United States v. Guillen

995 F.3d 1095

Opinion

FILED

United States Court of Appeals PUBLISH Tenth Circuit

UNITED STATES COURT OF APPEALS April 27, 2021

Christopher M. Wolpert

FOR THE TENTH CIRCUIT Clerk of Court

_________________________________

UNITED STATES OF AMERICA, Plaintiff - Appellee, v. No. 20-2004 ETHAN GUILLEN,

Defendant - Appellant.

_________________________________

Appeal from the United States District Court for the District of New Mexico (D.C. No. 1:17-CR-01723-WJ-1)

_________________________________

Melissa Ayn Morris, Assistant Federal Public Defender, Albuquerque, New Mexico, for Defendant–Appellant.

Tiffany L. Walters, Assistant United States Attorney (John C. Anderson, United States Attorney, with her on the brief), Albuquerque, New Mexico, for Plaintiff–Appellee.

_________________________________

Before TYMKOVICH, Chief Judge, BALDOCK, and CARSON, Circuit Judges.

_________________________________

BALDOCK, Circuit Judge.

_________________________________

After a young woman found a pressure cooker bomb hidden under her bed, law enforcement agents went to the home of the only person she said might want to harm

2

her: Ethan Guillen.1 The agents entered Ethan’s home, questioned him, and obtained consent from his father to search the residence. During the search, the agents found evidence in Ethan’s bedroom indicating his involvement with the pressure cooker bomb. When one of the agents confronted Ethan with the information and evidence they had collected, he confessed to making the bomb. The agent immediately provided the warnings required by Miranda v. Arizona, 384 U.S. 436 (1966), and Ethan proceeded to make more incriminating statements.

Ethan ultimately entered a conditional plea of guilty to possession of an unregistered destructive device and an attempt to damage or destroy a building by means of fire or an explosive. His plea agreement reserved the right to appeal the district court’s order denying his motion to suppress the physical evidence and incriminating statements resulting from the search of his home. Exercising that right, Ethan argues the district court should have suppressed the physical evidence found in his home because the agents’ warrantless entry and search of his bedroom violated his Fourth Amendment rights. He also contends the district court should have suppressed the incriminating statements he made after receiving Miranda warnings because the agents elicited them through coercion and used an impermissible two- step interrogation technique to end run around Miranda.

We have jurisdiction under 28 U.S.C. § 1291, and we conclude that the district court correctly denied Ethan’s suppression motion. No Fourth Amendment violation

1 The parties refer to Ethan Guillen; his father, Reynaldo Guillen; and his brother, Tyler Guillen, by their first names. In the interest of clarity, we do the same.

3

occurred because Ethan voluntarily consented to the agents’ entry into his home and because the agents reasonably relied on his father’s consent to search his bedroom. Ethan’s initial confession, which the district court suppressed, was inadmissible because the agents failed to provide Miranda warnings before they engaged in custodial interrogation. But the midstream Miranda warnings Ethan received were sufficient to advise him of his rights and render his voluntary postwarning statements admissible. For these reasons, we affirm the district court’s judgment.

I.

On May 31, 2017, law enforcement responded to a 911 call from “MC,” a young woman who had found an improvised explosive device under her bed. The device was a pressure cooker sealed with white duct tape and filled with black powder; homemade napalm; and various types of shrapnel, including nuts, bolts, and screws. A fuse ran through the pressure cooker’s release valve and connected to an electric soldering iron, which was plugged into a timer that was plugged into the wall with a power strip. The device was designed so that the timer would turn on the soldering iron, which would heat up, ignite the fuse, and cause an explosion. Fortunately, the bomb never detonated.

Special Agent Zachary Rominger, an agent with the Bureau of Alcohol, Tobacco, Firearms, and Explosives (ATF), interviewed MC and her mother. When asked if anyone would want to hurt or kill her, MC could think of only one person— her ex-boyfriend, Ethan. MC said she had dated Ethan for about six months. After they broke up, MC explained, Ethan continued to try and communicate with her

4

against her wishes. And at some point, MC’s school provided her with an escort to class in an effort to stop Ethan’s harassment.

After wrapping things up at MC’s home, law enforcement agents went to Ethan’s house. At approximately 9:43 p.m., ATF Special Agents Zachary Rominger and Derek Wright, FBI Special Agent Bomb Technicians Craig Greene and Michael Anthony, and Albuquerque Police Department Detective James Larranaga knocked on Ethan’s front door. FBI Supervisor Marco Gonzalez was also present, but he stayed out by the street when the other five agents approached the house. Detective Larranaga’s lapel camera captured on video, among other things, the moments leading up to the agents’ entry into Ethan’s home.

When the agents knocked on the door, Ethan and his brother, Tyler Guillen, answered. At the time, Ethan was eighteen years old; Tyler was twenty. The agents asked if they could come inside and talk. Tyler agreed, but Ethan asked if the agents had a warrant. After Agent Greene said they did not have a warrant, Ethan suggested they talk in the doorway instead. When the agents asked again if they could come inside, the brothers had a brief and largely inaudible discussion, during which Tyler asked Ethan, “Why do you care?” After the discussion, one of the brothers said “sure.” Agents Rominger and Greene testified that Ethan said “sure,” but Tyler testified he made that statement. Assessing the conflicting accounts and the lapel cam video, the district court deemed the agents’ testimony credible and found that Ethan was the one who said the word “sure.” At this point, Tyler put his hand on Ethan’s shoulder, and both brothers moved out of the doorway. Agent Greene

5

confirmed: “Are you inviting us in to talk?” One of the brothers responded, “Yeah, sure.” Then they all went inside the house.

Following a protective sweep of the residence, during which no evidence was gathered, Agents Rominger and Greene interviewed Ethan at the kitchen table while other agents spoke with Tyler in the hallway. Tyler told the agents that their father, Reynaldo, is a musician and was practicing with his band in Santa Fe that night. Reynaldo returned home about 18 minutes after the agents’ entry. He informed the agents he had recently bought a pressure cooker for Ethan and, at the agents’ request, looked for it. After Reynaldo could not find the pressure cooker, he asked Ethan where it was. Ethan said he had taken it to his mother’s house. Reynaldo then called Ethan’s mother to see if she had the pressure cooker. Ethan’s mother said she did not know whether the pressure cooker was at her house and told Reynaldo she would have to look for it.

While questioning continued, Reynaldo confirmed he owned the house and verbally consented to a search. He also signed a search consent form. At that point, the agents called in the Albuquerque Bomb Squad to assist. Shortly after the search began, the agents asked Reynaldo if he owned a soldering iron. Reynaldo said he did, but he couldn’t find it. The agents’ search did not uncover the soldering iron either, but they discovered a table on the back porch with burn marks and a piece of fuse burnt onto it. The bomb squad also found white duct tape matching the tape on the pressure cooker bomb, black duct tape, latex gloves, scissors, super glue, and zip

6

ties in a backpack on the floor of the master bedroom. The master bedroom was Ethan’s room.

Agents Rominger and Greene questioned Ethan at the kitchen table for about 50 minutes, during which time he repeatedly denied any involvement with making the pressure cooker bomb. After the search ended, the agents asked Ethan, who was then sitting on the couch in the living room, to return to the kitchen table. Agent Rominger then laid out the evidence discovered during the search, told Ethan it pointed to him, and asked if he created the improvised explosive device. Ethan hesitated, took a deep breath, and then said: “Yes, I made it.” Agent Rominger immediately read Ethan his Miranda rights.

Ethan acknowledged he understood his rights, continued to respond to the agents’ questions, and provided information about his involvement with making the device for the next 20 to 40 minutes. Among other things, Ethan described how he made the pressure cooker bomb, explained how he planted it under MC’s bed, and told the agents he wanted MC dead. He also took the agents into his room and showed them the items he used to build the device, including the white duct tape, gloves, and super glue stored in his backpack. Ethan never requested a lawyer or asked the agents to stop questioning him. And he answered all of their questions except one—whether he planned to make another device to kill MC.

Ethan was charged with possession of an unregistered destructive device, in violation of 26 U.S.C. § 5861(d), and an attempt to destroy a building by fire or explosive, in violation of 18 U.S.C. § 844(i). In district court, Ethan filed a motion

7

to suppress his self-incriminating statements and the physical evidence obtained from the search of his home. Ethan argued that the agents violated his Fourth Amendment rights by entering his house and searching his bedroom without a warrant and without his consent. He also contended that the agents violated his Fifth Amendment rights by questioning him without first providing Miranda warnings.

The district court suppressed Ethan’s pre-Miranda confession but denied his motion as to the remaining statements and evidence. Starting with the Fourth Amendment claims, the district court found that both Ethan and his brother voluntarily consented to the agents’ entry into their home. The district court also determined the search of Ethan’s bedroom was lawful because his father had apparent authority to consent to it.

Turning to the Fifth Amendment claims, the district court found the agents’

questioning moved beyond simply attempting to elicit information to custodial interrogation when they confronted Ethan with the evidence discovered during the search. Because Agent Rominger had not yet warned Ethan of his Miranda rights, the district court suppressed the initial confession Ethan made in response to this custodial interrogation. But the district court concluded that Agent Rominger’s administration of the Miranda warnings after that initial statement was sufficient to advise Ethan of his rights and render his postwarning statements admissible.2

2 Following the denial of his suppression motion, Ethan twice requested the district court to reconsider its factual finding that he consented to the agents’ entry into his home. In support, Ethan provided an audio enhancement of the previously submitted video taken on Detective Larranaga’s lapel camera and a forensic report

8

Ethan subsequently pleaded guilty, but he reserved the right to appeal the district court’s denial of his motion to suppress. This is his appeal.

II.

When reviewing the denial of a suppression motion, we view the evidence in the light most favorable to the government, accept the district court’s factual findings unless they are clearly erroneous, and review legal conclusions de novo. United States v. Cortez, 965 F.3d 827, 833 (10th Cir. 2020). Ethan argues the district court erred by not suppressing the physical evidence found in his home and his post- Miranda statements in the face of Fourth and Fifth Amendment violations.

A.

The Fourth Amendment guarantees the right of people to be “secure in their persons, houses, papers, and effects against unreasonable searches and seizures.” U.S. Const. amend. IV. Ethan contends the agents’ warrantless entry into his home violated that right for two reasons. First, he denies he gave consent to law enforcement to enter the residence. Second, he argues that even if he did consent to the entry, his consent was not voluntary. Ethan also claims the search of his bedroom and its contents was unlawful under the Fourth Amendment because his father lacked apparent authority to consent to the search. We address Ethan’s arguments in turn.

outlining the methods and procedures used to create the enhanced material. The district court denied both of Ethan’s motions to reconsider. Because Ethan has not specifically challenged the district court’s reasons for doing so, we need not address them here.

9

1.

We first consider whether the district court erred when it found Ethan voluntarily consented to the agents’ entry into his home. Voluntary consent is a longstanding exception to the general requirement that law enforcement officers must have a warrant to enter a person’s home. United States v. Warwick, 928 F.3d 939, 943 (10th Cir. 2019). The exception applies when the government proves (1) the officers received either express or implied consent and (2) that consent was freely and voluntarily given. United States v. Jones, 701 F.3d 1300, 1317 (10th Cir. 2012). Whether officers obtained valid consent to enter a home is a question of fact determined through the totality of the circumstances. Id. at 318. Accordingly, we review the district court’s finding of consent under the clearly-erroneous standard. United States v. Guerrero, 472 F.3d 784, 789 (10th Cir. 2007).

Ethan first argues that he never actually consented to the agents’ entry into his home. In doing so, however, he fails to meaningfully grapple with the district court’s factual findings. Ethan did initially object to law enforcement entering his home, but the district court found he subsequently consented to the agents’ entry by saying “sure” in response to their second request to come inside. In addition to its repeated viewings of the lapel cam video, the district court based its finding on credible agent testimony that Ethan was the one who said “sure.”

“This court is loath to second-guess a district court’s determination of a witness’s credibility.” United States v. Asch, 207 F.3d 1238, 1243 (10th Cir. 2000). We have no basis for doing so here. Neither Ethan’s nor Tyler’s mouth is visible on

10

the lapel cam video when the word “sure” is spoken. And Ethan has not shown the district court’s credibility determination is internally inconsistent. Warwick, 928 F.3d at 944. He points out that the district court also found Tyler’s testimony credible, but that credibility determination concerned Ethan’s privacy preferences, not who said “sure” during the initial encounter at the door. Thus, despite Tyler’s contrary testimony, the district court did not clearly err when it found Ethan orally consented to the agents’ entry.

The district court likewise did not clearly err when it determined Ethan impliedly consented to the entry by stepping away from the doorway and allowing the agents to enter the house. See Guerrero, 472 F.3d at 789–90 (“[C]onsent must be clear, but it need not be verbal. Consent may instead be granted through gestures or other indications of acquiescence, so long as they are sufficiently comprehensible to a reasonable officer.”). Again relying on his brother’s testimony, Ethan argues his movement away from the door was caused by Tyler shoving him out of the way. But Agent Rominger testified that Tyler did not physically remove Ethan from the doorway. Specifically, Agent Rominger explained that Tyler put his hand on Ethan’s shoulder, and then Ethan walked out of the way on his own accord. Consistent with the agent’s credible testimony, the district court determined there was no visible push on the lapel cam video and found that both brothers cleared the way for the agents to enter. The district court’s determination that Ethan consented through his actions is, at a minimum, not clear error.

11

Ethan next argues that even if he consented to the agents’ entry, his consent was not voluntary. Consent is voluntary if it is unequivocal and specific, freely and intelligently given, and not the product of duress or coercion. Warwick, 928 F.3d at 945. When examining the totality of the circumstances to determine the voluntariness of consent, some relevant considerations include

physical mistreatment, use of violence, threats, promises, inducements, deception, trickery, or an aggressive tone, the physical and mental condition and capacity of the defendant, the number of officers on the scene, and the display of police weapons. Whether an officer reads a defendant his Miranda rights, obtains consent pursuant to a claim of lawful authority, or informs a defendant of his or her right to refuse consent also are factors to consider in determining whether consent given was voluntary under the totality of the circumstances.

Id. (quoting Jones, 701 F.3d at 1318).

According to Ethan, any consent he gave was involuntary because he had no choice but to submit to the agents’ show of authority. More specifically, Ethan says his consent was necessarily coerced for several reasons, including (1) the presence of multiple officers, two of whom were in uniform and visibly armed; (2) the agents’ repeated requests to enter; and (3) his characteristics as a youth lacking prior experience with law enforcement. We are not persuaded.

The presence of multiple officers at the time Ethan consented increased the coercive nature of the encounter, but that factor is not dispositive. United States v. Iribe, 11 F.3d 1553, 1557 (10th Cir. 1993) (holding that the district court clearly erred when it found consent involuntary based on the presence of five officers because numerous factors indicated the resident voluntarily consented, including that

12

she was not coerced, frightened or otherwise threatened; she had a cordial conversation with officers spoken in low volume; and the officers made no promises or threats in an attempt to extract her consent). Most of the agents were dressed in plain clothes, and their weapons were concealed. And the two uniformed officers never drew their weapons or otherwise brandished them. Thus, aside from the mere number of officers present, no other evidence in the record suggests Ethan faced a display of force designed to overbear.

Ethan points to no credible evidence of coercive tactics such as physical mistreatment, use of violence, threats, promises, inducements, deception, or trickery. The agents spoke in a casual, rather than an aggressive, manner. And they never demanded entry or otherwise claimed any lawful authority to be admitted. Under these circumstances, neither the agents’ multiple requests to enter the Guillen residence nor Ethan’s initial objection rendered his subsequent consent involuntary. See United States v. Cruz-Mendez, 467 F.3d 1260, 1263, 1266–68 (10th Cir. 2006) (upholding district court’s finding of voluntary consent when the resident repeatedly refused the officers’ initial requests and told them they needed a warrant to search a home but subsequently consented to the search).

The record evidence also shows that Ethan’s consent was intelligently given.

It is true that the agents did not inform Ethan he could refuse their request to enter the house and that Ethan lacked prior experience with law enforcement. But Ethan was a legal adult who knew enough about his rights to ask the agents if they had a warrant to enter his home. See United States v. Carloss, 818 F.3d 988, 998–99 (10th

13

Cir. 2016) (explaining that an advisement of the right to refuse a warrantless entry is not a prerequisite for voluntary consent and that a prior refusal of consent showed awareness of the right to do so). He also was intelligent enough to build an improvised explosive device that an experienced bomb technician described as one of the most sophisticated devices he had ever seen in New Mexico. For these reasons, the district court’s finding that Ethan voluntarily consented to the agents’ entry into his home is not clearly erroneous.

2.

Ethan also challenges the search of his bedroom, backpack, and nightstand on the ground that his father, Reynaldo, did not have authority to consent to the search. Whether the agents reasonably relied on Reynaldo’s consent to search Ethan’s bedroom and its contents is a legal question we review de novo. United States v. Kimoana, 383 F.3d 1215, 1222 (10th Cir. 2004).

An officer may obtain valid consent to search from a third party with either actual or apparent authority over the subject property. United States v. Romero, 749 F.3d 900, 905 (10th Cir. 2014). Actual authority exists when a third party “has either (1) mutual use of the property by virtue of joint access, or (2) control for most purposes.” United States v. Rith, 164 F.3d 1323, 1329 (10th Cir. 1999). When actual authority is lacking, “a third party has apparent authority to consent to a search when an officer reasonably, even if erroneously, believes the third party possesses authority to consent.” United States v. Andrus, 483 F.3d 711, 716 (10th Cir.), decision clarified on denial of reh’g, 499 F.3d 1162 (10th Cir. 2007). The test for apparent

14

authority is objective: Would “the facts available to the officer at the moment warrant a [person] of reasonable caution [to believe] that the consenting party had authority over the premises?” Romero, 749 F.3d at 905 (quoting Illinois v. Rodriguez, 497 U.S. 177, 188 (1990)).

Critical here is the parent-child relationship between Ethan and his father.

When a child—even an adult child—lives in a parent’s home, the parent is presumed to have “control for most purposes over the property and therefore actual authority to consent to a search of the entire home.” Id.; see also Rith, 164 F.3d at 1331 (applying presumption and holding that the defendant’s parents had actual authority to consent to a search of their 18-year-old son’s bedroom). The presumption holds true unless “rebutted by facts showing an agreement or understanding between the [child] and the [parent] that the latter must have permission to enter the [child’s] room.” Rith, 164 F.3d at 1330–31. Relevant facts include a lock on the child’s bedroom door, an explicit or implicit agreement that the parent will not enter the room without the child’s consent, and payment of rent by the child. Id.

At the suppression hearing, Reynaldo testified that he had agreed to only enter Ethan’s bedroom with Ethan’s permission. Tyler also testified about this arrangement and described how Ethan guarded his privacy by habitually locking his bedroom door to keep everyone out when he was inside. The government does not argue those facts are insufficient to rebut the presumption of actual authority. Instead, it contends that Reynaldo had apparent authority because the agents who

15

searched the house were unaware of those facts and thus reasonably relied on the presumption of control established by Ethan’s relationship with his father. We agree.

At the time of the search, the agents knew Reynaldo owned the house and permitted Ethan to live there. See Romero, 749 F.3d at 906 (“An owner of a house is presumed to have control for most purposes of the entire house, including the bedroom of a [ ]child permitted to live there.”). Although the door to Ethan’s bedroom had a lock on it, which to some extent undermines Reynaldo’s apparent authority, the door was wide open when the agents arrived. Ethan never objected to the search. Neither did Reynaldo or Tyler. And at no point did any of the Guillens limit where the agents could look.

Even if, as Ethan maintains, the agents knew he occupied the master bedroom when the search began, they had no way of knowing Reynaldo only entered that room with Ethan’s permission. When Reynaldo voluntarily consented to the search of the home, he did not mention the agreement he had with Ethan about access to the master bedroom. Ethan and Tyler likewise said nothing about any privacy arrangement amongst the household members. Tyler’s statement that Ethan is “usually in his room doing his own thing” may shed light on Ethan’s personality, but it raises no doubt about Reynaldo’s control over his son’s bedroom. Because the agents reasonably believed Reynaldo had authority to consent, they were not required to make further inquiries “merely because one can imagine some way that additional facts might alter their analysis.” Id. at 907.

16

Under the totality of the circumstances, the facts known to the agents at the time of the search created an objectively reasonable perception that Reynaldo had authority to consent to the search of Ethan’s bedroom. If the agents had learned that Ethan habitually locked his bedroom door to keep everyone out when he was inside and that his father did not enter the bedroom without permission, they may no longer have been justified in relying on Reynaldo’s consent. But the district court found they did not know those facts when they searched Ethan’s bedroom. This after- acquired factual knowledge, therefore, has no bearing on the reasonableness of the agents’ belief in Reynaldo’s authority at the time of the search. Id. at 907–08.

Ethan also contends that even if Reynaldo had apparent authority over the master bedroom, the agents could not reasonably believe he had authority to consent to a search of the backpack and nightstand in that room. We reject that argument as well.

“Common authority over a residence does not necessarily imply common authority over all locations or objects within the residence.” United States v. Bass, 661 F.3d 1299, 1305 (10th Cir. 2011). But when general authority to consent exists, “we should not look for ‘metaphysical subtleties’ to define the boundaries of that authority.” Id. at 1306 (quoting Frazier v. Cupp, 394 U.S. 731, 740 (1969)). Thus, a “protected expectation of privacy may exist where the defendant has taken some special steps to protect his personal effects from the scrutiny of others, but does not unquestionably exist where the co-occupant has ready access (perhaps not theretofore

17

exercised) to the place searched.” Id. (quoting 4 Wayne R. LaFave, Search and Seizure § 8.3(f), at 168–69 (4th ed. 2004)).

Ethan took no “special steps” to protect the backpack or nightstand in his room from scrutiny. The door to Ethan’s bedroom was wide open when the agents arrived, and neither the backpack nor the nightstand was locked or otherwise secured. In fact, the backpack was found lying on the floor of the master bedroom, “hardly an object shouting ‘Do Not Enter.’” See id. What’s more, Reynaldo—whom the agents knew was a musician—stored musical equipment and clothes in the master bedroom, which would make it appear that he had ready access to the room and its contents. In light of those facts and the apparent authority Reynaldo generally had over Ethan’s bedroom, the agents reasonably believed Reynaldo’s authority to consent extended to the search of the backpack and nightstand in that room.

United States v. Salinas–Cano, 959 F.2d 861 (10th Cir. 1992), which Ethan cites to support his argument, does not persuade us otherwise. In Salinas–Cano, this court held that a resident of an apartment lacked the authority to consent to the search of a suitcase her boyfriend (i.e., the defendant) had left there. Id. at 865–66. Prior to the search, the girlfriend told the police she was not the co-owner of the suitcase and that it belonged exclusively to the defendant. Id. at 865. Because the couple spent five nights a week in separate residences, which suggests at least one of them desired his or her own space and the resulting privacy, we analyzed the case as one in which a host consents to the search of an object owned by a guest. See id. at 863; see also Bass, 661 F.3d at 1306. The guest-host relationship is far different from that of a

18

child living in his parent’s home, where the parent is presumed to have control for most purposes of the child’s bedroom. See Romero, 749 F.3d at 906.

The more instructive precedent here is United States v. Andrus, in which we held that a father had apparent authority to consent to the search of a computer located in his adult son’s bedroom. 483 F.3d at 720–22. In Andrus, we likened a computer to a suitcase, footlocker, and other containers that command a high degree of privacy. Id. at 718. But we concluded the facts known to the officers when the search began created an objectively reasonable perception that the father had authority over the computer. Id. at 722. Those facts included: (1) the defendant’s father owned the house and lived there with other family members; (2) the father paid the internet bill; (3) the computer was in the defendant’s bedroom, but the father had access to the room; (4) the computer was in plain view on a desk and appeared available for use by other household members; and (5) the father did not do or say anything to indicate he lacked control over the computer. Id. at 720–21.

The defendant in Andrus argued that the search was unreasonable because his computer was “locked” to third parties due to its password protection, a fact he said the officers would have known had they asked questions of his father prior to the search. Id. at 721. We rejected this argument and declined to place the onus on the officers to affirmatively ask the defendant’s father if the computer was password protected before they relied on his consent. Id. Without an affirmative statement by the father suggesting he did not have authority to access the defendant’s computer, the officers reasonably believed such authority in fact existed. Id. at 721–22.

19

As in Andrus, the surrounding circumstances known to the agents here would not cause a reasonable person to doubt Reynaldo’s consent and refrain from acting on it without further inquiry. Because the agents were reasonable in believing Reynaldo had the requisite authority to consent to a search of Ethan’s bedroom and its contents, no Fourth Amendment violation occurred.3 B.

Ethan also alleges violations of his Fifth Amendment rights due to law enforcement questioning him without providing Miranda warnings. Because the initial confession Ethan gave before Agent Rominger advised him of his Miranda rights was suppressed, Ethan’s argument focuses on his post-Miranda statements, which the district court deemed admissible. Ethan claims the Miranda warnings he received after his initial confession were insufficient to effectively advise him of his rights and render his postwarning statements admissible. In addition, Ethan says all of his incriminating statements, along with the waiver of his Miranda rights, were involuntary and obtained in violation of the Fifth Amendment.

1.

The Self–Incrimination Clause of the Fifth Amendment guarantees that “[n]o person . . . shall be compelled in any criminal case to be a witness against himself.” U.S. Const. amend. V. In Miranda v. Arizona, the Supreme Court concluded that

3 We need not address the government’s alternative argument that the evidence in Ethan’s bedroom inevitably would have been discovered because the agents would have obtained a search warrant absent consent.

20

“the process of in-custody interrogation of persons suspected or accused of crime contains inherently compelling pressures which work . . . to compel him to speak where he would not otherwise do so freely.” 384 U.S. 436, 467 (1966). The Court instituted measures to guard against this danger. Prior to custodial questioning, a suspect must be warned “that he has the right to remain silent, that anything he says can be used against him in a court of law, that he has the right to the presence of an attorney, and that if he cannot afford an attorney one will be appointed for him prior to any questioning if he so desires.” Id. at 479. Without these warnings, custodial confessions are presumed to be the product of coercion and are generally inadmissible for purposes of the prosecution’s case in chief. United States v. Patane, 542 U.S. 630, 639 (2004) (plurality opinion).

But Miranda warnings aren’t always required. They need only be given once an individual is in “custody” and subjected to “interrogation.” Cortez, 965 F.3d at 840 (quoting United States v. Jones, 523 F.3d 1235, 1239 (10th Cir. 2008)). Here, the government does not dispute that Ethan made his pre-Miranda confession in response to questioning that constitutes interrogation. Rhode Island v. Innis, 446 U.S. 291, 300–01 (1980) (explaining that interrogation encompasses both “express questioning” and “its functional equivalent,” including “any words or actions on the part of the police (other than those normally attendant to arrest and custody) that the police should know are reasonably likely to elicit an incriminating response”). So the remaining question is whether Ethan was in custody at that time. If he wasn’t, there is no basis to exclude any of his incriminating statements.

21

An individual is in custody for Miranda purposes when “a reasonable person in the suspect’s position would understand his or her situation as ‘the functional equivalent of formal arrest.’” Cortez, 965 F.3d at 840 (quoting United States v. Revels, 510 F.3d 1269, 1273 (10th Cir. 2007)). This is an objective, fact-intensive inquiry that focuses on the totality of the circumstances. Id. Several relevant factors inform our analysis, including: “(1) whether the circumstances demonstrated a police-dominated atmosphere; (2) whether the nature and length of the officers’ questioning was accusatory or coercive; and (3) whether the police made [the defendant] aware that [he or she] was free to refrain from answering questions, or to otherwise end the interview.” Revels, 510 F.3d at 1275.

When Ethan made his initial self-incriminating statement, the agents had not informed him that he was free to end the interview or decline to answer their questions. That is “a significant indication of a custodial detention.” United States v. Griffin, 7 F.3d 1512, 1518 (10th Cir. 1993). But it also is only one factor to consider. United States v. Zar, 790 F.3d 1036, 1048 (10th Cir. 2015) (finding the failure to advise the defendant that a three-hour interview was a consensual conversation troubling but concluding it “did not transform the in-home interview into a custodial interrogation”).

Another relevant factor is whether Ethan was questioned in a police-dominated atmosphere. After the agents obtained voluntary consent to enter the Guillen residence, two agents questioned Ethan at the kitchen table for about an hour before he made his initial confession. Revels, 510 F.3d at 1275 (“[T]he home is generally a

22

more familiar, comfortable atmosphere than a police interrogation room.”). During this time, Ethan moved freely about his home—going unaccompanied to the bathroom and getting a drink from the refrigerator—and was free to check his phone. Although a large number of law enforcement personnel were in the Guillen residence at various times, Ethan’s father and brother were also present in the home. In fact, Ethan’s father participated in the interview at least once. See United States v. Lamy, 521 F.3d 1257, 1263 (10th Cir. 2008) (concluding that an hour-long interview was not custodial when it was “conducted by two officers in a common area of [the defendant’s] home, during which his mother came and went from the room”). Ethan was not placed in handcuffs or otherwise physically restrained during the interview. And he was never subjected to or threatened with any physical mistreatment. Under these circumstances, we cannot say the agents questioned Ethan in a police- dominated atmosphere. See Revels, 510 F.3d at 1276 (concluding a home was dominated by police when officers “had recently breached [the defendant’s] front door with force, handcuffed her, and placed her prone on the hall floor”).

Whether the nature and length of the agents’ questioning was accusatory or coercive is a closer call. On the one hand, Ethan was not questioned for an especially long period of time before the Miranda warnings were given, and the agents neither spoke in a threatening tone nor displayed their weapons during the inquiry. The agents’ questioning, in short, did not rise to the level of coercion. Compare United States v. Eckhart, 569 F.3d 1263, 1276 (10th Cir. 2009) (finding circumstances to be noncoercive when the defendant “was never handcuffed or placed in a police cruiser

23

and no weapons were drawn” and “the officers were polite in their demeanor and did not use or threaten the use of force at any time”), with United States v. Perdue, 8 F.3d 1455, 1464, 1467 (10th Cir. 1993) (finding a coercive atmosphere when the defendant “was forced out of his car and onto the ground at gunpoint” and then questioned by officers “while police helicopters hovered above” and “while the officers kept their guns drawn on him and his pregnant fiancee”).

On the other hand, the questioning that elicited Ethan’s initial confession was plainly accusatory. After the search ended, Agent Rominger “pushed” Ethan by confronting him with the evidence discovered during the search. Specifically, Agent Rominger told Ethan that the information and evidence they had collected during their investigation indicated he made the pressure cooker bomb:

We know that you purchased a pressure cooker and it’s gone. We know that a soldering iron was used in this device, and your dad’s soldering iron is missing. White tape, like was found on the device, is found in the backpack. And there’s a table that looks like it has black powder burns, and there’s burns and fuses on that table.

ROA, Vol. III at 310. The accusatory nature of such questioning supports a conclusion that Ethan was in custody when he first confessed. See Rith, 164 F.3d at 1332.

Under the totality of the circumstances described above, Ethan was not in custody before Agent Rominger confronted him with the information and evidence discovered during the search. But the situation evolved when Agent Rominger pressed Ethan despite his repeated denials of involvement and then confronted him with the mounting information and evidence collected during the search. At that

24

point, a reasonable person in Ethan’s position would not have felt free to leave or otherwise end the interview. See id. (holding the defendant “was not in police custody until the point at which he was confronted with the illegal shotgun” seized during a search of his home). In other words, Ethan was in custody when he made his initial self-incriminating statement.

Resisting this conclusion, the government contends that the encounter turned custodial when Ethan confessed, not when Agent Rominger confronted him with the evidence discovered during the search. This is so, the government maintains, because the evidence Agent Rominger confronted Ethan with was less conclusive of criminal conduct than the sawed-off shotgun the officers used to confront the defendant in Rith. See id. That may be true, but it doesn’t change the outcome here.

It is difficult to ignore the effect that Agent Rominger’s accusatory questioning had on the nature of the interrogation under the circumstances Ethan faced. See Jones, 523 F.3d at 1240 (“[W]e must look to the totality of the circumstances and consider the police-citizen encounter as a whole, rather than picking some facts and ignoring others.”). With the search of the Guillen residence completed prior to Ethan’s second round of pre-Miranda questioning, there was little remaining for the agents to do other than to leave or place Ethan under arrest. Revels, 510 F.3d at 1277 (finding it “significant” that “the search had been completed prior to [the defendant’s] questioning”). Considering the evidence the agents had discovered, an arrest was likely, and—after being confronted with that evidence—a reasonable person in Ethan’s shoes would have recognized as much. Thus, Ethan

25

would have reasonably understood his situation as the functional equivalent of formal arrest when Agent Rominger elicited his initial confession.

Because the agents failed to advise Ethan of his Miranda rights before they subjected him to custodial interrogation, the district court properly suppressed his unwarned admission.

2.

After Ethan admitted he made the pressure cooker bomb, Agent Rominger immediately provided Miranda warnings. Ethan waived his right to remain silent and then proceeded to provide details about, among other things, how he created the device and planted it under MC’s bed. The question, then, becomes whether Agent Rominger’s tardy administration of Miranda warnings requires suppression of Ethan’s postwarning statements.

a.

The Supreme Court has twice addressed midstream Miranda warnings—first in Oregon v. Elstad, 470 U.S. 298 (1985), and again in Missouri v. Seibert, 542 U.S. 600 (2004). In Elstad, police officers went to the home of the defendant, an 18-year- old they suspected of breaking into a neighbor’s house, with a warrant to arrest him for burglary. 470 U.S. at 300. While there, one of the officers questioned the defendant about the burglary without first reading him Miranda warnings. Id. at 301. The defendant admitted he was present at the burglary, at which point the officers arrested him and took him to the police station. Id. About an hour later, the arresting officers informed the defendant of his Miranda rights. Id. The defendant then

26

waived those rights and gave a full confession detailing his role in the crime. Id. at 301–02.

The Court held that “a suspect who has once responded to unwarned yet uncoercive questioning is not thereby disabled from waiving his rights and confessing after he has been given the requisite Miranda warnings.” Id. at 318. “Though Miranda requires that the unwarned admission must be suppressed,” said the Court, “the admissibility of any subsequent statement should turn in these circumstances solely on whether it is knowingly and voluntarily made.” Id. at 309. The Court rejected the notion that an initial, unwarned statement creates a “lingering compulsion” based on “the psychological impact of the suspect’s conviction that he has let the cat out of the bag and, in so doing, has sealed his own fate.” Id. at 311. “[A]bsent deliberately coercive or improper tactics in obtaining the initial statement,” the Court reasoned, a “subsequent administration of Miranda warnings . . . ordinarily should suffice to remove the conditions that precluded admission of the earlier statement.” Id. at 314. The Court also explained that even when an initial, unwarned statement was actually coerced, the coercive effect could dissipate with the passage of time, a change in place, or a change in identity of the interrogators. Id. at 310–12. Because the defendant’s initial statement and waiver of his rights were voluntary, the Court held that his postwarning confession was admissible. Id. at 318.

The Court revisited the issue of midstream Miranda warnings in Missouri v.

Seibert. There, police officers arrested the defendant for setting a fire that killed a mentally ill teenager. 542 U.S. at 604–05 (plurality opinion). Unlike Elstad, where

27

“the officer’s initial failure to warn was an oversight,” id. at 614 (internal quotation marks omitted), the officers in Seibert “used a two-step questioning technique based on a deliberate violation of Miranda,” id. at 620 (Kennedy, J., concurring in the judgment). The interrogating officer deliberately chose to withhold Miranda warnings in accordance with a sanctioned interrogation technique he had been taught: “question first, then give the warnings, and then repeat the question until I get the answer that she’s already provided once.” Id. at 605–06 (plurality opinion) (internal quotation marks omitted). Employing this strategy, the officer questioned the defendant for 30 to 40 minutes until she confessed, gave her a 20-minute break, provided Miranda warnings, and then led her back over the same ground. Id. at 604– 05.

Five Justices admonished against this “question-first” technique designed to undermine Miranda and concluded that the defendant’s postwarning confession was inadmissible. Id. at 617; id. at 620–22 (Kennedy, J., concurring in the judgment). But no single rationale commanded a majority of the Court.

The plurality opinion, which four Justices joined, suggested the “threshold issue when interrogators question first and warn later is . . . whether it would be reasonable to find that in these circumstances the warnings could function ‘effectively’ as Miranda requires.” Id. at 611–12. According to the plurality, midstream Miranda warnings are not effective unless “a reasonable person in the suspect’s shoes would [ ] have understood them to convey a message that she retained a choice about continuing to talk.” Id. at 617. The plurality laid out five

28

nonexclusive factors to aid courts in determining “whether Miranda warnings delivered midstream could be effective”:

[1] the completeness and detail of the questions and answers in the first round of interrogation, [2] the overlapping content of the two statements, [3] the timing and setting of the first and the second, [4] the continuity of police personnel, and [5] the degree to which the interrogator’s questions treated the second round as continuous with the first.

Id. at 615.

Justice Kennedy concurred in the judgment, but he believed the plurality’s test “cut[] too broadly,” id. at 622 (Kennedy, J.), because it would apply in instances of “both intentional and unintentional two-stage interrogations,” id. at 621. Instead, Justice Kennedy articulated “a narrower test applicable only in the infrequent case . . . in which the two-step interrogation technique was used in a calculated way to undermine the Miranda warning.” Id. at 622. For Justice Kennedy, the first question would be whether the interrogating officer deliberately withheld Miranda warnings “to obscure both the practical and legal significance of the admonition when finally given.” Id. at 620. If the answer to that question is no, Justice Kennedy explained that the “admissibility of postwarning statements should continue to be governed by the [voluntariness] principles of Elstad.” Id. at 622. But if the answer is yes, postwarning statements “related to the substance of prewarning statements must be excluded unless curative measures are taken” to “ensure that a reasonable person in the suspect’s situation would understand the import and effect of the Miranda warning.” Id. Justice Kennedy gave two examples of such curative measures: (1) “a substantial break in time and circumstances between the prewarning

29

statement and the Miranda warning,” and (2) “an additional warning that explains the likely inadmissibility of the prewarning custodial statement.” Id.

Like Justice Kennedy, the plurality distinguished the police conduct in Elstad from the deliberate use of a question-first-warn-later interrogation strategy: “Although the Elstad Court expressed no explicit conclusion about either officer’s state of mind, it is fair to read Elstad as treating the living room conversation as a good-faith Miranda mistake, not only open to correction by careful warnings before systematic questioning in that particular case, but posing no threat to warn-first practice generally.” Id. at 615 (plurality opinion); see also id. at 620–21 (Kennedy, J., concurring in the judgment). But the plurality emphasized that its conclusion did not turn on the subjective intent of the interrogating officer. Id. at 616 & n.6 (plurality opinion). “Because the intent of the officer will rarely be as candidly admitted as it was here (even as it is likely to determine the conduct of the interrogation),” the plurality believed the focus should be “on facts apart from intent that show the question-first tactic at work.” Id. at 616 n.6.4 The four dissenting Justices likewise resisted Justice Kennedy’s attempt to focus the Court’s analysis on the intent of the interrogating officer. Id. at 623

4 Justice Breyer wrote separately to express his preference for a “fruits” rule and good-faith exception to two-step interrogations: “Courts should exclude the ‘fruits’ of the initial unwarned questioning unless the failure to warn was in good faith.” Seibert, 542 U.S. at 617 (Breyer, J., concurring). Although Justice Breyer joined the plurality opinion in full, he also stated that he agreed with Justice Kennedy’s opinion “insofar as it is consistent with [the application of a] good-faith exception” to the exclusionary rule. Id. at 618.

30

(O’Connor, J., dissenting). Because “[f]reedom from compulsion lies at the heart of the Fifth Amendment,” and “voluntariness is a matter of the suspect’s state of mind,” id. at 624, the dissenting Justices declined to recognize “an exception to Elstad for intentional violations,” id. at 625. They also rejected the plurality’s objective inquiry into the effectiveness of Miranda warnings because, in their view, it relied on the same “cat out of the bag” theory the Court rejected in Elstad. Id. at 627. Instead, the dissenting Justices would have evaluated the two-step interrogation and admissibility of the defendant’s postwarning statement under the voluntariness standard established in Elstad. Id. at 628.

b.

The splintered nature of Seibert has given rise to a debate over whether Justice Kennedy’s concurrence or the four-Justice plurality opinion provides the controlling standard for evaluating the admissibility of statements given after midstream Miranda warnings. In the past, we declined to pick a side in the debate and instead applied both tests to the facts of the case before us. See, e.g., United States v. Carrizales–Toledo, 454 F.3d 1142, 1151–53 (10th Cir. 2006); United States v. Crisp, 371 F. App’x 925, 929–33 (10th Cir. 2010) (unpublished). Today we stake out our position: Justice Kennedy’s concurrence is the binding opinion from Seibert.

Vertical stare decisis is absolute and requires us, as middle-management circuit judges, to follow applicable Supreme Court precedent in every case. So once the Supreme Court has adopted a rule, standard, or interpretation, we must use that same rule, standard, or interpretation in later cases. See Seminole Tribe of Florida v.

31

Florida, 517 U.S. 44, 67 (1996) (“When an opinion issues for the Court, it is not only the result but also those portions of the opinion necessary to that result by which we are bound.”). This is a straightforward task when the Court issues a single majority opinion that agrees on both the result and the reasoning. But how do we identify the legal rules of the road when the Justices whose votes were collectively necessary to the judgment disagree on the appropriate rationale to reach that judgment? We have a rule for that. And it’s simply stated, if not always easily applied.

In Marks v. United States, the Supreme Court instructed: “When a fragmented Court decides a case and no single rationale explaining the result enjoys the assent of five Justices, ‘the holding of the Court may be viewed as that position taken by those Members who concurred in the judgment[] on the narrowest grounds.” 430 U.S. 188, 193 (1977) (quoting Gregg v. Georgia, 428 U.S. 153, 169 n.15 (1976) (opinion of Stewart, Powell, and Stevens, JJ.)). The Marks Court, in other words, recognized an exception to the longstanding precept that precedent arises only when a decisional rule attracts a majority of Justices. Cf. Akhil Reed Amar, America’s Unwritten Constitution: The Precedents and Principles We Live By 356–61 (2012) (suggesting the majority rule for precedent formation has deep roots in Anglo-American judicial tradition). But Marks itself “did not elaborate on how to identify the narrowest grounds” for a splintered Supreme Court decision. Bryan A. Garner et al., The Law of Judicial Precedent 199–200 (2016).

We have clarified that a concurring opinion in a splintered Supreme Court decision is the narrowest under Marks, and thus produces a determinate holding,

32

when it is “a logical subset” of the other opinion(s) concurring in the judgment. Carrizales–Toledo, 454 F.3d at 1151 (quoting King v. Palmer, 950 F.2d 771, 781 (D.C. Cir. 1991) (en banc)). If the instances in which a concurring opinion would reach the same result as the splintered decision in future cases form a logical subset of the instances in which the other concurring opinion(s) would reach the same result, that opinion controls.5 EMW Women’s Surgical Ctr., P.S.C. v. Friedlander, 978 F.3d 418, 431 (6th Cir. 2020) (citing King, 950 F.2d at 781). And its precedential force is absolute: “The binding opinion from a splintered decision is as authoritative for lower courts as a nine-Justice opinion. . . . This is true even if only one Justice issues the binding opinion.” United States v. Duvall, 740 F.3d 604, 611 (D.C. Cir. 2013) (Kavanaugh, J., concurring in the denial of rehearing en banc) (internal quotation marks omitted).

Of course, it is not always possible to identify the opinion that “is a logical subset” of the other concurring opinion(s). Carrizales–Toledo, 454 F.3d at 1151 (quoting King, 950 F.2d at 781). Sometimes there is no discernable implicit consensus or “common denominator” among the Justices who support the Court’s judgment, id., “making Marks an exercise in chasing the wind,” United States v. Kratt, 579 F.3d 558, 562 (6th Cir. 2009). In that situation—“when the various

5 To be clear, Marks is not an invitation from the Supreme Court to read its splintered decisions like tea leaves in an attempt to divine how a future Supreme Court might decide a case. “[N]ose-counting” of that sort is an “exercise for litigators, not jurists.” People v. Lopez, 286 P.3d 469, 485 (Cal. 2012) (Liu, J., dissenting). The goal for a lower court under Marks is to determine how the principles articulated in a prior Supreme Court decision apply to the case before it.

33

opinions supporting the Court’s decision are mutually exclusive”—we do not apply Marks.6 Carrizales–Toledo, 454 F.3d at 1151.

Now back to Seibert. There is a coherent way to apply Marks here, which no doubt explains why most courts have reached the same conclusion: Justice Kennedy’s concurrence in Seibert is the controlling opinion from that case.

The Seibert plurality opinion replaces the central voluntariness standard of Elstad with an objective inquiry into whether midstream Miranda warnings were effective to apprise a reasonable person in the suspect’s shoes of his rights. According to the plurality, this effectiveness test—which considers non-exhaustive factors such as the timing and location of interrogations, continuity of police

6 The Supreme Court has twice sidestepped Marks because it did not find it “useful to pursue the Marks inquiry to the utmost logical possibility when it has so obviously baffled and divided the lower courts that have considered it.” Grutter v. Bollinger, 539 U.S. 306, 325 (2003) (quoting Nichols v. United States, 511 U.S. 738, 745–46 (1994)). And in its most recent encounter with Marks, the Court chose to ignore a possibly binding precedent—opting instead to address the underlying merits— without providing any justification whatsoever for doing so. See Hughes v. United States, 138 S. Ct. 1765, 1771–72 (2018) (declining to resolve which opinion in Freeman v. United States, 564 U.S. 522 (2011), is controlling under Marks). We, of course, don’t have that luxury. As now-Justice Kavanaugh explained during his time on the D.C. Circuit:

When the Supreme Court itself applies Marks, it is not bound in the same way that lower courts are bound by Marks to strictly follow the narrowest opinion from a prior splintered Supreme Court decision. That’s because the Supreme Court is free to reconsider or refine or tweak its own precedents—including splintered precedents—and it does so in appropriate cases. Lower courts, by contrast, are not free to reconsider or refine or tweak Supreme Court precedents. Marks is therefore even more important at the lower court level.

Duvall, 740 F.3d at 611 n.2 (concurring in the denial of rehearing en banc).

34

personnel, overlapping content of statements, etc.—applies in all cases involving sequential unwarned and warned statements. Under Justice Kennedy’s concurrence, an objective inquiry into change in time and circumstances between the unwarned and warned statements—what he dubbed “curative measures”—to determine whether midstream warnings were reasonably effective applies only in cases involving the deliberate use of a two-step interrogation technique calculated to circumvent Miranda. Simply put, the analysis of the Seibert plurality opinion and Justice Kennedy’s concurrence merge when a two-step interrogation was deliberately used to evade the requirements of Miranda, and the tests diverge when the interrogating officer(s) unintentionally performed a two-step interrogation.

The cases governed by Justice Kennedy’s approach thus form a logical subset of the cases governed by the plurality’s approach. Postwarning confessions made after objectively ineffective midstream Miranda warnings—determined by an inquiry into the change in time and circumstances between the prewarning and postwarning statements, and any other curative measures—are not always inadmissible, as the plurality concluded; they are inadmissible only when the interrogating officer(s) intentionally used a two-step interrogation strategy calculated to undermine Miranda. Whenever this narrower standard is satisfied and a confession is deemed inadmissible, the Seibert plurality would reach the same conclusion, because they would deem a confession made after objectively ineffective midstream warnings inadmissible in all cases where a two-step interrogation—intentionally or unintentionally—took place. In other words, Justice Kennedy’s legal standard and

35

the alternative standard endorsed by the Seibert plurality would necessarily point to the same result in any case where Justice Kennedy would deem an incriminating statement given after midstream Miranda warnings inadmissible.

Justice Kennedy’s Seibert concurrence, then, is both a logical subset of the plurality opinion and the narrowest grounds for the Court’s judgment. So it is the “controlling opinion” from Seibert, Marks, 430 U.S. at 193, and we must apply its reasoning as we would the reasoning of any other Supreme Court precedent, Duvall, 740 F.3d at 611 (Kavanaugh, J., concurring in the denial of rehearing en banc).

We are in good company in holding that Justice Kennedy’s Seibert opinion provides the controlling standard for assessing the admissibility of incriminating statements given subsequent to midstream Miranda warnings. The majority of our sister circuits have applied Marks to Seibert and reached the same conclusion. United States v. Capers, 627 F.3d 470, 476 (2d Cir. 2010); United States v. Ollie, 442 F.3d 1135, 1142 (8th Cir. 2006); United States v. Street, 472 F.3d 1298, 1313 (11th Cir. 2006); United States v. Courtney, 463 F.3d 333, 338 (5th Cir. 2006); United States v. Williams, 435 F.3d 1148, 1157–58 (9th Cir. 2006); United States v. Naranjo, 426 F.3d 221, 231–32 (3d Cir. 2005) (Alito, J., on the panel); United States v. Mashburn, 406 F.3d 303, 308–09 (4th Cir. 2005).

A string of Seventh Circuit decisions also supports our conclusion that Justice Kennedy’s concurrence set out the holding of Seibert. United States v. Stewart, 388 F.3d 1079, 1090 (7th Cir. 2004) (determining that Justice Kennedy’s Seibert concurrence controls under Marks); United States v. Stewart, 191 F. App’x 495, 497

36

n.2 (7th Cir. 2006) (unpublished) (“In Seibert, Justice Kennedy’s opinion concurring in the judgment provided the narrowest grounds for decision and thus constitutes the holding of the Court.”); United States v. Stewart, 536 F.3d 714, 718–20 (7th Cir. 2008) (applying Justice Kennedy’s Seibert opinion and clarifying the burden of proof and standard of review applicable to the determination of deliberateness). In United States v. Stewart, Judge Sykes ably summarized the opinions in Seibert and pointed out the common denominator for the Court’s judgment: “[A]s to deliberate two-step interrogations in which Miranda warnings are intentionally withheld until after the suspect confesses, the central voluntariness inquiry of Elstad has been replaced by a presumptive rule of exclusion, subject to a multifactor test for change in time, place, and circumstances from the first statement to the second.” 388 F.3d at 1090.

A few years later, a different panel of the Seventh Circuit, along with the Sixth Circuit, deviated from this majority view. In United States v. Heron, the Seventh Circuit recognized that “parts” of Justice Kennedy’s “reasoning could be construed as a narrower ground than the one described in [the] plurality opinion” because his concurrence “would carve out a smaller exception to Elstad.” 564 F.3d 879, 884 (7th Cir. 2009). But this time the court determined that Justice Kennedy’s concurrence did not provide the narrowest grounds for the judgment in Seibert because at least three members of the plurality and the four dissenting Justices rejected an intent- based approach. Id. at 884–85. Troubled by this disagreement, the court concluded

37

that Seibert is not amenable to a Marks analysis.7 Id. Falling in stride with Heron, the Sixth Circuit reached the same conclusion for the same reason. United States v. Ray, 803 F.3d 244, 270–72 (6th Cir. 2015) (holding that Seibert did not produce a binding holding and adopting the Seibert plurality opinion as the law of the circuit).

Under the logic of Heron and Ray, Justice Kennedy’s Seibert concurrence does not share a common denominator with, or cannot be viewed as a logical subset of, the plurality opinion because a majority of the Court disavowed consideration of the interrogating officer’s subjective intent. This seems entirely reasonable at first blush. After all, it is “usually a judicial decision’s reasoning—its ratio decidendi—that allows it to have life and effect in the disposition of future cases.” Ramos v. Louisiana, 140 S. Ct. 1390, 1404 (2020) (plurality opinion). What must give us pause, however, is that this novel approach to applying Marks—requiring a majority of the Justices to agree on a single underlying rationale before a binding rule can be discerned—cannot be squared with Marks itself.

7 Although the Seventh Circuit found Marks not workable in Heron, the court declined “to resolve once and for all what rule or rules governing two-step interrogations can be distilled from Seibert.” 564 F.3d at 885. Instead, the court applied both Justice Kennedy’s concurrence and the Seibert plurality opinion because they yielded the same result. Id. at 885–87. So while Heron appears to have created an intra-circuit split (and an inter-split circuit) on the precedential effect of Seibert, it is not clear exactly where the Seventh Circuit stands on the issue. See United States v. Lee, 618 F.3d 667, 678 (7th Cir. 2010) (“[T]wo tests have emerged from Seibert; this Court has yet to choose which test should govern.”). Two other circuit courts have avoided deciding the precedential effect of Seibert in cases in which the outcome would be the same under either Justice Kennedy’s concurrence or the plurality opinion. United States v. Straker, 800 F.3d 570, 617 (D.C. Cir. 2015); United States v. Widi, 684 F.3d 216, 221 (1st Cir. 2012).

38

Marks applied its eponymous rule in the context of a due process challenge to an allegedly ex post facto criminal punishment for transporting obscene materials. 430 U.S. at 189–91. At issue was the precedential effect of the Court’s prior splintered decision in Memoirs v. Massachusetts, 383 U.S. 413 (1966). In Memoirs, a three-Justice plurality concluded that sexually explicit literature was protected under the First Amendment unless it satisfied the three-part definition of obscenity laid out in Roth v. United States, 354 U.S. 476 (1957).8 Memoirs, 383 U.S. at 418– 19. Justices Black and Douglas concurred in the judgment, but they took the position that material could never be constitutionally banned as obscene. Id. at 421 (Black, J.); id. at 424–33 (Douglas, J.); see also Marks, 430 U.S. at 193 (explaining that Justices Black and Douglas stated “their well-known position that the First Amendment provides an absolute shield against governmental action aimed at suppressing obscenity”). Justice Stewart also concurred based on his view that only hardcore pornography could be suppressed. Memoirs, 383 U.S. at 421 (citing his dissenting opinions in Mishkin v. New York, 383 U.S. 502, 518 (1966), and Ginzburg v. United States, 383 U.S. 463, 498–99 (1966)).

8 The proper test, according to the three-Justice Memoirs plurality, was “[w]hether to the average person, applying contemporary community standards, the dominant theme of the material taken as a whole appeals to prurient interest. . . . Under this definition, as elaborated in subsequent cases, three elements must coalesce: it must be established that (a) the dominant theme of the material taken as a whole appeals to a prurient interest in sex; (b) the material is patently offensive because it affronts contemporary community standards relating to the description or representation of sexual matters; and (c) the material is utterly without redeeming social value.” 383 U.S. at 418 (cleaned up).

39

In sum, six Justices agreed that the obscenity conviction in Memoirs could not stand, but there was no majority agreement on the appropriate First Amendment standard to apply or the proper reasoning to reach that result. Marks, 430 U.S. at 192 (recognizing “the Memoirs standards never commanded the assent of more than three Justices”). Yet Marks held that the opinion of the three-Justice Memoirs plurality (which applied the Roth tests) “constituted the holding of the Court and provided the governing standards.” Id. at 194. That’s because in every case in which the Memoirs plurality would deem an obscenity prosecution unconstitutional under their test, Justices Black and Douglas (who ascribe to the broader view that obscene speech is always constitutionally protected) would necessarily agree with the plurality’s result.9

9 This commonsensical approach to applying Marks also explains how a holding could be derived from Furman v. Georgia, 408 U.S. 238 (1972), and applied in Gregg v. Georgia, 428 U.S. 153 (1976), in which the Marks rule originated. In Furman, five Justices agreed that the death penalty statutes at issue were unconstitutional—though for different reasons. Justices Brennan and Marshall concurred based on their view that the death penalty is always unconstitutional under the Eighth Amendment. 408 U.S. at 305–06 (Brennan, J.); id. at 369–71 (Marshall, J.). The other three Justices who concurred in the judgments likewise deemed the statutes invalid under the Eighth Amendment, but they left open the question whether the death penalty may ever be imposed. Justices Stewart and White concluded that the statutes violated the Eighth Amendment because they permitted capital punishment to be imposed arbitrarily and capriciously. Id. at 306 (Stewart, J.); id. at 310–11 (White, J.). For Justice Douglas, the statutes were unconstitutional because they applied disproportionately against minorities and the poor due to their discretionary aspect and the ability of wealthier defendants to obtain superior counsel. Id. at 255–57 (Douglas, J.).

Although no five Justices agreed on a single underlying rationale, the Gregg plurality concluded that the position taken by Justices Stewart and White was the narrowest grounds for the judgments in Furman and constituted the Court’s holding. Gregg, 428 U.S. at 169 n.15 (plurality opinion). Again, that’s because cases involving death penalty statutes that would be deemed impermissible under Justice Stewart and Justice White’s approach represent the universe of cases in which the rationales of each

40

See King, 950 F.2d at 781 (“Because Justices Black and Douglas had to agree, as a logical consequence of their own position, with the plurality’s view that anything with redeeming social value is not obscene, the plurality of three in effect spoke for five Justices: Marks’ ‘narrowest grounds’ approach yielded a logical result.”).

In other words, the Marks rule does not require agreement as to both the result and the reasoning of Justices in the majority to produce a determinate holding. If that were not so, Marks itself would be wrongly decided. There was no rationale common to a majority of Justices in Memoirs—the Memoirs plurality disavowed Justice Douglas and Justice Black’s view that the First Amendment absolutely prohibits suppressing obscenity—yet the Marks Court derived a binding rule from Memoirs. The Marks rule works the same way here. Justice Kennedy’s Seibert opinion is a logical subset of the plurality’s not because it agrees with all of the plurality’s reasoning but because in every case in which it would suppress a statement given after midstream Miranda warnings, the plurality opinion would reach the same result.

This is, of course, not to say that the respective rationales of a splintered Supreme Court decision are irrelevant. The way we identify the “narrowest grounds” of decision is by identifying differences in the reasoning of the split opinions necessary to the judgment. Only then can we determine which competing rationale

Justice whose vote was necessary to the judgments in Furman would point to the same result. “Selecting the opinions of Justices Stewart and White as the holding of Furman in Gregg was thus unproblematic.” King, 950 F.2d at 781.

41

(if any) is a logical subset of the other(s), thus yielding outcome convergence, such that it controls. But a fundamental divergence in the legal reasoning of a concurrence and the legal reasoning of a plurality opinion does not “necessarily destroy a subset- superset relationship between the two opinions.” Kratt, 579 F.3d at 562. Otherwise, “Marks would stand for little,” id., as there are multiple opinions in splintered cases “precisely because the Justices did not agree on a common rationale,” Duvall, 740 F.3d at 613 (Kavanaugh, J., concurring in the denial of rehearing en banc). The Supreme Court’s most recent application of Marks confirms this conclusion.

In Glossip v. Gross, 576 U.S. 863 (2015), the Court held that the three-Justice plurality opinion in Baze v. Rees, 553 U.S. 35 (2008), was controlling under Marks. Glossip, 576 U.S. at 879 n.2. The Baze plurality concluded that a challenged method of execution is cruel and unusual only if it poses an “objectively intolerable risk of harm” (i.e., a “substantial risk of serious harm”) in the face of a readily available alternative procedure. 553 U.S. at 50 (plurality opinion) (internal quotation marks omitted). Justices Scalia and Thomas concurred in the judgment upholding Kentucky’s lethal injection procedures, but they said a method of execution violates the Eighth Amendment only when it is “deliberately designed to inflict pain.” Id. at 94 (Thomas, J.). They also expressly rejected the plurality’s approach as fundamentally inconsistent with the original understanding of the Cruel and Unusual Punishments Clause, unsupported by Eighth Amendment precedent, and impracticable for lower courts to apply. Id. at 94, 103–05. As the opening line of Justice Thomas’s concurrence, which Justice Scalia joined, put it: “I write separately

42

because I cannot subscribe to the plurality opinion’s formulation of the governing standard.” Id. at 94.

A few years later, in Glossip, the Court applied Marks and clarified that the three-Justice plurality opinion in Baze set out the holding of the case because Justices Scalia and Thomas took the broader position. Glossip, 576 U.S. at 879 n.2; see also Bucklew v. Precythe, 139 S. Ct. 1112, 1121 (2019) (reaffirming that the Baze plurality opinion is controlling under Marks). Justices Thomas and Scalia reiterated their view that the Eighth Amendment “prohibits only those methods of execution that are deliberately designed to inflict pain,” but this time they joined the Court’s majority opinion because it correctly explained why the petitioners’ claim failed even under the “controlling opinion” in Baze. Glossip, 576 U.S. at 899–900 (Thomas J., concurring) (brackets omitted). Notably, the four dissenting Justices in Glossip argued that the Baze plurality opinion was not controlling under Marks because “none of the Members of the Court whose concurrences were necessary to sustain the Baze Court’s judgment articulated a similar view.” Id. at 971 (Sotomayor, J., dissenting).

The Glossip dissent was right about one thing: there was no rationale common to a majority of the Justices in Baze. Obviously, there is a fundamental distinction between legal reasoning that turns on whether pain is a product of some subjective deliberate design (Justice Thomas and Justice Scalia’s approach) and legal reasoning that looks objectively into whether there is a substantial risk of harm (the Baze plurality’s approach). Yet Glossip held that the opinion of the three-Justice Baze

43

plurality constituted the holding of the Court and provided the legal standard that governs Eighth Amendment method-of-execution challenges. That’s because— granting the reasonable assumption that an execution method deliberately designed to inflict unnecessary pain would also pose a substantial risk of inflicting such pain— the Baze plurality’s standard and the alternative approach endorsed by Justices Thomas and Scalia would necessarily point to the same result in any case where the plurality would find an execution method constitutionally permissible. Jackson v. Danberg, 594 F.3d 210, 222 (3d Cir. 2010) (holding the Baze plurality opinion controls under Marks because “any lethal injection protocol constitutionally acceptable to the plurality would invariably pass” Justice Thomas and Justice Scalia’s standard); see also Warner v. Gross, 776 F.3d 721, 729 n.6 (10th Cir. 2015) (citing Jackson and reaching the same conclusion about the holding of Baze).

From where we sit, the most definitive source of guidance as to the Marks rule’s application is the Supreme Court’s own application of the rule. The Court is, after all, not only the final arbiter of constitutional questions but also the inventor of the Marks rule. Thus, even without conscious or implicit agreement on an underlying rationale to reach a decision—indeed, even in the face of express disagreement—a concurring opinion can still reflect the narrowest grounds of the decision such that it controls under Marks. If that were not so, Glossip would be wrongly decided.

No matter how vigorously the Seibert plurality (and the dissent) may disagree with Justice Kennedy’s intent-based approach, his standard will produce results the plurality would necessarily agree with in any case suppressing statements given after

44

midstream Miranda warnings. Justice Kennedy’s opinion in Seibert therefore constitutes the holding of the case and provides the governing standard here. While there is admitted awkwardness in treating as precedential an opinion no other Justice joined, this is the settled practice when that opinion is the determinative one. See Marks, 430 U.S. at 193; Glossip, 576 U.S. at 879 n.2. And until the Supreme Court itself chooses to reconsider Marks, we cannot abandon it from below.

c.

Justice Kennedy’s Seibert concurrence provides the substantive standard that controls the outcome here, but his opinion is silent as to which party bears the burden of proving or disproving deliberateness and what that burden should be. Ethan argues that we should follow the lead of our sister circuits and require the government to prove that the agents’ failure to provide the requisite warnings was not part of a deliberate two-step interrogation strategy used to circumvent Miranda. See Capers, 627 F.3d at 479; Stewart, 536 F.3d at 719; Ollie, 442 F.3d at 1142–43. The government does not disagree with Ethan’s suggestion, and neither do we.

Although “the law generally frowns on requiring a party to prove a negative,”

there are several reasons why it is appropriate to burden the government with disproving deliberateness. Ollie, 442 F.3d at 1143. Placing that burden on the government is consistent with Supreme Court precedent requiring the government to prove by a preponderance of the evidence that the defendant waived his Miranda rights and voluntarily confessed before a confession may come into evidence. Colorado v. Connelly, 479 U.S. 157, 168–69 (1986) (waiver); Lego v. Twomey, 404

45

U.S. 477, 489 (1972) (voluntariness of confession). Putting the burden on the government to disprove deliberateness also makes sense as a practical matter because it will help ensure probative evidence, all or most of which will typically be in the government’s hands, is brought to the court’s attention. Ollie, 442 F.3d at 1143; Capers, 627 F.3d at 479. Accordingly, in cases involving sequential unwarned and warned confessions, the government bears the burden of proving by a preponderance of the evidence that the interrogating officer did not deliberately withhold the requisite warnings as part of a calculated strategy to foil Miranda.

One other word on applying Seibert. The question whether the interrogating officer made a conscious decision to withhold warnings as part of a deliberate strategy to circumvent Miranda will, of course, often turn on subjective evidence of the officer’s intent. But deliberateness may also be inferred from objective indications of subjective intent to frustrate Miranda. In Seibert, for example, the officer confronted the defendant with her inadmissible prewarning statements and pushed her to repeat them during the postwarning interview, which showed the deliberate two-step strategy at work. 542 U.S. at 621 (Kennedy, J., concurring in the judgment) (“The postwarning interview resembled a cross-examination.”); cf. Elstad, 470 U.S. at 316 (observing that the officers did not “exploit the unwarned admission”). Conversely, the lack of overlap between warned and unwarned confessions is evidence that the interrogating officer did not intentionally withhold the requisite warnings to undermine the suspect’s Miranda rights. Stewart, 536 F.3d at 722.

46

In sum, the standard for assessing the admissibility of confessions given subsequent to midstream Miranda warnings can be applied using a straightforward analysis: First, was the initial self-incriminating statement, though voluntary, obtained in violation of the defendant’s Miranda rights? If not, there is no need to go further. Cortez, 965 F.3d at 840. If the initial statement was obtained in violation of the defendant’s Miranda rights, has the government shown by a preponderance of the evidence that the questioning officer(s) did not engage in a deliberate two-step interrogation calculated to undermine Miranda? If so, the defendant’s postwarning statement is admissible so long as it, too, was voluntary. If not, the postwarning statement is inadmissible unless the objective circumstances show that a reasonable person in the defendant’s shoes would understand the import and effect of the midstream Miranda warnings.

d.

Turning back to Ethan’s appeal, we already concluded that the agents obtained his initial self-incriminating statement in violation of his Miranda rights. But a review of the evidence also leads us to conclude that the agents did not use a two-step interrogation technique “in a calculated way to undermine the Miranda warning.” Seibert, 542 U.S. at 622 (Kennedy, J., concurring in the judgment).

There is no subjective evidence of intent here—no testimony, for example, by the interrogating agents suggesting they intentionally withheld Miranda warnings to

47

deprive Ethan of his rights.10 Agent Rominger, the lead investigator who was responsible for the bulk of the interrogation, testified that he did not believe Ethan was in custody until his initial confession. But once Ethan admitted he made the pressure cooker bomb, Agent Rominger explained, he knew Ethan “was no longer free to leave” and believed Ethan would recognize as much. At that point—as soon as Agent Rominger believed Ethan was in custody—he immediately provided Miranda warnings.

Ethan attacks Agent Rominger’s explanation for deferring Miranda warnings—that he mistakenly believed Ethan was not in custody until he confessed— as unworthy of belief because the agent’s questions were crafted to elicit incriminating responses. Although Agent Rominger certainly sought a confession when he confronted Ethan with the evidence discovered during the search (i.e., engaged in interrogation for Miranda purposes), that does not undercut the agent’s subjective belief about Ethan’s custodial status. Cf. United States v. Chee, 514 F.3d 1106, 1113–14 (10th Cir. 2008) (concluding that the defendant’s “contention that he had to be given Miranda warnings once the investigative process moved to the point where [the agent] was trying to obtain a confession is simply incorrect”). The “task

10 As evidence of deliberateness, Ethan points to Detective Larranaga’s testimony that, in his experience, providing Miranda warnings can result in the suspect requesting an attorney, ending an interview, or refusing to talk further. But Detective Larranaga did not participate in the pre-Miranda questioning that resulted in Ethan’s initial confession. His thoughts therefore do not speak to the intent of the interrogating agents. Stewart, 536 F.3d at 722 (“The determination of whether a question-first strategy was deliberately used does not require an inquiry into the state of mind of every officer involved in the interrogation.”).

48

of defining custody is a slippery one, and policemen investigating serious crimes cannot realistically be expected to make no errors whatsoever.” Elstad, 470 U.S. at 309 (cleaned up). For that reason, one plausible explanation why an officer might legitimately wait to deliver Miranda warnings is that the “officer may not realize that a suspect is in custody and warnings are required.” Seibert, 542 U.S. at 620 (Kennedy, J., concurring in the judgment).

That is the plausible explanation here for the agents’ failure to administer Miranda warnings in a timely fashion. Unlike Seibert, Ethan was not arrested, taken to a police station, and questioned in an interrogation room. Id. at 604–05 (plurality opinion). Ethan’s interview began as consensual questioning in his home, and it did not turn into a custodial interrogation until Agent Rominger confronted him with the evidence discovered during the search. The difficulty in pinpointing the moment when Ethan’s interview became custodial supports a conclusion that the agents did not engage in a deliberate two-step interrogation to thwart Miranda. This conclusion is confirmed by the lack of any objective indicia that the agents set out to intentionally circumvent or undermine the protections the Miranda warnings provide.

Agent Rominger did not withhold Miranda warnings, solicit a full confession, and lead Ethan back through his confession again. Ethan’s prewarning confession was limited to his admission that he made the pressure cooker bomb. His postwarning confession was not just a repetition of that admission but contained far more details about, among other things, how he made the bomb, planted it under MC’s bed, and wanted MC dead. Had the agents intended to obtain a damning

49

confession first, give Miranda warnings, and then re-obtain the confession, as the officers did in Seibert, they would have asked about those incriminating details much earlier. The agents did not do so. Nor did they use Ethan’s initial admission to cross-examine or pressure him to answer their questions during the postwarning interrogation.

For these reasons, the government has met its burden of showing the agents did not engage in a deliberate two-step interrogation strategy to frustrate Miranda. That means the admissibility of Ethan’s postwarning statements is governed by the voluntariness principles of Elstad. Seibert, 542 U.S. at 622 (Kennedy, J., concurring in the judgment). So long as Ethan knowingly and voluntarily waived his Miranda rights and made both his prewarning and postwarning statements voluntarily, his postwarning statements are admissible. Elstad, 470 U.S. at 309, 318.

To be valid, a waiver of Miranda rights “must be made voluntarily, knowingly, and intelligently.” United States v. Smith, 606 F.3d 1270, 1276 (10th Cir. 2010). When assessing the voluntariness of a Miranda waiver, we look at: “[1] the suspect’s age, intelligence, and education; [2] whether the suspect was informed of his or her rights; [3] the length and nature of the suspect’s detention and interrogation; and [4] the use or threat of physical force against the suspect.” Id. We consider the same factors when determining whether an incriminating statement was voluntarily given. Id.; Carrizales-Toledo, 454 F.3d at 1153. “The essence of voluntariness is whether the government obtained the statements by physical or psychological coercion such

50

that the defendant’s will was overborne.” Carrizales-Toledo, 454 F.3d at 1153 (quoting Rith, 164 F.3d at 1333).

Ethan was an 18-year-old high school senior at the time of his arrest. Smith, 606 F.3d at 1277 (holding that the defendant, who was “nearly 21 years of age and had completed the tenth grade” validly waived his Miranda rights and voluntarily confessed). Although he was not on track to graduate that year, this clearly was not due to a lack of intelligence or education that would render him susceptible to coercion. Ethan taught himself how to build a sophisticated explosive device from homemade materials, built his own computer, and “displayed his fortitude” when he asked the agents if they had a warrant to enter his home. See Rith, 164 F.3d at 1333. The agents questioned Ethan in his home; he was not subjected to any physical punishments or threats; the interview was not unduly prolonged; and the agents’ questioning, though accusatory at one point, was done in a conversational tone and never rose to the level of coercion. Considering the relevant factors in light of the totality of the circumstances, there is no doubt that both Ethan’s prewarning and postwarning statements were voluntary.

Based on those same considerations, we conclude that Ethan voluntarily, knowingly, and intelligently waived his Miranda rights. According to Agent Rominger’s testimony, which the district court deemed credible, Ethan indicated he understood his Miranda rights and continued to speak with the agents. Nothing more was required for Ethan’s waiver to be valid. Berghuis v. Thompkins, 560 U.S. 370, 385 (2010) (“[T]he law can presume that an individual who, with a full understanding

51

of his or her rights, acts in a manner inconsistent with their exercise has made a deliberate choice to relinquish the protection those rights afford.”).

Because Ethan’s initial confession was voluntary, the voluntary statements he made after he received the Miranda warnings and waived his right to remain silent were properly admitted. The district court therefore did not err when it denied Ethan’s motion to suppress his postwarning statements.

III.

For all the reasons we have given, the judgment of the district court is AFFIRMED.