United States v. Walter Wooden

Reversed and remanded by published opinion. Chief Judge TRAXLER wrote the opinion, in which Judge MOTZ and Judge KEENAN joined. OPINION TRAXLER, Chief Judge: Approximately three months before Walter Wooden was to be released from federal prison, the government sought to commit him as a “sexually dangerous person,” 18 U.S.C.A. § 4248(a) (West Supp. 2012), under the civil-commitment provisions of the Adam Walsh Child Protection and Safety Act of 2006 (the “Act”), Pub.L. No. 109-248, 120 Stat. 587 (codified as amended in scattered sections of 18 and 42 U.S.C.A.). After an evidentiary hearing, the disti'ict court held that the government failed to prove Wooden suffered from pedophilia and failed to prove he would have serious difficulty refraining from re-offending. The court therefore dismissed the government’s petition and ordered Wooden released. The government appeals. For the reasons set forth below, we reverse the district court’s order and remand for reconsideration of the government’s petition on the existing record. I. The Act authorizes the government to civilly commit “sexually dangerous” federal inmates after the expiration of their sentences. 18 U.S.C.A. § 4248(a). An inmate is a “sexually dangerous person” if he has a prior act or attempted act of child molestation or sexually violent conduct and is “sexually dangerous to others.” Id. § 4247(a)(5). An inmate is sexually dangerous to others if he “suffers from a serious mental illness, abnormality, or disorder as a result of which he would have serious difficulty refraining from sexually violent conduct or child molestation if released.” Id. § 4247(a)(6). The Act’s mental illness and serious-difficulty-refraining requirements ensure that commitment is limited to inmates with a volitional impairment — inmates “whose mental illness renders them dangerous beyond their control.” United States v. Francis, 686 F.3d 265, 274-75 (4th Cir.2012); see also United States v. Hall, 664 F.3d 456, 463 (4th Cir.2012) (“[T]he serious difficulty prong of § 4248’s certification proceeding refers to the degree of the person’s volitional impairment, which impacts the person’s ability to re *443 frain from acting upon his deviant sexual interests.” (internal quotation marks omitted)). As the Supreme Court has explained, the requirement that the inmate suffer from a volitional impairment is of “constitutional importance” because it works to “distinguish[ ] a dangerous sexual offender subject to civil commitment from other dangerous persons who are perhaps more properly dealt with exclusively through criminal proceedings. That distinction is necessary lest civil commitment become a mechanism for retribution or general deterrence — functions properly those of criminal law, not civil commitment.” Kansas v. Crane, 534 U.S. 407, 412, 122 S.Ct. 867, 151 L.Ed.2d 856 (2002) (citation & internal quotation marks omitted). II. A. Wooden was born in March 1956 in Washington, D.C. Wooden’s intellectual capacity is somewhat limited. He repeated the third grade and was expelled from school in the eighth grade. See J.A. 112. A 1978 evaluation classified Wooden’s intellectual functioning as within the “dull normal range,” J.A. 75, and his I.Q. has been measured at 70, which qualifies as “borderline retarded,” J.A. 502. When Wooden was eight or nine years old, he was sexually assaulted by a man in his neighborhood. Wooden did not report the abuse. In January 1972 and again in April 1972, Wooden was adjudicated delinquent for committing rectal sodomy on a minor. In October 1973, Wooden again was adjudicated delinquent after sexually molesting a minor. A sentencing report prepared in connection with this offense indicated that Wooden had been arrested for sexual offenses against children at least six times since January 1972. In 1974, Wooden was charged as an adult and pleaded guilty to taking indecent liberties with a four-year-old child. Wooden was sentenced to ten years’ imprisonment, but he was paroled into the community in 1980. In 1984, after separate incidents involving an eight-year-old boy and a twelve-year-old boy, Wooden was convicted of various charges, including sodomy, and was sentenced to 25 years’ imprisonment. Wooden was paroled in 2000, but his parole was revoked in 2001 for failure to cooperate with the supervising officers. On July 25, 2002, Wooden was again paroled, and he was ordered to undergo long-term sex-offender treatment and testing. Dr. Ronald Weiner supervised Wooden’s court-ordered sex offender treatment. Wooden responded well enough to treatment that after about six months, Dr. Weiner recommended reducing the frequency of their sessions from once a week to once every other week. After about a year of treatment, Dr. Weiner recommended monthly sessions. In May 2005, probation officer John Taberski was assigned ■ to Wooden’s case. Taberski called and introduced himself to Wooden and told Wooden that he had a “maintenance” polygraph examination coming up. J.A. 452. Wooden repeatedly told Taberski that he would refuse to take the polygraph. Wooden would not explain his concerns about the polygraph, and Taberski encouraged Wooden to discuss his apprehensions with Dr. Weiner. Taberski reported Wooden’s reluctance to Paul Brennan, Taberski’s supervisor. In conversations with the probation agents, Wooden admitted that he had had contact with children in the community and that children visited Wooden’s apartment while he was alone there. See J.A. 452-54. When Taberski pressed Wooden about whether he was following the plans and using the relapse-avoidance skills developed in his treatment sessions, Wooden *444 became angry, saying, “ ‘What do you expect me to do; there’s nowhere for me to go.’ ” J.A. 455. Taberski called Dr. Weiner immediately after the home visit to report that Wooden had admitted to being around children not only in the community, but also in his apartment. Dr. Weiner met with Wooden on June 3, 2005, for a counseling session. Wooden told Dr. Weiner that he had been “placing himself in high risk situations” by hiring himself out as a babysitter and that he had been having “deviant sexual thoughts about children.” J.A. 67. Wooden eventually told Dr. Weiner that, months earlier, he had had sexual contact with a seven-year-old boy he knew from the neighborhood. Wooden told Dr. Weiner that the boy had followed him to the laundry room in the basement of his apartment building. After Wooden pulled down his pants and showed the boy his penis, the boy also took off his pants, and Wooden placed his penis against the boy’s buttocks without making any attempt to penetrate. Wooden reported that he then became upset with himself and changed his mind about engaging in sexual activity with the boy, who left after Wooden gave him five dollars. On June 7, 2005, Wooden, Dr. Weiner, and probation officers Brennan and Taberski met for an emergency meeting convened by Dr. Weiner. At the meeting, Dr. Weiner explained to the others that Wooden had admitted to molesting the seven-year-old boy, and Wooden agreed. Wooden refused to provide the boy’s full name, but he did provide the boy’s address. At a June 9 counseling session with Lesley Stamm, an associate of Dr. Weiner’s, Wooden said that he had met the boy about a year earlier and that “the boy initiated their encounters because the boy wanted to have sex with him.” J.A. 152. Wooden told Stamm that he fought an “internal battle” over whether he should touch the boy and that he “entertain[ed] fantasies” about sodomizing the boy. J.A. 152. When discussing his interactions with the boy, however, Wooden vacillated between admitting to actual sexual contact with the boy and claiming that he had only dreamed about having sexual contact with the boy. See J.A. 152. In the earlier conversations with Dr. Weiner and the probation officers, Wooden never claimed that he had dreamed the encounter. During a polygraph examination administered on June 20, 2005, Wooden admitted to having “deviant sexual thoughts about children in the past year” and to being sexually aroused in the presence of children in the past year. J.A. 68. Wooden acknowledged that he had engaged in sexual activity with a child in the past year and that he had attempted to engage in sexual activity with a different child. See J.A. 68, 153. According to the officer administering the polygraph, Wooden’s answers to those questions were “non-deceptive.” J.A. 153. In a session with Dr. Weiner the next day, Wooden acknowledged that he was not being completely truthful about his sexual contacts with children. He told Dr. Weiner that, a couple of years earlier, he had lured a boy into the basement for sexual purposes but changed his mind and did not go through with the offense. See J.A. 152, 529-30. Police officers investigating the incident interviewed Wooden, who told them the same story he had first told Dr. Weiner— that he placed his penis against the boy’s buttocks but did not attempt penetration. The officers also interviewed the seven-year-old boy identified by Wooden. The presence of the police scared the boy, who worried that Wooden was “real mad because he told on him.” J.A. 373. The boy would not explain why Wooden might be mad at him, and he denied that Wooden touched or molested him. The boy told the officers, however, that he was afraid to *445 be around Wooden, even though Wooden sometimes gave him money. The boy told the officers that one of his friends thought Wooden was “gay and told [the boy] not to go around him.” J.A. 373. In July 2005, Wooden was charged with violating the terms of his parole by having contact with minors and by having sexual contact with the seven-year-old boy. At his parole revocation hearing, Wooden testified that he had not actually molested the seven-year-old boy, but had only dreamed about the molestation. The hearing officer revoked Wooden’s parole, concluding (under a preponderanee-of-the-evidence standard) that Wooden did in fact have sexual contact with the boy. Wooden served the parole-revocation sentence at the Federal Correctional Institute in Butner, North Carolina. 1 While at Butner, Wooden sent a Christmas card to the seven-year-old boy. Wooden’s message on the card read, “Merry Christmas to you [name redacted], and if I was out there I would do a whole lot for you. Because it is Christmas and you are my friend. So have a very, very happy Christmas, and a happy New Year too.” J.A. 68. The card was signed, ‘Your Friend Walter.” J.A. 68. B. Approximately three months before Wooden’s scheduled release date, the government initiated commitment proceedings by filing a certification that Wooden was a sexually dangerous person within the meaning of the Act. The certification automatically stayed Wooden’s release pending the district court’s final determination, after an evidentiary hearing, of whether commitment was warranted. See 18 U.S.C.A. § 4248(a); Hall, 664 F.3d at 459. The hearing took place over two days in July 2011. Testifying at the hearing were Wooden himself, Wooden’s parole officer, and three expert witnesses — government experts Dr. Hy Malinek and Dr. Heather Ross, clinical psychologists specializing in forensic psychology and risk assessment for sexual offenders, and defense expert Dr. Terence Campbell, a psychologist specializing in forensic psychology. Wooden was one of the first witnesses called by the government; his direct examination began on the first day of the hearing and continued into the second day. On the first day, Wooden was a difficult and recalcitrant witness. There were often long pauses before Wooden answered a question, see J.A. 20, and Wooden frequently claimed he could not remember details of the crimes or tried to invoke his right to remain silent, see J.A. 386 (“I plead the Fifth. I ain’t tryin’ to go back to that.”). Nonetheless, the government was able to elicit some testimony from Wooden about his prior crimes, and that testimony also shed light on Wooden’s thought processes. Wooden testified that the victims *446 of his crimes, who were as young as four, wanted to have sex with him and that they came to him asking for sex. See J.A. 393— 95, 406, 426-27. He testified that his victims asked him for money and that these young children “understood” that they would have to have sex with Wooden before he would give them any money. J.A. 397. When Wooden was asked if, after reflecting on his early crimes, he had any thoughts about how he should have reacted, his response was, “I should have told all of them no.” J.A. 407. The testimony Wooden gave on the first day of the hearing was similar to the testimony he had given two weeks earlier, when he was deposed by the government. For example, Wooden testified at the deposition that “[a] whole lot of boys came up to me and wanted me to have sex with them back then. That’s the God honest truth, and my brothers and sisters — my brothers will tell you that, too.” J.A. 347. Wooden also testified that, at the time he committed his prior offenses, he believed the young boys wanted to have sex with him, and he testified that he still believed that to be true at the time of the deposition. See J.A. 345. And when asked in the deposition about a prior conviction involving a four-year-old boy, Wooden testified that the boy laid down on his stomach on his own, without Wooden touching or guiding him, because “he knew. I guess he knew that’s what I wanted.” J.A. 422. When Wooden retook the stand on the second day of the hearing, his demeanor was markedly improved. He did not claim memory lapses or assert a right to remain silent as he did on the first day, and he was much more responsive to the government’s questions. When the government asked Wooden about the disputed 2005 incident with the seven-year-old boy, Wooden responded: I told him it was a dream. I told Dr. Weiner that I had this dream. I had this dream about [name redacted], and I told Dr. Weiner, I said, I want him to find out was it a dream or not, because sometimes I have blackout spells. I wanted him to find out was it a dream or not because I didn’t want to hurt the boy. So the next thing I know he had me to go see — had me to go see the police investigator, had me to go see the investigator. I talked to the investigator, and the investigator asked me the same questions. I told the investigator the same thing. I said, I didn’t want to hurt the boy. I wanted them to find out was it a dream or not, because I didn’t know. J.A. 437 (emphasis added). Wooden testified that he sent the Christmas card to the boy because the boy was his friend. J.A. 442. The government asked Wooden if he felt he had changed as a person. Wooden responded that he had changed because he no longer thinks about boys and because he realized that he was at fault. See J.A. 446 (“Once before I was thinking that they wanted it, but it was something that I really wanted myself. I wanted that. I wanted to have sex with them kids.”). The government pointed out Wooden’s testimony from the day before that all of his victims wanted to have sex with him and asked Wooden why his testimony had changed. Wooden answered, I had nightmares all night. I couldn’t sleep. I couldn’t sleep. I couldn’t sleep. I kept tossing and turning and tossing and turning and tossing and turning. I couldn’t sleep. It kept on coming up to me. In my mind it kept coming up to me that it was my fault. I couldn’t sleep. I didn’t get no sleep. I couldn’t sleep. J.A. 446. The district court interjected, asking Wooden if he was saying that “because you want to get out” or “because you *447 actually have a conscience now that you didn’t have before?” J.A. 446-47. Wooden explained: I didn’t want to really answer them questions, because every time — every time when I hear about these boys my heart get to pounding. I get scared and it hurts. My heart pounds all the time when I hear about them boys. I was trying not to hear about them no more. I’m trying to do something positive. That’s why I didn’t want to answer his questions because I didn’t want to be up here on the stand crying, because it hurts. J.A. 447-48. After prompting by the district court, Wooden specifically acknowledged that it was wrong for him to have “molested them boys.” J.A. 448. Counsel for Wooden asked him three questions on cross-examination — whether Wooden had “any desire to have sex with young boys now”; whether Wooden believed that a young boy who asked him for money wanted to have sex with him; and whether Wooden believed “grown men should have sex with young boys.” Wooden answered, “No, ma’am,” to each question. J.A. 449-50. The government’s experts, Dr. Malinek and Dr. Ross, both testified that Wooden met the criteria for commitment under the Act. Prior to the hearing, the experts had submitted reports setting out their conclusions and recommendations in detail. Although Wooden refused to be interviewed by the government’s experts, they had access to extensive information about Wooden, including a transcript of the government’s deposition of Wooden taken two weeks before the hearing; court records and other information about his crimes; his prison disciplinary file; and his mental health records, including the treatment records and notes from Dr. Weiner’s sex-offender treatment program. Dr. Malinek testified that Wooden met the diagnostic criteria for pedophilia, which qualifies as a serious mental illness under the Act. When determining whether Wooden would have serious difficulty refraining from further sexual offenses, Malinek scored Wooden under several actuarial risk-assessment models and considered various “dynamic” risk factors that focus on the individual circumstances of the defendant under evaluation. 2 Wooden’s scores on the actuarial models all indicated either a high risk or very high risk of recidivism. And of the dynamic factors considered by Malinek, all but one (Wooden’s age) suggested that Wooden would have difficulty refraining from re-offense. Malinek therefore concluded that Wooden’s mental illnesses would cause him to have serious difficulty refraining from re-offending if released and that Wooden met the criteria for commitment under the Act. *448 Dr. Malinek’s conclusions and recommendations were premised in large part on Malinek’s determination that Wooden had, in fact, molested the seven-year-old boy in 2005. See, e.g., J.A. 103 (“[FJactoring heavily into my decision making in this case is the fact that Mr. Wooden molested a young boy ... while participating in sex offender treatment, and after he has been incarcerated for years due to similar crimes in the past. He admitted to sexual fantasies and arousal around children during 2005. This failure to stop himself speaks to a powerful volitional impairment.”). Dr. Ross’s conclusions were consistent with those of Dr. Malinek. Dr. Ross concluded that Wooden suffered from pedophilia, and her assessment of Wooden under two risk assessment ihodels also showed that he was at high risk of reoffending. Dr. Ross testified that the mental illnesses would make it difficult for Wooden to refrain from reoffending if released and that Wooden therefore qualified as a sexually dangerous person within the meaning of the Act. Like those of Dr. Malinek, Dr. Ross’s recommendations were premised on her view that the 2005 incident actually occurred. See J.A. 119— 20,123. Wooden’s expert, Dr. Terence Campbell, testified that Wooden did not meet the criteria for commitment under the Act. Although Campbell agreed that the pedophilia diagnosis may have been proper in the past, he testified that Wooden no longer satisfied the diagnostic criteria for pedophilia. Dr. Campbell also testified that Wooden was not impulsive, which Campbell viewed as foreclosing any possibility that Wooden had a volitional impairment. See J.A. 500 (“[IJmpulsiveness is a necessary condition for volitional impairment. ... If there’s no impulsiveness, there’s no volitional impairment.”). Because Wooden was not impulsive and thus was not volitionally impaired, Dr. Campbell testified that Wooden would not have serious difficulty refraining from further offenses if released. Dr. Campbell’s determination that Wooden could no longer be considered a pedophile was premised in large part on Wooden’s age, which was 55 at the time of the hearing. As Dr. Campbell testified, research has shown that “the prevalence and incidence of criminal behavior by adults decreases steadily with increasing age.” Robert A. Prentky et al., Sexually Violent Predators in the Courtroom: Science on Trial, 12 Psychol. Pub. Pol’y & L. 357, 375 (2006); see Testimony of Dr. Malinek, J.A. 573 (“[QJuite a few age related studies [have] documented an inverse relationship between age and recidivism.”). Dr. Campbell testified that only Wooden’s behavior, not his thoughts or feelings, was relevant to whether Wooden still suffered from pedophilia. See J.A. 498 (“Are we going to focus on what someone thinks? Are we going to focus on what someone feels? Or are we going to focus on what someone does. And in a legal proceeding, such as this one, diagnostic considerations dictate that we focus on what the Respondent does. What is his overt behavior.”). Campbell noted that there was no evidence of sexual misconduct by Wooden in prison since 2005, that Wooden had not been collecting child pornography or non-explicit pictures of children while in prison, and that, with the “one exception” of the Christmas card, J.A. 498, Wooden had not attempted to correspond with children. In Dr. Campbell’s view, the absence of these problematic and symptomatic behaviors is consistent with the relevant diagnostic data ... that the vast majority of previously convicted sexual offenders do not reoffend. Therefore, whatever psychopathology drove the original offending behavior resolves itself with the passage of time____[SJooner or later practically *449 all offenders stop offending. So age becomes a very important consideration, and Mr. Wooden’s age of 55 indicates that he’s at the point where he is beyond his pedophilia and he is not at a risk to reoffend. J.A. 498-99. Under questioning from the district court, Campbell acknowledged that there was no evidence that Wooden had ever collected or shown any interest in child pornography, but he still found the absence of a pornography cache significant, because “child molesters serving time who have no history of pornography possession out in the street [frequently] acquire their own collection while they’re behind bars because it affords them some kind of gratification for their sexual deviancy.” J.A. 558. After the district court observed that “the empirical evidence is that that’s not him[, b]ecause there are no incidents of him ever viewing or collecting child pornography,” J.A. 559, Campbell responded, “Correct. And he’s not doing it now. If he were doing it now then I would revise my opinion about pedophilia applying to him as a diagnosis now.” J.A. 559. Dr. Campbell testified that he had drawn no conclusion about whether Wooden actually molested the seven-year-old boy in 2005. Noting that there was evidence supporting either conclusion, Dr. Campbell explained that he was “hot a trier of fact” and that it would be “inappropriate for [him] to assume that position.” J.A. 517. Dr. Campbell’s opinion and recommendation to the court nonetheless were premised on the assumption that the incident did not happen; Campbell agreed that civil commitment would be warranted if Wooden had molested the boy in 2005. See J.A. 522. C. The district court in a written opinion rejected the government’s petition. The district court first held that application of the Act to Wooden violated the Due Process and Equal Protection Clauses of the Constitution. And on the merits of the certification question, the court concluded that the government had not proven by clear and convincing evidence that Wooden suffered from a serious mental illness or that he would have serious difficulty refraining from further acts of child molestation if he were released. The court thus ordered Wooden released. The district court agreed that Wooden had suffered from pedophilia “at some point in the past,” J.A. 39, but the court found that the government had not proven that Wooden still suffered from pedophilia at the time of the hearing. Relying on Dr. Campbell’s testimony, the district court explained that “under the proper circumstances,” pedophilia can “abate with time,” J.A. 39, and the district court ultimately concluded that pedophilia was “no longer a controlling and debilitating part of Mr. Wooden’s psychological makeup.” J.A. 39. Critical to this conclusion was the court’s determination that Wooden did not molest the seven-year-old boy in 2005. The court acknowledged that Wooden had initially admitted that he molested the boy, but the court gave “little weight” to that admission given Wooden’s “obvious and undisputed intellectual deficits.” J.A. 40. The court noted that the alleged victim of the 2005 incident, after being specifically and directly asked about the incident, denied that Wooden had touched him, which the court found significant in light of Dr. Campbell’s testimony that “studies of children’s response patterns to investigatory questions indicate that if they are directly asked, they do not deny, but tell.” J.A. 40 (internal quotation marks and alteration omitted). In the district court’s view, the government simply failed to present “sufficient credible evidence to sustain a finding *450 that the 2005 alleged assault occurred.” J.A. 41. After concluding that the 2005 incident did not happen, the district court found the evidence of current or ongoing pedophilia wanting. The court observed that Wooden had been free in the community for three years without re-offending, that Wooden had been making good progress in his treatment with Dr. Weiner before he reported the 2005 incident, and that Wooden cooperated with the investigation of the 2005 incident, and the court believed that these facts supported a finding that Wooden’s pedophilia had “subsided with time.” J.A. 42. The court took note of Dr. Campbell’s testimony that Wooden’s failure to collect child pornography in prison indicated that Wooden’s pedophilia had abated and of Dr. Malinek’s testimony “that some molesters are simply not interested in child pornography.” J.A. 42. Although the court found Malinek’s point to be “well taken,” the court explained that after considering that [Wooden’s] last act of child molestation occurred in 1983, that he was free in the community for three years without re-offending, that during the last six years while incarcerated there have not been any infractions involving any type of sexual deviancy, this Court finds that the Government has not shown by clear and convincing evidence that Mr. Wooden continues to suffer from pedophilia. J.A. 42. Although the district court could have rested its rejection of the government’s petition solely on its determination that Wooden was not a pedophile, see Francis, 686 F.3d at 275-76, the court also addressed the Act’s other requirements. In the district court’s view, proof of a volitional impairment that would make it seriously difficult for the inmate to refrain from reoffending, as required by § 4247(a)(6), is not enough for commitment under the Act. Instead, the court concluded that “[a] finding of dangerousness is also constitutionally required,” which the court viewed as requiring proof that Wooden “poses a risk of re-offense that is significant enough to justify a finding that [Wooden] is sexually dangerous and therefore can be preventively detained.” J.A. 18. Noting the conflicting expert testimony and choosing to credit Dr. Campbell’s testimony that Wooden had no volitional impairment because he was not impulsive, the district court held that the government had not proven that Wooden “currently manifests a serious mental illness, abnormality or disorder that impairs his volitional control such that he would have serious difficulty refraining from sexually violent conduct or child molestation if released.” J.A. 50. As to “dangerousness,” the district court stated that because none of the actuarial risk-assessments models showed a five-year recidivism rate of 50% or more, “the actuarial instrument scores alone cannot possibly satisfy the statutory threshold of clear and convincing evidence that Mr. Wooden would have serious difficulty refraining from engaging in sexually violent conduct or child molestation.” J.A. 52. The district court therefore concluded that Wooden was not a sexually dangerous person as defined by the Act, and the court dismissed the government’s petition and ordered Wooden released. This appeal followed. III. The government first challenges the district court’s constitutional rulings, arguing that application of the Act to Wooden is consistent with the Equal Protection and Due Process Clauses of the Constitution. These constitutional issues need not detain us long. The district court’s ruling on the constitutional issues rested entirely on *451 its analysis of those issues in an earlier civil-commitment case, United States v. Timms, 799 F.Supp.2d 582 (E.D.N.C. 2011). This court, however, has since rejected the district court’s constitutional analysis and reversed its decision in Timms. See United States v. Timms, 664 F.3d 436, 456 (4th Cir.2012), petition for cert. filed (U.S. June 4, 2012) (No. 11-10654). Because the district court’s constitutional analysis is foreclosed by our opinion in Timms, we reverse the district court’s determination that application of the Act to Wooden violated the Due Process and Equal Protection Clauses of the United States Constitution. IV. We turn now to the government’s challenges to the district court’s determination that Wooden did not qualify as a “sexually dangerous person” under the Act. The government first contends that the district court erred by concluding that Wooden no longer suffered from pedophilia. 3 The government also contends that the district court improperly interpreted the Act’s serious-difficulty element as requiring proof that Wooden was more likely than not to reoffend in the future if released. “[W]e review the district court’s factual findings for clear error and its legal conclusions de novo.” Hall, 664 F.3d at 462. A court reviewing for clear error may not “reverse a lower court’s finding of fact simply because [it] would have decided the case differently. Rather, a reviewing court must ask whether, on the entire evidence, it is left with the definite and firm conviction that a mistake has been committed.” Easley v. Cromartie, 532 U.S. 234, 242, 121 S.Ct. 1452, 149 L.Ed.2d 430 (2001) (citation and internal quotation marks omitted). “If the district court’s account of the evidence is plausible in light of the record viewed in its entirety, the court of appeals may not reverse it even though convinced that had it been sitting as the trier of fact, it would have weighed the evidence differently.” Anderson v. Bessemer City, 470 U.S. 564, 573-74, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985). A. Persuaded by the testimony of Dr. Campbell, the district court held that although Wooden may have suffered from pedophilia in the past, the government had not proven by clear and convincing evidence that Wooden, at the time of the hearing, continued to suffer from pedophilia. The government contends that the record is replete with evidence showing that Wooden still suffered from pedophilia, and the government argues that the district court committed clear error by ignoring the substantial amount of contradictory evidence. See, e.g., Francis, 686 F.3d at 273 (“A court commits clear error when it makes findings without properly taking into account substantial evidence to the contrary.” (internal quotation marks omitted)). Wooden, however, notes that expert opinion is critical when a court is called upon to determine whether an individual suffers from a mental illness. See Addington v. Texas, 441 U.S. 418, 429, 99 S.Ct. 1804, 60 L.Ed.2d 323 (1979) (‘Whether the individual is mentally ill and dangerous to either himself or others ... turns on the meaning of the facts which must be interpreted by expert psychiatrists and psychologists.”). Because there was conflicting expert testimony *452 about whether Wooden continued to suffer from pedophilia, Wooden argues that the district court’s decision to credit Dr. Campbell’s testimony over that of the government’s experts cannot be clearly erroneous. See, e.g., Anderson, 470 U.S. at 575, 105 S.Ct. 1504 (“[W]hen a trial judge’s finding is based on his decision to credit the testimony of one of two or more witnesses, each of whom has told a coherent and facially plausible story that is not contradicted by extrinsic evidence, that finding, if not internally inconsistent, can virtually never be clear error.”). As Wooden notes, a court reviewing for clear error “should be especially reluctant to set aside a finding based on the trial court’s evaluation of conflicting expert testimony.” Hall, 664 F.3d at 463 (internal quotation marks omitted). Nonetheless, while clear-error review is “deferential, it is not toothless.” In re Agnew, 144 F.3d 1013, 1014 (7th Cir.1998) (per curiam); accord Jiminez v. Mary Washington Coll., 57 F.3d 369, 379 (4th Cir. 1995) (district court’s factual findings are not “so sacrosanct as to evade review”). And as we will explain, our careful review of the evidence has left us “with the defi nite and firm conviction that a mistake has been committed.” United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 92 L.Ed. 746 (1948). (1) Pedophilia is a serious disorder characterized by “recurrent, intense sexually arousing fantasies, sexual urges, or behaviors involving sexual activity with a prepubescent child or children.” American Psychiatric Association, Diagnostic & Statistical Manual of Mental Disorders § 302.2, at 572 (4th ed., Text Revision 2000) (“DSM”). This definition, which “reflect[s] a consensus of current formulations of evolving knowledge in [the] field,” id. at xxxvii, makes it clear that pedophilia is characterized not only by child-centered sexual behavior, but also by child-centered sexual fantasies and urges. 4 As the government argues, the record contains substantial evidence showing that Wooden was still having intense and recurrent sexually arousing fantasies and sexual urges about prepubescent children, including Wooden’s admission to Dr. Weiner that he was having sexual thoughts about being around children, his admissions in the 2005 polygraph examination that he had been having deviant sexual thoughts about children and had been sexually aroused while in the presence of children, and his acknowledgement during his deposition that he had been having sexual thoughts about children. Moreover, Wooden’s testimony at his deposition and at the hearing was full of cognitive distortions 5 or “thinking errors” *453 common to sex offenders. For example, Wooden testified that the victims of his crimes wanted to have sex with him and that children who ask adults for money are really asking to have sex and will retaliate if rejected. See J.A. 406-07 (“It’s just like this. If a little kid tryin’ to get some money out of you and want to have sex with and you don’t have sex with him, [he] can go to anybody and say you did something. ...”). Wooden’s testimony about the 2005 incident likewise provides compelling evidence of the existence and power of his sexual fantasies and urges. Wooden testified that the incident was only a dream, but he also testified that he told Dr. Weiner about it because he was not completely certain it really was a dream: “[S]ometimes I have blackout spells. I wanted him to find out was it a dream or not because I didn’t want to hurt the boy.... I wanted them to find out was it a dream or not, because I didn’t know.” J.A. 437 (emphasis added). Assuming that Wooden was being truthful when he claimed that the incident was only a dream, 6 this testimony reveals that Wooden’s fantasies about sexually assaulting a seven-year-old boy were so strong that he was unable to distinguish his dreams from his actions. The possibility that he might have hurt the boy was so real and so troubling to Wooden that he voluntarily told Dr. Weiner about the dream, despite the consequences he might face. The district court, however, did not account for this evidence when considering whether Wooden was a pedophile. In fact, very little of this evidence is even mentioned in the district court’s order. The court did not mention Wooden’s admission that he was having sexually deviant thoughts about children or his admission of an attempted assault on another child. Moreover, while the district court when recounting Wooden’s testimony generally described the difficult and often non-responsive nature of Wooden’s testimony on the first day of the hearing, the court did not mention the substance of Wooden’s testimony that day. The court described Wooden’s second-day testimony that he was no longer interested in having sex with young boys and no longer believed that boys asking for money are really asking for sex. See J.A. 21-22. The court, however, did not acknowledge in any way that Wooden’s second-day testimony was completely contrary to his first-day testimony and his deposition testimony, despite *454 the court’s observations on the first day of the hearing that Wooden was “trying to reconcile profound guilt and sense of confrontation about his own behavior with doing something that detracts from his position,” J.A. 431, and that Wooden was “not telling the truth and ... not being candid” because he was “trying to assess the danger and the risk involved in being forthcoming,” J.A. 432. Although the district court might not have been required to accept that the evidence recounted above proved Wooden’s ongoing pedophilia, the court was required to at least consider the evidence, and account for it, when concluding otherwise. See Miller v. Mercy Hosp., Inc., 720 F.2d 356, 361 (4th Cir.1983) (“[T]he conviction of mistake may properly be based upon a conclusion that, without regard to what the ‘actual’ facts may be, the findings under review ... were made without properly taking into account substantial evidence to the contrary.”); accord Doe v. Menefee, 391 F.3d 147, 164 (2d Cir.2004) (Sotomayor, J.) (“We have found a district court’s factual findings to be clearly erroneous where the court has failed to synthesize the evidence in a manner that accounts for conflicting evidence or the gaps in a party’s evidentiary presentation.”). The need to acknowledge and account for this contradictory evidence is particularly acute given the district court’s observation (during a break from the government’s questioning of Wooden) that “[i]t is well established, prong one, that he is a pedophile.” J.A. 428. We do not suggest that the district court was somehow bound by this statement, but the statement does indicate that the court at least initially viewed the evidence as pointing toward a finding of pedophilia. Under these circumstances, the district court’s failure to acknowledge its initial views or explain why it disregarded the extensive evidence of Wooden’s continuing pedophilic fantasies and urges casts real doubt on the propriety of the district court’s determination Wooden no longer suffered from pedophilia. See Taylor v. Maddox, 366 F.3d 992, 1007-08 (9th Cir.2004) (“The process of explaining and reconciling seemingly inconsistent parts of the record lays bare the judicial thinking process, enabling a reviewing court to judge the rationality of the fact-finder’s reasoning.... [Fjailure to take into account and reconcile key parts of the record casts doubt on the process by which the finding was reached, and hence on the correctness of the finding.”). (2) As Wooden notes, the district court repeatedly explained in its opinion that it found Dr. Campbell’s testimony more credible than that of the government’s experts, and the court explicitly relied on Dr. Campbell’s opinion when concluding that Wooden was not a pedophile. Contrary to Wooden’s argument, however, that does not make the district court’s factual findings unreviewable. As the Supreme Court has explained, the trial judge may [not] insulate his findings from review by denominating them credibility determinations, for factors other than demeanor and inflection go into the decision whether or not to believe a witness. Documents or objective evidence may contradict the witness’ story; or the story itself may be so internally inconsistent or implausible on its face that a reasonable factfinder would not credit it. Where such factors are present, the court of appeals may well find clear error even in a finding purportedly based on a credibility determination. Anderson, 470 U.S. at 575, 105 S.Ct. 1504 (emphasis added). In our view, Dr. Campbell’s testimony was internally inconsistent and was otherwise deficient or problematic in so many respects that his opinion pro *455 vides no safe harbor for the district court’s factual findings. Dr. Campbell testified that the proper way to determine whether Wooden “at this point in time, here and now,” qualified for a pedophilia diagnosis was to “focus on what [Wooden] does. What is his overt behavior.” J.A. 498. Attempting to correspond with children was one of the overt behaviors that Campbell believed would evidence ongoing pedophilia, but Campbell attached no significance to the Christmas card Wooden sent to the boy involved in the 2005 incident and did not explain why this evidence of relevant overt behavior was not relevant after all. Dr. Campbell also ignored the evidence that Wooden, while on parole, was engaging in high-risk behavior by babysitting children. An important part of Wooden’s treatment under Dr. Weiner was learning “avoidance and escape techniques designed to help [Wooden] get away from problematic situations and high risk situations.” J.A. 317. The fact that Wooden, while in treatment, not only was spending unsupervised time around children but also was hiring himself out as a babysitter is highly relevant to the question of whether Wooden was a pedophile. Yet Dr. Campbell, who believed that behavior was all that mattered, ignored this evidence of highly relevant behavior. Dr. Campbell likewise failed to address the significance of another powerful piece of evidence of ongoing pedophilia — Wooden’s admission that, in addition to molesting the seven-year-old boy in 2005, he had previously attempted to engage in sexual activity with a different child. Given Campbell’s acknowledgement that commitment would be warranted if Wooden had molested the seven-year-old boy, the evidence of an additional attempted assault would seem to be highly relevant to Campbell’s opinion. Dr. Campbell, however, never explained why he disregarded it. Moreover, Dr. Campbell placed great weight on the fact that Wooden was not collecting child pornography, even though there was' no evidence that Wooden had ever been interested in child pornography and even after acknowledging that some “hands-on” sex offenders “never look at pornography because they get no gratification from [it].” J.A. 558. Dr. Campbell thus concluded that Wooden was no longer a pedophile by ignoring evidence of very troubling affirmative behavior by Wooden while at the same time emphasizing Wooden’s failure to engage in behavior he had never engaged in. Because Campbell insisted that behavior is all that matters, but then ignored, without explanation, all evidence of problematic behavior, his opinion was internally inconsistent, if not entirely implausible. Under these circumstances, the district court’s explicit crediting of Campbell’s testimony does not shield the court’s factual findings from our review. See Anderson, 470 U.S. at 575, 105 S.Ct. 1504. (3) The district court determined that Wooden was not a pedophile by relying exclusively on the opinion of Dr. Campbell. As discussed above, however, Dr. Campbell largely ignored all contradictory evidence, and his analysis was internally inconsistent, and the district court’s analysis of the issue suffers from the same deficiencies. The record in this case contains substantial evidence showing that Wooden’s pedophilia had not abated with age and that Wooden at the time of the hearing was still afflicted with and engaging in the child-focused sexually arousing fantasies, sexual urges, and behaviors that are characteristic of pedophilia. The district court did not account for or otherwise explain why it disregarded all evidence of Wooden’s *456 thoughts and thought-processes, nor did the court account for or otherwise explain why it disregarded the evidence of Wooden’s problematic conduct. Dr. Campbell, the expert whose opinion the district court found credible, testified that the evidence of Wooden’s continuing thoughts about offending against children and his abandoned attempt to carry out those thoughts would qualify Wooden as a pedophile under the DSM. See J.A. 556. The district court did not acknowledge this portion of Campbell’s testimony, nor did it explain why it rejected the consensus view of pedophilia reflected in the DSM. We fully understand that, as a reviewing court, we may not reverse the district court’s factual findings “even though convinced that had [we] been sitting as the trier of fact, [we] would have weighed the evidence differently,” as long as the court’s “account of the evidence is plausible in light of the record viewed in its entirety.” Anderson, 470 U.S. at 574, 105 S.Ct. 1504. Nonetheless, we have painstakingly reviewed the entire record, and the district court’s account of the evidence in this regard simply is not plausible. We are “definitely] and firm[ly]” convinced that the district court made a mistake by concluding that Wooden was not a pedophile. United States Gypsum Co., 333 U.S. at 395, 68 S.Ct. 525; see Mercy Hosp., 720 F.2d at 361 n. 5 (explaining that a factual finding is clearly erroneous if it is “against the great preponderance of the evidence” (internal quotation marks omitted)). Accordingly, we reverse as clearly erroneous the district court’s determination that Wooden did not suffer from pedophilia and thus did not have a serious mental illness as required for commitment under the Act. B. In an alternate holding, the district court explained that even if it could conclude that Wooden still suffered from pedophilia, the government nonetheless had not proven that Wooden would have serious difficulty refraining from re-offense if released. In reaching this conclusion, the district court discounted the testimony of the government’s experts because their opinions were based in part on their conclusions that the 2005 incident in fact occurred, a factual premise the district court had already rejected. The district court instead credited Dr. Campbell’s testimony on the issue, finding his analysis more persuasive because Dr. Campbell analyzed Mr. Wooden’s volitional capacity in an individualized and tailored manner, addressing the statutory issue head-on with the Barratt Impulsiveness scale. But Dr. Malinek, on the other hand, applied a more wooden historical analysis of Mr. Wooden’s volition, an analysis based largely on decades-old criminal convictions and an alleged 2005 sexual offense that the Court finds as a matter of fact did not occur. J.A. 49-50. The district court thus accepted Campbell’s view that Wooden was not volitionally impaired because he was not impulsive, and the court concluded that the government had not proven that Wooden “currently manifests a serious mental illness, abnormality or disorder that impairs his volitional control such that he would have serious difficulty refraining from sexually violent conduct or child molestation if released.” J.A. 50. The district court also held that the government failed to prove Wooden’s dangerousness, which the court defined as requiring proof that Wooden “poses a risk of re-offense that is significant enough to justify a finding that [Wooden] is sexually dangerous and therefore can be preventively detained.” J.A. 18. The district court concluded that the government could not make the required showing of dangerousness because none of *457 the risk-assessment models showed a five-year recidivism rate of more than 50%. On appeal, the government argues that the district court’s error in concluding that Wooden does not suffer from pedophilia prevented the court from properly assessing the serious-difficulty question. Moreover, the government contends that the district court’s analysis, like its analysis of the pedophilia question, failed to acknowledge or account for the substantial body of evidence showing that Wooden would have serious difficulty refraining from re-offense if released. The government argues that the court compounded those errors with its demand for proof of dangerousness that could only be satisfied by proof that Wooden was more likely than not to reoffend. The government therefore argues that the district court’s conclusion that Wooden would not have serious difficulty refraining from re-offending if released must be set aside as clearly erroneous. We agree. Although the district court based its conclusions on Dr. Campbell’s testimony, which the court again found to be more credible than that of the government’s experts, the deference generally due such decisions is not appropriate in this case. For reasons mirroring those discussed above, the many deficiencies in Dr. Campbell’s testimony and the district court’s analysis of the issue again leave us firmly and definitely convinced that the district court’s factual findings were mistaken. (1) Dr. Campbell concluded that Wooden was not impulsive based solely on Wooden’s score on the “Barratt Impulsiveness Scale,” a “self-report measure” that used Wooden’s answers to various questions to assess his level of impulsiveness. J.A. 541. And because Campbell found that Wooden was not impulsive, Campbell concluded that Wooden did not have a volitional impairment: So at this point in time if we ask does Mr. Wooden exhibit volitional control or volitional impairment, the answer would be no. Because impulsiveness is a necessary condition for volitional impairment. Remembering the old distinction of oxygen is necessary for human life but in and of itself it’s not sufficient. Impulsiveness is necessary for volitional impairment. If there’s no impulsiveness, there’s no volitional impairment. J.A. 500. As the district court recognized during the hearing, not all child molesters are impulsive or opportunistic; many are careful and deliberate, “grooming” their victims to gain their trust and affection before attempting to make sexual contact. See, e.g., United States v. Brand, 467 F.3d 179, 203 (2d Cir.2006) (explaining that “[c]hild sexual abuse is often effectuated following a period of grooming and the sexualization of the relationship” and that the defendant’s “sharing pictures, flirting, and attempting to gain affection ... constituted classic grooming behavior in preparation for a future sexual encounter” (internal quotation marks omitted)); J.A. 567 (district court asking Dr. Malinek if grooming behavior was “[planned or anticipated behavior, rather than impulsive behavior”). Under Campbell’s view, sex offenders who groom their victims would be categorically excluded from the Act because they are not impulsive and therefore could not possibly have any level of volitional impairment. Dr. Malinek flatly disagreed with this proposition, see J.A. 568-69, and the district court’s questions during the hearing suggested that it was likewise troubled by the premise of Campbell’s opinion, see J.A. 567-68. The district court, however, ultimately accepted Campbell’s view that volitional impairment was dependent on impulsiveness, without *458 explaining how it had resolved its earlier questions or even acknowledging the existence of those earlier questions. After accepting Campbell’s view that impulsiveness was the determinative question, the district court concluded that Wooden was not impulsive by relying on Campbell’s testimony that Wooden’s Barratt score of 48 was “far below the cutoff score of 74 for impulsiveness.” J.A. 499-500. In concluding that Wooden was not impulsive, however, the district court failed to account for the substantial evidence in the record showing Wooden’s impulsiveness. Wooden’s impulsiveness is apparent from the nature of his prior crimes, which were crimes of opportunity where Wooden took his victims to “the first place that he could find,” J.A. 337, without concern for whether he could be identified by his victims, who were all from his neighborhood. Wooden’s “abominable]” adjustment to institutionalization, J.A. 81, likewise provides compelling evidence of his impulsiveness. Wooden committed a “record high number of institutional rule violations,” J.A. 337, and was sanctioned for conduct including insolence towards staff members, work refusal, possession of gambling paraphernalia, destroying government property, stealing, threatening bodily harm, assault, and threatening to kill staff members, to list a very few. See J.A. 81-82. Wooden’s conduct led to serious sanctions, including administrative detention and multiple disciplinary transfers, but the conduct continued throughout his institutional career, including his time at FCI Butner serving the parole-revocation sentence. And as Dr. Malinek and Dr. Ross testified, Wooden’s disciplinary history was strong evidence of his impulsiveness. To the extent that the district court addressed this evidence at all, it was in the course of its rejection of the diagnosis of antisocial personality disorder made by the government’s experts. In that portion of the opinion, the district court suggested that Wooden’s criminal history was a “stale historical factor[ ]” entitled to little weight, J.A. 43, and the court mentioned Wooden’s prison disciplinary history only to note that the frequency of the infractions declined as Wooden aged, see J.A. 44. The nature of Wooden’s prior crimes may well be a historical factor, but it is by no means a stale or irrelevant one. When the question is whether an inmate suffering from pedophilia will have serious difficulty refraining from re-offending if released, consideration of the nature of his prior crimes provides a critical part of the answer. And while the decreasing frequency of Wooden’s prison infractions is also relevant, so too is the nature of those infractions. The district court’s failure to take this evidence into account substantially undermines the court’s factual determination that Wooden was not impulsive. (2) There is, moreover, an even more fundamental problem with the district court’s analysis. The district court ignored extensive evidence that, while perhaps not directly relevant to the question of impulsiveness, was directly relevant to the question actually posed by the statute: whether Wooden would have serious difficulty refraining from re-offense if released. Much of the previously discussed evidence ignored by the district court when concluding that Wooden was not a pedophile was also relevant to the serious-difficulty inquiry. And just as it did when answering the pedophilia question, the district court ignored this evidence when answering the serious-difficulty question. For example, the district court interrupted Wooden’s cross-examination of Dr. Ross to say, “[t]he fact that he’s sending a Christmas card to somebody that he has molested is a pretty strong indication of reoffending; don’t you think?” J.A. 490. The *459 district court in its written opinion, however, did not mention the Christmas card in its serious-difficulty analysis, did not acknowledge its previous view of the significance of the evidence, and did not explain why it no longer believed the evidence was relevant to the question of whether Wooden would have serious difficulty refraining from re-offense. The district court also failed to account for Wooden’s admissions that he was having deviant sexual thoughts about children, was sexually aroused in the presence of children, was placing himself in high-risk situations by hiring himself out as a babysitter, and had attempted to assault a child other than the seven-year-old involved'in the 2005 incident. Whether or not this evidence speaks to impulsiveness, it speaks directly to the serious-difficulty prong, and it should have been considered and accounted for by the district court. The district court likewise failed to acknowledge or consider the significance of Wooden’s dream about the 2005 incident. As previously discussed, Wooden testified that he told Dr. Weiner about the dream because he “want[ed] him to find out was it a dream or not, because sometimes I have blackout spells. I wanted him to find out was it a dream or not because I didn’t want to hurt the boy.” J.A. 437. The facts revealed by this explanation — that Wooden has blackout spells and has dreams about sexually assaulting a seven-year-old boy that are so vivid he cannot distinguish them from reality — are quite obviously relevant to the question of whether Wooden would have serious difficulty refraining from re-offense. The district court, however, simply ignored this evidence. The district court also failed to consider Wooden’s own testimony when determining whether Wooden would have serious difficulty refraining from re-offense. In our view, Wooden’s deposition testimony and his testimony on the first day of the hearing, and the cognitive distortions that testimony revealed, were highly relevant to the serious-difficulty inquiry. Those portions of Wooden’s testimony indicated that when Wooden committed his prior crimes, he believed that very young children were appropriate sexual partners, and that he continued to believe that at least through the first day of the hearing. As Dr. Malinek testified, “[t]he way people think often mediates their conduct. If he thinks that this is appropriate sexual activity or [that a young child is an appropriate] intimacy partner, then that certainly affects his behavior.” J.A. 323 (emphasis added). When considering the serious-difficulty prong, however, the district court declined to consider Wooden’s testimony. Wooden’s performance on the witness stand had led the district court and Dr. Malinek to believe that Wooden’s cognitive limitations were more significant than indicated in Wooden’s record. 7 . Dr. Malinek testified that Wooden’s limited intellectual functioning increased his risk of recidivism, given Wooden’s inability to learn from his experiences and his persistent belief that children are appropriate sex partners. The district court disagreed: *460 Mr. Wooden’s cognitive limitations merely compel this Court to discount Mr. Wooden’s testimony. Having thoroughly observed Mr. Wooden on the witness stand, the Court finds that Mr. Wooden is a poor historian, that his cognition is markedly impaired, and that he has difficulty in understanding and adequately responding to complex questioning. Mr. Wooden’s testimony, as a whole, is not entitled to significant weight. J.A. 48. Although Wooden is clearly cognitively impaired, that impairment cannot justify the district court’s decision to simply disregard Wooden’s testimony. Determining whether an inmate will have serious difficulty refraining from re-offending requires the court to “evaluate! ] the individual’s present mental condition and the likely prospective effect of that mental condition on his volitional control.” Francis, 686 F.3d at 275. This forward-looking inquiry, which attempts to predict the inmate’s “ability to refrain from acting in accord with his deviant sexual interests,” id.,' requires consideration of the grip strength of the mental illness on the inmate — the extent to which the inmate is controlled by the illness. Whether or not Wooden was a poor historian or was confused by complex questions, his testimony provided powerful evidence of his then-current thought processes, which provided critical insight into the degree to which Wooden would be able to control his deviant sexual interests should he be released. Determining the credibility of witnesses and the weight to be accorded their testimony, of course, is a matter for the district court as factfinder. And while Wooden’s cognitive limitations would provide a rational basis for the court to discount certain portions of Wooden’s testimony — for example, his claims that he enrolled in Howard University when he was 15 and went to business school when he was released from prison, J.A. 379, or his testimony that he learned about anal sex at a very young age by watching his brothers and sisters, J.A. 398, the district court went much farther here. The district court refused to consider Wooden’s testimony, which provided important evidence about the extent of Wooden’s cognitive impairments, and also refused to consider how Wooden’s cognitive impairments would affect Wooden’s ability to refrain from re-offending if released. The district court thus did not simply assign little weight to Wooden’s testimony, it assigned little weight to the fact of Wooden’s cognitive impairment. Because the district court did not consider this critical evidence or the other evidence showing the intensity and persistence of Wooden’s child-focused sexual fantasies, thoughts, and urges, the court’s account of the evidence is not “plausible in light of the record viewed in its entirety,” Anderson, 470 U.S. at 574, 105 S.Ct. 1504, and the court’s factual findings are not entitled to the deference typically required under dear-error review, see Menefee, 391 F.3d at 164 (explaining that a factual finding is clearly erroneous “where the court failed to weigh all of the relevant evidence before making its factual findings”); Taylor, 366 F.3d at 1007 (“Fact-finding is ... a dynamic, holistic process that presupposes for its legitimacy that the trier of fact will take into account the entire record before it.”). (3) In addition to these clear errors of omission, the district court erred by insisting that the government prove Wooden’s “dangerousness,” 8 which the court believed required proof of a greater-than-50% risk that Wooden would re-offend *461 within five years. See J.A. 52 (rejecting government’s evidence of Wooden’s actuarial risk of recidivism because “[n]one of the recidivism rates within the five-year window reaches or exceeds 50%.”). The district court’s insistence on proof of a greater than 50% risk of recidivism finds no support in the language of the Act. The Act requires the government to prove that the inmate will have serious difficulty refraining from re-offense, see 18 U.S.C.A. § 4247(a)(6), but it “does not ask the finder of fact to determine exactly how likely [the inmate] is to reoffend.” United States v. Hunt, 643 F.Supp.2d 161, 180 (D.Mass.2009). Recidivism rates are circumstantially relevant to the serious difficulty inquiry because offenders who continually expose themselves to punishment may be presumed to have the most difficulty refraining from sexual reoffending. But the ultimate question called for by the Act concerns the self-control of an individual, not the statistical rearrest patterns of a given population. Id. The court’s greater-than-50% requirement is likewise inconsistent with the Supreme Court’s refusal to give “lack of control a particularly narrow or technical meaning,” Crane, 534 U.S. at 413, 122 S.Ct. 867 (internal quotation marks omitted), and with the Court’s recognition that “in cases where lack of control is at issue, ‘inability to control behavior’ will not be demonstrable with mathematical precision,” id. (4) After a careful review of the entire record, we again are “left with the definite and firm conviction that a mistake has been committed.” Easley, 532 U.S. at 242, 121 S.Ct. 1452 (internal quotation marks omitted). The district court reached its conclusion that Wooden would not have serious difficulty refraining from re-offense by relying on a flawed expert opinion, by ignoring or otherwise failing to account for the substantial body of contradictory evidence, 9 and by disregarding per *462 haps the most relevant evidence on this issue — Wooden’s own testimony. See Mercy Hosp., 720 F.2d at 361 (“[T]he conviction of mistake may properly be based upon a conclusion that ... the findings under review ... were made without properly taking into account substantial evidence to the contrary.”). Moreover, it appears to us that the weight of the evidence in the record indicates that Wooden’s pedophilia would cause him to have serious difficulty refraining from re-offense if released. See United States v. Martinez-Melgar, 591 F.3d 733, 738 (4th Cir.2010) (“[C]lear error occurs when a district court’s factual findings are against the clear weight of the evidence considered as a whole.” (internal quotation marks omitted)); Mercy Hosp., 720 F.2d at 361 n. 5 (explaining that a district court’s factual finding is clearly erroneous if “the finding is against the great preponderance of the evidence” (internal quotation marks omitted)). The evidence established Wooden’s long history of acting on his pedophilic urges. He has been convicted (or adjudicated delinquent) five times, and by his own admission he offended many more times than he was caught. Each re-offense occurred within a relatively short time after Wooden returned to the community, which demonstrates that the threat of detection and incarceration have limited deterrent effect on Wooden. The evidence likewise establishes Wooden’s resistance to treatment. While he was undergoing intensive sex-offender treatment with Dr. Weiner, Wooden engaged in high-risk behavior by being alone with children, giving children money, and even hiring himself out as a babysitter. He continued to have deviant sexual thoughts about children, including the seven-year-boy at the center of the 2005 incident, and he admitted to Dr. Weiner that he came close to acting on these thoughts but changed his mind after luring a different boy into the basement. Wooden’s cognitive distortions and thinking errors about the appropriateness of children as sexual partners continued through the time of the hearing, as evidenced by his testimony that all of his victims initiated the contact and wanted to have sex with him. Moreover, the actuarial risk-assessment models all indicate that Wooden is at high risk of re-offending. Of the relevant dynamic factors, only one— Wooden’s age — indicates that Wooden’s risk might be lower. In this case, however, the generally observed inverse relationship between age and recidivism does little to overcome the evidence of Wooden’s continuing struggle with the child-focused fantasies, urges, and behaviors characteristic of pedophilia. Accordingly, after careful consideration of the record as a whole, we are constrained to reverse as clearly erroneous the district court’s determination that Wooden would not have serious difficulty refraining from re-offense if released. See Martinez-Melgar, 591 F.3d at 738; Mercy Hosp., 720 F.2d at 361 n. 5. V. To summarize, we hold that the district court erred in its conclusion that the application of the Act to Wooden violated the *463 Due Process and Equal Protection Clauses of the United States Constitution. We also conclude that the record does not support the district court’s determination that Wooden does not “suffer[] from a serious mental illness, abnormality, or disorder” because he no longer suffers from pedophilia, 18 U.S.C.A. § 4247(a)(6), nor does the record support the district court’s determination that Wooden would not have “serious difficulty refraining from sexually violent conduct or child molestation if released,” id., and we hereby reverse those factual findings as clearly erroneous. Accordingly, we reverse the district court’s judgment dismissing the government’s petition seeking to commit Wooden, and we remand the matter to the district court for reconsideration. On remand, the district court shall reconsider, on the basis of the existing record and in light of the questions about the district court’s original analysis and the concerns about the existing evidence raised in this opinion, whether Wooden is a sexually dangerous person within the meaning of the Act. REVERSED AND REMANDED Footnotes [1] The Act authorizes the government to seek commitment of inmates “in the custody of the Bureau of Prisons.” 18 U.S.C.A. § 4248(a). Although Wooden’s crimes all involved violations of the laws of the District of Columbia, defendants convicted in the District are committed to the custody of and serve their sentences at the place designated by the Attorney General of the United States, see D.C.Code § 24-201.26, and are "subject to any law or regulation applicable to persons committed for violations of laws of the United States consistent with the sentence imposed,” id. § 24-101 (a). The district court concluded that, by virtue of these statutes, Wooden was in the legal custody of the Bureau of Prisons and thus subject to commitment under the Act. See United States v. Joshua, 607 F.3d 379, 388 (4th Cir.2010) ("[U]nder § 4248 the word ‘custody’ refers not to physical custody or some qualified derivative but rather to legal custody. The statutory language 'in the custody of the Bureau of Prisons' therefore requires the BOP to have ultimate legal authority over the person's detention.”). Wooden does not challenge that ruling on appeal. [2] The actuarial models consider risk factors that have been shown to be predictive of recidivism. Sex offenders are scored under the model based on the presence or absence of the risk factors in that offender’s crimes, and the offender’s risk of recidivism is determined by reference to the known recidivism rates of released sex-offenders who received the same score under the model. See United States v. Hunt, 643 F.Supp.2d 161, 170-71 (D.Mass.2009). The actuarial models, however, "only gauge a risk of recidivism based upon the statistics of the particular group of sex offenders selected for comparison.” United States v. Hall, 664 F.3d 456, 464 (4th Cir.2012). "[Knowing the recidivism rate of a particular group does not mean that the individual under consideration poses the same chance of recidivism in the same time frame; his risk could be higher or lower than that of the group based upon the unique circumstances of his case.” Id. Accordingly, experts using these risk-assessment models also consider dynamic factors such as "the age of the particular offender, his participation in treatment, his compliance with such treatment, his history of reoffending after treatment, and his commitment to controlling his deviant behavior.” Id. [3] The government's experts concluded that Wooden also suffered from antisocial personality disorder. The district court held that the government had not proven that Wooden still suffered from the disorder or that the disorder was a "serious” mental illness as required by the Act. The government does not challenge those rulings on appeal. [4] The DSM is widely recognized as "the authoritative reference used in diagnosing mental disorders.” Young v. Murphy, 615 F.3d 59, 61 n. 1 (1st Cir.2010); see also J.A. 321 (testimony of Dr. Malinek describing the DSM as "[t]he universally accepted manual for diagnosing mental disorders” and the "Bible of diagnosis”). Although the DSM’s description of pedophilia is not controlling, it is persuasive. See Kansas v. Crane, 534 U.S. 407, 413, 122 S.Ct. 867, 151 L.Ed.2d 856 (2002) ("[T]he science of psychiatry ... informs but does not control ultimate legal determinations....”); see id. at 411, 414, 122 S.Ct. 867 (citing the DSM authoritatively); McGee v. Bartow, 593 F.3d 556, 575 (7th Cir.2010) (“Despite its limitations in a non-medical setting, the DSM is a highly influential and useful tool.”). [5] Cognitive distortions allow "sex offenders [to] explain their actions in a way to manage the impressions of others and in a way to make themselves more socially palatable.” United States v. Mitchell, 706 F.Supp.2d 1148, 1217 (D.Utah 2010) (internal quotation marks omitted). Such distortions "are an established, well understood phenomenon among sex offenders.” Id. (internal quotation marks omitted). [6] The district court’s conclusion that the 2005 incident did not happen was based not on a determination that Wooden was telling the truth, but on the court's view that the government had not carried its burden of proving that the incident happened. The district court did not make any explicit factual findings about the truth of Wooden’s statements, but the court's dismissive reference to Wooden's “shadowy admission” about the incident perhaps suggests the court found Wooden’s denial more credible than his admission. From our review of the record, it seems likely that Wooden was not being truthful and that he came up with the dream story in a futile attempt to get himself out of trouble— Dr. Weiner's contemporaneous treatment notes present Wooden's story as a factual admission of actual sexual contact with the boy and do not indicate that Wooden claimed he dreamed the incident, see J.A. 67-68; Wooden’s probation agent testified that during the emergency meeting on June 7, Wooden did not claim to have dreamed the incident, see J.A. 459; and the treatment notes of the session with Weiner’s associate show that even after he first claimed it was a dream, Wooden still vacillated between admitting the incident occurred and claiming it was only a dream, see J.A. 152. Nonetheless, our determination that the district court committed clear error is not dependent on whether Wooden was being truthful or whether the 2005 incident in fact occurred, and we therefore accept, for purposes of resolving this appeal, the district court’s factual conclusion that the 2005 incident did not occur. [7] Although previous testing pegged Wooden’s I.Q. at 70, the district court, drawing from its experience with Social Security cases involving claims of intellectual deficits, J.A. 543, believed Wooden's I.Q. was more likely to be in the 60s and that Wooden "manifests signs of mild retardation.” J.A. 571. In the government’s rebuttal case, Malinek testified that after observing Wooden’s testimony on the first day of the hearing, he thought Wooden “was clearly retarded based on how he presented. ... Today he was much more relevant and focused. He obviously is impaired.” J.A. 574. Malinek thought Wooden’s I.Q. was likely to be in the 70s, J.A. 574, but he agreed that it was likely below the "dull normal” range "identified in the 1978 evaluation.” J.A. 570. [8] As support for its view that "dangerousness” was a separate and necessary element *461 of the government’s case, the district court relied on the Supreme Court's observation in Kansas v. Crane that involuntary commitment statutes generally have been upheld when, among other things, "there is a finding of dangerousness either to one’s self or to others, and ... proof of dangerousness is coupled with the proof of some additional factor, such as a mental illness or mental abnormality.” Crane, 534 U.S. at 409-10, 122 S.Ct. 867 (emphasis added; internal alteration and quotation marks omitted); see Kansas v. Hendricks, 521 U.S. 346, 357-58, 117 S.Ct. 2072, 138 L.Ed.2d 501 (1997). "Dangerousness” in this context is simply shorthand for the danger posed "to the public health and safety” by “people who are unable to control their behavior.” Id. at 409, 122 S.Ct. 867 (internal quotation marks omitted). The Kansas statute at issue in Crane satisfied the dangerousness requirement by limiting commitment to persons suffering from a mental illness that makes the person "likely to engage in the predatory acts of sexual violence.” Id. at 352, 117 S.Ct. 2072 (internal quotation marks omitted). The Adam Walsh Act’s requirement of proof that the inmate "would have serious difficulty in refraining from sexually violent conduct or child molestation if released,” 18 U.S.C.A. § 4247(a)(6), likewise satisfies the dangerousness requirement referred to in Crane. The district court therefore erred by requiring the government to separately prove Wooden’s “dangerousness.” [9] In the instances where the district court did explain why it declined to credit contrary evidence, the reasons given by the court cannot withstand scrutiny. For example, when considering whether Wooden suffered from antisocial personality disorder, the district court indicated that the government’s experts failed to consider the relevance of Wooden’s age to the recidivism question. The government's experts, however, acknowledged the relevance of age and considered Wooden's age in their reports, see J.A. 92-93, 120-22, and Dr. Malinek in his testimony specifically identified Wooden's age as a “protective” risk factor — one that mitigates the risk of recidi *462 vism. See J.A. 340, 573-74. The court also rejected the views of the government experts because their opinions were erroneously based on their views that Wooden actually molested the seven-year-old boy in 2005. Neither Dr. Ross nor Dr. Malinek, however, based their opinions solely on their belief that the 2005 incident occurred; both experts gave multiple reasons to support their conclusion that Wooden still suffered from pedophilia and would have serious difficulty refraining from re-offense if released. Under these circumstances, the district court’s disagreement with one of their factual conclusions cannot justify the court's refusal to consider the balance of their testimony.