IS BASED ON THE UNCORROBORATED TESTIMONY OF A SINGLE WITNESS, A FULL RECANTATION BY THAT WITNESS, IF CREDIBLE, IS SUFFICIENT TO PROVE FACTUAL INNOCENCE BY CLEAR AND CONVINCING EVIDENCE
¶52 Ashby contends that the district court erred as a matter of law by applying a heightened burden of proof to Ashbys factual innocence petition because it was based on a recantation. To succeed on a factual innocence petition, the petitioner bears the burden of proof by clear and convincing evidence. UTAH CODE § 78B-9-404(1)(b). The clear and convincing evidence standard “demands the introduction of evidence that makes the existence of the disputed facts very highly probable.” Randolph v. State, 2022 UT 34, ¶ 84, 515 P.3d 444 (cleaned up). It “is an intermediate standard of proof that implies something more than the usual requirement of a preponderance of the evidence; and something less than proof beyond a reasonable doubt.” Brown v. State, 2013 UT 42, ¶ 69 n.68, 308 P.3d 486 (cleaned up).
¶53 Ashby contends that the district court incorrectly held her to a higher burden of proof because her factual innocence claim was based on a recantation. In denying the petition, the court reasoned that because Kevin “testified that he has been a prevaricator,” it would be difficult for his testimony to provide clear and convincing evidence of Ashbys innocence. The court further observed that, “[e]ven [if Kevin]’s testimony [was] reconcilable with the undisputed surrounding facts, it would be difficult to meet this burden of proof.”
¶54 In Ashbys view, the district courts observation that it would be difficult to meet the clear and convincing standard with recantation evidence shows that the court applied the wrong legal standard. Ashby points out that even under the higher burden of proof beyond a reasonable doubt, testimony from an admitted prevaricator is sufficient to uphold a criminal conviction. (Citing State v. Stricklan, 2020 UT 65, 477 P.3d 1251.) If the testimony of an admitted prevaricator is sufficient to prove that a defendant is guilty beyond a reasonable doubt, Ashby insists that it must be sufficient to prove factual innocence under the lower standard of clear and convincing evidence.
¶55 We agree with Ashby that where a conviction rests entirely on the testimony of a single witness, a credible recantation by that witness, standing alone, is sufficient to prove factual innocence by clear and convincing evidence. To the extent the district court considered it more difficult to prove factual innocence with a recantation than with other evidence, it inflated Ashbys burden of proof.
¶56 As the State acknowledges on appeal, if Kevins recantation is “believable, then the recantation would establish [Ashbys] factual innocence” by clear and convincing evidence. To prevail on her petition, Ashby is required to prove that she did not “engage in the conduct for which [she] was convicted.” UTAH CODE § 78B-9-401.5(2)(a). Ashbys conviction was based solely on Kevins statements. Although there was some evidence to corroborate Kevins account that he and Ashby bathed together naked,
2
no witnesses or other evidence supported Kevins claim that Ashby had engaged in the sexual conduct for which she was convicted. And Kevins recantation was complete and unequivocal. He did not merely claim to have no recollection of the abuse; he testified that he remembered lying about the abuse at Ashbys trial. As the State puts it, “in the context of Ashbys case, if Kevins recantation testimony—the new evidence—were found to be credible, then Kevins recantation would show by clear and convincing evidence that she did not engage in the conduct for which she was convicted and, therefore, that she is factually innocent.”
¶57 Accordingly, the only question before the district court was whether Kevins recantation was credible, or, to be more precise, whether it was sufficiently credible to meet the clear and convincing evidence standard. In other words, the court was required to determine whether it was “very highly probable” that the recantation was true, see Randolph v. State, 2022 UT 34, ¶ 84, recognizing that clear and convincing evidence is an “intermediate standard of proof” that, “by its very nature, tolerates some doubt,” Brown, 2013 UT 42, ¶ 69 n.68.
¶58 In the States view, that is where the difficulty lies. The State asserts that “recantations are inherently unreliable and therefore using them to prove innocence is necessarily a challenging proposition.” The State reads the district courts challenged comments as merely acknowledging that “a recantation, by its very nature, calls into question the credibility of the recanting witness.”
¶59 There is ample support for the proposition that recantations are viewed “’with extreme suspicion’ ” and have “ ‘long been disfavored as the basis for a claim of innocence.’ ” Case v. Hatch, 731 F.3d 1015, 1041–42 (10th Cir. 2013) (quoting Carriger v. Stewart, 132 F.3d 463, 483 (9th Cir. 1997) (Kozinski, J., dissenting)). “Skepticism about recantations is especially applicable in cases of child sexual abuse where recantation is a recurring phenomenon.” People v. Schneider, 25 P.3d 755, 763 (Colo. 2001).
¶60 To be sure, recantations must be carefully scrutinized. “The recanting witness is admitting that he or she has lied under oath. Either the original sworn testimony or the sworn recantation testimony is false.” State v. McCallum, 561 N.W.2d 707, 712 (Wis. 1997). But there is no presumption that the recantation—as opposed to the trial testimony—is false.
3
Determining which story to credit requires a careful examination of the retracting witnesss credibility under oath and the circumstances surrounding the recantation.
¶61 Other courts have identified various factors to consider in determining whether a recantation is credible. See, e.g., Schneider, 25 P.3d at 762; State v. Worley, 476 P.3d 1212, 1221–22 (N.M. 2020); People v. Nelson, 171 A.D.3d 1251, 1253 (N.Y. App. Div. 2019). For example, the New Mexico Supreme Court has instructed district courts to “analyze the following factors, none of which is dispositive on its own,” to determine if a recantation is credible:
(1) The original verdict was based upon uncorroborated testimony; (2) the recantation is corroborated by additional new evidence; (3) the recantation occurred under circumstances free from suspicion of undue influence or pressure from any source; (4) the record fails to disclose any possibility of collusion between the defendant and the witness between the time of the trial and the retraction; and (5) the witness admitted the perjury on the witness stand and thereby subjected himself or herself to prosecution.
Worley, 476 P.3d at 1221–22 (cleaned up). In addition to those five factors, “the district court must also weigh the credibility of the witness.” Id. at 1222.
¶62 Although these factors are neither exclusive nor dispositive, they are helpful in assessing whether a particular recantation is reliable. Recantations are viewed with suspicion due to the real concern that witnesses may be motivated to retract accusations for reasons other than the truth. Therefore, examining the circumstances in which the recantation occurred is critical to assessing whether it is credible. To the extent the district court raised Ashbys burden of proof, we remand for the district court to reexamine Kevins recantation under the correct legal standard.
II. THE DISTRICT COURT DID NOT CONSIDER THE RELEVANT CIRCUMSTANCES SURROUNDING THE RECANTATION OR MAKE AN EXPRESS CREDIBILITY FINDING
¶63 Where a recantation, if believed, would establish the petitioners factual innocence, the district court must determine whether it is “very highly probable” that the recantation is true. Randolph v. State, 2022 UT 34, ¶ 84, 515 P.3d 444. In making that determination, a court must consider all the relevant circumstances and assess whether the recanting witness testified credibly under oath.
¶64 Here, Ashby contends that the district court “did not weigh the evidence, resolve factual disputes, make specific factual findings about what occurred when Kevin was a child, or make credibility determinations based on the witnesss demeanor and delivery while testifying[,]” but instead “merely compared the content of Kevins testimony [at the evidentiary hearing] to evidence in the trial record ․ without ruling whether Kevin told the truth as an adult or as a child.” We agree that the district court did not consider all the relevant circumstances or make the factual findings necessary to allow us to review its ruling.
¶65 For starters, the district court did not address the circumstances surrounding Kevins recantation. And there is ample evidence in the record that weighs in favor of the recantations reliability.
¶66 One reason recantations are viewed with suspicion is the risk that the petitioner, or those close to the petitioner, may have persuaded the witness to falsely recant. Here, the record reveals there was no “possibility of collusion between the defendant and the witness between the time of the trial and the retraction.” State v. Worley, 476 P.3d 1212, 1221 (N.M. 2020). It is undisputed that Kevin had no contact with Ashby between the time of the CJC interview and his recantation. And there is no evidence that Kevin was ever contacted by anyone on Ashbys behalf. Kevin first disclosed that he had lied about Ashby in a conversation with his adoptive mother, who is Fathers sister-in-law. There is no evidence of any relationship between Kevins adoptive mother and Ashby, and, in any event, Kevin disclosed the information unprompted and on his own initiative.
¶67 Indeed, “the recantation occurred under circumstances free from suspicion of undue influence or pressure from any source.” Id. After Kevin disclosed that he had lied about Ashby, his adoptive mother arranged for him to meet with Dr. Goldsmith to assess whether he was “changing his story o[f] his own free will or, whether he has been pressured to change his narrative.” Dr. Goldsmith concluded that Kevin was not “under any external pressure to recant” and was “acting of his own free will.” His adoptive mother later helped Kevin figure out how to appear before the Board of Pardons, but Kevin “wrote the letter to the parole board all by [him]self.” No one reviewed it or made any suggestions. And his adoptive parents told Kevin, “You do not have to do this if you dont want to. This is your choice and your choice only.” Even the State conceded at the evidentiary hearing that Kevin was not “being coerced in any way” and is “doing this of his own free will.”
¶68 Further, there is no evidence that Kevin had a motive to falsely recant. In cases involving sexual abuse by a parent, a false recantation may be prompted by the disruption to the childs family life following the parents conviction. See, e.g., State v. Stricklan, 2020 UT 65, ¶ 101, 477 P.3d 1251 (holding that a jury could reasonably find that the victims “recantation was motivated not by a desire to set the record straight but to ameliorate the negative consequences of [the defendants] absence from their home”). But despite living in foster care for a time, Kevin did not recant his testimony as a child in hopes of being reunited with his mother. Instead, he did not come forward until he was nearly an adult and only after he was in a stable and loving adoptive home.
¶69 Kevin testified in his deposition that he no longer has any connection to Ashby and did not come forward with the expectation of “getting any kind of money or relationship” with his mom. And the State did not challenge Kevins sincerity, telling the district court that the State was not “questioning ․ [Kevins] motives in coming forward now.” In short, nothing in the record suggests that Kevin stands to gain by falsely recanting his trial testimony. To the contrary, according to Dr. Goldsmith, Kevin “was exposing himself to deep emotional risks by attempting to set the record straight.”
¶70 And finally, Kevin recanted under oath at the evidentiary hearing, exposing himself to prosecution for perjury if he testified falsely. Although Ashby could offer no evidence to corroborate Kevins retraction, there was also no evidence to corroborate Kevins original accusations of abuse. Ashbys conviction was based solely on the statements that Kevin now swears were false.
¶71 There may be other evidence relevant to whether Kevins recantation is credible, including the circumstances under which the original accusations were made. “For example, the court may consider any motive to fabricate the initial accusation; observations by witnesses of the recanting witness when making the initial accusations; the nature and detail of both the accusations and the recantation; and whether the accusations were made under oath.” People v. Schneider, 25 P.3d 755, 762 (Colo. 2001). But here the district court did not weigh the relevant considerations, resolve conflicting evidence, or make factual findings about the reliability of either the original accusations or the recantation.
¶72 Nor did the district court make findings about Kevins credibility as a witness. In deciding whether to credit a recantation, district courts must “consider the credibility of the witness recanting under oath.” Id. District courts are in a superior position to assess credibility because they have the opportunity to observe firsthand “a witnesss appearance, demeanor, and overall credibility.” Randolph, 2022 UT 34, ¶ 38; see also Sawyer v. Dept of Workforce Servs., 2015 UT 33, ¶ 13, 345 P.3d 1253. “Manner of expression, sincerity, candor and straightforwardness are just some of the intangibles available to the trial judge in evaluating the credibility of recantation testimony.” State v. Carter, 354 A.2d 627, 631–32 (N.J. 1976). For that reason, “when an assessment of credibility turns on observing a witness and [the witnesss] demeanor, we afford deference to the trier of fact that had the opportunity to assess the witnesss credibility.” Stricklan, 2020 UT 65, ¶ 100 n.18. But here, the court made no credibility assessment based on its observations of Kevins testimony. The order contains no findings about the manner in which Kevin testified, nor does it suggest that the court found Kevin to be an unbelievable witness based on intangible factors not apparent from the cold record.
¶73 In fact, the district court never expressly made any adverse credibility determination. Although the court labeled section eleven of its order, “The Courts Findings on Credibility,” it contains only one express credibility finding: that Kevins testimony about not knowing, as a child, that he was getting Ashby into trouble was “believable.” The court never made the inverse finding that Kevins other testimony was not believable. We do not necessarily require a district court to make express credibility findings where “the findings of ultimate facts implicitly reflect consideration of the believability of the witnesses’ testimony.” In re Adoption of McKinstray v. McKinstray, 628 P.2d 1286, 1289 (Utah 1981). But here, the court made no findings of ultimate fact regarding whether Ashby “engage[d] in the conduct for which [she] was convicted.” See UTAH CODE § 78B-9-401.5(2)(a).
¶74 Both parties agree that Kevins recantation, if believable, is sufficient to prove Ashbys factual innocence by clear and convincing evidence. Because the district court did not address the relevant considerations or determine the recantations veracity, we cannot evaluate whether Ashby met her burden.
III. THE DISTRICT COURTS DETERMINATION THAT THE RECANTATION WAS IRRECONCILABLE WITH THE “UNDISPUTED EVIDENCE” IS CLEARLY ERRONEOUS
¶75 The State argues that the district court did, in fact, find that Kevins recantation was not believable. It urges us to read the courts finding that Kevins testimony was “irreconcilable” with the “undisputed surrounding facts” as an implicit credibility finding. And, the State argues, Ashby has not met her heavy burden to overcome the deference that this court affords to the district courts factual findings.
¶76 For her part, Ashby argues that because “the district courts determination that Kevins testimony was ‘irreconcilable’ with ‘undisputed facts’ was made from comparing the cold record of the trial (for which it had not been present) with Kevins statements at the evidentiary hearing, the courts determination that Kevins testimony was ‘irreconcilable’ should be reviewed without any deference from this Court.” Even if we afford deference on review, Ashby contends that “the courts determinations that Kevins testimony is ‘irreconcilable’ are clearly erroneous because the court failed to consider all of the facts and made a determination against the clear weight of the evidence.” (Cleaned up.)
¶77 Generally, when “making factual findings, the trial court is in a unique position to assess the credibility of witnesses and weigh the evidence.” State v. Tripp, 2010 UT 9, ¶ 30, 227 P.3d 1251 (cleaned up). Accordingly, we “defer to the factual findings of the trial court unless the findings are clearly erroneous.” Id. When reviewing for clear error, “[t]he lower courts decision should be respected unless the court failed to consider all of the facts or reached a decision against the clear weight of the evidence.” State ex rel. E.R., 2021 UT 36, ¶ 32, 496 P.3d 58 (cleaned up).
¶78 The district courts determination that Kevins testimony could not be reconciled with the record rested on its determination that certain facts were “undisputed.” Specifically, the court deemed it undisputed that (1) Kevin had admitted to bathing nude with Ashby; (2) Kevin had never seen female nudity prior to his CJC interview; and (3) Kevin was under no pressure to make allegations against anyone when he was interviewed at the CJC. Because there is conflicting evidence in the record on each of these points, the district court clearly erred in treating these facts as “undisputed.”
¶79 First, the court found that Kevins testimony that he only recalled bathing with Ashby while they were wearing swimsuits was irreconcilable with the undisputed evidence that Kevin remembered bathing together naked. The undisputed evidence that the court cited was Kevins statements to Dr. Goldsmith and the Board of Pardons that Ashby washed him while he played with toys, that they washed each others backs, and that he had once asked to suck on Ashbys nipples.
4
The court inferred from those statements that Kevin had admitted to bathing naked with Ashby. Although the court might have drawn a reasonable inference that the two were naked, it was equally plausible that the two were wearing swimsuits when those events occurred. And Kevin was never asked to address those specific instances at the evidentiary hearing. Because competing inferences could be drawn from the evidence, the court clearly erred in finding that it was “undisputed” that Kevin knew that he and Ashby had bathed together naked and that Kevins testimony that he only recalled bathing together in swimsuits was irreconcilable with that fact.
¶80 Second, the court found that it was “undisputed” that Kevin “had not, at least up to the time of the CJC interview, ever seen pornography or female nudity other than [Ashbys].” As evidence, the court cited Kevins inability to explain how he acquired knowledge of female anatomy. But Kevin testified that, although he did not recall anyone showing him pornography as a child, he “could have looked at porn” and could imagine “sneaking and getting it on the internet.” When asked how he could have known as an eight-year-old child that a finger could be inserted into a womans vagina, Kevin responded, “I honestly dont know. I could have looked at porn or—because those things happen in pornography. I honestly am not sure how I became aware of that much female anatomy, but somehow I did. But it wasnt because of my mom.” The fact that Kevin, as an adult, could not specifically recall how he learned about female anatomy as a child does not establish, as an undisputed fact, that he had no exposure to pornography or other sources of such information.
¶81 The court also found that, even though Kevin correctly recounted as a child that his mothers genitals were hairless, Kevin “did not even admit to seeing [Ashbys] nude body.” But Kevin did not testify that he had never seen his mother nude. Although he testified that he only recalled bathing together in swimsuits, he was never asked whether he had observed his mother nude while she was dressing, showering, or engaged in other ordinary activities in the home. Because Kevin never denied seeing his mother nude, the evidence does not support the courts conclusion that his testimony was irreconcilable with his knowledge as a child.
¶82 Third, the court found that Kevins explanation that he lied to protect the neighbor boy was irreconcilable with the undisputed fact that Kevin was under no pressure to implicate anyone when he disclosed the alleged abuse in the CJC interview. Although the record supports the district courts finding that “nothing in the CJC interview applied any pressure on [Kevin] to implicate anyone of anything or to divert attention from the friend,” the court overlooked evidence that Kevin understood the purpose of the interview before it started. (Emphasis added.)
¶83 In his deposition, Kevin explained that he knew Father and Stepmother were upset because he had been acting out sexually with other children, his parents wanted to know why he was behaving that way, and they took him to a therapist because of his sexual behavior. He also knew going into the CJC interview that Father and Stepmothers concerns about his behavior were at least somewhat tied to him bathing with Ashby. And his statements in the interview confirm that he had been taught, prior to the interview, that bathing naked with his mother was “just really bad.” Kevin told Dr. Goldsmith that he knew that Father and Stepmother “believed that his sexual acting out behaviors were a result of inappropriate behaviors that took place while he visited [Ashby] on weekends.” So, Kevin explained, when he was asked where he learned sexual behaviors, he said “my mom” because he did not want to reveal that he had actually learned them from the neighbor boy.
¶84 Even if Kevin “had no idea that he was describing something that could be interpreted as abuse,” there was at least some evidence in the record that Kevin was preconditioned to attribute “sexual play” to Ashby. Given this competing evidence, the court clearly erred in finding that it was undisputed that Kevin was under no pressure to implicate Ashby at the time of the CJC interview.
¶85 Because there is conflicting evidence in the record, we reverse the courts finding that Kevins testimony is “irreconcilable” with the “undisputed evidence.” In reaching this conclusion, we do not mean to suggest that the evidence identified by the district court is not relevant to determining whether Kevins recantation is credible. Perceived discrepancies in a witnesss statements are appropriately considered in assessing a witnesss credibility. But such discrepancies “are not determinative of [a witnesss] credibility.” See State v. Kirby, 2016 UT App 193, ¶ 23, 382 P.3d 644 (addressing inconsistencies between a victims testimony at the preliminary hearing and at trial). And conflicting—even irreconcilable—evidence does not automatically defeat proof by clear and convincing evidence. In Brown v. State, 2013 UT 42, 308 P.3d 486, for example, we upheld a district courts determination of factual innocence by clear and convincing evidence even though there was inconsistent evidence in the record. In doing so, we noted that “the mere existence of contradictory, underlying evidence is of no consequence” because we rely on the district court “to judge credibility and resolve conflicting evidence.” Id. ¶ 63.
¶86 But here the district court did not purport to judge credibility or resolve the conflicts in the evidence. It treated the facts as undisputed and, in so doing, failed to consider all of the evidence. Because we conclude that any implicit credibility determination was based on a clearly erroneous factual finding that the recantation was “irreconcilable” with the “undisputed facts,” we vacate the district courts order and remand for further findings consistent with this opinion.
CONCLUSION
¶87 Kevins recantation, if credible, is sufficient to prove Ashbys factual innocence by clear and convincing evidence. But the district court denied the petition without weighing all the relevant evidence, assessing credibility, or making an ultimate finding on the recantations veracity. And, to the extent the court made an implicit credibility determination, it was based on a clearly erroneous finding that certain facts were undisputed. We therefore remand to the district court to determine whether Kevins recantation is credible.
FOOTNOTES
1
. A pseudonym.
2
. Specifically, Stepmother testified that she had gone to Ashbys house to pick up Kevin when he was between six and eight years old. Stepmother stated, “When I knocked on the door [Ashby] answered it in a towel and her hair was wet. And when I went inside, Kevin came out of the hallway and he was naked. And [Ashby] laughed and said, ‘Sorry, were just getting out of the tub.’ ”
3
. We note that some courts do apply something of a presumption in favor of the original testimony by requiring that a recantation be corroborated with additional newly discovered evidence. See, e.g., State v. McAlister, 911 N.W.2d 77, 87 (Wis. 2018). The State has not suggested that we adopt a similar rule.
4
. The district court did not cite Ashbys testimony at trial that she had bathed with Kevin when he was three or four years old, presumably because Kevin could not be expected to remember events that occurred so early in his childhood.
JUSTICE HAGEN, opinion of the Court: