NOTICE: This opinion is subject to motions for rehearing under Rule 22 as well as formal revision before publication in the New Hampshire Reports. Readers are requested to notify the Reporter, Supreme Court of New Hampshire, One Charles Doe Drive, Concord, New Hampshire 03301, of any editorial errors in order that corrections may be made before the opinion goes to press. Errors may be reported by e-mail at the following address: reporter@courts.state.nh.us. Opinions are available on the Internet by 9:00 a.m. on the morning of their release. The direct address of the court’s home page is: http://www.courts.state.nh.us/supreme.
THE SUPREME COURT OF NEW HAMPSHIRE
___________________________
Hillsborough-southern judicial district No. 2018-0575
THE STATE OF NEW HAMPSHIRE
v.
DOMINIC CARRIER
Argued: November 6, 2019 Opinion Issued: April 7, 2020
Gordon J. MacDonald, attorney general (Sean R. Locke, assistant attorney general, on the brief and orally), for the State.
Stephanie Hausman, deputy chief appellate defender, of Concord, on the brief and orally, for the defendant.
HANTZ MARCONI, J. The State brings this appeal pursuant to RSA 606:10, II(a), arguing that the Superior Court (Colburn, J.) erred in suppressing two sets of recorded statements made by the defendant, Dominic Carrier. See RSA 606:10, II (2001). The trial court ruled that the defendant was subject to custodial interrogation at the time he gave the first set of statements, and, because he was not given the warnings required by Miranda v. Arizona, 384 U.S. 436 (1966), those statements were obtained in violation of his right against self-incrimination. The court suppressed the second set of statements because it found that the State did not prove beyond a reasonable doubt that the defendant gave them voluntarily. We affirm and remand. I
The trial court found or the record reflects the following facts. As of November 19, 2017, the defendant, his mother, the alleged victim, and her father lived together in a Nashua apartment. The defendant’s mother and the alleged victim’s father were engaged to be married. The apartment is located in a multi-unit building with a partially enclosed front porch. At 5:00 a.m. on November 20, the defendant left for work. After he left, the alleged victim (AV), who was thirteen years old at the time, told her father that the defendant had entered her bedroom during the early hours of the morning and touched her genitals while she slept.
AV’s father called the police, and Officer Kekejian of the Nashua Police Department arrived at the residence around 7:00 a.m. Kekejian wore a full uniform, with badge and gun visible. AV’s father relayed his daughter’s allegation to Kekejian, who then contacted additional officers. Kekejian was directed to remain at the residence with AV, her father, and the defendant’s mother until the Child Advocacy Center (CAC) could conduct AV’s forensic interview later that day.
Kekejian and the family waited in the living room of the apartment. At 9:09 a.m., the defendant returned. He walked through the front door, which entered into the living room, without knocking. Kekejian immediately ordered the defendant to leave, then followed the defendant outside to the front porch and closed the door behind him, leaving the family inside. Kekejian stood in front of the door, preventing the defendant from reentering his home. He told the defendant that the police were “investigating a matter” and that the apartment was “being held as a scene.” Kekejian also pat-frisked the defendant. In addition, when the defendant began using his phone, Kekejian took it from him, gave no explanation for why he did so, and never returned it.
Kekejian proceeded to question the defendant about AV’s allegations. He asked the defendant “what he had done” that morning and whether he had entered AV’s bedroom. He also asked about the defendant’s living arrangements at the apartment. Kekejian did not tell the defendant that he could refuse to answer his questions, that he was not under arrest, or that he was free to leave. It was cold and windy that day, but the defendant did not have a coat.
This initial interaction on the porch lasted approximately ten minutes, until Officer Ciszek arrived. Ciszek arrived in a fully marked police cruiser and wore a police uniform. After Ciszek arrived, Kekejian called Detective Hallam. Hallam initially told Kekejian that he would “get in contact with” the defendant at a later time. Kekejian provided the defendant with Hallam’s contact information. As Kekejian and the defendant were discussing where the defendant would go in light of the weather and the fact that he would not be
2 allowed to return to the home, Hallam called Kekejian back. Hallam instructed Kekejian to ask the defendant to remain at the residence for an interview. The defendant said he would wait for Hallam; however, the defendant was not told that he was not required to wait for Hallam or that he could refuse to speak with him. Kekejian, Ciszek, and the defendant remained on the front porch until approximately 10:00 a.m., when Hallam arrived with Detective McIver. During that time, Kekejian again questioned the defendant about his whereabouts that morning while also making “small talk” about other topics, such as the defendant’s employment and cars. There is no evidence that either Kekejian or Ciszek left the porch during the time that the three men waited for Hallam and McIver.
Hallam subsequently arrived with McIver in an unmarked police car, a Chevrolet Impala, and parked across the street from the apartment. The Impala had government license plates and a police radio, but was otherwise similar to a non-police vehicle. Hallam and McIver wore dress shirts and dress pants, but their badges and guns were visible. Upon their arrival, Hallam and McIver walked to the front porch and asked the defendant to speak with them in their car. The detectives accompanied the defendant to the car and all three got inside, closing the doors behind them. Hallam sat in the driver’s seat, McIver sat in back, and the defendant sat in the front passenger’s seat.
Hallam turned on an audio recording device shortly after entering the car. The defendant stated he was “fine” with Hallam doing so. Hallam asked the defendant if he was “giving a statement voluntarily” and stated that he didn’t want the defendant to feel “forced” to speak with him. Hallam also demonstrated to the defendant that the doors to the car were not locked and said that he should “feel free” to leave. However, the defendant and Hallam were talking over each other when Hallam made all of these statements.
The detectives began questioning the defendant about AV’s allegations. The defendant denied that he had ever left his bedroom the previous night. However, over the course of the approximately one-hour interrogation, the detectives repeatedly accused the defendant of lying and of being “deceptive.” They told the defendant that they were not interested in hearing the defendant’s denials and that all they wanted to hear from the defendant were explanations for his conduct. The detectives told the defendant that they “wouldn’t be here” investigating AV’s allegations if they were not true, that AV had no reason to lie, and that they believed the defendant had touched AV’s vagina. Despite the nature of the detectives’ accusations, the trial court found that the overall tone of the interrogation was generally polite.1
1 The trial court based its findings regarding the interrogation in the car upon the audio recording. On appeal, the State has not provided us with this audio recording.
3 As the interrogation inside the Impala progressed and the detectives repeatedly accused the defendant of lying and of sexually assaulting AV, the defendant made incriminating statements. The doors to the car remained closed throughout this encounter. Kekejian and Ciszek were within viewing distance the entire time. The defendant was not given or offered a break at any point.
After the interrogation concluded, the defendant apparently used Hallam’s phone to call his father. The defendant arranged for his father to pick him up at a nearby convenience store. The defendant told the detectives he would be staying with his father and gave them his father’s address. The detectives then drove the defendant to the convenience store.
After AV’s CAC interview later that day, a warrant was issued for the defendant’s arrest. Hallam arrested the defendant at his father’s home at 8:20 p.m., and the defendant was brought to the Nashua police station. Hallam approached the defendant in the booking area around 10:00 p.m. and asked the defendant to speak with him. The defendant agreed and was brought upstairs to an interview room. Hallam read the defendant his Miranda rights, and the defendant said he would waive them. Less than twenty seconds after giving the defendant Miranda warnings, however, Hallam invoked the defendant’s statements in the Impala that morning, telling the defendant that there were “inconsistencies” in his first set of statements and that this time he wanted to “skip all that B.S. that we were doing out there.” Hallam continued to bring up the defendant’s first set of statements throughout this second interrogation; the defendant made additional incriminating statements.
The defendant was charged with one count of aggravated felonious sexual assault (AFSA). See RSA 632-A:2 (2016) (amended 2017, 2018). He moved to suppress the recorded statements he made inside the Impala as well as his subsequent recorded statements at the police station. The trial court held a hearing, at which it heard testimony from Kekejian and Hallam. The recorded statements themselves, as well as transcripts of each set of statements, were submitted to the trial court by agreement. As to the Impala statements, the court ruled that the defendant was in “custody” for Miranda purposes when, approximately sixteen minutes into the recording, Detective Hallam repeatedly accused him of lying. As to the statements at the police station, the court found that the State failed to meet its burden of proving beyond a reasonable doubt that the defendant gave those statements voluntarily. Accordingly, the court granted the defendant’s motion. The State filed a motion to reconsider, to which the defendant objected. The court denied the State’s motion, and this appeal followed.
4 II
We begin with the Miranda issue. The State argues that the defendant was not in custody at any point before he was formally arrested at his father’s home; thus he was not entitled to Miranda warnings before or at any point during the questioning in the Impala. The defendant contends that the trial court correctly ruled that he was in custody because a reasonable person in his position would have felt constraints on his freedom consistent with formal arrest. The parties raise their arguments under both the State and Federal Constitutions. We first address them under the State Constitution and rely upon federal law only to aid our analysis. State v. Ball, 124 N.H. 226, 231-33 (1983).
Before a defendant’s responses made during a custodial interrogation may be used as evidence against him, the State must prove, beyond a reasonable doubt, that it did not violate his constitutional rights under Miranda. State v. Marin, 172 N.H. 154, 159 (2019). Correspondingly, Miranda warnings are not required if the defendant was not subject to a custodial interrogation. State v. Carroll, 138 N.H. 687, 696 (1994); see State v. Sachdev, 171 N.H. 539, 548 (2018) (“[T]wo conditions must be met before Miranda warnings are required: (1) the suspect must be ‘in custody’; and (2) he must be subject to ‘interrogation.’”). Here, there is no dispute that Hallam and McIver interrogated the defendant in the Impala. The only issue before us is whether he was in custody when they did so.
Custody entitling a defendant to Miranda protections requires formal arrest or restraint on freedom of movement to the degree associated with formal arrest. In re B.C., 167 N.H. 338, 342 (2015); accord State v. Jennings, 155 N.H. 768, 772 (2007). In the absence of formal arrest, the court must determine whether a suspect’s freedom of movement was sufficiently curtailed by considering how a reasonable person in the suspect’s position would have understood the situation. Marin, 172 N.H. at 159; see also Stansbury v. California, 511 U.S. 318, 323 (1994) (per curiam) (“[T]he initial determination of custody depends on the objective circumstances of the interrogation, not on the subjective views harbored by either the interrogating officers or the person being questioned.”). To determine whether a reasonable person in the suspect’s position would believe himself to be in custody, the trial court should consider the totality of the circumstances of the encounter, including but not limited to such factors as the number of officers present, the degree to which the suspect was physically restrained, the interview’s duration and character, and the suspect’s familiarity with his surroundings. Marin, 172 N.H. at 159- 60. Custody analyses, however, are rarely based upon a static set of circumstances. Id. at 160; accord State v. McKenna, 166 N.H. 671, 677 (2014). Interrogations are fluid: what may begin as noncustodial questioning may evolve over time into custodial questioning. McKenna, 166 N.H. at 677.
5 The trial court’s findings regarding “‘the circumstances surrounding the interrogation’” are entitled to the deference we normally accord factual findings. Marin, 172 N.H. at 160 (quoting J.D.B. v. North Carolina, 564 U.S. 261, 270 (2011)); see also State v. Ford, 144 N.H. 57, 62 (1999) (explaining that “the trial court’s findings of historical facts relevant to the question of custody, that is, its determinations of ‘what happened,’” are entitled to the deference normally given to factual findings). However, because the ultimate determination of custody requires an application of a legal standard to historical facts, that determination is not merely a factual question, but a mixed question of law and fact. Marin, 172 N.H. at 160. “Unlike voluntariness, custody is a law- dominated mixed question in which ‘the crucial question entails an evaluation made after determination of the historical facts: if encountered by a “reasonable person,” would the identified circumstances add up to custody as defined in Miranda?’” Ford, 144 N.H. at 62-63 (brackets omitted) (quoting Thompson v. Keohane, 516 U.S. 99, 113 (1995)). Accordingly, although we will not overturn the trial court’s findings of historical facts unless they are contrary to the manifest weight of the evidence, we review the ultimate determination of custody de novo. In re E.G., 171 N.H. 223, 230 (2018).
In addition to challenging the trial court’s custody ruling, the State challenges certain factual findings made by the trial court. The trial court stated in its order that it “was left with the impression that Officer Kekejian was essentially ‘guarding’ the defendant on the porch.” The State asserts that this “conclusion[] either stretched reasonable inference to its limit or [was] wholly unsupported by the record” because “[t]he defendant never elicited this fact or any fact that implied it . . . during cross-examination.” We question whether the trial court’s conclusion that Kekejian was essentially “guarding” the defendant is a finding of historical fact rather than an application of facts to the legal question of whether a reasonable person in the defendant’s position would have felt restraints on his freedom of movement consistent with formal arrest. See McKenna, 166 N.H. at 677-79. Even assuming, however, that the State correctly characterizes this as a factual finding, it is not against the manifest weight of the evidence. See E.G., 171 N.H. at 230. The trial court’s “impression that Officer Kekejian was essentially ‘guarding’ the defendant” is supported by evidence in the record that, among other things, Kekejian ordered the defendant to leave his home, followed him outside, blocked the door to the home to prevent him from reentering, and did not leave the defendant’s side until Hallam and McIver arrived.
The State also takes issue with the court’s observation, in a footnote, that Kekejian did not disclose in his report that he took the defendant’s phone or pat-frisked him. The court noted that Kekejian acknowledged these omissions from his report only on cross-examination. The court’s observation is plainly supported by Kekejian’s testimony at the suppression hearing; our review of the transcript confirms that Kekejian did not acknowledge the omission of this information from his report until cross-examination. We have
6 reviewed the State’s remaining assertions regarding the trial court’s factual findings and find them to be without merit. See Vogel v. Vogel, 137 N.H. 321, 322 (1993). Thus, we turn to our custody analysis.
There are numerous circumstances surrounding the defendant’s encounter with the police that weigh in favor of a finding that he was in custody at some point during the first recorded interrogation. For example, “the degree to which the defendant’s movements were restrained” suggests he was in custody. Marin, 172 N.H. at 162; see Jennings, 155 N.H. at 773-74. “The lack of handcuffs or similar devices is not dispositive, . . . effective restrictions on a defendant’s movement” need not be derived from such devices, but instead “can be a product of verbal, psychological, or situational restraint.” Marin, 172 N.H. at 162 (quotation and brackets omitted). “When a defendant is ‘not permitted freedom of movement within [his] own home,’ and is escorted everywhere by the police, it weighs in favor of a finding of custody.” Id. (quoting United States v. Mittel-Carey, 493 F.3d 36, 40 (1st Cir. 2007)); see McKenna, 166 N.H. at 678-79.
Here, “although the defendant may not have been placed in handcuffs or any similar device, he was restrained from early on in the encounter.” Jennings, 155 N.H. at 773. Kekejian ordered the defendant to leave as soon as he walked in the front door to his home, immediately followed the defendant outside to the partially enclosed porch, shut the door behind him, and positioned himself in front of it to block the defendant’s reentry. Kekejian remained by the defendant’s side the entire time he was on the porch, and was joined by Ciszek, who occupied the porch almost the entire time the defendant did. Both officers remained with the defendant in the cold until Hallam and McIver arrived. Hallam and McIver then escorted the defendant to their vehicle, in view of Kekejian and Ciszek. We agree with the trial court that these actions “‘would have conveyed to a reasonable person . . . that the officers did not intend to allow the defendant to leave their sight.’” (Quoting McKenna, 166 N.H. at 678.); see Stansbury, 511 U.S. at 324-25.
In addition, Kekejian pat-frisked the defendant and seized his phone. These factors weigh in favor of custody. See Jennings, 155 N.H. at 773-74; see also Tankleff v. Senkowski, 135 F.3d 235, 244 (2d Cir. 1998). The seizure of the defendant’s phone is an especially weighty factor: “[b]y denying the defendant access to his . . . phone[], the police effectively ensured that he was dependent upon them for any . . . communication with the outside world.” Jennings, 155 N.H. at 774; accord Marin, 172 N.H. at 163-64; see State v. McMillan, 55 P.3d 537, 539-40 (Or. Ct. App. 2002) (officer’s actions of preventing defendant from using his cell phone would have reasonably suggested to defendant that “the officers could exert control over his means of communication”).
7 We next consider the “duration and character” of the interrogation. Marin, 172 N.H. at 160 (quotation omitted). “[T]he length of questioning can be a relatively undeterminative factor in the analysis of custody,” at least when police do not engage in “the ‘marathon’ routine of questioning a suspect,” with which “Miranda was most obviously concerned.”2 E.G., 171 N.H. at 237 (quotation omitted). Here, the defendant was questioned about AV’s allegations over a period of approximately two hours. See Jennings, 155 N.H. at 774 (concluding that an interview length of “nearly two hours” weighed in favor of custody); see also, e.g., Mittel-Carey, 493 F.3d at 40 (interrogation length of ninety minutes to two hours weighed in favor of custody). Kekejian began questioning the defendant about the sexual assault allegation shortly after ejecting him from his apartment, and continued to ask the defendant questions about his whereabouts that morning until Hallam and McIver arrived approximately an hour later. See Sachdev, 171 N.H. at 551 (concluding that trial court did not err in determining length of interview by length of “substantive questioning” rather than by length of overall interaction with police). Hallam and McIver then interrogated the defendant for an additional hour in the Impala. Thus, the duration of the police interrogation in this case suggests that the defendant was in custody. See Jennings, 155 N.H. at 774.
In evaluating the character of the interrogation, we consider the presence or absence of both accusatory questions and accusatory statements. McKenna, 166 N.H. at 681. Accusatory questioning weighs in favor of custody because such questioning often conveys that the questioning officer believes the defendant is guilty and that he or she intends to arrest. See id. at 681-82. Accusatory questioning stands in contrast to “questioning of a purely general nature,” which does not weigh in favor of custody. Id. at 682. In addition to accusatory questioning, the presence or absence of accusatory statements is relevant to our analysis because “‘a reasonable person understands that the police ordinarily will not set free a suspect when there is evidence strongly suggesting that the person is guilty of a serious crime.’” Id. at 683 (quoting State v. Muntean, 12 A.3d 518, 528 (Vt. 2010)). Therefore, “confronting the
2 In analyzing whether an interrogation is custodial, we note that the character of an interrogation may impact whether the interrogation’s duration is a reliable indicator of custody. See E.G., 171 N.H. at 237; State v. Dedrick, 132 N.H. 218, 225 (1989), abrogated on other grounds by Ford, 144 N.H. at 62-63, and State v. Spencer, 149 N.H. 622, 625 (2003). Although the brevity of an interrogation generally weighs against a finding of custody, Sachdev, 171 N.H. at 551, a particular interrogation might only have been brief because the officers’ questions were so accusatory that the defendant recognized the futility of resistance and confessed in short order, see E.G., 171 N.H. at 237. Indeed, “custody has been found in relatively brief interrogations where the questioning is of a sort where the detainee is aware that questioning will continue until he provides his interrogators the answers they seek.” Id. (quotation omitted); see also Dedrick, 132 N.H. at 225 (explaining that, when defendant was “accused of untruths,” “confronted with damning information,” and officers ignored defendant’s “vehement denials,” the character of interrogation signaled that period of questioning would continue until defendant confessed or asked to speak with an attorney). Thus, an interrogation’s character is often a more reliable barometer for custody than its duration.
8 defendant with evidence of guilt weighs in favor of custody.” Id. Finally, the extent to which the officers “raise[d] their voices or use[d] harsh language” is also relevant to our analysis of the interrogation’s character. Marin, 172 N.H. at 161.
In this case, as noted, Kekejian began questioning the defendant about AV’s allegations shortly after ejecting him from his home. Kekejian asked the defendant whether he had gone inside AV’s bedroom that morning, about his living arrangements at the home, and about “what he had done” that morning. Kekejian also told the defendant, while blocking him from entering the home, that it was “being held as a scene.” Later, when Kekejian, Ciszek, and the defendant were waiting on the porch for Hallam and McIver to arrive, Kekejian renewed questioning the defendant as to his whereabouts that morning. Although some of Kekejian’s questions were general or non-accusatory in nature, we conclude that, on balance, the character of Kekejian’s discussions with the defendant weighs slightly in favor of custody. Cf. E.G., 171 N.H. at 233-35 (contrasting accusatory questioning weighing in favor of custody from questioning that may lawfully occur during investigatory stop).
Weighing much more decisively in favor of custody is the overwhelmingly accusatory nature of the detectives’ questioning and statements inside the Impala. Hallam and McIver repeatedly and doggedly accused the defendant of dishonesty and of sexually assaulting AV. “On numerous occasions throughout the interrogation the [detectives] asked the defendant why he had sexually abused [AV] and posited reasons for his actions.” McKenna, 166 N.H. at 683. For example, at one point Hallam said,
you were rubbing her f**king vagina. What, what were you doing it for? What was the purpose? . . . I, I’ll tell you straight up, I know why [inaudible] is because I’m f**king turned on and s**t and you’ve got to be straight up with me. I know why guys f**king do that s**t so I need you to be straight up with me with why you were doing it man, because I’m not stupid.
As the above excerpt makes clear, the detectives’ “questions were premised upon the assumption that the defendant had committed the crime,” and would have communicated that assumption to a reasonable person in the defendant’s position. Id. at 682-83; see Stansbury, 511 U.S. at 325. In other words, the detectives’ questions “would have signaled [to] a reasonable [person] in the same circumstances that . . . as often as he made denials, they would renew their accusations.” Dedrick, 132 N.H. at 225. In addition, Hallam and McIver repeatedly confronted the defendant with AV’s allegations, telling him that they were certain that the allegations were true. See McKenna, 166 N.H. at 683 (discussing facts of Jennings); Jennings, 155 N.H. at 769, 774 (explaining that interrogation’s character weighed in favor of custody when “[w]ithin minutes of being inside the interview room . . . [a police officer] confronted the defendant
9 with [the victim’s] allegations of sexual assault and said he was certain she was telling the truth,” and when defendant was “repeatedly confronted . . . in this manner”). “Coupled with the control exercised by the police from the beginning of the encounter, this clear indication that the police believed the defendant to be guilty of sexual assault would have signaled to a reasonable person that his freedom of movement was curtailed to the degree associated with formal arrest.”3 Jennings, 155 N.H. at 774; see Dedrick, 132 N.H. at 225.
Several additional circumstances support the conclusion that the defendant was in custody during the Impala interview. “The number of officers present is a relevant factor in a custody determination — when multiple officers isolate and question a defendant, it is more likely that the defendant is in custody.” E.G., 171 N.H. at 237. Here, the defendant was in the presence of four visibly armed police officers, some of whom were in uniform. See Marin, 172 N.H. at 163 (fact that officers were in uniform and were visibly armed weighed in favor of custody); Jennings, 155 N.H. at 773 (“The fact that three officers and a prosecutor went to meet the defendant certainly bolsters the trial court’s custody determination.”). Although no more than two officers questioned the defendant at any given time, Kekejian and Ciszek were within viewing distance from the Impala. See McKenna, 166 N.H. at 685. We also note, as did the trial court, that the fact that the defendant had remained for an hour on a partially enclosed porch on a cold and windy day, without a jacket, suggests that a reasonable person in his position would believe he was in custody.
Moreover, although Hallam’s car was parked in front of the defendant’s home, a location which was familiar to the defendant, the car itself was an unfamiliar location. See Marin, 172 N.H. at 164 (noting that police cruiser was an unfamiliar location); cf. Jennings, 155 N.H. at 774 (fact that interview took place in “closed-door, confined atmosphere” suggested custody). In addition, the defendant was never told that he was not under arrest, and Kekejian did not tell the defendant he could refuse to answer his questions, that he could refuse to wait for Hallam, or that he could refuse to speak with Hallam. See McKenna, 166 N.H. at 679-80. These factors weigh in favor of custody.
To be sure, not all of the circumstances of the defendant’s encounter with the police suggest that he was in custody during the Impala interrogation. Here, “as in virtually every case, there are some factors that weigh in favor of a
3 We note that the trial court had the benefit of reviewing the recording of the interrogation inside the Impala. The court determined, based upon this audio recording, that “the tone of the interview was, for the most part, polite.” The State has not provided us with the recording on appeal. However, even assuming that the trial court’s finding as to the tone of the interview is supported by the recording, see Sachdev, 171 N.H. at 548-49, “[n]either the absence of hostility on the part of the [detectives], nor the polite tone of the interrogation, neutralizes the content or import of the accusatory questions and statements, nor diminishes the weight which we accord to them,” McKenna, 166 N.H. at 684.
10 finding of custody, and some that weigh against such a finding.” Marin, 172 N.H. at 160. Hallam and McIver wore plain clothes, not police uniforms. See Sachdev, 171 N.H. at 553. Neither Kekejian nor Ciszek physically prevented the defendant from leaving the porch, the defendant was not placed in handcuffs or other similar devices, and the porch of the defendant’s home was a familiar location. See Marin, 172 N.H. at 164; State v. Locke, 149 N.H. 1, 6-7 (2002). But see McKenna, 166 N.H. at 677 (“The location of questioning is not, by itself, determinative: a defendant may be in custody in his own home but not in custody at a police station.” (quotation omitted)).
The State emphasizes that Hallam told the defendant he should “feel free” to leave at the outset of the interrogation in the Impala. “[O]ur cases reflect that we have consistently regarded as a significant factor in our custody analysis whether a suspect is informed that he or she is at liberty to terminate the interrogation.” McKenna, 166 N.H. at 680. Thus, a “person who is clearly advised that he is free to leave is ordinarily not in custody.” Jennings, 155 N.H. at 775 (emphasis added). Here, however, the defendant was not clearly advised that he was free to leave. Rather, based upon the recording of the Impala interrogation, the trial court found that Hallam and the defendant were talking over each other when Hallam stated that the defendant should “feel free” to leave. As such, the trial court discounted the effectiveness of Hallam’s statement. In light of the State’s failure to provide us with this recording on appeal, we must assume that the recording supports the trial court’s determination as to this statement’s effectiveness. See State v. Thiel, 160 N.H. 462, 464 (2010).
Finally, we do not view as especially significant, in the context of this case, Hallam’s statement to the defendant that he did not want him to feel “forced” to speak with the detectives, or Hallam’s invocation of the word “voluntarily” at the interrogation’s outset. In light of all the circumstances of the defendant’s encounter with the police that morning, “we cannot turn a blind eye toward a custodial relationship simply because the police made a[n] . . . attempt to clothe their custody of the defendant in the language of voluntariness.” Jennings, 155 N.H. at 772 (brackets omitted).
There are times when actions speak louder than words. Ultimately, police conduct should be judged in terms of what was done rather than what the officer may have called it at the time. Here, as noted above . . . , many indicia of custody were present, and the accusatory nature of the interview is undeniable. By the time the police confronted the defendant in the closed-door interview . . . with both [the victim’s] accusations and their own certainty that the accusations were true, a reasonable person in the defendant’s position would have found his freedom curtailed to the extent that Miranda warnings were required. Thus, we think it implausible that the defendant could have risen from his seat and
11 freely exited the interview . . . in the middle of an escalating period of interrogation and gone along his merry way . . . .
Id. at 775 (quotations and citation omitted).
Based upon the totality of the circumstances, we conclude that a reasonable person in the defendant’s position would have believed himself to be in custody at the point specified by the trial court in its suppression order. Therefore, we hold that the defendant was in custody for Miranda purposes at that point, and the defendant’s statements made after the interrogation became custodial must be suppressed. Because the defendant prevails under the State Constitution, there is no need to analyze the parties’ arguments under the Federal Constitution. See Jennings, 155 N.H. at 776; Ball, 124 N.H. at 237.
III
We turn now to the defendant’s statements at the police station. The trial court suppressed these statements because it found that the State did not prove beyond a reasonable doubt that they were voluntarily given. The State argues that the court erred in so finding. Again, we first address the voluntariness issue under the State Constitution and rely upon federal law only to aid our analysis. See Ball, 124 N.H. at 231-33.
Our State Constitution requires the State to prove beyond a reasonable doubt that a defendant’s confession is voluntary. State v. Ruiz, 170 N.H. 553, 559 (2018); see N.H. CONST. pt. I, art. 15. To be voluntary, a confession must be the product of an essentially free and unconstrained choice and not be extracted by threats, violence, direct or implied promises of any sort, or by the exertion of any improper influence or coercion. Ruiz, 170 N.H. at 560. We will not overturn a trial court’s determination that a confession was not voluntary unless it is contrary to the manifest weight of the evidence, as viewed in the light most favorable to the defendant. See State v. Cloutier, 167 N.H. 254, 258 (2015). 4
4 The State, quoting Ruiz, asserts that we must review the evidence of voluntariness in the light most favorable to the State. We stated in Ruiz that the trial court’s ruling – that the defendant’s confession was voluntarily given – would “not be overturned unless it was contrary to the manifest weight of the evidence, as viewed in the light most favorable to the State.” Ruiz, 170 N.H. at 560; see also Cloutier, 167 N.H. at 258 (specifying that we “will not overturn a trial court’s determination that a confession is voluntary unless it is contrary to the manifest weight of the evidence, as viewed in the light most favorable to the State” (emphasis added)); State v. Rezk, 150 N.H. 483, 486 (2004) (same). That standard of review applied in Ruiz because the defendant was the appealing party with the burden of demonstrating that the court’s voluntariness ruling was erroneous. See Ruiz, 170 N.H. at 555, 559-60. See generally Gallo v. Traina, 166 N.H. 737, 740 (2014) (“[A]ppealing parties . . . have the burden of demonstrating reversible error.”). Here, however, the trial court ruled that the defendant’s second set of statements was not voluntarily given, and the State is the party appealing that ruling. “A determination of the voluntariness of a confession is a question of fact for the trial court to decide . . . .” State v. Chapman, 135 N.H. 390,
12 When, as in this case, a defendant’s post-Miranda confession is preceded by an earlier confession that violated his Miranda rights, we have articulated five factors which guide our analysis under part I, article 15 as to whether the second confession was voluntary in light of the Miranda violation: (1) the time lapse between the initial confession and the subsequent statements; (2) the defendant’s contacts, if any, with friends or family members during that period of time; (3) the degree of police influence exerted over the defendant; (4) whether the defendant was advised that his prior admission could not be used against him; and (5) whether the defendant was advised that his prior admission could be used against him. State v. Fleetwood, 149 N.H. 396, 405- 06 (2003); accord Ruiz, 170 N.H. at 560. No single factor is dispositive. Ruiz, 170 N.H. at 560.
The State argues that the trial court committed legal error when it “looked beyond the five factors” highlighted above in analyzing the voluntariness of the defendant’s second set of statements. The State further contends that the trial court committed legal error by applying the “cat out of the bag” theory to the defendant’s claim, a theory which we rejected under the New Hampshire Constitution in State v. Aubuchont, 141 N.H. 206, 208-09 (1996).
We begin with the latter contention. “Under Federal constitutional law prior to . . . Oregon v. Elstad, 470 U.S. 298 (1985), . . . an admission or confession of guilt obtained from an accused person in violation of the Miranda requirements was presumed to taint any subsequent confession made by the accused . . . .” Com. v. Smith, 593 N.E.2d 1288, 1291 (Mass. 1992). This presumption came to be known as the “cat out of the bag” theory, based on the United States Supreme Court’s use of that phrase in United States v. Bayer, 331 U.S. 532, 540 (1947). See State v. Smith, 834 S.W.2d 915, 919 (Tenn. 1992). See generally Katherine E. McMahon, “Cat-Out-of-the-Bag” & “Break- in-the-Stream-of-Events”: Massachusetts’ Rejection of Oregon v. Elstad for Suppression of Warned Statements Made After a Miranda Violation, 20 W. New Eng. L. Rev. 173, 185-87 (1998). In Elstad, however, the United States Supreme Court disclaimed the notion that a Miranda violation renders subsequent confessions, obtained after the administration of Miranda warnings, presumptively involuntary. See Elstad, 470 U.S. at 314. Instead the Court concluded that, “absent deliberately coercive or improper tactics in obtaining the initial statement, the mere fact that a suspect has made an unwarned admission does not warrant a presumption” that a subsequent, post-Miranda confession was involuntary. Id.
399 (1992) (quotation omitted). That factual determination, made after considering the totality of the relevant evidence, “is entitled to stand” unless the appealing party demonstrates that it is contrary to the manifest weight of the evidence as viewed in the light most favorable to the appellee, which, in this case, is the defendant. Id.
13 In Aubuchont, the defendant argued that, in interpreting part I, article 15 of the New Hampshire Constitution, we should reject the Elstad Court’s holding. Aubuchont, 141 N.H. at 208. “The defendant assert[ed] that instead, we should follow those jurisdictions which hold that, pursuant to state constitutional law, ‘extraction of an illegal, unwarned confession from a defendant raises a rebuttable presumption that a subsequent confession, even if preceded by proper Miranda warnings, is tainted by the initial illegality’ . . . .” Id. at 208-09 (quoting Smith, 834 S.W.2d at 919). We declined to apply such a presumption under our State Constitution. Id. at 209. Instead, we explained that we would “apply our traditional part I, article 15 due process voluntariness inquiry and ask whether[,] considering the totality of the circumstances, the second confession [was] the product of an essentially free and unconstrained choice.” Id. (quotations and citation omitted).
We refined our approach to determining the voluntariness of post- Miranda-violation confessions in State v. Fleetwood. See Fleetwood, 149 N.H. at 402-06. In Fleetwood, unlike in Aubuchont, “there was one continuous interrogation by the same detectives at one location interrupted by only a fifteen-minute break to comply with Miranda.” Id. at 404; see Aubuchont, 141 N.H. at 206-07. These facts, we said, “br[ought] into sharp focus the concern that Elstad may ‘give a green light to law enforcement officers to ignore the requirements of Miranda until after such time as they are able to secure a confession.’” Fleetwood, 149 N.H. at 404 (quoting United States v. Carter, 884 F.2d 368, 373 (8th Cir. 1989)). In light of this concern, we articulated specific factors that the traditional part I, article 15 voluntariness inquiry “‘necessarily encompasses’” as applied to post-Miranda-violation confessions:
“[T]he time lapse between the initial confession and the subsequent statements; the defendant’s contacts, if any, with friends or family members during that period of time; the degree of police influence exerted over the defendant; whether the defendant was advised that her prior admission could not be used against her; or whether the defendant was told that her previous statement could be used against her.”
Id. at 405-06 (quoting United States v. Wauneka, 770 F.2d 1434, 1440-41 (9th Cir. 1985)) (brackets and ellipsis omitted); accord Ruiz, 170 N.H. at 560. We did not hold in Fleetwood, however, that courts should consider only these factors to the exclusion of all others. See Fleetwood, 149 N.H. at 402-08. Rather, we reaffirmed that, “as we stated in Aubuchont . . . , the proper analysis is a totality of the circumstances test.” Id. at 405.
In this case, the trial court properly conducted a totality of the circumstances analysis. The court discussed and applied the factors articulated in Fleetwood, as well as other factors we have found relevant when conducting voluntariness analyses, such as whether the police complied with
14 Miranda in obtaining the second confession, see State v. Bilodeau, 159 N.H. 759, 764 (2010), and whether the police made promises, threats, or engaged in displays of force, see Ruiz, 170 N.H. at 560. In ruling that the State failed to prove beyond a reasonable doubt that the defendant’s second confession was voluntary, the court accorded significant weight to “the manner in which Detective Hallam utilized the defendant’s ‘prior confession in obtaining [the] second confession.’” (Quoting Wauneka, 770 F.2d at 1440.)
However, contrary to the State’s assertions on appeal, the trial court’s decision to accord significant, even determinative, weight to this factor did not equate to the application of a presumption that the defendant’s second confession was involuntary due to the earlier Miranda violation. Rather than presuming that the defendant’s unwarned confession rendered the second involuntary, the trial court carefully considered the specific manner in which Hallam deployed the defendant’s first confession to obtain the second. The court noted, for example, that “Hallam began the second interrogation by incorporating the defendant’s [first confession] into his questions,” which the court found “even more troublesome” in this particular case because Hallam “was only able to extract the first confession after repeatedly accusing the defendant of lying and ignoring his consistent denials.” Because the trial court’s decision to accord substantial weight to this circumstance in its analysis of the totality of the circumstances did not amount to the application of a legal presumption, we reject the State’s argument that the trial court erred as a matter of law by applying the “cat out of the bag” theory in considering the voluntariness of the defendant’s second set of statements.
The State further argues that the trial court erred as a matter of law because the manner in which a police officer uses a defendant’s first, unwarned confession to obtain a second is, essentially, legally irrelevant to the question of whether the second was involuntary, insofar as such a consideration requires law enforcement to predict whether a judge would grant a motion to suppress the first confession on Miranda grounds. We disagree. Although we have cautioned, in analyzing the voluntariness of post-Miranda- violation confessions, that it would be “impractical to require the police to determine the admissibility of an unwarned confession,” Ruiz, 170 N.H. at 561- 62 (quotation omitted), we have never gone so far as to hold that “the manner in which the officers utilized this prior confession in obtaining a second confession,” Wauneka, 770 F.2d at 1440, is irrelevant to whether the second was given voluntarily, see Aubuchont, 141 N.H. at 209-10.
Indeed, in Aubuchont we affirmed the trial court’s finding that the defendant’s post-Miranda-violation confession was voluntary. Aubuchont, 141 N.H. at 209. In reaching that conclusion, however, we considered the fact that the officers “scheduled the [second] interview to ‘go over’” the defendant’s first, unwarned statements. Id. While we were not convinced that “this [fact] alone . . . g[a]ve rise to the conclusion that the second confession was involuntary,”
15 we nevertheless considered this factor under our “traditional part I, article 15 due process voluntariness inquiry.” Id. (emphasis added); see Wauneka, 770 F.2d at 1440. There is ample justification for the trial court’s consideration of Hallam’s use of the first confession to obtain the second as a relevant factor in the totality of the circumstances. See Aubuchont, 141 N.H. at 209; Wauneka, 770 F.2d at 1440 (noting that courts should evaluate how officers used a confession obtained in violation of Miranda to obtain a second confession); see also Joshua I. Rodriguez, Note, Interrogation First, Miranda Warnings Afterward: A Critical Analysis of the Supreme Court’s Approach to Delayed Miranda Warnings, 40 Fordham Urb. L.J. 1091, 1121 (2013) (“Most courts evaluate whether post-Miranda questioning referenced pre-Miranda statements and ask whether the police confronted the suspect with her prior statements.”).
Finally, the State argues that the trial court’s determination is against the manifest weight of the evidence. Viewing the evidence in the light most favorable to the defendant, we disagree. See Cloutier, 167 N.H. at 258. The trial court found, and we concur, that “[t]his case presents a close call.” On the one hand, for example, over eleven hours lapsed between the defendant’s first and second confessions. During most of that time the defendant was not in police custody, spending at least some of that time with his father and at his father’s home. These facts tend to support a finding of voluntariness. See Aubuchont, 141 N.H. at 209.
On the other hand, based on the trial court’s review of the video recording of the second set of statements, the court found that “less than twenty seconds after Detective Hallam finished the Miranda process,” he brought up the defendant’s first set of inculpatory statements.5 See id. Almost immediately after Hallam administered Miranda warnings, he told the defendant that he “wanted to talk to him again,” and that “we went over this already and . . . we talked about it in the car out at the . . . scene.” Hallam told the defendant that there were “inconsistencies” in the defendant’s first set of statements, and that Hallam thought the defendant had “led [him] in the wrong direction” despite the fact that the defendant had “said [he] gave [Hallam] the truth the first time, 100 percent.” Hallam told the defendant, “[i]t is what it is,” and that, because the defendant had “been arrested,” he wanted the defendant to be “100 percent honest” this time. Hallam continued to bring up the defendant’s first set of inculpatory statements at multiple points throughout the second interrogation, telling the defendant to “skip all that B.S. that we were doing out there,” and that they “had . . . a 50 minute conversation about B.S. until we got to the truth . . . .” We agree with the trial court that Hallam’s use of the first set of statements in obtaining the second weighs in favor of a
5 On appeal, the State has not provided us with this video recording. Therefore, we must assume that the evidence was sufficient to support the trial court’s factual findings as they relate to the second interrogation. See Sachdev, 171 N.H. at 548-49.
16 finding that the latter statements were not voluntary. See id.; see also United States v. Gonzalez-Lauzan, 437 F.3d 1128, 1139 (11th Cir. 2006) (concluding that second confession was not involuntary because, in part, “the officers did not have prewarned incriminating statements with which to cross-examine [the defendant] in order to pressure him to repeat them”).
After balancing these and other factors, the trial court was left with “a reasonable doubt as to the voluntariness” of the defendant’s second set of statements. Accordingly, the court determined that the State had failed to carry its burden to prove that these statements were voluntary. See Ruiz, 170 N.H. at 559 (noting that the State has the burden of proving voluntariness beyond a reasonable doubt). From our review of the record, viewed in the light most favorable to the defendant, we cannot say that the trial court’s determination is against the manifest weight of the evidence. See Cloutier, 167 N.H. at 258. Accordingly, we affirm the court’s ruling as to the involuntariness of the defendant’s second set of statements and its order suppressing those statements. Because the defendant prevails under the State Constitution, there is no need to address the parties’ arguments under the Federal Constitution. See Jennings, 155 N.H. at 776; Ball, 124 N.H. at 237.
Affirmed and remanded. HICKS, BASSETT, and DONOVAN, JJ., concurred.