United States v. Specialist JOSHUA D. LEWIS

UNITED STATES ARMY COURT OF CRIMINAL APPEALS Before WOLFE, SALUSSOLIA, and ALDYKIEWICZ Appellate Military Judges UNITED STATES, Appellant v. Specialist JOSHUA D. LEWIS United States Army, Appellee ARMY MISC 20180260 Headquarters, Fort Hood Douglas Watkins, Military Judge Lieutenant Colonel Joseph M. Fairfield, Staff Judge Advocate For Appellant: Colonel Tania M. Martin, JA; Lieutenant Colonel Eric K. Stafford, JA; Captain Catharine M. Parnell, JA; Captain Allison L. Rowley, JA (on brief); Colonel Steven P. Haight, JA; Lieutenant Colonel Eric K. Stafford, JA; Captain Catharine M. Parnell, JA; Captain Allison L. Rowley, JA (on reply brief and brief on specified issue) For Appellee: Lieutenant Colonel Christopher D. Carrier, JA; Captain Patrick G. Hoffman, JA; Captain Benjamin A. Accinelli, JA; Captain Benjamin J. Wetherell, JA (on brief); Lieutenant Colonel Christopher D. Carrier, JA; Major Jack D. Einhorn, JA; Captain Benjamin A. Accinelli, JA; Captain Benjamin J. Wetherell, JA (on brief on specified issue). 26 October 2018 --------------------------------------------------------------------- OPINION OF THE COURT AND ACTION ON APPEAL BY THE UNITED STATES FILED PURSUANT TO ARTICLE 62, UNIFORM CODE OF MILITARY JUSTICE --------------------------------------------------------------------- WOLFE, Senior Judge: The United States appeals the ruling of a military judge suppressing statements made by the accused. 1 The military judge found none of the accused’s interrogations included coercion, but the statements from each interrogation were 1 We have jurisdiction over this appeal under Article 62, UCMJ. The parties raise no jurisdictional issues to our attention nor have we independently identified any. Unlike our reviews under Article 66, UCMJ, our review is limited solely to questions of law. LEWIS—ARMY MISC 20180260 involuntary. We affirm the military judge’s ruling regarding one of the two statements, and reverse as to the other. The accused is charged with the sexual assault of Miss ZC, a child between the age of 12 and 15. 2 As we discuss in more detail below, military law enforcement questioned the accused three separate times. During the first interview, appellant made statements both before and after receiving an Article 31, UCMJ / Miranda 3 rights advisement. A month later, the accused waived his rights and was again questioned. At a third session, the accused was again questioned after waiving his rights. At trial, the defense filed a motion to suppress the statements and derivative evidence. The military judge’s initial ruling was narrowed upon reconsideration. The government, in turn, only appeals portions of the amended ruling. In other words, the legal issues presented to this court are narrower than the whole story may otherwise suggest. While we limit our holding to the issues properly presented, we provide a broader factual picture for context. BACKGROUND A. An unusual report 4 On about 11 May 2017, an unknown woman approached a Charge of Quarters (CQ) desk at Fort Hood, Texas. The woman had a dog with her and was “wielding” a baseball bat. She then alleged to the CQ that someone had “touched” her daughter. The woman wore a green jumpsuit and a battle dress uniform (BDU) 5 jacket with a last name on the jacket. She told the CQ that she was ex-military but did not otherwise identify herself. The CQ escorted the woman to the accused’s unit, but she departed the area after she received a text message from the accused. 2 The conduct was charged as a sexual assault by bodily harm, in violation of Article 120 and as a sexual assault of a child, in violation of Article 120b. 3 Miranda v. Arizona, 384 U.S. 436 (1966). 4 Unless otherwise stated, our factual summary is taken from either the factual findings of the military judge or appellee’s brief. 5 The Army began phasing out the BDU on 14 June 2004 with a final wear-out date for both active and reserve forces of 30 April 2008. Dep’t of the Army, All Army Activities Message 004/2008, Military Uniforms and Accessories Wear Out Dates (January 2008) (ALARACT). 2 LEWIS—ARMY MISC 20180260 The allegation, brief as it was, was conveyed to military law enforcement. At this point, investigators did not know the identity of the woman, the name or age of any victim, or any specifics regarding the alleged offense. Although the limited record in this interlocutory appeal does not explain why, it is clear that law enforcement somehow knew that the woman’s accusation was directed at the accused. B. The three statements This brief accusation touched off the first of three interrogations all of which the military judge suppressed. Although this appeal is limited to the latter two statements, it is clear from the military judge’s ruling that his suppression of the second and third statements is related to police conduct during the first. Accordingly, even though the suppression of the first statement is not before this court as the government has not appealed the matter, we discuss it in some depth. 1. The first statement Based on the unknown woman’s report, the accused was escorted to Army Criminal Investigation Command (CID) at Fort Hood. Consistent with standard procedures, the accused was searched for officer safety and his personal belongings, to include a cell phone, were placed in a locker. The accused was then questioned by an investigator from Military Police Investigations (MPI). Although the test is an objective one, United States v. Swift, the questioning agent at least subjectively believed the alleged touching was sexual. 53 M.J. 439, 446 (C.A.A.F. 2000) (“Whether a person is a suspect is an objective question . . .”). After taking some initial biographical information, and prior to any rights waiver, the investigator asked, “Real quick, I had a crazy lady come in and report something, I don’t know who she is, she mentioned something about a daughter, so do you happen to know someone whose mom is crazy?” The accused provided the name of the woman who had made the unusual report, and who is the mother of the alleged victim, Miss ZC. The accused then asked the investigator, “What’s going on?” The investigator responded, “Well you mentioned the name right off the bat.” The accused then stated he was just trying to get accurate information because he thought he and the woman had settled the situation. The investigator asked, “Is there a situation?” The accused responded, “They thought something happened between me and their daughter.” 3 LEWIS—ARMY MISC 20180260 The investigator returned to asking the accused about biographical data, but then asked the accused, “Do you want to tell me about the story?” The accused then made a statement admitting that he had touched Miss ZC’s leg two years earlier, making Miss ZC uncomfortable. During a break in the interview, and outside the presence of the accused, the investigator discussed with other agents whether she should give a cleansing statement to the accused. The agents decided not to give a cleansing statement. 6 The military judge found that the investigator did not advise the accused of his rights before this time because she “wanted to get the accused’s story and to get the identity and contact information of the victim” and that she “was concerned that if she advised him of his rights he might invoke.” The military judge would reasonably conclude that the investigator acted in implied bad faith. After obtaining some additional biographical information from the accused, the investigator advised the accused of his rights under Article 31. She began by stating, “You mentioned a story, and I didn’t ask any questions, I’m not allowed to ask questions, until I advise you of your rights, so we’ll go through that first, and then you’ll tell me your story again.” 7 When explaining the difference between someone “accused” of a crime and someone “suspected” of a crime, she stated “suspected means I can talk to you.” 8 The accused waived his rights and agreed to talk to the investigator. The military judge found the accused’s demeanor when being advised of his rights and waiving them was “willing” and “that he was inquisitive.” 6 That the agents did not give a cleansing statement is certainly relevant to the issue of voluntariness we later discuss. However, that the agents discussed giving a cleansing statement outside the presence of the accused is not likely relevant to a voluntariness determination. “Events occurring outside of the presence of the suspect and entirely unknown to him surely can have no bearing on the capacity to comprehend and knowingly relinquish a constitutional right.” Moran v. Burbine, 475 U.S. 412, 422 (1986). 7 The quoted language is from the military judge’s findings of fact. The emphasis is ours. 8 This appears to be a reference to the fact that the Sixth Amendment right to counsel – and the associated limitations on police questioning – attaches at preferral of charges. See United States v. Wattenbarger, 21 M.J. 41, 43 (C.M.A. 1985). However, from the accused’s perspective, defining what it meant to be a “suspect” in this manner was likely confusing, as it was incomplete. 4 LEWIS—ARMY MISC 20180260 The accused was then questioned for about forty minutes. The accused admitted to touching Miss ZC’s thigh, rubbing her leg, making her uncomfortable, and that she was fifteen at the time. 2. The second statement After the first interrogation – and now knowing the name of the complainant – agents from Army CID then investigated the case. Miss ZC was interviewed. She alleged the accused was driving her home and, against her repeated protestations, forced down her jeans and inserted two fingers in her vagina. The agents were also informed that Miss ZC had reported allegations of abuse to high school officials in Virginia, and subsequently to local police. As the alleged offense took place in Killeen, Texas, Miss ZC and her mother were advised they would have to report the crime in Killeen. According to the accused’s First Sergeant, the accused told him that the woman had driven for twenty hours to report the incident. It appears that this is how ZC’s mother ended up at the Fort Hood CQ desk. The second interrogation took place one month after the first and was conducted by a special agent from Army CID. At the second interrogation, the accused was immediately advised of his rights both verbally and in writing. The accused again waived his rights, stating he understood them. When asked whether he wanted to make a statement and talk to the CID agent, the accused asked where he should sign. The military judge found that this second exchange consisted primarily of open ended questions, asked by an agent in a “calm voice and demeanor throughout.” Throughout the interview, “the accused was cooperative and inquisitive.” The agent did not ask about vaginal penetration, but instead asked “what story [the accused] had heard regarding the victim.” The accused responded that he heard he had “fingered her.” The accused related that he gave Miss ZC a ride to her home, that Miss ZC was worried because she was out past curfew, and that he had rubbed her thigh to reassure her. 3. The third statement The third interrogation was with a polygrapher from Army CID. The accused again was read and waived his rights. Initially the accused was talkative and inquisitive. However, when the polygrapher asked about whether the accused had vaginally penetrated Miss ZC, the accused became “overwhelmingly sad and then admitted to penetrating Miss ZC’s vagina with his finger after she had told him no.” He stated he had done this in an attempt to convince Miss ZC to have sex with him. From the record, it appears that appellant’s statements were made before the administration of the polygraph and that the “instrumentation” part of the polygraph was never conducted. 5 LEWIS—ARMY MISC 20180260 C. The military judge’s rulings The military judge suppressed all three statements by the accused. The military judge also found that law enforcement learned the identity of the victim only because of the suppressed statements of the accused. Accordingly, the military judge also suppressed the identification of Miss ZC. This ruling covered not only suppressing the accused’s statements identifying Miss ZC but also extended to the suppression of any testimony that Miss ZC may have offered. At the government’s request, the military judge reconsidered his ruling. The government presented evidence that a CID agent who was unaware of the accused’s statements was still able to identify Miss ZC using ordinary investigative techniques. On reconsideration, the military judge found that law enforcement would have inevitably discovered the identity of Miss ZC and her mother. However, the military judge again ruled that the accused’s statements from all three interrogations would be suppressed. 9 For all three interrogations, the military judge found there was no coercion. Indeed, the military judge found that the accused’s “appearance is one of willingness and voluntariness . . . .” Nonetheless, the military judge found the accused’s free will to have been overborne and the statements to have been involuntary. 9 Essentially, after the government lost the initial suppression motion, they showed what investigative steps would have been taken had they not discovered the identity of Miss ZC in the first interrogation. Whether such a post hoc endeavor is cognizable in determining an inevitable discovery question is not before us. See, e.g., United States v. Eppes, 77 M.J. 339, 347-49 (C.A.A.F. 2018) (discussing the standard for inevitable discovery). However, to the extent this issue is subject to future litigation, we would note the correct standard for suppressing live witnesses as derivative evidence. United States v. Ceccolini, 435 U.S. 268, 276 (1978) (“Witnesses are not like guns or documents which remain hidden from view until one turns over a sofa or opens a filing cabinet”); see also United States v. Campbell, 41 M.J. 177 (C.M.A. 1994); United States v. Mancini, ACM 38783, 2016 CCA LEXIS 660 (A.F. Ct. Crim. App. 7 November 2016) (unpub.); People v. Mendez, 28 N.Y. 2d 94 (1971) (analyzing whether witness testimony should be suppressed when name of witness was obtained during illegal wiretap). 6 LEWIS—ARMY MISC 20180260 The military judge’s first ruling on involuntariness rested on several facts and conclusions of law. The military judge found that the accused was 23 years old, 10 was a junior-enlisted soldier (an E-4), had six years of military service, and had a high school education. These findings are unchallenged on appeal and are well supported by the record. The military judge also found that the accused had a GT score of 92 and was of “low average or below average intelligence.” Although the government asserts these findings are clearly erroneous, we conclude they are supported by the record. However, three other aspects of the military judge’s decision warrant a more detailed discussion: First, the military judge found the accused had been diagnosed with an adjustment disorder. Although this diagnosis was made six months after the last interview, the military judge found it was a reasonable inference that an adjustment disorder diagnosis would have been applicable at the time of the interrogations. Second, the military judge found that the accused was, at all relevant times, in custody. Third, the military judge’s consideration that police misconduct in the first interrogation bled into the second and third interrogations. ANALYSIS We review a military judge’s ruling on a motion to suppress for an abuse of discretion and consider the evidence in the light most favorable to the party that prevailed at trial. United States v. Rodriguez, 60 M.J. 239, 246-47 (C.A.A.F. 2004). “A military judge abuses his discretion if his findings of fact are clearly erroneous or his conclusions of law are incorrect.” United States v. Olson, 74 M.J. 132, 134 (C.A.A.F. 2015) (citation and internal quotation marks omitted). These standards also apply to interlocutory appeals under Article 62, UCMJ. United States v. Michael, 66 M.J. 78, 80 (C.A.A.F. 2008); see also United States v. Mitchell, 76 M.J. 413, 417 (C.A.A.F. 2017). 11 10 In his first ruling, the military judge found that the accused was “a 24 year old specialist [E-4] with 6 years of service and a high school diploma.” In his second ruling, the military judge found that the accused was 23 years old. We do not believe this discrepancy is significant. 11 The overwhelming majority of cases that come before this court fall under Article (continued . . .) 7 LEWIS—ARMY MISC 20180260 A confession is inadmissible as a matter of due process if under the totality of the circumstances it was involuntarily obtained. Voluntariness turns on whether the “defendant’s will was overborne” when he gave his statement, and the test for this is whether the statement was a “product of an essentially free and unconstrained choice by its maker.” Schneckloth v. Bustamonte, 412 U.S. 218, 225-26 (1973); Culombe v. Connecticut, 367 U.S. 568, 602 (1961). A defendant’s mental condition is a relevant factor in determining whether a confession was voluntary, but “this fact does not justify a conclusion that a defendant’s mental condition, by itself and apart from its relation to official coercion, should ever dispose of the inquiry into constitutional ‘voluntariness.’” Colorado v. Connelly, 479 U.S. 157, 164 (1986). Accordingly, “coercive police activity is a necessary predicate to the finding that a confession is not ‘voluntary.’” Id. at 167. Coercive activity includes “trickery, psychological pressure, or mistreatment.” Withrow v. Williams, 507 U.S. 680, 708 (1993) (O’Connor, J., concurring in part and dissenting in part). Furthermore, whether rights warnings were given is a relevant factor but “does not . . . dispense with the voluntariness inquiry.” Dickerson v. United States, 530 U.S. 428, 444 (2000). “Determination of whether a statement is involuntary ‘requires more than a mere color-matching of cases.’ It requires careful evaluation of all the circumstances of the interrogation.” Mincey v. Arizona, 437 U.S. 385, 401 (1978) (quoting Reck v. Pate, 367 U.S. 433, 442 (1961)). Thus, we must consider “both the characteristics of the accused and the details of the interrogation.” Dickerson, 530 U.S. at 434 (quoting Schneckloth, 412 U.S. at 226). Relevant factors include the defendant’s age and education, the length of detention, whether the defendant was advised of his rights, and the nature of the questioning. Schneckloth, 412 U.S. at 226. Our superior court has provided guidance for cases such as this where there are multiple admissions by an accused and the voluntariness of a second or subsequent statement is challenged on the grounds that it is tainted by an earlier, illegally obtained statement: Where a confession is obtained at a lawful interrogation that comes after an earlier interrogation in which a confession was obtained due to actual coercion, duress, or (. . . continued) 66, UCMJ. Under that article, this court has an independent duty to review the entire record, and may make factual findings that are contrary to those of the trial court. Given our habitual practice, caution is appropriate when our role is limited to issues of pure law under Article 62, UCMJ. See, e.g., United States v. Stellato, 74 M.J. 473, 482 (C.A.A.F. 2015). 8 LEWIS—ARMY MISC 20180260 inducement, the subsequent confession is presumptively tainted as a product of the earlier one. On the other hand, where the earlier confession was “involuntary” only because the suspect had not been properly warned of his panoply of rights to silence and to counsel, the voluntariness of the second confession is determined by the totality of the circumstances. The earlier, unwarned statement is a factor in this total picture, but it does not presumptively taint the subsequent confession. United States v. Cuento, 60 M.J. 106, 108-09 (C.A.A.F. 2004). In this case, the military judge found that there was no coercion in any of the interrogations but did find implied bad faith during the first interrogation. The military judge then correctly applied the second of the two tests articulated above in Cuento. Therefore, we review whether the military judge erred as a matter of law in concluding that the accused’s statements made during the second and third interrogations were involuntary based on the totality of the circumstances. United States v. Bubonics, 45 M.J. 93, 94-95 (C.A.A.F. 1996) (voluntariness of a confession is a question of law). A. Diagnosis of the Accused The military judge found that a diagnosis of adjustment disorder made six months after the first interrogation was present at all three interrogations. The government argues this finding is clearly erroneous. We agree. But the impact of this issue is questionable, given that the military judge made a related and unchallenged finding that this diagnosis generally does not affect decision making. On 8 November 2017, a board convened pursuant to Rule for Courts-Martial (R.C.M.) 706 issued a report evaluating the accused’s mental competency and responsibility. 12 The board determined the following diagnosis existed at the time of the evaluation: “DSM-5 309.28 (F43.23) – Adjustment Disorder with Mixed Anxiety and Depressed Mood.” 12 At the time of the alleged offense (August 2015), the board diagnosed appellant with “other problems related to employment” and “relationship distress with spouse.” The board determined that the accused did not suffer from a severe mental disease or defect at the time of the alleged criminal conduct. The board also determined that the accused “is not presently suffering from a mental disease or defect rendering him unable to understand the nature of the proceedings against him or to conduct or cooperate intelligently in his defense.” 9 LEWIS—ARMY MISC 20180260 “A mental impairment is a factor to be considered in determining the voluntariness of the challenged confession only if government overreaching is also shown.” United States v. Campos, 48 M.J. 203, 207 (C.A.A.F. 1998). In the Fifth Amendment context, our voluntariness inquiry “is not concerned ‘with moral and psychological pressures to confess emanating from sources other than official coercion.’” Connelly, 479 U.S. at 170 (quoting Oregon v. Elstad, 470 U.S. 298, 305 (1985)). “The voluntariness of a waiver of this privilege has always depended on the absence of police overreaching, not on ‘free choice’ in any broader sense of the word.” Id.; see also United States v. Diaz, 59 M.J. 79 (C.A.A.F. 2003) (applying Connelly to the Article 31 context). The adjustment disorder diagnosis was raised sua sponte by the military judge. In his initial ruling, the military judge acknowledged that the diagnosis in the evaluation was made six months after the first interrogation, but found that it is “a reasonable presumption that the accused suffered adjustment disorder at the time of the [ ] interview.” The military judge then found that the diagnosis was valid during all three interrogations. During the motion to reconsider, the government presented evidence from an expert psychologist that an adjustment disorder would not affect a person’s “ability to make good decisions.” On reconsideration, the military judge found that the diagnosis received by the accused “generally does not affect decision making, but it does affect mood and the ability to cope with additional stressors.” The government argues to this court that it was clearly erroneous for the military judge to find as fact that an adjustment disorder diagnosis that was diagnosed on 8 November 2017 also existed six months earlier without any additional facts. The accused responds that we must defer to the factual findings of the trial court. Certainly, there are chronic mental health conditions for which backdating a diagnosis by six months could be a reasonable inference. There may be other conditions which, because they are acute, such a presumption may be strained or unreasonable. The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-V) defines the diagnosis of adjustment disorder as “[t]he development of emotional or behavioral symptoms in response to an identifiable stressor(s) occurring within 3 months of the onset of the stressor(s).” Am. Psychiatric Ass’n, Diagnostic and Statistical Manual of Mental Disorders (5th ed. 2013). Here, there was no evidence introduced as to what was the “identifiable stressor” that formed the basis of the diagnosis. Nor was there expert testimony explaining a basis to retroactively infer that the diagnosis applied six months earlier. 10 LEWIS—ARMY MISC 20180260 For the military judge’s inference to be correct, it would appear the stressor (and the associated symptomology) would have to predate the first interrogation. Thus, based on the record, we find insufficient evidence to support a conclusion that the accused suffered from an adjustment disorder at the time of the first interrogation. A finding of fact is clearly erroneous “when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.” United States v. Martin, 56 M.J. 97, 106 (C.A.A.F. 2001) (quoting United States v. United States Gypsum Co., 333 U.S. 364, 395 (1948)). Although we agree with the government that this finding is clearly erroneous, we note the military judge specifically found, consistent with government expert testimony, that a diagnosis of adjustment disorder generally does not affect decision making. Accordingly, even if the military judge’s finding over the accused’s diagnosis was not clearly erroneous, it would have minimal impact on our overall analysis in light of his additional finding related to decision making. B. Custody The military judge found that at all relevant times the accused was in “custody.” The military judge then used his determination that the accused was in custody when assessing the accused’s voluntariness. We find the military judge erred in his custody determination. Whether the accused was in custody “is a de novo question of law to be decided on the basis of facts found by the factfinder.” United States v. Catrett, 55 M.J. 400, 404 (C.A.A.F. 2001); see also United States v. Schake, 30 M.J. 314, 318 (C.M.A. 1990) (“This is largely a question of fact, although the ultimate conclusion is a legal one.”). That is, while “custody” may be a question of law, it is one that often turns on facts. The Supreme Court has explained that in the Miranda context “custodial interrogation” means “questioning initiated by law enforcement officers after a person has been taken into custody or otherwise deprived of his freedom of action in any significant way.” Miranda, 384 U.S. at 444; see also Mil. R. Evid. 305(b)(3); Yarborough v. Alvarado, 541 U.S. 652, 661 (2004). In determining whether someone is in custody, however, it is not enough to say that they are not immediately free to leave without delay. Persons briefly detained by police during traffic stops or Terry stops 13 are certainly not free to leave, 13 See Terry v. Ohio, 392 U.S. 1 (1968). 11 LEWIS—ARMY MISC 20180260 but nor are they “in custody” in the Fifth Amendment sense because their lack of freedom of action does not rise to the level associated with a formal arrest. See, e.g., Berkemer v. McCarty, 468 U.S. 420, 438-40 (1984) (Terry stop is not Miranda custody). In making a custody determination, courts consider: (1) whether the person appeared for questioning voluntarily; (2) the location and atmosphere of the place in which questioning occurred; and (3) the length of the questioning. United States v. Evans, 75 M.J. 302, 306 (C.A.A.F. 2016). The military judge found that each time the accused was interrogated he was escorted by a superior from his unit to CID. The military judge further found that in accordance with standard operating procedure the agents searched him for officer safety, and required him to place his personal belongings in a locker. The military judge found that during the second and third interrogations the accused’s “appearance is one of willingness and voluntariness.” (Although, obviously, the military judge did not find the accused’s statements to actually be voluntary). During the second interrogation, the questioning was conducted in a calm voice and calm demeanor. The interrogations were “not lengthy” and did not involve “coercive tactics” or “inhumane conditions.” These factual findings have clear support in the record. 14 After summarizing his findings of fact, the military judge concluded as follows: “While [the accused] may have been allowed to leave if he insisted, a reasonable person of the accused’s age, experience, education, diagnoses, and military service would not have felt he was at liberty to terminate the interrogation and leave.” We discuss several aspects of the ruling. First, the military judge appears to have deviated from the custody test required by Miranda for Fifth Amendment custody determinations. The Supreme Court has rejected “place[ing] upon the police the burden of anticipating the frailties or idiosyncrasies of every person whom they question.” Alvarado, 541 U.S. at 662 (citing Berkemer, 468 U.S. at 442) (quoting People v. P., 21 N.Y.2d 1, 9-10 (1967)). 14 When assembling the record for interlocutory appeal, the government included several “allied papers” that contained matter that was never submitted to the military judge. For example, police reports indicate that the third interrogation (a planned polygraph) was scheduled in advance by prior arrangement with the accused. However, if this information was never presented to the military judge, it is legally irrelevant to our determination as to whether the military judge erred in his ruling. 12 LEWIS—ARMY MISC 20180260 The Court explained that the objective test for making a custody determination is different than the test for determining whether a suspect’s will has been overborne. Id. at 666-68. The suspect’s age and other life experiences are relevant to the latter, but not the former. Id. at 667-68. The Court explained: There is an important conceptual difference between the Miranda custody test and the line of cases from other contexts considering age and experience. The Miranda custody inquiry is an objective test . . . . The objective test furthers the clarity of Miranda’s rule, ensuring that the police do not need to make guesses as to the circumstances at issue before deciding how they may interrogate the suspect. ... [T]he objective Miranda custody inquiry could reasonably be viewed as different from doctrinal tests that depend on the actual mindset of a particular suspect, where we do consider a suspect’s age and experience. For example, the voluntariness of a statement is often said to depend on whether the defendant’s will was overborne, a question that logically can depend on the characteristics of the accused. The characteristics of the accused can include the suspect’s age, education, and intelligence, as well as a suspect’s prior experience with law enforcement. Id. (cleaned up by omitting internal citations and quotations). Although the Court would later allow consideration of the fact that the suspect is a young child, J.D.B. v. North Carolina, 564 U.S. 261 (2011), the Court reemphasized the objective nature of the test, and that a custody determination is not to be made on factors “unknowable” to the police at the time: By limiting analysis to the objective circumstances of the interrogation, and asking how a reasonable person in the suspect’s position would understand his freedom to terminate questioning and leave, the objective test avoids burdening police with the task of anticipating the idiosyncrasies of every individual suspect and divining how those particular traits affect each person’s subjective state of mind. J.D.B., 564 U.S. at 272. 13 LEWIS—ARMY MISC 20180260 Thus, we find the trial court erred by considering the accused’s age, education, and military service in making a custody determination. The court strayed further by including the accused’s diagnoses – diagnoses that would not be knowable by any person until six months after the first interrogation – in making a custody determination. The requirements of Miranda are triggered by custody equivalent to a formal arrest, Berkemer, 468 U.S. at 440, and the police cannot be expected to follow their Miranda obligations if the determination of custody includes considering a diagnosis that was made six months after the interrogation. Second, the military judge did not appear to distinguish between the three interrogations and how an objective determination of custody might change at each instance. There were significant breaks in time between the interrogations. After each interrogation, the accused was free to leave. 15 A conclusion that a person reasonably believed he was not free to leave becomes less and less tenable after each prior interrogation has ended and the accused was left to go about his business. Thus it may be that the custody question in the third interrogation might be answered differently than the first. Third, the military judge appears to have applied his custody determination to both his analysis under the Fifth Amendment and his analysis under Article 31(b). The military judge’s initial ruling focused on the Fifth Amendment, while the ruling in reconsideration was based on a violation of the accused’s Article 31(b) rights. In articulating that he found the accused’s statements from the second and third interrogations involuntary under Article 31, UCMJ, the military judge included his determination that the accused “was in custody and subjected to custodial interrogation.” Whether an interrogation was “custodial” is not legally relevant to a determination of voluntariness under Article 31, UCMJ. While the underlying facts may be highly relevant to both conclusions (e.g. if the accused was handcuffed or threatened), that an interrogation was custodial (as a question of law) is not the relevant inquiry for an Article 31, UCMJ, analysis. Accordingly, we cannot separate out any error from the military judge’s Fifth Amendment analysis from his analysis under Article 31. Fourth, in making custody determinations, other courts have specifically considered “whether the officers used coercive tactics such as hostile tones of voice, the display of weapons, or physical restraint of the suspect’s movements.” United States v. Willaman, 437 F.3d 354, 359-60 (3d. Cir. 2006); see also Evans, 75 M.J. at 306 (courts should consider “atmosphere” of interrogation). In analyzing the 15 That the suspect was free to leave at the end of questioning was described by the Supreme Court in Alvarado as an “objective fact[] [that is] consistent with an interrogation environment in which a reasonable person would have felt free to terminate the interview and leave.” 541 U.S. at 664-65. 14 LEWIS—ARMY MISC 20180260 question of custody, the military judge did not appear to weigh his specific factual findings: (1) that the interrogations were conducted without the use of harsh or coercive tactics, and (2) that the second interrogation was conducted with a calm voice and demeanor. Although the military judge did not claim otherwise, we note that no court has created a per se rule that all military law enforcement interrogations are custodial interrogations under the Fifth Amendment. The circumstances of the accused’s interrogations are present at many, if not most, interrogations by law enforcement. Indeed, the military judge found that the pat down and removal of the accused’s phone was for officer safety and consistent with standard operating procedure. While the military context may be relevant to a custody analysis, 16 each case must be decided by the facts. Not every soldier escorted to law enforcement, or who is processed for an interrogation, will be in custody for purposes of the Fifth Amendment. Soldiers, for example, are often escorted to appointments with defense counsel, or finance, or for other reasons. And while those are not wholly analogous situations with an interrogation by law enforcement (at all), we would still nonetheless never describe such an escorted soldier as having limitations on freedom similar to those of a formal arrest. To be sure, a soldier told to report to a superior’s office is not free to walk out the door in the sense that they are not free to ignore superior military authority. Nor is a civilian detained by civilian law enforcement during a Terry stop free to ignore police authority. Berkemer, 468 U.S. at 438-40. Whether either case results in custody will depend on the surrounding circumstances. Although we find some gaps in the military judge’s custody determination, our main concern is that the legal determination of whether a suspect is in custody is not particularly relevant to the question at issue: voluntariness. Although the same facts (e.g. if the accused was handcuffed) will be relevant to both legal questions, that the accused was in custody is not a prerequisite to finding a statement involuntary. C. Voluntariness: Was the cat out of the bag? As we have previously stated, the government has not appealed the military judge’s suppression of the first interrogation. This court is jurisdictionally barred from revisiting that part of the military judge’s ruling, and no part of this opinion 16 See, e.g., Mitchell, 76 M.J. at 417-18. In Mitchell, the CAAF found the accused was in custody when his presence in his commander’s office was involuntary, the location and atmosphere suggested custody, and the accused was surrounded by law enforcement officers who were backed by the authority of the accused’s commander. 15 LEWIS—ARMY MISC 20180260 should be confused with any attempt to do so. However, the decision to suppress the second and third interrogations relied, in part, on the degree to which errors in the conduct of the first interrogation weighed on the question of whether the accused’s will was overborne in the second and third interrogations. It is through this lens in which we revisit the first interrogation. Consistent with case law, the military judge considered the degree to which appellant’s admissions during the first interrogation weighed on the voluntariness of the second and third interrogations. See Cuento, 60 M.J. at 108-10. This is sometimes referred to as having let the “cat out of the bag.” Missouri v. Seibert, 542 U.S. 600, 615 (2004). The government attacks two parts of the military judge’s reasoning. 1. The “spontaneous” statement After the first interrogation, an entry was made in the investigatory file that labeled the accused’s pre-warning statements as “spontaneous.” The military judge correctly found this label to be inaccurate – explicitly finding that the admissions were made in response to an unwarned interrogation. In suppressing the third interrogation, the military judge relied on the fact that the agent had prepared for the interrogation by “reviewing a summary of the statements in the case, which included a misleading reference to a ‘spontaneous’ statement.” All parties agree that in preparing for the third interrogation, which was to be a polygraph, the polygrapher reviewed a summary of the statements that had been made in the case. However, the summary did not include a reference to the spontaneous statement. The parties appear to agree that the military judge’s reliance was error. The government asks this court to find the military judge’s finding to be clearly erroneous as it is clearly contradicted by the record and otherwise unsupported. By contrast, the defense points to the fact that the military judge’s reference to the “spontaneous statement” was not found in the “facts” portion of the judge’s ruling. Instead, it is located in the “analysis” part of the ruling. At least here, we do not think it important where in the military judge’s ruling the factual finding was located. What matters more is whether the military judge relied on the fact in reaching a legal conclusion. An erroneous factual conclusion (e.g. that the 1st of April was a Tuesday) is relevant only if it affects the legal question at issue. And here, it appears that the military judge relied on an erroneous 16 LEWIS—ARMY MISC 20180260 fact in his analysis of whether to suppress the accused’s statements arising from the third interrogation. 2. The use of prior statements as the “basis of denials” In suppressing the third interrogation, the military judge found that the polygrapher “knew and used the accused’s prior statements as a basis for his denials for the polygraph.” The polygraph examination was never conducted. Thus it is unclear, and the parties dispute, what was meant by the military judge. In this Article 62 appeal, we are required to assess the evidence in a light most favorable to the prevailing party. See United States v. Pugh, 77 M.J. 1, 3 (C.A.A.F. 2017). Thus here, when a record can reasonably be understood two different ways, we assume a meaning that is favorable to the accused unless we direct a remand to clarify the issue. However, regardless of what the military judge was referring to, we find that the military judge was focused on the incorrect question. When assessing whether the accused’s free will was overborne, the question is not what the agent read in preparing for an interrogation. Rather, the question is whether he used that knowledge in a manner that would affect the voluntariness of the accused’s rights waiver and admissions. Thus we see the military judge’s analysis to be incorrectly focused on what the agents knew. An agent’s knowledge of an accused’s earlier, illegally obtained statement is relevant only to the extent that it is accompanied by a finding or an inference that the knowledge was used in a manner that undermined the voluntariness of the accused’s decision to repeat the admission. Indeed, the military judge’s decision to suppress the second interrogation contained just such a finding. In analyzing the second interrogation, the military judge made specific findings that the agent had referenced the first interrogation when informing the accused of his rights. The record supports the military judge’s conclusion. Indeed, in our listening, the agent tells the accused: I know you were in here earlier, you talked to us, so I’m going to ask some additional questions -- before we do it . . . we have to go through your rights advisement because we want you to be aware of your rights. 17 The agent then thanked the accused for agreeing to come back to CID. 17 This quoted language is taken from the record, not the military judge’s findings of fact. The emphasis is ours. 17 LEWIS—ARMY MISC 20180260 We understand the military judge’s ruling to be that by referencing the accused’s statements in the first interrogation, and framing the second interrogation as consisting of “additional” questions, the agent linked the two interrogations and risked that the accused’s rights waiver in the second interrogation would not be viewed as a voluntary act independent of the first interrogation. That is, that the cat was already out of the bag. The danger was greater as the side comment was made just prior to the accused’s rights advisement and decision to waive his rights. While a different military judge may have come to a different conclusion, the military judge’s interpretation here was a reasonable one. Indeed, it is for this reason that we leave the military judge’s suppression of the statements made during the second interrogation intact. D. There are substantive differences between the second and third interrogations In determining whether the military judge abused his discretion in suppressing the second and third interrogations, we are required to consider the totality of the circumstances. Both interrogations shared some commonalities. 1. Commonalities between the second and third interrogations During both the second and third interrogations, the accused was 23 years old, an E-4 with six years’ experience in the military, and of low average intelligence. In neither the second or third interrogation was the accused given a cleansing statement. Additionally, in both instances, the accused was escorted to the interrogation by a superior before being processed for an interrogation. On balance this evidence provides some weight in favor of finding his will was overborne. On the other hand, in both interrogations the accused was properly advised, both verbally and in writing, of his rights under Article 31, UCMJ, and the Fifth Amendment. The military judge found that after the accused stated he understood his rights, he then waived them. Although the military judge used slightly different language in describing the manner in which the interrogations were conducted, the military judge found that neither of the interrogations were of long duration or involved coercion. All three interrogations took place on different days separated by weeks. See Seibert, 542 U.S. at 615 (the timing between the two interrogations is a consideration when determining voluntariness). The three interrogations were conducted by three different members of law enforcement. Id. (finding that the “continuity of the police personnel” factored in favor of finding the suspect’s will being overborne). The military judge found that in both interrogations the accused was “inquisitive and appeared to be acting rationally and voluntarily.” 18 LEWIS—ARMY MISC 20180260 2. Differences between the second and third interrogations The second and third interrogations differed in several respects. First, the accused’s admissions in the third interrogation far exceeded those made during the first and second interrogation. See id. (The Court considered the degree to which the first and second statements had “overlapping content.”). In the second interrogation, the essential admission by the accused was that he had rubbed Miss ZC’s leg to reassure her, and this had made her uncomfortable. In the third interrogation, appellant repeated that he had rubbed her leg, but also admitted to putting his finger inside her vagina. In his sworn statement, he described it as follows: We talked a little while in the car. While in the car I rubbed her leg. I then pulled her pants down. I put my finger inside her vagina and asked if she was uncomfortable. She told me she was and I stopped. She then pulled up her pants and left the car. I drove away. The accused further admitted that he did these acts in an attempt to convince Miss ZC to have sex with him. Thus, the “cat” was not wholly out of the bag at the time of the third interrogation. Appellant’s statements in the first two interrogations that he rubbed Miss ZC’s leg for reassurance placed him in the car with Miss ZC. They were not confessions. See Mil. R. Evid. 304(a)(1)(B-C). Appellant’s statement in the third interrogation that he placed his fingers in the vagina of a girl he knew was underage for the purpose of convincing her to have sex with him was a confession. In the same light, in Cuento, our superior court looked differently at an interrogation that referenced unwarned statements admitting to an “accidental touching” and those that referenced substantive admissions. 60 M.J. at 108-10. When assessing voluntariness and the weight to give the failure to provide a cleansing statement, we think it is relevant to consider the degree to which appellant repeated or expanded on the prior statement. See Seibert, 542 U.S. 600. Here, there is not much “overlap” between the two interrogations. Second, as a matter of logic, by the time of the third interrogation the accused had more familiarity with the military justice system than he had during the first and second. Appellant cites to Bubonics for the proposition that we should consider whether an accused has “been involved with military justice before the night of his apprehension and interrogation” when assessing the totality of the circumstances. 45 M.J. at 96. However, a factor that may weigh in favor of finding the first interrogation involuntary becomes less powerful when the accused is being advised of his rights for the third time over the course of many weeks. 19 LEWIS—ARMY MISC 20180260 Third, and again as a matter of logic, the third interrogation was farther removed from the first interrogation than the second. Whereas in Seibert the second interrogation took place only 15 to 20 minutes after the first, here the third interrogation took place about 21 days after the second interrogation and 52 days after the first interrogation. 18 Although not exactly the same issue of law, the Supreme Court has held that the police may re-initiate contact with a suspect who has invoked his right to counsel after only a 14-day break. Maryland v. Shatzer, 559 U.S. 98 (2010). The passage of time, and the associated attenuation of any taint from the first interrogation, weighs against finding the accused’s statements in the third interrogation were involuntary. Fourth, and perhaps most importantly, the accused himself was asked questions about factors that would bear light on the voluntariness of the third interrogation. A sworn statement, signed by the accused, was attached to the defense motion to suppress and was considered by the military judge. We quote it at length: Q. Was that the only time you had sexual contact with her? A. Yes. Q. How many times did she tell you to stop that night? A. Once when my hand was on the outside of her clothing on her thigh and again when my finger was inside her vagina. Q. You lied in your previous statement, Why? A. I was under a lot of pressure. Q. Why are you telling the truth now? A. I want the case over. I want to do the right thing to make this right so it will go away. Q. How do you feel now that you have told the truth? 18 The first interrogation was on 15 May 2017. The second interrogation was on 15 June 2017. The third interrogation was on 6 July 2017. 20 LEWIS—ARMY MISC 20180260 A. I feel the same. I just want this to go away and I know being honest will be the fastest way to make this go away. Perhaps concerned that the accused had a misperception about the effect of his admissions, the agent immediately asked a series of follow up questions: Q. Did anyone promise you anything to say the things you have said in this statement? A. Noe [sic] Q. Did anyone tell you that you would get in less trouble for saying the things you have said in this statement? A. No. Q. Do you understand that I am not the person who makes a decision as to what happens in this case and to you? A. Yes. Q. How were you treated during this interview? A. Great. Q. Were you allowed to take breaks and get something to eat or drink any time you wanted? A. Yes. Q. How much sleep did you get last night? A. 4-6 hours which is more than usual. Q. When was the last time you had anything to eat? A. Last night. 19 Q. Do you feel that you were deprived of anything prior to and/or during this interview? 19 The accused arrived at CID at 0800 in the morning. 21 LEWIS—ARMY MISC 20180260 A. No. Q. What is your GT score? A. 100. 20 Q. Did you fully understand what is going on today during this interview? A. Yes. Q. Is there anything that would prohibit you from completely understanding what we talked about today? A. No. Q. Are you admitting to the things you have admitted in this statement for any other reason than they are the truth? A. No. The accused did not testify or explain how his will was overborne. While the accused is not required to testify at a suppression motion, and we give no weight to his decision not to testify, the effect was to leave his sworn statement about the voluntariness of his admissions unrebutted. E. Conclusion In his ruling on reconsideration, the military judge correctly stated that this case is fact dependent. Having found that the military judge applied the wrong law in determining that the accused was subjected to a custodial interrogation, that the error was incorporated into the military judge’s voluntariness assessments under both the Fifth Amendment and Article 31, because we find one fact to have been unsupported by the record, and based on the totality of the circumstances, we set aside the military judge’s ruling suppressing the third interrogation. To be clear, the ultimate question does not turn on whether the accused was in custody or not. Rather, the question is whether the accused’s “will” was “overborne and his capacity for self-determination critically impaired” as a result of agent conduct. Schneckloth, 412 U.S. at 225-26 (quoting Culombe, 367 U.S. at 602). We 20 Based on other evidence, the military judge would find as fact that the accused’s GT score was 92. 22 LEWIS—ARMY MISC 20180260 conclude that under the totality of the circumstances the accused’s statements made during the third interrogation were voluntary. While we acknowledge that some factors, such as the lack of a cleansing statement, weigh in favor of finding that the accused was more vulnerable to having his will overborne, we find little evidence in the record to support a conclusion that the accused’s will was actually overborne. 21 CONCLUSION The government’s appeal pursuant to Article 62, UCMJ, is GRANTED in part and DENIED in part. The government’s appeal is denied as to the suppression of the accused’s second interrogation. The government’s appeal is granted as to the suppression of the accused’s third interrogation. The record will be returned to the military judge for action not inconsistent with this opinion and R.C.M. 908(c)(3). Judge SALUSSOLIA and Judge ALDYKIEWICZ concur. FOR THE FOR THE COURT: COURT: MALCOLM MALCOLM H. H. SQUIRES, SQUIRES, JR. JR. Clerk of Court Clerk of Court 21 Where the Supreme Court has found statements involuntary, the circumstances surrounding the interrogation have been much worse than those here. In Mincey, the defendant was suffering “unbearable” pain from a gunshot wound while unable to speak because of a tube in his mouth; he also could not provide coherent answers to questions, and he asked for a lawyer repeatedly over the course of a four-hour interrogation. 437 U.S. at 396-401. In Blackburn v. Alabama, the defendant endured an eight to nine-hour interrogation in a small room surrounded by three police officers and “was insane and incompetent at the time he allegedly confessed.” 361 U.S. 199, 204, 207 (1960). The accused’s circumstances do not come close to those of Mincey and Blackburn. 23