United States v. Tymond Preston

751 F.3d 1008, 2014 WL 1876269, 2014 U.S. App. LEXIS 8825

• Court: Court of Appeals for the Ninth Circuit
• Decided: May 12, 2014
• Docket: 11-10511
• Status: Published

Opinions

OPINION

BERZON, Circuit Judge:

Today we consider the voluntariness of a confession given by Tymond Preston, an intellectually disabled eighteen-year-old. To elicit this confession, the police, among other tactics, repeatedly presented Preston with the choice of confessing to a heinous crime or to a less heinous crime; rejected his denials of guilt; instructed him on the responses they would accept; and fed him the details of the crime to which they wanted him to confess. Under the totality of the circumstances, including Preston’s intellectual disability, we conclude that the confession that resulted from this questioning was involuntarily given and should not have been admitted at trial.

I

Background,

A. The Accusation

At the time of the underlying events, Tymond Preston was eighteen years old. Preston has an IQ of sixty-five, which the Supreme Court recognizes as within the range of intellectual disability.¹ See Atkins v. Virginia, 536 U.S. 304, 309 n. 5, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002) (describing an IQ below seventy-five as within the range of mental retardation).² He attended special education classes beginning in elementary school, and did so until he dropped out of high school. Preston’s mother said that a doctor told her that Preston had a “small brain, like a five-year-old.” Psychological evaluations conducted during the course of this litigation show that Preston has “exceptionally limited linguistic ability,” and “significant problems with verbal communication and comprehension.”³ The district court found that he had “deficits in general linguistic and academic skills and low IQ.” Before the offense addressed in this appeal, Preston had been arrested twice for minor juvenile offenses but had never been convicted of any offense or adjudicated delinquent.

Preston lived with his mother and father on the Navajo Nation. An extended family of relatives, including his paternal aunt, resided next door. The neighboring households had a long-standing feud. As recounted by its participants, the feud took some unusual turns. For example, Preston told investigators that the neighbors had used “-witchcraft” to paralyze his father for three months.

Some time on Wednesday, September 23, 2009, Preston’s eight-year-old neighbor — a child in the household with whom Preston’s family was feuding, and the grandchild of Preston’s aunt — entered and later left Preston’s house. That evening, the child reportedly told his grandparents and uncle that Preston “ ‘put his pee-pee in [my] butt,’ ” and that “his butt was hurting.” The grandparents called the police.

When asked about this assertion the next day by Carli Moncher, a forensic interviewer for the Safe Child Center at Flagstaff Medical Center, the child spun a lengthy, fantastical, largely incomprehensible narrative. The tale, in brief, was: Preston had come to his house the day before and threatened to kill him with a knife; the child locked Preston in his bedroom, and Preston escaped by going out the window; the child called 911, climbed on top of a shack, jumped off of it, and ran away; Preston followed the child’s tracks and found him hiding in the bushes; the child ran home and hid while watching a movie with his sister; he and his sister climbed out the window and then down a cliff on a ladder, while Preston jumped over the cliff in a “monster truck”; the child then got into a fight with Preston and kicked him “in the balls,” and Preston “fell out the window”; Preston came to his school in the monster truck, and the police followed, chasing Preston in cars and helicopters and shining a yellow light at Preston and his house; the child and his sister climbed onto the shack and jumped off; Preston jumped over and broke his leg; Preston then came to his house and was “trying to fuck [his sister’s] butt,” at which point the child tried to beat up Preston, hitting him in the head and face; the rape was eventually forestalled when the child’s kittens began scratching Preston; and the child then also “took a lot” of knives and threw them at some robbers, hitting one “right ... in the heart,” and killing him. As the district court observed after hearing this story, “[m]any of these details are obviously not factual.”

As to the alleged sexual contact, the boy stated that Preston pulled the child’s underwear down, “put his penis in my butt,” and touched the boy’s “balls” and “butt” with his mouth. He also stated that Preston told him to “suck his balls,” and when the child did, “white stuff got on my shirt” and “on my lips.” The boy described the shirt he was wearing as red, and said it was at his home in a bag. He further alleged Preston had tried to “cut ... my balls” and that “next he cut his balls.” He also said that Preston had started the attack by dragging him by his shirt and choking him.

On the same day he gave this account to Moncher, the boy was interviewed by a nurse practitioner. The nurse asked him, “Did [Preston] put something on his dick?” and the child responded, “He just put on the dick wearing,” which the parties assume refers to a condom. When asked what happened to the “dick wearing,” the child said that “[i]t got white stuff on it,” and that Preston “threw it away.”

Most of the child’s allegations of sexual abuse were not corroborated by physical evidence. Although swabs were taken from various parts of the child’s body, including his lips, anus, and genitals, the forensic examiner found no evidence of semen on the child. The child had a “normal genital and anal exam,” and no signs of injuries, bruises, or trauma of any kind, although he complained of pain during the anal exam. No red shirt was found. Skin cells were found on the child’s underwear⁴ with DNA from multiple contributors, including at least one male and potentially a female; Preston, the child’s grandmother and at least one more relative could not be excluded as the source of some of the DNA.⁵

B. The Confession

About a week after the boy reported that he had been assaulted, Federal Bureau of Investigation (“FBI”) Special Agent James Kraus and Navajo Nation Criminal Investigator Greg Secatero (“the officers”) went to Preston’s home to question him about the allegations. To aid the investigation, the FBI had obtained Preston’s Navajo Nation certificate of Indian Navajo blood, which revealed that he was eighteen years old. The officers noticed the notation, as they commented to Pres7 ton, “you’re a young guy, 18 years old.”

Kraus and Secatero questioned Preston outside his house, next to Kraus’s vehicle. They began recording their interaction with Preston within “one or two minutes” of approaching him. The interrogation lasted about forty minutes.

The officers quickly became aware of Preston’s mental disability. A short time into the questioning, Preston told them that “I’m not ... all there,” and that “I have problems with my head, like a tumor.” Recognizing that Preston may be impaired, Secatero asked him, “Are you disabled right now?” Preston did not understand the question. “What’s that, disabled?,” he asked. When Secatero explained that “disabled mean[s] you’re not able to take care of yourself or you’re not able to ... get a job,” Preston agreed that he was disabled. He also explained that he had not finished high school.

The agents told Preston that they were investigating a “molestation” that took place “last Friday.”⁶ Preston said that he was not home the prior Friday. All parties now agree that Preston was in fact not home on the Friday in question, as he routinely visits a cousin on Fridays. But the officers rejected as false — over and over again — Preston’s accurate responses to the “Friday” assertions, and repeated throughout most of the interrogation that the incident did take place on Friday. Secatero said, for example, ‘We just don’t buy it,” and “Friday, I know you remember you were here”; and insisted “you have to remember what happened Friday”; Kraus, too, informed Preston that it was “not disputable” that the child had been at Preston’s house on Friday, and stated, incorrectly, that “we have a bunch of people that said you were over here.” Eventually, Preston was told, “the fact is, we know you were here,” and “[tjhere’s other witnesses putting you here, so there’s no denying” it. Preston eventually stopped disagreeing with the assertion that he was home on Friday.⁷

Aside from the “Friday” colloquies with Preston, the officers spent much of the interrogation engaged in an extended dialogue about types of sexual offenders. Secatero told Preston that “there’s two type[s] of people,” and that he was “just trying to figure out which ... one” Preston was. The first type, the officers explained, was “predators, sexual predators,” who “prey on little kids” and are “coldhearted.” This type of person is “the monster that ... everybody’s scared of,” who is “molesting all these little kids,” and to whom “we don’t show any sympathy.” The second type, they went on, was not a monster but just did a “one-time thing,” maybe because he was “curious” or “drunk.” The officers repeatedly told Preston that they “want[ed] to know what kind of person” he was, the “kind of guy that does it all the time or the guy that just” had a “bad day.” Secatero and Kraus minimized the culpability of the second type of person, stating that these individuals could “move on” and be given help, because what transpired was “just a misunderstanding” — but only if these individuals were truthful. Kraus assured Preston that “if nothing happened, that’s cool,” but also “if something just a little bit happened, ... that’s cool, too.” Kraus later acknowledged that the officers meant to convey that admitting guilt could minimize the consequences Preston faced, and that Preston could get some sympathy if he was a one-time offender rather than a serial predator.

When Secatero asked Preston directly, “which person are you? Are you that type where you prey on little kids?,” Preston answered that he was not. Secatero then asked several times if Preston had done “a one-time thing.” Preston gave confused, equivocal responses, including “I don’t know, probably, but I don’t know. But I don’t fuckin’ do that shit,” and “Something probably like that, one time. But I’m not, like, whatever you ... guys are trying to think, that what I’m trying to do, like with everything, but, but fuck, I — I ain’t like that.”

During this time, the officers told Preston that he was “not arrested” and “not in custody,” but also informed him that “[a]fter the interview, you’re free to go,” (emphasis added), indicating that he was free to stop talking to them only when they terminated the interview.⁸ The two interrogators also conveyed, repeatedly, that Preston had to tell them something, or they would keep coming back until he did. For example, Kraus told Preston: “We don’t want to come back here later”; “we’re needing to figure out something because we don’t have to come back here again and again”; “we got to figure out exactly what happened and without you saying anything, that’s — that’s not helping anyone, you know, because then we’re going to have to keep coming back and— and — until we figure exactly what out”; “we don’t want to come back and say, hey, man, you lied to us.” Secatero later said, “you got to tell me what — what happened Friday” and “you have to remember what happened Friday.” (emphasis added). At the hearing on Preston’s motion to suppress, Kraus acknowledged that he meant Preston to understand that the officers would keep coming back until Preston admitted what happened with the child.

Eventually, more than twenty minutes into the questioning, Preston acquiesced in the assertions that he was in fact at home on Friday — even though he was not. So, when the officers said, “Friday---- Six days ago. Do you remember [the child] being over here?” Preston at first responded, “No,” but then switched gears, stating, “[T]here’s a bunch of little kids over here, and plus that other guy, but ... I don’t know what happened that day.” The officers continued, “so Friday, I know you remember you were here ... you said that there were kids over here, and then what? Did you guys go inside or what?” Preston responded, “I was inside just by myself.”

Having gotten Preston to stop protesting that he was not at his house on Friday, Secatero and Kraus proceeded to ask Preston a series of questions that required him to choose between two incriminating alternatives. For example:

“[I]s it because you wanted to have sex? ... Or is it he’s the one that came onto you?”

“[I]s it something ... where you forced the issue or is it something that he wanted?”

“[D]id he pull away or did you pull out?”

“Did you ... put your penis in all the way or just a little bit?”

“Did you do it a lot or just that one time?”

In each case, presented with two incriminating choices, Preston chose the less incriminating one.

The officers also asked a number of leading and suggestive questions that introduced facts Preston himself never mentioned until the officers brought them up. Preston agreed to some of these facts.

Eventually, Preston nodded when asked if he put his penis inside of the child’s “butt,” stated he did this for “five, six seconds,” and that the child then “walked out,” “said, I’m going to tell on you,” and parted crying.” Kraus then wrote out a statement summarizing the admissions the officers had elicited from Preston.⁹

The officers twice misled Preston about the statement’s purpose. First, they promised Preston that they would not “tell this to anybody,” and the statement would never leave the U.S. Attorney’s file. Later, they told him that the statement they would write up was just an apology note to the child, a way to say “sorry” to the child: “Do you want to write any — usually what we do is we write a statement. If, like, you wanted to say sorry or something like that. You could definitely do that. And we can provide that to him.” Preston’s response to that suggestion was equivocal — that he would say “I’m sorry for what I did, but they’re just trying to accuse me of that shit. But fuck, I’m not ... like that.” Ignoring Preston’s equivocation, Kraus told Preston he would “just summarize” their conversation in the apology note. According to Kraus, Preston was never informed that “he might be signing something that could be used later in a court of law.”

The “summary” — Preston’s confession— was a brief gathering of details chosen by the officers, and written out in Kraus’s hand. Many of the details selected were those the officers had fed Preston. For example, Kraus was the first to bring up the possibility that Preston used a condom; Kraus testified that he did so because he understood the child had discussed a condom during the forensic interview. At first, when Kraus asked, “Did you use a condom?” Preston responded “I don’t know.” Only after the officer persisted, saying, ‘Yeah, you did?,” did Preston finally acquiesce, by nodding. Kraus later reinforced the point, “[Y]ou had a condom on, correct?” While Kraus was writing the statement, he reiterated, “you pulled your penis out, and you put a condom on?,” to which Preston nodded assent. The detail that Preston wore a condom was then included in the written statement.¹⁰

Similarly, the officers also fed Preston the detail that he “unzipped” his pants, which was included in the written statement. Initially, when Kraus asked Preston, ‘You just unzipped your zipper?,” Preston responded, “I don’t know.” The officers then affirmatively repeated several times, “you unzipped your pants”; Preston eventually indicated assent to this fact.

As Kraus was writing the summary, he repeated back the information that he was including in question form and asked Preston periodically if it was correct. Preston never revised, corrected, or countered Kraus’s version.

After Kraus finished writing the statement, Kraus informed Preston, “I’m going to have you sign this.” Preston signed.

C. The Proceedings

Preston was charged in federal district court with aggravated sexual abuse of a minor in violation of 18 U.S.C. § 2241(c), for which the mandatory minimum prison sentence is thirty years. He moved to suppress the use of his confession. After a hearing, the district court denied the motion.

Thereafter, Preston and the government agreed that Preston would waive his right to a jury trial if the government would reduce his charge to the lesser offense of abusive sexual contact, which carries no minimum prison sentence, and recommend that he receive no more than a fifteen-year sentence. After a three-day bench trial, the district court found Preston guilty of that charge. Preston was sentenced to fifty months’ imprisonment and a lifetime term of supervised release.

Preston appealed the conviction on a number of grounds, including that the confession was involuntary. A panel of this court held the confession properly admitted; Judge Noonan dissented. United States v. Preston, 706 F.3d 1106 (9th Cir.2013). Preston filed a petition for panel rehearing and rehearing en banc. A majority of the active judges on the court voted to rehear the case en banc. United States v. Preston, 727 F.3d 894 (9th Cir.2013).

II

Legal Principles Regarding Voluntariness of Confessions

The right against compulsory self-incrimination is “the mainstay of our adversary system of criminal justice, and ... one of the great landmarks in man’s struggle to make himself civilized.” Michigan v. Tucker, 417 U.S. 433, 439, 94 S.Ct. 2357, 41 L.Ed.2d 182 (1974) (internal quotation marks and citations omitted). “This principle, branded into the consciousness of our civilization by the memory of the secret inquisitions, sometimes practiced with torture, which were borrowed briefly from the continent during the era of the Star Chamber, was well known to those who established the American governments.” Culombe v. Connecticut, 367 U.S. 568, 581, 81 S.Ct. 1860, 6 L.Ed.2d 1037 (1961). “Its essence is the requirement that the State which proposes to convict and punish an individual produce the evidence against him by the independent labor of its officers, not by the simple, cruel expedient of forcing it from his own lips.” Id. at 581-82, 81 S.Ct. 1860.¹¹

A

In implementing this bedrock constitutional value, our focus is on “whether [the] defendant’s will was overborne by the circumstances surrounding the giving of [the] confession,” an inquiry that “takes into consideration the totality of all the surrounding circumstances — both the characteristics of the accused and the details of the interrogation.” Dickerson v. United States, 530 U.S. 428, 434, 120 S.Ct. 2326, 147 L.Ed.2d 405 (2000) (quoting Schneckloth v. Bustamonte, 412 U.S. 218, 226, 93 S.Ct. 2041, 36 L.Ed.2d 854 (1973)) (internal quotation marks omitted) (emphasis added).

“Each of these factors, in company with all of the surrounding circumstances — the duration and conditions of detention (if the confessor has been detained), the manifest attitude of the police toward him, his physical and mental state, the diverse pressures which sap or sustain his powers of resistance and self-control- — • is relevant.” Culombe, 367 U.S. at 602, 81 S.Ct. 1860; see also Withrow v. Williams, 507 U.S. 680, 693-94, 113 S.Ct. 1745, 123 L.Ed.2d 407 (1993). Thus, the voluntariness inquiry “is not limited to instances in which the claim is that the police conduct was ‘inherently coercive,’ ” Miller v. Fen-ton, 474 U.S. 104, 110, 106 S.Ct. 445, 88 L.Ed.2d 405 (1985) (quoting Ashcraft v. Tennessee, 322 U.S. 143, 154, 64 S.Ct. 921, 88 L.Ed. 1192 (1944)), but “applies equally when the interrogation techniques were improper only because, in the particular circumstances of the case, the confession is unlikely to have been the product of a free and rational will,” id. (citing Mincey v. Arizona, 437 U.S. 385, 401, 98 S.Ct. 2408, 57 L.Ed.2d 290 (1978)). Ultimately, the voluntariness “determination ‘depend[s] upon a weighing of the circumstances of pressure against the power of resistance of the person confessing.’ ” Dickerson, 530 U.S. at 434, 120 S.Ct. 2326 (quoting Stein v. New York, 346 U.S. 156, 185, 73 S.Ct. 1077, 97 L.Ed. 1522 (1953), overruled in part on other grounds by Jackson v. Den-no, 378 U.S. 368, 84 S.Ct. 1774, 12 L.Ed.2d 908 (1964)) (alteration in original).

B

These principles have particular application where, as here, the individual interrogated is of unusually low intelligence. “What would be overpowering to the weak of will or mind might be utterly ineffective against an experienced criminal.” Stein, 346 U.S. at 185, 73 S.Ct. 1077. So, although low intelligence does not categorically make a confession involuntary, it is “relevant ... in establishing a setting” in which police coercion may overcome the will of a suspect. Procunier v. Atchley, 400 U.S. 446, 453-54, 91 S.Ct. 485, 27 L.Ed.2d 524 (1971). The American Bar Association’s Criminal Justice Mental Health Standards summarize this point well: “Official conduct that does not constitute impermissible coercion when employed with nondisabled persons may impair the voluntariness of the statements of persons who are mentally ill or mentally retarded.” ABA Criminal Justice Mental Health Standards, Standard 7-5.8(b), available at http://www.americanbar.org/ pubhcations/criminal_justice_section_ arehive/crimjust_standards_mental healthtoc.html. Similarly, the Seventh Circuit observed that “a finding of involuntariness cannot be predicated solely upon” the defendant’s mental state, but “his mental state is relevant to the extent it made him more susceptible to mentally coercive police tactics.” Smith v. Duckworth, 910 F.2d 1492, 1497 (7th Cir.1990) (internal quotation marks omitted).¹²

Accordingly, we cannot resolve this case by labeling the questioning either inherently coercive or not. Instead, we must evaluate the law enforcement tactics used in conjunction with Preston’s serious intellectual disability.

C

Also consistent with the directive that we must consider “all the surrounding circumstances,” Dickerson, 530 U.S. at 434, 120 S.Ct. 2326, in determining the voluntariness of a confession is that there is “no talismanic definition of ‘voluntariness,’ mechanically applicable to the host of situations where the question has arisen,” Schneckloth, 412 U.S. at 224, 93 S.Ct. 2041. Because there is no “single controlling criterion,” no single factor, such as length of interrogation, can be dispositive. Id. at 226, 93 S.Ct. 2041.

Doody v. Ryan, 649 F.3d 986 (9th Cir.2011), an en banc opinion of this court, provides guidance for carrying out the multivariate inquiry essential to the voluntariness inquiry: We may not “tick[] off the list of circumstances rather than actually considering them in their totality.” Id. at 1011. So it is not enough for courts to “list[ ] the circumstances of [an] interrogation separately on a piece-meal basis.” Id. (internal quotation marks omitted). Courts must “weigh, rather than simply list,” the relevant circumstances, and weigh them not in the abstract but “against the power of resistance of the person confessing.” Id. at 1015-16 (internal quotation marks omitted).

D

One final point as to the nature of the voluntariness inquiry, and one that is easy to elide: In evaluating the voluntariness of a confession under the totality of the circumstances, we are not trying to determine whether the suspect told the truth when he confessed. “[C]onvictions following the admission into evidence of confessions which are involuntary, i.e., the product of coercion, either physical or psychological,” are tenuous not simply “because such confessions are unlikely to be true.” Rogers v. Richmond, 365 U.S. 534, 540-41, 81 S.Ct. 735, 5 L.Ed.2d 760 (1961). “As important as it is that persons who have committed crimes be convicted, there are considerations which transcend the question of guilt or innocence.” Blackburn v. Alabama, 361 U.S. 199, 206, 80 S.Ct. 274, 4 L.Ed.2d 242 (1960). We exclude coerced confessions “because the methods used to extract them offend an underlying principle in the enforcement of our criminal law.” Rogers, 365 U.S. at 540-41, 81 S.Ct. 735. That basic principle is that “[o]urs is the accusatorial as opposed to the inquisitorial system.” Watts v. Indiana, 338 U.S. 49, 54, 69 S.Ct. 1347, 93 L.Ed. 1801 (1949). “To maintain a fair state-individual balance, to require the government to shoulder the entire load, to respect the inviolability of the human personality, our accusatory system of criminal justice demands that the government seeking to punish an individual produce the evidence against him by its own independent labors.” Miranda v. Arizona, 384 U.S. 436, 460, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966) (internal quotation marks and citations omitted). We recognize that “important human values are sacrificed where an agency of the government ... wrings a confession out of an accused against his will.” Blackburn, 361 U.S. at 206-07, 80 S.Ct. 274. The prohibition on the coercion of confessions “also turns on the deep-rooted feeling that the police must obey the law while enforcing the law; that in the end life and liberty can be as much endangered from illegal methods used to convict those thought to be criminals as from the actual criminals themselves.” Spano v. New York, 360 U.S. 315, 320-21, 79 S.Ct. 1202, 3 L.Ed.2d 1265 (1959).

Because of these considerations, the question whether a confession was voluntary is “to be answered with complete disregard of whether or not [the confessor] in fact spoke the truth.” Rogers, 365 U.S. at 544, 81 S.Ct. 735. A “coerced confession is inadmissible under the Due Process Clause even though statements in it may be independently established as true.” Watts, 338 U.S. at 50 n. 2, 69 S.Ct. 1347. Thus, although “coerced confessions are forbidden in part because of their ‘probable unreliability,’” Lego v. Twomey, 404 U.S. 477, 484 n. 12, 92 S.Ct. 619, 30 L.Ed.2d 618 (1972), the voluntariness inquiry focuses not on the truth or falsity of the confession, but on the coercive nature of the interrogation — again, taking into account the particular circumstances of the suspect.¹³

E

The government would have us depart to a considerable degree from the well-established set of principles just outlined, by determining first whether the police’s conduct here was inherently coercive, and, if not, holding the confession voluntary without regard to the likely impact on an individual with Preston’s mental characteristics.

The government is correct that in Derrick v. Peterson, 924 F.2d 813 (9th Cir.1991), a panel of our court stated that the defendant’s individual characteristics “are relevant to our due process inquiry only if we first conclude that the police’s conduct was coercive.” Id. at 818 (emphasis added). To the extent Derrick held that the issue of police coercion during an interrogation must be considered without regard to the suspect’s individual characteristics, it simply cannot be reconciled with the Supreme Court’s totality-of-the-circumstances analysis applicable to the voluntariness inquiry; with the Court’s specific directives, already surveyed, concerning the role of individual characteristics- — -including mental characteristics — in the voluntariness inquiry; or with our fairly recent en banc decision, Doody. Moreover, Derrick, and the government’s argument relying on Derrick, rest on an evident misreading of Colorado v. Connelly, 479 U.S. 157, 107 S.Ct. 515, 93 L.Ed.2d 473 (1986).

For its holding that individual characteristics may be considered only if the court first finds impermissible police coercion, Derrick relied principally upon Connelly. See Derrick, 924 F.2d at 818 (discussing Connelly, 479 U.S. at 163 & n. 1, 167, 107 S.Ct. 515). But Connelly concerned a confession by an individual who spontaneously approached an officer, and “without any prompting,” admitted “that he had murdered someone.” 479 U.S. at 160, 107 S.Ct. 515.¹⁴ Connelly’s holding was simply that “[a]bsent police conduct causally related to the confession, there is simply no basis for concluding that any state actor has deprived a criminal defendant of due process of law.” Id. at 164, 107 S.Ct. 515. It was on that premise that Connelly held the confession voluntary, explaining “that while mental condition is surely relevant to an individual’s susceptibility to police coercion, mere examination of the confessant’s state of mind can never conclude the due process inquiry.” Id. at 165, 107 S.Ct. 515 (emphasis added). Ultimately, the Court held that “coercive police activity is a necessary predicate to the finding that a confession is not ‘voluntary,’ ” id. at 167, 107 S.Ct. 515, thereby reiterating its earlier rejection of the proposition that “a defendant’s mental condition, by itself and apart pom its relation to official coercion, should ever dispose of the inquiry into constitutional ‘voluntariness,’ ” id. at 164, 107 S.Ct. 515 (emphasis added). In so ruling, Connelly reaffirmed, rather than departed from, the established law — that “[wjhile the state of the accused’s mind” was “certainly [a] factor[ ] to be evaluated in assessing the ‘voluntariness’ of an accused’s responses, [it was] not in and of [itself] determinative.” Schneckloth, 412 U.S. at 227, 93 S.Ct. 2041.

For these reasons, to the extent that Derrick held that the issue of police coercion during interrogations must be evaluated without regard to the individual circumstances of the suspect, it cannot be reconciled with prior opinions of this Court or with binding Supreme Court precedent. As previously noted, Derrick is irreconcilable with our post -Derrick en banc opinion in Doody, and so already is not binding authority in our court. Because the three judge panel in this case relied upon Derrick despite our holding in Doody, we now explicitly hold that Derrick no longer is good law and overrule it, as well as subsequent opinions that have relied upon it. See Amaya-Ruiz v. Stewart, 121 F.3d 486, (9th Cir.1997) (citing Derrick); United States v. Chischilly, 30 F.3d 1144, 1151 (9th Cir.1994) (citing Derrick).

Ill

Voluntariness of Preston’s Confession

We now turn to whether, under all of the circumstances- — -including Preston’s age, intellectual disability, and lack of sophistication, and the interrogation techniques used — there was coercive police action which overbore Preston’s will and rendered his confession involuntary. In doing so, we consider the district court’s factual account of what happened during the interrogation under the clearly erroneous test. United States v. Wolf, 813 F.2d 970, 974 (9th Cir.1987) (citing United States v. McConney, 728 F.2d 1195, 1200 (9th Cir.1984) (en banc)). We note that there can be no dispute as to what happened during the interrogation itself. We have the audiotapes and transcript of the interrogation, and so “are not consigned to an evaluation of a cold record, or limited to reliance on the detectives’ testimony.” Doody, 649 F.3d at 1009.

We review de novo the district court’s conclusion that the confession was voluntary. Wolf, 813 F.2d at 974. “Although sometimes framed as an issue of ‘psychological fact,’ the dispositive question of the voluntariness of a confession has always had a uniquely legal dimension.” Miller, 474 U.S. at 115-16, 106 S.Ct. 445 (internal citation omitted). “The notion of ‘voluntariness’ is itself an amphibian” and “purports at once to describe an internal psychic state and to characterize that state for legal purposes.” Culombe, 367 U.S. at 604-05, 81 S.Ct. 1860. Ultimately, “the admissibility of a confession turns as much on whether the techniques for extracting the statements, as applied to this suspect, are compatible with a system that presumes innocence and assures that a conviction will not be secured by inquisitorial means as on whether the defendant’s will was in fact overborne.” Miller, 474 U.S. at 116, 106 S.Ct. 445. Thus, the voluntariness determination “reflects deep, even if inarticulate, feelings of our society” about the acceptability of the imposition of certain interrogation methods on a particular person. Haley, 332 U.S. at 603, 68 S.Ct. 302 (Frankfurter, J., concurring). This “focus on the constitutional acceptability of the government conduct rather than merely on the defendant’s state of mind at the time of the confession requires us to consider legal concepts in the mix of fact and law and to exercise judgment about the values that animate legal principles.” Wolf, 813 F.2d at 974 (citation and internal quotation marks omitted).

In short, because of the interweaving of factual, legal, and judgmental considerations, our voluntariness review is de novo.

Applying de novo review, we conclude that, in light of the totality of the circumstances, including Preston’s individual characteristics, his confession was involuntary.

A

(1) We begin with “[cjonsideration of [Preston’s] reduced mental capacity,” a factor that is “critical because it [may] render[] him more susceptible to subtle forms of coercion.” N. Mariana Islands v. Mendiola, 976 F.2d 475, 485 (9th Cir.1993), overruled on other grounds by George v. Camacho, 119 F.3d 1393 (9th Cir.1997) (en banc).

Preston was eighteen, with an IQ of sixty-five. The two officers realized early in the interrogation that Preston suffered some sort of intellectual disability, as his initial responses gave them cause to believe that he had an impairment. They therefore inquired directly if he was “disabled.” Preston did not understand the word “disabled,” and so asked its meaning. That he had to ask for an explanation of a common word itself suggests the extent of his cognitive impairment. After the officers explained the word’s meaning, Preston agreed that he was disabled, elaborating that he was not able to complete his schooling as a result.

The psychological evidence introduced at the suppression hearing and trial confirmed, in spades, that Preston has many significant deficits in mental functioning. The psychologists’ evaluations demonstrated that Preston has a “very impaired” ability to understand “everyday interpersonal exchanges as well as ... formal legal” exchanges. “[A]ny English verbal material must be repeated, reinforced, and then revisited.” Without such repetition, “he may easily confuse the content of a conversation and give ... spurious responses” or be misled. Thus, “[h]is relatively poor verbal linguistic fluency is likely to result in misunderstanding of directions or translate into delayed, unconventional, or inappropriate responses in verbal settings.” Preston also finds “complexity ... confusing” and has trouble understanding abstract terms.¹⁵ He has difficulty following “simultaneous communication,” such as from two individuals speaking at once. Where there are two messages, Preston has trouble “sorting ... out” what they are saying “and deciding how to respond.” “[T]o set up the potential for him to understand something, you have to use rather simple, concrete terms.”¹⁶

During psychological testing, Preston expressed “distress of not knowing the answers to some questions or possibly consequent to difficulty understanding some of the questions he was asked.”¹⁷ Dr. DiBacco opined that Preston’s intellectual impairment made him more susceptible to the interrogative pressure. He testified that such distress at “not understanding] everything that is being said to him” creates situations in which “he may want to acquiesce to what he thinks people want him to say.” Even if a questioner asks him about “something that was impossible,” Preston may agree because he “thought that that’s what they wanted to hear.”

(2) These traits — being “easily confused,” “highly suggestible and easy to manipulate” — are consistent with characteristics of the intellectually disabled in general. See Richard A. Leo, Police Interrogation and American Justice 232 (2008). Studies show that “subjects with IQs well below average, such as those who are borderline or mentally handicapped, tend to be markedly more suggestible.” Gisli H. Gudjonsson, The Psychology of Interrogations and Confessions: A Handbook 382 (2003). “Lack of intelligence may render” the intellectually disabled “unable to understand what is being said” in an interrogation. President’s Panel on Mental Retardation, Report of the Task Force on Law 33 (1963). They “are highly susceptible to leading, misleading, and erroneous information,” and it is very “easy to get them to agree with and repeat back false or misleading statements, even incriminating ones.” Leo, supra, at 232. “Research shows that witnesses with mental deficiencies are highly influenced by questions that contain leading and misleading information.” Saul M. Kassin & Gisli H. Gudjonsson, The Psychology of Confessions: A Review of the Literature and Issues, 5 Psychol. Sci. Pub. Int. 33, 53 (2004).

Summarizing the evidence regarding how the intellectually impaired respond to contemporary police interrogation methods, several scholars have listed “seven common characteristics” of such people, including (1) “unusual[] susceptibility] to the perceived wishes of authority figures”; (2) “a generalized desire to please”; (3) difficulty “discerning] when they are in an adversarial situation, especially with police officers,” who they generally are taught exist to provide help; (4) “incomplete or immature concepts of blameworthiness and culpability”; (5) “[d]efieits in attention or impulse control”; (6) “inaccurate views of their own capacities”; and (7) “a tendency not to identify themselves as disabled” and to “mask[] their limitations.” Morgan Cloud et al., Words Without Meaning: The Constitution, Confessions, and Mentally Retarded Suspects, 69 U. Chi. L.Rev. 495, 511-13 (2002). These scholars. pronounced it “now ... beyond reasonable dispute” that the ABA was correct in stating “that the increased vulnerability of a mentally disabled suspect, and his or her naiveté, ignorance, confusion, suggestibility, delusional beliefs, extraordinary susceptibility to pressure, and similar considerations may make it possible for law enforcement officers to induce an involuntary statement by using techniques that would be acceptable in cases involving mentally typical suspects.” Id. at 509 (internal quotation marks and citation omitted). As a result of these traits, “[m]entally disabled individuals ... are ... long known to be vulnerable to coercion.” Brandon L. Garrett, The Substance of False Confessions, 62 Stan. L.Rev. 1051, 1064 (2010). The Supreme Court has so recognized, noting that “mental condition is surely relevant to an individual’s susceptibility to police coercion.” Connelly, 479 U.S. at 165, 107 S.Ct. 515.

We turn now to another consideration in our voluntariness analysis, the techniques the officers used when interrogating Preston.

As the Supreme Court pointed out in Miranda, most police officers have been trained in psychological techniques designed to induce confessions from reluctant suspects. 384 U.S. at 448-59, 86 S.Ct. 1602. Outside of the custodial setting, these techniques have never been held inherently coercive in the sense that the resulting confessions are necessarily involuntary, and we are certainly not so holding today. And the officers’ interrogation of Preston was not lengthy and did not take place hidden from public view in the confines of a police station, as was true in many cases in which courts have deemed the resulting confession involuntary. See, e.g., Reck v. Pate, 367 U.S. 433, 441, 81 S.Ct. 1541, 6 L.Ed.2d 948 (1961); Doody, 649 F.3d at 1009. But, as we have recognized, there is “no single litmus-paper test for constitutionally impermissible interrogation,” and no individual feature of the interrogation is determinative of whether a confession is voluntary. Culombe, 367 U.S. at 601, 81 S.Ct. 1860. “[A]s interrogators have turned to more subtle forms of psychological persuasion,” and away from physical coercion, “courts have found the mental condition of the defendant a more significant factor in the ‘voluntariness’ calculus.” Connelly, 479 U.S. at 164, 107 S.Ct. 515.¹⁸ It simply “takes less” in terms of sophisticated police interrogation techniques “to interfere with the deliberative processes of one whose capacity for rational choice is limited than it takes to affect the deliberative processes of one whose capacity is not so limited.” Smith, 910 F.2d at 1497.

Among the police tactics used here were several recommended by a manual on police interrogation, see Fred E. Inbau, John E. Reid, Joseph P. Buckley & Brian C. Jayne, Criminal Interrogation and Confessions (5th ed. 2013) (“Reid manual”),¹⁹ from which both the officers who interrogated Preston were trained. The officers, however, sometimes disregarded the manual’s cautions about the tactics they used.

For example, using one of the recommended approaches, the two officers asked Preston a number of questions that presented him with two alternatives as to how the crime was committed. See id. at 293-303. “Both alternatives are highly incriminating, but they are worded in such a way that one alternative acts as a face-saving device whilst the other implies some repulsive motivation.” Gudjonsson, supra, at 19. In this instance, Preston was asked to choose, for example, whether he was a monster — a sexual predator who repeatedly preys on children — or if the abuse of the child was a one-time occurrence.

These questions were derived from similar exemplars in the Reid manual. Reid manual, supra, at 296-97, 298. The manual, however, suggests that the inculpatory alternatives technique recommended may be unduly coercive when used for suspects of seriously impaired mental ability: it trains agents in the alternative questioning method with the understanding that “no innocent suspect, with normal intelligence and mental capacity, would acknowledge committing a crime merely because the investigation contrasted a less desirable circumstance to a more desirable one and encouraged the suspect to accept it.” Reid manual, supra, at 303 (emphasis added). The psychological evidence regarding Preston’s intellectual disabilities confirms this assessment by indicating that he is confused by complexity, abstraction, and multiplicity, and likely to acquiesce in suggestions made by the questioner. As a result, recognizing that where one is asked “a or b,” one can answer “neither one,” rather than acquiescing in one or the other, could well have exceeded his intellectual abilities.

A second questioning technique the officers used with Preston was repeated pressure to change answers inconsistent with guilt and adopt answers evidencing guilt instead. Repeatedly rejecting Preston’s denials or equivocations, Kraus and Secatero asked him the same questions over and over until he finally assented and adopted the details that the officers posited.²⁰ Such acquiescence and willingness to “shift” answers in response to interrogative pressure is common for the intellectually disabled, who, when presented with leading or suggestive questions, “frequently seek to conform to the perceived desires of the interrogator.” Stanley L. Brodsky & Allyson D. Bennett, Psychological Assessments of Confessions and Suggestibility in Mentally Retarded Suspects, 33 J. Psychiatry & L. 359, 363 (2005).²¹

Identifying a third technique that the officers used, Dr. DiBacco testified that “[tjhere were a number of times” during the officers’ interrogation of Preston “that the desired response was embedded in the question.” The agents carried out this technique by asking Preston questions that contained details about the allegations already made. For example, when Kraus first asked if Preston used a condom— which Kraus knew had been alleged during the child’s forensic interview — Preston responded that he did not know. Only after Kraus told Preston the correct answer— “Yeah, you did?” — did Preston adopt that detail. The same pattern was repeated to pressure Preston to adopt the suggested fact that he had unzipped his pants.

Particularly strong evidence of Preston’s suggestibility is that he adopted answers that were demonstrably false.²² Most tellingly, the officers rejected Preston’s repeated denials of being home on Friday, saying “you telling us ... you’re not being here ... We just don’t buy it,” “the fact is, we know you were here,” “there’s no denying” it, and so “you have to remember what happened Friday.” Eventually, Preston stopped denying that he was home on Friday, instead acquiescing that “there’s a bunch of little kids over here, and plus that other guy,” and that he “was inside just by myself’²³ — even though it is undisputed that Preston in fact was not at home on that day.²⁴

That Preston attempted to tell the officers what they wanted to hear is also clear from some of his other shifting responses. At first, when the officers asked if any of the children came inside his house on Friday, Preston denied it. As the officers repeatedly asked this question, communicáting that his initial responses were not what they wanted, his answer shifted from “nobody came inside” to “[i]t’s just like what you guys said, that guy came in.”

The agents coupled the techniques of alternative questioning, providing suggestive details, and repetitious and insistent questions with other techniques that the Reid manual specifically cautions against. The Reid manual specifically warns that the questioning “should not be, in any way, based on leniency if the more understanding alternative question is accepted.” Reid manual, supra, at 300 (emphasis added). It also cautions that when questioning people of low intelligence, investigators should avoid offering promises of leniency or using deceptive interrogation techniques due to the vulnerability of this group. Id. at 332-33, 352, 429.

Kraus and Secatero ignored these admonitions. They told Preston that if he were not the “kind of guy that [abuses children] all the time,” but instead was the type of “guy that just” had “a bad day” and did it once, then he could simply “move on.” The message conveyed was that Preston would not be punished if he admitted to being a one-time child rapist — which was, of course, not true.

The officers misled Preston in other ways as well, telling him that his written confession was just an apology note to the child, that they would not tell anyone else what he said, and that the confession would never leave the “folder” or the United States Attorney’s Office. At the same time, they told Preston that he was free to leave only after he finished answering their questions, and threatened that they would keep returning until Preston did so. In this way, the police paired the prospect of relentless questioning with false promises of leniency.²⁵ Such tactics, in combination, would be hard for a person of Preston’s impaired intelligence to withstand or rationally evaluate.²⁶

Assuredly, interrogating officers can make false representations concerning the crime or the investigation during questioning without always rendering an ensuing confession coerced. See, e.g., Frazier v. Cupp, 394 U.S. 731, 739, 89 S.Ct. 1420, 22 L.Ed.2d 684 (1969). But false promises stand on a different footing. In particular, the Supreme Court has observed that “the test of voluntariness” is “whether the confession was extracted by any sort of threats or violence, or obtained by any direct or implied promises, however slight, or by the exertion of any improper influence.” Hutto v. Ross, 429 U.S. 28, 30, 97 S.Ct. 202, 50 L.Ed.2d 194 (1976) (per curiam) (quoting Bram v. United, States, 168 U.S. 532, 542-43, 18 S.Ct. 183, 42 L.Ed. 568 (1897)) (internal quotation marks and formatting omitted). In Henry v. Kernan, 197 F.3d 1021, 1027-28 (9th Cir.1999), for example, we concluded that detectives’ deceptive tactics overcame the defendant’s will and rendered his statement involuntary when the detectives made deliberately misleading comments “intended to convey the impression that anything said by the defendant would not be used against him for any purposes.”

The types of deception used here, which primarily related to considerations extrinsic to the suspect’s guilt or innocence, are particularly problematic when used on a person with an intellectual disability. Intrinsic falsehoods, which relate to the facts of the crime itself or of the investigation— such as falsely informing a suspect that the victim had survived and identified the suspect — do “not lead [the suspect] to consider anything beyond his own beliefs regarding his actual guilt or innocence, his moral sense of right and wrong, and his judgment regarding the likelihood that the police had garnered enough valid evidence linking him to the crime.” Holland v. McGinnis, 963 F.2d 1044, 1051-52 (7th Cir.1992). But here, the police did not simply inflate the amount of incriminating evidence against Preston. Instead, they suggested falsely that if he confessed, his admissions would not be used against him — he could “move on” after apologizing to the child, rather than being punished. This approach “interject[ed] the type of extrinsic considerations” more likely to “distort[] an otherwise rational choice of whether to confess or remain silent.” Id. at 1051.²⁷ The intellectually disabled are more susceptible to such extrinsic deception tactics.

“Because of their cognitive deficits and limited social skills, the mentally retarded ... often lack the ability to appreciate the seriousness of a situation.” Steven A. Drizin & Richard A. Leo, The Problem of False Confessions in the Post-DNA World, 82 N.C. L.Rev. 891, 919-20 (2004). “Under interrogation, they are not likely to understand that the police detective who appears to be friendly is really their adversary or to comprehend the long-term consequences of making an incriminating statement.” Jon B. Gould & Richard A. Leo, One Hundred Years Later: Wrongful Convictions After a Century of Research, 100 J.Crim. L. & Criminology 825, 847 n.119 (2010). They fail “to understand the context in which interrogation occurs, the legal consequences embedded in the rules or the significance of confessing.” Cloud et al., supra, at 501. In particular, research shows that the intellectually disabled are “significantly more likely ... to believe the suspect will be allowed to go home after making a confession” to a serious crime. Gudjonsson, supra, at 326. So being told falsely that, after a confession, one could simply “move on,” or that the confession would be kept confidential, is likely to have a considerably greater impact on a person with serious intellectual impairments, such as Preston, than on an individual of normal intelligence.

IV

Analysis and Conclusion Regarding Voluntariness

Considered all together, the various factors here — Preston’s severe intellectual impairment; the police’s repetitive questioning and the threats that it would eontinue without end; the pressure placed on Preston to adopt certain responses; the use of alternative questions that assumed his culpability; the officers’ multiple deceptions about how the statement would be used; the suggestive questioning that provided details of the alleged crime; and the false promises of leniency and confidentiality — leave us convinced that Preston’s will was overborne and his statement involuntary. “The line, between proper and permissible police conduct and techniques and methods offensive to due process is, at best, a difficult one to draw, particularly in cases such as this where it is necessary to make fine judgments as to the effect of psychologically coercive pressures and inducements on the mind and will of an accused.” Haynes v. Washington, 373 U.S. 503, 515, 83 S.Ct. 1336, 10 L.Ed.2d 513 (1963). But we have long been aware that even “subtle coercion ... can have an extraordinary effect on one of low mental capabilities.” N. Mariana Islands, 976 F.2d at 485. We must be mindful to protect the constitutional rights of all members of our society, not just those of normal intelligence and cognitive functioning. Even if we would reach a different conclusion regarding someone of normal intelligence, we hold that the officers’ use of the methods employed here to confuse and compel a confession from the intellectually disabled eighteen-year-old before us produced an involuntary confession.

Accordingly, we conclude that the district court erred in admitting Preston’s confession.²⁸

V

Sufficiency of the Evidence

As a final matter, Preston argues on appeal that the evidence was constitutionally insufficient to establish essential elements of the crime charged. Specifically, Preston contends that the government failed to introduce sufficient evidence to prove that he “engaged in sexual contact with the victim’s buttocks or anus with [his] penis,” (alteration in original), or that he acted with “an intent to abuse, humiliate, harass, degrade, or arouse or gratify the sexual desire of any person,” 18 U.S.C. § 2246(3).

The Double Jeopardy Clause of the Fifth Amendment provides that no person shall “be subject for the same of-fence to be twice put in jeopardy.” “It has long been settled, however, that the Double Jeopardy Clause’s general prohibition against successive prosecutions does not prevent the government from retrying a defendant who succeeds in getting his first conviction set aside, through direct appeal or collateral attack, because of some error in the proceedings leading to conviction.” Lockhart v. Nelson, 488 U.S. 33, 38, 109 S.Ct. 285, 102 L.Ed.2d 265 (1988). But the Supreme Court has recognized an exception to the government’s right to retry a defendant without offending the Double Jeopardy Clause where the conviction is overturned for insufficient evidence. Burks v. United States, 437 U.S. 1, 11, 98 S.Ct. 2141, 57 L.Ed.2d 1 (1978). This exception recognizes that the “Double Jeopardy Clause forbids a second trial for the purpose of affording the prosecution another opportunity to supply evidence which it failed to muster in the first proceeding.” Id. As Preston did not “ ‘waive[ ]’ his right to a judgment of acquittal by moving for a new trial,” id. at 17, 98 S.Ct. 2141, we must address the sufficiency of the evidence question even though we are remanding for a new trial.

Viewing all the evidence introduced at trial, including the erroneously admitted confession, Lockhart, 488 U.S. at 40-41, 109 S.Ct. 285, in the light most favorable to the prosecution, a “ ‘rational trier of fact could have found the essential elements of the crime beyond a reasonable doubt,’ ” United States v. Rios, 449 F.3d 1009, 1011 (9th Cir.2006) (quoting Jackson v. Virginia, 443 U.S. 307, 319, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1979)). There was sufficient evidence to support that the sexual act took place, including Preston’s confession, the child’s statements to his grandparents and uncle, and the child’s repeated complaints of pain. Further, contrary to Preston’s contention that his denial of being motivated by a sexual “urge” undermines entirely a finding of the requisite intent, we have long recognized that a jury may draw inferences from circumstantial evidence to find the requisite intent. See, e.g., Ngo v. Giurbino, 651 F.3d 1112, 1114-15 (9th Cir.2011). There is little doubt that the jury had ample circumstantial evidence, including the evidence of the act itself, to find that Preston intended to gratify his sexual desire.

Finally, Preston’s confession itself was sufficiently corroborated by independent evidence. See United States v. Norris, 428 F.3d 907, 914-15 (9th Cir.2005) (citing United States v. Lopez-Alvarez, 970 F.2d 583, 592 (9th Cir.1992)). Corroborating evidence need “not independently establish any element beyond a reasonable doubt.” United States v. Delgado, 545 F.3d 1195, 1206 (9th Cir.2008) (internal quotation marks and citation omitted). Here, the child’s statements to his relatives and the forensic interviewer, his complaints of pain, and the DNA evidence together “support the essential facts admitted” by Preston “sufficiently to justify” an “inference of their truth” by the factfinder. United States v. Corona-Garcia, 210 F.3d 973, 978 (9th Cir.2000).

Thus, we conclude that sufficient evidence supported Preston’s conviction at the first trial, and we remand for a new trial.²⁹

REVERSED AND REMANDED.

1 . Disability advocates now use the term "intellectual disability" rather than "mental retardation.” Congress has adopted the term "intellectual disability” for use in federal statutes. See Rosa's Law, Pub.L. No. 111-256, 124 Stat. 2643 (Oct. 5, 2010). We therefore use the term "intellectual disability,” except where the earlier locution “mental retardation” is used in the cited authority.

2 . The Supreme Court is currently reviewing the constitutionality of Florida's scheme for identifying individuals with intellectual disabilities in capital cases, which mandates that a person must have an IQ below seventy to be classified as intellectually disabled. See Hall v. Florida, - U.S. -, 134 S.Ct. 471, 187 L.Ed.2d 316 (2013).

3 .Preston underwent two psychological evaluations. First, he was examined for his competency to stand trial by Dr. Daniel C. Cady, who provided a report, admitted as an exhibit at trial. The defense retained Dr. John DiBacco, who examined Preston and testified regarding his confession. Dr. DiBacco has a Ph.D. in psychology from Vanderbilt University, is an adjunct professor of clinical psychology at Arizona State University, and is the senior consultant to Maricopa County's Child Protective Services department.

4 . One swab was used to collect a sample from various parts of the child’s underwear, including from the outside waistband. So the DNA found may have come from the outside waistband of the underwear.

5 . The DNA of Preston’s brother, mother, and father, all of whom also live next door to the child and presumably have similar DNA profiles to Preston, was never compared to the DNA collected from the underwear.

6 . One of the forensic psychologists who examined Preston observed that he has difficulties with abstract terms. The examining psychologist observed, “I would suspect that he’d have a tough time defining molestation, forensics.” Both terms were used in the interrogation.

7 .Toward the end of thé encounter, the officers realized that the alleged incident had not occurred on Friday, and changed the day in the confession they wrote to Thursday — still incorrect.

8 . The officers testified that, before they began recording the conversation, Kraus told Preston that he was free to leave or stop talking at any time. But Kraus also testified that, when he told Preston, “After the interview, you’re free to go,” he was merely restating what he had told Preston before the interview.

9 . At the suppression motion hearing, Secatero represented that Preston wrote out the statement, although Kraus "dictated it for the defendant.” At trial, Kraus acknowledged that he penned the statement.

10 . No condom was introduced at trial; the investigators had not looked for one.

11 . In Miranda v. Arizona, 384 U.S. 436, 86 S.Ct 1602, 16 L.Ed.2d 694 (1966), the Supreme Court, recognizing “that the coercion inherent in custodial interrogation blurs the line between voluntary and involuntary statements,” and presents too great a risk that any resulting statement is compelled, required that suspects be given certain advisements of their rights prior to custodial interrogation as a condition precedent to the admissibility of any statement into evidence. Dickerson v. United States, 530 U.S. 428, 435, 120 S.Ct. 2326, 147 L.Ed.2d 405 (2000). But ‘‘[t]he requirement that Miranda warnings be given does not, of course, dispense with the voluntariness inquiry” for suspects in custody. Id. at 444, 86 S.Ct. 1602. Concomitantly, the voluntariness standard applies to suspects not in custody. See, e.g., Beckwith v. United States, 425 U.S. 341, 348, 96 S.Ct. 1612, 48 L.Ed.2d 1 (1976).

Preston does not argue that he was in custody during the interrogation. We proceed on the premise that he was not.

12 . The Supreme Court has made similar observations when considering the interrogation of children, recognizing, ''[t]hat which would leave a man cold and unimpressed can overawe and overwhelm a lad in his early teens.” Haley v. Ohio, 332 U.S. 596, 599, 68 S.Ct. 302, 92 L.Ed. 224 (1948).

13 . Although probable truth does not demonstrate that a confession was voluntary, we note that there is abundant research that the intellectually disabled "are more likely to confess falsely for a variety of reasons." Jon B. Gould & Richard A. Leo, One Hundred Years Later: Wrongful Convictions After a Century of Research, 100 J.Crim. L. & Criminology 825, 847 (2010). "[0]f the first 200 DNA exonerations in the U.S., 35% of the false confessors were 18 years or younger and/or had a developmental disability.” Saul M. Kassin et al., Police-Induced Confessions: Risk Factors and Recommendations, 34 Law & Hum. Behav. 3, 19 (2010). Another recent study found "69% of exonerated persons with mental disabilities were wrongly convicted because of false confessions.” Id. In holding that mentally retarded offenders may not be executed constitutionally, Atkins noted the "possibility of false confessions” by such' individuals, and that "at least one mentally retarded person who unwittingly confessed to a crime that he did not commit” had been given a death sentence. Atkins, 536 U.S. at 320 & n. 25, 122 S.Ct. 2242.

Recognizing the heightened likelihood of false confessions by intellectually disabled suspects does not contravene Rogers' directive that truth or falsity is not part of the voluntariness inquiry. Rather, just as I.D.B. v. North Carolina, - U.S. -, 131 S.Ct. 2394, 2401, 180 L.Ed.2d 310 (2011), cited the "heightened risk of false confessions from youth” as a reason to consider a child's age— not the truth or falsity of the child's confession — as part of the Miranda in-custody requirement, our observation regarding false confessions by the intellectually disabled similarly informs the importance of carefully taking into account the intellectual disability of the suspect — not the truth or falsity of the confession — as part of the totality-of-the-circumstances voluntariness inquiry.

14 . In later psychological evaluations, the defendant was found to have chronic schizophrenia and revealed that he had confessed after hearing "God's voice" tell him "either to confess to the killing or to commit suicide.” Connelly, 479 U.S. at 161, 107 S.Ct. 515.

15 . Preston's difficulty understanding language and abstract concepts was evident at the hearing on his jury trial waiver. After Preston told the district court that he believed that to be convicted, “[mjaybe five, six” of the jurors would have to decide that he was guilty, the district court explained that that was incorrect and that "[i]n order for you to be found guilty, all twelve of the jurors must agree that you are guilty of the charges brought against you by the government beyond a reasonable doubt.” The district court then asked Preston if he could be found guilty of the charges if "eleven jurors thought you were guilty and one juror still thought you were not guilty.” Preston replied, "Yes.”

16 . The district court's finding that Preston’s "responses and verbal demeanor demonstrate that he understood the questions” was clearly erroneous. The district court reached that conclusion by listening to the audio recording and reading the transcript of the interrogation, which we have done as well. The audiotape and transcript demonstrate that Preston’s answers were often nonresponsive or muddled, evidencing clear confusion. Preston’s most common responses were "I don't know,” "I don’t remember,” and "Huh?” For example, when Kraus asked Preston if he remembered the child being over at his house on Friday, Preston responded, "No. All I know is it's like — there's a bunch of little kids over here, and plus that other guy, but — I don't know, but — I don’t know what happened that day, but I wasn't on drugs, under the influence or nothing.” Also, the officers sometimes themselves posed garbled questions, such as "Well, he — okay, but he said— you know, and I — yeah, he probably wants— so did it happen that day?” Preston did not ask for them to repeat the question, but instead, much as he often did, responded, "I don’t know. Probably.”

17 .People with cognitive deficits generally have "tendencies to mask or disguise their cognitive deficits and to look to others — particularly authority figures — for the appropriate cues to behavior.” Richard A. Leo, Police Interrogation and American Justice 232 (2008). Thus, a "disabled person may feel compelled to answer a question, even if the question exceeds his ability to answer.” Morgan Cloud et ah, Words Without Meaning: The Constitution, Confessions, and Mentally Retarded Suspects, 69 U. Chi. L.Rev. 495, 513 (2002).

18 . Culombe v. Connecticut, 367 U.S. 568, 624-25, 634, 81 S.Ct. 1860, 6 L.Ed.2d 1037 (1961), for example, considered the defendant’s “mental age of nine or nine and a half,” and that he was "suggestible and subject to intimidation,” in finding that his "will was broken.” Reck v. Pate, 367 U.S. 433, 442, 81 S.Ct. 1541, 6 L.Ed.2d 948 (1961), held a confession involuntary after observing that the nineteen-year-old "Reek’s youth, his subnormal intelligence, and his lack of previous experience with the police make it impossible to equate his powers of resistance to overbearing police tactics with those of” other individuals. Payne v. Arkansas, 356 U.S. 560, 567, 78 S.Ct. 844, 2 L.Ed.2d 975 (1958), concluded that the confession of a “mentally dull 19-year-old youth” was coerced. Fikes v. Alabama, 352 U.S. 191, 196-97, 77 S.Ct. 281, 1 L.Ed.2d 246 (1957), determined that a confession was involuntary in part because the suspect was "uneducated” and "of low mentality, if not mentally ill,” and so "was a weaker and more susceptible subject than” in other cases.

19 . The "Reid method,” named for the manual of which Reid was a coauthor, is widely used by law enforcement agencies. See Miranda, 384 U.S. at 449 n. 9, 86 S.Ct 1602; Douglas Starr, The Interview: Do Police Interrogation Techniques Produce False Confessions?, The New Yorker, Dec. 9, 2013, at 42.

Unless specified, the cited portions of the fifth edition of the manual, released in 2013, are unchanged substantively from the fourth edition, which was the most recent version when the events at issue took place.

20 . Psychologists have found that an interrogator's explicit and implicit negative feedback, such as “repeating the same questions several times because the answers given are not acceptable to the interrogator,” can cause a suspect to “shift” or "adapt” his answers to responses acceptable to the questioner. Gudjonsson, supra, at 347-48 (internal quotation marks and citation omitted).

21 . Research shows that in general, “the mentally retarded are eager to please.” Leo, supra, at 232. "They tend to have a high need for approval,” and thus are "prone to being acquiescent,” particularly with "authority figures.” Id. This desire for approval "is manifested in an acquiescence response bias, a tendency to say 'yes.' " Kassin & Gudjonsson, supra, at 53. A "common phenomenon” among the intellectually disabled "is the mental process of 'cheating to lose,' that is, accepting blame so that others will not be angry.” Cloud et al., supra, at 511-12. "If an authority figure such as a police officer makes it clear to the individual that he wants a confession, even an innocent disabled person may confess so a law enforcement officer will not become angry with him.” Id. at 512.

As a result, studies show that “people with mental retardation [are] more likely to yield to leading questions and change their answers in response to mild negative feedback.” Kassin et al., supra, at 21.

22 . Although the introduction of the incorrect facts here was not intentional, the Reid manual suggests that officers may introduce such falsities to verify the accuracy of a confession: "[o]ne method for checking the authenticity of a voluntary confession, or one that seems to be the result of mental illness, is to introduce some fictitious aspects of the crime and test whether the suspect will accept them as actual facts relating to the occurrence.” Reid manual, supra, at 349.

23 . The district court’s finding that Preston’s admissions were not necessarily tied to the previous Friday is clearly erroneous. As explained above, Preston admitted that there were "a bunch of little kids over here” in response to questions about whether the victim was at his home six days before, on Friday. Preston eventually admitted that the child had come in when asked again, "So what happened Friday? ... So you were inside in there. Did the kids come inside then?”

The district court's suggestion that the officers' occasional references to the correct day of the month dissipated any confusion created by the numerous, repeated references to the wrong day of the week is also clearly erroneous. Twice, the officers referred to the "the 23rd,” but they always coupled this date with the incorrect day of the week, which they emphasized more than the date. In the first instance, Kraus said, "Did he come over here on? Was it the 23rd, Friday? Yeah? Yeah, the 23rd, which would have been, what, a week ago?” Secatero responded, “Yeah.” Kraus continued, "Yeah, a week ago Friday.” Secatero affirmed, "Last Friday.”

Later, when Kraus was writing the confession, he said to Preston, "basically, what you’re saying is, on the 24th, which was last Friday, you were inside and — •" Secatero interjected that the "23rd is the Friday.” After Kraus agreed, Secatero corrected himself, saying "Or — well, Thursday.... Thursday evening, and the interview [of the child] took place on the 24th, Friday. It was last Thursday ... Yeah, last Thursday.” Kraus then continued to Preston, “So on Thursday, you were hanging out inside, and who else was— who else was there?” Preston responded, "Huh?” The references to the correct day of the month embedded within these exchanges could hardly have cleared up the confusion.

24 .Researchers have observed that this type of persistence can be particularly hard to resist for suspects who are naturally compliant and eager to please, such as the intellectually disabled. Welsh S. White, False Confessions and the Constitution: Safeguards Against Untrustworthy Confessions, 32 Harv. C.R.-C.L. L.Rev. 105, 124-25 (1997) (internal citations omitted). "When an interrogator repeatedly tells a suspect that there is not the slightest doubt of his guilt and brushes aside the suspect’s attempt to give a different version of the facts, the suspect is even more likely to acquiesce.” Id. at 125.

25 . We point out the officers’ statements that Preston was free to leave only after the questioning was finished and that the officers would return until he answered their questions not to suggest that Preston was in custody for Miranda purposes, but to show some of the pressures the officers put upon Preston to overbear his will.

26 . Studies show that "people with significant intellectual impairment do not fully appreciate the legal consequences for suspects of making self-incriminating admissions during questioning.” Gudjonsson, supra, at 326; see id. (describing a "mean IQ score of 68” as a "significant intellectual impairment”). So a person such as Preston would not be able to see past the officers’ lies about the use of his apology letter or the punishment that he could face to understand "the implications or consequences of his statements in the way a person of normal intelligence” would. President’s Panel on Mental Retardation, supra, at 33.

27 . This danger is illustrated by Lynumn v. Illinois, 372 U.S. 528, 83 S.Ct. 917, 9 L.Ed.2d 922 (1963), in which the Supreme Court deemed police trickery of this sort, taken to an extreme, to render a confession involuntary. In Lynumn, the police told a young mother, who had no prior experience with the police and no reason to believe the police were not capable of carrying out their threats, that if she were charged with a crime, she would probably lose her welfare benefits and custody of her children, but that if she cooperated, they "would go light with her.” Id. at 533-34, 83 S.Ct. 917. The Court deemed it “clear that a confession made under such circumstances must be deemed not voluntary, but coerced.” Id. at 534, 83 S.Ct. 917; see also Holland, 963 F.2d at 1052 (describing Lynumn and explaining that this sort of pressure "distorted the suspect’s rational choice (i.e., is it wise or morally right to confess given [her own understandings of her guilt or innocence and moral sense of right and wrong]?) by introducing a completely extrinsic consideration: an empty but plausible threat to take away something to which she and her children would otherwise be entitled”).

28 . The government has not argued that the admission of Preston’s confession constituted harmless error. See Arizona v. Fulminante, 499 U.S. 279. 295-96. 306-12. 111 S.Ct. 1246, 113 L.Ed.2d 302 (1991). We therefore do not consider whether the admission of the confession contributed to his conviction.

29 . We need not reach Preston's other grounds for appeal, including that his jury trial waiver — which he entered with the understanding that his confession would be admitted at trial' — was not voluntary, knowing, and intelligent. Even if the prior jury trial waiver was voluntary, because we have found error entitling Preston to a retrial, Preston's earlier consent to a bench trial, made prior to this appeal, does not carry over to any later retrial. See United States v. Mortensen, 860 F.2d 948, 950 (9th Cir.1988); see also United States v. Groth, 682 F.2d 578, 579-80 (6th Cir.1982); F.M. Davies & Co. v. Porter, 248 F. 397, 398 (8th Cir.1918).