STATE OF OREGON, PETITIONER V. MICHAEL JAMES ELSTAD No. 83-773 In the Supreme Court of the United States October Term, 1983 On Writ of Certiorari To the Court of Appeals of the State of Oregon Brief For the United States As Amicus Curiae Supporting Petitioner TABLE OF CONTENTS Interest of the United States Statement Summary of argument Argument: Respondent's written confession should not be exluded from evidence A. Neither of respondent's statements was compelled B. Respondent validly waived his Miranda rights before making the second statement C. An otherwise admissible statement should not be excluded from evidence solely because it is the product of a statement obtained in violation of Miranda when there is no substantial probability that the first statement was compelled Conclusion QUESTION PRESENTED Whether respondent's voluntary and otherwise admissible confession must be excluded from evidence on the ground that it was the product of a prior voluntary statement by respondent that is inadmissible under Miranda v. Arizona, 384 U.S. 436 (1966). INTEREST OF THE UNITED STATES In this case a criminal defendant voluntarily made two successive inculpatory statements. Because the first statement was made under conditions that, according to a concession by the State, rendered it inadmissible under Miranda v. Arizona, 384 U.S. 436 (1966), the court below ordered the suppression of the second statement as well. This Court's analysis and resolution of the question whether the second statement is admissible is likely to have a significant effect on the admissibility of voluntary statements, and perhaps other evidence allegedly derived from a violation of Miranda, in federal criminal prosecutions. STATEMENT Two police officers went to respondent's residence during the afternoon and were met at the door by respondent's mother. When they asked to see respondent, she directed them to a bedroom where respondent was resting. One officer went to the kitchen of the home with respondent's mother, while the other asked respondent to dress and then accompanied him to the living room. Respondent's mother was told that the officers had a warrant to arrest respondent on a charge of burglary, but respondent was not told that the officers had a warrant or that he was under arrest. Pet. App. 2; J.A. 9, 18-20, 27-28. In the living room, the officer asked respondent if he was aware why the officers were there. Respondent said he was not. The officer then asked respondent if he knew a Mr. Gross. Respondent replied in the affirmative and added that he had heard of a "robbery" at the Gross residence. The officer then suggested to respondent that be believed respondent may have been involved in the burglary, and respondent replied: "I was there." Pet. App. 2; J.A. 19-20 28-29. Respondent was then transported to the stationhouse. The trip took approximately one hour because the arresting officers were diverted to answer another call, and a different officer completed the journey with respondent. At the stationhouse, respondent was, for the first time, given the warnings specified by Miranda v. Arizona, 384 U.S. 436 (1966). Respondent completed and signed a written form that had the Miranda warnings printed on it; he acknowledged in writing that he understood his rights and wished to discuss the crime with the police. He then gave the arresting officers a full written confession (J.A. 40-42) that detailed his role in the burglary. Pet. App. 2; J.A. 10-14, 21-23, 32-33, 38. The State did not seek to use at trial the statement respondent made at his home. The trial court, after a hearing, denied respondent's motion to suppress the subsequent written confession. The court first remarked that in its view the statement given at respondent's home would have been inadmissible because it was not preceded by Miranda warnings. But the court explained that the subsequent statement was admissible for the following reasons (J.A. 37-38): (T)he evidence shows that there was no taint between the first statement "I was there" and the second. * * * There was a break in the events * * * and I think the taint -- there's no showing of any taint whatsoever. There is a showing by the State that the statements were made freely and voluntarily. After a trial without a jury, respondent was convicted of first-degree burglary. He was sentenced to no more than five years' imprisonment and required to make restitution (J.A. 48). The Oregon Court of Appeals reversed (Pet. App. 1-6). It noted that the State conceded that respondent's first statement was inadmissible, and stated (id. at 3): "Our inquiry here is whether there was a sufficient break in the stream of events between that inadmissible statement and the written confession to insulate the latter statement from the effect of what went before." The court acknowledged that "the absence of actual compulsion * * * may be a factor in the analysis of the 'total circumstances of each case' that we must make in these successive interrogation cases" (id. at 4 (citation omitted)), but it ruled that this was not "a factor of sufficient significance" to render the second statement admissible (id. at 3-4). The court reasoned as follows (id. at 4): Regardless of the absence of actual compulsion, the coercive impact of the unconstitutionally obtained statement remains, because in a defendant's mind it has sealed his fate. It is this impact that must be dissipated in order to make a subsequent confession admissible. In determining whether it has been dissipated, lapse of time and change of place from the original surroundings are the most important considerations. See Westover v. United States, 384 U.S. 436, 496 * * * (1966). The Court of Appeals then noted that respondent "remained in custody without an opportunity to consult an attorney" after he made the first statement, and that "the same two officers who had elicited the original confessions were present at the subsequent interrogation" (Pet. App. 4). For these reasons, the court stated, it did not "find the absence of actual threats and compulsion sufficient" to warrant admitting the second statement into evidence (ibid). The Court of Appeals also rejected the State's argument that respondent's first statement revealed so little incriminating information that it could not be viewed as the cause of the second statement, which was far more inculpatory. The court explained (Pet. App. 5): "(The first) statement may not literally have admitted complicity in the burglary, but given the manner in which the questioning officer put the inquiry * * * (it) was significantly incriminatory. The cat was sufficiently out of the bag to exert a coercive impact on (respondent's) later admissions." A concurring judge joined the court's opinion because he believed the case was controlled by state law precedents but expressed his regret that the "metaphor (of 'the cat is out of the bag') now rules where analysis was once sovereign" (Pet. App. 5). The Oregon Supreme Court denied a petition for review without opinion (Pet. App. 7). SUMMARY OF ARGUMENT The opinion of the Oregon Court of Appeals suggests three possible bases for ordering the suppression of respondent's written confession. We submit that none of them justifies that result. A. It is plain that neither of respondent's statements was actually "compelled" within the meaning of the Self-Incrimination Clause of the Fifth Amendment. The officers used no threatening or oppressive tactics at any time. The Court of Appeals' suggestion that respondent's first statement coerced his second statement by letting "the cat out of the bag" reflects a misunderstanding of the notion of coercion: when a suspect makes a voluntary statement because the authorities have confronted him with incriminating evidence, that statement may be subject to suppression for some reason, but it is not correct to say that it has been coerced. B. The Court of Appeals also suggested that respondent's second statement was inadmissible, even though it was obtained following a waiver of Miranda rights, because the prior questioning at respondent's residence -- questioning that, the State conceded, did not comply with Miranda -- undermined respondent's capacity to give a subsequent valid waiver of his Miranda rights. We doubt that the State's concession was sound; we believe that respondent, who was not told he was under arrest, was not "in custody" within the meaning of Miranda when he was questioned at his residence. See Beckwith v. United States, 425 U.S. 341 (1976). But assuming that respondent was "in custody" when he made his first statement, the most that can be said is that the questioning constituted a borderline violation of Miranda. This was not remotely sufficient to cause respondent not to comprehend the meaning of his Miranda warnings when he did receive them, or to impair his ability to make a fully voluntary decision whether to invoke his rights. C. The most likely explanation of the court of appeals' decision is that the court considered respondent's second statement to be derived from -- that is, to be a "fruit" of -- the prior, inadmissible statement. This Court has not resolved the question whether evidence derived from statements that are inadmissible under Miranda should ever be suppressed, and we have taken the position that the Miranda exclusionary rule should be confined to the statement obtained without proper warnings and should not extend to any fruits of such a statement. But it is unnecessary for the Court to resolve that broad question here, because there are narrower grounds for holding respondent's written confession admissible. To begin with, respondent's written confession cannot be a suppressible "fruit" unless it was the causal product of his earlier remark, and the trial court appears to have been correct in its determination that the first statement did not produce the second. But even assuming that respondent's written confession was a product of his initial remark, the second statement should not have been suppressed because there is no substantial likelihood that the first statement was actually coerced. We do not dispute that when the first statement has been compelled, a subsequent statement that is a product of the first statement should also be inadmissible. We also recognize that the Miranda exclusionary rule is designed to deter abusive police questioning practices, and that it reflects in part the difficulty of determining reliably the voluntariness of statements obtained after prolonged incommunicado police interrogation. But these considerations underlying the Miranda rules cannot possibly be undermined by admitting into evidence an alleged fruit of a prior statement when -- as here -- there is no substantial likelihood that the prior statement was actually compelled. Admitting respondent's written confession is entirely consistent with the values underlying the Self-Incrimination Clause. When respondent decided to provide the State with the testimonial evidence it now seeks to use against him, he acted voluntarily and with an understanding of his right to remain silent if he chose to do so. At bottom, respondent's claim is not that his written confession was the product of coercion but that it was the product of his ignorance about the evidentiary status of his first statement. But such ignorance has never been thought to amount to coercion, and it should not be grounds for suppressing evidence when the only action by the police that was even arguably improper was a technical violation of Miranda. ARGUMENT RESPONDENT'S WRITTEN CONFESSION SHOULD NOT BE EXCLUDED FROM EVIDENCE Before respondent gave the written confession at issue here, he was advised of his Miranda rights and waived those rights in writing. There is no suggestion that the officers used oppressive or abusive tactics in obtaining the confession. Ordinarily, the Self-Incrimination Clause of the Fifth Amendment would present no obstacle to the use of such a confession against a criminal defendant. The rationale of the Oregon Court of Appeals' contrary ruling is somewhat unclear, but in our view none of the reasons suggested by the court's opinion justifies the suppression of respondent's written confession. A. Neither Of Respondent's Statements Was Compelled 1. The evil at which the Self-Incrimination Clause is aimed is compelled self-incriminating testimony. In this case, it appears to be beyond dispute that neither of respondent's statements was the product of actual compulsion. The trial court specifically found that "(t)here is a showing by the State that the statements were made freely and voluntarily" (J.A. 38). The Court of Appeals not only did not overturn this finding but appeared to agree that there was an "absence of actual threats and compulsion" (Pet. App. 4). There is no evidence whatever that the officers used coercive tactics either at respondent's residence or at the stationhouse. Although the State has conceded that respondent was "in custody" within the meaning of Miranda at the time of his first statement, it is far from clear -- as we will explain -- that even this concession was sound. But assuming that respondent was "in custody," the most that can be said is that the first statement was obtained through a technical violation of Miranda. There is no suggestion that the police used abusive or oppressive tactics. And there is no suggestion that the police actions at the stationhouse were coercive or overbearing. Indeed, the stationhouse interrogation was plainly not irregular in any way; respondent received Miranda warnings and signed a written waiver. 2. The Court of Appeals suggested that respondent's first statement, by letting "the cat out of the bag," had a "coercive impact" that led to the second statement (Pet. App. 5). This appears to have been merely a careless use of the word "coercive" that does not reflect a determination that respondent's written statement was actually the product of compulsion within the meaning of the Self-Incrimination Clause. In any event, it is clear that even when officers deliberately use a suspect's prior statement to cause him to make a second statement, the second statement has not been compelled within the meaning of the Fifth Amendment. This of course does not mean that the second statement will always be admissible into evidence: if the first statement was involuntary, we do not dispute that the subsequent statement should ordinarily be suppressed, /1/ and if the first statement is inadmissible for some other reason, the second statement may also be subject to suppression. But it is incorrect to say that a subsequent statement that is merely induced by officers through the use of a prior statement, and is not itself the product of coercion, has been compelled within the meaning of the Fifth Amendment. If a person speaks because he has been threatened with criminal punishment or physical violence, then, of course, his statement has been compelled. If his will has been overborne by such psychological tactics as isolation or relentless interrogation, then, too, the statement will be considered involuntary. On the other hand, if a person confesses because he has assessed the government's evidence against him and has decided that it is pointless to persist in his claims of innocence, his confession is the fully voluntary "product of a rational intellect and a free will" (Blackburn v. Alabama, 361 U.S. 199, 208 (1960)). Similarly, if a suspect is confronted with incriminating evidence and decides that it is in his interest to try to explain why the evidence does not necessarily inculpate him, his explanation is not compelled within the meaning of the Self-Incrimination Clause. Cf. Oregon v. Mathiason, 429 U.S. 492, 495-496 (1977); Beckwith v. United States, 425 U.S. 341 (1976). These propositions are equally true when the evidence with which a suspect is confronted is his own prior statement. See McMann v. Richardson, 397 U.S. 759, 769 (1970). If the officers in this case had prompted respondent to confess by confronting him with a recording of an intercepted telephone coversation in which he planned the burglary, the confession might or might not be admissible -- that would depend in part on the legality of the interception -- but there is not doubt that it would be voluntary. The situation is not materially different when a suspect is confronted, not with an intercepted conversation, but with statements he has previously made to the officers themselves. It follows that, whatever the admissibility of respondent's written confession, that confession was not compelled, even if it was the product of the prior statement "I was there." B. Respondent Validly Waived His Miranda Rights Before Making The Second Statement Even a voluntary statement, of course, is inadmissible if it has been obtained in violation of the prophylactic rules established by Miranda. Miranda requires that unless a person in custody has been warned of his right to remain silent and his right to counsel and has made a voluntary and intelligent waiver of those rights, any statements he makes in response to interrogation cannot be used against him. 384 U.S. at 444-445. As we have noted, respondent's second statement, on the surface, appears to have been obtained in circumstances that fully satisfied Miranda. Respondent was warned of his rights, and he executed a written waiver. Nothing about the interrogation atmosphere, and nothing in the behavior of the police or respondent, indicates that the waiver was not voluntary and intelligent. Certain portions of the court of appeals' opinion, however, suggest that the court believed that the events at respondent's residence somehow vitiated his subsequent waiver of Miranda rights. For example, the Court of Appeals emphasized that only an hour elapsed between the two statements and that the same officers elicited both. This suggests that the court believed that the atmosphere created by the officers at the arrest scene -- where, under the State's concession, Miranda was violated -- carried over and invalidated respondent's subsequent waiver of his Miranda rights. This reading of the court's opinion is reinforced by its citation of Westover v. United States, a case decided with Miranda (see 384 U.S. at 494-497) in which this Court held that the atmosphere created by a particular course of custodial interrogation without warnings -- during which the suspect did not confess, so there was no "fruits" question -- carried over and undermined a waiver given, after warnings, in subsequent questioning. To the extent the Court of Appeals concluded that respondent's waiver of his Miranda rights was invalid, it plainly erred. It is instructive to compare this case with Westover. Westover was interrogated at length, without warnings, while being held in custody for 14 hours by local police. During this period, he did not confess. Federal agents then took over the interrogation and, at the outset of their questioning, warned him of his rights. Two hours later, he confessed to the federal agents. 384 U.S. at 494-496. This Court ordered the confession suppressed because the "intense and prolonged interrogation" (Michigan v. Mosely, 423 U.S. 96, 107 (1975)) that preceded the warnings had the effect of undermining the voluntary and intelligent character of the suspect's waiver of his rights (see Miranda, 384 U.S. at 495-497). The contrast between Westover and this case could hardly be greater. Westover was subjected to lengthy, relentless interrogation while being held at a stationhouse; respondent was asked a few questions in his own living room. This case resembles Westover far less than it resembles Beckwith v. United States, supra, where the Court had little difficulty in concluding that no Miranda warnings were required. Indeed, in Beckwith the questioning was much more prolonged than it was here (see 425 U.S. at 342-344). To be sure, Beckwith was not in custody, and the State has conceded that respondent was in custody -- and was therefore entitled to Miranda warnings -- at the time he made his first statement. But we doubt that this concession was legally sound. The basis of Miranda was the Court's conclusion that warnings are needed to dispel coercive psychological forces that are created when a suspect is placed in custodial surroundings. See 384 U.S. at 448-458. It therefore seems evident that the test for whether a suspect is "in custody" within the meaning of Miranda must depend on whether a suspect would reasonably perceive that he has been subjected to formal arrest or its equivalent (see California v. Beheler, No. 82-1666 (July 6, 1983), slip op. 4) -- not on the arresting officers' uncommunicated subjective intentions. Cf. INS v. Delgado, No. 82-1271 (Apr. 17, 1984), slip op. 5. See, e.g., United States v. Harris, 611 F.2d 170, 173 (6th Cir. 1979) (defendant was not in custody because "(w)hile the agents might have prevented the defendant from leaving the room, they never informed the defendant of that intention"); United States v. Smith, 441 F.2d 539, 540 (9th Cir. 1981) ("The unexpressed intent of the officer to detain Smith had he attempted to leave does not create a custodial interrogation."); United States v. Hall, 421 F.2d 540, 544-545 (2d Cir. 1969), cert. denied, 397 U.S. 990 (1970), quoted in Steigler v. Anderson, 496 F.2d 793, 799 (3d Cir.), cert. denied, 419 U.S. 1002 (1974). See also United States v. Kennedy, 573 F.2d 657, 660 (9th Cir. 1978) ("The relevant inquiry * * * is whether a reasonable person would have believed himself to be" under arrest.). Here, there is little basis for a conclusion that respondent could have reasonably believed that he was under arrest. Respondent's mother was told that the officers had an arrest warrant, but there appears to be no evidence that respondent was told, before he made his statement, that the officers had a warrant or that they had come to the house for the purpose of arresting him. The officers testified (J.A. 10, 27) that they subjectively considered respondent to be under arrest at the time they interviewed him at this residence, but that is immaterial if the officers did not lead respondent to believe that he was under arrest. Accordingly, the most that can be said is that the brief questioning at respondent's residence constituted a borderline violation of Miranda. This was plainly insufficient to draw into question the validity of respondent's subsequent waiver of his Miranda rights. Miranda requires that a suspect, after having been advised of his rights, "knowingly and intelligently waive() his privilege against self-incrimination and his right to * * * counsel" (384 U.S. at 475). It is easy to see why Westover's ability to make such a waiver was impaired by the previous course of interrogation. /2/ But there is no reason to suppose that respondent even knew that he had been subjected to custodial interrogation. The interrogation in issue -- a few questions, asked in an unthreatening manner in his living room with his parents nearby -- could not possibly have impaired his ability to decide whether he wished to cooperate with the authorities. When he later received Miranda warnings, therefore, he had no reason to doubt that he was indeed free to terminate questioning or seek counsel if he wished. The court below identified nothing else about "'the particular facts and circumstances (of the) case, including the background, experience, and conduct of the accused'" (North Carolina v. Butler, 441 U.S. 369, 374-375 (1979), quoting Johnson v. Zerbst, 304 U.S. 458, 464 (1938); see Edwards v. Arizona, 451 U.S. 477, 486 n.9 (1981)) that might draw the validity of the waiver into question. For these reasons, the Court of Appeals manifestly erred in suggesting that respondent's waiver of his Miranda rights was not effective. The most likely explanation of this error is that the court confused this issue with the analytically distinct question whether respondent's second statement was "tainted by" or was the "fruit of" the first statement. That is, the court probably believed not that the prior questioning impaired respondent's ability to give a valid waiver but that respondent's previous answer -- his statement "I was there" -- rendered involuntary his subsequent decision to waive his Miranda rights and speak. But as we have explained, this reasoning rests on a misconception. If respondent was induced to waive his Miranda rights by the knowledge that he had already made an incriminating statement to the officers, that did not make his waiver any less voluntary. Indeed, the fact that the authorities already possess incriminating evidence might well be a good reason for a suspect to decide to waive his rights and speak to them. The only substantial question concerning the admissibility of respondent's second statement is, therefore, not whether the waiver was valid but whether the admission of that statement is inconsistent with the "fruits" doctrine and the policies underlying it. C. An Otherwise Admissible Statement Should Not Be Excluded From Evidence Solely Because It Is The Product Of A Statement Obtained In Violation Of Miranda When There Is No Substantial Probability That The First Statement Was Compelled Miranda generally requires the exclusion from the prosecution's case-in-chief of a statement obtained through custodial interrogation without the proper warnings and waiver. But this Court has never decided whether evidence derived from such a statement -- such as physical evidence or the testimony of witnesses whose existence is disclosed by the statement -- must also be suppressed. In Michigan v. Tucker, 417 U.S. 433 (1974), the Court held that the testimony of a witness discovered through a statement obtained in violation of Miranda was admissible; but the basis of the decision in Tucker, while highly relevant here, did not resolve the broad question whether the "fruits" of a Miranda violation may generally be admitted. Massachusetts v. White, aff'd by an equally divided Court, 439 U.S. 280 (1978), and Massachusetts v. Meehan, cert. dismissed as improvidently granted, 445 U.S. 39 (1980), concerned the admissibility of physical evidence derived from a statement obtained in violation of Miranda, but the Court did not decide the question in either case. The question presented in this case -- the admissibility of a subsequent statement by the suspect alleged to be the fruit of a prior statement that is inadmissible under Miranda -- is obviously closely related to the question presented by fruits consisting of physical evidence or the testimony of witnesses. Indeed, in one respect at least, the argument for admitting the fruit is strongest in this context. Once a suspect has made a statement that is inadmissible under Miranda, he is powerless to prevent the authorities from uncovering any physical evidence or the testimony of witnesses to which his inadmissible statement might lead; but he can prevent them from deriving any subsequent statement simply by refusing to speak further. In the Brief for the United States as Amicus Curiae in Michigan v. Tucker, supra, we urged that the Miranda exclusionary rule should be limited to the improperly obtained statements themselves and should not extend to fruits. /3/ We adhere to that position, but we do not repeat the argument here because we believe that the decision of the court below can be reversed on the narrower grounds -- either that the written confession was not in fact a product of the prior, inadmissible statement or that, even if it was, there is no justification for excluding it when the initial Miranda violation involved no substantial risk of actual compulsion. 1. Whatever the correct scope of the "fruit of the poisonous tree" doctrine in this case, respondent's second statement cannot be suppressed as a fruit unless it is causally connected to the first, inadmissible statement. That is, if it is established that respondent would have given his written confession even if he had not previously made an incriminating statement to the officers, the written statement cannot be considered a fruit and excluded from evidence on that basis. The trial court found that "the evidence shows that there was no taint between the first statement 'I was there' and the second (statement). * * * (T)here's no showing of any taint whatsoever" (J.A. 37-38). There are several reasons to accept this determination. First, respondent's first statement did not, in fact, let the cat out of the bag; it is implausible to suppose that respondent actually believed that he had nothing more to lose once he made the first brief remark. It requires little legal sophistication to realize that an admission that one was present at the scene of a crime is not tantamount to an admission that one was involved in it. A layman, in fact, is probably likely to overestimate the extent to which he can escape criminal liability or punishment by showing that he was only complicit in an offense and was not the instigator or the principal actor. Nor is there any evidence that respondent was aware of the probative value of oral statements to police officers; many suspects are not, and Miranda itself requires a warning that a statement may be used against an accused precisely because the Court assumed that many suspects will be ignorant of this fact (see 384 U.S. at 468-469). Second, there is no evidence that the officers used the first statement as part of their efforts to obtain the second statement. It appears that the officers did not mention the first statement to respondent during the stationhouse questioning or attempt to convince him that he had nothing more to lose by providing a full account of his actions. Finally, as we have explained, respondent's second statement was preceded by a fully valid waiver of Miranda rights. This means, among other things, that respondent was specifically told that he did not have to speak to the officers, and that he understood this warning. His situation was therefore materially different from that of a suspect who may believe that he has some obligation to explain away his prior statement. Respondent's valid waiver further diminishes the likelihood that he was induced to speak by the knowledge that he had already made a statement to the officers. 2. Nonetheless, for the remainder of our discussion we will assume that respondent's written confession was the product of his prior inadmissible oral statement; that is, we assume that respondent would not have given the written confession had he not already made the limited admission that he was present at the scene of the burglary. It does not follow that the written confession must be suppressed. On the contrary, we submit that because it is clear that the first statement, although inadmissible under Miranda, was voluntary, the second statement should not be suppressed merely because it was a product of the first. As the concurring opinion below suggested, the proliferation of figures of speech in this area -- "fruit of the poisonous tree," "attenuation of taint," "cat out of the bag" -- has tended, to a degree to obscure analysis. The problem is further complicated because a confession may be excluded not only as the product of a prior coerced confession -- a Fifth Amendment issue -- but as the product of a prior Fourth Amendment violation. See, e.g., Brown v. Illinois, 422 U.S. 590 (1975); Wong Sun v. United States, 371 U.S. 471 (1963). But "(t)he considerations underlying the Fifth Amendment 'fruits' doctrine are not necessarily the same as those relevant in the Fourth Amendment context" (Pillsbury Co. v. Conboy, No. 81-825 (Jan 11, 1983), slip op. 7-8 n.3 (Blackmun, J., concurring)), and the "fruits" analysis that is familiar and relatively well-established in the Fourth Amendment area cannot simply be carried over bag and baggage to a case like this. See United States v. Bayer, 331 U.S. 532, 541 (1947) (The Fourth Amendment exclusionary rule concerns "evidence of a quite different category and do(es) not control this question."). a. i. We do not dispute that when the authorities induce a suspect to make a further statement by using a previous statement that was actually compelled, the second statement is also inadmissible. In his separate opinion in Darwin v. Connecticut, 391 U.S. 346, 350-351 (1968), a case that involved an involuntary confession obtained through custodial interrogation, Justice Harlan gave an explanation of the basis of this rule: /4/ A principal reason why a suspect might make a second or third confession is simply that, having already confessed once or twice, he might think he has little to lose by repetition. If a first confession is not shown to be voluntary, I do not think a later confession that is merely a direct product of the earlier one should be held to be voluntary. It would be neither conducive to good police work, nor fair to a suspect, to allow the erroneous impression that he has nothing to lose to play the major role in a defendant's decision to speak a second or third time. In consequence, when the prosecution seeks to use a confession uttered after an earlier one not found to be voluntary, it has, in my view the burden of proving not only that the later confession was not itself the product of improper threats or promises or coercive conditions, but also that it was not directly produced by the existence of the earlier confession. The primary objective of any "fruits" rule in this context, therefore, should be to exclude statements that are the product of prior compelled statements. But it is of course well established that not every statement obtained in violation of Miranda has been compelled within the meaning of the Self-Incrimination Clause. See, e.g., Tucker, 417 U.S. at 444-446; Harris v. New York, 401 U.S. 222 (1971); Johnson v. New Jersey, 384 U.S. 719, 729-730 (1966); Miranda, 384 U.S. at 457. This objective, therefore, does not necessitate a rule excluding from evidence the fruits of statements that are inadmissible solely because they have been obtained in violation of Miranda. ii. In addition, the Court has made it clear that the Miranda rules serve a deterrent purpose. See, e.g., Oregon v. Hass, 420 U.S. 714, 723 (1975); Tucker, 417 U.S. at 446-447. But it is important to be clear about what the Miranda exclusionary rule is designed to deter. This is one respect in which it is incorrect to draw a close parallel between the Miranda rules and the Fourth Amendment exclusionary rule. The Fourth Amendment exclusionary rule (including its fruits doctrine) is designed to deter all unreasonable searches, to the extent that is practicable, because all unreasonable searches violate the Constitution. Miranda, however, is not designed to deter all custodial interrogation without warnings. Police officers do not violate any constitutional provision, threaten any constitutional value, act improperly in any way -- or, indeed, violate any stricture of the Miranda opinion -- if they conduct non-abusive custodial interrogation without giving Miranda warnings. All that Miranda requires is that the resulting statements not be admitted into evidence. This is because the Fifth Amendment privilege -- which is the foundation of Miranda -- does not prohibit the authorities from compelling incriminating statements; the Self-Incrimination Clause is violated only when a compelled statement is used against the speaker during a criminal presecution. It follows that even if officers deliberately decide to conduct questioning without warnings in circumstances that they know might be deemed custodial, they have not engaged in an improper practice so long as they have not acted abusively. Cf. Massiah v. United States, 377 U.S. 201, 206-207 (1964). In fact, it is not difficult to imagine instances -- if a suspect in custody is thought to know the location of a kidnapping victim or a bomb that is about to explode, for example -- where the far better police practice might be to conduct the interrogation without warnings, thus rendering inadmissible any statement that is obtained but increasing the likelihood that the suspect will speak. Accordingly, the primary "deterrence" objective of the Miranda exclusionary rule -- including any rule concerning the fruits of a Miranda violation -- is not to deter police from questioning suspects in custody without warnings, but rather to deter abusive police practices. b. As we have noted, this Court has not yet resolved the question whether fruits of voluntary statements obtained in violation of Miranda are generally admissible. See Tucker, 417 U.S. at 446. But United States v. Bayer, supra, strongly suggests that the Court would not require the suppression of every statement that is the product of a prior voluntary statement inadmissible under Miranda. In Bayer, the authorities questioned a suspect in custody under circumstances that were assumed to violate a prophylactic rule that served purposes similar to those of Miranda -- the rule of McNabb v. United States, 318 U.S. 332 (1943), which excluded evidence obtained from a defendant who was not brought before a magistrate without undue delay (compare 384 U.S. at 463 with 318 U.S. at 342-345). Subsequently, the authorities obtained a further statement under circumstances that were not themselves suspect but that left little doubt that the second statement was the product of the first. /5/ This Court reversed a ruling that the second statement was inadmissible, reasoning as follows (331 U.S. at 540-541): Of course, after an accused has once let the cat out of the bag by confessing, no matter what the inducement, he is never thereafter free of the psychological and practical disadvantages of having confessed. He can never get the cat back in the bag. The secret is out for good. In such a sense, a later confession always may be looked upon as fruit of the first. But this Court has never gone so far as to hold that making a confession under circumstances which preclude its use, perpetually disables the confessor from making a usable one after those conditions have been removed. It would appear to follow from the holding in Bayer that subsequent statements that are the product of statements obtained in violation of Miranda may nevertheless be admissible on at least some occasions. Consequently, leaving aside the possibility of a general rule that all such fruits are admissible, application of the fruits doctrine must involve some inquiry into the particular facts of the case. We note that such an inquiry does not impair the certainty and rigidity that are sometimes thought to be a principal virtue of Miranda: /6/ while it is generally desirable to have relatively clear and mechanical rules to guide police behavior, a doctrine governing the admissibility of fruits -- in this or other contexts (e.g., United States v. Ceccolini, 435 U.S. 268 (1978)) -- is not a guideline for police behavior but a principle for courts to apply in delineating rules of evidence. c. We submit that exclusion of an otherwise admissible voluntary statement as a fruit of a Miranda violation is justifiable, if at all, only when there is a substantial probability that the Miranda violation involved actual compulsion. In our view, a rule admitting the second statement when the first was clearly not compelled cannot undermine the purposes of Miranda and the Self-Incrimination Clause. /7/ i. The first task of a "fruits" doctrine in this area, as we noted, is to exclude the fruits of compelled statements. This objective is plainly not jeopardized by a rule allowing in evidence the fruits of clearly voluntary prior statements like respondent's first statement in this case. To go further, and exclude every statement that is the product of a prior statement inadmissible under Miranda, no matter how clearly voluntary the prior statement, would unnecessarily suppress much highly probative evidence. The government may use voluntary statements obtained in violation of Miranda in significant ways in which it could not use compelled statements -- for example, to impeach a defendant's testimony at trial. Compare Mincey v. Arizona, 437 U.S. 385, 398 (1978), and New Jersey v. Portash, 440 U.S. 450 (1979), with Harris v. New York, supra, and Oregon v. Hass, supra. Similarly, there is no sufficient reason to preclude the government from using the fruits of statements that were obtained in violation of Miranda but are plainly not coerced. ii. The decision in Miranda was largely rooted in the Court's concern that the traditional voluntariness standard was difficult to apply to incommunicado stationhouse interrogation in a way that gave confidence in the reliability of the ultimate determination. See, e.g., 384 U.S. at 445, 461. The only witnesses to such questioning, generally, were the suspect and the police officers, neither of whose testimony was entirely reliable. Moreover, the inquiry into whether the suspect's will was overborne by coercive interrogation tactics sought to answer a subtle, elusive, subjective question about state of mind -- a question that was made even more difficult because the interrogating officers often employed sophisticated psychological strategies, and because the correct answer depended upon nuances that were impossible to reconstruct in court. See id. at 448-458. Under these circumstances, there appeared to be an unacceptably high risk of error in the determination of voluntariness, with a consequent danger that the use of the confession would in fact involve an undetected infringement of the constitutional privilege against self-incrimination. These considerations, when combined with the principle that the fruits of compelled statements are inadmissible, afford a rational basis for excluding the fruits of a Miranda violation if the violation consists of incommunicado police interrogation of a kind that creates an appreciable possibility of undetected coercion. But they provide no justification for excluding the fruits of a Miranda violation when the questioning occurs in a setting in which -- as with the questioning at respondent's residence -- it can be reliably determined that the initial statement was not actually compelled. /8/ iii. Permitting the admission of respondent's written confession will not encourage police officers to engage in practices that are even arguably abusive. To begin with, many questionable police practices in this area can be deterred under the rules governing the waiver of Miranda rights, without resort to a fruits doctrine at all. For example, if police officers conduct a prolonged stationhouse interrogation without warnings, in order to obtain a statement (itself inadmissible under Miranda) that can be used to elicit a second confession after the suspect has been given proper warnings, Westover would ensure the failure of their strategy; the waiver that preceded the second statement would plainly be invalid. In all likelihood, the officers would have to release the suspect or allow him substantial contacts with outsiders before they could attempt to take advantage of the prior statement. Knowing this, the officers would be unlikely to engage in such a practice in the first place. But even if it is thought that it will sometimes be necessary to rely on a Miranda fruits exclusionary principle in order to deter abusive police practices, there is no justification for suppressing a subsequent statement derived from a prior inadmissible statement when -- as here -- there is no substantial probability that the first statement was coerced. The abusive police practices with which Miranda is concerned are those that are likely to create a substantial risk of coercion. To the extent that the decision to admit a fruit depends on whether there was a substantial probability of coercion, law enforcement officers have no incentive to create a situation that is even arguably coercive. iv. Finally, admitting respondent's written confession does not offend against any of the values underlying the Self-Incrimination Clause; in no meaningful sense can it be said that if respondent's written confession is admitted he will have been compelled to testify against himself. Indeed, the admission of a subsequent statement in a case such as this seems even more clearly consistent with the principles of the Self-Incrimination Clause than the admission of other fruits of a Miranda violation, such as physical fruits or a witness's testimony. That is because, in this case, the incriminating evidence against respondent resulted not solely from a decision to speak made without benefit of Miranda warnings, but from a subsequent "exercise of 'a rational intellect and a free will'" (Mincey, 437 U.S. at 398 (citation omitted). Moreover, before respondent spoke at the stationhouse he made, as we have shown, a fully effective waiver of his Miranda rights. In other words, before respondent decided to produce the testimonial evidence that the State now seeks to use against him, he fully understood that he did not have to speak, and he voluntarily and intelligently chose to do so nonetheless. It is difficult to see why the principles of the Self-Incrimination Clause should require the suppression of a statement given in these circumstances. At bottom, the complaint of a suspect in respondent's position is not that he was coerced but that he did not know the evidentiary status of his prior statement. If respondent knew that his first statement was inadmissible, he could not possibly contend that his second statement should be suppressed. See McMann, 397 U.S. at 768. But the Self-Incrimination Clause does not require the exclusion of statements that are made because a suspect is not fully aware of what evidence can and cannot be used at his trial. Cf. Beheler, slip op. 5 n.3; Schneckloth v. Bustamonte, 412 U.S. 218, 226-227, 234 (1973). If, for example, police officers induced a defendant to confess by confronting him with another party's inadmissible hearsay statement, the confession would not be suppressed. There is no reason for a different result in this case, where coercion is, similarly, wholly lacking, and the inadmissibility results from what can fairly be described as at most a technical violation of Miranda. /9/ CONCLUSION The judgment of the Oregon Court of Appeals should be reversed. Respectfully submitted. REX E. LEE Solicitor General STEPHEN S. TROTT Assistant Attorney General ANDREW L. FREY Deputy Solicitor General DAVID A STRAUSS Assistant to the Solicitor General MAY 1984 /1/ See pages 19-21, infra. Justice Harlan, in his opinion in Darwin v. Connecticut, 391 U.S. 346, 350-351 (1968), drew this distinction between a subsequent statement that is itself voluntary and is inadmissible only because it is the product of a prior, involuntary statement, and a subsequent statement this is involuntary because the conditions that compelled the first statement persist. See also McMann v. Richardson, 397 U.S. 759, 769 (1970). /2/ Westover was subjected to the "inherently compelling pressures" (Miranda, 384 U.S. at 467) and the "badge of intimidation" (id. at 457) of the custodial environment for an extended period; the mere giving of warnings could reasonably be thought unlikely to restore his capacity to decide whether he wished to discuss his alleged offense with the authorities. Similarly, having been subject to the "interrogator's imprecations * * * that the interrogation will continue until a confession is obtained" (id. at 468), it is more than possible that Westover would not fully believe, when he was finally given warnings, that he actually was free to terminate questioning if he wished. /3/ Copies of our brief in Tucker have been sent to the parties. We also relied on our brief in Tucker as the basis of part of our submission in New York v. Quarles, No. 82-1213 (argued Jan. 18, 1984). /4/ This rule is well established in cases in which the source of compulsion is irregular police practices during custodial interrogation. How and whether the rule should apply when the source of compulsion is legal process (for example, compelled grand jury testimony under a grant of immunity) are difficult and uncertain questions. See generally Pillsbury Co. v. Conboy, supra. On the one hand, under the rule of Counselman v. Hitchcock, 142 U.S. 547 (1892), the privilege against self-incrimination extends not only to a compelled statement but to evidence derived from it. See also Kastigar v. United States, 406 U.S. 441, 453-454 (1972); Murphy v. Waterfront Commission, 378 U.S. 52, 79 (1964). It might be said that a subsequent voluntary statement by the witness can be "derived" from his prior compelled testimony just as physical evidence, or the testimony of another person, can be so derived. On the other hand, a subsequent voluntary statement by the witness is, by hypothesis, the product of an independent, deliberate decision to speak. The nexus between the compulsion and the derivative evidence is, therefore, much different from the case of derived physical evidence. In addition, a subsequent statement is the product not just of the prior statement but of the witness's ignorance about the evidentiary status of the prior statement. If a witness knew that his compelled statement could not be used against him, he would surely not be heard to claim that a subsequent voluntary statement was derived from the compelled statement. (Witnesses who are compelled to speak by legal process, unlike those compelled by police practices, are typically told that they have use immunity and therefore are seldom ignorant of the fact that their statement is inadmissible. This is presumably why questions parallel to those presented by this case seldom arise in the context of testimony compelled by judicial process.) The fact that ignorance and not just compulsion is the source of the subsequent statement suggests that due process doctrines concerned with fairness -- in addition to, or perhaps instead of, doctrines concerned with compelled self-incrimination -- are the source of the rule excluding subsequent voluntary statements derived from prior compelled statements. Justice Harlan's remarks, quoted in text, suggest as much. /5/ The Court explained (331 U.S. at 540) that after his first statement the suspect volunteered some additional facts, and the meeting at which the second statement was obtained "was to incorporate the whole story in one statement. * * * (The suspect) requested the original statement and read it before making the second. The (second) statement is labelled a 'supplementary' statement and is 'basically' the same as the earlier one but went into more detail." /6/ We believe these virtues are substantially overstated. The question whether a suspect is "in custody" for Miranda purposes is not one that can be resolved by the use of a bright-line standard. See Florida v. Royer, No. 80-2146 (Mar. 23, 1983), slip op. 15 (plurality opinion). See also United States v. Harris, 611 F.2d at 172; United States v. Jimenez, 602 F.2d 139, 143 (7th Cir. 1979); United States v. Kennedy, 573 F.2d at 660; United States v. Del Soccorro Castro, 573 F.2d 213, 215 (5th Cir. 1978); Steigler v. Anderson, 496 F.2d at 799-800; United States v. Hall, 421 F.2d at 541. Moreover, under Harris v. New York, 401 U.S. 222 (1971), in order to decide whether a statement obtained in violation of Miranda can be used for impeachment, a court must inquire into the particular circumstances involved to determine if the statement was voluntary. /7/ The admission of fruits in these circumstances is not inconsistent with Harrison v. United States, 392 U.S. 219 (1968). In that case, the government introduced into evidence at trial three confessions made by the defendant. The defendant then testified in his own behalf. On appeal, the confessions were held to be inadmissible, although not because they had been coerced. On retrial, the government was permitted to introduce a transcript of the testimony the defendant had given at the first trial. This Court, finding that the defendant's testimony at the first trial was at least in part the fruit of the improper admission of his confessions at that trial (id. at 224-226), held the testimony inadmissible. In Harrison, however, the Court explicitly relied not just on the need to deter the government from obtaining confessions improperly, but on "'the imperative of judicial integrity'" (392 U.S. at 224 n.10 (quoting Elkins v. United States, 364 U.S. 206, 222 (1960)). Harrison is crucially different from this case (even assuming that respondent's second statement was the product of his earlier utterance) because in Harrison the trial court erred in sustaining the government's efforts to have the confessions admitted into evidence and as a result of that error the government was able to obtain a further damaging statement. There is an interest in preventing the defendant from being prejudiced by an erroneous judicial ruling that is quite distinct from the interests implicated in the present case. /8/ It is true, of course, that a suspect has the right to Miranda warnings whenever he is subjected to custodial interrogation, even in circumstances in which there is little or no risk of actual compulsion. But this aspect of Miranda is based, we believe, on the need to specify an identifiable line that will facilitate administration of the rule by the police, not on the concern about undetected coercion that is at the heart of Miranda. As we have noted (pages 24-25, supra), the need to define a clear standard is a far less important concern in the context of expounding a "fruits" doctrine. /9/ It appears that the admission into evidence of respondent's written confession would constitute no significant innovation in the way Miranda is now applied by the courts of appeals. In cases in which a defendant made two confessions, the first of which was voluntary but inadmissible under Miranda, the courts of appeals appear consistently to have inquired only into whether the second statement was preceded by a valid waiver. They have, in general, not independently considered whether the second statement should be suppressed as a fruit. See, e.g., United States v. Packer, 730 F.2d 1151, 1157-1158 (8th Cir. 1984); Bartlett v. Allen, 724 F.2d 1524, 1527-1528 (11th Cir. 1984); United States v. Bowler, 561 F.2d 1323, 1326 (9th Cir. 1977); United States v. Toral, 536 F.2d 893, 896-897 (9th Cir. 1976); Knott v. Howard, 511 F.2d 1060, 1061 (1st Cir. 1975); United States v. Hale, 397 F.2d 427, 429-431 (7th Cir.), cert. denied, 393 U.S. 1067 (1968); United States v. Knight, 395 F.2d 971, 973-975 (2d Cir. 1968), cert. denied, 395 U.S. 930 (1969).