HOWES, WARDEN v. FIELDS, 565 U.S. 499

Justice Alito delivered the opinion of the Court. The United States Court of Appeals for the Sixth Circuit held that our precedents clearly establish that a prisoner is in custody within the meaning of Miranda v. Arizona, 384 U. S. 436 (1966), if the prisoner is taken aside and questioned about events that occurred outside the prison walls.

Our decisions, however, do not clearly establish such a rule, and therefore the Court of Appeals erred in holding that this rule provides a permissible basis for federal habeas relief under the relevant provision of the Antiterrorism and Effective Death Penalty Act of 1996 (AEDPA), 28 U. S. C. § 2254(d)(1). Indeed, the rule applied by the court below does not represent a correct

interpretation of our Miranda case law. We therefore reverse. I While serving a sentence in a Michigan jail, Randall Fields was escorted by a corrections officer to a conference room where two sheriff’s deputies questioned him about allegations that, before he came to prison, he had engaged in sexual conduct with a 12-year-old boy. In order to get to the conference room, Fields had to go down one floor and pass through a locked door that separated two sections of the fa-eility. See App. to Pet. for Cert. 66a, 69a.

Fields arrived at the conference room between 7 p.m. and 9 p.m. and was questioned for between five and seven hours. At the beginning of the interview, Fields was told that he was free to leave and return to his cell. See id., at 70a. Later, he was again.told that he could leave whenever he wanted. See id., at 90a. The two interviewing deputies were armed during the interview, but Fields remained free of handcuffs and other restraints. The door to the conference room was sometimes open and sometimes shut. See id., at 70a-75a.

About halfway through the interview, after Fields had been confronted with the allegations of abuse, he became agitated and began to yell. See id., at 80a, 125a. Fields testified that one of the deputies, using an expletive, told him to sit down and said that “if [he] didn’t want to cooperate, [he] could leave.” Id., at 89a; see also id., at 70a-71a. Fields eventually confessed to engaging in sex acts with the boy.

According to Fields’ testimony at a suppression hearing, he said several times during the interview that he no longer wanted to talk to the deputies, but he did not ask to go back to his cell prior to the end of the interview. See id., at 92a-93a. When he was eventually ready to leave, he had to wait an additional 20 minutes or so because a corrections officer had to be summoned to escort him back to his cell, and he did not return to his cell until well after the hour when he generally retired. At no time was Fields given Miranda warnings or advised that he did not have to speak with the deputies.

The State of Michigan charged Fields with criminal sexual conduct. Relying on Miranda, Fields moved to suppress his confession, but the trial court denied his motion. Over the renewed objection of defense counsel, one of the interviewing deputies testified at trial about Fields’ admissions. The jury convicted Fields of two counts of third-degree criminal sexual conduct, and the judge sentenced him to a term of 10 to 15 years of imprisonment.

On direct appeal, the Michigan Court of Appeals affirmed, rejecting Fields’ contention that his statements should have been suppressed because he was subjected to custodial interrogation without a Miranda warning. The court ruled that Fields had not been in custody for purposes of Miranda during the interview, so no Miranda warnings were required. The court emphasized that Fields was told that he was free to leave and return to his cell but that he never asked to do so. The Michigan Supreme Court denied discretionary review.

Fields then filed a petition for a writ of habeas corpus in Federal District Court, and the court granted relief. The Sixth Circuit affirmed, holding that the interview in the conference room was a “custodial interrogation” within the meaning of Miranda because isolation from the general prison population combined with questioning about conduct occurring outside the prison makes any such interrogation custodial per se. The Court of Appeals reasoned that this Court clearly established in Mathis v. United States, 391 U.

S. 1 (1968), that “Miranda warnings must be administered when law enforcement officers remove an inmate from the general prison population and interrogate him regarding criminal conduct that took place outside the jail or prison.” 617 F. 3d 813, 820 (CA6 2010); see also id., at 818 (“The central holding of Mathis is that a Miranda warning is required whenever an incarcerated individual is isolated from the general prison population and interrogated, i. e.[,] questioned in a manner likely to lead to self-incrimination, about conduct occurring outside of the prison”).

Because Fields was isolated from the general prison population and interrogated about conduct occurring in the outside world, the Court of Appeals found that the state court’s decision was contrary to clearly established federal law as determined by this Court in Mathis. 617 F. 3d, at 823. We granted certiorari. 562 U. S. 1199 (2011).

II Under AEDPA, a federal court may grant a state prisoner’s application for a writ of habeas corpus if the state-court adjudication pursuant to which the prisoner is held “resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States.” 28 U. S. C. § 2254(d)(1). In this context, “clearly established law” signifies “the holdings, as opposed to the dicta, of this Court’s decisions.” Williams v. Taylor, 529 U. S. 362, 412 (2000).

In this case, it is abundantly clear that our precedents do not clearly establish the categorical rule on which the Court of Appeals relied, i. e., that the questioning of a prisoner is always custodial when the prisoner is removed from the general prison population and questioned about events that occurred outside the prison. On the contrary, we have repeatedly declined to adopt any categorical rule with respect to whether the questioning of a prison inmate is custodial. In Illinois v. Perkins, 496 U.

S. 292 (1990), where we upheld the admission of un-Mirandized statements elicited from an inmate by an undercover officer masquerading as another inmate, we noted that “[t]he bare fact of custody may not in every instance require a warning even when the suspect is aware that he is speaking to an official, but we do not have occasion to explore that issue here.” Id., at 299 (emphasis added). Instead, we simply “rejected] the argument that Miranda warnings are required whenever a suspect is in custody in a technical sense and converses with someone who happens to be a government agent.” Id., at 297.

Most recently, in Maryland v. Shatzer, 559 U. S. 98 (2010), we expressly declined to adopt a bright-line rule for determining the applicability of Miranda in prisons. Shatzer considered whether a break in custody ends the presumption of involuntariness established in Edwards v. Arizona, 451 U. S. 477 (1981), and, if so, whether a prisoner’s return to the general prison population after a custodial interrogation constitutes a break in Miranda custody. See 559 U. S., at 102-103.

In considering the latter question, we noted first that “[w]e have never decided whether incarceration constitutes custody for Miranda purposes, and have indeed explicitly declined to address the issue.” Id., at 112 (citing Perkins, supra, at 299; emphasis added). The answer to this question, we noted, would “depen[d] upon whether [incarceration] exerts the coercive pressure that Miranda was designed to guard against — the ‘danger of coercion [that] results from the interaction of custody and official interrogation.’” 559 U. S., at 112 (quoting Perkins, supra, at 297).

In concluding that our precedents establish a categorical rule, the Court of Appeals placed great weight on the decision in Mathis, but the Court of Appeals misread the holding in that case. In Mathis, an inmate in a state prison was questioned by an Internal Revenue agent and was subsequently convicted for federal offenses. The Court of Appeals held that Miranda did not apply to this interview for two reasons: A criminal investigation had not been commenced at the time of the interview, and the prisoner was incarcerated for an “unconnected offense.” Mathis v. United States, 376 F. 2d 595, 597 (CA5 1967).

This Court rejected both of those grounds for distinguishing Miranda, see 391 U. S., at 4, and thus the holding in Mathis is simply that a prisoner who otherwise meets the requirements for Miranda custody is not taken outside the scope of Miranda by either of the two factors on which the Court of Appeals had relied. Mathis did not hold that imprisonment, in and of itself, is enough to constitute Miranda custody. Nor, contrary to respondent’s submission, see Brief for Respondent 14, did Oregon v. Mathiason, 429 U. S. 492, 494 (1977) (per curiam), which simply restated in dictum the holding in Mathis.

The Court of Appeals purported to find support for its per se rule in Shatzer, relying on our statement that “[n]o one questions that Shatzer was in custody for Miranda purposes” when he was interviewed. 559 U. S., at 112. But this statement means only that the issue of custody was not contested before us. It strains credulity to read the statement as constituting an “unambiguous conclusion” or “finding” by this Court that Shatzer was in custody. 617 F. 3d, at 822.

Finally, contrary to respondent’s suggestion, see Brief for Respondent 12-15, Miranda itself did not clearly establish the rule applied by the Court of Appeals. Miranda adopted a “set of prophylactic measures” designed to ward off the “‘inherently compelling pressures’ of custodial interrogation,” Shatzer, supra, at 103 (quoting Miranda, 384 U. S., at 467), but Miranda did not hold that such pressures are always present when a prisoner is taken aside and questioned about events outside the prison walls.

Indeed, Miranda did not even establish that police questioning of a suspect at the station house is always custodial. See Mathiason, supra, at 495 (declining to find that Miranda warnings are required “simply because the questioning takes place in the station house, or because the questioned person is one whom the police suspect”). In sum, our decisions do not clearly establish that a prisoner is always in custody for purposes of Miranda whenever a prisoner is isolated from the general prison population and questioned about conduct outside the prison.

I — 1 b-i hH Not only does the categorical rule applied below go well beyond anything that is clearly established in our prior decisions, it is simply wrong. The three elements of that rule— (1) imprisonment, (2) questioning in private, and (3) questioning about events in the outside world — are not necessarily enough to create a custodial situation for Miranda purposes. A As used in our Miranda case law, “custody” is a term of art that specifies circumstances that are thought generally to present a serious danger of coercion.

In determining whether a person is in custody in this sense, the initial step is to ascertain whether, in light of “the objective circumstances of the interrogation,” Stansbury v. California, 511 U. S. 318, 322-323, 325 (1994) (per curiam), a “reasonable person [would] have felt he or she was not at liberty to terminate the interrogation and leave.” Thompson v. Keo-hane, 516 U. S. 99, 112 (1995).

And in order to determine how a suspect would have “gauge[d]” his “freedom of movement,” courts must examine “all of the circumstances surrounding the interrogation.” Stansbury, supra, at 322, 325 (internal quotation marks omitted). Relevant factors include the location of the questioning, see Shatzer, supra, at 112-114, its duration, see Berkemer v. McCarty, 468 U. S. 420, 437-438 (1984), statements made during the interview, see Mathiason, supra, at 495; Yarborough v. Alvarado, 541 U.

S. 652, 665 (2004); Stansbury, supra, at 325, the presence or absence of physical restraints during the questioning, see New York v. Quarles, 467 U. S. 649, 655 (1984), and the release of the interviewee at the end of the questioning, see California v. Beheler, 463 U. S. 1121, 1122-1123 (1983) (per curiam). Determining whether an individual's freedom of movement was curtailed, however, is simply the first step in the analysis, not the last. Not all restraints on freedom of movement amount to custody for purposes of Miranda.

We have “decline[d] to accord talismanic power” to the freedom-of-movement inquiry, Berkemer, supra, at 437, and have instead asked the additional question whether the relevant environment presents the same inherently coercive pressures as the type of station house questioning at issue in Miranda. “Our cases make clear... that the freedom-of-movement test identifies only a necessary and not a sufficient condition for Miranda custody.” Shatzer, supra, at 112. This important point is illustrated by our decision in Ber-kemer v. McCarty, supra.

In that case, we held that the roadside questioning of a motorist who was pulled over in a routine traffic stop did not constitute custodial interrogation. Id., at 423, 441-442.

We acknowledged that “a traffic stop significantly curtails the ‘freedom of action’ of the driver and the passengers,” and that it is generally “a crime either to ignore a policeman’s signal to stop one’s car or, once having stopped, to drive away without permission.” Id., at 436. “[F]ew motorists,” we noted, “would feel free either to disobey a directive to pull over or to leave the scene of a traffic stop without being told they might do so.” Ibid.

Nevertheless, we held that a person detained as a result of a traffic stop is not in Miranda custody because such detention does not “sufficiently impair [the detained person’s] free exercise of his privilege against self-incrimination to require that he be warned of his constitutional rights.” 468 U. S., at 437. As we later put it, the “temporary and relatively nonthreatening detention involved in a traffic stop or Terry stop does not constitute Miranda custody,” Shatzer, 559 U. S., at 113 (citation omitted). See Terry v. Ohio, 392 U. S. 1 (1968).

It may be thought that the situation in Berkemer — the questioning of a motorist subjected to a brief traffic stop — is worlds away from those present when an inmate is questioned in a prison, but the same cannot be said of Shatzer, where we again distinguished between restraints on freedom of movement and Miranda custody. Shatzer, as noted, concerned the Edwards prophylactic rule, which limits the ability of the police to initiate further questioning of a suspect in Miranda custody once the suspect invokes the right to counsel.

We held in Shatzer that this rule does not apply when there is a sufficient break in custody between the suspect’s invocation of the right to counsel and the initiation of subsequent questioning. See 559 U. S., at 112-114. And, what is significant for present purposes, we further held that a break in custody may occur while a suspect is serving a term in prison. If a break in custody can occur while a prisoner is serving an uninterrupted term of imprisonment, it must follow that imprisonment alone is not enough to create a custodial situation within the meaning of Miranda.

There are at least three strong grounds for this conclusion. First, questioning a person who is already serving a prison term does not generally involve the shock that very often accompanies arrest. In the paradigmatic Miranda situation— a person is arrested in his home or on the street and whisked to a police station for questioning — detention represents a sharp and ominous change, and the shock may give rise to coercive pressures. A person who is “cut off from his normal life and companions,” Shatzer, supra, at 106, and abruptly transported from the street into a “police-dominated atmosphere,” Miranda, 384 U.

S., at 466, may feel coerced into answering questions. By contrast, when a person who is already serving a term of imprisonment is questioned, there is usually no such change. “Interrogated suspects who have previously been convicted of crime live in prison.” Shatzer, 559 U. S., at 113.

For a person serving a term of incarceration, we reasoned in Shatzer, the ordinary restrictions of prison life, while no doubt unpleasant, are expected and familiar and thus do not involve the same “inherently compelling pressures” that are often present when a suspect is yanked from familiar surroundings in the outside world and subjected to interrogation in a police station. Id., at 103 (internal quotation marks omitted). Second, a prisoner, unlike a person who has not been sentenced to a term of incarceration, is unlikely to be lured into speaking by a longing for prompt release.

When a person is arrested and taken to a station house for interrogation, the person who is questioned may be pressured to speak by the hope that, after doing so, he will be allowed to leave and go home. On the other hand, when a prisoner is questioned, he knows that when the questioning ceases, he will remain under confinement. Id., at 124, n. 8. Third, a prisoner, unlike a person who has not been convicted and sentenced, knows that the law enforcement officers who question him probably lack the authority to affect the duration of his sentence. Id., at 113-114.

And “where the possibility of parole exists,” the interrogating officers probably also lack the power to bring about an early release. Ibid. “When the suspect has no reason to think that the listeners have official power over him, it should not be assumed that his words are motivated by the reaction he expects from his listeners.” Perkins, 496 U. S., at 297. Under such circumstances, there is little “basis for the assumption that a suspect . . . will feel compelled to speak by the fear of reprisal for remaining silent or in the hope of [a] more lenient treatment should he confess.” Id., at 296-297.

In short, standard conditions of confinement and associated restrictions on freedom will not necessarily implicate the same interests that the Court sought to protect when it afforded special safeguards to persons subjected to custodial interrogation. Thus, service of a term of imprisonment, without more, is not enough to constitute Miranda custody. B The two other elements included in the Court of Appeals’ rule — questioning in private and questioning about events that took place outside the prison — are likewise insufficient.

Taking a prisoner aside for questioning — as opposed to questioning the prisoner in the presence of fellow inmates— does not necessarily convert a “noncustodial situation ... to one in which Miranda applies.” Mathiason, 429 U. S., at 495. When a person who is not serving a prison term is questioned, isolation may contribute to a coercive atmosphere by preventing family members, friends, and others who may be sympathetic from providing either advice or emotional support.

And without any such assistance, the person who is questioned may feel overwhelming pressure to speak and to refrain from asking that the interview be terminated. By contrast, questioning a prisoner in private does not generally remove the prisoner from a supportive atmosphere. Fellow inmates are by no means necessarily friends. On the contrary, they may be hostile and, for a variety of reasons, may react negatively to what the questioning reveals.

In the present case, for example, would respondent have felt more at ease if he had been questioned in the presence of other inmates about the sexual abuse of an adolescent boy? Isolation from the general prison population is often in the best interest of the interviewee and, in any event, does not suggest on its own the atmosphere of coercion that concerned the Court in Miranda. It is true that taking a prisoner aside for questioning may necessitate some additional limitations on his freedom of movement.

A prisoner may, for example, be removed from an exercise yard and taken, under close guard, to the room where the interview is to be held. But such procedures are an ordinary and familiar attribute of life behind bars. Escorts and special security precautions may be standard procedures regardless of the purpose for which an inmate is removed from his regular routine and taken to a special location. For example, ordinary prison procedure may require such measures when a prisoner is led to a meeting with an attorney.

Finally, we fail to see why questioning about criminal activity outside the prison should be regarded as having a significantly greater potential for coercion than questioning under otherwise identical circumstances about criminal activity within the prison walls. In both instances, there is the potential for additional criminal liability and punishment.

If anything, the distinction would seem to cut the other way, as an inmate who confesses to misconduct that occurred within the prison may also incur administrative penalties, but even this is not enough to tip the scale in the direction of custody. “The threat to a citizen’s Fifth Amendment rights that Miranda was designed to neutralize” is neither mitigated nor magnified by the location of the conduct about which questions are asked. Berkemer, 468 U. S., at 435, n. 22. For these reasons, the Court of Appeals’ categorical rule is unsound.

IV A When a prisoner is questioned, the determination of custody should focus on all of the features of the interrogation. These include the language that is used in summoning the prisoner to the interview and the manner in which the interrogation is conducted. See Yarborough, 541 U. S., at 665. An inmate who is removed from the general prison population for questioning and is “thereafter ... subjected to treatment” in connection with the interrogation “that renders him ‘in custody’ for practical purposes . . . will be entitled to the full panoply of protections prescribed by Miranda.” Berkemer, 468 U.

S., at 440. “Fidelity to the doctrine announced in Miranda requires that it be enforced strictly, but only in those types of situations in which the concerns that powered the decision are implicated.” Id., at 437; see Shatzer, 559 U. S., at 108; Ma-thiason, supra, at 495.

Confessions voluntarily made by prisoners in other situations should not be suppressed. “Voluntary confessions are not merely a proper element in law enforcement, they are an unmitigated good, essential to society’s compelling interest in finding, convicting, and punishing those who violate the law.” Shatzer, supra, at 108 (internal quotation marks and citations omitted). B The record in this case reveals that respondent was not taken into custody for purposes of Miranda.

To be sure, respondent did not invite the interview or consent to it in advance, and he was not advised that he was free to decline to speak with the deputies. The following facts also lend some support to respondent’s argument that Miranda's custody requirement was met: The interview lasted for between five and seven hours in the evening and continued well past the hour when respondent generally went to bed; the deputies who questioned respondent were armed; and one of the deputies, according to respondent, “[u]sed a very sharp tone,” App. to Pet. for Cert. 76a, and, on one occasion, profanity, see id., at 77a.

These circumstances, however, were offset by others. Most important, respondent was told at the outset of the interrogation, and was reminded again thereafter, that he could leave and go back to his cell whenever he wanted. See id., at 89a-90a (“I was told I could get up and leave whenever I wanted”); id., at 70a-71a. Moreover, respondent was not physically restrained or threatened and was interviewed in a well-lit, average-sized conference room, where he was “not uncomfortable.” Id., at 90a; see id., at 71a, 88a-89a. He was offered food and water, and the door to the conference room was sometimes left open.

See id., at 70a, 74a. “All of these objective facts are consistent with an interrogation environment in which a reasonable person would have felt free to terminate the interview and leave.” Yarborough, supra, at 664-665. Because he was in prison, respondent was not free to leave the conference room by himself and to make his own way through the facility to his cell. Instead, he was escorted to the conference room and, when he ultimately decided to end the interview, he had to wait about 20 minutes for a corrections officer to arrive and escort him to his cell.

But he would have been subject to this same restraint even if he had been taken to the conference room for some reason other than police questioning; under no circumstances could he have reasonably expected to be able to roam free. And while respondent testified that he “was told ... if I did not want to cooperate, I needed to go back to my cell,” these words did not coerce cooperation by threatening harsher conditions. App. to Pet. for Cert. 71a; see id., at 89a (“I was told, if I didn’t want to cooperate, I could leave”). Returning to his cell would merely have returned him to his usual environment.

See Shatzer, supra, at 113 (“Interrogated suspects who have previously been convicted of crime live in prison. When they are released back into the general prison population, they return to their accustomed surroundings and daily routine — they regain the degree of control they had over their lives prior to the interrogation”).

Taking into account all of the circumstances of the questioning — including especially the undisputed fact that respondent was told that he was free to end the questioning and to return to his cell — we hold that respondent was not in custody within the meaning of Miranda. * * * The judgment of the Court of Appeals is Reversed. Fields testified that he left his cell around 8 p.m. and that the interview began around 8:30 p.m. App. to Pet. for Cert. 77a. Both the Michigan Court of Appeals and the Sixth Circuit stated that the interview began between 7 p.m. and 9 p.m.

See 617 F. 3d 813, 815 (2010); App. to Pet. for Cert. 54a. The Court of Appeals stated that the interview lasted for approximately seven hours, see 617 F. 3d, at 815, a figure that appears to be based on the testimony of one of the interviewing deputies, see App. to Pet. for Cert. 123a. Fields put the number of hours between five and five and a half, saying the interview began around 8:30 p.m. and continued until 1:30 a.m. or 2 a.m. See id., at 77a. The Michigan Court of Appeals stated that the interview ended around midnight, which would put the length of the interview at between three and five hours.

Fields testified that his normal bedtime was 10:30 p.m. or 11 p.m. See id., at 78a. Indeed, it is impossible to tell from either the opinion of this Court or that of the court below whether the prisoner’s interview was routine or whether there were special features that may have created an especially coercive atmosphere. The state-court decision applied the traditional context-specific analysis to determine whether the circumstances of respondent’s interrogation gave rise to “the coercive pressure that Miranda was designed to guard against.” Maryland v. Shatzer, 559 U. S. 98, 112 (2010).

The court first observed: “That a defendant is in prison for an unrelated offense when being questioned does not, without more, mean that he was in custody for the purpose of determining whether Miranda warnings were required.” App. to Pet. for Cert. 56a (internal quotation marks omitted; emphasis added). In this case, the court noted, the “defendant was unquestionably in custody, but on a matter unrelated to the interrogation.” Ibid. The Sixth Circuit concluded that the state court thereby limited Miranda in a way rejected by Mathis v. United States, 391 U.

S. 1 (1968), and “cur-tailfed] the warnings to be given persons under interrogation by officers based on the reason why the person is in custody.” Id., at 4-5. We think the better reading is that the state court merely meant to draw a distinction between incarceration and Miranda custody. This reading is supported by the state court’s subsequent consideration of whether the facts of the case were likely to create an atmosphere of coercion. App. to Pet. for Cert. 56a. Respondent did not testify to the contrary. The following colloquy occurred at his Miranda hearing: “Q.

You’re not generally allowed to just roam around Lenawee County Jail on your own, are you? “A. No, I never have. “Q. So wouldn’t it make sense to you, since you had that experience, that in fact you would have been escorted just like you were escorted ... into this conference room? “A. That makes common sense. “Q. So when they said that you were free to leave and you get up — could get up and go and all you had to do was tell them you wanted to go, in your mind, did you understand that to mean that somebody would come get you and take you back to your cell? “A.

But that doesn’t give me freedom to just get up and walk away. “Q. I understand it doesn’t— “A., So, no. “Q. The question is this, sir, not whether you had freedom to get up and walk away, but did you understand that what that meant was that a jailer would come get you and— “A. No— “Q. —take you back to your cell? “A. I did not understand that. “Q. You didn’t? “A. No. “Q. Why not? That’s how you got there. “A. Because I did not know if a jailer would take me back or if one of those gentlemen would take me back. “Q.

But you understood that, if you asked, one of them or a jailer would take you back to your cell? “A. I assumed that. “Q. And you believed that to be true? “A. I assumed that.” App. to Pet. for Cert. 91a-92a.